MTA Comments on CSA 2010
Feb/22/10 09:02
Download this document as a PDF here.
Federal Motor Carrier Safety Administration
United States Department of Transportation
Comprehensive Safety Analysis 2010 (CSA2010)
(Docket #FMCSA-2004-18898)
Submitted By:
Minnesota Trucking Association
2277 Highway 36 West, #302
Roseville, Minnesota 55113
(651) 646-7351
In May 2009, the Comprehensive Safety Analysis (CSA) 2010 pilot test became a reality for interstate motor carriers in Minnesota. Within days, hundreds of safe motor carriers received warning letters stating that they were deficient, that a case file had been opened, and that they were being monitored for compliance. Many motor carriers in the State had not heard of this test until they received a warning letter.
Now, with several months experience under our belts, Minnesota motor carriers are qualified to offer observations and recommendations based on their real world experience in a pilot program that has become so much more than just the testing of a model.
Minnesota motor carriers are pleased to be able to submit these comments through the Minnesota Trucking Association’s CSA 2010 Task Force (Attachment A). Overall, we support efforts to make our nation’s highways safer. CSA 2010 has the potential to help identify motor carriers who may employ the services of unsafe drivers. With this in mind, we offer the following comments intended to help make CSA 2010 an effective and reliable tool, not only for FMCSA, but also for motor carriers nationwide.
The observations and recommendations that follow come from the people who have the largest interest/stake in the CSA 2010 initiative – the motor carriers themselves. Generally, their concerns are along the lines of:
Enforcement – Inconsistencies by state and in the way violations are noted
Interpretation of regulations varies between states (and within states) resulting in some things being considered violations in one state but not in another. Some violations that may result in a citation in one state may not in another. Some states have more aggressive motor carrier enforcement programs and motor carriers who travel those states are subject to more scrutiny than those who don’t.
In a recent study by a motor carrier whose primary lane is between Minnesota and Pennsylvania, the motor carrier discovered that the states of Indiana and Ohio accounted for 47% of its violations but only 19% of its miles. Yet in nearly every instance, the same drivers which hauled their loads safely through Minnesota, Wisconsin, and Illinois were said to be driving unsafely in Indiana and Ohio. Additionally, this real-life study indicated that these same states did not implement a citation vs. warning system in similar ways. Indiana wrote 42% warnings where as Ohio wrote 78% warnings. The enforcement of violations differs dramatically from state to state. And within that enforcement, the use of warning tickets differs even more dramatically. Before we can introduce warning tickets as reliable data, there must be some standard of enforcement, some method for dismissal, and some lower level of weighting them.
There is little doubt that a pattern of citations for an individual driver is an indicator of unsafe driving behavior which is likely to correlate to preventable accidents. The value of warning tickets though is much less clear. The use of warning tickets varies greatly from enforcement area to enforcement area. Further, warning tickets, when they are issued, always connote an offense which is not egregious enough to warrant a full citation. Below is a list of concerns with the strict tabulation of warning tickets on equal grounds as full citations.
We believe warning tickets ought to be maintained separately, studied and valued at a later date if warranted. In the mean time, we believe citations and/or out-of-service violations serve as a more accurate measure of driver and carrier safety.
We do not believe warning tickets have no value. We, carriers, believe that by evaluating each of them, we can identify safety concerns. CSA 2010 is not the place to be evaluating them.
DataQ – Motor carriers inability to have inaccurate data changed or removed from the CSI database
With the advent of DataQ, motor carriers have been given an efficient means to challenge and have removed from the record, in a timely manner, out-of-service violations noted or cited in error. CSA 2010 changes that.
Because ALL violations of the regulations that make up CSA 2010, actual or alleged, go into the CSA 2010 calculation, motor carriers are very much attuned to actual violations and those noted or cited in error. If a driver is cited for an alleged violation but the driver or motor carrier can demonstrate that a violation did not exist, a court will dismiss the charge. When the motor carrier then challenges the violation through DataQ, many states refuse to remove the violation from the record. So, while legally, the violation does not exist, the driver and motor carrier continue to be penalized for it.
