Caspar Weinberger
Caspar Weinberger; drawing by David Levine

The most important theme of President Reagan’s proposals for defense has little to do with military strategies or concepts, but rather with sheer quantities: the budget stands for more. The administration’s program is the clearest possible expression of a faith that spending more money for defense, without being particular about where or how, will make the nation more secure.

The Reagan military program has so far put forward no new concept of building weapons or organizing soldiers. With the exception of a plan to build a navy that will supposedly be large enough to challenge the Soviet fleet in “high threat environments” off the Soviet shore, it offers no coherent view of how and where force might be applied. Indeed, beyond its proposals for weapons to be built in fiscal year 1982, which begins this fall, the administration has not even revealed just what it intends to spend the money for. It has revealed only that it intends to spend a lot.

In the last budget revisions it prepared before leaving office, which included substantial increases in spending for defense, the Carter administration had proposed to raise the Total Obligational Authority for the military1 by 7.8 percent (after inflation) in fiscal year 1981, and by a further 5.3 percent in 1982. Reagan proposes to increase the budget by 12.4 percent and 14.6 percent, respectively, in the same two years.2 During the following four years, the administration intends to increase the defense budget by at least 7 percent, after inflation, every year, until it reaches $367.5 billion in 1986.3 If these intentions are realized, they will amount to a 50 percent increase in military spending, as measured in constant dollars, within five years.

This plan calls for nothing less than the largest peacetime military buildup in American history; indeed, it exceeds even the cost of America’s most recent war. As Lester Thurow recently pointed out in these pages,4 if the cost of the Vietnam buildup between 1965 and 1970 were converted to 1980 dollars, it would come to about $59 billion. Measured on the same scale, the increase in military expenditures over the next five years will be nearly three times as large. Lyndon Johnson refused to raise taxes to pay the bill for his war; ever since he has shared the blame with OPEC for causing the ruinous American inflation of the 1970s. Ronald Reagan now proposes to finance his much larger expansion while reducing taxes by 30 percent.

The gravest effects of this increase may be felt not on any battlefield but in the rest of the American economy, which needs to find new sources of capital to finance industrial regeneration. But if such concerns are put to one side and the new defense program is judged strictly on its own terms, does it then make sense? If the purpose of Reagan’s military budget is to increase the readiness, flexibility, and usable power of the American military, it must be judged a likely failure, for it threatens to repeat on a more titanic scale a pathological pattern that has dominated American military spending for at least the last twenty years.

That pattern is most succinctly described as the unwillingness to match assumptions to reality—the reluctance to bring to military plans an awareness of the uncertainties the future holds. Those uncertainties may be economic, diplomatic, or purely technical. In building weapons and laying plans that ignore the possibility of the unexpected, the nation has produced a military force that is needlessly vulnerable to changes in the situations in which it prepares to fight, and that has undue difficulty adapting to an enemy’s changes in tactics or plans.5 Nothing in the Reagan proposals attempts to correct this pattern, and several of Reagan’s plans would make it worse. That is why four years of “more” may produce American armed forces even worse suited to the real world than are the ones we now possess.

The outlines of the destructive pattern were suggested by Vice Admiral Earl B. Fowler, Jr., the commander of the Naval Sea Systems Command, in congressional testimony this spring about the Trident submarine.6 The Trident is an extremely large vessel, longer than the Washington Monument is tall. It is powered by a nuclear generator and is designed to carry nuclear missiles for retaliation against the Soviet Union. In the early Seventies, the Nixon administration approved the Tridents, which each carry twenty-four missiles, as eventual replacements for the Polaris and Poseidon submarines, which each carry sixteen. The forty-one Polaris and Poseidon subs we now have will begin reaching their scheduled retirement ages in the first few years of this decade. Nine Tridents have already been authorized, the first of which was to go into service no later than April 1979.

