Requiem for the Constellation

Five and a half years ago, I covered the announcement that the Italian FREMM class frigates would be adapted into the USN's new frigate, which was named the Constellation class after one of the original frigates a few months later. The idea was that by buying an existing design, it would reduce the risk of major cost overruns and help get the USN out of the shipbuilding doldrums it had been in ever since it was saddled with the LCS.

Unfortunately, none of this seems to have worked out. The program is currently 36 months behind schedule and well over budget, and the originally-planned 85% commonality has dropped to 15%, while the Navy and the shipbuilder, Fincantieri Marinette Marine (FMM), have pushed ahead with construction well before the design is even completed. Things have gotten so bad that Secretary of the Navy John Phelan that four of the six ships currently on order are being cancelled, and that the Navy will look elsewhere for a small surface combatant.

Now, this all leads to the question of what went wrong and why. Obviously, the adaptation process proved much more difficult than expected, to the point that several people affiliated with the program claimed it would have been easier to design a new ship. I actually wondered if they were basically planning to do that when it was first announced, making a new design that looked like a FREMM a la the Super Hornet, but it appears that they took the task seriously. But there were obviously going to be adaptation issues, and the design going from 85% the same to 85% different over 3-4 years implies that there was some very fundamental problem with how the program was conceived when the contract was signed, and several media reports hint that the big issue was survivability standards, most likely the design's shock resistance.

https://youtu.be/5DuJaGFkCmg At this point, the obvious thing to do is to blame USN gold-plating, and say that things would have been fine if we'd just stuck to the Italian standards. This is a more complicated issue than it seems. The USN institutionally tends to demand a higher level of survivability than many others, lessons learned in blood, but on the other hand, it's not entirely clear that the tradeoff is worth the 3-year delay. Most likely the people involved kept telling themselves it wouldn't be that long, and I'm sure the decision to stick with traditional standards was influenced by one of the main criticisms of the LCS being its poor survivability.

But leaving aside the object-level question of what they should have done and if the Italian version was good enough, there's still the fact that everyone involved, both at Naval Sea Systems Command (NAVSEA) and FMM, completely misunderstood what bringing the ship up to USN standards would entail. I don't have the visibility into the early contract period to know for sure who bears more of the blame, but suspect based on what we've seen that the problem here was mostly at FMM. If there had been major requirements changes after contract award, that would have been pretty notable (and is frequently called out as a bad thing that causes increased expenses on other programs) but there's been effectively no discussion of such changes anywhere. If FMM was unduly optimistic about its ability to meet the contract in the first place, then we would see more or less what we have seen.

USS St Louis at FMM before launch

In theory, this kind of thing should have been discouraged by the way the contract is set up. Constellation was bid using what is known as a fixed-price incentive contract. As you might expect from the name and a basic knowledge of government procurement, it's not actually for a fixed price. Instead, there are two main prices in this kind of contract, the target price and the ceiling price. The target price is what the contractor wants to hit if things go well, with diminished profit in case of cost overruns up to the ceiling price, which is the maximum the government will pay for delivery of the ship. At the moment, we're well above the ceiling price, so in theory, FMM is having to eat a bunch of losses on this. In practice, it looks like there has been a reasonable amount of money added to the contract to deal with the effects of inflation and workforce issues, but nothing much for requirements changes, which normally are negotiated separately. I tried to get a sense of how Fincantieri was doing financially on all of this, but their annual reports were unhelpful. In any case, the normal practice if the contractor screws up this badly is to throw them under the bus quite publicly, as has happened with, for instance, Boeing and the KC-46 program. Instead, the Navy has taken the bulk of the blame here, which suggests that FMM has powerful political interests on its side.

