An Exploratory analysis of traffic accidents and vehicle ownership decisions using a random parameters logit model with heterogeneity in means

In this study, we present a methodology that addresses a driver’s decisions pertaining to a damaged car following a traffic accident. The choice set includes three alternatives: to replace the damaged car with a new one, replace the damaged car with a used one, and repair the damaged car. A random parameter (mixed) logit model with heterogeneity in the means is specified and estimated to gain more insight into the driver’s decision-making process following a traffic accident.The estimation results of the random parameter (mixed) logit model with heterogeneity in the means indicate that a wide range of explanatory variables affect the driver’s decision to replace a damaged car with a new or used one. For example, as the repair cost to the car value ratio increases, the likelihood of the driver replacing the damaged car with a new one increases. The marginal effect of not repairing a damaged car is twice that of replacing a damaged car with a new car. Additionally, traffic accidents causing critical injuries to drivers decrease the likelihood of replacing a damaged car with a used car relative to accidents that do not generate critical injuries. Other explanatory variables that are likely to increase the propensity of drivers to replace damaged cars with new ones include the car age, type of insurance policy, total household annual income, driver age, and number of workers at the household. Clearly, a driver’s decision to repair or replace a damaged car should not be vi...
Source: Analytic Methods in Accident Research - Category: Accident Prevention Source Type: research