Macro models in traffic safety and the DRAG family: literature review

Abstract

In this text, a literature review is given of the models in the DRAG family. These are macro-economic traffic safety models that describe the demand for road use, the accidents and their severity. The DRAG structure typically decomposes the number of victims into exposure, frequency and severity, which makes it possible to search for evidence of risk substitution among these three dimensions. The algorithm uses a multiple equation regression structure, enriched with Box-Cox transformations, that corrects for auto-correlation and heteroscedasticity in the residuals. In the text, the DRAG models are situated within the existing models for traffic safety. Also, macro models in general in traffic safety are discussed, and the major characteristics are highlighted. The original DRAG model is introduced, together with the extensions added in DRAG-2, and followed by an overview of other models in the DRAG family. Apart from the model descriptions, the use of the DRAG models is placed in the context the current research.