Union Pacific and Norfolk Southern file final STB response with unprecedented shipper protections

The Analysis

You land a container in Los Angeles and the real cost starts after it comes off the ship. The railroad leg to your customer's door is where this Union Pacific and Norfolk Southern merger hits home, and the headline number of 3.5 billion dollars in shipper savings is the least useful part of the story. I have stood on both the UP ramps in the West and the NS yards in the East, and when two railroads that never shared a single through-line promise to become one, the question is never the savings. The question is which gateway jams first. Let me put the map on the table.

The two carriers filed their final answer to the Surface Transportation Board on 27 July 2026. On paper it looks like a love letter to shippers. They attach four voluntary promises. Committed Gateway Pricing, which today covers about 134,000 shipments a year, gets roughly doubled to 258,000. For the first time the program reaches agricultural and bulk unit-train customers. Shippers who today have only two Class I railroads to pick from, the so-called two-to-one and three-to-two lanes, keep a fallback option. A Targeted Access Program opens the door to temporary alternative service while the two networks stitch together. And if the promised benefits lag, there is a new lane to dispute rates. The applicants then paint the big picture: single-line service on more than 10,000 lanes, 88,000 new county-to-county lanes, 2.1 million truckloads a year pushed from road to rail, and an estimated 3.5 billion dollars in annual shipper savings, with the deal closing around the middle of 2027.

Put that on a map and here is what changes. Union Pacific owns the West. Its tracks run from the Pacific to the Gulf and up through the middle of the country. Norfolk Southern owns the Southeast and the industrial Northeast, the old coal and auto heartland. Until now, a box that lands in Long Beach and needs to reach a distribution center outside Atlanta crosses a seam. Someone has to hand it from one railroad to the other, and that handoff is a node where things slow down, where a chassis goes missing, where a ramp gets congested.

The merger promises to erase that seam on more than 10,000 lanes. One bill of lading, one railroad, from the port gate to the customer dock. I have watched those handoffs eat two or three days and a repositioning fee that nobody puts in the quote until the box is already sitting there. So the route logic is real. The corridor gets simpler, and simpler corridors are worth something to a freight forwarder who has to explain to a client why a box is late.

Committed Gateway Pricing is worth a word on its own, because the name hides the mechanism. In practice it is a commitment that a shipper gets access to the railroad's gateway ramps at a published rate without being pushed to the back of the queue when capacity is tight. For an importer whose boxes arrive in waves, that queue-jump is often worth more than the rate itself, because a box that waits at a ramp waits in inventory dollars, not railroad pennies. Doubling the eligibility from 134,000 to 258,000 shipments a year means the carriers are opening that ramp access to a larger slice of shippers, and extending it to agricultural and bulk unit-train customers for the first time is a deliberate signal to the shipper groups that lobby the STB. None of this is free for the railroad. Ramp capacity is a physical thing. Promising more shippers a slot at the gateway means either the ramp gets bigger or the slot gets thinner, and the merged company is the one that decides which.

The STB is the toll booth on this road, and the supplemental information request is the booth asking to see your papers again. When a regulator sends that request, it is rarely a formality. It is a signal that the board wants more before it grants the largest rail combination in a generation. The carriers answered on 27 July with the four commitments precisely because the request gave them a deadline and a stage. Read it that way and the voluntary promises are not generosity, they are the price of getting waved through. A forwarder who has stood at enough regulatory crossings knows the pattern: the bigger the deal, the sweeter the pre-approval concession, and the softer the enforcement once the gate closes behind you.

The 88,000 new county-to-county lanes sound like abstraction until you plan a route. Today, reaching a county that sits off both networks means a drayage leg, a handoff, and a final short haul that three different parties touch. The promise of 88,000 new direct county pairs means some of those awkward routes become single-line, which removes a handoff and a billing party. For a forwarder building a routing guide, that is real option value, because every handoff you delete is a place your box can disappear for three days. But again the catch: those lanes only help if they are yours, and the carriers decide which county pairs go single-line first. The aggregate number is a map of intention, not a guarantee of service on your specific lane.