Example – alleged violation of a Texas Rule – A motor carrier was ticketed for the movement of a permitted load a half hour before official sunrise –both driver and pilot car escort documented the time and that the time was legal. The driver went to court, and the citation/ticket was dismissed. Since there wasn’t actually a violation, it was challenged on DataQ. Texas will not remove the “violation,” and it remains on the CSA 2010 record.
Going back to the warning ticket example, the issuing officer and the court assume that warning tickets carry no penalty. Because there is no penalty, the courts will not hear a plea of innocence. Since the defendant cannot have his day in court and be found guilty or not guilty, a warning ticket becomes proof of guilt in CSA 2010 and states will not remove the violation. While we believe there can be some correlation between warning tickets and unsafe behavior, we do not believe this correlation will hold for those drivers, who in fact, did not commit the act for which they were warned.
Since warning tickets will be used by FMCSA to determine a carrier’s fitness, regardless of their validity, carriers will be forced to use the same warnings to eliminate drivers, whether innocent or guilty. Stated simply, warnings which cannot be challenged because they are not convictions can force falsely accused drivers to lose their livelihood. While we believe in erring on the side of safety, if there is no way for a warning to be expunged, then we must recommend against their use in CSA 2010.
We propose that carriers be given the opportunity to appeal and correct any inaccurate data prior to any interventions being administered and/or release of any Safety Measurement System (SMS) data.
Peer Groups – Grouping motor carriers more accurately
Under the current CSA 2010 model, motor carriers are “lumped” together by number of trucks or number of inspections, depending on the BASIC being calculated. This skews the results because enclosed trailer operations (dry vans and reefers) get compared to open equipment operations (flatbed specialized). Over-the-Road (OTR) carriers are ranked against Less-than-Truckload (LTL) carriers that are generally exposed to fewer roadside inspections based on delivering areas.
The Peer Groups under SAFESTAT are composed of both interstate and intrastate carriers. Because CSA 2010 only includes interstate carriers, peer group sizes have been greatly reduced. (Example- Under SAFESTAT, a 1200 unit motor carrier is grouped with 25,000 other motor carriers. Under CSA 2010, there are 700 motor carriers in that peer group. This resulted in one violation causing a 30+ point jump in the motor carrier’s BASIC ranking and a Deficient Assessment for that basic.) All motor carriers should be ranked in CSA 2010 as they are now in SAFESTAT.
Because of the “Bell Curve” ranking of motor carriers, it stands to reason that even if ALL motor carriers improve, one third of motor carriers will always have “deficient” assessments no matter how safe the industry becomes.
In addition to supporting changes to correct the concerns listed above, we support the recommended changes filed by the American Trucking Associations (ATA) in their September 15, 2009 CSA2010, DOCKET #FMCSA-2004-18898.
Those recommendations include:
Accessing the data – Motor carriers, drivers and the public
Past driver behavior is perhaps the best predictor of future driver behavior, and thus, the best way to prevent a crash is for the motor carrier to have complete access to a driver's safety record. In addition to the motor vehicle record, a motor carrier must have access to a driver's drug and alcohol testing history, all of the enforcement actions or warnings issued to a driver during roadside inspections, and a driver's accident history to strengthen the driver qualification process. The language in 391.11(6) should be changed to force a prospective driver to provide all CSA 2010 relevant data to a motor carrier prior to qualification. All driver data should be made available to motor carriers prior to or at least in conjunction with deployment of CSA 2010 nationally.
Further, we recommend that FMCSA keep access to scores and records limited to motor carrier and FMCSA use only. As has been demonstrated by SAFESTAT, public access to records has allowed SAFESTAT to be used for purposes it was never intended for.
