Opinion: FMCSA Refuses to Recognize CSA Defects

By Steven G. Bryan

CEO

Vigillo

In January, FMCSA released its Crash Weighting Report. The four-year study was conducted due to pressure from Congress and the trucking industry. The concern is that the Crash Indicator, one of seven scores in the Compliance, Safety, Accountability program patently misrepresents a carrier’s level of responsibility in certain DOT-reportable crashes.

Some examples of DOT-reportable crashes that create punitive CSA scores for carriers with no logical connection to a carrier’s future safety performance are:

• Intoxicated driver (of second vehicle) rear-ends a legally parked truck.

• Individual takes his own life by stepping or jumping into the path of a truck.

• Tornado or extreme weather event encounters a truck.

• Roadway fails, causing crash involving a truck.

FMCSA conducted this study to determine whether some system could be implemented to remove such incidents and to focus CSA on preventable driver or carrier behavior to make our highways safer.

The study looks at whether police accident reports, perhaps combined with fatality accident reports, new weighting and a dose of statistics, could improve upon the Crash Indicator in a cost-effective, timely manner.

The report winds along for 143 pages and, as much as I enjoy a good tussle over the ins and outs of the Wilcoxon-Mann-Whitney test, this boils down to basic fairness to all concerned, as well as the credibility of CSA.

FMCSA has positioned itself to conclude that:

• PARs do not provide sufficient, consistent and reliable information to support crash-weighting determinations.

• The crash-weighting determination process does not significantly improve prediction of crash risk.

• The crash-weighting process would be costly, would not meet the two-year timeline and would be questionable in its ability to protect the rights of all involved.

FMCSA annually provides in excess of $165 million in taxpayer dollars to states for commercial vehicle enforcement. These funds are allocated to an assortment of training and other programs and to about 14,000 law enforcement officers who partner with FMCSA to conduct commercial vehicle enforcement in the states.

Point No. 1: If $165 million does not result in PARs sufficient to filter out the obviously nonpreventable crashes, insert a full stop here, do not proceed further, the crash data are not fit for a statistical model and must be removed. FMCSA has the burden to show that data it uses are reliable for the intended purpose, not the other way around. It offends all standards of our expectations of government that they spend millions of dollars on a program, declare the result is insufficient to the task, then use the flawed data anyway in a manner that punishes businesses and costs them millions of dollars in lost revenue.

Point No. 2: In the study, FMCSA now applies new weightings to certain crashes (single vehicle) and comes to the conclusion that the statistical model does not improve significantly. There is an old saying in statistics called GIGO (garbage in, garbage out). FMCSA layers the crash data, which never were designed to identify crash responsibility, on top of some new weightings that it has manufactured and says, “Look, it’s no better.” Garbage on top of garbage.

Point No. 3: FMCSA introduces a “requirement” that the public be given the opportunity to weigh in on every DOT-reportable crash — more than 200,000 annually — through a comment process, or lose their rights. Who makes this stuff up? We have a judicial system that protects the rights of anyone injured by another, whether it involves a truck, airplane, drug, or any product that may cause injury. There is no judicial determination in any aspect of CSA, whether it’s the issuance of any of almost 900 different violations that occur daily across the country, or in which crashes FMCSA decides to count in its scoring system. In fact, just last year, FMCSA was forced to acknowledge finally that we have a judicial system and to remove adjudicated citations from CSA scores. FMCSA seems to continue to believe it sits above our judicial system. Note to FMCSA: You don’t. Anyone injured by a truck has all the recourse our legal system provides. CSA does not preclude any possible recourse for damages. Inserting this public comment requirement on every DOT-reportable is FMCSA’s way of rendering it impossible to implement.

Anyone who has been following CSA for the past seven years has witnessed an FMCSA that refuses to acknowledge the many defects of the CSA methodology and data collection process. Study after study has pointed out where FMCSA uses flawed data, flawed methodology and flawed logic to support its stance of absolute refusal to accept ideas for improvement. This Crash Study is the latest example of FMCSA’s unwillingness to listen.

For those who advocate keeping all crashes, you are going to lose CSA completely. There are efforts in Congress that are going to result in CSA scores being completely removed from public view. Ask yourself, “Do I really want to be this inflexible?”

In the early days of CSA, most carriers were willing to accept it and had many ideas on how to improve it. FMCSA would not hear a word of it. I fear now that CSA will be pulled back, and the public will lose any view at all of carrier safety.

These comments were submitted to FMCSA earlier this month. Bryan is the founder of Vigillo, a Portland, Oregon-based data analytics company serving the trucking industry. It is a provider of CSA scorecards and score management tools to 2,000 motor carriers employing 1 million drivers.