An Assessment of the Reliability of the NHTSA New Car Assessment Program

An August 4, 1994 news release of the U.S. Department of Transportation [Ref. 1] states that “The National Traffic Safety Administration (NHTSA) has determined that there is a positive relationship between scores it provides to consumers in its new car and light truck crash tests and the fatality risk in real-world crashes. ‘Consumers can use this crash test information in their purchase decisions. We are pleased that our latest analysis verifies that the information continues to provide a valid indication of the crash protection that various makes and models provide,’ said NHTSA Acting Administrator Christopher Hart.” These assertions are said to be based on an NHTSA technical report [Ref.2]. In this referenced report, however, it is stated that the “results do not guarantee that an individual make-model with low NCAP (New Car Assessment Program; the crash test) scores will necessarily have lower fatality risk than other make-models with higher NCAP scores”. The contradiction between the news announcement and the referenced underlying technical report is apparent. The questions this paper addresses are: 1) does the NCAP program provide “valid indication of crash protection”; and 2) is it true that the “consumers can use this crash information in their purchasing decisions”? Our analysis indicates that 1) the NCAP results have not been shown to provide a valid indication regarding the crash protection of specific vehicles; and 2) consumers will frequently make the wrong decisions if they use this information when making their purchasing decisions. In addition, we found that other researchers have reported similar findings regarding the lack of utility of NCAP, and the public statements of NHTSA are in conflict with the cautions expressed in their underlying technical documentation.