Thank you, admiral, for the kind introduction. It’s an honor to be here. It’s great to be back at APL, and to have a chance to share my views with you. So again, thank you for inviting me today. As usual, I should note that these views are my own and not necessarily those of my employer.
By now, I think most of you know that I don’t sugar-coat my assessments. In the debate on defense issues, there’s already plenty of cheer leading, so you don’t need me to add to it. And if all you ever hear is cheer leading, you can drift into self-deception and fail to take actions to better prepare for the future. So I try instead to describe things as I see them, on the theory that you’ 11 find it helpful, if not always comforting.
Seeing things as they are isn’t doom and gloom, and it isn’t pessimism. If you hear someone describing it that way, you should take it as a possible sign of cheer leading intoxication, and organize an intervention to get that person into rehab, before it’s too late.
Some things worth noting
Now, having said that, I want to start today by noting some things that deserve positive recognition. There are a lot of things I could mention, so these are just a few examples.
First, I think the submarine community has put together what looks like a promising general approach for reaching the CNO’s Virginia-class cost target. The target figure appears somewhat arbitrary, and I don’t understand why attack submarines appear to be the only type of ship that has to meet its target as a condition for all ships of that type to be kept in the shipbuilding plan. And the Senate authorizes have directed that the plan be better defined in terms of specific goals and benchmarks. But the general approach looks promising.
It’s worth noting that the target is a goal that the Navy has set for itself, and that Congress, while taking it into account, can choose to set it aside if it likes. It’s important to note this, because there’s been some discussion on the Hill about how and when the target can be met, and when these discussions get going, it can be easy to forget that this is an internal Navy goal, and not one that has to control congressional action.
Within the Navy’s cost-reduction plan is an idea to reduce installation costs and shorten construction time by not installing certain elements of the combat system that are simply going to be removed and replaced during the PSA. There might be some challenges in implementing this idea, but assuming they can be overcome, this strikes me as a real innovation in the shipbuilding process that might be applicable to other shipbuilding programs as well.
Another item worth mentioning is the provision in the Virginia- class contract, and also the CVN-21 contract, for the Navy to front the investment cost of shipyard improvements that can lead to reductions in recurring production costs.
Beyond this, the Navy has also mentioned the possibility of making some adjustments to the teaming agreement that could reduce Virginia-class construction costs by $25 to $80 million per boat. I understand why the two yards might be reluctant to reopen the agreement, because one of them would lose some work, and because changing the agreement once might be viewed as the first step on a slippery slope to further changes. But in light of the potential savings, it would be helpful to hear from the yards why this idea, in their view, is bad or wouldn’t work.
Finally, as many of you know, the submarine community in recent years has shifted to an open-architecture strategy that permits frequent and affordable combat system upgrades. This looks like a possible standard against which to judge open-architecture efforts elsewhere in the fleet.
Full plate of challenges
Now, when Admiral Emery invited me to speak, he told me he looked forward to hearing my challenges. I thought that was admirable, because the submarine community already has a lot of challenges on its plate, without me adding any more.
One of those will be getting to the CNO’s cost target. A second will be maintaining, with 48 boats, the same level of forward deployments that has been maintained in the past with more than 50 boats. A third will be making sure that existing boats remain in service for 33 years, even though the Force is being used intensively. And a fourth and continuing challenge will be managing a new- construction supplier base consisting largely of single sources, some of whose business situations might be fragile.
Beyond those four challenges, there are at least three others already on the Navy’s plate that I want to address in the remainder of my talk. As I go through them, and particularly as I go through the second and third, I’ll throw a couple of additional challenges on the plate.
Submarine Design and Engineering Base
The first of the three challenges is finding a way to maintain the submarine design and engineering base. This challenge was foreseen, and is now upon us, and yet there is no firm plan in place to address it.
Addressing this challenge was one of the reasons behind last year’s proposal in the House for starting design work on a new SSN to succeed the Virginia-class design. That proposal was being opposed by the Navy even as I was mentioning it to you last year, and it didn’t survive conference. In terms of up-front cost vs. downstream break-even point, I can understand why someone might decide to not support this option. But it would have preserved the design and engineering base for several years, and now something else needs to be done instead.
There are some other options out there, such as expanding the scope of the Virginia-class redesign effort, or designing a new ASDS. These options would help, but they likely wouldn’t be sufficient, due to limits on the volume of work they would provide and the number of skill areas that they would engage.
Another option would be to design a diesel boat for Taiwan. This would offer a greater volume of work, and it would engage a large number of skill areas. But there currently is uncertainty over whether and when this project will occur, making it hard at this point to confidently incorporate it into a plan for preserving the design and engineering base. In addition, this option would not preserve certain skills, such as those related to nuclear propulsion plants, and it could raise concerns regarding the potential for unintended technology transfer.