By the time of Admiral Fowler’s testimony in March 1981, the first Trident, the Ohio, had still not been delivered. The purpose of Fowler’s statement was to explain the troubles that had plagued the Trident’s construction at the shipyards of the Electric Boat company, a subsidiary of General Dynamics, in Connecticut. He described first a series of major technical failures. For example, navy inspectors found that Electric Boat had stocked a substandard grade of steel, with inferior crack-resistance and “weldability,” in bar sizes that could have been used in as many as 126,000 places in the Trident.7 In December 1979, the navy discovered several incomplete welds on a different kind of submarine also under construction at Electric Boat; soon afterward, it discovered that the company had no record of ever having inspected tens of thousands of welds on the Trident and other vessels. A vast re-inspection program was begun. Out of 36,149 welds that originally required inspection on the Trident, inspection records were missing for more than one quarter. When the company re-examined such welds as were still accessible, 2,772 of them, one third of those examined, proved to be defective.8


Because of such technical failings and other “unforeseen” occurences, the Tridents under construction at Electric Boat have fallen far behind schedule. Admiral Fowler predicted that the Ohio would not go into service before the end of this year, more than two and a half years behind the contract schedule, and fully four years behind Electric Boat’s original estimate that it could, with its “best effort,” deliver the vessel by December 1977. When the contracts for the project were originally signed, they called for seven Tridents to be at sea by the end of August 1983. According to Admiral Fowler’s latest estimates, the navy will have only two Tridents by that time, and will be one month away from receiving its third. On the average, the eight Tridents under construction at Electric Boat will each be more than two and a half years late.

As work has fallen behind schedule, the cost of the submarines has also soared. When Electric Boat submitted bids for the first Trident, it estimated that construction would require 14.6 million manhours. Since then, its estimate has risen to 23 million manhours, an increase of 60 percent.9 Even before the latest “slippage” (as the navy calls it) in the Trident’s schedule, the estimated cost of the first vessel went up by 35 percent. (From $1.47 billion in 1982 dollars to $2.053 billion.) Because of the slippage, the final cost is sure to be higher still.

The defense industry has known snafus and overruns before, but rarely has their effect been so apparent. The troubles at Electric Boat have undermined the fundamental premises upon which the decision to build Tridents was made. No matter which side one chooses in the metaphysical arguments about nuclear deterrence, there is little dispute about the purpose of nuclear-missile submarines: they are meant to be the least vulnerable nuclear system. Because they would be able to survive any conceivable first strike and deliver a devastating retaliatory blow, they would (in the logic of nuclear “scenarios”) give any aggressor reason to pause before launching an attack. But to do so, they must be at sea, not in a shipyard, and there must be enough of them to ensure that the Soviet Union could not disable the force if, by fluke, it found and destroyed a few.

One argument made by those who opposed the Trident was that it would be dangerous to reduce the number of submarines in the fleet by building such large, expensive vessels; they said this would, in effect, put too many eggs in too few baskets. Those who favored the Trident said that its greater range and operating efficiency would more than make up for the smaller quantity. But the basis of the argument would have changed had both sides realized how long it would take the Tridents to go into operation, and how expensive they would become.

By the fall of 1981, seven Polaris submarines will have been retired, and several more are scheduled to follow. Because of the Trident’s delays, the US now has fewer nuclear missile submarines at sea than at any time since the 1960s; because of rising costs in the project, the navy is rethinking whether it can ever afford the full fleet of fourteen or more Tridents. If the logic of nuclear deterrence means anything (which is a separate question), the choice of the Trident and its subsequent difficulties has at least temporarily diminished American security.10 They have also limited America’s flexibility in nuclear planning; if the government finally decides that the MX missile system is too cumbersome to make sense, it can’t quickly increase production of Tridents as the replacement. These harmful effects have arisen from the system’s vulnerability to uncertainties of cost, construction, and scheduling.

