Arms Wrestling

What to buy and why


Flexible Defense
By James S. Robbins
The United States' strategic goals during the Cold War were reactive; the containment of communism was the primary goal. But the retreat of communism allows the United States to pursue more-active goals, reflected in three basic strategic priorities:

1) Maintaining stability by making aggressive war difficult for states to pursue profitably;

2) Preserving the freedom of commerce through open sea lanes, open skies, and open borders where possible; and

3) Preserving and extending liberal democracy and opposing the extension of political repression.

An underlying assumption of these goals is that liberal states, of varying complexions, are generally peaceful and don't pursue aggressive military policies. The people of such states recognize that commerce, not coercion, is a preferable method of intercourse. Were the international system composed only of liberal states, armed conflict wouldn't be a significant issue. However, not all states are liberal democracies, and most don't share classical-liberal assumptions, values, or principles. It is the illiberal states which pose future threats to the United States and to the free world and which must be deterred or fought and, in time liberalized.

The United States should not intervene militarily in all conflicts. Nor should it rule out intervention in any conflict. Rather, we should judge intervention by U.S. interests. If American interests are threatened directly, through a possible armed attack on the American homeland, threat to citizens or property abroad, or similar circumstances, intervention (proportional to the threat) is mandatory. If the threat is indirect, the United States may (not must) intervene if:

1) U.S. action will promote the cause of freedom; and

2) U.S. action is militarily and politically feasible.

The feasibility test is an important check. For example, the United States won't invade China to impose democracy primarily because it has no chance of succeeding. On the other hand, the United States shouldn't intervene in a country even when militarily feasible with the intention of overthrowing its government simply because it is an illiberal state. While such action may be militarily feasible, in most cases it would not meet with international approval. Furthermore, there is more to democratization than simply overthrowing a dictatorship, and rash action by the United States could do more harm than good in the long run. The United States isn't committed to freeing every oppressed people, certainly not unilaterally.

The United States is now the world's dominant military power. This isn't because American force levels have increased; in fact, they have decreased since 1989. But the collapse of the Soviet Union has shifted the correlation of forces toward the United States. Because of this, the United States can reduce the size of its military establishment while still maintaining its dominant position.

Disarmament is only practical to a point. The Soviet Union is no longer a threat, but other dangers are emerging, such as the growth of regional powers in the developing world, increased problems of nationalism and ethnic strife, the proliferation of missile technology and weapons of mass destruction, and the continuing inclination of dictators to attack their neighbors. Radical disarmament may create an environment in which such threats develop more quickly. Our strategy should therefore look not only to imminent dangers but to preventive action against potentially dangerous future developments.

The Cold War force structure was centered on fighting the Soviet Union, primarily in Europe. Future forces will have to be designed to meet less predictable, more wide-ranging situations. One important role will be conventional deterrence and demonstration. The United States should make it clear that its armed forces will be used if necessary to pursue strategic goals. This will reduce attacks against America and its allies and citizens abroad.

The primary arm for this mission would be, as it has traditionally been, the Navy. A large, active, and visible naval presence worldwide will serve as an important check on the aspirations of leaders who are contemplating threats against the United States, its interests, or allies. Carrier battle groups would continue to be the centerpiece of the Navy, with a force of nine carrier battle groups and naval air wings.

If deterrence fails and conflict ensues, the United States should have the necessary force available to prevail against an adversary state. Naturally, the specific use of force will depend on the particulars of the crisis, but because many types of conflict are possible, U.S. forces should be flexible. With fewer forward bases and the possibility of future threats arising unexpectedly in remote regions, fast-attack forces will be an important component of the armed force.

Naval assets would probably be first on the scene and may be the only forces necessary if the conflict is contained. Three Marine divisions should be retained to serve as the inland element, using high-technology weapons and naval air support to make up for their relative lack of firepower. The Air Force would play a parallel though diminished role, since ground bases won't always be available.

