Time To Fix Military Procurement -- Again
Momentum is building for real reform of one of government's most intractable problems.
Saturday, April 25, 2009
by Sydney J. Freedberg Jr.
The Defense Department's process for buying weapons is byzantine, tedious, and yet a matter of life and death. A system that produced $640 toilet seats was bad enough when the enemy was the Soviet Union, a slow-moving target with bureaucratic troubles of its own. Against the low-budget innovators who attack Americans with box cutters and roadside bombs, the Pentagon has somehow managed to spend more than $100 billion a year on procurement and still fail to deliver the right gear on time.
"Our enemies are adapting faster than we are," said Andrew Krepinevich, a retired Army officer who heads the independent Center for Strategic and Budgetary Assessments. Improvised explosives became the No. 1 killer of U.S. troops in Iraq soon after the 2003 invasion, he noted, but "it takes us until about 2007 until we get a sufficient number of [mine-resistant ambush-protected vehicles] in the field -- and that takes an extraordinary action by the secretary of Defense to overrule the acquisition system."
On April 6, that secretary, Robert Gates, announced the biggest cutbacks and cancellations since then-Secretary Dick Cheney began the post-Cold War drawdown in 1991. Gates also promised one massive new investment: the addition of 13,000 civil service positions to the Pentagon procurement workforce, which was slashed in half during the 1990s and never recovered, even after the procurement budget that employees administer rebounded back to 1980s-buildup levels after 9/11.
Cutting home-state spending at weapons factories while expanding the federal workforce is hardly a way to win over members of Congress. Yet key Capitol Hill leaders have been cautiously approving. Meanwhile, Congress is working on its own procurement reforms, which Gates, in turn, cautiously endorsed in his speech. A bipartisan bill to toughen standards on cost estimation and weapons testing has passed the Senate, where it was sponsored by the two top members of the Armed Services Committee, Michigan Democrat Carl Levin and Arizona Republican John McCain. The influential House Budget Committee chairman, Rep. John Spratt, D-S.C., has introduced a companion bill.
The House's elder statesman on defense, Armed Services Committee Chairman Ike Skelton, D-Mo., is reserving judgment on the McCain-Levin plan. "The Senate bill has some flaws," he told National Journal. Skelton is awaiting recommendations from the ad hoc subcommittee he set up to propose procurement reforms, led by New Jersey Democrat Robert Andrews. (See "A Lawmaker Looks at Reform," p. 42.) "They're going to do it right and not rush the process," Skelton said. He and Andrews announced their own initial proposal on the 23rd.
Even some defense-industry officials acknowledge, guardedly, that the pressure and prospects for change are unprecedented. One veteran Hill staffer now working in aerospace, said, "For the first time in probably 15 years, I've seen a coincidence of thinking in Congress and in the administration that suggests an appetite for big changes in acquisition policy -- capital 'A,' capital 'P.' "
Much of the willingness to change, however, stems from the unprecedented breakdown of the system. In March, the Government Accountability Office surveyed 96 "major defense acquisition programs" and found overruns totaling $296 billion.
How did this happen? Both parties have kicked the can down the road for almost two decades since Cheney staged the last round of major cuts. The Clinton administration trimmed programs across the board during the drawdown, while the second Bush administration boosted them across the board after 9/11, but in both cases officials let a lot of troubled projects limp along without either canceling them outright or funding them adequately. Until April 6, no one had the combination of political will and strategic vision to kill not one program here or there but a half-dozen at once. Meanwhile, weapons bought during the Reagan buildup are wearing out.
"The problem grew much, much worse under the [George W.] Bush administration," said Anthony Cordesman, a top defense analyst at the Center for Strategic and International Studies. "If you don't make hard decisions for nearly a decade -- which is what took place both at the end of the Clinton administration and in the Bush administration... your problems are cumulative."
Part of the breakdown, however, long predates the Bushes, or Clinton, or even Reagan. The procurement system has been delivering weapons over budget and behind schedule since at least the Korean War. "People were saying that in the second Eisenhower administration," Cordesman said. "People have been talking about exactly the same reforms ever since."
Making a fix that sticks, this time around, requires understanding why previous attempts have so consistently failed. Past reformers have fired one would-be silver bullet after another, with each brainstorm tending to contradict the previous one: centralizing procurement decisions, then decentralizing them; imposing strict contracting rules, then freely granting waivers; building parts to strict military specifications, then buying from commercial vendors on the open market; outsourcing major decisions to corporate "lead system integrators," and then bringing them in-house again; holding contractors to a fixed price on some contracts, and reimbursing them on others for overruns on a "cost-plus" basis; developing some technologies in prototype before starting large-scale production, and launching others in "spirals" between one production lot and the next.
Impervious to all of these procedural changes, the underlying problems remained. As long ago as 1967, Norman Augustine, then a Pentagon official and later a defense-industry executive, calculated the trend lines at which each generation of weapons was growing more expensive than the last and predicted, only half in jest, that by 2054 the entire Defense Department budget would cover just one airplane. "I'll never live that down," Augustine, now retired, told National Journal with a chuckle. "Unfortunately, it's still true." (See "A Businessman Looks at Reform," p. 44.)
