Rebuttal to Peltzmans Findings—Robertson and Joksch

Robertson proceeds to offer a modified version of Peltzman s prediction equation in which the accident cost variable is dropped, the youth variable is changed from share of the population which is young to the share of youthful drivers in fatal accidents , the fatality rate for occupants is multiplied by the proportion of fatal accidents which involved passenger cars and the nonoccupant rate was backed out from the adjusted total fatality rate. Using the modified equation Robertson finds that actual fatality rates during the regulatory period from 1966 to 1972 are well below projected rates and that the measured contribution of vehicle safety standards is quite compatible with earlier technological analysis. However, as Peltzman points out that each of the corrections made by Robertson is a matter of choice and that each is 

Evidence Based on Crash Involvements Rates—Robertson and On  

Robertson appears to have found strong evidence that the safety standards did not increase nonoccupant risk and even made nonoccupants safer. One 

The surprising results found by Robertson have been examined more closely by Lloyd Orr. Orr reanalyzes the same FARS data using the technological approach implicit in Robertson s research. The difference is Orr does not accept Robertson s use of truck rates and the car-truck dummy variable in the regressions. He gives two reasons  

---