Measuring the effect of police surveillance on the prevention of traffic accidents

An approach for estimating the effect of preventive measures taken against traffic accidents is presented. A simple stochastic model as to the process of accident occurrence is provided and the relationship among vulnerable behavior, collisions, and police surveillance is formalized. Here vulnerable behavior is defined as the state prior to the collision in the process of accident occurrence, and police surveillance is selected as one of the preventive measures against the accident.

The procedures by which the model can be applied in practice are shown for the case of intersection accidents, and the effects of some intersection surveillance forms on accident prevention are calculated. The results suggest that police surveillance which affects vulnerable behavior reduces accidents and that each collision category differently has the most effective surveillance form.

It is concluded that vulnerable behavior, which increases the likelihood of accident occurrence, provides a more sensitive safety measure than accidents.