Both the navy and General Dynamics argue that this is merely one unfortunate and isolated case. The Secretary of the Navy said that he would exclude General Dynamics from the next round of submarine contracts and would perhaps even look to overseas shipyards. The company, for its part, claimed that ceaseless “change orders” from Admiral Hyman Rickover made the job impossible, by forcing the company to rip out completed sections and start all over again. On each side the assumption was that none of the problems could have been foreseen.


Indeed, naval shipbuilding has long been a source of special misery for the Pentagon. Yet the implication of several serious studies under review is that “surprises” of this sort are precisely what must be expected in future military programs, all the more so when the US places its emphasis on weapons that are large, expensive, and technologically complex. To fail to prepare for them is to fall into the error of the Trident planners, in which basic strategic and military calculations can be undone by “unexpected” problems in costs, scheduling, and budget constraints.

In The Defense Industry, Jacques Gansler, an experienced defense official, explains why large cost increases can be expected to continue. Gansler is a former deputy assistant secretary of defense for “material acquisition” who has presented a dense but quite valuable anatomy of the collaboration between government and business to produce weapons. One of his central points is that defense contracting is a “thin” industry; for any given military product, there are at best a handful of firms capable of accepting the government’s business. Only one company makes castings for tank hulls, 11 two make aluminum tubing, one makes airframe bearings, two make periscope lenses, two make titanium “skins” for airplane wings, and only one makes the titanium extrusions found in the most complex and expensive planes.12 These companies have little extra capacity for production, and few other firms can surmount the technical and bureaucratic obstacles to entering the field. As a result, when demand goes up, as it has modestly done in the last few years and will dramatically do in the next few, prices go up faster.

Only the energy industry has had a higher rate of inflation than defense contractors. Earlier this year, the Defense Science Board’s special panel on industrial responsiveness tabulated the inflation rates for crucial sub-units of the defense industry. The panel showed not only that they were very high but that they were also increasing. In 1978, the cost of an aircraft part known as the hydraulic actuator was 14 percent higher than in 1977; in 1980, it was 68 percent higher than in 1979. The price of titanium rose by 2 percent in 1978, and by 38 percent in 1980. Hook-up wire for electronic systems rose by 3 percent in 1978 and 126 percent in 1980. A dozen items on the list—including electrical connectors, torque motors, resistors, and heat exchangers—more than doubled in price during 1980 alone.13 All in all, the price of parts and materials for defense contractors rose by 34 percent last year, and defense labor costs went up by 16 percent.14

Yet it is this same industry—already overstrained, about to cope with unprecedented levels of demand—in which the administration is assuming that inflation will virtually disappear in the next few years. The administration’s projections assume that inflation in the defense industries will be about 7 percent in 1982, and will decline steadily to 5 percent by 1986. The General Accounting Office has estimated that inflation among defense contractors will average 13.3 percent through the next six years. The likelihood of unhappy “surprise” is great.

Unexpected troubles in scheduling are likely as well. Like prices, production delays in the defense industry are bad and getting worse. In 1977 the “lead time”—or gap between the placement of an order and the production of a finished item—for the airframe for a military airplane was 95 weeks. By 1980, it had become 199 weeks. For an item known as a speed brake actuator, the lead time rose from 43 weeks in 1977 to 112 weeks in 1980. Titanium supplies, used extensively in high-speed aircraft, went from a 40-week lead time in 1976 to 104 weeks in 1980. The integrated circuits that are so profligately used in sophisticated aircraft suffered a 25-week delay in 1978 and 62 weeks in 1980. And so on down a long list.15 What is more important than the absolute length of the delays is that, as with the Trident, they are often longer than predicted. In 1977, the Pentagon predicted that the lead time for F-16 fighter planes would be 28 months; it is now closer to 42 months. The Defense Science Board’s report contains chart after chart revealing a similar pattern.

