War offers ample opportunities for most varieties of foolishness. Among these, there is one sort of folly to which war is especially well suited: the lust for glory. One can hardly ever be sure about a commander’s motives in any one case, but there are familiar signs of that lust: a readiness to accept a challenge to fight under unfavorable circumstances; a preference for taking action independent of allies or colleagues; an unreasoning predisposition for offense rather than defense; and an effort to seize a decisive role in winning victory. Examples come easily to mind. Antony accepted Agrippa’s challenge to fight by sea at Actium, though he was stronger by land. In 1421 the Duke of Clarence violated the orders of his brother, King Henry V, and died attacking five thousand French troops with 150 mounted men-at-arms and no archers. To recapture the glory he had won by riding around McClellan’s army in search of its flank during the defense of Richmond in 1862, J.E.B. Stuart in June and July of 1863 led his cavalry on a wild ride through Maryland and Pennsylvania, even though it left the Army of Northern Virginia without the reconnaissance it needed in the week before the Battle of Gettysburg. Admiral William F. Halsey Jr. commanded the Third Fleet to chase Japanese battleships and carriers while other Japanese battleships threatened American soldiers landing on the beaches of Leyte Island.

Though there always will be soldiers and sailors “seeking the bubble reputation, even in the cannon’s mouth,” it seems that the vainglory of individual commanders has lately become less dangerous in war, as improvements in the technology of communications and surveillance have increased the ability of commanders to control subordinates. But there is a continuing danger from an institutionalized vainglory. Sometimes a branch of the military may try to maximize its opportunity for glory, turning its back on other less glamorous tasks that are really needed. This can become an ideology, like the French army’s doctrine in 1914 of “l’attaque à outrance.” The military may even adopt weapons that serve more to enhance its glory than the likelihood of victory, and weapons themselves may become imbued with a glamour that stands in the way of sensible decisions about their use. One can find instances throughout history, and they extend unfortunately to the present day, with dangerous effects on our current defense policy.


On February 1, 1917, Germany began a program of unrestricted submarine warfare. The effect on British shipping was devastating. During the first three months German U-boats sank 844 ships, at a cost of only ten of their submarines. According to Winston Churchill, “That was, in my opinion, the gravest peril that we faced in all the ups and downs of that war.”

It should have been obvious that the solution to the U-boat threat was to require merchant ships to sail in convoy. As Churchill later explained in The World Crisis,

The size of the sea is so vast that the difference between the size of a convoy and the size of a single ship shrinks in comparison almost to insignificance. There was in fact nearly as good a chance of a convoy of forty ships in close order slipping unperceived between the patrolling U-boats as there was for a single ship; and each time this happened, forty ships escaped instead of one.

(This is also the reason that fish of many species swim in schools.) Furthermore, forty merchant ships can be guarded by a much smaller number of destroyers or other escorts, while it would be impossible to send an escort with each merchant ship sailing alone. Even before sonar became available, a submerged submarine could be found when it attacked, by tracing back the track of the submarine’s torpedo. It is true that a convoy presents a great many more targets than a single ship, but even with all those targets a single U-boat can destroy only a few ships before it exhausts its torpedoes or is destroyed or driven off by the escorts. U-boats in World War II learned to call in other U-boats to the attack when a convoy had been found by using radio to communicate with their headquarters in occupied France, but this technology was not available in World War I. (Nor is it available to fish.)

For several months after the start of unrestricted submarine warfare, while British ship losses mounted, the Admiralty continued to reject the use of convoys. For the Royal Navy in World War I, convoy duty was inglorious.1 Arthur Marder, the leading historian of naval warfare in the early twentieth century, has explained in From the Dreadnought to Scapa Flow that

the strange dogma had emerged in the pre-war generation that to provide warship escorts to merchant ships was to act essentially “defensively” (because it protected ships from attack), which was ipso facto bad, and that to use naval forces to patrol trade routes, however futile the result, was to act “offensively” against the warships of an enemy, and this was good.

In 1915 British Vice Admiral H.F. Oliver, chief of the War Staff, had explained to the secretary of the War Cabinet, Sir Maurice Hankey, that the correct antidote to the U-boat menace was not convoy, but “hunting.”