The most sufficient option for preserving the design and engineering base would be to design a new nuclear-powered submarine, and since designing a new attack boat has been rejected, that leaves the option of bringing forward the start of design work on the next SSBN. At this point, it appears to me that this option will likely font a part, and perhaps a large part, of the strategy going forward. The Senate version of the defense authorization bill recommends $10 million to start this project.
The expanded Virginia-class redesign, the news ASDS, or the Taiwan diesel boat, if they happen, could be added on top of the accelerated SSBN to make the solution more robust. And upon completing the SSBN design, the Navy could then turn to designing a new SSN. An approach along these lines could preserve the design and engineering base for a number of years.
SSN Procurement Rate and Projected SSN Shortfall
I’ll get back to the next SSBN later, but I want to tum now to the second issue, which is the SSN procurement rate and the projected shortfall in the SSN Force. I’m going to spend most of my time today on this.
This is an issue that has been building in Navy force planning for a long time. I first testified on it I l years ago, in ’95. At the time, I said the Navy was starting on a trajectory that could reduce the SSN Force to 4 I boats by the mid-2020s. Eleven years later, that projection remains pretty close to the mark.
Over the last 11 years, I’ve testified and reported on this issue on many occasions. And finally, the issue has attracted some attention. But now that the Navy has finally acknowledged the projected SSN shortfall, it’s also saying that, because of budget constraints, there’s not much that can be done about it, at least not without seriously disrupting other programs.
It reminds me of something that somebody once said years ago, in a spirit of dark humor, about the process for developing a new weapon. He said the process has two basic stages, the first being “it’s too early to tell,” and the second being “it’s too late to stop it.”
When the Force bottoms out at 40, it’ll be missing one boat out of every six that it’s supposed to have. That’s a pretty deep shortfall to manage. And the bottom will happen just as the SSGNs are scheduled to leave service, so the SSGNs won’t be around to compensate at that point.
Now it’s true that the Force will be substantially below 48 for only a certain number of years. But potential adversaries can know in advance when that will occur, and make plans to take advantage of it.
The Navy says the requirements in the 313-ship plan are for 2020, and they can change after that. If the SSN requirement goes down after 2020, it could reduce or eliminate the projected shortfall. But the requirement could also go up after 2020, which would make the shortfall worse. The Navy can’t know at this point which way the requirement might change after 2020, so the argument about changing requirements doesn’t get the Navy off the hook.
The House version of the defense authorization bill addresses the projected shortfall by recommending $400 million in FY07 funding to support the acceleration of 2-per-year Virginia-class production to FY09. That would produce a force that bottoms out at 43 boats rather than 40, and it would reduce the total shortfall period from 14 years to about 8 years.
This funding, however, is competing against two other major options for FY07 shipbuilding plus-ups. One of those is to fund two additional LCSs, and the other is to accelerate the ninth LPD from FY08 back into FY07. Both of these items are on the Navy’s unfunded requirements list, while the Virginia-class acceleration is not.
Although the Virginia-class acceleration requires less funding in FY07 than the other two options, it requires a lot more funding over the FYDP. The 2 additional LCSs are about $500 million, while the acceleration of the single LPD actually reduces funding requirements over the FYDP. The $400 million for Virginia-class acceleration, in contrast, is only the first increment of an additional $7.4 billion that would be required over the FYDP for the three extra boats.
The Senate version of the authorization bill funds the accelerated LPD rather than the Virginia-class acceleration. Some Members have been quoted in the press stating that, in their view, retaining the $400 million for Virginia-class acceleration in the final version of the bill will be an uphill battle.
30-Year Shipbuilding Plan
The issue of SSN procurement and force size encompasses two challenges. The first is to procure all the SSNs that are in the Navy’s 30-year shipbuilding plan. That’s going to be a challenge, because the plan as a whole may simply not be executable.
The Navy says that for the plan to be executable, 4 things need to happen. First, O&M spending needs to remain flat. Second, Mil Per spending needs to remain flat. Third, R&D spending needs to go down and stay down.
All three of these things, the Navy says, are necessary for the Navy to increase the shipbuilding budget to the higher level that the Navy is planning.
And the fourth thing that needs to happen is that the ships have to come in at their estimated costs.
Now, each of these four things poses a risk. Past DOD efforts to control O&M spending have not been fully successful. The Navy doesn’t have complete control of MilPer spending, because that can be affected, for example, by decisions that Congress makes on pay and benefits. And while it may be feasible for R&D spending to go down over the next few years because a number of systems are transitioning from development to procurement, it may be difficult to keep R&D spending at that reduced level over time, because the Navy at some point will likely want to start development of other new systems. And as many of you probably know, CBO believes that the ships in the plan will cost substantially more to build than the Navy estimates.
Now, some of the 4 required things might happen. But I don’ t know of anyone outside the Navy who has studied the situation who thinks that all 4 of them are going to happen.