The Army will play a less important part in overall defense, with the reduction of most of its NATO complement. Some have suggested that the Army should also be "lightened," that heavy armored forces are anachronistic and should be abandoned. This would be a grave error. If the United States must use force abroad, it must have the capability to use offensive force, to give theater commanders options to pursue whatever manner of military action they see fit. Operation Desert Storm offers a good example of the results that can be achieved by heavy forces.

Lack of armor would limit our troops to defensive roles, blunting the attacks of adversary heavy forces but unable to pursue decisive victory. The Army could be scaled back to 10 divisions at various states of readiness: three armored, four mechanized infantry, one cavalry, one airborne, and one air mobile. The smaller elite forces of all branches would be retained for special operations.

To maintain our current technological advantage, the United States should continue research and development on the next generation of "smart" weapons. As the Gulf War demonstrated, high-technology weapons and defensive systems mean fewer allied casualties. Smart weapons also cause fewer civilian casualties, which isn't a military necessity but is a value the United States should pursue. Civilians in illiberal states ought not to sacrifice their lives to pay for the sins of their juntas. Furthermore, the sheer margin of technological superiority could have a deterrent effect on potential aggressors, especially if the United States has the ability to target the entire enemy command structure (often including the leaders themselves).

The United States should continue to work for the elimination of nuclear, chemical, and biological weapons worldwide. Increasingly powerful conventional explosives and more-accurate delivery systems are rendering nuclear weapons obsolete for most strategic purposes. A small nuclear force could be maintained for purposes of deterrence against recalcitrant countries, but complete multilateral denuclearization would be preferable. Nuclear forces—submarines, bombers, and missiles—would be partly eliminated and partly converted to conventional roles.

As missile technology proliferates, the threat of purposeful or accidental launch against the United States increases correspondingly. Therefore, the United States should pursue political and diplomatic measures to slow the spread of these weapons, but it should also continue efforts to build some form of ballistic-missile-defense system. Because of the unlikelihood of a massed ICBM attack, a simple and inexpensive system should offer adequate protection.

The current system of military alliances requires substantial review, especially regarding costly force deployments overseas against threats that no longer exist. But the alliance structures themselves should be continued as consultative bodies and frameworks for cooperative action. Europe is an important forward base for U.S. forces, and naval bases in Europe will continue to service our fleets in the Atlantic and Mediterranean. NATO also represents a significant political link between Europe and the United States, which will become more important as Europe moves toward political unity.

Other alliances serve similar ends. By maintaining bases and predeploying materiel in regions worldwide, the United States will be better able to react to future crises.

The wars of the 20th century were large, violent, and total. The decline of ideological conflicts will usher in an era of less violent, limited warfare. U.S. defense strategy should take this into account but the United States should continue its diplomatic and political involvement on the world stage. A return to isolationism in any form will bring exactly what it brought the last time we tried it: conflict. It is in our interest to maintain American involvement in world affairs now to prevent being forced into large-scale wars later. It is an investment we cannot afford not to make.

James S. Robbins is assistant professor of diplomacy at the Fletcher School of Law and Diplomacy.

Beyond Intervention
By Ted Galen Carpenter
Dramatic changes in the international political and military environment create both the opportunity and the need for the United States to adopt a new security strategy. For the first time in more than 50 years, we don't confront a powerful state, such as Nazi Germany or the Soviet Union, that could pose a grave threat to America's security, nor is a new threat of that magnitude likely to emerge in the foreseeable future.

The absence of a would-be global hegemon is a crucial development that should fundamentally alter U.S. defense policy. Although the post–Cold War world may sometimes be a disorderly place, without a powerful rival to exploit the turmoil, most conflicts will be parochial. The United States can, therefore, afford to view them with detachment, intervening only as a balancer of last resort in the unlikely event that a conflict cannot be contained by powers in the affected region and is expanding to the point that it jeopardizes America's own security.

