This paper quantifies relationships between accident rates and congestion on California urban freeways. Data for a number of freeway sections in the San Francisco Bay Area are analyzed and numerous alternative accident‐model specifications are developed and tested. Accident rates are compared under conditions of queuing and no queuing and for a variety of geometric and traffic conditions. No matter how the data are stratified, the same picture emerges: In the presence of queuing, the average accident rate is about two‐three times higher than the rate when there is no queuing. This leads to the development of a simple accident‐rate estimation procedure that is recommended for use in cost‐benefit analyses of the impacts of freeway improvements on accident rates.
No takes yet. Share an insight, caveat, or question.
Edward C. Sullivan (1990) studied this question.