This paper identifies some determinants of possible injuries from crashes that occur at signalized intersections using a conditional probability model. It uses longitudinal data for one city and calculates the marginal effects of these determinants. Among its findings, an airbag deploying in a crash reduces the conditional probability of sustaining possible injury. The size of this decrease is similar for the front airbag or both side airbags deploying. When crashes occur because a driver suddenly falls ill, or because a driver has an existing medical condition that flares up, the results show that the conditional probability of sustaining possible injury increases by 0.1414 and 0.1217, respectively. In comparison the conditional probability of sustaining possible injury increases by 0.0706 when the crash is because of impairment from drugs or alcohol. Additionally, the probability of sustaining possible injury in the event of a crash is 0.0435 and 0.0459 higher in passenger cars and vans respectively than in other vehicles. Comparing these results to those in the traffic safety literature, the marginal effects from national data are similar to those from local level data. (A) Reprinted with permission from Elsevier.
Abstract