ATA's American Transportation Research Institute (ATRI) released a landmark report in late October that identified a quantitative model for using past driver-safety performance data to predict future crash involvement. According to ATRI's findings, for example, drivers convicted of an improper or erratic lane change violation within the previous 12 months were 100% more likely to be involved in a crash in the next 12 months than drivers without such a conviction.

ATRI identified 587,000 drivers who had participated in roadside inspections between February 1 and April 30, 2004. Information from the Motor Carrier Management Information System and Commercial Driver License Information System databases was used to track the number of crashes, inspection violations and traffic convictions incurred by these drivers from February 1, 2001 to April 30, 2004.

As a group, these drivers were involved in 46,100 crashes and 2.2-million roadside inspections, including nearly 5-million individual violations.

ATRI then defined the before and after time frames: May 1, 2002 - April 30, 2003 was designated the “history” period, and May 1, 2003 - April 30, 2004 the “future” period. For each violation/conviction type, future crash involvement rates were collected for all drivers with the violation. The rates were compared to those for drivers without the violations.

The final model was developed using a statistical method that combined the individual comparisons.

ATRI should be applauded for its efforts, since the study could have far-reaching implications for fleets.

It's refreshing that this study confirms our internal research. Based on these findings, we have created programs aimed at helping front-line supervisors identify and manage at-risk driver behavior.

The study could provide the impetus for the enforcement community to reallocate “safety oversight” resources. ATRI provides a number of recommendations, including increased targeting of problem driver behaviors for roadside inspections, improved data collection and development of “best practices” problem-driver enforcement strategies.

However, some important questions have been left unanswered — answers that are essential if we're to meet our goal of reducing the total number of truck crashes. First, we have yet to identify the issues that impede our ability to correct problem driver behavior. How do we know who will benefit from time-consuming and costly correction efforts?

Second, we need to assess the effectiveness of front-line supervisors' coaching and intervention strategies. Behavioral scientists have found that safety improvements are only operationalized when those on the front lines are involved. We need to support front-line supervisors by identifying the kinds of best-practice approaches that are most effective.

Finally, FMCSA could use this report as an invitation to develop a carrier-level driver quality measure. Thanks to the ATRI study, we now have even more evidence that at-risk driver behavior leads to increased truck crashes. By definition, carriers that tolerate such behavior are more likely to have trucks involved in crashes. If we can develop a carrier driver-quality measure, we could identify those in need of increased oversight and mandatory safety improvement initiatives.

I urge you to read a copy of the report, which is available at

Jim York is the manager of Zurich Service Corp.'s Risk Engineering Transportation Team, based in Schaumburg, IL.