The American people should repudiate efforts to preserve a global-interventionist role for the United States and should instead insist on a policy of strategic independence. That policy would avoid military commitments unless there was a serious threat to America's own vital security interests, which would be defined quite narrowly. A vital interest ought to have a direct, immediate, and substantial connection to America's physical survival, political independence, or domestic liberties.

Moreover, those who argue that an interventionist enterprise is necessary should bear the burden of proof. Their policy prescription would automatically entail tangible and immediate costs and risks to the American people as opposed to the hypothetical costs and risks of inaction. They should have to present compelling reasons for us to adopt their position.

Strategic independence would in no way constitute isolationism. Not only should the United States maintain extensive diplomatic, cultural, and economic relations with the rest of the world, but it must have capable military forces and be prepared to take decisive action if a serious threat to American security does emerge. A restrained defense strategy doesn't mean that U.S. policy makers must wait until bombs are falling on American cities to respond.

The most dangerous threat to American security in the post–Cold war era will probably be the proliferation of nuclear weapons. Not only does proliferation appear to be accelerating, but the regimes that seem most determined to acquire nuclear arsenals are precisely the ones that the United States would least like to have them—the Irans, Libyas, and North Koreas of the world.

That situation has important defense-policy implications for the United States. First, it means that Washington must be doubly cautious about intervening in regional disputes that don't have a clear and compelling relevance to America's security. The only thing worse than gratuitous meddling in parochial conflicts when the parties are armed with conventional weapons is meddling when one or more of the parties has nuclear weapons.

Second, the United States must maintain a credible strategic arsenal—approximately 2,000 weapons deployed on submarines and long-range bombers to ensure the survivability of a retaliatory force—as a deterrent. Finally, the dangers posed by proliferation greatly increase the need for an effective shield against ballistic missiles. Indeed, the ABM program is the one area in which U.S. military spending should be modestly increased during the 1990s.

America's post–Cold War defense forces should be geared to neutralize threats to American security, not to intervene in regional conflicts or to protect Cold War–era allies that are now capable of defending themselves. This means concentrating resources on long-range retaliatory forces and strategic defenses.

In addition to preserving a "strategic dyad" of eight or nine Trident missile submarines and a bomber fleet (50 to 60 B-1Bs and B-2s), the United States should maintain a 250-ship Navy (including six carrier battle groups) and 16 tactical air wings (nine Air Force, two Marine, and five Navy). Ground units (both Army and Marine) are the least relevant arm of the military in a post–Cold War setting and should be pared to an active-duty force of 250,000.

Most important, over the next five years, the United States should withdraw and demobilize the forces stationed in Western Europe, Japan, and South Korea. With the disintegration of the Soviet Union, it is absurd to protect those nations from secondary threats that they should be able to handle with their own military resources. The Bush administration contemplates keeping 150,000 troops in Europe indefinitely—plus reinforcements stationed in the United States—at an annual cost to American taxpayers of more than $90 billion.

Washington's security commitments in East Asia (primarily to Japan and South Korea) cost another $40 billion annually. Yet Japan has the world's second-largest economy and can easily afford to spend more on defense, if necessary. South Korea has twice the population of and an economy 10 times as large as its only conceivable adversary, North Korea.

If the United States jettisons its obsolete alliances and adopts a force structure appropriate to the conditions of the post–Cold War world, it can protect the republic's security with an active-duty force of approximately 875,000 persons at a cost (after a five-year transition period) of $125 billion a year (measured in 1992 dollars). That sum is barely half of the Bush administration's projected 1997 budget.

Even at $125 billion, the military expenditures of the United States would still be three to four times larger than the probable 1997 defense budget of any other member of the Group of Seven (G-7) industrial powers. The cuts may seem radical in the context of the elephantine Cold War–era budgets that have come to be considered "normal," but with the end of the Cold War, we must change our ideas of what constitutes normal defense spending.

