(The views expressed here are solely those of the author and do not necessarily represent the views of FreightWaves or its affiliates.)
A broker has a load and three carriers willing to haul it. Which one is safe? For all the data our industry supposedly drowns in, there is still no dependable way to answer that question. And after this spring, the inability to answer it carries consequences that are harder to shrug off.
Two readings of the Supreme Court’s decision in Montgomery v. Caribe Transport II are making the rounds, and both are wrong. The first suggests the Court made brokers liable whenever a carrier they hired crashes. The second, a reaction to the first, says nothing of substance changed. As usual, the truth sits between them.
Start with what the Court did not do. It did not announce a new duty of care, declare brokers automatically liable, or set any national standard for carrier vetting. It resolved a split among the federal circuits over one question, whether the Federal Aviation Administration Authorization Act preempts state-law negligent selection claims against brokers. The Ninth Circuit had said the statute’s safety exception preserves those claims, the Seventh had said it does not, and the Supreme Court unanimously sided with the Ninth. A plaintiff still has to prove the broker was negligent under ordinary state negligence principles, and a broker that vetted reasonably still has every right to prevail in the case. The substantive law of negligence did not change.
But anyone telling you the decision changed nothing is overcompensating. In roughly half the country, courts had treated FAAAA preemption as a complete defense, a way to get a negligent selection suit dismissed at the earliest stages of litigation and saving the broker substantial sums in litigation defense costs. That defense is gone. In those jurisdictions, the claims now survive the motion to dismiss and head much deeper into litigation, including expensive discovery, unless they settle sooner. And because ordinary care is a question of fact for a jury rather than a question of law a judge can resolve on the papers, there’s a real risk these cases run all the way to trial. That’s not nothing.
So the significance of Montgomery is not a new rule of liability. It’s that the decision removed the procedural exit ramp that let many brokers avoid thinking too hard about carrier selection. At the same time, it left exposed a serious problem our industry has grappled with for decades. There is no dependable way to gauge carrier safety at the moment of selection.
A reliable signal, built on the wrong foundation
Predictably, the scramble to fill the void is underway. Brokers, through the Transportation Intermediaries Association, have petitioned FMCSA for a federal carrier selection standard and a public list of high-risk carriers. The Truckload Carriers Association has urged the agency to modernize how it determines safety fitness. And opportunists from within our industry and without are pitching all kinds of paid “solutions” to the glaring information gap.
I share the urgency, but I don’t share the instinct about where to find the answer.
Most of what is being floated leans on the two things I trust least when it comes to measuring carrier safety: FMCSA’s Safety Measurement System (SMS) data and its safety fitness determinations stemming from compliance reviews and focused audits.
Take the data first. The most persuasive critic of using SMS to judge carrier safety is FMCSA itself. After three trucking associations sued over the public display of CSA data, the agency settled in 2011 and posted a disclaimer that still sits on the SMS website today. It tells prospective consumers of that data the scores are performance data the agency uses to decide whom to monitor, that a flag is “not intended to imply any federal safety rating” under 49 U.S.C. 31144, and that readers “should not draw conclusions about a carrier’s overall safety condition” from the data. Congress drove the point home in the FAST Act of 2015, pulling the property carrier percentiles from public view and ordering a National Academy of Sciences review of the system. The Government Accountability Office had already found, in 2014, that most of the violations feeding the scores occur too rarely to correlate reliably with crash risk and that the agency lacks enough inspection data to compare most carriers with confidence.
To drive this point home, when the American Transportation Research Institute ran CSA scores against actual crash data back in 2012, it found the Driver Fitness BASIC inversely correlated with crashes, with carriers that scored better crashing more, not less. And FMCSA’s own Volpe Center told the agency’s advisory committee the same thing. Read that again. A metric where a better grade predicts more wrecks is not measuring safety. It’s measuring whether carriers are checking compliance boxes or not.
So, when folks urge FMCSA to publish SMS percentiles or otherwise use them as a barometer for carrier safety, they are asking the agency to brand carriers with the very data it has spent many years warning the public not to use this way, and for good reason. In short, that does not fix the data problem; it exacerbates it.
