The effectiveness of safety belts in preventing fatalities to drivers and to passengers seated in the right front seat of cars involved in crashes is determined using the double pair comparison method described in a just issued paper. All available complete years (1975 through 1983) of data in the Fatal Accident Reporting System (FARS) are used. Only cars of model year 1974 or later are included, so that all occupants coded in FARS as using any restraint system are presumed to be using the three point integrated lap/shoulder belt. By safety belt effectiveness we mean the reduction in fatalities that would occur in an unbelted population if all its members were to become belt wearers without any other changes. The double pair comparison method focuses on crashes by cars containing a "subject" occupant (a driver or right front passenger) and an "other" occupant (any other occupant in the car except the subject occupant). The other occupant acts essentially to estimate exposure, so that fatality risk to the subject when belted and unbelted may be compared relatively free from biasses due to confounding factors. In this study, drivers and right front passengers are subject occupants; by choosing a variety of other occupants (of different ages, seating positions, and belt use) 46 essentially independent estimates of safety belt effectiveness and associated standard errors are obtained. By taking a weighted average of the 46 estimates, an estimate of overall safety belt effectiveness is obtained which is considered more precise and free from bias than prior estimates. It is found that if all presently unbelted drivers and right front passengers in the U.S. were to use lap/shoulder belts, but not otherwise change their behavior, fatalities to this population would be reduced by (41 plus or minus 4)%, where the error limits indicate one standard error. This finding is in agreement with a recent major review and synthesis by the National Highway Traffic Safety Administration (NHTSA) of prior knowledge on the subject which concluded that effectiveness was in the range 40%-50%. Combining the present determination with the NHTSA value generates a value (43 plus or minus 3)%.

  • Corporate Authors:

    General Motors Corporation

    Research and Development Center, 30500 Mound Road
    Warren, MI  United States  48090
  • Authors:
    • Evans, Leonard
  • Publication Date: 1985-6-24

Media Info

  • Features: Figures; References; Tables;
  • Pagination: 28 p.

Subject/Index Terms

Filing Info

  • Accession Number: 00495590
  • Record Type: Publication
  • Source Agency: National Highway Traffic Safety Administration
  • Report/Paper Numbers: GMR-5088, HS-039 470
  • Files: HSL, TRIS, USDOT
  • Created Date: Jul 31 1990 12:00AM