Public access to records only causes confusion and poor policy decisions by shippers and anyone else who will use the information to make decisions on safety of carriers. The access will only accomplish a wide-range of users making their own safety conclusions without being safety “experts” causing unintended and undue harm to carriers with scores above certain thresholds
If for some reason, a decision is made to make CSI available to the public, the final system needs a “refined” view for public uses containing good descriptions and warnings about the use of the data
Safety Fitness Determinations
FMCSA has proposed that a motor carrier’s roadside performance, as evidenced by CSA 2010, will determine its safety rating. Proposed safety ratings are Continue to Operate, Marginal, and Unfit. Current safety ratings are Satisfactory, Conditional, and Unsatisfactory and are assigned as a result of a compliance review.
A Satisfactory rating is the “Gold Standard” by which the public, shippers, insurers, and others judge motor carriers. A Satisfactory rating indicates that the motor carrier has met or exceeds a threshold of compliance as determined by the FMCSA and the public is protected.
A Conditional rating indicates that the motor carrier is in substantial compliance, at least with the regulations deemed most important in protecting the public and has demonstrated that it can continue to operate while it works to improve in the other areas.
Unsatisfactory indicates that the motor carrier is not in compliance, chooses not to comply and is a danger to the public.
We don’t see how changing the words will improve truck safety or reduce crashes. In fact, we believe that changing the safety ratings will cause confusion and may cause safe motor carriers to lose business and experience increased costs of doing business. This is especially true with the “Marginal” rating. Marginal has a negative connotation while Conditional is more neutral and allows the motor carrier the chance to improve.
We recommend that the FMCSA keep the current rating system. If for some reason a change must be made, we propose that FMCSA go to a two tier published rating system (Continue to Operate and Unfit) and keep the “marginal” carriers internal to FMCSA while interventions are being pursued.
Additionally, the impact of motor carrier’s ability to obtain insurance or to self-insure, should be studied in light of any changes to SFD’s and how they are calculated.
Determining the Crash Indicator
According to the CSA 2010 methodology, the Crash Indicator is determined by an algorithm that incorporates “Accidents” as defined by 390.5 and the number of trucks a motor carrier claims to operate.
Every day in this country, commercial motor vehicles (CMVs), are involved in crashes. More often than not, the cause of the crash lies with the driver of a passenger vehicle involved. There are many instances where the CMV was legally parked, unoccupied, and someone in a private passenger vehicle loses control and hits or somehow involves the CMV in their crash. All of these “recordable crashes” are used in the Crash Indicator calculation.
We do not believe a carrier or driver should be penalized for accidents which are determined to be non-preventable.
This may be the most important factor for reducing accidents and identifying unsafe behavior.
The very word preventability is causal by definition. If you are unable to prevent something, then your future actions will not reduce it. If your future actions will not reduce it, it falls outside what we hope to accomplish. Specifically, if a car crosses the centerline and crashes into oncoming traffic does this indicate that the oncoming traffic is culpable? Should we counsel oncoming traffic to never behave in a similar fashion again? Should we punish the oncoming traffic or the manufacturer of the oncoming automobiles? If we punish the oncoming traffic, are we likely to see fewer accidents in which a car crosses the centerline and crashes into oncoming traffic? It is clear that the action that needs correction is not the action of oncoming traffic, but the action of the driver of the car who crossed the centerline and collided with the oncoming traffic.
If we are content to measure this inaccurately, then we are choosing to ignore the very reason for CSA 2010. When we consider that almost all statistics verify that 65% to 75% of accidents involving commercial vehicles are the fault of a non-commercial vehicle, it is easy to understand how a very safe carrier may be made to look unsafe if we record accidents in which the other party was at fault. If data is to have statistical significance, the data must correlate to correctable issues. If two thirds of the data does not correlate with correctable issues, then the correlation coefficient loses critical value. This is especially true if neither the preventable nor the unpreventable accident data is static.
Two very different carriers may produce identical statistics.
