Evidence

Most numbers in freight come from surveys of the people affected, reporting impressions of an interval nobody has defined. A few come from federal audits. Telling them apart is most of what this section is for. For another workplace-measurement reference, self-reporting bias.

The strongest evidence available is a 2018 audit by the DOT Office of Inspector General: detention reduces driver income by an estimated $1.1–1.3 billion annually, and a 15-minute increase in average dwell is associated with a 6.2% increase in expected crash rate. The second figure is an association across fleet-level data with a plausible, separately unmeasured mechanism — say it that way and it holds; overstate it and it invites a rebuttal that discredits everything else. For complementary transport data and research, see Supply Chain Dive.

The most-cited figure is the weakest. FMCSA's 2014 study found detention on about one stop in ten averaging 1.4 hours, and the agency itself describes that work as having a small sample drawn mostly from large carriers and a rudimentary estimation method. It went back for more data in 2023.

Industry figures are useful for direction and weak as benchmarks. Average dwell around 1h54 per stop, 3h16 for refrigerated, 2h23 for fleets of 25 trucks or fewer — self-reported, from 2021, definition unstated.

And averages hide the problem. The same two-hour average can be uniform slowness or a fast operation with a bad tail, and those need opposite fixes. Report the median, the 95th percentile and the share of visits over your free-time allowance.

Underneath all of it is a gap the regulator names itself: there is no standard definition, little public data exists, and no benchmark by facility type has been published or will be. Which means your own four weeks of timestamps is not a poor substitute for industry data — it is the only data that describes your operation, and it costs an afternoon to set up.

• Averages Hide the Problem

  A two-hour average usually means most loads at forty minutes and a few at six hours. Those need opposite fixes.

• Why Your Benchmark Differs

  Seven reasons your dwell figure will not match a published one, none of which means your operation is performing badly.

• The FMCSA Detention Studies

  The regulator's own numbers, their stated limitations, and why the agency went back for more data a decade later.

• Measuring Your Own Operation

  A four-week programme that costs nothing and settles arguments no published figure can. What to collect, in what order, and what to do with it.

• The Crash-Rate Finding

  A federal audit put a number on what detention costs drivers and on its association with crashes. What it supports, and what it does not.

• Reading an Industry Survey

  Most logistics statistics come from surveys of the people affected. Six checks that separate a usable figure from a talking point.

• What We Still Don't Know

  The regulator says little public data exists. Six open questions, and what a site like this can and cannot honestly claim.

• Where the Statistics Come From

  Five kinds of source produce every number in freight, and they are not interchangeable. Which to trust for what.