US freight capacity
Three federal enforcement actions are removing drivers from the licensed pool faster than carriers can replace them. Rates are rising on flat-to-falling volume — and a Supreme Court ruling is about to squeeze the brokered long tail.
For most of the post-pandemic cycle, the consensus on truckload freight was demand-led: rates would recover only when shippers returned. That read broke in the second quarter of 2026. Carriers, brokers, and third-party indices now describe a market that is tightening from the supply side — fewer qualified drivers, not more freight.
The enforcement stack
Three distinct federal actions, each on a hard date, are removing capacity from the licensed pool.
English Language Proficiency entered the North American Standard Out-of-Service Criteria on 25 June 2025. The April 2026 edition made it a permanent, nationwide inspection standard. A failing driver is placed out of service — the truck stops, not just the driver.
A non-domiciled CDL restriction took effect on 16 March 2026, limiting issuance, renewal, transfer, and upgrade to H-2A, H-2B, or E-2 visa holders. An early wave removed roughly 13,000 drivers; a state audit found a 53% issuance-violation rate.
Visa revocations added another shock. Jim Filter, CEO of Schneider National, told analysts on the Q2 2026 call that approximately 30,000 drivers had visas revoked, impacting cabotage capacity and accelerating the market's supply-led recovery.
How large is the removal? J.B. Hunt published a model projecting 5% to 12% of CDL holders — 214,000 to 437,000 drivers — exiting over two to three years. That figure is a named carrier's estimate, not a government count; J.B. Hunt's economics improve if truckload tightens. A disinterested third-party measure puts effective capacity down 5.5% in 2025. The direction is confirmed; the magnitude spans a wide range.
Driver-constrained, not demand-led
Filter closed the attribution gap on Schneider's 30 July 2026 earnings call. "We would now categorize the market as driver-constrained." He named the enforcement stack — non-domicile CDL usage, English language proficiency, illegal cabotage, entry-level driver training, ELD tampering — and sized the runway: roughly half of the non-compliant capacity remains to exit through next year.
Reed Loustalot, chief marketing officer at Truck Parking Club and a former Echo Global Logistics executive with no public-equity book, corroborated the mechanism on Bloomberg's Odd Lots podcast: capacity has "structurally been chopped out of the market," and "most of the talking heads were saying that demand returning was the only thing that was going to drive rates up."
We would now categorize the market as driver-constrained.
Jim Filter, Schneider National, Q2 2026
The rate prints match a supply story, not a volume one. Schneider's network revenue per truck per week rose 16% year-over-year; average price renewals accelerated to double digits. Knight-Swift's truckload segment operating income jumped 96.3% to $89.1 million on revenue growth of just 2.8% — operating leverage on price, not volume.
Cass truckload linehaul vs shipments, August 2026
Cass Information Systems, August 2026. Linehaul 153.9; twentieth consecutive year-on-year increase. Shipments 1.038 — first year-on-year gain after a 42-month downturn. Cass still hesitates to call it a major demand recovery.
C.H. Robinson, a freight broker whose purchased-transportation cost rises with truckload inflation, absorbed spot costs up roughly 29 to 30% year-over-year in Q2 while holding average gross profit per load flat. It lowered its market volume assumption from 0% to 5% growth to negative 3%. Freight volume is expected to shrink while rates rise — the supply-driven signature, from a party with no incentive to talk rates up.
Carriers are not adding capacity into the tightness. Schneider raised 2026 EPS guidance to $0.90–$1.10 from $0.70–$1.00 while cutting net capex to $350–400 million from $400–450 million. Werner cut fleet-growth guidance to 16–18% from 23–28% while raising net capex — replacement and age reduction, not expansion.
Derek Leathers, Werner’s chief executive, told a Deutsche Bank industrials summit in Chicago that the supply-led recovery is still in the early stages. July’s “little snippets of news” do not worry him. The administration, he said, is “not backing off its crackdown on bad actors.” Werner’s third-quarter forecast is a 10 to 13 percent year-on-year increase in rate per mile. The company is looking to grow the one-way fleet after nearly cutting it in half since the end of 2022. That would be the first named unit response after the second-quarter cut-capex, replacement-only print. Driver scarcity may still cap the rehire. Watch whether the quarter actually adds trucks. C.H. Robinson’s August North American truckload note left the 2026 dry-van cost-per-mile forecast unchanged at plus 34 percent year on year, tightening “driven more by capacity constraints than by a significant increase in freight demand.” Rail intermodal carloads were about 5 percent higher in July — a partial volume offset for truckload, not a Union Pacific or Norfolk Southern re-rate. Schneider still carries the hangover from a lost dedicated customer. Do not chase that name on this tape.
