Twelve years of evidence: modelling the injury severity of single-vehicle collisions pre- and post-20mph (32 km/h) implementation in Edinburgh and Glasgow

This article presents a comprehensive evaluation framework for assessing the collision severity implications of two competing 20mph schemes in the cities of Edinburgh and Glasgow, UK. To achieve this, road traffic collision severity data are statistically analysed to provide a comprehensive overview of road safety pre- and post-20mph implementation in each case city. Advanced discrete outcome models that account for unobserved heterogeneity, namely, Random Parameters Ordered Probit Models with allowances for Heterogeneity in the Means (RPOPHM) of Random Parameters were estimated to analyse the collision-, casualty- and vehicle-specific determinants of collision severity across different speed limit scenarios: Edinburgh pre- (1) and post-20mph (2) and Glasgow pre- (3) and post-20mph (4). The estimation of four separate models facilitated intracity (in other words, pre- versus post-20mph in each case city) and intercity comparisons of collision severity determinants. In terms of intracity findings, the results suggest that the citywide enforcement of 20mph speed limits, as in Edinburgh, has reduced the risk of vulnerable road users, and especially pedestrians, being involved in serious or fatal collisions, relative to other road users. Conversely, the Glasgow models suggest that the Glasgow 20mph scheme, which was less radical and more targeted, has not significantly altered the disproportionately high risk of pedestrians being involved in severe collisions. Policy recommendations are provided, specifically in terms of how varying 20mph approaches may affect existing road safety inequalities.