The paper analyses the geographical pattern(s) of road mortality in western Europe for three standard levels of data aggregation. A new adjusted road mortality ratio is proposed. Unlike most mortality ratios, the standardization uses population density rather than age and sex. The hypothesis to be tested is that changes in the level of spatial aggregation of the data produce significant differences in the values of some basic descriptive statistics, and hence in operational conclusions. The paper shows that population density works as a proxy for structural factors that may be difficult to change by policies and that working with 264 regions confirms the north-south divergences in terms of road safety in Europe but also stresses the local or regional differences due to political, geographical, economic, environmental and statistical disparities. Safe countries include unsafe regions, and there are safe regions in unsafe countries. Decision makers should be aware of this when comparing countries. (Author/publisher)
Abstract