The Capacity Shortage We Keep Creating

Freight is getting more expensive, available capacity is shrinking, and the industry is preparing to spend its way out. Before we start buying more trucks, maybe we should figure out what happened to the hours we already bought.

The October Logistics Managers' Index came out on Tuesday, November 3, and offered a fairly ugly combination for anyone who has to ship anything. Transportation capacity registered 38.0, warehousing capacity 43.5, and inventory costs 88.8. Those are diffusion indexes, not shares of trucks or warehouse space: both capacity readings signal contraction, while inventory costs are rising rapidly. Overall logistics activity hit 70.7, its second-highest reading since March 2022.

Meanwhile, DAT reports that national dry van spot linehaul averaged $2.23 per mile for the week ending October 31, 32.6% above a year earlier, excluding fuel. That's a specific spot-market measure, not the change in every shipper's contract rate, but it tells you what happens when trucks get harder to find at the margin.

There is a predictable playbook for this environment: lock up carriers, pay to protect service, add trailers, reserve overflow space and start telling everyone we're short on truck drivers again. Some of that will be necessary. But the playbook leaves out an inconvenient participant in the capacity equation: the facilities buying the transportation.

It's a remarkable arrangement. A shipper spends months negotiating a competitive freight rate, then its own operation makes that freight less profitable to haul. When the carrier starts rejecting the load, everybody calls it a capacity problem.

A truck can exist and still be unavailable

A truck sitting outside a gate isn't missing from the national fleet. The driver isn't necessarily missing from the labor force, either. But until that visit is finished and the driver can make the next appointment, neither is available to move the next load. Multiply that across enough facilities, and the distinction between how many trucks exist and how much freight those trucks can move starts to matter more than the headcount.

This is the point David Correll of MIT FreightLab made to Congress in November 2021. Working from electronic records for roughly 4,000 over-the-road drivers at two carriers, collected in periods from 2016 to 2020, he found about 6.5 hours of driving in the average working day against an 11-hour legal driving maximum. He challenged the industry's instinct to diagnose every capacity problem as a driver shortage.

Correll then did an exercise that should be taped to every transportation manager's monitor. Using the then-estimated shortage of 80,000 drivers and a workforce benchmark of 1.8 million, he calculated that adding roughly 18 minutes of productive driving per driver per day could provide the equivalent driving capacity. Not 18 minutes off every visit. Not 18 minutes of saved waiting automatically converted into miles. Eighteen more minutes of driving over the course of the day, in the arithmetic of a specific 2021 comparison.

The 11-hour rule is a daily legal ceiling, not a realistic nationwide driving-time target. And Correll's long-haul sample and his 1.8-million-driver workforce comparison weren't identical populations. It wasn't a forecast for 2026, either. It was a way to put an actual size on the capacity we might recover by no longer wasting driver time.

Say a driver gets 30 minutes of avoidable facility time back across a day. Not all of it turns into miles. Schedules, hours-of-service limits and the next available load take their share. If 60% does, that's 18 more minutes of driving. The 30 and the 60 are mine, picked to show the arithmetic, and the arithmetic is the point: fixing facilities across a whole network can matter on the same scale as adding labor.

The hours are real, even when no detention invoice arrives

The American Transportation Research Institute estimated that U.S. for-hire trucking lost more than 135 million hours to driver detention in 2023. That research measured detention under its own methodology; it doesn't count every wasted minute in every type of trucking. It does establish that time stranded at facilities isn't some rounding error in the freight system.

And time is the one asset the industry cannot put back into inventory. A trailer can be repositioned. A driver can be re-dispatched. A missed pickup window can erase the next load from that driver's day altogether. At a national scale, those little failures don't remain little, especially when transportation capacity is already contracting.

This isn't about pretending there's no labor problem, no regional imbalance, no highway congestion and no legitimate cost of adding trucks. Those are real constraints, and no responsible operator would count every minute cut from dwell as another minute of revenue-generating driving. But neither should anyone buy the driver-shortage diagnosis before measuring how much of the existing fleet's day is consumed by shipper-controlled friction.

