This research project describes and analyses the effects of measures taken to improve accident spots in the road network in urban areas. The project was based upon accident data and additional local information from 128 accident spots in urban areas where measures for improvement had been taken between 1988 and 1992. The data on 106 locations were collected by the competent local police and construction authorities in North Rhine-Westphalia, Hamburg, Lübeck and Berlin. The data on another 22 investigation locations in Baden-Württemberg came from the Landesamt für Strassenwesen (Department for Road Construction, Traffic Safety and Transport Systems of the Land of Baden-Württemberg). The investigation subjects were selected irrespective of the occurrence of accidents or of the local boundary conditions. 95 % of the random sample consisted of junctions while the remaining accident spots were road sections. The measures for improvement which the project dealt with were divided into classes according to the type of action carried out. This resulted in 37 different kinds of action undertaken which were then brought together into ten main classes. In order to assess the effectiveness of the measures taken, the absolute number of accidents and casualties, the absolute accident costs and the relative parameters (accident rate and accident cost rate) were compared for the same lengths of time prior to and after the action and investigations were carried out into the causal connection between the measures taken and the types of accident occurring. The statistical analysis of the absolute accident figures indicated a decrease at three-quarters of the accident spots. However, this difference was only of statistical significance for one-third of the accident spots. In total, the number of personal injury accidents decreased by 38 %, the number of accidents with major material damage went down by 37 % and the number of accidents with minor material damage sank by 27 %. Additional statistical procedures were used in order to search for a connection between the local parameters of the accident spots and the occurrence of accidents. However, the size of the random sample proved to be too small in view of the many and diverse measures and possible influencing variables. A comparison of the accident spots with regard to the most frequent types of accidents permitted statements to be made about the area in which certain measures could sensibly be taken. It has proven to be the case that traffic-safety work aimed at improving specific accident spots promises the greatest success when based on precise knowledge of the local conditions. The investigation into the economic viability of the measures showed that all the measures which were included in the project could be justified as being economically meaningful. (A)
Samenvatting