A single freight train can replace three hundred trucks on the highway. And yet, in most of the world, most overland freight still moves by truck. So Daniel sent us this one — he's asking how important rail freight actually is in today's global logistics picture, what the real modal split looks like between truck and rail for overland routes, and what drives the decision for logistics operators choosing between them. There's a lot of tension here between rail's theoretical efficiency and its real-world friction.
That tension is exactly where this gets interesting. Let's start with the numbers, because the global variation is wild. In the United States, rail carries about seven to eight percent of freight ton-kilometers. The EU sits around eighteen percent. China and Russia are up near fifty percent. Same planet, same basic physics — steel wheel on steel rail is always more efficient than rubber on asphalt — but completely different outcomes.
What's actually producing that spread? Is it just infrastructure spending, or is something deeper going on?
It's three things layered on top of each other. Infrastructure, obviously — China has poured one point two trillion dollars into rail infrastructure since twenty ten, much of it freight-corridor-specific. But underneath that, you've got geography and industrial structure. Russia moves enormous quantities of bulk commodities over vast distances — coal, ore, timber — and those are exactly the cargoes rail was born to carry. The US has a similar bulk-commodity backbone. Europe has a lot more manufactured goods moving shorter distances between dense urban clusters, which favors truck.
The modal split isn't really a policy choice so much as a reflection of what a country's economy physically looks like.
But policy is a huge lever. The EU has an explicit target of thirty percent rail freight share by twenty thirty under the Green Deal. They're not going to hit it — they're stuck at eighteen and the trend line is basically flat — but the fact that they set the target tells you it's a political priority. The US has never set a modal-shift target for freight. It's been market-driven, for better or worse.
The market, left to its own devices, chose trucks.
The market chose trucks for most things and rail for the things only rail can do. Which brings us to the economics, because this is where the rubber meets the road, so to speak. The fundamental fact is that rail is three to four times more fuel-efficient than truck per ton-mile. The US Department of Energy pegs rail at about four hundred fifty ton-miles per gallon versus truck at about one hundred thirty. That's a massive efficiency gap. But fuel is only one line item.
Right, because the train doesn't back up to the loading dock.
And this is the thing most people miss when they look at the raw efficiency numbers. Rail requires transshipment at both ends — the first mile and the last mile. Your cargo gets trucked from the origin to a rail terminal, loaded onto a train, moved hundreds or thousands of miles, unloaded at another terminal, and trucked to the final destination. That drayage at both ends adds one to two days and fifteen to twenty-five percent to the total cost for most shipments under five hundred miles.
There's a break-even distance where rail's line-haul efficiency finally overcomes the drayage penalty.
That distance is roughly five hundred miles. Below five hundred miles, truck almost always wins on total door-to-door cost and transit time. Between five hundred and fifteen hundred miles, rail becomes competitive, especially for heavier cargo. Above fifteen hundred miles, rail has a clear cost advantage — but only if the route actually exists.
Which is a bigger "if" than people realize.
So let's make this concrete. Imagine a shipment of auto parts from Detroit to Dallas — about eleven hundred miles. By truck, that's two days door-to-door at roughly two dollars fifty cents per mile. By rail, you're looking at four to five days at about a dollar twenty per mile for the line haul, but then you add drayage trucking at both ends — another four hundred to six hundred dollars total. Your net savings come out around thirty percent, but your transit time more than doubles.
The logistics operator is making a judgment call: is this cargo time-sensitive enough that the savings aren't worth the delay?
That judgment depends entirely on what's in the container. Auto parts are a great example because they sit right on the boundary. If it's a just-in-time supply chain feeding an assembly line, you pay the truck premium every time. A line shutdown costs far more than the freight savings. If it's replenishment stock for a parts warehouse, rail makes perfect sense.
How does that actually play out in practice, day to day? Who's making that call and what does the conversation sound like?
Usually it's a logistics manager staring at a screen at seven in the morning with a coffee that's gone cold. They've got a purchase order for twelve pallets of brake calipers that need to be at the plant by Thursday. The truck quote is forty-two hundred dollars and guarantees Thursday delivery. The rail quote is twenty-nine hundred, but the intermodal ramp is forty miles from the plant, the train schedule means it arrives Wednesday — or maybe Friday if there's a delay in Chicago — and then you still need a drayage carrier with availability. The logistics manager has been burned twice this quarter on rail delays. They click "approve" on the truck quote and move on to the next problem.
The theoretical thirty percent savings evaporates against the reality of "I cannot be the person who shut down the assembly line.
