During the late 1970s, when the prime rate soared toward 20 percent and inflation seemed chronically beyond control, the beforethefact estimates often turned out to be too low. When they were, Pentagon officials went back to Congress for supplemental appropriations, to make up the difference. For the most part, Congress approved the requests. But during the 1980s exactly the opposite has happened. The estimates turned out to be too generous, but this time there was no afterthefact rectification. The Pentagon did not offer to give the windfall profits back.
The estimates proved to be high not merely because inflation fell so fast but also because of a quirk in the indexing formula. Since fiscal year 1983 the Pentagon has "protected" part of its budget against inflation by means of a 30 percent "kicker." If the expected inflation rate for defensecommodity purchases was five percent, the rate allowed for major weapons purchases would be 30 percent higher, or 6.5 percent. The 30 percent kicker, based on the assumption of fastrising defense prices, was one of the "Carlucci initiatives" for better management of the Pentagon, developed by the Deputy Secretary of Defense, Frank Carlucci, and promulgated by Secretary Caspar Weinberger early in his term.
As a management incentive, the 30 percent kicker has its drawbacks, since it indicates to companies and military planners that continual cost growth is the norm. In addition, technicalities in the indexing system are such that when a weapon's cost rises, the increase can easily be defined as inflation, and therefore not anyone's fault, rather than as the result of mismanagement, waste, cost overrun, goldplating, or other culpable behavior. Nonetheless, the Pentagon's overall costs have not risen as fast as expected, leaving the military with more "protection" than Congress, or even the White House, intended it to have.
At the request of Senators David Pryor, Charles Grassley, and Thomas Eagleton, the General Accounting Office attempted to calculate the size of this inflation dividend. In a report wryly titled "Potential for Excess Funds in DOD," released in March, it estimated that the surplus inflation payments from 1982 to 1986 amounted to the almost unbelievable sum of $39.5 billion, not counting overestimates in the price of fuel. Through the miracle of compound interest each past year's inflation windfall becomes part of the budget baseline and in the next year is itself "protected" against inflation. If the GAO's calculations are correct, the overestimate of inflation—not the price of the weapons themselves, and not even the inflation in weapons costs, but simply the inflation bonus—comes to about $165 per person in the United States.
Even more remarkable, and at first even harder to believe, the GAO contended that no one could really tell what had happened to the money. In the past few years the Pentagon's "unobligated balances"—sums previously appropriated but not yet committed—have mysteriously increased, from $28 billion in 1981 to $63 billion now. Some of this was to be expected—when a big weapons system is approved by Congress, the money goes into unobligated balances until the Pentagon lets the contracts. But the increase is more than can be explained that way. Perhaps the difference is what became of the inflation dividend, the GAO said. At the moment, it is impossible to say for certain, because the Pentagon's auditing system is not designed to reveal clear connections between money approved through the budget and money spent. Some of the money may have been absorbed in various kinds of cost growth; some may be sitting idle, waiting to be "protected" against future inflation; some is eligible for "reprogramming" to other purposes by Congress. What is clear is that the military has received $40 billion that no one intended it to have.
Beyond the inflation dividend, and in seeming contradiction to it, is a more basic explanation for the spendup—buildup gap: everything costs more than it used to, so those budget dollars don't go so far. (The higher prices have nothing to do with general inflationary pressures, which the index and "kicker" were meant to offset.)
The average military airplane of the late 1970s cost about $20 million. Its counterpart today costs over $27 million. (And that's not counting two of today's biggestticket items, the B1 bomber and the C5 transport.) Tactical missiles cost roughly $40,000 apiece in the late 1970s and cost $80,000 now. Helicopters have gone from about $5 million apiece to about $8 million.
Some of these increases occur when the military shifts from an older, less expensive system to a more costly replacement. But the costs go up even when the purchase list does not change. Last fall Representative Les Aspin, the chairman of the House Armed Services Committee, reported on the fifteen major weapons whose costs could be tracked from 1977 to 1985. (That is, weapons that had been started before 1977 and were still in production in 1985.) When paving the way for new purchases, military planners always forecast an upward "learning curve," which will drive down the price of weapons as more of them are made. But Aspin's report suggested that if there is a learning curve, it has a downward slope.
