Page images
PDF
EPUB

their own internal estimates so as not to exceed DOD budget amounts. Frequently, cost-type contracts are used at this early stage of the acquisition cycle so that the ultimate effect is a cost overrun which is absorbed primarily or entirely by the Government.

OSD 22-6: Prior to issuance of an RFP, DOD should informally consult with potential contractors in order to aid in contractor understanding of the project and to allow contractor input. Studies have indicated that some cost underestimation results when contractors do not fully understand the scope and complexity of the program.

OSD 22-7: DOD should propose revision of the SAR reporting format in order to make these reports a more useful management information tool.

Savings and Impact Analysis

Better estimates of weapons systems costs is one of the factors necessary to reduce instability in the weapons acquisition process. We project that improvements in weapons cost estimates would ultimately reduce by 3 to 10 percent the number of major programs which enter full scale production.

We have estimated that approximately $5.2575 billion in annual savings (in then-year dollars) could accrue from an overall reduction in instability in the acquisition process (see Issue OSD 23). This estimate includes the savings which would result from improvements in estimating weapons costs, which in itself would lead to reduced instability costs.

Implementation

The Secretary of Defense should direct that procedures for estimating weapons systems costs be revised as discussed above. In addition, contractors should be better informed about the complexity of programs at the bidding stage. Finally, greater incentives should be built into contracts. to preclude deliberate underbidding and to discourage cost

overruns.

II.

ISSUE AND RECOMMENDATION SUMMARIES (CONT'D)

B. WEAPONS ISSUES (CONT'D)

OSD 23: INSTABILITY IN THE WEAPONS ACQUISITION PROCESS

Summary Recommendation

The Department of Defense (DOD) should commit to a stable five-year spending plan for the acquisition of weapons systems at economical production rates. DOD should focus the attention of Congress on any significant increase in costs that would result from proposals to change the plan. Critical to achieving such a plan for program stability is the DOD's ability to relate financial affordability to proposed new systems early in the budgetary cycle. This would avoid accumulating systems that cannot be funded in economic production quantities during the entire production cycle.

Financial Impact

$5.2575

billion
annually
(after full
implementa-
tation)

Potential Savings: Cost growth due to insta-
bility on 25 major programs studied recently
is estimated to be $27 billion per year.
While it is not possible to control instabi-
lity completely, the improvements recommended
herein should save at least 5 to 10 percent
of the annual weapons costs (approximately
$70 billion) or approximately $5.2575 billion
(in then-year dollars) once full implementa-
tion is in place.

Background

For purposes of this report, program instability is defined as any event, except inflation, which causes actual program costs to exceed original estimated costs. Program instability would, therefore, include poor initial cost estimates, stretched and delayed production schedules, engineering changes, quantity changes and poor contractor performance.

Issue OSD 21 estimates savings available in RDT&E phases through limiting new starts. Issue OSD 22 discusses the problems associated with poor cost estimation, particularly at the time of the Milestone III decision to go into full-scale production. This issue treats the excess costs incurred in the production phase due to instability. Because inadequate cost estimates contribute to instability, the savings attributable to both issues OSD 22 and OSD 23 are combined.

Methodology

The Task Force conducted over 100 interviews with personnel in OSD, the services, and private industry. It also reviewed various articles, conference reports, books, many internal DOD reports and reports of the General Accounting Office (GAO).

The concerns addressed by this issue and OSD 22 were also addressed by the President's Private Sector Survey Procurement/Contracts/Inventory Management Task Force in their issues PROC 4 (Multi-year Contracting), PROC 6 (Planning), and PROC 8 (Cost Estimating and Scheduling). That task force estimated that $8.8 billion could be saved in DOD weapons procurement in the first three years through the adoption of their recommendations.

Findings and Conclusions

The OSD Task Force traced the program cost estimates for 25 major weapons systems, based upon an analysis initiated by the Defense Systems Management College. That analysis reviewed the cost growth factors in the acquisition process of major weapons systems.

The 25 major programs that we reviewed were initiated between 1971 and 1978. They have been affected by program changes, technological changes, quantity changes, timing changes and unanticipated inflation. We attempted to

measure the cost growth in these programs through the end of 1981, recognizing that a significant factor in seeming "cost growth" could be due to low cost estimations performed at the onset (see Issue OSD 22).

The 25 programs had an average age of seven years as of year-end 1981, but the weighted average was somewhat less when the size of programs was considered.

We found that the year-end 1981 total of the cost estimates for these 25 programs was 323 percent of the total development estimates made at the time the respective programs were initiated. Our analysis of the cause of cost escalation revealed the following:

[blocks in formation]

the

Even when unanticipated inflation was removed, cost growth due to instability factors for these 25 programs was $192.1 billion, or 183 percent of the original development estimate. The total annual increase in program costs, adjusted for year of entry for each program, was $26.9 billion, or 28.8 percent by weighted average (see Exhibit II-8 at the end of this issue).

This does not imply that the program measured at yearend 1981 was the identical program on which the development estimate was based. It may be, to the extent that the original estimate was unduly low, but other changes must be analyzed as well. These results do demonstrate that affordability cannot be adequately anticipated in the face of such unreliable data. It also demonstrates the high cost of not freezing the design or proceeding to approve the program before the true costs of technological improvements to the design are known.

On the basis of our review of this data, other studies reviewed, and discussions with DOD personnel, the OSD Task Force concluded that DOD has difficulty relating financial affordability to the production of proposed new systems sufficiently early in the decision-making process.

As a

result, more systems are put into production than can be funded in economical production quantities during the product cycle of each system, resulting in cost growth for reasons such as stretched production schedules and quantity changes. In this connection, the Selected Acquisition Report (SAR) of June 30, 1981, shows that 40 of 47 programs have experienced changes in the number of units to be procured, and 41 have experienced schedule changes.

On average, the production phase of a weapons system spans about five years and represents 80 percent of the total weapons system cost. The remaining 20 percent of the weapons system cost is represented by prior RDT&E expenditures, including full-scale development, which may have required up to 10 or 15 years before the system was ready for production (Exhibit II-4 on page 145 illustrates the process).

Because most of the funds are spent for procurement, and not development, a careful determination of whether a program will proceed into production or be cancelled should be made at Milestone III. At this point, the determination should be made regarding:

O

O

O

whether the particular weapons systems is still required to meet the threat as perceived early on in the R&D work program;

whether the weapons system reflects successful and effective culmination of the preceding R&D work;

whether the R&D development should be cancelled and written off as an unsuccessful venture;

whether the cost of the weapons system in econom-
ical production quantities over the entire life
cycle is affordable, in relation to other systems
presently in production or being considered by
DOD.

At the present time, we do not believe that the processes of weapons system selection, internal DOD budgeting for production, and Congressional appropriations are sufficiently interrelated so that an orderly and economical process of weapons system acquisition management may operate. Weapons systems should be acquired at the lowest reasonable cost per unit, which frequently means higher annual rates of production. However, there appear to be many systems in production at uneconomically low production rates that cannot be terminated because of the pressures of various constituencies such as the services, contractors,

« PreviousContinue »