The trouble is that in the absence of long-range aviation and radar (and often even with them) it is very difficult for patrols to find submarines cruising in the open ocean. In the days when I used to work on antisubmarine warfare as a member of JASON, a group of academics serving as defense consultants, my efforts were mostly limited to solving equations, but in the early 1960s I did once go out on antisubmarine maneuvers. I was on a destroyer escort searching for a World War II diesel submarine that had left Key West that morning. Our ship and a destroyer searched all day, using passive and active sonar and magnetic detection devices, but it could not find the sub. At the end of the day the submarine had to send up a radio beacon to tell us where it was. Remembering this, when I later read the estimate of Rear Admiral William Sims (the American admiral in charge of US naval forces in Europe during World War I) that for patrols to find submarines, which can submerge at will, requires about one destroyer per square mile of ocean, I thought that Sims had been too optimistic. And of course there are a lot of square miles in the ocean. According to Sims’s estimate, to find a submarine cruising somewhere in the western approaches to the British Isles would have required 25,000 destroyers.

As aggressive patrolling clearly failed to counter the U-boats and shipping losses grew, the Admiralty became desperate. When Sims arrived in London on April 10, 1917, he was invited to meet with the First Sea Lord, Admiral Sir John Jellicoe. Jellicoe told him that if shipping losses continued at their present rate Britain would have to leave the war, and that the Admiralty had no idea what to do about it. Sims cabled back to Washington that “briefly stated, I consider that at the present moment we are losing the war.”

Pressure from Sims, Hankey, and Lloyd George finally forced the Admiralty to try using convoys. As an experiment, a convoy of merchant ships was sent from Gibraltar to Britain, and all ships arrived safely on May 20, 1917. The next day the Admiralty at last decreed that all merchant shipping to or from Britain must travel in convoys, and the rate of shipping losses dropped sharply.

In World War II it was the strategic air forces of Britain and the US that tried to play a glorious but unrealistic role—to win the war by themselves through strategic bombing, i.e., bombing aimed at the enemy’s industry and population. There has been a great debate about the efficacy of strategic bombing in World War II. But however useful strategic bombing may have been, it could not have won the war by itself, at least not until the advent of nuclear weapons. In the invasion of Normandy it was surely necessary that all available bombers and their fighter escorts should come under Eisenhower’s authority, so that he could divert them as needed from strategic bombing to support of ground troops and the interdiction of German reinforcements. Nevertheless, James Doolittle and Arthur Harris, the commanders of the American Eighth Air Force and the British Bomber Command, put up a strong resistance to this arrangement, or to any pause in strategic bombing, and they managed to enlist Churchill’s support.2 To gain control over the strategic air forces during the critical phase of the invasion Eisenhower had to state that as long as he was in command he would accept no other arrangement, thus implicitly threatening to resign.

Even Stuart’s failure to keep in touch with Lee before Gettysburg may have represented an institutionalized hubris rather than mere personal vainglory. Other Civil War cavalry generals also liked independent action. In August 1864 the commander of the Confederate Army of Tennessee, John B. Hood, sent his cavalry corps, under General Joseph Wheeler, to raid the railroad lines that were supplying Sherman’s army in Georgia, with the expectation (according to Richard McMurry’s book Atlanta 1864) that Wheeler would be gone for only a few days. As it happened, nobody in the Army of Tennessee would see Wheeler’s troops again until mid-October. And in February 1864 the Union cavalry general Judson Kilpatrick talked Lincoln into allowing his cavalry to make a disastrous raid into Richmond in order to distribute pamphlets and free some prisoners. As Bruce Catton observed, “It was born of a romantic dream and it was aimed at glory, and glory was out of date.”


We have recently had another example of the military distaste for unheroic roles. Just as the Royal Navy preferred “hunting” to convoy duty in World War I, and the Allied air forces preferred strategic bombing to ground support in World War II, and the cavalry preferred independent action to the support of infantry in the Civil War, so in recent years the United States Army has preferred to plan for fighting battles without worrying about how to govern conquered territory. The Army in World War II had an effective Division of Military Government. It was established in the Office of the Provost Marshall in July 1942, long before there were any captured Axis territories to govern. It was this division and the personnel whom it trained at the Charlottesville School of Military Government that made it possible for the United States later to govern Japan and parts of Germany and Italy in an orderly way, without encountering widespread looting, rioting, or guerrilla attacks.

In the years after the war, responsibility for military government was relocated in the Civil Affairs branch of the Army. Support for this branch was allowed to dwindle, and Civil Affairs survived several attempts to disband it as a separate unit, until in 1987 it finally found a home in the Special Operations Command. There it had to fight off attempts to divert its remaining funds and personnel slots to Special Forces. At the end of the 1980s, an Army-commissioned report, in a chapter called “Pruning Non-Essentials,” asked the questions “Should 7,000 reservists continue to be trained to govern occupied nations? Is there a need for those trained in the administration of art, archives, and monuments to preserve the culture of occupied territories?”3 Civil Affairs became known as a dead end for career officers.