Carrier A may have eight preventable accidents and zero non-preventable accidents. Carrier B has zero preventable accidents and eight non-preventable accidents. Data collected under CSA 2010 would grade these carriers equally with eight recordable accidents each. The FMCSA would either completely miss the poor performance of Carrier A, or it would unfairly judge the performance of Carrier B. Either of these mistakes would likely result in less safe highways because preventability is not considered.
Preventability
In the vast majority of cases, the preventability is not subjective. Almost all accidents fall into a category.
We, as an industry, believe this is important because we do not like being held responsible for accidents caused by others. The FMCSA should respect the theory that blame should not be placed upon the innocent.
Mileage as the normalizing factor
Carrier A has 100 trucks and runs 10,000,000 miles per year. Carrier B has 100 trucks but has parked 20 of them against the fence and runs 8,000,000 miles per year. Each carrier has five accidents. They look identical on CSA 2010 – 100 trucks, five crashes. However, looking at the crash frequency using vehicle mileage as the exposure, Carrier A has.5 accidents per million miles while Carrier B has.62 accidents per million miles (25% greater than Carrier A).
The US Department of Transportation, various State Departments of Transportation, the insurance industry and the motor carrier industry all use mileage as the exposure basis in determining accident frequency. We recommend FMCSA incorporate fleet mileage into the CSA 2010 algorithm when calculating the Crash Indicator. Fleet miles are more frequently available through quarterly fuel tax reporting or annual vehicle registration applications.
Inspection Point Values – Credit for all inspections and representative severity weights
Within the CSA 2010 methodology, there is no credit given a driver or motor carrier for their efforts to operate safely. “Clean” inspections, inspections in which no violations were found, should be assigned positive “credit” weighting to help offset the “debit” scoring effect that violations have. Currently, “clean” inspections are only counted in the relevant inspections grouping which is used as a divisor in the calculation, heavily diluting/discounting the safety efforts of the motor carrier as represented by clean inspections.
With regard to violations discovered during an inspection, CSA 2010 states that a severity weight is assigned to each regulation, reflecting its relevance to crash risk. Severity weights are numbered 1 to 10 with 10 representing the highest crash risk. Based on the experience of motor carriers in Minnesota, the severity weighting of regulations is not consistent with the severity of a violation or its relevance to crash risk.
Examples include: 383.31(a) Failing to notify licensing jurisdiction of a traffic law conviction, or 383.31(b) Failing to notify motor carrier of conviction within 30 days each have a severity weight of six. 391.11(b)(4) operating CMV without corrective lenses or hearing aid has a severity weight of two. According to CSA 2010, not being able to see or hear while operating a CMV has a lower risk of crash than forgetting to notify the state licensing bureau of a traffic violation. (Driver Fitness BASIC)
177.835(i) Improper transport of explosives (class 1) has a severity weight of two, while 393.104(f)(3) Loose or unfastened tiedown (even an extra tiedown beyond what the securement formula requires) has a severity weight of ten. So, if the load securement formula requires five chains and the driver uses six to be safe, one loose chain out of the six will result in a ten point violation even though the requirement of five chains is met. (Improper Loading/Cargo Securement BASIC)
Close examination of the Improper Loading/Cargo Securement BASIC severity weights indicate a system that targets open-equipment, specialized carriers versus the remainder of the trucking population. In addition, violations of state permit standards – curfew, lights, amber lights, flags, signage placement, weight and measurement discrepancies are being rolled into the Improper Loading/Cargo Load Securement BASIC of the Pilot Program. Many of these “load securement regulations” are in fact non-securement issues and should not warrant a severity weight of 10. Non load securement issues need to be purged from this BASIC as they only confuse it. It stands to reason that more accurate Peer Grouping would help to remove the negative bias that the system seems to have for open equipment carriers.
