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

In a 1970 paper, Donald Weber presented a stochastic model of the automobile accident process. In that paper, Weber took age and gender differences into account in a deterministic fashion by means of some ad hoc rational functions of time. The present paper, on the other hand. deals with these differences explicitly though a stochastic model. by using a Markov process to describe how an individual’s accident likelihood varies over time. The latter approach is much more satisfying since it recognizes that accident likelihood's do vary among individuals with otherwise identical risk characteristics. In the light of current controversies over risk classification, this paper undoubtedly will be an important contribution to the literature. and it is certainly a timely and thought-provoking one.