A September 1 pass put a third measurement on the same spine. DAT’s dry-van report for the week ending August 21 printed spot linehaul at $2.21 a mile, minus fuel — down four cents, or 1.6 percent, week on week, and still up 35.6 percent, or fifty-eight cents, from a year earlier. That is 23.8 percent above the nine-year seasonal average of $1.78. Load posts were flat on the week and up 23.7 percent year on year. Truck posts fell 2.4 percent on the week and 28.4 percent from a year earlier. The load-to-truck ratio moved to 9.88 from 9.64; a year ago it was 5.72. ATA for-hire truck tonnage fell 1 percent in July, to 113.5, and sat 0.5 percent below July 2025. DAT quoted Bob Costello: the tightening owes almost entirely to capacity leaving, not to freight demand coming back. DAT’s own thirty-five-day forecast has dry-van at $2.19 in late September — still about fifty-four cents above the year-ago print. C.H. Robinson left the 2026 dry-van forecast at plus 34 percent; July route-guide depth was 1.41. A minus-28.4-percent truck-post series is not a CDL-holder census. It does not settle J.B. Hunt’s 5-to-12-percent exit against the 5.5-percent third-party cut. Do not re-rate Werner, Knight-Swift, or Schneider on a four-cent seasonal dip. The next catalyst is still the third-quarter print in late October.
Labor Day week added a fourth tape, not a re-rate. FreightWaves SONAR, as of September 1, had the Truckload Rejection Index at 14.19 percent against a 6.05 percent 2025 average. Spot sat at $3.37 a mile; contract at $2.69 — about 18 percent year on year. The spot-to-contract spread had come back near flat after a May-to-July round trip. Same capacity-removal spine as DAT’s seasonal dip and Werner’s 10-to-13-percent third-quarter rate-per-mile guide. Do not re-rate Werner, Knight-Swift, or Schneider on it.
A September 8 FreightWaves note put the holiday jump on the same tape. Tender rejections moved back above 14.5 percent into Labor Day — the strongest holiday increase since 2021 — then settled around 14 percent as of September 8. Spot was about $3.44 a mile, up roughly 2 percent month on month. Contract was about $2.72 plus fuel, around 20 percent year on year. Julie Van de Kamp: “This is still a capacity-driven market.” Same spine. A holiday print is not a re-rate.
After the holiday the volume faded and the rate did not. DAT’s dry-van report, dated September 8, had spot linehaul at $2.21 a mile, up 33.6 percent from a year earlier, and a load-to-truck ratio of 11.47 against 6.68 a year ago — trucks trickling back faster than freight. FreightWaves, the next day, had truckload volumes down 15 percent on Wednesday as the surge left, and the National Truckload Index still at $3.44 a mile, up 47 percent year on year. Capacity trickling back is not a contract-rate collapse. Do not re-rate Werner, Knight-Swift, or Schneider on it. The third-quarter prints are still late October.
Mid-September, C.H. Robinson put a later print on the same spine. Truckload contract rates had come off the July peak and were still about 30 percent above 2025. That is the broker’s cost, not a carrier’s revenue, and it matches DAT’s year-on-year spot still running a third higher. Off a peak is not a collapse. Do not re-rate Werner, Knight-Swift, or Schneider on it.
DAT’s dry-van report for the week of September 12 printed spot linehaul at $2.20 a mile — down a penny on the week, still up 34.2 percent from a year earlier, and still near the top of the historical range. The August monthly is the sharper tape: van spot fell twenty cents, to $2.19, the steepest July-to-August drop in DAT’s sixteen-year history, and moved back below contract — $2.19 against $2.41. It is still more than 30 percent above August 2025. Year-on-year capacity removal is intact. The August print is a seasonal record pullback, not a contract-rate collapse. Do not re-rate Werner, Knight-Swift, or Schneider on it. The third-quarter prints are still late October.
