GAO Report Details the Challenges of DoD Weapon System Programs

The Government Accounting Office’s (GAO) recently released the results of its sixth annual review of Defense weapons system programs. The results showed significant, widespread, and growing problems in these programs.

The 72 programs that were examined had an average 21-month delay in delivering their systems’ initial operating capability (IOC). Furthermore, their total acquisition costs grew 26 percent from their initial cost estimates, producing an increase of $295B in the programs’ total costs. To put that number in perspective, the $295B of increased program costs represents the total tax revenue derived from 26,663,051 individual tax returns.

Even more troubling, the results of this annual review showed that the problems in the performance of these programs are growing. GAO compared the results from the FY2007 portfolio of programs with those from FY2005 and FY2000. The comparison showed that problems in the measures of program performance were increasing, particularly between FY2005 and FY2007. The charts on the right show the changes GAO identified in the programs’ schedule delays and cost growth. These results indicate that not only are the programs’ problems growing but that they’re growing at an increasing rate.

The fact that these problems stretch across many programs and several years shows that they are not explainable simply as the management failings of individual programs and program managers. What then explains the programs’ persistent performance problems?

In its report, GAO points to “the lack of widespread adoption of a knowledge-based acquisition process within DOD despite polices that support such a process. Our assessment of 72 weapon systems shows that DOD programs continue to proceed through critical junctures with knowledge gaps that expose programs to significant, unnecessary technology, design, and production risks.”

The report goes on to point to three key areas in which a lack of program knowledge has contributed to program management problems:

• Is there a match between program resources and requirements? An especially important consideration is whether the technologies incorporated into the system’s design are mature enough to support the system’s performance.
• Is the design stable and adequate to meet user and program requirements?
• Is the production process mature? For example, has the production contractor demonstrated its ability to manufacture the system within specifications and in a manner consistent with the program’s technical, cost, and schedule requirements?

Sources cited in the Washington Post article about the GAO’s findings, add to the list of factors that drive the programs' performance problems. The problems are also caused, they feel, by:

• “Too many programs chasing too few dollars”, possibly resulting in overly optimistic initial estimates of program budgets, schedule and technical results.
• The excessive length of time needed to design, develop, and produce a system.
• A frequent failure to freeze the requirements and configuration of the system being developed.

The Defense weapon system programs are by no means unique in dealing with these kinds of program management challenges. For example, in 2006 the Census Bureau awarded a $595M contract to develop handheld computers to be used by enumerators to manage and conduct follow-up field interviews as part of the 2010 census. However, as reported in Federal Computer Week, the program did not begin with a clear requirements specification and as a result suffered a scope creep that significantly increased the program’s cost and risk. The Commerce Department therefore decided in April to cancel the program and revert back to using paper and pencil to collect the data in 2010. Though some of the systems developed in the program will still be used to validate address locations during the census’ planning, much of the value of the investment made in the program so far will be lost as will the potential cost savings the handheld computers could have provided. The switch back to paper is expected to add $3B in costs for the 2010 census.