Only two of the fifteen weapons got cheaper over the years—the Sidewinder missile and the Harpoon missile. In all the other cases the price went up. A Phoenix missile cost 80 percent more in the mid1980s than it had in the late 1970s. An F15 fighter plane cost about 50 percent more, a TOW or Sparrow missile about 40 percent more. Even during the spendup the defense budget could not keep up with soaring unit costs, so of course the number of weapons purchased went down. Has there ever been a commercial manufacturing industry in which production went up—and unit price went up too? How many computers would have been sold if memory chips still cost $300 apiece, instead of $5?
BUT TO SAY that the weapons cost more is only to raise another question. What drives their prices?
Perhaps the answer is simply increased profit. Without question, military contracting became much more profitable after 1981 than it had been for many years before. For American manufacturers as a whole, the early 1980s were a time of sorrow, caused mainly by export problems (which in turn were caused mainly by the overvalued dollar, which was mainly because of a high rate of government spending and borrowing, which leads us back to the spendup). But during this gloomy era the defense business was remarkably good. To determine just how good, the Navy Department commissioned an audit of profits among twentytwo major defense contractors. According to the report, which was released last summer, companies that undertook both government contracts and commercial work found it far more lucrative to sell to the government. In 1984 sales to the government (mainly the Defense Department) represented 42 percent of Boeing's business but accounted for 94 percent of its profits. For Lockheed the government represented 85 percent of sales and 95 percent of profits. For McDonnellDouglas the government was 69 percent of sales and 98 percent of profits. Raytheon made less than half its sales to the government but made 79 percent of its profits through those sales.
Perhaps profits rose because of contracts that perversely reward companies for jacking costs up, not holding them down. Perhaps prices and profits rose because the government tried to buy so much so quickly, creating bottlenecks. Perhaps the sky'sthelimit atmosphere of defense spending in the early 1980s tempted some contractors to grab for all they could get. Perhaps the government wanted to ensure that the companies making up the "defense/industrial base" remained solvent and financially strong. Whatever the combination of reasons, the result was clear: American manufacturers could make more money producing weapons than doing practically anything else, such as competing with the Japanese.
It was not always this way. In the late 1970s government business was less profitable for these companies than commercial sales. In those days the companies' typical "return on sales"—operating profit divided by sales—was ten percent for private business and eight or nine percent for government contracts. In 1981 the positions reversed. Ever since then government business has been more profitable (nine percent return on sales), and commercial sales less (seven or eight). By 1984 the major defense contractors earned an 11 percent return on assets in their commercial sales. For government contracts their return on assets was 26.7 percent. The FMC corporation, which manufactures the controversial Bradley Fighting Vehicle, received a 54 percent return on assets on government business in 1984.
The life history of the one weapon Caspar Weinberger has canceled after it went into production, the antiaircraft gun called Sergeant York or DIVAD, helps explain where six years of heavy military spending have led. DIVAD earned a profit for its manufacturer, Ford Aerospace—under the cancellation agreement the government will pay for 146 DIVADs, although Ford delivered only sixtyfive. But corporate profitability was of course not the explicit reason that DIVAD was built. As Gregg Easterbrook first explained in these pages, DIVAD was an attempt to automate a function that human beings had previously performed: aiming antiaircraft guns at planes as they maneuvered and dodged. DIVAD was costlier than previous systems, it was more delicate and temperamental—and it could not do the job. Radar and computers could not predict a pilot's next maneuver, and they could not even guess as well as human gunners could. Last December a Nicaraguan contra shot down a Soviet Mi8—exactly the kind of superhelicopter DIVAD was designed to confront—with an SA7 missile that probably cost around $10,000 to manufacture. The one American F111 that went down during the raid on Libya was hit by gunfire, not some advanced radarguided system. DIVAD represented "modernization" for its own sake, divorced from the realities of war.
