can it be controlled?
University of Wisconsin — Oshkosh
Overruns, indeed doubling of original cost estimates, have become a way of life in the field of research and development.The problem plagues commercial as well as government-sponsored efforts, but the government-sponsored efforts have had the more alarming repercussions. The strain of significant cost growth has caused concern not only in the government agencies involved but also in the economy as a whole. The billions of dollars involved in such cost growth lead to:
- Panic reprogramming of funds.
- Lack of funds for other deserving programs within our economy.
- Inflationary pressure to the extent that supplemental appropriations are obtained without corresponding revenues, and to the extent that such cost growth represents cost and price increases.
Factors Contributing to Cost Growth
Why does cost growth occur? There are a host of factors which contribute to the problem. Most attempts to deal effectively with these factors and thereby improve the underlying situation have fallen far short of expectations. Some of the most significant factors leading to cost growth are summarized below.
1. Early Contractor Cost Estimates
Much of the over-optimism surrounding early cost estimates stems from contractors’ anxiety to establish an early advantageous position, often through unsolicited proposals. Contractors often find it hard to resist low initial estimates when they have come to depend upon government business. Although the report of the Blue Ribbon Defense Panel implores contractors to take a more realistic attitude toward cost estimating and performance, this appears to be a vain appeal.1
2. Early Government Cost Estimates for Planning
The Logistics Management Institute has provided a good summary of the kinds of bias that tend to affect early estimates. They note that when the optimism of a prospective contractor is combined with the desire of a given military organization, in competition with others, to bring a specific weapon into its inventory, there is a compound tendency to underrate probable costs.2
3. Early Government Cost Estimates for Budgeting
There is a natural human tendency to understate the total life cycle impact of a new major system when seeking initial budgetary support. Since the ability to contract effectively for an engineering development program is contingent upon an adequate appropriation, low budgetary estimates tend to compound the problem of cost growth through inefficient stretch-out of the effort, panic reprogramming of funds, or both.
4. Cost Estimates for Contracting
It is a common practice for contractors to ask their internal organizations for detailed engineering cost estimates for their portion of work on the end items specified in a request for proposal (RFP). Subcontractor cost estimates are also compiled. In most cases the total “first cut” of such estimates far exceeds the figure which either management or the customer had in mind. Frequently the order is given by contractor management to arbitrarily slash each cost elementby a certain percentage, perhaps as much as 50% to 75%. This happens especially often in the case of competitive, cost reimbursement types of contracts. After one or two rounds of this process, an “acceptable figure” is arrived at for inclusion in the proposal. In the case of sole-source procurements, the process may be reversed, that is, contingency factors may be added to the basic estimate.
Government cost estimating teams supposedly take an independent look at the problem through preparation of an Independent Government Cost Estimate (IGCE), prepared before solicitation of quotations. In practice, the IGCE, if made, may be known by the contractors, and conversely, earlier contractor estimates may be known by the IGCE team members. When they are now known, the government estimates often vary over 100% higher or lower than the dollar figure in the contract. As a result, contracting officers often ignore the IGCE.
Estimates for contracting are therefore heavily influenced by known previous estimates and budgetary data. The only exception to this practice appears to be the efforts of the “should cost” teams for sole source procurements.
5. Contract Negotiation and Award Procedures
Although negotiated procurements do not require it, the lowest bidder usually does receive the award. For cost reimbursement competitive contracts, there is less pressure on the contractor to bid realistically. Each contractor, in addition to presenting a sound technical proposal, strives to underbid the other competitors. Ironically, the folly may be compounded even further by the government negotiation team, which may establish a contract negotiation objective below the lowest bidder's estimate.
6. Contractor Performance
There is no question that inefficiencies exist within most contractors’ plants. A “good” original cost estimate can look very poor in the final analysis due to poor performance by the contractor. One aspect of the cost estimating problem therefore entails the performance of contactors after contract award.