There are currently 892 regulations that have been assigned severity weights. Of these, 424 regulations are found in the Unsafe Driving, Controlled Substance and Alcohol, Fatigued Driving, Driver Fitness, and Vehicle Maintenance BASICs. Within that 424, there are 14, or 3%, that have been assigned a severity weight of 10, indicating the greatest crash risk.
The remaining 468 regulations assigned severity weights are found in the Improper Loading/Cargo Securement BASIC. Of those 468 regulations, 122, or 26%, have a severity weight of 10. That seems out of proportion with the overall program.
We believe that the entire severity weight assignment should be reviewed by a panel including safety experts from the motor carrier industry and only those regulations, violations of which are true predictors of crash risk should be included in the methodology. Also, severity and time weighting must be applied based on date of the event, not the date the data was entered into the system. There are still many states that are delinquent in data entry.
Hazardous Materials – A separate BASIC?
It is interesting to note that of the 468 regulations that make up the Improper Loading/Cargo Securement BASIC, 353 are hazardous materials (HM) regulations. Along with the 14 HM regulations cited in the Unsafe Driving BASIC and the 3 HM regulations cited in the Driver Fitness BASIC, there are a total of 370 regulations cited that deal with the transport of hazardous materials. We believe that a Hazardous Materials BASIC be developed and added to CSA 2010.
Commercial vehicle drivers – qualifying new and disqualifying poor performers
The stated purpose of CSA 2010 is to improve truck safety and reduce crashes by allowing the FMCSA and its State partners to contact a larger number of motor carriers and drivers, addressing safety problems before crashes occur. Motor carriers have the same objective not only because running safe is good business, but it is the right thing to do.
Good, safe motor carriers address safety problems with their drivers in an effort to reduce crashes. Occasionally, that means discharging a driver who does not respond to safety training and other interventions applied by the motor carrier. Unfortunately, because CSA 2010 data stays with motor carriers for 24 months, those who do the right thing by eliminating unsafe drivers are held responsible even after the driver has been disciplined out of the motor carrier’s system. Safe motor carriers should get credit for doing the right things and credit could be given by removing the discharged driver’s violations from the motor carrier’s CSI. Drivers should take their record with them and not leave them with the motor carrier.
It is clear that CSA 2010 will broaden the number and nature of carrier enforcements actions, but it is unclear how that process will apply to drivers. Will drivers receive some sort of percentile ranking? Experienced drivers may not score as well as inexperienced drivers because of more exposure to roadside events. Will a high CSA 2010 score on a driver involve specific enforcement on the driver, up to and including disqualification? If that is the case, how will 391.15 be changed?
We believe that until the Driver Safety Measurement System (DSMS) is available, and motor carriers have access to ALL of the information needed to make the right qualification decisions, CSA 2010 should not be deployed and should not go beyond its current pilot program stage.
In summary, we recommend that FMCSA must convene a panel of motor carrier safety experts, along with law enforcement (CVSA), FMCSA, and the contractor responsible for program development, to consider these issues and recommendations:
Respectfully Submitted,
Jan W. Markison
Chairman, CSA 2010 Task Force
Minnesota Trucking Association
Attachment A
Minnesota Trucking Association CSA 2010 Task Force
Jan Markison, Daly Agency (Chair)
Geoff Baker, McFarland Truck Lines, Inc.
Jim Fagan, Anderson Trucking Service, Inc.
Jeff Good, Spee Dee Delivery Service, Inc.
Jim Haberkorn, Dedicated Logistics, Inc.
John Hausladen, Minnesota Trucking Association
Bruce Johnson, C.H. Robinson Worldwide, Inc.
Gary Johnson, RAIR
Pete Kane & Stephen Kaufman, Kane Transport, Inc.
Keith Klein, Transport America
Al Koenig, Midwest Specialized Transportation, Inc.
Brian McLaughlin, PeopleNet
Roger Nordtvedt, Land O’ Lakes
Dan Oren, Dart Transit Company
Rita Rue, Lakeville Motor Express