Some of the highest technology, and the severest problems, are in the Air Force, whose top civilian and military leadership Gates ousted last year. The service's F-22 fighter plane has become a poster child for cost overruns and delay. The program began in the 1980s as the Advanced Tactical Fighter with a plan to build 750 aircraft at $35 million each to sweep Soviet planes out of the skies over Europe. After the Soviet threat evaporated and the defense budget started declining in 1991, the program limped along through successive delays -- which meant paying contractors' fixed overhead costs for longer than planned -- and through reductions in quantity -- which meant applying those overhead costs to ever-fewer airplanes. In his speech, Gates finally put the F-22 program out of its misery at just 187 aircraft costing on average $185 million apiece to build, or more than $350 million each counting the years of research-and-development costs.
While the F-22, a high-performance stealth fighter, represents the expensive pinnacle of American technological ambition, more-modest Air Force programs have had troubles of their own. In 2007 and '08, the service awarded two major contracts, each protested by the losing bidder and overturned by the GAO: the CSAR-X combat search-and-rescue helicopter -- which Gates canceled on April 6 and the KC-X midair refueling tanker -- which Gates will put out for bids again this summer.
In each case, an understaffed and undertrained procurement workforce made basic mistakes. "When you lay down a requirement for a new system, you try to establish the 'key performance parameters,' or KPPs," said retired Gen. Lawrence Skantze, who was chief of Air Force Systems Command (disbanded in the workforce cuts of the 1990s). "You shouldn't have more than 10, for the most part," Skantze said. "In the case of the KC-X, they had 37 KPPs and then they broke them down into 800 subrequirements." Yet the plethora of requirements did not make them any less ambiguous or more comprehensive, and planners somehow omitted a basic "concept of operations" for how the tanker was to be used, Skantze found in his review for the Air Force.
"You have to go back to square one," Skantze said. "A lot of what happens in the front end is, everybody wants to get going, and if you've got any questions, 'We'll straighten that out later.' But oh, you pay for it."
The perverse incentives to overpromise at the outset of a major weapons program, setting it up to underdeliver later, arguably account for the most fundamental, intractable problems in the defense procurement system. One independent study by the Rand think tank showed that the actual costs of weapons systems exceeded original estimates, on average, by 60 percent.
"I know for a fact that the estimates that we give are the best estimates we can give," said one defense procurement expert with experience in both business and government, who asked to remain anonymous. "Now, should we add 60 percent to our multibillion-dollar development estimates that we give at the outset? The answer of course is, we wouldn't get picked."
Like a contractor bidding to renovate a home, a defense contractor has every incentive to lowball the initial estimate to get the job and then jack up the price through change orders later. But unlike you or me looking to redo a kitchen or basement -- a decision we will have to pay for and live with -- a Pentagon program manager has every incentive to go along with the contractor.
In complex defense programs, overruns may take years to show up, often after the original program manager has moved on. But the immediate, pressing task, on which the program manager's job depends, is to persuade one's superiors to keep the program in the budget. And those superiors, in turn, have every incentive to accept lowball estimates so that their total budget looks smaller to their bosses, all the way up through the Defense secretary, the Office of Management and Budget, and ultimately Congress.
Bottom line: It is Congress that creates the perverse incentives by voting to fund programs that pull this 'lowball now, overrun later' bait and switch. "When rigorous cost estimates are rendered at the outset of a program, it makes it less likely that the program will be funded at all," said Loren Thompson, chief operating officer of the Lexington Institute and a consultant close to industry insiders. "By beginning a program with very optimistic estimates, you not only can get it funded but you also can almost guarantee that down the road there will be more jobs and more rewards for the political system in keeping it going. Whenever we have a cost overrun on a program, that money goes somewhere: It goes to a political district. And so the system doesn't actually punish cost overruns most of the time."
The obvious solution to fluctuating cost estimates is to fix the price paid to the contractor in advance -- obvious and, given the inherent uncertainties of inventing high-tech weaponry to order, also impractical. "Ultimately, you cannot get it under control until you go primarily to fixed-price contracting, but you can't go to fixed-price contracting when you're pushing the envelope in R&D, because, by its nature, since you're in development, you don't know how much it's going to cost," said John Lehman, Navy secretary during the Reagan buildup and an adviser to McCain's 2008 presidential campaign.
Even on simpler systems with proven technology, much of the slip in cost and schedule results from the Defense Department's constant second-guessing of what it wants to build. "No company who has any semblance of sound management would bid fixed-price on a program that does not have the design settled, and hardly any of them do now, because everybody can write change orders," Lehman said.
Procurement veterans call this phenomenon "requirements creep," the military's tendency to up the ante for what a system is supposed to do after the contract has been signed. On the nominally low-cost Littoral Combat Ship program, for example, the Navy was issuing 75 change orders a week at one point, including a total overhaul of the standards for damage resistance and fire safety -- arguably necessary, but requiring costly do-overs that doubled the cost.
Every additional requirement is somebody's bright idea, and each might be justifiable on its own terms. But as the bright ideas accumulate, the impact on cost and schedule is exponential. And the irony is that reforms often add extra layers of supervision, which means that more officials and organizations feel entitled, even obligated, to impose their own ideas, and on and on.
Experts who agree on little else join in warning that neither Congress nor the Pentagon can simply reorganize its way out of the problem. Ultimately, process matters less than people.
"The secretary has the tools he needs," said William Perry, Defense secretary under Clinton and probably the last person who successfully made large-scale changes to the weapons acquisition system. "What is required is the political will to act. The only person who can bring discipline to that system is the secretary of Defense, and with this speech that Secretary Gates made recently, he was making an important step in that direction. If he doesn't do it, nobody does."
This is the 14th in an ongoing series looking at an issue on President Obama's agenda. The entire series can be found at NationalJournal.com/agenda. Next week: Medicare financing.