There is now just one active-duty Civil Affairs unit, the 96th Civil Affairs Battalion (Airborne), headquartered at Fort Bragg; the remaining 95 percent of Civil Affairs personnel are reservists. In Afghanistan there are now only about two hundred Civil Affairs personnel, as compared with about 15,000 military government soldiers in the American Zone of Germany soon after the German surrender in World War II. A colonel (not in Civil Affairs) who is just back from Iraq tells me that there are about two thousand Civil Affairs officers there (not all in military government), leaving few anywhere else, and that although they are doing good work, there are not nearly enough of them. Unfortunately the Defense Department’s priorities do not seem to have changed. Later this year it plans to close the ten-year-old Peacekeeping Institute of the Army War College.


Histories of military technology often describe how the adoption of newly developed weapons is delayed by cultural influences. (Perhaps the most striking example was the refusal of the Tokugawa shoguns to allow guns in Japan.) But in these histories, when new weapons are finally adopted, it is always because of their objective effectiveness, rather than something as irrational as a lust for glory.

In his influential 1962 book Medieval Technology and Social Change, the distinguished medieval historian Lynn White follows this pattern. As an example of new military technology, he chose the combination of stirrup and lance. The foot-stirrup was unknown in classical antiquity. As White tells the story, it was invented in China sometime around the fifth century AD. Nomadic horsemen brought the stirrup through Central Asia to Constantinople. Nowhere along this journey did it have much effect on how battles were fought. But when the stirrup reached Western Europe late in the seventh century its effect was explosive. The stirrup allowed a horseman with “couched” lance, i.e., one held between the body and the upper arm, to deliver the whole force of a charging horse at the tip of his lance, without losing his seat. White argued that suddenly it was necessary to recruit horsemen who could afford the armor and war horses needed to fight in this way, horsemen who could only be paid by the grant of lands. Starting in 732, Charles Martel, the great mayor of the palace of Merovingian France, began to seize church lands and grant them to his cavalry. A new class of feudal knights and nobles was born.4

I have my doubts about this story. There is no question that stirrups are a good thing, as everyone knows who has ever tried to get on a horse. I don’t doubt that the cavalry charge with couched lance became a popular tactic after 1100. But was it adopted because it was effective?

The use of the stirrup that White thought was “above all” important in tipping the balance in favor of cavalry over infantry, enabling a mounted knight to charge with couched lance, which is also the one tactic in which the stirrup is indispensable, does not seem to have been very effective against infantry. In such a charge each knight can only kill one foot soldier before a general melee begins in which the lance is useless. A “destrier,” a horse strong enough to carry an armored knight and bred and trained to charge an enemy, was enormously expensive. As the Chorus says in King Henry the Fifth, “they sell the pasture now to buy the horse.” When we consider also the cost of the knight’s and horse’s armor, and of the necessary grooms and remounts, killing a foot soldier this way seems to have been hardly cost-effective. It would be like tanks without guns attacking infantry by running them down, one by one. Doubtless the tanks would defeat an equal number of foot soldiers, but in a war between two countries with equal resources, the side that used its resources this way would surely lose.

Also, foot soldiers were not defenseless in the Middle Ages. They could stand behind pointed stakes set in the ground, as the English archers did at Agincourt and other battles of the Hundred Years’ War. They could protect themselves with long pikes, as Italian foot soldiers did to repulse the cavalry of Emperor Frederick I at Legnano in 1176, or fight on marshy ground, as Flemish pikemen did in repulsing the French cavalry at Courtrai in 1302, or on hilly ground, as the Swiss did in defeating Leopold of Austria at Morgarten in 1315. And the foot soldiers could fight back with missile weapons. A bolt from a crossbow or an arrow from a long bow could penetrate all but the best armor. It was English archers who defeated the French knights at Crécy, Poitiers, and Agincourt.

I do not know of a single European battle in the Middle Ages that was won by a charge of cavalry with couched lances against a line of foot soldiers. No clear examples are mentioned by Charles Oman in his classic two-volume A History of the Art of War in the Middle Ages. Granted, foot soldiers all but disappeared from Eu-ropean battlefields in the high Middle Ages, from 1100 to 1300, but apparently not because they had been defeated by the lance and stirrup.5

Oddly, White mentions only one European battle to prove the superiority of the new military technology, the Battle of Hastings on October 14, 1066. It is true that the Normans under Duke William were mostly mounted at Hastings, and they defeated King Harold’s English army, which fought on foot. But the English soldiers were probably outnumbered by the Normans, and they were exhausted. After defeating an invasion by a Norwegian army under Harald Hardraade at Stamford Bridge on September 25, they had arrived in London on October 6, having marched the two hundred miles from York, and they then marched another fifty-eight miles to the neighborhood of Hastings. What is more, it is not clear that the Norman cavalry at Hastings ever charged English foot soldiers with couched lance.6 The only reliable testimony to the wea-pons used at Hastings is the Bayeux tapestry, embroidered not long after the battle.7 Looking over reproductions of the whole tapestry, I see plenty of lances in the hands of Norman cavalry, but that proves nothing. Lances had been widely used by cavalry for stabbing and throwing since long before the invention either of stirrups or feudalism. Many of the lances shown in the Bayeux tapestry are held above the head, as if they are about to be thrown.

