Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations
|This report is an archived publication and may contain dated technical, contact, and link information|
Publication Number: FHWA-HRT-05-051
Date: October 2005
Crash Cost Estimates by Maximum Police-Reported Injury Severity Within Selected Crash Geometries
PDF Version (424 KB)
PDF files can be viewed with the Acrobat® Reader®
TARGET CRASH TYPES AND COST LEVELS
Based on the past work by Miller et al. the needs of the red-light-camera evaluation effort, and projected needs in future FHWA safety studies, the decision was made to estimate human capital and comprehensive costs for each of 22 geometry categories.(5) (For a detailed listing of the crash geometries and the definitions and names, see appendix B.)
As noted earlier, the goal of this crash-cost estimation process is to produce a cost for each police-reported crash severity level (i.e., KABCO level) within each of the 22 crash geometries. However, since KABCO levels are much broader than AIS levels, the cost of injury within any KABCO level for a given crash geometry might differ depending on speed limit or urban/rural location. For example, the severity and thus the cost of A-injury angle crashes at rural higher-speed intersections may be greater than A-injury angle crashes at urban intersections. Given this fact, it was desirable to categorize these 22 geometries further by speed limit and either urban or rural location.
Unfortunately, examination of documentation for the databases to be used by Miller indicated no urban/rural indicator in one of the critical files. However, speed limit variables were present. Analyses using the Fatality Analysis Reporting System (FARS), National Accident Sampling System (NASS), General Estimates System (GES), and Highway Safety Information System (HSIS) data from two States were then conducted to compare crash-related speed limits to various urban versus rural designations. There was significant overlap of limits within urban and rural designations in all three files. Based on the distributions and on the need to have sufficient samples sizes in all the subcategories, cost estimates were categorized by locations with speed limits of 72 kilometers per hour (km/h) (45 miles per hour (mi/h)) and below versus 80 km/h (50 mi/h) and above.
A third issue concerned the levels of police-reported severity for which crash cost should be estimated. Initially, the desire was to develop a human capital and comprehensive cost estimate for each level of crash severity (i.e., each KABCO level) within each speed limit category within each of the 22 crash geometries. (See appendix B for definitions of the crash geometries.) In addition, since safety studies sometimes are based on limited data samples in which there are very few fatalities or serious injuries, it was desirable to develop costs when some of the KABCO levels are combined, and costs where crash types are not separated. Preliminary analysis of the databases used in the cost development indicated some problems due to small samples within the most detailed cells. Based on the needs of the RLC evaluation and the available sample sizes, the following levels of costs were ultimately developed. In each case, "cost estimate" refers to an estimate of both human capital cost and of comprehensive cost, and each geometry is always further subdivided by the two speed limit categories, unless otherwise noted.
In summary, the analyses were designed to produce both the human capital and comprehensive costs of crashes with 22 crash types, with and without categorization by two speed limit categories. Crash severity levels within each crash type were defined to allow a variety of different levels of analysis for future studies.
Topics: research, safety, stop red light running program
Keywords: research, safety, crash geometries, red light running
TRT Terms: traffic accidents, accident data, cost estimating