Beyond the issue of cost, there is an important moral consideration. The lives, freedoms, and financial resources of the American people are not—or at least, should not be—available for whatever foreign-policy objectives suit the whims of the national-security bureaucracy. The U.S. government has a fiduciary responsibility to protect the security and liberty of the American republic, not to pursue grandiose schemes of world leadership. It doesn't have either a moral or constitutional writ to implement the political elite's activist agenda internationally any more than it has one to do so domestically. Those who are serious about the principles of limited government cannot in good conscience support a global-interventionist foreign policy.

Ted Galen Carpenter is director of foreign-policy studies at the Cato Institute.

Haste Makes Waste
By Murray Weidenbaum
Events since 1989, at home and abroad, reinforce the need for a "new look" by the United States in setting national-defense strategy. The most sensible approach is a step-by-step arms reduction closely keyed to firm evidence of continuing ease in world tensions. Hasty defense cutbacks not geared to the changing threats to our national security are foolish. They could place the United States in a needlessly vulnerable position. The history of the past half century shows frequent and wasteful start-and-stop cycles of defense spending.

It makes good sense to cooperate with the ex-Soviet republics to cut back on the size of both military establishments. Yet the United States should be prepared to reverse course if a new leadership in Russia takes a more aggressive position or if serious strategic threats arise elsewhere. Effective reversibility thus requires a strong and diversified defense industrial base as well as alert and well-trained reserve components.

The most reasonable basis on which to plan our own national security is to expect continued and substantial reductions in U.S. military spending over the coming five years, if not longer. The cutbacks will be more than the very modest 1-percent to 2-percent annual declines of the late 1980s. We can plan on defense reductions of 5 percent a year or more. Another way of looking at it is that a high level of military spending will continue for the indefinite future, but it will be much lower than the peak rates of the 1980s.

This substantial, but curtailed, level of defense outlay necessitates a variety of adjustments, many of them painful to the people directly involved. Yet the general magnitude of the change will be less (as a share of GNP) than was the case following the end of the Korean War or the Vietnam War.

To fit within fiscal reality, Congress and the president have to take those difficult actions that the government has tended to avoid since the mid-1980s—substantially reducing the number of aircraft, missiles, and ships the Defense Department buys and the number of people in uniform. The mismatch between the military's wishes and congressional appropriations can only be resolved in one way—by cutting the planned spending to fit the budget cloth.

Coming to grips with the budgetary challenge will also reduce the great uncertainty that currently hangs over planning in the defense industry. That uncertainty adversely affects defense businesses and their employees. Some of the cuts that should be made are obvious. It is only bureaucratic inertia that has blocked them. For example, the Navy still plans to spend $1 billion to build and outfit a series of additional "home ports" originally designed to support a 600-ship fleet.

The current fleet is less than 500 ships, and further reductions are clearly in the cards. The support for continuing expensive and needless construction comes from the cities where the new ports are scheduled to be built. We must be on our guard against born-again military enthusiasts. Local interest groups find it easy to confuse pork with patriotism. Do you want to convert a congressional dove to a hawk? Easy; just try canceling a defense contract in his or her district.

The national interest requires a basic downsizing of the three military services. The end of the Soviet threat to Western Europe means a smaller Army. The reserves should be assigned the major portion of the task of defending our interests in Europe in the event of a future crisis.

The Air Force can also be cut because the United States will play a substantially smaller role in NATO. Nevertheless, we will still need a strong Air Force. Intercontinental ballistic missiles are still very much operational all throughout the region formerly controlled by the Soviet Union. And no one knows who will wind up controlling them.

Finally, the Navy should shift more of its vessels to reserve status and also reduce the number of aircraft carriers. Nevertheless, America needs a substantial tactical force of Navy and Army units to deal with potential crises in the developing world.

This smaller U.S. military force should be based mainly in the continental United States. It would require much less infrastructure support than the current armed forces, although airlift and sealift capability would be important considerations. Moreover, the quality and morale of military personnel must be maintained at high levels despite the reduction in their overall size and budgets. That takes pay and perks at levels competitive with the civilian economy. Visions of peace dividends shouldn't obscure the need to maintain an adequate corps of professionals in the armed forces and key reserve units.