Under fixed price contracts, the cost burden of such inefficiencies should be the responsibility of the contractor. In practice, this has not proven to be the case under some of the principal fixed price incentive contracts, such as the contracts for the Lockheed Aircraft Corporation heavy logistics transport aircraft, the C-5A, and the General Dynamics all-purpose fighter aircraft, the F-lll.
For cost reimbursement type contracts, the principal cost burden for such inefficiencies is the responsibility of the Government.
The Armed Services Procurement Regulations (ASPR 7.402.2, September 30, 1970) make mandatory the inclusion of the Limitation of Cost Clause in all cost reimbursement type research and development contracts which do not provide the cost sharing. Although the wording of the clause appears harsh, there are many factors which tend to reduce the effects of the language. Of greatest impactis that, since the Government is interested in obtaining a finished product for test and possible later production and deployment, additional funds are supplied in the great majority of cases where the original cost limitations will be exceeded. Thus cost growth beyond the original cost limitation can occur, with Government approval, under all types of cost reimbursement contracts, through the workingof the Limitation of Cost Clause.
Cost growth tends to occur in steps. In other words, the contractor notifies the Government not once, but several times under the provisions of the Limitation of Cost Clause. In each case the motivations of the contractor (and often those of the agency involved) are somewhat similar to the original motivation of each, i.e., the full cost growth potential is not estimated for fear of adverse reaction and possible cancellation of the program.
7. Contract Changes
Another major factor that makes for cost growth under all types of Government contracts is the changes clause (ASPR 7-103.2, September 30, 1970). The neophyte in space and weapon systems programs is generally unaware of, and unable to comprehend, the magnitude of changes which are sequentially incorporated into a development effort. It is not unusual for several thousand change proposals to be in the flow of paper at a particular moment in time. A large weapon system program is fortunate if it can live with less than 15,000 proposed changes in its lifetime. Not all proposed changesare approved, obviously, but for those programs whcih have set out to restrict changes to a bare minimum, success has been elusive. Douglas Aircraft Corporation's (now McDonnell-Douglas) commercial passenger aircraft, the DC-8, and Lockheed's heavy logistics military transport, the C-5A, are notable examples of programs which were unable to restrict changes to a bare minimum in spite of objectives to do so. One major missile program was said to have involved one change per hour, of which some two-thirds had impact upon cost. In such an environment, the ultimate system may bear little resemblance to the system which was originally estimated.
8. Follow-on Contracts
After the Government has contracted for a portion or all of the original development on a system, it can find itself in what amounts to a sole source environment. In this environment, cost estimating motivations of the contractor may change significantly especially if the follow-on efforts entail fixed-price type contracts. The contractor may seek to compensate for his early losses or low-fee efforts by bidding higher than justified by realistic costs.
It is then up to the Government to evaluate critically all data referred to on Department of Defense (DD) Forms 633 (Contract Pricing Proposal) and to adopt a penetrating “should cost” analysis.
Pressure also exists to develop alternative sources and to “break-out” the procurement of components. This process has its own set of problems which often result in even greater life cycle costs for a system.
9. Other Factors
In addition to the factors already discussed, there are a number of other factors which contribute to cost growth patterns. Inflation, traditional habit patterns, delays, and corrections of deficiencies, are but a few of the additional factors which cause a system to greatly exceed its original cost estimate.
In most cases a combination of circumstances causes cost growth. It is virt ually impossible to isolate a single cause. It is therefore unrealistic to expect a single improvement to rectify the problem.
Attempts to reduce cost growth in the form of more thorough analysis of proposed systems and their cost estimates, more definitive contacting arrangements, more intensive management and control contracts, and more intensive record keeping of actual performance seem not to have resulted in significant reduction of cost growth.
Improving Early Government Cost Estimates
The problem is not simply one of improving the accuracy of cost estimates, for this is but one aspect of a much more complex problem. If the process is to be improved at any point, however, early government cost estimates represent one of the more promising points at which critical reviews and adjustments can separate fact from fantasy early enough to affect important decisions on large new systems.
