The impact of the safe intersections program: a longitudinal perspective

Abstract

Background. The Safe Intersections (Intersecciones Seguras) program was implemented in Mexico City in 2019 to improve road safety for pedestrians on specific street corners. In this document, we evaluate the impact of this program on reducing the number of pedestrian-vehicle collisions and its fatality in the street corners where it was implemented. Methods. We used openly available data from different ministries of Mexico City’s government on the number of pedestrian-vehicle collisions between 2017 and 2020 to generate a geo-referenced, longitudinal database. We then created a control group using propensity score matching methodology. We estimated a multiphase longitudinal model, defining the date of program implementation as the change point. Results. The variables that affected the program allocation were the number of pedestrian-vehicle crashes, the intersection on a main road, the number of streets that converged, and the presence of specific political parties. The multiphase model estimated that, on average, the Safe Intersections program reduced the number of pedestrian-vehicle crashes by 0.014 each additional bimester after the program was implemented. Also, the treated intersections diminished the proportion of lethal pedestrian-vehicle crashes by 0.001 per bimester once the program was implemented. The reduction on the fatality rate, however, was not statistically significant.