U.S. Rail Freight Surges 7.8% Year-over-Year as Intermodal Volume Climbs 13.6% for Week 22

The Analysis

Eight consecutive weeks. That is the number I read first in the Association of American Railroads release for the week ending June 6, not the 7.8%. One strong week is weather. Eight in a row is a system that has been running above its own plan for two months, and a system running above plan for two months has already spent the slack it was counting on for peak.

The arithmetic inside the headline is more interesting than the headline. Total U.S. rail traffic up 7.8% year over year, intermodal up 13.6%, carloads up 1.0%. Those are two different businesses averaged into one figure. Boxes are flooding in because retailers are pulling imports forward through West Coast ports ahead of tariff deadlines. Bulk is barely moving because domestic manufacturing is only starting to come back. Blend 13.6% and 1.0% and you get a 7.8% that nobody on a dock has ever actually experienced.

That blend also hides a transfer of physical capacity. A railroad has one crew pool, one locomotive fleet, one set of yard tracks and a fixed number of ramp spots. Every extra intermodal lift takes a switch crew and a track segment away from a carload shipper. The intermodal surge is not adding capability to the network. It is borrowing capability from the bulk side, and the bulk side is being paid 1.0% to hand it over.

Follow one box and you can see where the queue forms. It comes off a ship at Los Angeles or Long Beach, gets grounded or goes straight to an on-dock rail loader, waits for a well car, rides to a ramp in Chicago or Alliance or Memphis, gets grounded again, waits for a chassis, gets picked up by a drayage truck, and then waits once more at a distribution center gate for an appointment that may be two days out. Five waits. The ship is not the problem in that chain, and the ship has not been the problem for a while.

The first physical capability to run short is not track. It is parking. A ramp is a parking lot with a crane and a schedule, and its capacity is measured in slots, not in trains. When intermodal volume runs 13.6% above last year, the ramp is asked to hold 13.6% more boxes on the same asphalt, and the dwell clock starts on every one of them. The second capability is chassis. Pools are sized to a normal mix, and a normal mix assumes boxes leave as fast as they arrive. The third is the drayage driver, who will not sit four hours at a gate for a two-hour move. The fourth is the receiving dock, which is a human constraint dressed up as a scheduling problem.

The causal chain runs backwards from the headline, and this is the part worth holding onto. Retailers accelerated shipments through West Coast ports to inland distribution centers ahead of tariff deadlines. That decision was made weeks ago. What the railroad association printed for the week ending June 6 is a receipt, not a forecast. The containers behind that 13.6% were bought, booked and loaded before anybody saw this number, which means the freight is already committed and the network has to absorb it whether or not the ramps are ready.

Who hurts first depends on what they own. An importer with its own chassis pool, contracted drayage and a dedicated receiving door pays a little more and complains a little. A mid-size importer renting chassis from the pool and using a broker's drayage gets hit twice, once on availability and once on per-diem that starts before anyone calls to tell them. A bulk shipper with a private siding reads a headline about a rail boom and then finds that its empty cars arrive late and its spotting window has moved, because the crew that used to switch its yard is now switching an intermodal ramp. Exporters take the same cut from the other end. Their inland leg shares ramps and drayage with the import boxes, and the import boxes pay better.

The timing is not mysterious. The figures land with a week's lag. Ramp dwell and chassis shortages show up ten to twenty-one days after the volume does. The distribution center appointments break after that, when trailers arrive faster than the doors can take them. Peak season is described in the release as intensifying, which is a polite way of saying the next six weeks are the test and the buffer is already gone.

Here is what the coverage leaves out. A streak of eight weeks is reported as strength, and it is, but a streak is also a measure of consumption. Every week above plan eats a little of the reserve that exists for the week something breaks. Nobody publishes ramp dwell or chassis availability next to the volume number, so the industry reads a demand story while the invoice is being written in accessorials. And the 7.8% blend does real damage to bulk shippers, who are told the network is booming while their own service quietly deteriorates. Rail is not one product. Averaging it into one number is a reporting convenience that costs somebody money.

A small calculation, and the numbers here are mine, not the article's. Suppose you move forty FEU a week from the West Coast to a Midwest ramp, and your ramp dwell goes from two days to four. Assume chassis per-diem of thirty-five dollars a day, which is an assumption for illustration, not a figure from the release. Forty boxes times two extra days times thirty-five dollars is twenty-eight hundred dollars a week, roughly eleven thousand a month, and none of it appears on a freight invoice. Add a missed appointment at the distribution center and a redelivery charge lands on top. That is the shape of the bill in a front-loaded market. It is small per box and enormous per quarter.

The counterfactual is worth stating plainly. If the import pull-forward had been spread evenly across the second quarter, the same boxes would have moved through the same ramps, just without the compression. The network would still have been busy. It would not have been over its plan for eight straight weeks. Compression is the whole story, and compression is what turns a rate problem into a capacity problem.

The mistake I see most often is treating rail as the cheap leg and trucking as the expensive one. On a normal lane that holds. In a front-loaded market, the last three hundred miles decide the cost. A box that sits on a ramp for four days because nobody booked an appointment is not cheap, it is just cheap on the linehaul line of the invoice. The drayage, the per-diem, the redelivery and the lost sales at the distribution center all live somewhere else.

So the action for today is not a strategy review. It is a phone call to your drayage provider and your ramp contact before Friday, asking for dwell by lane for the last six weeks. If a lane moved more than a day, that lane is your problem this month, and you fix it by booking appointments and chassis before you tender, not after.

↑ Back to the key points

— By Ömer Kaya