Bending the Technology-Cost Curve

Aug. 1, 1983

What law of nature says that costs must always rise In some technological endeavors, they don’t.

Midway through the biggest weapon system modernization effort in thirty years, the Air Force faced grimly up the fact that its acquisition strategy wasn’t working.

Through the 1970s and into the 1980s, fewer numbers of systems than intended rolled off the production lines and into service, the Air Force paid more for them than had been planned, and development time was nearly double that of predecessor systems.

To determine exactly what was happening – and why – Air Force Systems Command conducted an in-depth analysis of past acquisitions, encompassing four decades and 109 different systems. That study, the Affordable Acquisition Approach (A3), revealed how program instability, cast overruns, and the traditional way of reacting to them have robbed the Air Force of wing after wing of new aircraft it might otherwise have had flying today. In addition, it showed how a continuation of the historical pattern might lead to a shortfall of twenty-three percent of worse in the projected weapons buy between now and FY ’88. (For a detailed report on A3, see “The Costly Alternative to Controlling Cost,” Air Force Magazine, June ’83.)

“We didn’t invent a new truth here by any means, but we did bring it out to where it hits you between the eyes,” says General Robert T. Marsh, AFSC Commander. “We can’t just continue indefinitely in the same old direction or the force structure will shrink to that one airplane Calvin Coolidge talked about, with pilots taking turns flying it.”

Many factors have contributed to escalating cost, but the pivotal one is program instability. From concept to initial operational capability, it now takes an average of nearly twelve years to field a weapon system. Thus, in it development, a typical system will have been through the administrations of at least two different Presidents, 2.4 full cycles of the Five-Year Defense Plan, and a dozen federal budget debates. Program directors will have come and gone, and there will have been some shifting in perceived requirements and relative priorities with in the air Force.

This situation practically invites instability. The high-value programs have suffered most. The B-1 bomber, for example, was started, stopped, and restarted. The MX missile has been subjected to a long string of agonizing reappraisals.

The Agreement That Wasn’t

To make matters worse, the Air Force often assumed more internal agreement about the nature of emerging systems than actually existed. That has been demonstrated starkly by AFSC’s new approach to baselining, which requires formal agreement on content of a program from the user, tester, builder, trainer, and maintainer. “The frustration level of my staff has mounted because it is so hard to get everyone to agree about a program baseline,” General Marsh says. “But I think the difficulty of the task attests to the need for it. We used to assume agreement. We now know we didn’t have it, and how hard it is to get.”

Loose and ephemeral baselines go side by side with faulty cost estimates because, as Brig. Gen. Daniel B. Geran, AFSC Comptroller, says “it is impossible to estimate what you can’t define.”

In the 1970s, defense budgets dropped, and procurement authority for force modernization fell far short of what the Air Force had expected to get. That, combined with unforeseen inflation rates, constant tinkering with baselines, and a variety of factors such as technical problems and underestimation of actual costs, left the Air Force with less buying power than it had counted on. The typical response – to cut back on quantities and stretch out acquisitions over more time – added to the problem. Inefficient production rates and additional overhead expenses drove unit costs up.

Costs and Budgets

The A3 study indicates that the Air Force cannot carry out its intended acquisition program unless it can bring cost growth to zero – down from the prevailing average of more than five percent a year – and at the same time get the full procurement authority called for its FY ’83-88 projections. General Marsh is not among those who see such an objective as being impossible to attain.

“I’ve been in this business too long to lay out what I think are unachievable aims,” he says. “In the past, we didn’t achieve the procurement authority we had planned on. Instead of eliminating some things, we took that reduced obligation authority and spread the hurt over all of our programs, including the well ones, and made them all sick.

“The striking lesson that came out of A3 is that we can’t continue to do that. If we have to take cuts, if we don’t get the program the President has asked for, then we’ve got to take that reduction in the smart way. Don’t shallow the base – narrow the base. That’s the first rule.”

He says that the Air Force leadership is prepared for the possibility that this strategy may involve more than killing weak or marginal programs. Some healthy and critically needed ones may have to go, too.

“What it says is that you might have to cancel your lowest priorities,” General Marsh continues. “We have proposed a program that we think to be the minimum essential. We need all of it. We feel strongly about that, and the President feels strongly about it, and we’re trying to defend that program so I’m not going to stand here and tell you what programs are going to be canceled. But if it comes down to the spades and we don’t get the program we’re asking for, we’re ready to approach that in a sensible fashion.”