We've already seen what an operating change can release

In Jake Koppinger's recent Yards First edition, he describes one Primo Brands site where the drop-and-hook turn time improved from 48 minutes to 24 after its driver workflow was standardized. Across the sites where the workflow went live, YardFlow reported roughly 5% more volume while dock-office headcount stayed flat. Those are company-reported operational observations, not proof that every facility will get a 5% lift or that a 24-minute turnaround gain creates 24 minutes of new driving everywhere.

But consider what that result means in the context of the current market. The bottling operation wasn't necessarily short on product, and the carrier didn't necessarily need another truck to make that particular move possible. The constraint was in getting available freight onto an available truck and getting the truck back into circulation. I explored the plant-side version in the last MMYQB, The Bottling Line Wasn't The Bottleneck. The industrywide consequence is what deserves attention now.

Here's the claim I'd be willing to test: if shippers and receivers eliminated avoidable dwell at scale, and the saved time consistently turned into usable truck movements, we could recover capacity comparable to a substantial number of additional drivers. In some markets, that might cover the entire shortfall. I think the industry has been far too quick to assume it needs more drivers before asking how efficiently it uses the ones it has. But the magnitude in 2026 remains an open empirical question, not something a 2021 calculation can settle.

Put the capacity audit before the capacity purchase

If I were responsible for a major shipper's network this week, I'd want to see three numbers by facility before I signed off on another round of capacity premiums: the actual time a truck spends on property; the portion of that time we can eliminate without compromising safety or service; and the portion of the improvement that translates into an on-time next move. Not simply average dwell. Break out live loading from drop-and-hook, inbound from outbound, and the ugly tail of visits that wrecks a dispatcher's entire afternoon. Identify which delays come from appointments, paperwork, labor coverage, trailer location and release readiness, then follow the driver far enough to know whether the improvement changed an outcome.

A gate system that saves ten minutes and leaves the trailer unavailable for another two hours hasn't solved the whole problem. A beautifully optimized dock schedule that makes the driver wait outside the fence hasn't either. The measure that matters is whether the operation releases usable capacity back to the network.

There's a lot of debate about how many more drivers America needs. Fair enough. But there's a less comfortable number that shippers and receivers can start measuring themselves, right now: how many drivers' worth of productive hours their own facilities take off the road.

The LMI is telling us capacity is getting tight. DAT is telling us the marginal truck is getting expensive. Correll's arithmetic tells us why counting CDLs is only part of the answer. If we're going to keep talking about a driver shortage, the people operating freight facilities ought to come with a number too: how many productive driver-hours are we returning to the market, not just how many minutes did we trim from our own averages?

Before you pay for another truck, account for the hours you're taking from the trucks already here.

Sources

1. Logistics Managers' Index, October 2026, released November 3, 2026: the-lmi.com. Readings are diffusion indexes; 50 is the line between contraction and expansion.
2. DAT, "Dry van report: spot rates climb as month-end freight meets shrinking capacity," November 2, 2026: dat.com. National dry van spot linehaul, fuel excluded, week ending October 31, 2026.
3. David H.C. Correll, prepared statement to the U.S. House Committee on Transportation and Infrastructure, November 17, 2021, printed page 40: govinfo.gov. MIT CTL's summary: ctl.mit.edu.
4. ATRI, "New Research Documents Substantial Financial and Safety Impacts from Truck Driver Detention," September 10, 2024: truckingresearch.org.
5. Jake Koppinger, Yards First, "The trucks are already here. Turn them.", October 18, 2026: yardflow.ai. The 48 to 24 minute turn time is YardFlow's measurement at one Primo Brands site; the roughly 5% is YardFlow's analysis of Primo Brands' operating data for the sites running the driver workflow.
6. MMYQB, "The Bottling Line Wasn't The Bottleneck," October 26, 2026: yardflow.ai; "Yards are part of your freight rate," September 14, 2026: yardflow.ai.
7. The 30 minutes and the 60% conversion in the illustration are assumptions chosen to show the arithmetic, not measurements by MIT, ATRI or YardFlow.