But there's other evidence that this program has been heavily beset by politics from the beginning. One decision that has attracted a lot of criticism was the choice to go ahead and start construction before the design was complete. Now, this sort of concurrency between design and build isn't necessarily a bad idea if you follow good engineering practices. There's no need to have the detailed design for the mast done when you lay the keel, and waiting around for it is just going to delay things. FFG(X) did not follow good engineering practices. Construction on FFG-62 began in August of 2022, and 14 months later, construction work was being done on modules that were still less than 75% design-complete. While bureaucracies often make stupid decisions, there are some decisions so bad that they can only be understood as the result of the bureaucracy optimizing for a different objective. In this case, that objective was undoubtedly keeping the program "on schedule" in the short term, presumably because someone didn't want them rocking the boat.

More evidence of this sort of political top cover comes from a recent GAO report, which describes the Navy as accepting incomplete or nonexistent design documents (p.18) and giving them 50% credit for completion, allowing the design to stay "on schedule" despite the fact that some of the documents were "without any design content from the frigate shipbuilder". I regularly work with similar document deliverables for the government, and if I proposed sending in something blank so that we met our deadlines, I'm not sure who would be more upset, my management or the government. Again, this is such bad practice that it can only come via serious political pressure to avoid rocking the boat. Eventually, things got so bad that the Navy's senior leadership had to acknowledge the reality of the problem, with a 12-month delay coming out in January 2024, followed by 3 more months in March and then a total of 36 months in April. At the time, it was blamed on workforce retention issues and supply chain hangups, but that's a pretty obvious lie when you're slipping a delivery from 2026 to 2029.

But what sort of political interests could drive this?¹ Well, the first and most obvious answer is that Wisconsin is a swing state, so there was apparently political pressure even back in 2019/2020 to award the FFG(X) contract to FMM, and to avoid pulling this contract in ways that might cause serious job losses. Second, the yard is in a district that was represented by Mike Gallagher, who was a major player on the House Armed Services Committee until he resigned in 2024 to go into business. Third, there has been a lot of interest in bringing in foreign shipyards to fix America's broken shipbuilding base, particularly under SecNav Del Toro, who was in charge when the worst of this was happening. If we threw Fincantieri under the bus, no matter how deserved it was, other partners might be scared off.

But at least someone in Washington has finally woken up and cancelled the program, right? Unfortunately, the political forces that have shaped the program from the start continue to exercise their malign influence. Despite headlines about the program being cancelled, SecNav Phelan has chosen to keep building the first two units, effectively to preserve jobs at FMM, while saying that the Navy will examine other options for Small Surface Combatants. This makes absolutely no sense. Yes, the design phase of this program has been unusually bad even for defense projects, but it's mostly over, and we are going to have to complete it anyway if we are to build so much as one ship. And after we have built that one ship, we then can build others without having to do design again. Starting over means that we have to do design again, and even if everything goes absolutely as perfectly as it is possible to imagine, that's still several years of delay relative to just building follow-on Constellations. Also, we will be stuck with a two-ship force that requires a dedicated engineering support staff. As we've seen with the discourse around the F-35, many people are suggesting actions that might have made sense if executed a number of years ago, but will just drive up cost and delay the delivery of systems to the front lines today.

Bertholf, the lead National Security Cutter

Phelan has since announced that the replacement will be based on the HII National Security Cutter, which was probably the runner-up design in 2020. Besides the political aspects of the decision (FMM was running out of work with LCS going away, while Ingalls, which builds Burkes and most of the 'phibs, still had plenty) the major downside is that it is considerably smaller than the FREMM, thus limiting what systems we can put aboard. And while he's talking about getting something in the water by 2028, that seems like obvious nonsense. Ingalls generally has a reputation for competence, and they are unlikely to have messed up the way FMM did, but 3 years from order to launch is what you'd expect from a proven design in series construction, not the first of a class from a standing start. If they do actually try to make it happen, then the result will undoubtedly cost more and be later than if they'd just done it the right way. This mess will undoubtedly haunt the USN for decades, and is proof that we shouldn't appoint a Secretary of the Navy who doesn't even own a rowboat.²