Just one panel, number 65, shows what may be a couched lance being used to attack an English foot soldier, but there is no way to tell if this is the result of a cavalry charge. I find it telling that none of the lances shown on the Bayeux tapestry have the kind of handhold (such as the conical handholds known as vamplates that are shown on later illustrations of tournaments) that would allow a knight on a charging horse to drive the lance into an enemy. Without such a handle, a lance striking anything would just slide through the knight’s armpit, stirrups or no stirrups. In any case, the English were defeated at Hastings more by Norman archers than by knights. The Bayeux tapestry shows an Englishman, generally believed to be Harold, struck in the eye by an arrow.

So if the cavalry charge with couched lance was an ineffective tactic against infantry, why did the lance become the standard weapon of medieval cavalry, and why did cavalry dominate the battlefields of the high Middle Ages? It seems likely to me that, instead of the new class of feudal nobility being called into being to take advantage of the combination of stirrup and couched lance, feudalism arose for other reasons,8 and the cavalry charge with couched lance became the favorite tactic of the feudal knights and nobles because it gratified their desire for military glory. Landowners could see themselves as military heroes, not merely as recruiting sergeants of the prince. The great expense of this tactic was an advantage, not a drawback, because peasants could not afford a destrier and armor. The reason that foot soldiers were not used in battle is not that they would have been ineffective, but that they were common. If this all sounds too illogical to be true, recall Oman’s verdict on the knights of the high Middle Ages: “A feudal force presented an assemblage of unsoldierlike qualities such as have seldom been known to coexist.”

The cavalry charge with couched lance found its real use in tournaments. These gorgeous rituals gave skillful knights like England’s William Marshall the opportunity to become international celebrities. It is sometimes said that tournaments became popular in the twelfth century as training grounds for real battle,9 but it may be equally true to say that cavalry battles with infantry excluded served as training grounds for tournaments.


It is not hard to find examples closer to the present of weapons that were adopted more for glory than for military effectiveness. Between the world wars, officers of the US Army Air Corps imagined that land-based aircraft would not only be able to de-stroy the enemy’s industry by strategic bombing, but would also be able to conquer enemy fleets. Air Corps General William Mitchell predicted that air power “will not only dominate the land but the sea as well.” Just as medieval knights were supposed to make plebeian foot soldiers unnecessary, bombers would make navies irrelevant.

For bombers to act independently at sea beyond coastal waters without having to be carried to the attack by naval vessels, they would have to have a very long range and carry a large bomb load, just as they would for strategic bombing. Such an airplane would have to be large, so it could not be used as a dive bomber, and it would therefore have to release its bombs in level flight. To avoid antiaircraft fire it would have to bomb from a high altitude. After a competition among several aircraft manufacturers, the Boeing corporation in January 1936 received a contract for a bomber to meet these requirements. It would eventually be named the B17, or “Flying Fortress,” and after the war it would become the glamorous star of the movies Memphis Belle and Twelve O’Clock High.

The B17 met all technical specifications of the Department of the Army, but it could not fulfill the purposes for which it was planned. After the attack on Pearl Harbor, B17s were sent to Midway Island. As a Japanese fleet approached Midway in June 1942, it was spotted by B17s. In fifty-five sorties between June 3 and 5, the B17s dropped 315 bombs, but despite optimistic reports from the pilots, not a single one of these bombs struck a Japanese ship. Antiaircraft fire from the Japanese fleet kept the B17s above 20,000 feet, and at that altitude it was just too hard to drop a bomb on a target as small as a ship. Also, like medieval foot soldiers dodging an attack by knights with couched lances, ships can maneuver out of the way of bombs dropped from high altitude. As Samuel Eliot Morison remarked, in his History of United States Naval Operations in World War II, the B17s would have done more damage by tracking the Japanese ships than by trying to bomb them, but their pilots were trained only for bombing.

