Do we know why the number of traffic fatalities is declining? If not, can we find out?

Abstract

The number of traffic fatalities has declined by 80-90% from the all-time peak number in many highly motorized countries. It would be of great interest to identify factors that explain this decline. Unfortunately, this is difficult, and an ideal method does not exist. This paper discusses some less-than-ideal methods. Multivariate statistical analyses are unlikely to be informative because they are likely to be affected by both multicollinearity and omitted variable bias. This implies that they will always include both too many and too few predictor variables, a problem that is impossible to solve. Historical reconstruction is another possible method. It explains fatality reductions over time by known factors that are related to fatalities and for which sufficient information about their presence or uptake is available. Studies based on historical reconstructions show consistently that safer road user behavior is a major contributor to reducing fatalities, followed by infrastructure and vehicle related safety measures. However, there is no way of establishing causality. The relative importance of different types of factors is highly dependent on the amount of information available. In a strict sense, there is therefore no prospect of providing a scientific explanation of the decline in traffic fatalities. In a less strict sense, historical reconstructions still may give an impression of relative contributions of some known factors. For example, the most recent Norwegian study identified factors that may explain more than half of the reduction of road traffic fatalities after 2000.