Methodology

HaulReport is an independent directory built on public data published by the Federal Motor Carrier Safety Administration (FMCSA) at data.transportation.gov. This page explains exactly how our figures are produced and where their limits lie.

Verifiable facts vs. independent estimates

We separate two kinds of information on every carrier page:

• Verifiable facts — authority status, inspection counts, crash counts, registration dates. These come directly from FMCSA datasets and are reproduced, not derived.
• Independent estimates — reconstructed BASIC percentiles. This percentile is an independent estimate produced by ranking FMCSA's publicly published BASIC measure within a safety-event peer group. FMCSA does not publish percentiles for property carriers; this figure is HaulReport's reconstruction and may differ from any internal FMCSA value. Bands describe percentile ranges among peers, not safety judgments. It is provided for general informational purposes only; it is not a consumer report and may not be used for any purpose governed by the Fair Credit Reporting Act (including credit, insurance, employment, or tenant screening), and it is not a safety rating, a fitness determination, legal advice, or a recommendation to do business with any carrier. Always verify against official FMCSA sources before relying on it.

How the percentile estimate is reconstructed

FMCSA publishes the absolute BASIC measure for property carriers but does not publish the relative percentile. We reconstruct the percentile by ranking each carrier's published measure within a safety-event peer group, and we validate that reconstruction against passenger carriers, for whom FMCSA does publish percentiles.

Until a reconstruction passes our validation gate (≥90% of estimates within ±5 points of the published passenger percentiles), we do not publish reconstructed property percentiles as fact — pages show the verifiable facts and clearly-labeled estimates only.

Crash statistics figures (the /data pages)

The truck accident statistics pages aggregate the public FMCSA MCMIS Crash File — the official record of FMCSA-recordable truck and bus crashes. Every figure is a count of reported crash records, summed one row per crash. These counts reflect reporting volume and exposure (fleet size and miles traveled), not fault, cause, crash risk, or safety performance. They do not represent all U.S. crashes and are not an FMCSA or CSA score.

State figures show where crashes were reported, not where a carrier is based. Recent periods are provisional and may rise as late reports arrive, so the most recent year on each page is marked provisional. We exclude the current partial year from the trend window, and we apply a minimum record count before a jurisdiction gets its own page (avoiding thin, small-sample pages). Per-state severity is not charted because the source tow-away flag overlaps the fatal and injury flags and is not a mutually-exclusive bucket. Figures carry the snapshot date they were captured; refer to the live FMCSA source for the current record.

What this is not

These estimates are not official FMCSA or CSA scores, not a safety rating, not a fitness determination, and not legal advice. FMCSA's own measures carry a data-sufficiency caveat; ours inherit it and add the uncertainty of reconstruction.

See also Data sources and About HaulReport.