Step back and this is the first time a single railroad could plausibly offer single-line service from a Pacific port to a Southeastern factory without a handoff. The West has long been a duopoly of BNSF and Union Pacific. The East has been CSX and Norfolk Southern. The middle is where they all meet and swap. A UP-NS combination does not give you coast-to-coast by itself, but it gives you the western half plus the eastern heartland in one hand, which is most of the lane a China importer actually cares about. That is why this merger, more than past ones, changes the route calculus for trans-Pacific freight. The corridor you plan today with two billing parties becomes a corridor you plan with one, and the question of who to call when it breaks changes from two phone numbers to one, which is not always an upgrade.

Why now, and why these concessions? The STB asked for more information and the carriers answered with sweeteners because they need the board to say yes. This is not charity. A merger this size, two of the six Class I railroads in the country becoming one, walks into a regulator that has spent years hearing from shippers who hated the last big rail consolidations. The memory of past service meltdowns after mergers is still warm, and the board knows that a bigger railroad is a harder railroad to discipline later.

So the carriers pre-empt the objection. They hand the board a package that looks like protection, timed to the supplemental information request, timed to keep the closing target of mid-2027 alive. The upstream push is the regulator's skepticism and the shipper coalitions that show up at every STB hearing with stories about being held hostage by a single line. The carriers are buying approval with promises, and the timing tells you the promises are priced into the deal, not bolted on after.

Now who actually feels this. The clear winners are the big intermodal players, the ones moving containers in volume from the coast to the interior. Single-line service cuts a handoff, and for a China-based importer landing boxes in California and feeding Midwestern or Southeastern DCs, that is the leg that mattered. Agricultural and bulk unit-train customers are new to the protected pricing, and that is a real win for a commodity shipper who never had a seat at this table. The two-to-one and three-to-two shippers, the ones whose only two Class I options were exactly these two, keep a fallback. That sounds like nothing until you remember that on those lanes, losing one of your two railroads means you are captive to the survivor.

The losers are the shippers who sit on a lane where UP and NS already overlap and compete today. After the merger that competition is gone, and the voluntary promise of a fallback only works if the fallback railroad actually has capacity and appetite on that lane. Smaller regional shippers who never touch the intermodal ramp barely move. So the savings are not evenly spread. They tilt toward the big boxes and the bulk trains, and they are paid for, partly, by the lanes that lose a competitor. A forwarder routing for a client has to read the specific lane, not the headline.

When does any of it reach your cost? Not tomorrow. The deal closes around mid-2027 at the earliest, and the savings only show up after the networks actually merge and the single-line lanes light up. That is twelve to eighteen months of waiting for the benefit. The risk, though, shows up sooner. The integration window, the months when two giant operating systems are being welded together, is exactly when service tends to stutter. The carriers know this, which is why the Targeted Access Program is written to switch on during integration. But a temporary alternative service is only as good as the railroad you switch to, and every other Class I will be fielding the same panic calls from the same stranded shippers.

For an importer the exposure is a timing one. The ocean leg and the drayage to the ramp are unchanged. The domestic intermodal leg is where the gain lands and where the disruption also lands, on the same track. If you have ever lived through a railroad's bad integration quarter, you know the box does not care about the press release.

Here is the part the headlines skip. Everyone repeats the 3.5 billion dollars and calls it unprecedented protection. But read the fine print on the word voluntary. These four commitments are promises the carriers made to the board, not conditions the STB imposed with a hammer. A voluntary commitment is only as solid as the carrier's reputation and the board's willingness to enforce it after approval. Once the merger is granted, the leverage shifts to the merged company.

The Targeted Access Program, the thing meant to save you during the messy integration, only triggers during integration, the precise window when capacity is tightest and every alternative railroad is full. The 3.5 billion is the carriers' own estimate, not an independently audited number, and buried inside it is an assumption that 2.1 million truckloads willingly move to rail. That is a capacity-absorption bet, not a saving. If the trucks do not come, the saving does not exist.

And for an Asia-to-US importer, the ocean freight, the part of the journey you actually shop around, does not move a cent. The only place this touches you is the US domestic leg, and there the real risk is not that you miss a discount. It is that the bigger, less accountable railroad stumbles while it is learning to be one company, and your box sits on a ramp you no longer have a second option for.

