For the past week, the press has been reporting on the Government Accountability Office’s (GAO’s) new report on Department of Defense (DOD) cost growth. (“Defense Acquisitions: Assessments of Selected Weapons Programs,” GAO-09-326SP.)
The media has summarized the report’s findings – conveniently prepackaged in a one page précis by GAO – and quotes various reactions to the report. Prominent among the latter are the comments of John Young, the Pentagon’s Under Secretary for Acquisition, Technology, and Logistics – a position, known as “Acquisition Czar,” created by the “Goldwater-Nichols Department of Defense Reorganization Act of 1986,” (Public Law 99-433). It was supposed to help reform what in the 1980s was generally regarded as a “broken” DOD acquisition system.
The GAO report is important. It documents over time the cost of weapons development and procurement in the Pentagon. As measured by GAO, cost growth is today higher than ever before – well above the amounts recorded in the 1980s. While GAO does not say so, another way to characterize its new report is that it documents the failure of previous DOD acquisition system reforms, such as that by the aforementioned, and generally acclaimed, Goldwater-Nichols Act.
“Acquisition Czar” John Young’s reaction to the latest GAO report was made available in a memo he sent to Secretary of Defense Robert Gates on March 31. It was leaked to the press as quickly as April 1. Young’s memo is a good measure of just how broken the DOD acquisition process is, and it helps to identify, inadvertently, the root cause.
To sort through the wreckage of the current state of affairs, it is useful to recapitulate the GAO report – both its strengths and weaknesses, the latter ignored by the press.
Key elements of the GAO report follow. My own comments are sub-bulleted in parentheses.
* Since 2003, major DOD programs have grown from a total number of 77, costing $1.2 trillion, to 96 costing $1.6 trillion.
(Most of the new programs can readily be identified in DOD’s Selected Acquisition Reports [SARs] which are the basis for the GAO analysis. It is astonishing how few of these “new” weapons have anything to do with what used to be called the “Global War on Terror.”)
* Cumulative cost growth is higher now than reported in GAO’s 2003 report. The 2008 report found it totaled $301 billion; using a slightly different set of programs, the 2009 report found it totaled $296 billion.
* Measuring just R&D (rather than R&D and procurement combined) found 40% cost growth in 2008 and 42%; in 2009.
* Programs experienced schedule delays in 2008 averaging 21 months; in 2009 the average delay was 22 months.
* GAO surveyed 96 programs with questionnaires. Only 33 of those 96 programs reported to GAO that they “were going to test a fully configured, integrated, production-representative prototype.” (p. 18) Of those 33, only 17 reported to GAO they plan to complete such testing before full production.
(It seems that 63 of 96 programs were unable or unwilling to assert to GAO that they planned “to test a fully configured, integrated, production-representative prototype” at any point.)
(“Fly before You Buy” [fully testing before full production] was supposed to be one of the major reforms encouraged in the 1980s under Goldwater-Nichols. Instead, it appears to be mostly dead, and “Less Flying before Buying” now seems to be the practice. Was Goldwater-Nichols a step backwards on this important idea?)
* Less than a quarter of programs reporting to GAO said they use independent Cost Analysis Improvement Group (CAIG) estimates to measure the cost of their own programs; 70% of the programs reporting generate and use their own cost estimates.
(In other words, three out of four times, the in-house advocates are permitted to estimate their own costs. Their license to do so is essential for the initial “buy-in” of acquisition programs at “affordable” prices.)
(Like Congressional Budget Office [CBO] cost estimates, CAIG estimates are routinely criticized by program advocates as too high, but they almost always ultimately prove to be too low, once real costs are known.)
* Most growth is in the ten most expensive programs. (FCS, SSN-774, CVN-21, D-5, P-8A, F-18E/F, F-22, F-35, C-17, V-22) There, total acquisition costs grew by 13%.
(13% cost growth sounds modest but as GAO pointed out, these programs also experienced a reduction in the quantities delivered of 32%. That means that we pay more to get fewer weapons, and as the annual increments of additional cost for fewer programs build up, both the shrinkage and the cost expansion become stunning. These trends also show up in dramatically increasing unit costs, that GAO measured as increased between 38% and 127%.)
The GAO data adds up to a pretty horrific picture. Nevertheless, a major GAO conclusion is, “There have been some modest improvements in DOD’s acquisition outcomes.” (p. 1) This assertion is based on three findings, none of which are convincing:
* A $5 billion reduction in total measured cost growth occurred in 2009 ($296 billion) from 2008 ($301 billion). However, the sample of programs also changed from 2008 to 2009, which GAO also explains is part of the reason for the statistically insignificant (1.6%) reduction.
