The Air Force, Army, and Navy have identified fourteen weapons programs where they will employ two management techniques to prevent massive cost overruns, according to an April 22 Department of Defense (DoD) memo obtained by Inside Defense.com.
Both the Air Force and the Navy picked the F-35 Joint Strike Fighter (JSF) program as one of their five example programs. The Air Force has plans to use the F-35A conventional take-off model; the Navy has plans to use the F-35C carrier version; and the Marine Corps (which is part of the Department of the Navy) is planning on using both the F-35C and the F-35B short take-off and vertical landing model (STOVL). We’ve reproduced the DoD’s table of the fourteen programs here:
|Joint Strike Fighter (F-35)
|Global Hawk Blocks 30 & 30 (GH BLK 30 & 30)
|Space Based Infrared System (SBIRS)
|Evolved Expendable Launch Vehicle (EELV)
|Advanced Extremely High Frequency (AEHF) Satellite System
|Joint Air Ground Missile (JAGM)
|Black Hawk (UH-60M)
|Ground Combat Vehicle (GCV)
|Paladin Product Improvement (PIM)
|Joint Strike Fighter (F-35)
|Presidential Helo (VXX)
|Littoral Combat Ship (LCS)
|Ohio Replacement Program
The memo by Ashton Carter, the Pentagon’s chief procurement official, is part of the DoD’s attempt to improve management of its weapons portfolio. For decades, with relatively few exceptions, each new generation of weapons is increasingly more expensive per unit (even after adjusting for inflation) than the last, takes longer to develop, and leads to a much smaller force structure that is more expensive to maintain.
Carter's initiative is not necessarily an attempt to save taxpayer dollars, although obviously that can be a side effect. For instance, you can spend an equal amount of taxpayer dollars on a canceled/poorly-managed program that produces few units as you can on a well-managed program that produces large numbers of units. Carter wants the latter, rather than the former.
POGO is very familiar with the two subjects of the memo, which are “will-cost” and “should-cost” management of programs: they are associated with two Pentagon whistleblowers POGO worked with in our early years. In some ways, the two prongs seem contradictory, but if implemented correctly and in concert, they should save money—or at least lead to better managed programs. But there are serious risks, which we will explore towards the end of the piece.
One of the prongs of Carter’s strategy is “should-cost.” Industrial engineer Ernest Fitzgerald, a now-retired Air Force acquisition official, was an aggressive proponent of should-cost beginning in the 1950s until his retirement in 2006. POGO founder Dina Rasor collaborated with Fitzgerald closely in POGO’s early years in the 1980s, when POGO was known as the Project on Military Procurement.
The essence of should-cost is “what work should cost after the fat is squeezed out,” as Fitzgerald put it in his 1989 book, The Pentagonists. Carter detailed some of what he dubbed the “ingredients of should-cost management” in his April 22 memo. Some of the ingredients he listed are:
- “Scrutinize each contributing ingredient of program cost and justify it. Why is it as reported or negotiated? What reasonable measures might reduce it?”
- “Particularly challenge the basis for indirect costs in contractor proposals.”
- “Benchmark against similar DoD programs and commercial analogues (where possible), and against other programs performed by the same contractor or in the same facilities.”
- “Promote Supply Chain Management to encourage competition and incentivize cost performance at lower tiers.”
- “Identify items or services contracted through a second or third party vehicle. Eliminate unnecessary pass-through costs by considering other contracting options.”
As always, the devil is in the details. Fitzgerald, in his 1989 book, said the DoD had embraced the term “should-cost” but not the substance of the hard-nosed, on-the-floor analysis of contractor costs that he recommended. Carter’s April memo states there are three ways to approach should-cost, and leaves it up to program managers and other service acquisition officials to determine which approach or combination of approaches to use. Each service is supposed to produce their first should-cost progress report to Carter on November 1, 2011, so it remains to be seen how the DoD will actually use should-cost.
The second prong of Carter’s strategy is “will-cost,” which is an effort to better budget and plan for weapons programs using more realistic independent cost estimates rather than rosy, mostly contractor-derived estimates.
