The Catapult That Was Already Half Built

A presidential memorandum signed in the middle of August gives the Secretary of War sixty days to deliver a plan for replacing the Electromagnetic Aircraft Launch System and the Advanced Weapons Elevators on the carrier CVN-81 with steam and hydraulic equipment. That plan is due in October. Nothing has been cancelled yet, no contract has been modified, and the three carriers already delivered or under construction keep the electromagnetic system.

Read the document as a procurement instrument rather than an engineering one and the number that matters is not the launch speed. General Atomics, which builds both the launch system and the arresting gear, has said its hardware for that ship is close to half complete. The American production line for steam catapults went cold after the last Nimitz class carrier was commissioned in 2009.

From the Battlefield to the Balance Sheet

What a reversal of this kind actually spends is qualification. In defense manufacturing the expensive part is almost never the metal. It is the years of testing, documentation, tooling and trained labor that turn a design into something a shipyard is allowed to install, and that investment is specific to the design. Change the design late and the qualification does not transfer. It is written off.

That is the real cost sitting inside this memorandum, and it runs in both directions. Half-built electromagnetic hardware and the supplier base behind it become surplus to that hull. At the same time, steam catapults are not sitting on a shelf. The vendors, the castings, the piping specialists and the people who knew how to build and certify the system dispersed more than fifteen years ago, and standing that capability back up is itself a first-of-a-kind program with its own qualification bill. The ship class was also designed around a power plant that produces less steam than the older carriers did, so the change is a rework of the ship rather than a swap of one box for another.

For anyone holding defense suppliers, the lesson is that requirement stability, not technology risk, is the variable that decides returns in this sector. A supplier can execute perfectly and still lose the program to a decision made above it.

The Dual-Use Reality Check

This is not a defense pathology. Any capital-intensive industry that builds to a customer specification runs the same exposure, and the reshoring buildout is full of it. A supplier tooling up for a battery chemistry, a reactor design or a chip package is making a bet on a specification staying put for the decade it takes to earn the money back. Automotive suppliers learned this expensively when electrification timelines moved, and the machine shops now adding capacity against aerospace backlogs are making the same bet on a different customer.

The transferable part is the habit of asking what a supplier's capacity is actually worth if the program changes. Equipment that can be reconfigured, a workforce with certifications that travel, and a customer list with more than one name are worth a premium that does not show up in an order book. A backlog concentrated in one design for one buyer looks identical to a diversified one on paper and behaves nothing like it when a memorandum lands.

The Capital Signal

The capital signal is that the plan due in October is the document to read, not the memorandum. Sixty-day planning directives frequently come back recommending a narrower version of what was asked, and the shape of that recommendation, along with whether the Navy actually modifies the shipbuilding contract, is what converts this from a policy statement into money moving. Until then no credible cost estimate exists, and the figures circulating are characterizations rather than official work.

The structural read is that carrier programs have become a live demonstration of a broader condition. Long-cycle industrial assets are being ordered by a buyer whose requirements now move faster than the assets can be built, and the risk of that mismatch sits with suppliers rather than with the buyer. That argues for paying attention to contract structure, to how much of a supplier's revenue is termination-protected, and to whether a company has been paid for work it has already performed. Those are unglamorous questions, and in this environment they matter more than the specification argument that gets the headlines.

Signal: Late design reversals destroy qualification, not hardware, and qualification is where the money in defense manufacturing actually sits.