Good and Bad Drivers--A Markov Model of Accident Proneness [Discussion]

A series of studies by the Department of Transportation in the 1960's and early 1970's first popularized analysis of automobile insurance experience through accident proneness models. Unlike many actuarial approaches which tend to be empirical and. emphasize practical interpretations, accident proneness models tend to be heavily theoretical and often prove difficult to interpret on a non-technical level. If one is not already familiar with the various Poisson process models found in actuarial literature, Dr. Venezian's paper might be extremely slow going. Dr. Venezian's summary of other authors' models in the introduction is remarkable in its terseness without omitting the salient points. In the same style, the author points out many strengths and implications of his own proposed model. This is a paper which requires careful reading, working out details derived by the author, for full appreciation and understanding.