Modernization—an emphasis on more expensive, delicate, and theoretically capable weapons—is the theme of the spendup years, and it explains why our forces have not grown larger and are, on average, getting older. For more than a generation the American arsenal has evolved toward smaller numbers of more costly and delicate pieces of equipment, each of which is in principle able to do far more than the preceding version. The stated reason for this progression is not simply America's technological superiority but also the Soviet Union's numerical edge. Because they have more tanks and planes, we build fewer but more complex tanks and planes—and then we're more outnumbered than before. When American military spending intensified, starting in the late Carter years, the Pentagon decided not to deviate from this pattern but to intensify it, at full speed. Our military planners have staked their future, and ours, on a smaller, more fragile, but only partly modernized force.
This kind of modernization is driven by the habits of a peacetime military, in which weapons can be judged on their theoretical potential, rather than on how they stand up to dirty, dusty wartime conditions when used by confused troops. But it also reflects a deliberate policy choice, based on American's comparative advantage over the Soviet Union, and on the historic importance of technology in war. In one sense modernization is the only thing that matters in combat. Cavalry does not count for much these days, no matter how numerous the horses. Nuclear propulsion allows missile submarines to remain submerged and undetected for months. The newest versions of the Sidewinder missile are more effective than the old ones; advances in engine technology and control systems make newer fighters like the F16 more maneuverable than their predecessors.
But often, as in the case of DIVAD, modernization is confused with complexity and indiscriminate automation, which may or may not make battlefield sense. The incentives of the procurement system tempt soldiers to stop thinking about the uncertainties of combat and to concentrate on getting budgets approved. In its push for more-modern systems—even though they are more expensive and can't be bought in large numbers—the Reagan Administration is betting that all of them represent the right kind of modernization. Aegis cruisers, with elaborate radar and missile systems that can theoretically deter any airborne attack, will be purchased at some $1 billion apiece—even though the system has never been subjected to realistic, seaskimming saturationattack tests. (Nearly all of its tests have been against single attackers.) The F15 has been America's highcost, firstline, modern fighter plane. Its high maximum speed and extensive electronic systems account for much of its $20millionplus unit cost, but the Air Force has said they are necessary if the plane is to prevail in aerial combat. Now some F-15s are being built in a $40 million "Strike Eagle" bomber version, which is so laden with electronic gear and external fuel tanks that it is no longer useful as a fighter, the purpose for which the F-15 was originally designed. The Air Force will buy Lockheed's C5B transport planes, at $270 million apiece, rather than much cheaper altered versions of the Boeing 747. The Army will forge ahead with the modernized Bradley Fighting Vehicle, a "tank supplement," or personnel carrier, that costs as much as a tank. The vehicle it is replacing, the M113, costs about $130,000 (that's right—thousands, not millions), or less than one tenth as much as the Bradley.
By placing our chips on a complex, costly force, the Administration is finally gambling with our survival, if the machines don't work when called on. But this gamble creates problems even in time of peace. More than five years ago a Pentagon analyst named Franklin Spinney said that complex, modernized weaponry had forced the military into a selfaggravating cycle of shrinking forces and decaying readiness. Generation after generation of complex machinery had cost more than predicted and turned out to be unexpectedly hard to maintain. Because costs kept rising, the machinery was bought in smaller quantities and therefore became more expensive still. As costs still kept rising, lastminute cutbacks always had to be made, and they typically came out of the readiness accounts. The result was a shrinking force with chronic readiness problems.
Spinney's findings were derided as "merely historical in nature" by the Reagan Administration, but his predictions about cost growth and readiness cutbacks have been borne out. Weinberger's latest "posture statement," or apologia pro vita sua, is full of laments about "unexpected" budget constraints and "unforeseen" pressure to reduce operating funds. The place to make the cuts, the statement suggests, is in the readiness accounts. After six years of big budgets, readiness never really rose and is about to ratchet down. The Congressional Budget Office recently pointed out that the military is already falling behind in its "war reserve stocks of secondary items," the material that would keep tanks, airplanes, and soldiers going if we went to war. Since 1980 these war reserves have increased by 106 percent—but the "objectives," or requirements, have increased by 118 percent. "These increases in objectives do not result from changing assumptions regarding the scope or length of a future war," the CBO report said. "Rather ... [the] increases stem from the advent of new weapons that require much more expensive spare parts. ... Indeed, this may be one of the less visible ways in which more complex weapons add to DoD costs."
The gamble will pay off if the forces never have to be used. Our main hope is that the Russian military thinks like ours, and is impressed by the spendup.