It therefore is considered worthwhile to determine what leading government cost estimators and analysts identify as the principal problems of cost estimating and analysis and what they believe would improve the overall systems acquisition process.
In a recent survey over 450 cost estimators and cost analysts within the U.S. Army Material Command, the U.S. Air Force Systems Command, the U.S. Naval Material Command, and the National Aeronautics and Space Administration were asked to indicate what they considered to be the three most important problems to be solved in cost estimating and analysis. Analysis of their responses showed that these managers and practitioners of cost estimating considered the lack of an adequate data base to be the most important obstacle to their work.
Listed next most often was the lack of adequate, complete, or firm systems definitions, statements of work, specifications, and drawings. The respondents indicated that they cannot do an adequate job of cost estimating and analysis on many of the tasks assigned because they cannot clearly determine the scope and nature of the work to be done. In such an atmosphere, estimates will be greatly influenced by the judgment of the estimator as to what is required, probably resulting in underestimating more often than over-estimating the scope and nature of the work.
The third often listed problem was the lack of qualified, experienced, and specialized personnel; the lack of new, young personnel; and the lack of continuity of personnel. It is rare to find university courses devoted to cost estimating and analysis. Those that do exist are most often oriented to construction work. It is generally through long experience in procurement, accounting, or contract administration that one enters the field of cost estimating and analysis within the DOD-NASA environment. It is therefore difficult to find new, young personnel with suitable backgrounds. Moreover, rarely does an analyst become involved with a system at all stages throughout its life cycle. Cost estimating and analysis assignments tend to be for short periods of time, often on a team basis. The lack of continuity of personnel and the lack of in-depth familiarity with any one system is therefore a severe handicap.
When all of the problems mentioned were grouped into seven categories, a somewhat different order of problems resulted. Problems involving cost estimating techniques, tools, methodology, and procedures were; paramount among the seven categories. Many cost estimators and analysts do not believe there is sufficient knowledge concerning cost drivers, the impact of variables, risk and uncertainty, cost differences among manufacturing techniques, projection of future costs, and the impact of work phasing interruptions. They point to a lack of standardized formats, work breakdown structures, accounting systems, cost estimating relationships, terminology, procedures, regulations, and thecost estimating methodology in general. Thus, the state-of-the-art in cost estimating and analysis appears to need considerable improvement according to those who are closest to the subject. The second largest grouping centered about historical data problems, already discussed.
The third largest grouping entailed problems of completing a particular cost study.Many respondents believe they are not given adequate time to prepare their studies, that they cannot adequately analyze the data as it is submitted, and that their working relationships are not good with personnel in other organizations at their same level and above. These types of problems will probably persist to some degree. If many of the problems of methodology and historical data can be solved, however, perhaps these problems will also become much less significant.
The agencies differed somewhat in their emphasis upon problems, but these differences appeared to be largely attributable to the differing missions, emphasis, personnel mix, and current state-of-the-art within the agencies. None of the agencies seemed very concerned with the number of people to do the job or the adequacy of management guidance. In spite of minor differences, there was surprising unanimity regarding the principal problems of cost estimating and cost analysis. That is, the lack of adequate methodology and the lack of good historical data represent the paramount concerns of all agencies.
If average cost growth is to be reduced significantly in the years to come, original cost estimates must become much more realistic. It is not possible to say with confidence that costgrowth will be eliminated if improvements to cost estimating and analysis are undertaken. It can be surmised, however, that as original cost estimates become more realistic, greater pressure will existto stay within those original estimates. Managers within governmental agencies and within contractorplants can then be held more accountable for their performance. In that type of environment, American managers have historically performed well.
The only way in which public confidence in the acquisition process can be restored is to improve the actual process itself. The first step in improving the process should be to improveoriginal cost estimates.
(1) See Blue Ribbon Defense Panel, Report to the President and the Secretary of Defense (Washington, D.C.: Government Printing Office, July, 1970), pp. 85-86.
(2) Logistics Management Institute, Improved Cost Estimating Techniques (Washington, D.C.: Logistics Management Institute, December 1970), p. 16.