Now the audit model. Currently, a carrier earns a safety rating (satisfactory, conditional or unsatisfactory) essentially one way, through a compliance review, and FMCSA performs roughly eleven thousand a year against a carrier population north of seven hundred fifty thousand. That ratio tells the story. It’s why most carriers have no rating at all, why ratings issued in the 80’s and 90’s are still on the books, and why the agency’s safety rating system draws so much ire. Any “fix” that relies on these ratings or contemplates some kind of expansion of them is severely short-sighted.
A market for lemons
This is not, at bottom, a measurement problem, and we keep getting stuck because we treat it like one. It’s an information problem, and it has a name. In 1970 the economist George Akerlof described what he called the market for lemons, leveraging used cars to make the point. When a buyer cannot tell a good car from a bad one, he will not pay extra for quality he cannot verify, so good cars go unrewarded, manufacturers of those cars quietly leave the market, and overall quality sinks across the board. The insight earned Akerlof a share of the 2001 Nobel Prize in economics.
Carrier selection is a lemons market. A broker cannot reliably see that a carrier is safe at the moment it books loads, so safe carriers struggle to prove their worth and unsafe ones hide in the crowd. We’ve spent years trying to solve it from the inside, refining a scoring system and an audit process never really built for that job. The more useful move is to notice that other safety-critical industries faced a similar problem and handled it differently. Notably, none did it by building a bigger government grading system.
To be fair, reaching outside the usual toolbox is not a completely novel idea. The FAST Act directed FMCSA in 2015 to build a Beyond Compliance program crediting carriers for safety investments above the minimum, and the agency’s own advisory committee recommended that a third party, not FMCSA, administer it. The National Academies pointed toward a statistical model borrowed from other fields. Each time, though, the conversation snapped back to SMS as the engine, keeping us in a decades-long feedback loop that has produced no fruit worth eating.
We should go further. And we should start by asking ourselves a foundational question. Do we want a system that actually improves highway safety and keeps unsafe carriers off the road, or do we want one where we simply check boxes and carry on with business as usual? The honest answer is that, to date, we’ve only built and pitched systems that do the latter. “Do what you’ve done and get what you’ve got,” is the lesson we should take from our work to date. What we really need is not some new way to slice SMS data or safety ratings. It’s a different institutional structure that likely doesn’t lean on these at all.
How other industries handled it
Construction handled a similar problem with the Experience Modification Rate. Every contractor carries a single number built from its actual workers’ compensation loss history, calculated by a neutral bureau from claims data it cannot massage. Below 1.0 means better than average for its class, above 1.0 means worse. And it has teeth. Many general contractors and public owners will not let a subcontractor bid if its EMR sits above a set threshold. Critically, no one asked OSHA to grade every contractor. The insurance data already existed, and the industry standardized it into a figure buyers could rely on in making purchasing decisions.
Aviation handled it with the IATA Operational Safety Audit. Airlines pass a standardized audit of their safety management systems on a periodic basis, run by accredited auditors. The government is not the primary grader. Major airlines won’t partner with or sell seats on another airline that hasn’t passed the audit, so the commercial pressure makes it effectively mandatory, and regulators including the FAA then recognize the credential to streamline their oversight. Registered carriers have run accident rates several times lower than the rest of the industry, a gap that has held for many decades.
Maritime handled it by rating risk and matching oversight to it, and our own Coast Guard does it here. Through its Port State Control program, the Coast Guard scores foreign vessels calling at U.S. ports on a risk matrix built from the operator’s record, the flag, the performance of the classification society that surveys the ship, and the vessel’s own history, then sends its limited inspectors to the worst performers first. Its Qualship 21 program runs the other side of that logic, rewarding the vessels with the strongest records with lighter, less frequent exams. And on the domestic side the Coast Guard goes further still. A towing company can earn its certificate either through direct Coast Guard inspection or through a safety management system audited by an accredited third party under Coast Guard oversight.
Three industries, one pattern. A trustworthy signal built from real safety outcomes, a market that requires it because the parties relying on it have money at stake, and a government that sets the floor and polices the scorekeepers.
