MEASURING THE CONTRIBUTION OF RANDOMNESS, EXPOSURE, WEATHER, AND DAYLIGHT TO THE VARIATION IN ROAD ACCIDENT COUNTS

Road accident counts are influenced by random variation as well as by various systematic, causal factors. To study these issues, a four-coun try, segmented data base has been compiled, each segment consisting of monthly accident counts, along with candidate explanatory factors, in the various counties (provinces) of Denmark, Finland, Norway, or Swed en. Using a generalized Poisson regression model, we are able to decom pose the variation in accident counts into parts attributable to rando mness, exposure, weather, daylight, or changing reporting routines and speed limits. To this purpose, a set of specialized goodness-of-fit m easures have been developed, taking explicit account of the inevitable amount of random variation that would be present in any set of accide nt counts, no matter how well known the accident generating Poisson pr ocess. Pure randomness is seen to ''explain'' a major part of the vari ation in smaller accident counts (e.g. fatal accidents per county per month), while exposure is the dominant systematic determinant. The rel ationship between exposure and injury accidents appears to be almost p roportional, while it is less than proportional in the case of fatal a ccidents or death victims. Together, randomness and exposure account f or 80% to 90% of the observable variation in our data sets. A surprisi ngly large share of the variation in road casualty counts is thus expl icable in terms of factors not ordinarily within the realm of traffic safety policy. In view of this observation, it may seem unlikely that very substantial reductions in the accident toll can be achieved witho ut a decrease in the one most important systematic determinant: the tr affic volume.