Road safety and road infrastructure expenditure: A bivariate analysis

Abstract

A considerable body of earlier literature has analysed the determinants of road safety outcomes at the aggregate level. A sub-set of that literature has sought to assess the effect of road expenditure, both investment and maintenance spending, on road safety. The results have been inconclusive and, in part, counterintuitive. To clarify that relationship, this paper undertakes a bivariate analysis using European panel data. Three main results emerge. First, there is only limited evidence of longer-term relationships (common stochastic trends) among different measures of the variables. Second, causality tests on stationary variables suggest that the direction of causality depends on how the variables are measured. Third, unrestricted Vector Autoregression (VAR) analyses suggest that there is a positive relationship from some measures of road safety outcomes to road infrastructure investment, and that an increase in one real measure of road infrastructure investment reduces fatalities permanently, if with a lag. While not modelling the true and complex data generating processes, these bivariate analyses help address the salient empirical issues emerging from earlier literature, and the results provide a detailed description of the relationship between road safety and infrastructure expenditure.