In voting lower appropriations for defense, Congress should avoid setting in motion a new stop-and-go cycle in military spending. Serious threats to national security are changing in form, but they surely continue. Possibilities range from the Middle East to the nuclear buildup of North Korea. The effectiveness of the large amounts of money and resources devoted to defense will be enhanced by lowering the peaks in the military budget and raising the valleys.

None of this is reason to stop the movement to a lower level of defense spending. The prosperity of the United States doesn't require any particular amount of military activity. In fact, the productivity and competitiveness of the American economy will suffer if the government uses defense spending to prop up the prosperity of any region or industry.

Further, unemployed defense workers shouldn't receive special benefits. They should be treated as generously as—but not more or less than—say, people who lose their jobs because of sluggish housing sales resulting from a change in the government's monetary policy.

The scientists, engineers, and technicians that constitute a large fraction of the defense industry don't need federal "make work" programs. They are among the most mobile members of the labor force—geographically, industrially, and occupationally. While defense is a national problem, economic adjustment is largely local. In any event, the private sector is better equipped to deal with adjustment to defense cutbacks.

Many defense-oriented communities take the position that the nation owes them something special because of their "contributions" to national defense. But we must dismiss such obviously self-serving views. Just recall the vigorous lobbying efforts those same communities made to get the Pentagon to locate the defense contract in their locality in the first place.

However, cuts in defense spending should be accompanied by a massive dose of deregulation in the military procurement process. Congress should strip out the host of special provisions that require military contractors to act like government bureaus doling out benefits to designated classes of beneficiaries. That would reduce the overhead costs of defense contractors. It would also increase their ability to transfer new technology between civilian and military products.

Murray Weidenbaum is director of the Center for the Study of American Business at Washington University in St. Louis and the author of Small Wars, Big Defense (Oxford University Press).

Invest in Higher Machine IQ
By Bart Kosko
Most money for weapons goes to develop them. Much less money goes to research them and draw their blueprints. In fiscal 1990, Congress spent $81 billion on weapons procurement, $88 billion on "operation and maintenance," and $75 billion on military personnel. It spent $37 billion on all R&D and testing and most of that on development and testing. In contrast, Congress spent $12 billion on NASA and $2 billion on the National Science Foundation.

This suggests a rule of thumb for massive R&D cuts: Up the R a little, slash the D a lot. In 1991, Defense Secretary Dick Cheney called for stopping the development cycle at the prototype level. That is the right idea and the right first step. As long as the United States retains the capability to produce state-of-the-art weapons, it's not necessary to actually manufacture them unless new threats to our security emerge. Cutting production will mean the end of large aerospace firms, at least those that get 50 percent or more of their revenue from defense contracts.

A new market could soon form. There are and will be hundreds of thousands of laid-off aerospace workers and researchers who can consult, form small firms, and start think tanks to supply the new demand for defense research. A report by the Los Angeles County Aerospace Task Force claimed that by 1995 L.A. County alone may lose up to 420,000 aerospace jobs. The new defense-research market can help the United States keep its lead in the design of smarter weapons.

There is now a reason to favor research over development: Machine IQs are on the rise. Emerging technology could make U.S. weapons obsolete unless we keep up. We are seeing smarter cars, smarter satellites, even smarter washing machines. Part of this IQ boost comes from advances and shrinkage in sensors and computer chips and materials. Defense firms and funding agencies will no doubt support these areas. But most of the IQ boost comes from an area that most people have not heard of. This area is a new branch of machine intelligence that looks at the world from a view called model freedom.To grasp model freedom it helps first to see what it is not.

Science paints a picture of the world with math models. Each model uses equations to turn inputs into outputs. Newton's law of gravity is a simple math model. It takes as inputs the mass of two objects and their distance and gives as output their gravitational attraction. The models of econometrics use many equations to tie outputs to inputs. New data on interest rates and GNP and employment and the money supply tune the equations.

