FMCSA’s approach to crash reviews is pointless – or worse
Since the Federal Motor Carrier Safety Administration first began publishing motor carriers’ raw safety data nearly two decades ago, the trucking industry has reminded the agency and Congress of the flaws in that data and the dangers of its misuse. While FMCSA has made strides in improving the accuracy and timeliness of inspection and crash data it gets from individual states, one issue remains unresolved and, perhaps, unresolvable: The publication of crashes that could not have been prevented. FMCSA now has an idea for addressing crashes that isn’t just inadequate. It’s potentially worse than the status quo.
Carriers rightly are indignant that recordable accidents – those involving a fatality; injury requiring medical treatment away from the scene; and sufficient damage to require a tow – are published without the context of who caused the accident or even whether the commercial motor vehicle driver could have done anything to prevent the crash. It’s a big deal because plaintiffs’ attorneys mine the data in crash litigation to try to paint a picture to juries of a reckless carrier. In rare cases, that might be the truth, but not often.
As required by the recent FAST Act, FMCSA now proposes to test a process for selectively scrubbing its Motor Carrier Management Information System (MCMIS) of non-preventable crashes. Selective is an understatement. Based on the analysis of TransAdvise, our transportation research affiliate, at most 11% to 12% of recordable crashes would qualify for removal even if a non-CMV driver, if applicable, was convicted of one of four limited violations in every single case.
Realistically, then, far fewer than 10% of crashes would be removed – a pittance given that multiple studies (even at least one sponsored by the AAA Foundation) have pegged the automobile driver at fault 70% or more of the time in automobile-truck crashes. Meanwhile, the existence of any process for reviewing crashes makes the remaining data seem even more reliable – at least to juries. So it’s quite possible that adopting an ineffective process for reviewing crash preventability could hurt carriers more than having no process at all.
Crashes that qualify
The crashes covered by FMCSA’s proposed program fall into two broad categories – those where another driver was convicted of certain specific offenses and those involving single-vehicle accidents under certain limited conditions. A crash would be considered not preventable if documentation established that a (CMV was struck by a motorist who was convicted of one of four offenses or a related offense:
Driving under the influence;
Driving in the wrong direction
Striking the CMV in the rear
Striking the CMV while it was legally stopped
For these crashes, the RDR would have to include evidence of a conviction as defined in 49 CFR 383.5 and 390.5 as well as all available law enforcement reports, insurance reports from all parties involved in the crash and other relevant information.
Carriers also could submit requests for data review (RDRs) through DataQs when the crash did not involve other vehicles in certain situations, such as crashes in which an individual committed suicide by stepping or driving in front of the vehicle; the vehicle was incapacitated by an animal in the roadway; or the crash was the result of an infrastructure failure.
Digging into the data
It doesn’t appeal that hard data is readily found on some of these types of crashes. The crash data available in MCMIS includes virtually nothing about the circumstances of the crash beyond the number of fatalities and injuries and whether there was a towaway. The best seems to be available in FMCSA’s Large Truck and Bus Crash Facts 2014 (https://www.fmcsa.dot.gov/safety/data-and-statistics/large-truck-and-bus-crash-facts-2014). While this report is available on FMCSA’s website, the data comes from two NHTSA reporting systems. The NHTSA databases capture critical data elements not captured in FMCSA’s databases, especially whether the truck struck the passenger vehicle or vice versa.
Using the NHTSA data in the FMCSA report, data is available on the following types of crashes covered by FMCSA’s proposed crash preventability review program:
Truck struck from behind
Other vehicle driving the wrong way
Other driver driving under the influence
Truck strikes animal in single-vehicle accident
As for the other types – driver hitting a legally parked truck and single-vehicle accidents involving suicide or infrastructure failure, published data appears unavailable. However, these all seem to be relatively rare events and probably together represent an essentially negligible portion of the recordable crashes in MCMIS.
Based on the analysis' assumptions and methodology, the data shows the following:
Again, the 9.3% of crashes involving another vehicle would be removed only if the carrier can document that another driver was convicted of one of the four offenses listed above. But a driver could escape conviction for several reasons, and documentation might not always be easy to retrieve even when a conviction occurs. (One reason could be the death of the other driver, but the numbers above exclude situations when the driver of the passenger vehicle died.)
Another aspect of FMCSA’s program raises a concern: The agency says that if a carrier is unsuccessful in challenging a crash as non-preventable, FMCSA may label the crash as preventable in MCMIS and may also increase the severity weighing in the Safety Measurement System. This could be a disincentive for carriers to challenge crashes that likely were – but not certainly – preventable.
To be fair to FMCSA, the agency calls this a pilot demonstration program to remove the “less complex” preventable crashes, leaving open the possibility of expanding the program later. Still, carriers should assume that the agency will go no further any time soon.