The Air Force, with no real control over budget authority, is concentrating its energies on eliminating cost growth. In the wake of A3 systems Command has elevated cost reduction to the top priority in its 1984 corporate plan.

All this hits just as the Air Force is defining its long-range directions and requirements in such activities as Air Force 2000 and AFSC 1990. Continuation of business as usual, General Marsh says, will lead to “the day when we might not be able to afford both the weapons we need and the forces necessary to man them.”

Technology Can Help

In Project Cost, the broad effort to address the problems identified in A3, Systems Command is emphasizing affordability, stability, management, and rigorous application of solid business procedures. Beyond that, General Marsh believes, a proper technology investment strategy can contribute not only to operational capabilities needed for the future, but also to making those capabilities more affordable than now.

“I don’t think that it’s a law of nature that the weapon system cost curve has to continue rising,” he says. “I think that technology is the factor that can bend the curve back down, and we’ve got to put it to work for us.

“That cost curve has been going out of sight, but it is not typical across all technological endeavors. This wristwatch is a lot cheaper than its predecessor, and it’s a lot better. And that calculator is a lot better than its predecessor and a lot cheaper. There are lots of examples where technology has produced more capability at lower costs than the predecessor capability. We’ve, therefore, got to bring technology to bear in our weapon systems not only to produce improved capability but, perhaps more important, to reduce the cost per unity of capability.”

Computational technology, where very-high-speed integrated circuitry (VHSIC) can lead to less maintenance and lower life-cycle costs for avionics systems, is a prime example.

“Composites very definitely offer the promise of lower cost airframes,” General Marsh adds, “Technology is ringing the cost of propulsion – thrust per pound – down, and it is going to continue to do so. Composite applications to turbine engines will help bring that down. In our investment strategy, we have to give equal importance to those technologies that offer promise of reducing cost as we do to those technologies that offer the promise of increased capabilities.”


The three elements in Project Cost are affordability, stability, and management.

“Affordability,” General Marsh says, “is sort of a fundamental notion. First, we in this command have got to do a better job of making sure the air force understands the total implications of an initial decision to acquire a weapon system. We’ve got to do a better job of cost estimating so we can make a decision on whether a system is affordable or not.”

A related issue is the propriety of the costs being estimated.

“We’ve got a cost-based philosophy of procurement,” General Marsh says. “We generally try to satisfy ourselves that the costs are actually being incurred and that they are legitimate. We then add profit to that, and that’s the price of the undertaking. If you’re going to reduce the price of an item, you’ve got to get at the cost base.

“Now if you accept this thesis, and I contend that it’s true, then industry is not fundamentally motivated to reduce the cost base. In fact, you could say that our approach is a disincentivizing one, because profit is generally a percentage of cost.

“That leads us to do everything within our power to reduce that cost base. That’s why we seek out things like incentives to a contractor for productivity, because that will reduce costs. That’s why we subsidize the manufacturing technology efforts of the aerospace industry, because that will reduce costs. And that’s why I’m very much interested in their labor rates, in how many executives are apportioned to an effort, and in overhead of all kinds, because that comes back to us as cost.” (See box, p. 48.)


The classic “horror example,” as General Marsh calls it, from A3 is what happened to procurement of the original increment of 729 F-15 fighters. To stay within budget, the Air Force departed from its original plan of acquiring F-15s at an average rate of 144 aircraft a year and stretched the program out from six years to nine. (See chart on next page.) That added $2 billion – the price of an additional wing of F-15s – to the program cost.

The Air Force had hoped to stabilize the F-15 program by putting it on multiyear procurement, thus eliminating the ups and downs and unpredictability’s of year-by-year acquisition. (That approach has worked well with the F-16 fighter program, where nearly a quarter of a billion dollars has been saved already as a result of multiyear procurement. The cost of procuring forty-four KC-10 aircraft was reduced by about $606 million by using multiyear procurement rather than annual contracting.)

But such is not to be – at least not yet – for the F-15 program, where the horror story is not over. In May, the House Armed Services Committee rejected the Air force’s request for F-15 multiyear procurement and cut the FY ’84 buy to thirty, down from the Air Force’s proposal to acquire forty-eight.

This perpetuates the pattern that A3 warned against. True, less money is spent in the near term – $1.4 billion rather than $1.7 billion – but unit price for the smaller buy rises from $25.2 million to $30.9 million.

