Estimating the potential for savings on production programs is difficult, if not impossible. A key premise of this analysis is that there are too many production programs chasing the available production budget. It is hypothesized that affordability reviews and overprogramming limits imposed at the time of new system starts would reduce the number of programs entering the system and, therefore, would relieve the intense pressures for program stretchouts and volume reductions. The effect would be to permit program production to proceed at more economical rates and provide the required unit volumes in a more timely manner.

Given these assumptions, the financial implications of this proposal will be RDT&E savings early in the program and efficiency savings on program production. However, DOD makes no attempt to track actual program costs versus an ideal program profile. The savings which might result by attempting efficient program rates, therefore, are not available and would have to be developed program by program. We believe these savings to be considerable, given the large number of programs which have suffered delays and cutbacks. The financial implications of even modest efficiency savings would obviously be considerable.

In total, not including any potential savings for production efficiency, it is estimated that implementation of the proposed limits would save the DOD and the U.S. Government a minimum of about $1.1145 billion (in then-year dollars) per year in RDT&E spending. These savings could be used to accelerate other RDT&E efforts or to improve production efficiency. The projected RDT&E savings, while substantial, are believed to be modest relative to the potential savings on production programs and the corollary improvements in force effectiveness resulting from fielding defense systems earlier and in the required volumes.

The OSD Task Force analyses in Issues OSD 22 and OSD 23 provide alternative approaches to measuring the production savings that would derive from reducing instability, which is caused, in major part, by an excessive number of new starts at the RDT&E stage.

Implementation

The Secretary of Defense should revise DOD Directive 5000.1 and Instruction 5000.2 to require complete affordability reviews (through production) of all major programs from the time new starts are approved.

In addition, DOD Instruction 7000.3 should be revised to require the SARS to project system costs for all major defense systems through the final production phase, from the time each new start is approved.

Finally, the Secretary of Defense should issue a directive limiting total overprogramming at each phase in the acquisition cycle to fixed percentages of the total DOD funds that are likely to be available through the final production phase.

The Secretary of Defense should establish procedures to ensure more accurate estimates of weapons costs in order to permit better planning and reduce cost overruns. Specifically, provision should be made for establishing a new production cost baseline at Milestone III. The Secretary should also establish procedures to create a policy-oriented data base for use by DOD management. The Department of Defense (DOD) Comptroller should be assigned the responsibility for analyzing the affordability of proposed weapons systems and for monitoring actual costs versus estimated costs. Internal DOD budget data should not be disclosed to contractors at the pre-award stage, because such disclosure encourages contractors to submit bids that are less than their own internal cost estimates. In addition, prior to issuing a Request for Proposal (RFP), DOD should consult with industry in order to give contractors a better understanding of the complexity of the project. Finally, in order to give contractors more incentive to estimate costs more realistically, contracts should be used which require contractors to absorb a greater share of cost overruns.

Financial Inpact

Potential Savings: Improved cost estimates which, in most cases, would be higher than existing cost estimates would result in the elimination of an estimated 3 to 10 percent of major programs prior to entry into production. Improvement in the estimation of weapons systems costs is a necessary condition for reduction in instability in the weapons acquisition process. We have not attempted to calculate discrete savings attributable to our recommendations regarding each of these two issues. Estimated savings from the reduced instability recommendations (Issue OSD 23) include potential savings attributable to better estimation of weapons systems costs.

Background

DOD Directive 5000.1 and DOD Instruction 5000.2, recently revised, specify the phases in the major systems acquisition process. The intent of Directive 5000.1 is to delineate prescribed decision points in the acquisition process which require approval of a system, based on such items as technical feasibility, affordability, acquisition time, readiness and system flexibility prior to its moving to the next phase in the acquisition process.

Specifically, decisions are required by the Secretary of Defense for major systems new starts (based on the recommendations of the Defense Resources Board) and based on the recommendations of the Defense Systems Acquisition Review Council (DSARC) at the three so-called milestones. At Milestone I, the concept selection must be approved before the system moves into the demonstration and validation phase. Approval is based on preliminary concepts, cost estimates, schedules, objectives and affordability estimates. At Milestone II, after preliminary data on the system have been defined somewhat in the demonstration and validation phase, the decision is made regarding full-scale development. Milestone III is the production decision point. Approval to proceed into production is contingent upon the appropriate service's demonstration that financial resources are available or can be programmed to complete the system development, produce efficiently, and to support the deployed system (Exhibit II-4 on page 145 illustrates the process).

to

DOD Directive 5000.4 provides for independent evaluation of weapons system cost estimates by the OSD Cost Analysis Improvement Group (CAIG). The CAIG provides the DSARC with a review and evaluation of both the independent cost estimates of the services and the program office cost estimates. Independent cost analyses by the services must be prepared by organizations in each service separate from the control and direction of the program office responsible for acquisition of the system.

Methodology

Interviews were conducted with DOD and industry representatives. Internal DOD data were examined, and various studies, conference reports and articles were reviewed.

4

Findings

Despite the various independent estimates and reviews of program cost estimates which are required, DOD has been continually faced with significant cost overruns on weapons systems acquisitions. The Selected Acquisition Report (SAR) of June 30, 1982, indicates that for 39 programs, cost estimating errors alone (exclusive of other elements contributing to cost growth, such as unanticipated inflation or engineering and quantity changes) amounted to $10.8 billion, or about 9 percent of the original estimated cost of $124.6 billion.

One reason for cost estimating errors is that the measurement of a program cost overrun compares actual costs to a baseline cost estimate that is prepared early in the life of the development of the weapons system. The baseline cost estimates for major programs reported in the SAR are established at Milestone II, prior to the system's entering the full scale development phase. At this early stage in the development of the weapons system, it is too soon to establish a baseline of estimated program costs against which actual costs will be measured to determine the extent of cost overrun or underrun. Up to Milestone II, only the demonstration of the system and validation of requirements has occurred. Little or no production has occurred, so there is little experience for forecasting costs with any degree of accuracy.

Another part of the problem in cost underestimating lies with contractor underbidding in the initial development phases of the program. The OSD Task Force believes that the problem of intentional underbidding by defense contractors is compounded by defense contractors' underestimation of the actual complexity of producing the system. Based on three major weapons programs, the following revealing information is disclosed by Joseph P. Large in "Bias In Initial Cost Estimates How Estimates Can Increase The Cost of Acquiring Weapons Systems":

The differential between columns one and three is greater than the difference between columns one and two. This suggests that deliberate understatements of cost are a less important cause of cost growth than failure to appreciate project difficulty.