A third kind of uncertainty exaggerates the effects of the other two: the amount of money that will be available to the military to pay for new equipment. On its face, this would seem to be the variable that the administration would be in the best position to control. Compared with the soaring costs of such programs as food stamps and Social Security, the sustained 7 percent annual growth rate for defense that the Reagan budget projects may seem easily attainable. But one Pentagon official, Franklin C. Spinney, has presented evidence that points toward the opposite conclusion. To assume that the military will get that much money, Spinney says, is to assume away the constraints that have applied for thirty years.

Spinney’s findings are contained in a study called Defense Facts of Life, which has been in gestation for several years in the Program Analysis and Evaluation division of the Pentagon and, through the urging of Senator Sam Nunn, was finally made public early this spring. Spinney’s findings about the defects of complex weapons—the fighter planes that rarely fly, the missiles that rarely work—have been picked up in the press, but not his careful historical analysis of how military planners chronically deceive themselves about the resources at their disposal.

Spinney’s approach was to prepare a thirty-year history of American military budgets, beginning in the early days of the Korean War and ending in 1980. The first point such a review makes clear is that defense spending has been remarkably constant throughout those years. When all figures have been adjusted to remove the effects of inflation, a graph of military budgets from 1950 to 1980 shows a series of zigs and zags around an essentially constant level. Measured in 1981 dollars, the United States has spent roughly $150 billion on the military every year. There was a rise for the Korean War, another rise for Vietnam—but always corresponding falls. In fact, Spinney says that during this period, defense spending has never risen for more than three consecutive years.

Spinney then calculated the relative probabilities of increases and decreases in the defense budget in any given year. During more than half of the thirty years he reviewed, the defense budget declined, in real terms, from the level of the year before. For three quarters of the years, it has either declined or has risen by less than 5 percent; in only three years out of thirty has it risen by as much as 7 percent above the previous year’s level. The Reagan projections, once again, are that the budget will rise by 14 and 12 percent in the first two years, and by 7 percent in each of the next four. To extrapolate from Spinney’s findings, the statistical probability of that happening is less than one in a million. The actual chances are probably less.

The competition for capital from the rest of the economy will be more acute during the early 1980s than during the 1950s and 1960s. While the invasion of Afghanistan and the troubles of Poland may have increased the public’s willingness to pay more for defense, there were similar stimuli—the Korean War, the invasion of Hungary, the “missile gap,” and the Cuban missile crisis, to name a few—during the period covered by Spinney’s study. Even the administration’s own political tacticians have essentially conceded that the defense budget cannot grow as now planned. In return for their votes in favor of cutting taxes for 1983 and 1984, congressmen are asking for details of where federal spending can be reduced in those years. The budget director, David Stockman, has said privately that before long the cuts must extend to defense.

What this means is that the Reagan administration’s five-year plans for defense will have the same illusory quality that has been typical of military budgets over the last few decades. The Five Year Defense Plan (known as FYDP), with its long-range predictions for spending, was introduced in the McNamara era as a way of imposing rational analysis on defense spending. In fact, it has become an instrument of self-deception, since the only real decisions are those for the first year, while the projections for years two through five represent sheer hope. Jacques Gansler gives this example: In its FYDP released in 1973, the navy predicted that it would be able to build five nuclear attack submarines in 1978. By 1975, the navy was predicting two. When the budget for 1978 was released, it authorized one submarine.16

What is the significance of such uncertainties in cost, schedule, and available budget? Each involves a question of quantity, but together these quantities can have a significant effect on quality.

In its decisions about military hardware, the US often faces the choice between the complex and the simple path. The complex alternatives emphasize rarefied technical developments—powerful radar and computer systems in aircraft, sophisticated sensors in guided missiles, high-acceleration engines in tanks—which are inherently more expensive and less reliable than simpler systems. Not only are their purchase and maintenance costs higher in absolute terms, but also, as Spinney shows, costs of complex systems are more likely to be substantially worse than anyone expected.

