What actually moves a CSA score

CSA gets discussed as though it were a single number, like a credit score. It is not. FMCSA's Safety Measurement System sorts the last two years of roadside inspections and crashes into seven separate categories, updates them every month, and scores each one on its own — so a carrier can sit comfortably in six while the seventh quietly climbs toward a level that attracts an intervention.

The stakes for a small fleet are direct, though not quite the way people assume. Since the FAST Act of 2015, FMCSA no longer shows a property carrier's percentiles to the public: the carrier sees its complete results when logged in, and so does enforcement. What stays public is the inspection and crash record and the measures behind most of the categories — which is what a broker, a shipper or an insurer can read before they tender or quote.

The seven categories, which FMCSA calls BASICs, are unsafe driving, hours-of-service compliance, driver fitness, controlled substances and alcohol, vehicle maintenance, hazardous materials compliance, and the crash indicator. Each is measured independently.

Two consequences follow, and both are easy to miss:

• A clean record in six categories does not offset a bad one in the seventh. They do not average.
• Your percentile is relative to carriers with a similar number of inspections, not to the whole industry — FMCSA groups carriers that way before ranking them. A small fleet with few inspections moves further on each individual violation than a large one does.

Each violation carries a severity weight from 1 to 10 within its category, reflecting how strongly it is associated with crash risk, so a brake defect and a paperwork error are not equivalent events. One inspection can add at most 30 severity points to a category.

The severity is then multiplied by a time weight: 3 if the violation was recorded in the past six months, 2 if it was six to twelve months ago, and 1 if it is older than that but inside 24 months. After 24 months it drops out of the measurement.

This is the part worth internalising, because it changes what to do about a bad score:

1. 1A bad stretch is not permanent. Sustained clean inspections genuinely dilute it, and the improvement compounds as the old violations age.
2. 2A recent violation hurts now, not eventually. The gap between an issue found in your yard and the same issue found at a scale is the entire difference.
3. 3Doing nothing is not neutral. Scores drift as your record ages relative to everyone else's.

Vehicle maintenance is the category most within your control on any given morning, and much of what gets written up there is visible before the truck leaves:

• Lamps and reflectors that don't work (§ 393.9) — the easiest class of defect to clear in a walkaround.
• Tires with exposed ply or belt, separation, a flat or audible leak, or a cut down to the cord (§ 393.75).
• Brake components out of spec, and brakes that don't operate (§§ 393.47, 393.48).
• Cargo securement, especially on flatbed and step deck (§ 393.100 and the sections after it).
• Wipers, mirrors and the horn (§§ 393.78, 393.80, 393.81) — the small items that read as neglect to an inspector.

The reason these matter beyond the violation itself is what an inspector infers. A truck presenting several minor defects invites a closer look at everything else, and a full vehicle-and-driver inspection finds more than a driver-only one.

Hours of service is the other common source, and with electronic logging the failures are usually process rather than falsification. The carrier has to review unidentified driving on its ELDs and either assign it to the right driver or annotate why it is unassigned (§ 395.32). Authorized personal use — personal conveyance — and yard moves are special driving categories the driver has to select and annotate on the ELD (§ 395.28). Form-and-manner errors nobody reviews show up the same way: on a report, months later.

Inspection data is not infallible. Violations get recorded against the wrong carrier, duplicated, or entered with errors, and an unchallenged mistake sits on your record for its full life.

FMCSA's DataQs system is the formal route for disputing federal safety data. It requires evidence and it takes time, which is why an incorrect violation that nobody challenges keeps costing a carrier at every broker check and every insurance renewal for the two years it stays in the measurement.

Review your inspection records on a schedule rather than after a problem. Finding an error eighteen months late means you carried it for eighteen months.

1. 1Pull your inspection records and read the new entries, including the clean ones.
2. 2Sort violations by category and look for the same defect repeating — a repeat is a process failure, not bad luck.
3. 3Check anything you do not recognize against your own records, and dispute what is genuinely wrong.
4. 4Feed the repeats back into the pre-trip inspection the driver is already required to make before driving (§ 396.13). A defect found twice at roadside and never in the yard means the walkaround is not happening.

Almost all of this is boring, none of it requires a consultant, and it is the difference between a score that drifts upward and one that quietly improves while the trucks keep running.