The trouble is that someone has to guess at the equations. Newton guessed well, and Einstein guessed better. In most cases, we have no idea what the equations look like. The world is too nonlinear. We don't know the math to control a helicopter in flight when it loses a rotor blade or to tell a mine in the sea from a rock or even to back up a truck and trailer to a loading dock. The new techniques of neural networks and fuzzy logic can perform these tasks. Like humans they work with no math model.

Model freedom lets the data tell their own story. It lets systems learn from experience. A neural net starts out as a black box with a tangled web of synapses inside. You show it a stream of inputs and outputs and the web changes. In time, when you give it an input it gives the right output. You show it a sonar picture of a rock and it says a rock and not a sea mine. It learns some complex math model in its tangled web of synapses but no one knows what it is or cares.

A fuzzy system works the same way but it learns, or you give it, a set of common-sense rules: If the water is dirty, then add more detergent. If the trailer turns to the right a little, then turn the steering wheel to the left a little. If the scud speeds up a lot, then move the cross hairs to the right a lot.

In theory, these model-free systems can model anything. So in theory neural nets and fuzzy systems can learn any relation between inputs and outputs, between cause and effect. In practice, we are still working on it. Smart fuzzy appliances and decision aids in Japan, which owns this market, came to over $2 billion in 1991. In the United States, Motorola, General Electric, and Rockwell have started large programs in smart fuzzy products. NASA has held workshops on neural nets and fuzzy logic and worked them into new designs for space shuttles and Mars rovers.

Defense firms and agencies have been slow to work with the model-free strains of machine intelligence. Cruise missiles and fighter aircraft still use 1960s control and scene-matching software. In some cases, you can prove theorems about how these simple math models work. The new model-free black boxes work well but do not leave much of an audit trail of how they work. They have no math model and that means no math guarantees. You have to trust simulations.

That and their newness has not set well with many defense agencies. But it has not stopped Sony from building fuzzy TVs or stopped Honda from putting fuzzy transmissions and brakes and fuel ignitions in its new cars. And future weapons systems—many with high-speed swarms of smart missiles and robocraft—will defy any man's math model. The future is model free whether generals feel comfortable with it or not.

Investing in higher machine IQs has a better chance that it will lead to commercial spinoffs than does investing in the later stages of weapons R&D. First, the commercial market has adapted to much of the new smart technologies. It stands ready to use the next advances. There is little market for stealth and other technologies grown in the defense hothouse.

Second, the problems of smart weapons are much like those of smart products. They come down to how we sense data and how we use them to control actions at the cognitive level. If you can tell friend from foe by a laser radar pulse, you might use the same pattern-recognition net to tell cracked bottle caps from good ones or to tell a safe loan from a risky one or to tell a healthy cell from a cancerous cell. If you can control a swarm of cruise missiles as they talk and hit targets and get hit and reassign themselves, you might use the same fuzzy-logic rules to schedule tasks in a supercomputer or to guide a platoon of small cars as the cars cruise down a freeway and join or exit the platoon.

Few weapons programs favor simulation over production, software over hardware. One exception is the little-known Air Defense Initiative or "Air Wars." This program started in the late 1980s in the shadow of the Strategic Defense Initiative. SDI tried to put the roof on a house before the sides were up. Even the best Star Wars shield could not stop a cruise missile or swarm of them lobbed at the White House from a close submarine or bomber. ADI deals with these "air-breathing" threats of the future, the sort of pinpoint strikes a future Azerbaijan or Mexico might launch against its neighbor. The Air Force wants to fund an ADI simulation testbed to test new ADI weapons systems and organization schemes in "virtual" battles. So far, the giant SDI program has taken most of the ADI monies.

Today the Defense Advanced Research Projects Agency funds most of the research on smart weapons. DARPA has favored mainstream computer and "artificial intelligence" approaches over the new model-free systems. From the 1960s to the 1980s DARPA kept the new A.I. field afloat and kept that infant industry an infant. Hundreds of millions of dollars led to a few A.I. breakthroughs and no A.I. commercial products that you can find on the shelf or in the office or in the kitchen.