What this could look like in trucking
I won’t pretend to have a fully-developed solution or that these external approaches translate cleanly to trucking. Obviously, we need to think carefully about our approach. The point is that the directions come into sharper focus once we stop looking so much inward. If we actually want to assess carrier safety and, in so doing, weed out unsafe carriers, we need solutions that are purpose-built for doing so.
Two pieces do most of the work. A score that measures what a carrier has done, and a credential that measures how a carrier operates. They answer different questions, they cover each other’s gaps, and neither asks the government to do something it cannot.
Start with the score. Picture something a broker pulls up in seconds, the way it already pulls a carrier’s authority and insurance. Not a percentile that ranks a carrier against a shifting peer group, but a measure built from the carrier’s own safety outcomes, similar to an EMR or credit rating. The inputs are the things that actually track with crash risk. How often the carrier has crashed, adjusted for the crashes it could not have prevented. What its roadside inspections turn up, weighted toward the violations serious enough to put a truck or a driver out of service rather than the paperwork issues that do not. And, where it can be obtained, its insurance loss experience, which is the closest thing the industry has to a running tally of real-world crash data. Every one of these is an outcome, not a compliance box check, and that’s the whole point. The violations that feed SMS occur too rarely to predict crashes, as the GAO found. Outcomes do not carry that defect, because they are the very thing we are trying to predict.
The score would need to account for things like exposure (i.e., a fleet running fifty million miles a year and a one-truck operation running fifty thousand cannot be compared on raw counts) and age (i.e., a safety record should also fade with time, so the score weights recent experience more heavily than old, the way an experience modification rate rolls on a multi-year window and the way a credit score lets an old delinquency fall off).
The hard case is the carrier with too little history to measure, and the score would have to treat it honestly. The construction model handles it. A contractor below a size threshold does not receive an individual experience modification rate at all; it carries the class average until it has generated enough loss history to stand on its own. The carrier score could behave the same way. A brand-new or very small carrier is not handed a number the data doesn’t support. It sits at a neutral baseline, plainly marked for lack of data, until it builds a record. And because the absence of a score is itself a handicap for a small carrier trying to win freight, the credential described next gives it a way to prove itself in the meantime.
A neutral party computes the score and publishes it, the way a credit bureau computes a credit score. It’s not a government score, and it is not a vendor’s proprietary one. It’s a standardized number built to a published method, so a carrier that disputes its score can see exactly what produced it and contest the inputs. Transparency of method is what the National Academies urged for SMS and never got. It should be baked in here. What a broker sees is not a mishmash of incoherent data, it’s a confidence level showing how much data stands behind it, and the components that drive it, so a demonstrably safe carrier is distinguishable from one nobody knows enough about yet.
Of course, a score built from claims and inspection history has one unavoidable gap. It cannot speak for the carrier that has no history yet. A second piece could fill that gap, and it’s borrowed from aviation and maritime.
Airlines do not earn their safety standing only by tallying past crashes. They submit to a standardized audit of their safety management system on a periodic basis, performed by accredited auditors against a published standard. The audit doesn’t ask whether the airline had a bad week or month. It asks whether the airline runs a disciplined operation, with real processes for the things that produce safety. Trucking has every one of those things to examine. Whether the carrier timely pulls motor vehicle records and queries the Drug and Alcohol Clearinghouse before it puts a driver in a cab. Whether its drivers are qualified and its files complete. Whether it runs a maintenance program that discovers defects before they lead to accidents. Whether its hours-of-service and electronic logging are managed or simply installed. None of this novel. It’s the same ground a compliance review already covers, asked as a forward-looking question about how the carrier runs rather than a backward hunt for violations.
The credential would be voluntary, and it’s an answer for the carriers the score serves least. A new entrant with no record, or a small fleet that will take years to build one, can sit for the audit and earn a mark that says an accredited examiner looked at how this operation runs and found it sound. For a five-truck carrier shut out of a broker’s roster because it has no score yet, that’s critical. The audit could be tiered, light and mostly remote for a tiny fleet, deeper for a large or complex one, and renewed on a fixed cycle so it stays current. Scores and credentials then work as a pair. The score reports what a carrier has done. The credential reports whether it is built to keep doing it well. A carrier can be strong on one and weak on the other, and a buyer (i.e., shipper or broker) learns more from seeing both than from either alone.