* GAO also asserts that “newer programs [less than 5 years old], on average, have not yet shown the same degree of cost and schedule growth.” (p. 6) But we know from past work, such as from retired Pentagon insider Chuck Spinney in several of his highly detailed analyses, that significant cost growth rarely shows up early, i.e. in “new” programs. Instead, it typically starts to ramp up in initial procurement in late developmental stages, when reality starts imposing itself on phony DOD cost estimates. Even, GAO says in its own report “.most cost growth does not materialize until later – after the critical design review.” (p. 10) The GAO finding that less cost growth might be occurring in “new” acquisition programs is unsubstantiated, even meaningless.
* GAO also finds improvement in “newer programs [that] are beginning with higher levels of technology maturity.” (p. 14) But it turns out that this assertion is based on just one of five programs being reported (by DOD, not independently assessed by GAO) as having “fully mature critical technologies . demonstrated in a realistic environment.” (p. 14-15) And later, GAO contradicts itself in saying that the number of ‘immature technologies being accepted in programs” has not changed. (p. 16)
The poorly supported GAO finding of “modest improvements” in the collapsing DOD acquisition system appears to be pasted onto the data and analysis in the report. The evidence for it is too scant; it is even contradicting. One has to wonder why such a flimsy, but significant, conclusion was made.
Speaking of flimsy evidence, we come to Acquisition Czar John Young’s March 31 memo about this report. (The memo has been circulated on the internet by several parties, but a link to it does apparently not exist; copies can be obtained from the author.)
In it, Young starts with an assertion that “The new GAO report continues to sensationalize the assessed cost growth of $296 billion in 96 programs.” He then finds three reasons to discount the cost growth GAO found, characterized by me as follows:
* It’s not cost growth if higher costs resulted from more systems being bought. Thus, if the advocates of the program helped to get it initially sold on the promise that a smaller number, at less total cost, were planned than was ultimately found to be needed, then it’s not “cost growth.” This exception ruled out $95.7 billion in GAO findings.
* It’s not cost growth if higher costs resulted from fewer programs being bought. As we know from painful history, higher expenditures often result in fewer purchases: unit costs get so out of control that the total costs of the program increase while the number to be bought decreases. By excepting this most pernicious from of cost growth, John Young rules out $72.2 billion in GAO findings.
* It’s not cost growth if it only grows by 10%. This arbitrary exception rules out another $57 billion in GAO findings. As noted above, behind modest sounding “non-cost growth” serious problems can lurk.
Ergo, Young concludes, “66 programs have performed reasonably well.”(p. 3 of the Young memo.)
As for the rest, $166.6 billion in 28 programs, there was indeed cost growth, but some of the factors (low cost estimates, changing requirements, optimistic schedules, immature design, etc.) causing it “are totally out of the control of acquisition program managers and do not necessarily indicate a broken defense acquisition management and oversight process.” (p. 3)
And there you have it: most cost growth is not cost growth and that which did occur is nobody’s fault. The Tooth Fairy did it.
On page 8 of the GAO report, there stands a remarkably simple and straightforward recommendation to deal with the cost growth problem: “The surest way to improve.is by reducing the number of underperformaing programs, by either completing or canceling them.”
This is good advice. Hopefully, when he announces what is being ballyhooed as “major” changes and reductions in the DOD program, Secretary of Defense Gates will follow it.
Moreover, we can learn – if we want to – that new regulations, even rewritten statutes, will not change a system where the people running it are determined not to change nor to accept responsibility for bad decisions. Reading over newly proposed legislation, such as that proposed by Senators Carl Levin, D-MI, and John McCain, R-AZ, in the Armed Services Committee (and recently watered down there by various committee-adopted amendments), we can see a replay of past history. Pretending reform in the face of serious problems has lots of precedence in the Pentagon and Congress. We can expect more of the same, just as we got from the 1986 Goldwater-Nichols reforms, unless genuine leaders and reformers emerge either in the Pentagon or Congress. The system will continue to seep fewer weapons for gushers of additional cost. And, or course, no one in the Pentagon or Congress will be responsible for this. For the responsible party, check under your pillow.
WINSLOW T. WHEELER spent 31 years working on Capitol Hill with senators from both political parties and the Government Accountability Office, specializing in national security affairs. Currently, he directs the Straus Military Reform Project of the Center for Defense Information in Washington. He is author of The Wastrels of Defense and the editor of a new anthology: ‘America’s Defense Meltdown: Pentagon Reform for President Obama and the New Congress’.