Citation

Alnawmasi N, Mannering F. Anal. Meth. Accid. Res. 2022; 34: e100205.

Copyright

(Copyright © 2022, Elsevier Publishing)

DOI

10.1016/j.amar.2021.100205

PMID

unavailable

Abstract

In recent years, US States have raised their maximum interstate speed limits from 70 mi/h to 75 mi/h, 80 mi/h and even 85 mi/h. However, understanding the effect that these higher speed limits have had on the frequency and severity of crashes using traditional before and after analyses has been difficult due to possible temporal shifts in driver behavior, and potential changes in vehicle safety technology and highway safety features. Using multi-year data from before and after higher speed limits were instituted on Kansas freeways, random parameters models of crash frequency and resulting injury severity were estimated. Regarding the frequency of crashes, the findings showed that the higher speed limits did not have a significant effect in the mean number of crashes on the 253 studied roadway segments. For injury severity, model-estimation results in one- and two-vehicle crashes show that the factors affecting driver-injury severities have changed before and after the speed limit increase, but changes were also observed in the years before the speed limit increases and the years after. However, using pre-speed-limit-increase model estimation results to predict post-speed-limit-increase injury-severity distributions it was found that the aggregate effect of the changing influences of explanatory variables on average injury severities was relatively small. While the injury-severity estimation results make it difficult to attribute any temporal shifts in parameter values to the increased speed limit, there was a significant increase in the probability of rollover crashes that suggests the higher speed limits may have had some contributory effect on injury severities in single-vehicle crashes.

Language: en

Keywords

Crash frequency; Increasing speed limits; Injury severity; Temporal instability; Unobserved heterogeneity in means and variances