The telling effect is seen in projecting this schedule out over the 696 aircraft remaining in the F-15 procurement in FY ’84 and beyond. The air Force had proposed acquiring forty-eight in FY ’84, seventy-two in FY ’85, and then stabilizing production at ninety-six aircraft a year through 1991. That would result in a total cost of $24.2 billion and an average unit cost of $28.9 million over the life of the buy.

Horror Projection

The House ruling does not address production schedules in the out-years. If, however, the remaining F-15s were bought at a rate of thirty a year, the procurement would not be completed until 2007, at a total cost of $46.3 billion and a unit cost of $56.5 million. (It would mean stretching out the procurement for an additional sixteen years and more than doubling the total cost. Figured at the original unit price, that difference in total cost could buy an additional 760 airplanes.)

Other cost factors pale in comparison to the effects of instability. The air Force will continue to press its analyses and emphatically as it can, and in the meantime work hard on internal measures to promote stability. General Marsh believes that increased program stability may be the best tool has for controlling and reducing costs.

“The real illness in the business and the real cause of unconstrained cost growth that typified our programs over a period of ten to twenty years was in the stability area – in not laying out a program and adhering to it,” he says. “Instead of doing that, we cut production rates as we went along, taking the management reserve out of programs, delaying and stretching out programs.”

Congress willing, the Air Force does not intend to do that again.

Systems Command is about one third of the way through establishing firm baselines for its existing programs. Baselines define a program in terms of cost, schedule, technical content, and supportability. All future programs will be baselined as they enter full-scale development. The difficulty in getting all interested parties in ongoing programs to agree and sign up to baselines has been an eye-opener. The Air Force had not previously realized there was so much divergence of opinion about programs in progress. Resolving those differences should eliminated some of the instability problems later on – especially since anyone proposing changes will have to come armed with heavy ammunition.

“One of the root causes of cost growth is that we historically haven’t done a good job of defining the program in pre-FSD [full-scale development],” says Major General Melvin F. Chubb, Jr., AFSC Deputy Chief of Staff for Systems.

Realistic baselines depend on good cost estimates. Actually, AFSC’s cost estimating has gotten better since the 1950s and 1960s, but more than a third of the estimates still are hitting low of the mark. Systems Command has begun twenty-one separate actions to improve this average. Initiatives range from greater use of independent estimates to “murder boards,” where cost estimators present their findings and methodology to colleagues fro critical evaluation.

Program Management

“Programs often run into problems because they don’t get off to a good start,” General Marsh says. He notes that the historical trend is that up to eight-five percent of a system’s eventual cost can be laid to decisions made before it entered full-scale development. Only about three percent of the cost is expended in that period, though, and the crucial decisions are often made by people with limited experience in program management. Seasoned managers move in when the program approaches full-scale development.

“An we reduce costs simply by putting our most experienced people on the early phases of programs?” General Marsh asks. “We don’t know for sure yet.”

The answer will be coming, because AFSC has already asked its product divisions to put more experienced people – although not necessarily higher-ranking ones – on programs early.

“Further,” General Marsh says, “we are looking at longer assignments for program directors to ensure that they are in the position long enough actually to be accountable for cost-control efforts.”

Accountability begins early. When a new manager is assigned, he is given a reasonable time to study the history and status of his program. Before assuming control, he must formally concur with the baseline, just as everybody else involved with it has already done. After that, he’s responsible for it. This change of accountability is a contrast to bygone practice, when program managers were freer to blame their problems on the messes they inherited. They also felt freer to fiddle with baselines.

If, in the future, a program does get into trouble, the program manager will be required to submit at least one workable solution that can be implemented without diverting money from some other program.

The Bottom Line

“The bottom line of all this,” General Marsh says, referring to A3 and to the analyses since then, “has been that we in the Air Force have not been able to acquire as much capability for the dollars spent as we should have been able to acquire.”

The consequences of this situation, if not corrected, could go well beyond this sort of lost capabilities. In his personal assessment that prefaces the AFSC corporate plan for 1984, General Marsh puts it this way:

“There appears to be an increasingly questioning attitude about defense spending, both in Congress and in the public at large. This attitude has been heightened by the debate over major systems such as Peacekeeper, laser weapons, and by the major issues surrounding arms control. The perception that defense is growing ever more expensive, and that defense funds are not spent wisely has also contributed.

“Defense dollars will be increasingly hard to get, and our performance will be even more carefully scrutinized to assure that each and every defense dollar is spent efficiently. Every cost overrun, every schedule slip, and every instance of less than expected performance will weaken our credibility with the public and reduce congressional confidence and support.

“A direct link between poor management and a weakened national defense is not hard to establish.”