One idea is to start a new civilian DARPA to nurture new smart technologies and to compete with the Cold War DARPA. A lesser-cost idea is to strip DARPA of its top-secret projects and put it under civilian control. That might make it act more like Japan's Ministry of International Trade and Industry. A still lesser-cost idea is to abolish DARPA outright.

Bart Kosko is an assistant professor of electrical engineering at the University of Southern California and author of the Prentice-Hall textbooks Neural Networks and Fuzzy Systems and Neural Networks for Signal Processing.

Defense Trust Busting
By Benjamin Zycher
The collapse of the Soviet Union and of state socialism in Eastern Europe signals the end of the great 20th-century struggle for hegemony in Europe, a contest that had its origins in the German quest for power that grew in the late 19th century, culminating in the First and Second World Wars. That in the end this hegemony will be enjoyed by a Germany defeated in both blood baths is ironic but hardly affects the vital interests of the United States.

In the narrower context of U.S. defense planning and organization, the end of the Cold War provides a rare opportunity to restructure substantially the U.S. defense establishment, which more accurately should be termed the Socialist Defense Cartel. This reflects one of the great missed opportunities of the Reagan administration—the failure to use the large Pentagon budget increases to offset inevitable bureaucratic opposition to reform of the defense cartel.

Of more central interest are the appropriate goals of the U.S. defense establishment in a post–Cold War world. These are: preservation of a core defense program as an insurance policy; development and deployment of missile defense; expansion of a vigorous antiterrorism capability; deployment of a substantial preemption capability; and preservation of a highly credible and stabilizing second-strike nuclear capability.

Core Defense Insurance. Invoking uncertainty about the future, the Bush administration supports only minor reductions in the size and cost of the U.S. military force structure. Such uncertainty, of course, is an omnipresent feature of life; that hardly implies that a force structure appropriate for an uncertain world with a Cold War is appropriate also for an uncertain world without one. That life is uncertain can be used to justify anything. But because resources are limited always and everywhere, judgments must be made about the seriousness of alternative risks. Whatever one's view of U.S. interests in Europe, the end of the Cold War must imply that we should cut U.S. military forces subtantially.

On the other hand, unforeseen risks can arise. That future risks are undeniably smaller, but still real, means that preservation of a smaller but substantial defense establishment is appropriate. This force must be sufficiently large to deal with realistic threats, say, to Mexico and must be expandable on relatively short notice.

In budget terms, this requires substantial resource commitments to training and equipment but for smaller forces. Equipment whose acquisition involves long lead times should be stockpiled, with production lines kept open and expansion capacity maintained. Further, the United States should maintain a large training program for reserves to allow for a quick expansion of our forces should that prove necessary. In total, the defense budget should be cut to about $150 billion (in 1991 dollars) over a period of five years at most.

Missile Defense. "Nonproliferation" is an unrealistic goal. The embargoes upon which it depends require that all potential sellers agree not to sell, an arrangement that is unstable for the same reasons that any cartel will tend to find its members cheating. In a world with rapid evolution in technology and with rapid dispersion of technological expertise, it is foolish to assume that many regimes will fail to acquire substantial nuclear, biological, or chemical weapon technology and delivery capabilities.

Such a world may well be one in which future Saddam Husseins, of which there is an infinite supply, will be able to threaten greater damage to the United States than the United States can threaten in retaliation. Certainly, the exchange of political damage between elected U.S. leaders and foreign autocrats is unlikely to prove salutary from our perspective.

Therefore, the United States should develop and deploy missile defenses, whether ground or space-based, capable of defending against attacks by dozens or perhaps hundreds of warheads. Such a system should be made expandable because the military capabilities of even small regimes will tend to grow over time, increasing the danger to the United States.