Much of the choice between the different approaches depends on predictions—how much each weapon will cost, how long it will take to procure it, how many of the weapons the US can afford to buy and maintain. (This is not even to mention predictions about how well the weapons will work, which is an equally tangled question.)17 If the assumptions turn out to be over-optimistic, as they have time and time again, the Pentagon often chooses needlessly complex systems that leave it with a smaller force than expected (because the weapons cost more); with fewer weapons in the field (because of long lead times); and with less money available to maintain the weapons, train crews, or stock spare parts (because the “unexpected” budget cuts often come first from operating funds); and with no way to minimize the damage by changing course quickly (because of the tremendous costs and lead times involved). The Trident exemplifies all of these problems. The military is left, that is, with a force that few would have chosen had they allowed for unforeseen consequences.

In the climactic part of his analysis, Spinney explains how these systematic mis-predictions create a degenerative cycle at the Pentagon. Based on its overestimates of how much money is available, the military makes the initial payments for costly, complex systems (e.g., Tridents, nuclear aircraft carriers, “smart” missiles). Based on its underestimates of the cost of new weapons and the inevitable shrinkage of the budget, the military ends up with fewer of the weapons than it originally hoped for (e.g., any of modern weapon systems). Based on the surprising escalation of maintenance costs for complex systems, the military must move even more of its money from the “investment” account to the “operations” account, and therefore can afford to buy even fewer weapons. And based on their inability to learn from this experience, the planners prepare yet another FYDP, which assumes yet another increase in funding, and which is used to justify yet another complex system.

The best way to tell whether an administration has succumbed to this pattern is to see whether the increases it proposes for buying new weapons are matched by the increases it allows for operations and maintenance. (To use a homely comparison, this is like asking whether a family that has budgeted money to buy a horse has also budgeted to feed it.) Between 1981 and 1982, the Reagan administration plans to increase the investment budget by 35 percent, largely to pay for costly, complex systems, such as the nuclear aircraft carrier, the MX missile, a bomber like the B-1, and a number of new electronic systems. But it predicts that its maintenance costs need only rise by 4.2 percent.18 It will be “surprised” by anything more—and because of the rigidities created by large, slow, lumpy investments, it will not be able to improvise its way out of trouble when the surprises occur.

What has been forgotten in this program, which purports to add so greatly to American strength? Apart from questions of appropriate levels of military technology or more realistic planning systems, at least three crucial elements of American security in the next few years have been largely ignored.

The first concerns men, not machinery—the incentives that convince soldiers to face risks and obey orders, the bonds that connect a military force to the nation it is meant to defend. Throughout the Reagan program, the assumption is that these are purely economic issues: if pay is raised for the volunteer force, all questions about the character of the officer corps, the morale of the enlisted force, and the social representatives of the army will be moot. It is hard to find anyone who has studied the history of combat, or who has commanded troops, who would agree. Writers (and former soldiers) such as William Hauser,19 Richard Gabriel, and Paul Savage,20 and the anonymous army officer known as “Cincinnatus”21 have all emphasized that the character and instincts of military leaders overshadow differences in machinery in determining success or failure in combat. Charles Moskos of Northwestern, among others, has explained why recent attempts to make military work resemble civilian life, as in today’s volunteer army, defy all historical evidence about the kinds of loyalty and shared sacrifice that are essential in a fighting force.22 These are questions of spirit, motivation, ideals; the administration proposes to answer them by raising military pay by 14 percent.

