Does 3,675 = 4,000?
New DOT Math?

December 15, 2011

In her November 30 testimony, Federal Motor Carrier Safety Administrator Anne Ferro told the House Oversight Committee on Government Regulation that, "last year, 2010, nearly 4,000 people died in crashes involving large trucks. By the department's estimates, nearly 500 of those would have been related to a fatigued driver."

Later, under questioning from Chairman Jordan, Ms. Ferro related that the data was preliminary and showing an uptick in 2010. When a congressman asked about the 4,000 deaths, Ms. Ferro corrected him, saying in her testimony that it was "upwards of 4,000" and "approaching 4,000."

On December 8, the National Highway Transportation Safety Administration released the data for its 2010 preliminary report. In 2010, 32,855 people were killed in accidents of all kinds. Overall highway deaths dropped to their lowest level since 1949.

Large truck-related fatalities did rise to 3,675.

Still, 3,675 does not equal 4,000. It is a stretch of the imagination to think that 3,675 is "upwards of 4,000." It could be, if one thinks liberally, that going from 3,360 large truck-related deaths in 2009 to 3,675 in 2010 is "approaching 4,000."

We could quibble. The number did rise. But how about the rate? That is, what about the rate per million miles, or the rate by number of registered trucks? Well, you have to dig deeper than the summary report to learn about that. Too bad the deep data is not yet available to dig out those rates.

Keeping in mind that this number represents the people who died in large truck crashes, including truck drivers and the drivers and passengers of other vehicles. Of these deaths, 529 were truckers, 337 of whom were killed in single-vehicle accidents.

The estimated number of injuries sustained in large-truck crashes rose 12% from 2009 to 19,000.

In reading through the preliminary report there's no way to tell how many of these truck-related deaths were due to driver fatigue. And driver fatigue is what the HOS rules are all about.

The NHTSA Fatality Analysis Reporting System (FARS) provides some robust data on crash rates. The latest data available in the FARS encyclopedia is for 2009, so we will have to wait to dig into the latest and greatest data.

It is fascinating to see what is included under the category of large trucks. For most people, that term brings to mind an image of a large Class 8 tractor-trailer rig. But the spectrum of large trucks is much broader, as this list from the FARS reports shows. The category of large trucks starts with step vans, graduating up through single-unit trucks in four different classes.

The FARS data provides the absolute number of accidents and fatalities by detailed vehicle type. But it is always difficult to understand absolute number data when you don't have a frame of reference. Looking at absolute numbers doesn't really give you the context of scale. To truly grasp the numbers, and the impact of the numbers, you need a ratio, perhaps the number of fatalities by vehicle type divided by the total number of registered vehicles by vehicle type. With ratios you start to be able to better understand through context, understanding the total population of the trucks in each class and the fatal accidents created by each class.

In many cases, single-unit trucks below 26,000 pounds GVRW are operated by drivers who are not subject to the hours-of-service rules. As the data becomes available, it will be very interesting to learn whether the increase in fatalities is driven by tractor-trailers or light single-unit trucks. With the data presented today, we just don't know.

It takes a fair amount of effort to dig into the data. In the HOS debate, the data and its validity are becoming the center points of the argument. Each side continues to deride the data used by the opposition. Once you start digging for the data, you start to discover a few inconvenient truths.

The fatality data for 2010 is still preliminary. Developed through manual review of accident reports gathered by state highway and local police departments, the FARS data is far from pristine in quality.
In her November 30 testimony, Ms. Ferro talks about an estimated 500 truck-related vehicle deaths attributed to driver fatigue. Listen carefully to her testimony as she explains how the estimate is based on a survey that indicates that 13% of truck-related fatalities are due to driver fatigue.

More troubling is how estimation data is used as fact in many of the arguments. In an effort to understand the accident rates in comparison to the number of vehicles registered, we had to rely on data from a variety of sources. We present below the 2009 medium and heavy-duty truck registration data, courtesy of RL Polk and Company matched up to the 2009 FARS data.

It is important to understand how the data is used to argue the different points. This table could be used to make a point, that GVW Class 8 Truck Tractor units account for two thirds of fatal crashes while only accounting for 31% of the total registered Heavy Duty Trucks on the road. That is a true statement based on the data in the table.

But consider this: single-unit trucks, the trucks that are mostly not subject to the HOS rules, account for a third of the fatal crashes.

Fun in Front of the House Oversight & Government Reform Subcommittee

FMCSA HOS Listening Session