If one or more of them don’ t happen, then the 30-year plan will come under pressure, and perhaps fall apart completely. This won’t necessarily happen in the near term, because the more expensive part of the plan doesn’t start until a few years from now, when the Navy starts trying to get 1.5 cruisers and destroyers and 2 submarines per year. But if the plan begins to come under pressure at that time, the goal of procuring 2 submarines per year will likely be reassessed.
Adding 8 more SSNs
Assuming that all the SSNs in the 30-year plan can somehow be procured, the second challenge is to reduce or eliminate the projected SSN shortfall by adding up to 8 more SSNs to the plan between now and FY22.
In attempting this, one of the most significant barriers you’ll face will be the projected shortfall in cruisers and destroyers. It may come as a surprise to you, but the SSN shortfall isn’t the only projected shortfall in the 313-ship plan- and it’s not even the biggest, because there’s an even bigger projected shortfall in cruisers and destroyers.
The cruiser-destroyer force-level goal is 88 ships. That implies a 35-year steady-state rate of about 2.5 ships per year. But the 30- year shipbuilding plan procures an average of only about 1.75 ships per year.
As a result, the cruiser-destroyer force, which has a lot of ships today, will eventually fall to a low of about 62 ships. That’s 26 ships below the goal. And in the long run, the cruiser-destroyer force never gets back to 88- it only gets back to 70, and then plateaus.
In more than 20 years of tracking Navy force-structure and procurement planning, this may be the biggest disconnect I have seen. This part of the Navy’s plan is just completely broken- and the Navy right now has no announced plan for fixing it. Indeed, the Navy barely even acknowledges the problem. There was a mention of it in the draft version of the report on the 30-year shipbuilding plan, but it was deleted from the final version. In long-term Navy force planning, this problem is the unacknowledged elephant in the room.
If you haven’t heard about this cruiser-destroyer shortfall before, that’s understandable, because the Navy isn’t going out of its way to alert people to it. That’s similar to a situation in ’95, when I first testified on the SSN shortfall. The Navy back then wasn’t talking much about that projected shortfall either, so it came as news to a lot of people.
The Navy might be operating on the theory that the cruiser destroyer shortfall is so far in the future that there’s plenty of time for someone else to do something about it. That’s a risky game to play, because past experience shows that once you get onto a certain trajectory, it can be difficult to change it, and that the longer you wait to do something about the situation, the harder it becomes to do anything about it. The cruiser-destroyer shortfall is no further in the future today than the SSN shortfall was when I first began warning about it in ’95. And look where we are now with the SSNs.
Now, why am I spending so much time talking to you about a projected shortfall in cruisers and destroyers? Why is this a concern for you? Well, it’s a concern for you because, as weird as it may sound, this shortfall, by certain measures, is now a more pressing long-term force-structure problem than the SSN shortfall:
- The SSN shortfall will peak at 8 ships, but the cruiser-destroyer shortfall will peak at 26 ships.
- The SSN shortfall will peak at about 17% of the requirement, but the cruiser-destroyer shortfall will peak at about 30%.
- And while the SSN force will eventually get back up to its force-level goal, the cruiser-destroyer force never will, and it will remain about 20% below the required number.
As I mentioned earlier, eliminating the SSN shortfall will require adding 8 SSNs to the shipbuilding plan between now [FY07] and FY22, or an average of one-half additional submarines per year. Eliminating the cruiser-destroyer shortfall, however, will require adding 26 ships between now and FY39, or about 0.8 additional cruisers and destroyers per year.
Consequently, if opportunities do arise to add ships to the shipbuilding plan, supporters of cruisers and destroyers are going to have at least as strong a force-structure argument as supporters of submarines, if not a stronger one. In other words, submarines might not be the first in line for extra ships-cruisers and destroyers might be.
Unless the Navy closes the projected shortfall in cruisers and destroyers, this logic will become more and more prominent over time, which will make it more and more difficult to add submarines back into the plan.
To close the projected cruiser-destroyer shortfall, the Navy can do one or more of the following. First, it can reduce the requirement to something less than 88. But it’s not clear what would permit the requirement to be reduced.
Second, it could try to extend cruiser-destroyer service Jives beyond 35 years. But it’s not clear whether that would be feasible or cost effective.
Third, it could add cruisers and destroyers to the shipbuilding plan. But that would put more pressure on other parts of the plan, including submarines.
And fourth, it could take actions to reduce the procurement cost of its planned cruisers and destroyers.
If you need to rely primarily on this last option, then a possible goal would be to reduce average cruiser-destroyer unit procurement costs by about 31 %. This would permit the funding now planned for cruiser-destroyer procurement through FY39 to procure an additional 26 ships.
It’s in this connection that I have suggested that the Navy consider the option of transitioning over the next several years from the DD(X) design to a smaller cruiser-destroyer design of about 11,000 tons.
