REGRESSION ANALYSIS OF TIME SERIES DATA ON MOTOR VEHICLE DEATHS IN THE UNITED STATES.

THIS STUDY UPDATES PELTZMAN'S (SEE IRRD 228191 AND 222069) TIME SERIES ANALYSIS. IT USES LEAST-SQUARES REGRESSION AND DATA FROM 1947 THROUGH 1980 TO ESTIMATE AN EXPANDED VERSION OF PELTZMAN'S MODEL. INDEPENDENT VARIABLES CONSIDERED ARE THE SIZE OF VEHICLE AND A VARIABLE WHICH MEASURES THE IMPACT OF FEDERAL SAFETY STANDARDS FOR MOTOR VEHICLES ON HIGHWAY TRAFFIC DEATHS. IT USES DEATH FREQUENCY AS THE DEPENDENT VARIABLE AND INCLUDES AMOUNT OF DRIVING AS AN ADDITIONAL REGRESSOR. THREE CATEGORIES OF MOTOR VEHICLE DEATHS ARE ANALYSED: TOTAL DEATHS, VEHICLE-OCCUPANT DEATHS, AND PEDESTRIAN DEATHS. NOTE IS TAKEN OF THE IMPACT OF VARIOUS FACTORS ON EACH OF THESE CATEGORIES. DETAILS ARE GIVEN OF THE MODEL DERIVED AND DATA USED FOR THE STUDY. MANY OF THE STATISTICAL RESULTS OF THE STUDY CORROBORATE PELTZMAN'S EARLIER FINDINGS. VEHICLE SIZE IS FOUND TO HAVE A STATISTICALLY SIGNIFICANT INVERSE EFFECT ON TOTAL AND VEHICLE-OCCUPANT DEATHS. THE RESULTS ALSO INDICATE THAT THE RATIO OF RURAL TO URBAN DRIVING SIGNIFICANTLY INCREASES BOTH TYPES OF DEATH, AND THAT THE NUMBER OF HIGHWAY DEATHS IS REDUCED BY MANDATED SAFETY FEATURES.