Title: Analysis of road mortality in digital age using Bayesian ecological model: the case of Tunisia

Authors: Karim Kammoun; Aymen Ghédira; Chaker Ben Saad; Nesrine Bouhamed

Addresses: Higher Institute of Industrial Management of Sfax, University of Sfax, Sfax, Tunisia ' School of Economics and Commerce, DEFI Laboratory, University of Tunis, Tunis, Tunisia ' Higher Institute of Transport and Logistics, University of Sousse, Sousse, Tunisia ' Faculty of Economics and Management of Sfax, University of Sfax, Sfax, Tunisia

Abstract: While awareness of the public health burden of road insecurity is recent, the idea that it is developing countries, particularly in Africa, that experience high road deaths is older. Tunisia is an example of this. In this context, our article proposes recommendations through the study of the road mortality rate in Tunisia based on population density and belonging to a geographical unit. To do so, we used the Bayesian ecological regression model whose parameters are adjusted by Gibbs sampling. The analysis shows that the variation in road mortality risk is highest at the delegation level but lowest at the district and governorate levels. An estimated elasticity of −0.25 at the district level means that a 10% increase in population density can lead to a 2.5% decrease in road deaths. Bayes Relative Risk Mapping could help identify areas with high road mortality and strengthen road safety decision making.

Keywords: road safety policy; regional analysis; road mortality; population density; Bayesian method.

DOI: 10.1504/WRITR.2020.111063

World Review of Intermodal Transportation Research, 2020 Vol.9 No.4, pp.393 - 409

Received: 21 Feb 2019
Accepted: 15 Oct 2019

Published online: 23 Oct 2020 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article