1. Home
  2. TRID
  3. View Record
https://bit.ly/TRBpodcasts

Determining the Safety Effects of Differential Speed Limits on Rural Interstate Highways Using Empirical Bayes Method

A differential speed limit (DSL) is defined as being one limit for automobiles and a different limit for commercial motor vehicles (trucks) whereas a uniform speed limit (UNI) is defined as a single limit for cars and trucks. Because states enact a DSL solely in order to improve safety, assessment of DSL’s safety impacts is of significant importance to the transport community. Previous before-and-after studies could not fully investigate DSL’s impact on crashes due to the limited periods of time used in these studies. A different genre of studies based on the comparison of safety effects at different physical sites, such as I- 64 in the western portion of Virginia (UNI) and the adjacent section of I-64 in the eastern portion of West Virginia (DSL) were also inadequate because of the limited data available at the time. Thirteen years have passed since the enactment of the Surface Transportation and Uniform Relocation Assistance (STURA) Act, rendering a new set of data available for further study regarding the safety effects of DSL. Using the Empirical Bayes method for before-after safety analysis, this study developed a multivariate crash estimation model (CEM) using the before treatment years data and predicted what the safety would have been if there was no DSL enactment for the after treatment years. This study used data from seven states, which either kept the same speed limit strategy since 1990, or changed their strategy at least once. Six types of crashes (total number of crashes, total number of fatal crashes, total number of rear-end crashes, total number of crashes with truck involved, total number of fatal crashes with truck involved, total number of rear-end crashes with truck involved) were selected for analysis. The evaluations of DSL implementation was then carried out by comparing the predicted “would have been” crash counts and the actual crash counts of the after treatment period. A nonlinear relationship was found between crash counts and section length, and between crash counts and the annual average daily traffic (AADT). The results vary for different types of crashes through different states. The results, however, generally showed that as time passed, the actual total numbers of crashes for the after period were greater than the predicted “would have been” after total numbers of crashes. Whether this difference was caused only by the policy change of DSL or other factors that contribute to differing safety conditions is therefore not conclusive.

  • Record URL:
  • Supplemental Notes:
    • This research was sponsored by the U.S. Department of Transportation, University Transportation Centers Program.
  • Corporate Authors:

    University of Virginia, Charlottesville

    Center for Transportation Studies, P.O. Box 400742
    Charlottesville, VA  United States  22904-4742

    Mid-Atlantic Universities Transportation Center

    Pennsylvania State University
    201 Transportation Research Building
    University Park, PA  United States  16802-4710
  • Authors:
    • Sun, Xin
    • Garber, Nicholas J
  • Publication Date: 2002-5

Language

  • English

Media Info

  • Media Type: Web
  • Edition: Research Report
  • Features: Appendices; Figures; References; Tables;
  • Pagination: 117p

Subject/Index Terms

Filing Info

  • Accession Number: 01046058
  • Record Type: Publication
  • Report/Paper Numbers: UVACTS-14-5-36
  • Files: UTC, TRIS, ATRI
  • Created Date: Apr 11 2007 8:12AM