How Carrier Safety Scores Are Calculated, and Why the Method Is Contested
The safety score you can look up on a trucking company is built from roadside inspection results, and the method used to turn those results into a number has been argued about for more than a decade. The argument is not academic. It decides which carriers get investigated, which lose contracts, and which appear safe to anyone checking.
Here is how the scoring works, what the objection to it is, and what it means for choosing who carries your vehicle.
What the Score Is Built From
Every roadside inspection produces a record: what was checked, what violations were found, how severe each one is. Those records are grouped into categories covering unsafe driving, hours of service compliance, driver qualification, vehicle maintenance, controlled substances and crash history.
Within each category, violations are weighted by severity and by how recent they are, then the carrier is compared against others of a broadly similar size. Carriers above a threshold are prioritized for intervention.
Two features of that design cause most of the trouble. The comparison is relative, so a carrier’s score moves when other carriers’ scores move. And it depends entirely on being inspected.
The Data Sufficiency Problem
This is the objection that has never gone away, and it is a real statistical problem rather than industry grumbling.
A carrier with two trucks might be inspected three times in two years. A carrier with two hundred trucks might be inspected four hundred times. Both get a percentile. The small carrier’s number is built on so little data that a single bad inspection can move it dramatically, while the large carrier’s is stable.
Inspection frequency also varies by geography. A carrier running corridors with heavy enforcement accumulates a thick record; one running quieter states accumulates a thin one, for the same standard of operation. Neither is necessarily safer.
The Proposed Fix, and Why It Was Rejected
Independent review has repeatedly pointed toward a more statistically rigorous model, one that treats inspection results as indicators of an underlying, unobserved safety level and handles carriers with sparse data more honestly than a straight percentile does. Item response theory, the approach used in educational testing, is the best known candidate.
Regulators declined to adopt it, and the stated reason is worth quoting fairly: it is too opaque. A carrier cannot look at such a score and see which violation moved it or what to fix. A percentile is crude but legible; a latent-trait model is defensible but effectively a black box to the person being scored.
That is a genuine trade off between statistical accuracy and the transparency owed to a business being regulated, and reasonable people land on both sides of it.
What a Compliance Review Adds
Separately from the scoring, a carrier can be subject to a compliance review: an on site audit of records and operations that produces the formal safety rating of satisfactory, conditional or unsatisfactory. That is a much richer assessment than a percentile built from roadside stops.
The catch is that most carriers have never had one, because reviews are resource intensive. So for the majority of companies the inspection derived score is all that exists, which is precisely why its statistical weaknesses matter so much.
What This Means for You
Three practical consequences, and they hold regardless of how the methodology argument resolves.
Read the underlying data, not just the percentile. The out of service rate, the raw inspection count and the crash count tell you more than a ranking that depends on who else got inspected.
Weight the inspection count. A carrier with a handful of inspections has a score with wide error bars. Neither a good nor a bad number from a thin record should carry much weight.
Remember what it does not measure. None of this covers damage claims. Loading a car badly is not a roadside inspection item, so a carrier can have an excellent safety record and still handle vehicles carelessly. Safety data describes the operation of the truck, not the care taken with what is on it.
How the ratings themselves work is covered in how carrier safety ratings work, and the way some operators escape a bad record entirely in chameleon carriers.
Why It Matters Beyond Regulation
The score is not only a regulatory instrument. Shippers screen carriers on it, insurers price on it, and brokers use it to decide who they will dispatch to. A carrier that crosses a threshold can lose contracts long before any enforcement action arrives.
That is why the accuracy question has teeth. A small carrier whose percentile swung on two inspections is not just facing extra scrutiny; it may be losing work on the strength of a number built from very little evidence. It is also why some operators respond by re registering under a new identity rather than fixing the underlying problem.
A final practical note. If a company will name the assigned carrier before collection, you can run this check yourself on the actual truck coming to your address rather than on the broker. That is the single most useful thing any of this data enables, and it costs five minutes.
Common Questions
What are safety scores built from? Roadside inspection and crash records, weighted by severity and recency, grouped into categories and compared against similar carriers.
Why is the method disputed? It is a relative percentile that depends on inspection volume, so carriers with little data get unstable scores.
Why was a better model rejected? Regulators judged it too opaque: a carrier could not see what moved its score or what to fix.
What should I actually read? Out of service rate against the national average, the number of inspections behind it, and the crash count.
Does a good score mean my car will not be damaged? No. Loading damage is not an inspection item and is not measured at all.
Get a Quote
Call Ship A Car, Inc. at (866) 821-4555 with both zip codes and your date range, or price the move with the instant calculator.
