Why do traditional freight indices show demand down when the article claims freight is actually strong?

The Cass Freight Index and similar measures rely heavily on LTL (less-than-truckload) and general freight tracking, which miss specialized industrial shipments. Data center construction moves via dedicated truckload, air freight (for high-value electronics), and heavy equipment channels. Air freight imports of GPUs and electrical goods are up 17% year-over-year, and primary metals volumes (steel, switchgear) are elevated, but these don't roll into headline indices the same way general trucking data does. Dr. Miller estimates true freight volume growth at 1–1.5% above year-ago levels once data center activity is factored in.

What commodities are driving this hidden freight demand?

GPU shipments, computer equipment, and electrical goods destined for data center facilities arrive via expedited air freight, then move to construction and distribution sites by truck. Steel, switchgear, and electrical equipment for data center infrastructure and related power/cooling systems move via traditional freight. Heavy equipment manufacturers like Caterpillar and Cummins report higher volumes this year, reflecting demand for generators, transformers, and related infrastructure.

When will carriers add meaningful new capacity to meet this demand?

Dr. Miller projects new carrier capacity entry will not be significant until mid-2027. Long-distance dry van employment bottomed at approximately 494,000 jobs in February 2025 and has edged up to around 500,000, a modest recovery. Historical precedent shows a 9–12 month lag between rate improvement and new carrier entry; the market shifted in December 2025, so mid-2027 represents the expected inflection point. Near-term capacity will remain constrained.

What is the biggest risk to continued freight demand?

Federal Reserve rate-hiking cycles. Historically, trucking demand drops materially within 3–6 months of rate hikes beginning. With the 30-year Treasury yield at its highest level since 2007 and producer price inflation likely to worsen, the Fed may sustain a tightening cycle. Miller projects expansionary cycle dynamics running into early 2027, but expects spot rates to trend softer around mid-to-late 2026 if the Fed maintains hawkish policy.

How long will the data center freight boom last?

The article suggests a typical 18–21 month expansionary cycle from December 2025, pointing to running room into early 2027. However, Federal Reserve policy poses significant downside risk. Spot rates are expected to soften mid-to-late 2026, with contract rates following after the next major RFP season. The boom is likely to moderate sharply in late 2026 if rate-hiking cycles persist or economic conditions deteriorate.

Demand Planning

High Impact

Data Center AI Boom Quietly Driving Freight Demand Higher

Aug 18, 2026

Get tomorrow's supply chain signal

Daily supply-chain brief. Free, unsubscribe anytime.

The signal

The article reveals that data center construction tied to AI infrastructure expansion is generating significantly more freight demand than conventional industry indices capture. Dr. Jason Miller's analysis shows that while traditional metrics like the Cass Freight Index report a 4.5% year-over-year decline, underlying supply chain indicators, including a 17% surge in air freight imports of GPUs and electrical equipment, elevated primary metals volumes, and strong order books from heavy equipment manufacturers, suggest actual freight growth of 1–1.5% above year-ago levels. This divergence reflects how specialized industrial activity can mask overall market weakness, creating pricing pressure and capacity tightness despite headline softness.

The demand picture becomes clearer when examining the data center supply chain: GPU shipments and electrical components arrive via air, then move to distribution sites and construction locations by expedited truckload. Steel, switchgear, and heavy equipment for infrastructure projects follow traditional freight channels. This fragmented demand pattern escapes measurement in standard trucking indices but materially tightens capacity, explaining tender rejection rates near 13% and improved rates that are beginning to attract modest driver recruitment. The article projects that meaningful new carrier capacity will not materialize until mid-2027, consistent with historical nine-to-twelve-month lags in carrier entry cycles.

The biggest downside risk is Federal Reserve policy: historical patterns show that rate-hiking cycles produce trucking demand declines within 3–6 months. With the 30-year Treasury at 2007 highs and inflation signals worsening, demand could reverse mid-to-late 2026, making the current window of tight capacity and elevated rates likely temporary. Supply chain professionals should prepare for a demand cliff in late 2026 while managing current rate escalation and capacity constraints driven by AI infrastructure investment.

Frequently Asked Questions

What This Means for Your Supply Chain

Simulation Suggestion

this month

What if Federal Reserve rate hikes trigger a 20–30% drop in freight demand by Q3 2026?

Simulate a demand shock where overall freight volumes decline 20–30% from current levels starting in mid-2026, driven by Fed tightening and reduced shipper utilization of expedited trucking. Model the impact on capacity utilization, tender rejection rates, and contract rate negotiations in H2 2026. Factor in how data center construction activity might decelerate or shift to inland regions.

Run this scenario

Simulation Suggestion

this week

What if air freight delays push more GPU shipments into expensive expedited ground?

Simulate a scenario where international air cargo delays (port congestion, capacity constraints) force an additional 15–20% of GPU and electrical component shipments to rely on expedited domestic trucking instead of standard LTL. Model the impact on spot rates, tender rejections, and carrier capacity utilization across regional lanes serving data center hubs (Virginia, Iowa, Arizona, Texas).

Run this scenario

Simulation Suggestion

strategic

What if data center construction timelines accelerate by 6 months due to AI demand urgency?

Simulate compressed project schedules where major data center construction projects front-load material delivery and heavy equipment shipments by 6 months (into Q3–Q4 2025 instead of Q1–Q2 2026). Model the resulting capacity crunch in expedited truckload, heavy haul, and primary metals freight. Assess how this pulls forward rate escalation and tender rejection peaks.

Run this scenario

Go deeper:

Supply Chain Simulation: How It Works and How to Choose Software

What discrete-event, agent-based, and Monte Carlo simulation actually do, how to evaluate simulation software, and what it costs.

Demand PlanningHigh Impact

Freight Growth Masks Consumer Collapse: Data Centers Masking Broader Demand Crisis

Sep 29, 2026

Shipping & FreightHigh Impact

AI Data Centers Create 'Capacity Tax' for Traditional Shippers

Jul 30, 2026

Demand PlanningHigh Impact

AI Data Center Boom: Freight Demand Bubble or Structural Growth?