Cass August printed. Truckload linehaul, excluding fuel and accessorials, was 153.9 — up 11.3 percent from a year earlier and 0.7 percent on the month. July had been plus 8.6. Shipments were 1.038, up 2.1 percent year on year and 5.6 percent on the month. That is the first year-on-year shipment gain after a forty-two-month downturn. July’s volume-down, rate-up pair did not hold on a year-on-year basis. Cass still hesitates to call it a major improvement in freight demand — two-year stacked shipments are still down 7.4 percent. FreightWaves counted it as the twentieth consecutive year-on-year linehaul increase, the largest since June 2022. ATA August tonnage is still unpublished; the calendar date is September 22.
The vault’s earlier reading that FMCSA had no driver-supply analysis is stale. The February 13, 2026 final rule estimates about 200,000 non-domiciled CDL holders and says the remaining roughly 194,000 will exit over about five years as credentials renew — about 5 percent of 3.8 million active interstate CDLs in 2024. That is the regulator’s arithmetic, not only J.B. Hunt’s two-to-three-year 5-to-12-percent model. Realized DHS and DOT counts through August 31 are still far below that stock: more than 30,000 illegally issued licenses canceled, more than 28,000 English-proficiency out-of-service orders since June 2025. Direction is confirmed. Magnitude still spans a range. Do not re-rate Werner, Knight-Swift, or Schneider on an August shipment print.
DAT’s dry-van report for the week ending September 18 printed spot linehaul at $2.17 a mile — down three cents, or 1.2 percent, on the week, still up 32.8 percent, or fifty-three cents, from a year earlier, and 19.7 percent above the nine-year seasonal $1.82. Load-to-truck sat at 11.22. The thirty-five-day forecast was about $2.15 by late October. DAT’s own headline treated Cass August shipments as having turned positive — the plus 2.1 percent already on this page. A three-cent ease on a still-elevated year-over-year premium is tape, not a contract-rate collapse. Do not re-rate Werner, Knight-Swift, or Schneider on it.
The next prints are dated into a late-October Werner call — working listed October 28, a street date, not an IR notice. ATA August tonnage, on the MTS calendar, September 22. Cass September October 13. Confirm is linehaul still up year on year in September, with Cass still refusing a demand-led read. Break is linehaul negative month-on-month, or year-on-year fading hard, or capacity coming back.
The broker liability fork
A separate capacity chain runs through the brokered market, not the licensed driver pool. On 14 May 2026, the Supreme Court decided Montgomery v. Caribe Transport II, LLC unanimously: negligent-hiring claims against freight brokers are not preempted by the Federal Aviation Administration Authorization Act. The shipment that injured Shawn Montgomery was brokered by C.H. Robinson.
Defense counsel predict higher insurance premiums and settlement values. C.H. Robinson's CFO Damon Lee said on the Q2 call that coverage runs through end-2026 and insurance expense "will likely rise" — a dated catalyst at the 2027 renewal cycle. Underwriters are beginning to gate coverage on carrier-selection policy.
The near-term capacity effect is harder to see. C.H. Robinson's North American Surface Transportation operating margin was 40.9%, up 280 basis points year-over-year, with volume growth outpacing the Cass Freight Shipment Index for the thirteenth consecutive quarter. The cost has not landed yet. Reed Loustalot argued that if the brokerage model comes under structural pressure, the long tail of small carriers who source freight primarily from brokers is in trouble too — but no broker has yet disclosed a change in carrier mix.
Rail as the structural escape
Union Pacific's proposed merger with Norfolk Southern would create the first coast-to-coast single-line US railroad. The Surface Transportation Board accepted the revised application on 28 May 2026 but held proceedings in abeyance; supplemental information was due 27 July 2026, with completion targeted in the first half of 2027.
The truck-to-rail conversion economics are now quantified. Domestic intermodal runs about 30% cheaper than truckload on a contract basis — well beyond the 10% to 15% discount J.B. Hunt says is typically needed to pull freight off the road. Truck costs rose 16.0% versus rail at 0.7%. Intermodal pricing is expected to rise 3% to 5% as truckload capacity tightens. The merger's own filing claims 2.1 million truckloads removed from roads.
Schneider's intermodal renewals moved only from low-single to mid-single digits — materially weaker than truckload's double digits. The rail re-pricing is a lagging leg, not a coincident one.