And that's why rail's market share for manufactured goods keeps slipping. The cost savings are real on a spreadsheet, but the career risk of a missed delivery is also real, and spreadsheets don't capture that.
Then there's the stuff where there's no decision to make at all.
Coal from Wyoming's Powder River Basin to a power plant in Georgia. Eighteen hundred miles. Unit trains of a hundred twenty-plus cars, no transshipment, straight from mine to plant. Rail cost is roughly three cents per ton-mile. Truck would be about fifteen cents per ton-mile. There is no trucking option — it would be economic suicide. That's why bulk commodities make up sixty to seventy percent of US rail tonnage. Coal, grain, ore, chemicals, petroleum products. These are heavy, low-value-per-pound, predictable flows. Rail was designed for them.
The US rail network was built around exactly those flows. The Class I railroads — BNSF, Union Pacific, Norfolk Southern, CSX — they're private companies running private infrastructure optimized for long-haul bulk.
Which is both a strength and a weakness. The US has about a hundred forty thousand route miles of rail, down from two hundred fifty-four thousand in nineteen sixteen. Seventy percent of it is single-track. That means when something goes wrong — a derailment, a crew timeout, a signal failure — you can't just route around it. The whole line stops.
East Palestine in twenty twenty-three was the dramatic version of that, but I imagine the day-to-day congestion is the real story.
The day-to-day is where the capacity constraint bites. The US rail network is, by global standards, under-invested and over-optimized for a narrow set of cargo types. The Class I railroads have been enormously profitable running this model — operating ratios in the low sixties, meaning they keep thirty-five to forty cents of every revenue dollar as profit. But that profitability came from focusing on the most efficient lanes and cargoes and shedding everything else.
Which brings us to Precision Scheduled Railroading.
The operating philosophy that reshaped North American rail starting around twenty eighteen. The idea is simple: run fewer, longer trains on fixed schedules, minimize yard time, maximize asset utilization. And it worked, by the metrics that matter to shareholders. Asset velocity improved twenty-five percent. Operating ratios dropped. Stock prices went up.
The shippers hated it.
Some of them still do. Because PSR means the railroad runs trains when it's efficient for the railroad, not when the shipper wants their cargo moved. On-time performance for some shippers dropped ten to fifteen percentage points. If you're a grain elevator in Kansas and the train shows up two days late during harvest, that's a real problem. Some cargo shifted to truck not because truck was cheaper, but because truck was reliable.
The railroad optimized itself into losing business.
It optimized itself into being very profitable at a smaller set of services. Whether that's a good long-term strategy is an open question. But it highlights something important: modal choice isn't just about cost per ton-mile. It's about reliability, frequency, and whether the service actually exists for your lane.
Let's talk about Europe, because the picture there is completely different.
Fragmented is the word. The EU has national rail networks built by individual states over a century and a half, with different electrification standards, different signaling systems, different loading gauges, and in some cases — looking at you, Spain and Portugal — different track gauges entirely. A train crossing from Germany into France needs to switch locomotives or use multi-system units. Crossing from Poland into Belarus, you're dealing with a gauge break where they literally lift the cars and swap the bogies.
Wait, literally lift the cars? How does that even work?
It's called a gauge-changing facility, and it's exactly as slow as it sounds. The train pulls into a shed. Hydraulic jacks lift each car — or sometimes the entire trainset — the old bogies are rolled out, new bogies of the correct gauge are rolled in, and the car is lowered onto them. The process takes hours for a full freight train. There are some variable-gauge systems that can shift wheel spacing on the fly at low speed — Spain's Talgo system is the best-known — but those are passenger-focused and not widely deployed for heavy freight.
Every border crossing is a multi-hour production.
Which adds time and cost at every border. The EU has been trying to harmonize this for decades with limited success. The Rail Freight Corridors initiative was supposed to create seamless cross-border paths, and it's helped, but the fundamental problem is that national infrastructure was designed for national priorities. France built its rail network radiating from Paris. Germany built a polycentric network connecting multiple industrial regions. They don't connect to each other elegantly.
A shipment from Barcelona to Berlin has to navigate two gauge changes, multiple signaling systems, and at least one locomotive swap.
Each of those transitions is a point where the shipment can sit for twelve hours waiting for a crew, a slot, or paperwork clearance. The average speed of a cross-border freight train in the EU is about eighteen kilometers per hour when you factor in all the dwell time. A truck on the same route averages sixty-plus.
Yet the EU wants thirty percent rail share.
