Better Method Would Improve Identification of High-Risk Motor Carriers, GAO Reports
A report sent to Congress on June 11 by the U.S. Government Accountability Office estimates the Federal Motor Carrier Safety Administration could improve its ability to identy high-risk carriers by 9 percent if it uses a different method -- a negative binomial regression model. FMCSA now uses SafeStat, a system built on expert judgments rather than statistical approaches, and SafeStat is better than random selection, but it is not the best method, GAO said.
The report was requested by Reps. James Oberstar and John Mica, chairman and ranking member, respectively, of the U.S. House of Representatives Transportation and Infrastructure Committee, and the chair and ranking member of the committee's Highways and Transit Subcommittee. They asked GAO to evaluate SafeStat's ability to identify high-risk carriers after a bus carrying elderly patients caught fire during the Hurriance Rita evacuation in 2005 from the Houston area, resulting in 23 deaths. The company had a record of driver violations but was a low priority for compliance review from SafeStat, according to the cover letter of the report, which is available at www.gao.gov.
The report also found that late reporting of crash data by states had only a small effect on SafeStat's ability to identify high-risk carriers in 2004. If all states had reported their data within 90 days, as FMCSA requires, 299 more carriers (a 6 percent increase) would have been identified as posing high crash risks, GAO found.