Integrity is critical to the audit piece. Carriers cannot pay for the outcome. We accept this distinction everywhere else. You pay a home inspector, but you cannot buy a clean report on a house with a cracked slab. You pay an accountant to audit your books, but cannot buy a passing opinion on numbers that don’t add up. The integrity rests on four things. The standard is fixed and public, so nothing turns on the auditor’s mood. The auditor is accredited and can lose that accreditation for deviating from the standard. The carrier cannot shop from auditor to auditor for a better result. And the auditor’s fee does not ride on whether the carrier passes. Every credible certification in the economy works on these principles.
So how does it play out in practice? Run three carriers through it. The brand-new authority shows up as unrated for lack of data. At its option, it sits for a tiered, mostly remote audit, and if it runs a clean shop it earns the credential within weeks and competes on day one while its score fills in over time. The established carrier with a strong record carries a high score and a current credential, so brokers book it with confidence, its insurance reflects the lower risk, and the FMCSA, seeing nothing alarming, spends none of its scarce review time there. The reward for running safe is lighter friction, the incentive today’s system fails to really give. And the carrier that starts to slip shows it first in the outcomes, a rising crash rate and worsening inspections, which drives down its safety score. Brokers fall away and insurers reprice long before any federal audit commences, and most will correct in light of market pressures. The few that do not are exactly the ones the outcomes data punishes and flags for a compliance review.
Pulling it together
Every conceivable approach to carrier safety vetting will be met with a healthy dose of criticism. I get it. I’ve criticized many proposed systems myself.
Perhaps the sharpest criticism to the approach I’ve posited above is one I’d raise myself. It goes like this: “In response to a system everyone agrees is broken, you propose to keep the broken part, SMS, and build a whole new system, next to it, and the new system doesn’t even do the thing the public most needs, which is to get unfit carriers off the road.” That’s a fair critique.
To be clear, I’m not proposing to keep SMS. I’m proposing to retire the job it does badly. SMS was built as a government triage tool and then misused as a market tool. The outcomes score does the market work better, on data that actually tracks with carrier safety. As the score proves out, it should take over that role and leave SMS to whatever narrow purposes it actually serves, if any.
The same outcomes data is better targeting for the reviews that do shut carriers down, and the credential is a verification the government no longer has to perform on its own. FMCSA sets and enforces the regulatory floor as it has always done, and then certifies/oversees the groups that perform the market functions of scoring and auditing carriers.
Another objection, and this matters most to brokers, is that even a sound signal does not by itself create a safe harbor, and a plaintiff’s lawyer will wave any low number at a jury anyway. That’s true. A durable safe harbor, the kind that immunizes a careful broker, would almost certainly require an act of Congress. There’s no getting around that. But short of that, there’s no denying that a broker’s reliance on a purpose-built carrier safety vetting system would be infinitely more defensible than the current approach of relying on data that says virtually nothing about actual crash risk, if it can be read to mean anything at all.
In sum, a standardized, defensible, outcomes-based signal beats a world where brokers are judged against an undefined standard using data the FMCSA itself disclaims.
The opening we shouldn’t waste
To be sure, Montgomery did not rewrite the law of broker liability. But it is an inflection point for carrier vetting and highway safety. For too long, we’ve put lipstick on a pig, and what has that gained us? By many accounts, a rising trend of fatalities stemming from crashes involving large trucks and buses. So what exactly do we have to lose by thinking outside the box for once? Carriers, brokers, shippers, the FMCSA and the public at large share a reason to take this up together, renewed by the recent publicity surrounding Montgomery. The worst use of this moment would be to spend it staring into the same box. The better one is to look up and out.
Brandon Wiseman is a partner with the Childress Law Firm and president of Trucksafe Consulting. The views expressed here are his own.
The post After Montgomery, Everyone Wants to Rate Carrier Safety. We’re Looking in the Wrong Place. appeared first on FreightWaves.