Driven by the Emissions Trading System extension to transport and the broader decarbonization push. And the logic is sound — rail produces about seventy-five percent fewer CO2 emissions per ton-mile than truck. But here's the catch: that advantage depends heavily on electrification. About sixty percent of EU rail is electrified, which is decent. In the US, it's about one percent. American rail runs on diesel.
The green case for rail in the US is...
It's still better than truck, but it's not the slam dunk the headline numbers suggest. Diesel locomotives emit significant CO2 and, critically, particulate matter — especially around rail yards in urban areas, which tend to be in lower-income neighborhoods. There's an environmental justice dimension here that the "rail is green" narrative often skips over.
Like adopting a feral cat. Seems straightforward until you look at the details.
actually a very good analogy for rail policy generally.
Let's pivot to the disruptors. You mentioned autonomous trucking earlier, and that seems like it could upend this entire calculus.
This is the thing that keeps rail executives up at night. The rail industry's cost advantage over truck rests substantially on labor efficiency — one train crew of two or three people moves ten thousand-plus tons, while one truck driver moves about twenty-five tons. But if you remove the driver from the truck equation, the math shifts.
How much of trucking's cost is the driver?
Roughly thirty to forty percent of the per-mile cost of trucking is labor. Autonomous trucks also don't have hours-of-service restrictions — they can run twenty-two hours a day with only refueling stops. Aurora, Waymo Via, and a few others are targeting commercial driverless operations on highway lanes by twenty twenty-eight to twenty thirty.
A truck that never sleeps, with no driver to pay, on routes where rail currently wins on cost.
The break-even distance for rail versus autonomous truck might push out to eight hundred or a thousand miles. That eats into a big chunk of rail's intermodal business. But there's a countervailing force: capacity. One double-stack intermodal train carries the equivalent of about two hundred eighty trucks. Even with autonomous driving, you can't fit two hundred eighty additional trucks onto Interstate 80 without massive congestion.
The highway becomes the bottleneck before the economics flip.
And that's why the intermodal growth story is so important. Containerized rail — double-stack trains moving shipping containers from ports to inland distribution hubs — has grown about forty percent in the US since twenty fifteen. The Alameda Corridor in Los Angeles is the poster child here. Twenty miles of rail express line connecting the ports of LA and Long Beach to the inland rail yards. It moves about twenty-five percent of all US container imports, bypassing two hundred-plus street-level crossings and cutting truck traffic on I-710 by thirty percent.
Twenty miles of rail replacing how many truck trips?
And the expansion of the Panama Canal in twenty twenty-four and the opening of the Port of Lazaro Cardenas in Mexico in twenty twenty-five are creating entirely new rail corridors. Containers from Asia land on the Mexican Pacific coast, get loaded onto trains, and arrive in Chicago or Kansas City without ever touching a US port.
That's a fascinating strategic shift. Bypassing West Coast port congestion entirely.
It changes the modal calculus. A shipment of electronics from Shenzhen to Chicago via LA-Long Beach port to Chicago rail ramp takes fourteen to eighteen days door-to-door. All-truck from LA to Chicago takes three to four days but costs two and a half times more. The rail option wins for cost-sensitive, time-tolerant cargo — which is most retail goods.
We've got a landscape where rail dominates bulk commodities by default, competes on long-haul containerized freight, and loses on short-haul and time-critical shipments. What about the driver shortage? That seems like a structural advantage for rail.
The American Trucking Associations estimated eighty thousand unfilled driver positions in twenty twenty-five, projected to reach a hundred sixty thousand by twenty thirty. It's a brutal job — long hours away from home, sedentary, high-stress, and the pay hasn't kept pace with the demands. Rail doesn't have this problem to nearly the same degree, but it has its own labor issues.
Crew fatigue and union work rules.
The unions have been fighting PSR implementation tooth and nail because it often means longer trains with the same crew size, tighter schedules, and less recovery time. There have been some high-profile fatigue-related incidents. The Federal Railroad Administration has been pushing for updated hours-of-service rules. So neither mode has a clean labor story.
On the environmental side, you mentioned carbon pricing as a potential driver of modal shift.
The EU's Emissions Trading System now covers transport, which effectively puts a price on the carbon advantage rail offers. California's Advanced Clean Fleets rule is pushing toward zero-emission trucks, which are heavier and more expensive than diesel equivalents — at least for now. If carbon pricing spreads, rail's seventy-five percent emissions advantage becomes a direct cost advantage. But again, only if the rail network is electrified.
