Modeling railroad trespassing crash frequency using a mixed-effects negative binomial model

A better understanding of rail trespass crashes is needed as more than 400 trespassing related fatalities occur along rail tracks each year in the United States (U.S.). The objective of this research was to investigate factors associated with the occurrence of rail trespass crashes. Yearly crash frequency for counties in the U.S. with train tracks was modeled using a Mixed-effects Negative Binomial Model based on 2012-2016 datasets from the Federal Railroad Administration, the U.S. Census Bureau and National Historical Geographic Information System. Results revealed that key factors affecting rail trespassing crashes include county population density, length of rail tracks in a county, median age and male proportion of the county population, and average train traffic within a county. The findings provided useful information on improving public safety along railroad tracks.