President Donald Trump (R) has ordered a conversion of the Navy’s carriers from electronic catapults to steam catapults for launching its aircraft. There are a couple of points about that.

Chief of Naval Operations Admiral Daryl Caudle:

If we were to go back to steam and hydraulics with the elevators, we need to really understand the total cost of that. That will be a significant design change.

Absolutely. Any design change, especially one done after production has begun, has costs, beginning with the cost of removing the hardware or the complete system that’s being changed, designing the replacement hardware or system to fit into the space—or more expensively, also altering that space to accommodate the replacement—and building and installing the replacement. In the present case, an added cost results from the fact that no one makes steam catapult systems anymore. These have to be gutsed up from scratch, including (re)building the factories that make them.

On the other hand, it’s critical to not be married to sunk costs in considering a change. Those costs are sunk for a reason: they’ve already been incurred and won’t be recouped. They’re expensive to walk away from, but it can be even more expensive to continue sinking money into a suboptimal system just because money has already been spent on it.

Better is the enemy of good enough is certainly true, but there’s an equally valid flip side of that: good enough can’t be allowed to be the end game; better really is an improvement. Existing sunk costs can’t be allowed to be in the way of better. What matters is long term performance of the two systems, the good enough vs the better. Cost is only one component of that assessment, and cost is more than just dollars. In the present case, cost includes which system gets more aircraft—manned or unmanned—into the air at the faster rate with the longest time between breakdown and the shortest time to repair.

And a bonus point, about which Acting Navy Secretary Hung Cao had this in his memo to the Navy:

To be clear, the debate on whether we do this has ended. Our task now is to provide the plan on how to implement.

Slow-walking the implementation or other forms of passive resistance can only run up the costs of the implementation, and it should get those resisters dismissed from the Navy or fired from their civilian DoD jobs.