Vehicle safety tests, rankings, curb weight, and fatal crash rates: automatic emergency brakes associated with increased death rates
Robertson, L. S.
Show abstract
The European New Car Assessment Programme (Euro NCAP), the Insurance Institute for Highway Safety (IIHS), and the U.S. National Highway Traffic Safety Administration (NHTSA) each publish safety ratings of new passenger vehicles based on crash test results and crash avoidance technology soon after they are introduced into the market. IIHS alone singles out vehicles for "Top Safety Pick". The Institute also periodically lists driver death rates of vehicles during their first few years of use, accompanied by assertions that larger, heavier passenger vehicles are safer. Median death rates by vehicle size in the Institutes data do not support the assertion. This study examines the association of vehicle weight to the risk of all deaths in fatal crashes where specific makes and models were involved, controlling statistically for the Institutes vehicle safety ratings as well as NHTSA ratings of full-frontal crash tests and rollover propensity, lane retention warnings, adaptive cruise control, and automatic braking. Increased weight is slightly related to a lower fatal risk in the 2014-2017 models but not the 2018-2019 models. Lane retention warnings and IIHSs higher ratings of crashworthiness are related to reduced death risk. Adaptive cruise control is not associated with fatal crash risk. Automatic emergency braking (AEB) technology rated "superior" is a criterion for an IIHS "Top Safety Pick" but it is correlated to higher fatal crash involvement of vehicles that have the technology as optional or standard equipment. IIHS tests of AEB systems are conducted only at low speeds. The statistical results are not definitive but suggest that a rating system based on modeling death risk prediction from more detailed data from various tests such as Euro NCAP tests of AEB systems could better inform consumer choice of vehicles.
Matching journals
The top 6 journals account for 50% of the predicted probability mass.