Such a ship would be about 25% smaller than the current DD(X), but it would still be about the same size as the nuclear cruisers of the 1970s, and about 1,800 tons larger than the DDG-51.
As you might imagine, this suggestion has not put me at the top of the DD(X) program office Christmas card list.
The Navy has argued that an 11,000-ton ship can’t meet all the requirements that are to be met by the DD(X). That’s true, but a cruiser-destroyer force that falls to 62 ships, and then grows back to no more than 70, would fall well short of the Navy’s requirement for a force of 88. You can’t talk about one side of this situation without talking about the other.
Now, I can only suggest a general strategy to the surface community. I don’t have the technical resources to flesh out a smaller cruiser-destroyer design in detail, or to show how such a ship, though less capable in some respects, could still fill the surface community’s most urgent capability gaps.
But you have those resources. And you’ve got some recent experience doing something like this. When you shifted from the Sea wolf design to the smaller Virginia-class design, you reduced certain areas of performance while holding others constant. And in the end, you came up with a design that is less expensive to build on an apples-to-apples basis, but still capable enough to meet your future needs.
So if you want a challenge from me, here it is: Help your surface colleagues. Help them close their projected shortfall by assisting them in substantially reducing the average unit procurement cost of planned cruisers and destroyers. And by substantially, I mean as close to 31 % as possible.
When you return to your place of work, make it one of your first to-do items to get that process in motion. Because unless the surface community’s force-structure problem is fixed, it’s going to be more difficult to think about fixing yours – or even to think about getting 2 submarines per year, should the shipbuilding plan come under pressure.
Now, when you go to see your surface colleague and you tell him, “Hi, I’m so and so from the submarine side of the house, and I’m here to help,” he might look at you a bit funny. He might ask what it is you want to help him with.
And when you tell him that you’re going to help him close the projected cruiser-destroyer shortfall, don’t be surprised if he isn’t even aware of the problem, because the focus in the surface community has been on getting the DD(X) into serial production, and a lot of surface people consequently might not have had a chance to look beyond this near-term objective. And when you tell him that you’re going to help him close the shortfall by assisting him in substantially reducing the average procurement cost of cruisers and destroyers, don’t be surprised if he gets defensive and takes umbrage. Don’t let that bother you – that’s just a side effect of DD(X) cheer leading intoxication. You ‘II need to ease him out of it gently.
But whatever reaction you get, you’ 11 need to persevere. Because remember, this isn’t just a problem for them, it’s a problem for you.
Procuring Replacement SSBNs
I want to finish by turning to my final item, about the next SSBN, and here I want to make two comments.
First, the 30-year plan calls for procuring these ships at a rate of 1 per year. I’m not sure that will be affordable. My sense is that you should look at the option of getting 2 every 3 years, or better yet, 1 every other year. That could mean getting the first ones earlier than currently planned, and the last ones later than currently planned, which in tum implies extending the lives of the final Ohio-class SSBNs, if possible, beyond the currently planned figure. So there’s another challenge for you.
Second, don’t assume at the outset that the new SSBN will use a D-5 sized missile. Instead, do some analysis to understand the implications of missile size for total program cost. Take a D-5 sized missile, a smaller missile perhaps about the size of the C-3, and maybe some other size that might reflect the new knee of the curve in missile technology. Then examine each option against the mission set to see how missile size affects the total sum of missile development and procurement costs, platform development, procurement, and O&S costs, and shore infrastructure costs.
Now, it may tum out that a D-5 sized missile results in the lowest total program cost, or at least a cost that isn’t substantially higher than the other options. But at least you’ll have an analytical basis for your decision, and some confidence that you didn’t overlook a less expensive approach.
Understanding this issue will be important not only for justifying your design to OSD and Congress, but also because of the link with the UK. The UK needs to replace its own SSBNs, and because their ships will age out sooner than ours, they need to make key decisions before we do. Since the UK will likely build fewer platforms, the up- front cost of developing a new missile of a different size may loom larger in their total-cost calculation than it does in ours, which might incline them more strongly to sticking with a D-5 sized missile.
If so, and if it also turns out that a D-5 sized missile would mean a substantially higher total cost for us, then you’ll need to have a dialogue with the UK on the issue. The UK’s views deserve respect, but their decision on preferred missile size should not predetermine the U.S. decision, if that would mean a substantially higher cost for the U.S. program. If there needs to be a dialogue with the UK to resolve differences on preferred missile size, then you’ll need that study to show the UK why the best solution for them might not be the best solution for us, or for the two countries jointly.
I said earlier that the submarine community already has a lot of challenges on its plate, without me adding any more. But I’ve added a couple anyway, because I thought they were important for you to consider.
As always, I hope you found my comments helpful, at least in helping to move your own thinking forward. Thank you