Flip the assumption and the whole rosy picture inverts. If the STB blocks the deal or hangs harder conditions on it, the savings never arrive and the two railroads stay separate and competitive, which for many lanes is the better outcome. If the merger closes but the integration goes the way past rail mergers have gone, with terminals choking and trains running late, then the single-line convenience is eaten by unreliability and the fallback option is worthless because every other railroad is also at capacity. The condition that flips the conclusion is simple: the promise only pays if the merged railroad runs better than two railroads did separately. That is not guaranteed. A monopoly that promises nicely is still a monopoly, and the morning after approval is when you find out whether the promise was a condition or a press release. The board can say it extracted protections, but a promise you cannot enforce is a promise someone else wrote.

Let me put a number on it the way a forwarder would, with the assumptions out in the open. Suppose you are a mid-size importer landing forty containers a month at a US West Coast port and inlanding them by rail to a Southeastern DC. Suppose today that move crosses the UP to NS seam and costs you, call it, a handoff delay of two to three days plus a repositioning charge on maybe one box in five. Suppose the merger removes the handoff on your lane. The saving is real but it is measured in days of inventory and the occasional repositioning fee, not in the headline billions.

If your domestic intermodal spend runs four thousand dollars a box, shaving the handoff might trim a low single-digit percentage off that leg, and only after mid-2027, and only if your lane is one of the 10,000 that goes single-line. The billions are an aggregate. Your invoice is a detail. Do not let the aggregate hypnotize you, because the aggregate was written by the people asking you to applaud.

The interchange nodes are where I would keep my eyes. Chicago is the busiest rail junction in North America and both roads meet there; Memphis is another seam where east meets west. When two networks merge, the first place the smoothing shows up, or fails to, is these handoff points. A single-line promise does not help you if the junction itself is the bottleneck, because the merger does not add a single new track at Chicago on day one. It adds a new owner. So the route question is not only whether the seam disappears on paper. It is whether the merged railroad invests in the junction faster than the old two did separately, or slower, because now there is no competitor shaming it into action. I have seen junctions where the only thing that kept the trains moving was the fact that two railroads were embarrassed to be beaten by each other. Take that away and the node can quietly rot.

There is a second flip worth naming. Suppose the merger closes and the integration is smooth, the best case. Even then, the voluntary commitments sit in a drawer unless a shipper complains and the board acts. A promise you have to litigate to use is a weaker promise than a condition written into the order. So the real test of these protections is not whether they were announced, but whether a mid-size importer with forty boxes a month has the muscle to invoke them when the merged railroad drags its feet. The big shippers will. The small ones will read about it in a trade press release. That asymmetry is the part the 3.5 billion headline papers over.

Run a second case. Suppose instead of forty boxes a month you move four hundred, and your lanes already sit inside the UP and NS footprints rather than crossing them. Then the handoff removal matters less, because you were mostly single-carrier already, but the loss of competition on overlapping lanes matters more, because you may have been using the two as a pricing check on each other. For that shipper the merger is a net negative dressed as a neutral, and the 3.5 billion has nothing to say to them. The point of the arithmetic is not to cheer or boo. It is to locate yourself on the map before the gate closes, because the same merger is a different deal depending on which lane you stand on.

The physical nodes where UP and NS already touch each other are the places to watch once the deal is live, and they belong in the same breath as the interchange story. Chicago is the obvious one, the busiest rail junction on the continent, where every western railroad and every eastern railroad already crosses. Memphis is another, a gateway where north-south and east-west freight meets. St. Louis, New Orleans, and the Ohio corridor are smaller seams but still seams. Today, freight crossing these points changes hands. After the merger, the handoff becomes internal, which should be smoother, but the junction itself does not get a single new track on day one. So the node that was a bottleneck under two owners becomes a bottleneck under one owner, and the only thing that changed is who is embarrassed when it jams. I have always said, do not cheer a merger for removing a seam until you have watched the junction in week twelve.

So what do you actually do while the two railroads court the regulator and the clock runs to 2027. You do not renegotiate your ocean contract on the back of this. You watch the seam, you keep your options open on the domestic leg, and you time your commitments so you are not the shipper who discovers the fallback was a press release. The moves below are the ones I would make if those forty boxes a month were mine, and the one rule underneath all of them is that a promise from a railroad you cannot leave is not protection, it is a sentence with a smiley face.

↑ Back to the key points

— By Ömer Kaya