Critique of Sam Peltzman's study : The effects of automobile safety regulation

The variables used in Peltzman's analysis were reviewed. It was concluded that some of them were arbitrarily chosen, that some were correlated, and that important factors were omitted. This may cause spurious and biased correlations. Peltzman's time series regression equations were reconstructed and found unstable, which makes them useless for predictions which are one basis for Peltzman's conclusions. The cross-sectional analyses were found to be unvalidated. Their results for important factors disagreed with those from the time series analysis. Peltzman's conclusions on the role of young drivers were compared with their actual accident involvement and found to disagree. To illustrate the misuse of trend models, a “model” for pedestrian deaths was constructed which leads to conclusions contradicting Peltzman's.