Evaluating Performance of Highway Safety Projects

2016Final or Interim Report 10/26/2015 -10/26/2016PDFTech ReportFHWA-ID-16-257P#1520350 041 SC16Highway safetyPerformance measurementProgramming (Planning)Safety programsIdaho Transportation DepartmentUnited StatesIdahoIdaho. Transportation DepartmentLoudon, WilliamSchulte, Robert J.DKS Associates, Inc.Idaho. Transportation Department. Division of Engineering Services, Contracting Services, Research ProgramLocal Highway Technical Assistance Council (Idaho)United States. Federal Highway AdministrationUS Transportation CollectionThe purpose of this project was to investigate and document methods that the Idaho Transportation Department (ITD) and Local Highway Technical Assistance Council (LHTAC) can use to evaluate the performance of safety projects that have been implemented. The MAP-21 Act requires states to establish a safety project evaluation process and use the results for setting priorities for future safety projects. ITD and LHTAC currently lack formalized safety project evaluation processes. A literature review was conducted to determine other states' current approaches to safety project evaluation, as well as state-of-the practice methods in this area. Based on this review, two evaluation methods contained in the Highway Safety Manual (HSM), the Empirical Bayes method and the Shifts-in-Collision-Type-Proportions method, were proposed for testing. The Empirical Bayes method was applied for three safety-related highway segment improvement types \u2013 seal coats, rockfall mitigation, and inlaid centerline markings. The Shifts-in-Collision-Type-Proportions method was applied for the same improvement types, plus intersection turn lanes. The Empirical Bayes method indicated significant safety benefits associated with seal coat and rockfall mitigation improvements, and lesser benefits with the inlaid centerline markings. The Shifts-in-Collision-Type Proportions method resulted in average shifts in proportion for the fatal/injury crash type ranging from -0.07 to -0.25 for the highway segment improvements and 0.16 for intersection turn lanes. This study demonstrated that these methods can be applied using currently available crash and roadway data. The number and type of safety projects that could be evaluated was limited by the lack of detailed construction history data on project location and type. Alternative approaches for future automation or semi-automation of the evaluation process include the development of custom software, implementation of commercial-off-the-shelf software, and enhancement of the spreadsheet-based approach used in this study