The second issue is the military protection of energy sources, and here the main question concerns the balance among different means. Part of the administration’s justification for large increases in naval construction is the need to “project power” in troubled regions such as the Middle East—to land troops if necessary, to sustain them through long combat, to forestall the need for warfare through sheer intimidating presence. Nearly everyone who has studied the Rapid Deployment Force (RDF) organized by the Carter administration agrees that it needs strengthening, both in the plans for its use and its equipment,23 if it is to be at all plausible as a threat. It is difficult to predict the future of the Middle East with such confidence as to deny the potential need for a more effective RDF. But if a government defines its goal, as this one often seems to, as being able to guarantee through military means that the oil will not be stopped, then there is no limit to the corresponding cost. No matter how many carriers or tanks might be available, military tacticians are pessimistic about the possibilities of keeping all the oil flowing if local groups, especially terrorists or guerrillas, were intent on cutting it off.24

A more reasonable approach would be to match improvements in the RDF with efforts that are equally urgent, if not more so, to reduce dependence on imported oil. The administration has so far put most of its emphasis on the first half of that combination, because of the two ideologies on which it relies. One is reassertion of American power overseas, which makes it eager to build a larger military force. The other is dependence on the market, which makes Reagan resist any steps toward energy independence except raising the price of gas and oil. Using several billion dollars to fill the strategic petroleum reserve rather than to build an aircraft carrier would offend the first belief, and imposing a tax on gasoline would offend the second. But if American security is really the issue, these measures deserve as much attention as anything else in the defense budget.

The third issue is controlling nuclear weapons. Reagan argues that the way to limit nuclear weapons is to convince the Soviet Union that we are intent on building them. Such a view lies behind the MX missile project, the proposals for a new nuclear bomber, and especially the reports, which appear with increasing degrees of official sanction, that the US is considering building the antiballistic missile (ABM) systems that were prohibited by the first Strategic Arms Limitation treaty. The books by Desmond Ball and Gerard Smith, read together, suggest the way such a calculation can miscarry.

Ball’s subject is the rapid increase in American nuclear forces during the Kennedy administration. This was by far the most dramatic expansion in American nuclear weapons; by 1967, Robert McNamara said that the resulting nuclear arsenal was “both greater than we had originally planned and in fact more than we require.”25 The momentum to sustain such a buildup, Ball says, came from bureaucratic politics, from pure technical ambition, from deep symbolic feelings about “weakness” and “strength”—from many things, in short, other than a rational assessment of American security or of the steps most likely to add stability to nuclear deterrence.

In Doubletalk, Gerard Smith describes a period six to eight years later, after the Russians had significantly expanded their own nuclear forces. A similar pattern emerges. The same combination of pressures—bureaucratic, technical, symbolic—lay behind two nuclear innovations that emerged at the end of the 1960s: antiballistic missile systems, designed to destroy enemy warheads as they descended toward their targets, and multiple independently targetable reentry vehicles, now widely known as MIRVs, which gave each rocket the theoretical ability to destroy not just one but as many as ten or even fourteen targets.

The SALT negotiators succeeded in agreeing to prohibit further ABM systems. The ABM had always been a dubious proposition practically. From the point of view of strategic theory, moreover, it threatened to destroy nuclear deterrence by encouraging sneak attacks.26 In the other case, the negotiators failed. The US went into the talks with a head start in MIRV technology, and it was not eager to trade that advantage away. The meetings ended with no agreement on limiting MIRVs; ten years later, the enormous increase in Soviet MIRVs is the principal worry of American nuclear planners. What began as a “sweet” technical achievement, and a demonstration that we were strong, ended as a threat to the stability of deterrence.

There is a clear parallel to our current situation. When all the detailed arguments over nuclear scenarios are finished, it is generally admitted that the pressure to build new systems now, before negotiating, arises from the desire to create symbols. Few serious people will claim that the MX missile or the ABM is necessary for the physical security of the United States. Their most sophisticated proponents speak instead of “perceptions”: the systems must be built to demonstrate to our enemies, to our allies, and (probably most important, but this is rarely admitted) to ourselves that we have the will to prevail. Smith and Ball remind us that, while the symbolic aspects of nuclear armaments are important, there are certain technical barriers that, when breached, are difficult to repair. The ban on ABMs represents such a barrier; the current indifference to the perils that lie beyond it is one more illustration of the fundamental danger posed by the Reagan administration: its inability to distinguish between “more” and “better” defense.

This Issue

May 28, 1981