Which in North America, it almost entirely isn't.
One percent electrification. The US would need a Marshall Plan-scale investment to electrify its freight rail network. The Class I railroads have shown no appetite for it — they're private companies, electrification has a decades-long payback period, and their shareholders want quarterly returns. In Europe, electrification is happening faster because the national networks are publicly owned and the EU is subsidizing the transition.
Let's step back and talk about what actually determines the rail-versus-truck decision for someone making these calls day to day.
It boils down to three variables. Distance, cargo value density, and time sensitivity. If your shipment is over five hundred miles, low value per pound, and not time-critical, rail is almost always cheaper. If any of those conditions flip — shorter distance, higher value density, tighter delivery window — truck wins.
Give me the value density piece in more detail.
Think about what you're shipping. A ton of coal is worth maybe fifty dollars. A ton of iPhones is worth, I don't know, a million dollars. The transportation cost as a percentage of total value is negligible for the iPhones — you'll pay for the fastest, most secure mode because inventory carrying cost and obsolescence risk dominate. For the coal, transportation might be forty percent of the delivered cost, so you obsess over every cent per ton-mile.
That inventory carrying cost is the hidden variable most people don't think about. If your cargo is worth a lot, every day in transit is money.
Let's say you're shipping a container of consumer electronics worth two hundred thousand dollars. At a ten percent annual inventory carrying cost, that container is costing you about fifty-five dollars a day just to exist in transit. Rail adds two to three days versus truck, so that's a hundred ten to a hundred sixty-five dollars in additional carrying cost. Suddenly the rail freight savings don't look as compelling.
High-value goods almost always go truck, regardless of distance.
Or air, if it's really high-value and time-critical. But for overland, yes. The other piece is that high-value goods tend to move in smaller, more frequent shipments to closely match demand. Rail wants big, predictable, consistent volumes. Trucking handles variability better.
The infrastructure constraint you mentioned — that rail is often simply unavailable for specific lanes.
This is the part that frustrates shippers the most. Even where rail makes perfect economic sense, the service might not exist, or it might exist but be so unreliable that you can't build a supply chain around it. The US rail network has been shrinking for a century. Thousands of miles of branch lines have been abandoned. If your facility isn't on a Class I main line or a short line that connects to one, you don't have a rail option.
Short lines being the smaller railroads that feed the Class I networks.
About six hundred short line and regional railroads in the US, operating roughly fifty thousand miles of track — about a third of the total network. They're the capillaries. Without them, the Class I arteries don't get fed. And short lines are perpetually undercapitalized, often operating on track that hasn't seen major investment in decades.
You could be a manufacturer five miles from a short line that can only handle ten-mile-per-hour speeds and twenty-year-old locomotives that break down every other week. That's not a viable logistics option.
That scenario is more common than people think. There's a chicken processor in rural Georgia I read about that's three miles from a short line, but the track is so poorly maintained the railroad imposes a five-mile-per-hour speed restriction for the first twelve miles. The truck option is faster, more reliable, and — when you factor in the transloading and delay risk — not meaningfully more expensive. The rail option exists on paper and nowhere else.
Which creates a fragility in the system that isn't obvious from the outside.
The system works until it doesn't. The twenty twenty-two freight rail service crisis — where multiple Class I railroads essentially stopped providing reliable service to large swaths of their network — was a wake-up call. The Surface Transportation Board held emergency hearings. Shippers testified about grain rotting in silos, fertilizer not reaching farms during planting season, factory slowdowns due to parts shortages. All because the railroads had cut crews and mothballed locomotives too aggressively under PSR.
Some of that freight shifted to truck permanently.
Once you redesign your supply chain around truck, you're not coming back to rail unless there's a compelling reason. The switching costs are real. You've built relationships with trucking companies, you've sited your distribution centers near highways rather than rail spurs, you've adjusted your inventory levels to faster transit times. Modal choice has hysteresis.
There's a Herman word.
It's the right word. The system's state depends on its history, not just current conditions. Even if rail service improves, the shippers who left may not return.
Let's talk about China, because the fifty percent rail share is so far outside the Western experience.
China is a different category entirely. The state decided rail freight was a strategic priority and funded it accordingly — one point two trillion dollars since twenty ten. They've built dedicated high-speed freight corridors. They've integrated rail terminals into major industrial parks. They've used administrative measures to push freight off highways and onto rail during pollution crises.
It's not a market outcome.