Antiterrorism. Rarely if ever does terrorism pose a threat to the survival of a state. But responses to terrorism carry important implications for the security of individual citizens, for future terrorist activities, and for public morale under democracy. As a central principle, attacks upon innocent lives and property must be resisted vigorously, and nations harboring or sponsoring terrorist organizations must be subjected to certain and disproportionate retaliation.

Retaliation is necessary because free societies cannot be on the defensive always and everywhere, and more generally because hiding under the bed is not consistent with preservation of national self-respect. This means that resources must be invested in commando units, transportation, delivery systems, special weaponry designed to minimize collateral damage, and the like.

Preemption. Conditions will arise under which hostile regimes will develop weaponry threatening vital or important U.S. interests. Major ground operations or other actions in response may appear too cumbersome or otherwise inappropriate. Therefore, the United States should have a capability for "surgical" preemptive strikes designed to remove such threatening weapons. A small long-range bomber force, coupled with development of highly accurate cruise missiles, whether ground-, sea-, or air-launched, will provide an important form of insurance.

Nuclear Deterrence. Nuclear threats will not disappear, and preservation of effective deterrence against potential first strikes is an important defense goal remaining from the Cold War. So that potential aggressors will see an effective first strike as futile, presentation of an effective second-strike capability requires some combination of missile defense and deployment of large numbers of retaliatory weapons.

Maintenance of deterrence stability—reduction of incentives for first strikes by potential adversaries—requires that the U.S. strategic force not be seen as a first-strike force. This consideration puts limits on the number of strategic missiles carried by submarines and a premium on nuclear-armed cruise missiles and bombers, which are too slow to pose a first-strike threat. Multiple-warhead weapons and other such destabilizing strategic assets should be limited sharply or eliminated.

Finally, let us recognize that the U.S. defense establishment now may be the largest centrally planned economy in the world. Pentagon decision makers, when designing weaponry or when specifying performance characteristics, have weak incentives to respond to shifts in consumer preferences. The "consumers" are the soldiers in the foxholes, the airmen facing dogfights and anti-aircraft fire, and so on. A good proxy for these ultimate users may be the theater commanders charged with the task of winning battles. But without a profit motive or competition, Defense Department decision makers have few incentives to follow the preferences even of theater commanders.

As a result, the Pentagon often has promoted and shielded from criticism and competition weapon designs with dubious combat features and effectiveness. The performance of the M-16 rifle during the Vietnam War provides a classic example of such bureaucratic turf protection: the M-16 was redesigned by the Army from the AR-15, which had been developed by a private inventor. To protect its own design bureaus, the Army fought against it for years, and then redesigned it, resulting in a far less reliable and effective weapon.

During the 1991 Gulf War, the Air Force A-10 "Warthog," however ugly, performed brilliantly in support of Army and Marine ground operations, as it was designed for that specific purpose. For years, the Air Force tried to eliminate funding for the A-10 precisely because it supports the other services, and thus yields few bureaucratic benefits for the Air Force. The Air Force wanted to use F-16s and other more glamorous aircraft for ground support, even though their great speed makes them far less suitable for such missions. One crude way to get weapons and other equipment that conform more to user preferences is to give the users a larger voice or a direct veto in design decisions.

Because the military services have tasks that are defined sharply with little overlap, each service is in effect a monopolist in its defined missions. This is closely analogous to a cartel which divides submarkets among its members. The Army, for example, is prevented from flying fixed-wing aircraft, giving the Air Force a monopoly in the provision of much close air support for ground operations.

An important reform would have the services compete. For example, the Army and the Marines could be required to compete on a much broader scale in "production" of ground combat operations. The Army and the Air Force could compete in the provision of close air support. The Navy could be forced to compete on a much broader scale with the Coast Guard.

The 1990s have brought the dawn of a new age, but the U.S. defense establishment is structured to deal poorly with smaller but more numerous threats—or with tighter budgets. Defense is too important to be left to the Cold Warriors.

Benjamin Zycher is vice president for research at the Milken Institute for Job and Capital Formation and a visiting professor of economics at UCLA. He was formerly a defense analyst with the RAND Corp.