This article describes ways in which official accident statistics are either inaccurate, or are interpreted in such a way as to give misleading conclusions about road safety. The question is also raised as to which class of road or area is safer: one which has the least accidents, or one with the least fatalities. Accident statistics can be inaccurate because: a) the classification of serious injuries overestimates the numbers of such injuries by something like 75%; b) slight injuries are grossly under recorded; and c) the differences in the ratio of fatality to injury over time or between areas could just be due to the number of police available to report them and not to actual differences. Conclusions about the safety of roads based on accident statistics could also be correct for other reasons. For example: a) a statistically safe road could be so dangerous that it is avoided; b) the lack of a measure for pedestrian activity compatible to vehicle miles travelled could lead to crowded areas being wrongly considered less safe than one sparsely populated, but with faster roads; and c) statisticians tend to use injury rather than fatality figures. Measures taken to decrease the former could actually increase the latter.
Samenvatting