Too often, the Pentagon, Congress, and the White House (and sometimes many of the country’s closest allies, as in the case of the JSF) “buy-in” to a program based on unrealistically low program cost estimates, only to be “surprised” years later. But even as estimates are readjusted upwards and reality sinks in—when it is discovered you can’t develop and build the weapon at the unrealistically low cost—programs are rarely canceled. Instead, the numbers of units are cut, meaning the estimated average per unit cost goes up because you don’t get economies of scale in production and development costs are divided by fewer units. This in turn further drives the per unit costs up: the dreaded program “death spiral” begins.
More realistic will-cost estimates are associated with Chuck Spinney, who was a Pentagon reformer and acolyte of Air Force Col. John Boyd. In the late 1970s and early 1980s, Spinney developed a very rigorous, data-based critique of how the Pentagon mismanages its weapons portfolio.
The new Cost Assessment and Program Evaluation (CAPE) office at the Pentagon, created by the Weapons System Acquisition Reform Act of 2009, is the latest attempt to impose improved cost estimates on the DoD.
If both should-cost and will-cost parts of Carter’s strategy are implemented aggressively by hard-nosed analysts at the beginning of programs, in theory, taxpayers could be surprised—programs could come in on time and on budget, or perhaps even under budget and faster than expected.
How would this happen?
Aggressive front-end independent cost estimates could force program managers to drop unnecessary weapons requirements, especially those based on immature technologies (this is one of the biggest drivers of cost and schedule increases in the development phase of a program). Furthermore, Pentagon budget planners could create appropriate space in the research and procurement budgets based on more accurate estimates of program costs, which would help stabilize funding for programs over the long run. Unstable funding also tends to increase costs and lengthen programs.
Meanwhile, should-cost teams could help get the DoD’s prime contractors to operate more efficiently (the defense contracting industry, particularly the major prime contractors, is relatively uncompetitive compared to commercial markets). Since will-cost is based largely on the DoD’s historical experience with weapons programs and many of those were poorly managed, you don’t want “did cost = will cost” as Ernie Fitzgerald put it. Ash Carter echoed Fitzgerald in a memo last fall:
While ICE [independent cost estimate] Will Cost analysis is valuable and credible, it does not help the program manager to drive leanness into the program. In fact, just the opposite can occur: the ICE, reflecting business-as-usual management in past programs, becomes a self-fulfilling prophesy. The forecast budget is expected, even required, to be fully obligated and expended.
But there’s a flip side to this as well. Should-cost can be used to make a contractor and government program manager buy time and conceal serious long-run cost problems. This is most likely to happen during the development phase of a program when cost estimates are skyrocketing and small batches of the weapons are being procured, known as “low rate initial production,” or LRIP. To come full circle to a program I mentioned at the beginning of this piece, this is a concern with the JSF program. As Aviation Week editor Bill Sweetman wrote:
"The cost of production is being reassessed," [JSF program manager Vice Admiral David Venlet] said. Venlet noted that the LRIP-4 price was settled in September 2010 and the contract was signed in November, "and we'll do everything we can to beat that schedule." There will also be a team working alongside the negotiators to examine Lockheed Martin's proposal and look for changes that would help meet the "should cost" goal set by Pentagon acquisition chief Ashton Carter.
Lockheed Martin, Venlet reminded us, says that it can come in below the predictions of the Cost Assessment & Program Evaluation (CAPE) office. But, he said "if LRIP-4 goes to its ceiling price" [that is, the cost above which the contractor team eats the overruns] "that means a decrement of their fee. Lockheed Martin can't lose money for very long. For LRIP-4 and LRIP-5, they need to have a good return."
One colleague emailed me to ask why Venlet was (apparently) giving Lockheed Martin a "get out of jail free card" on LRIP overruns. My interpretation is different: I think that what Venlet was saying was that he's not about to sign an LRIP-5 contract that makes Lockheed Martin look good, but that's a recipe for trouble two years down the line, when it blows through its ceiling and the supply chain starts bleeding cash like a stuck pig. (Emphasis POGO's)
In sum: You use independent cost estimate will-cost to help plan budgets over the long run, tame unreasonable expectations at the onset, and control for the contractors’ and military services’ rosy expectations of technological performance, integration, and cost. You use should-cost to help negotiate better contracts with a relatively uncompetitive defense industry. Ideally, you beat the will-cost expectations, but there are risks.
This is the theory, at least. While it’s good that there’s some high-level emphasis on both should-cost and will-cost, in Washington it almost always comes down to implementation.