THE EFFECT OF SEAT BELT USAGE RATES ON THE NUMBER OF MOTOR VEHICLE-RELATED FATALITIES Richard A. Derrig, Ph.D., 1 Maria Segui-Gomez, M.D., Sc.D., 2 Ali Abtahi, M.P.H. 2 1 Automobile Insurers Bureau of Massachusetts 2 Injury Control Research Center Harvard School of Public Health ABSTRACT The effectiveness of seat belt usage in reducing mortality and morbidity among traffic accident victims has been well established. Population usage rates have been increasing from eleven percent in 1980 to sixty-eight percent in 1995, as measured by observational surveys sponsored by the National Highway Traffic Safety Administration (NHTSA). Safety incentive grants from NHTSA to the States with higher than average usage rates are expected to total $500 million during 1999-2003. Longitudinal annual motor vehicle-related fatality levels are analyzed by state to estimate the effect of the population seat belt usage rate on fatalities in the presence of known confounders such as alcohol use and youthful drivers. Consideration of alternative multivariate methodologies applied to fifteen years of data shows that the population usage rate is associated with a small effect on fatalities that for most methodologies is not statistically significant. Such a result calls into question the NHTSA policy of basing the incentive program on overall seat belt usage rates. 1 To whom all correspondence should be addressed: 101 Arch St. Boston, MA 02110-1103 tel (617) 439-4542, fax (617) 439-6789 e-mail: richard@aib.org March 17, 2000