It's an industrial policy outcome. And it works for China's economic structure — huge volumes of coal and steel moving long distances from interior production regions to coastal manufacturing hubs, then containers moving finished goods back to interior consumption centers. The flows are balanced in a way that American rail flows often aren't.
Empty backhauls being the profitability killer.
If your train goes from Chicago to Los Angeles full of auto parts and comes back empty, you're only making money on half the round trip. The US has persistent trade imbalances that create directional freight imbalances. The West Coast ports import far more than they export, so containers pile up and railroads have to reposition empties. That's pure cost.
How big a drag is that, actually? The empty container problem.
It's enormous. Roughly twenty to twenty-five percent of all container moves by rail in the US are empty repositioning. That's thousands of trains a year hauling nothing but air and steel boxes, just to get the containers back to where they're needed. The railroads charge for it, obviously — it's built into the rate structure — but it consumes capacity that could be earning revenue. If you could wave a wand and balance US trade flows, you'd instantly increase effective rail capacity by something like fifteen percent without laying a single mile of new track.
That's a staggering inefficiency hiding in plain sight.
It's almost never discussed in the modal-shift conversation, because you can't solve it with transportation policy. It's a trade policy problem that manifests as a logistics problem.
What's the actionable picture for someone managing logistics? What should they actually do with all this?
First, treat the rail-versus-truck decision as a function of three variables: distance, value density, and time sensitivity. If you're over five hundred miles, under roughly five dollars per pound in cargo value, and your delivery window is measured in days not hours, rail should be on the table.
If any of those doesn't hold, start with truck.
Start with truck and only consider rail if there's a compelling cost reason. Second, evaluate rail on specific corridors, not as a blanket alternative. Rail makes sense from LA to Chicago. It probably doesn't make sense from Atlanta to Miami. The service quality and transit time vary enormously by lane. Talk to shippers who actually use that corridor, not just the railroad's marketing materials.
The marketing materials will tell you every lane is great.
The marketing materials are aspirational fiction in some cases. Third, watch the regulatory and technology trends. Autonomous trucking could flip the economics within five years — if you're making a long-term logistics investment, model what happens when truck driver costs drop by thirty percent. On the flip side, carbon pricing could make rail's environmental advantage a hard cost advantage, especially in Europe and California.
Electrification is the wild card.
Rail's environmental advantage is real but contingent on electrification. A diesel train is better than a diesel truck per ton-mile, but an electric train powered by a clean grid is in a different league entirely. Europe is moving in that direction. The US is not. That divergence will shape modal choice for decades.
The future isn't rail versus truck. It's which kind of rail versus which kind of truck, in which regulatory environment, on which corridors.
The countries that invest in rail electrification and intermodal infrastructure now will have a logistics advantage in the twenty-thirties. China already has it. Europe is trying. The US seems content to let the market sort it out, which means the market will continue to choose trucks for most things and rail for the things only rail can do.
Which brings us to the big open question. Will autonomous trucks kill long-haul rail?
I don't think they kill it. But they compress it. They push rail back into the niches where it's truly irreplaceable — bulk commodities, the densest intermodal corridors, the longest hauls. The middle ground — the five hundred to thousand-mile container moves — that's where autonomous trucks could eat rail's lunch.
Unless highway congestion becomes the binding constraint.
Which it will, on certain corridors. I-80, I-40, I-5. You can't quadruple truck volumes on those highways. At some point, the congestion cost outweighs the driver savings, and rail's capacity advantage reasserts itself. The Alameda Corridor worked because the alternative was gridlock.
The future is messy and corridor-specific, which is honestly the most honest conclusion we could reach.
Logistics is always messier than the consultants' slide decks suggest. The modal split we have today reflects a century of infrastructure decisions, geography, industrial structure, and regulation. It won't change overnight. But it will change.
Now: Hilbert's daily fun fact.
Hilbert: In the nineteen seventies, a surge in commercial kelp harvesting along the Yukon River delta inadvertently collapsed the local population of ribbon seals, not because of the kelp removal itself, but because the harvest barges' engine noise disrupted the seals' underwater mating calls, which they produce by inflating a specialized air sac in their throat. The seals abandoned the delta within three seasons and never returned.
...right.
The Yukon River has ribbon seals?
This has been My Weird Prompts. Our producer is Hilbert Flumingtop. If you enjoyed this, leave us a review wherever you get your podcasts — it genuinely helps. Submit your own prompts at myweirdprompts dot com. I'm Corn.
I'm Herman Poppleberry. We'll be back next week with something else entirely.