The confidence of the Army Air Force in its ability to bomb enemy ships arose in part from a dramatic demonstration staged in 1921 by Mitchell, in which Army bombers had successfully sunk a number of surrendered German warships, including the dreadnought Ostfriesland. But the German ships were at anchor, not maneuvering, and since they were not firing at the aircraft, it was safe to bomb them from low altitude. The four Japanese carriers at Midway were sunk not by B17s but by short-range Navy SBD dive bombers, which had been carried to the battle by the Enterprise and Hornet. During 1943, B17s were mostly withdrawn from the Pacific theater, though a few were kept for less showy activities like dropping life rafts to shipwrecked sailors.

The B17 also encountered trouble even in its primary role of strategic bombing. To have any chance of hitting targets like factories and railroad marshaling yards without our present satellite-based global positioning system, bombers had to attack in daylight. But flying over enemy territory in daylight made them vulnerable to attack by enemy interceptor airplanes as well as antiaircraft artillery. For this reason the design of the B17 had sacrificed part of its bomb capacity to enable it to carry exceptionally heavy armament: thirteen machine guns, manned by six crew members. (The maximum bomb load of the B17 was less than that of the British Lancaster10 or the American B24. The B24 proved more useful than the B17, but it never attracted the same public attention.) The Air Force expected that B17s flying in tight formation would be able to fight off interceptors by themselves and reach targets beyond the range of American fighter escorts. Experience did not bear this out. In a daylight raid on August 17, 1943, a force of 315 unescorted B17s attacked a ball-bearing factory at Schweinfurt and a Messerschmitt factory at Regensburg, and lost sixty bombers and their precious crews. At that rate of loss, the Eighth Air Force could not have continued daylight strategic bombing for more than a few more raids without losing almost all of its strength. Daylight strategic bombing deep into Germany had to be suspended, until long-range escort fighters like the P51 became available in February–March 1944.

Despite all disappointments about the ineffectiveness of the B17 in performing the tasks for which it had been planned, it embodied the dreams of glory of the Air Force, and became a symbol of American air power. Over 12,000 B17s were built during the war. It is true that B17s along with other bombers did take an important part in the war by attacking German synthetic oil plants and transportation facilities after long-range escorts became available, and above all in forcing Germany to divert artillery and fighter aircraft from battlefields to air defense. But the huge commitment of American resources to the B17 limited other production. Allied war planning was continually constrained by a shortage of landing craft, but never by a shortage of B17s.


Since World War II it has been rockets and nuclear weapons that have the glamour that used to surround bombers or lances. There is no doubt about the extraordinary effectiveness of modern high-technology weapons in certain circumstances. But it always needs to be asked at which tasks are they effective, and which of these tasks actually need to be accomplished?

In the decades since nuclear weapons were used in war, there has developed a healthy conviction that these weapons should not be used again for anything but deterrence. In a remarkable international bargain, a number of countries agreed in 1970 that they would not develop nuclear weapons at all, in exchange for a pledge by the US and other nuclear powers to deemphasize the role of nuclear weapons and to work toward their elimination. Now the Bush administration has turned its back on the Non-Proliferation Treaty by calling for work on a new generation of low-yield nuclear weapons. This would add nothing to our nuclear deterrent, but might actually be used in fighting wars. Prominent in the administration’s wish list is a “robust nuclear earth penetrator,” a nuclear weapon that could penetrate into the earth to attack subterranean bunkers and laboratories.

The political pressure for these developments comes in part from the national weapons laboratories, especially Sandia and Los Alamos, whose leaders would like to return to the glory days of designing and testing new nuclear weapons. (The weapons labs are powerfully aided by Senator Pete Domenici, the chairman of the subcommittee that deals with appropriations for nuclear programs. Domenici represents New Mexico, where Sandia and Los Alamos are located.) The earth penetrator is also actively supported by the Air Force, which would drop it. There also seems to be a widespread sentiment in Congress and the public that it is inconsistent with American dignity to accept any limitations on our military technology.

As with the couched lance, it is hard to think of a plausible mission in which new low-yield nuclear weapons would actually be effective. As I have discussed previously in these pages,11 a nuclear weapon carried underground by a robust earth penetrator would only be able to attack targets that are not buried too deeply and whose location is precisely known, and any such attack would inevitably produce large quantities of radioactive fallout at the surface, possibly killing large numbers of civilians.

The problem is not so much that the money spent on new nuclear weapons would be wasted. The cost of these programs is in the range of tens or hundreds of millions rather than billions of dollars. The problem is, rather, that the United States would be violating the 1970 Non-Proliferation Treaty and encouraging a new round of nuclear weapons development throughout the world. For a nation with an enormous lead over the rest of the world in non-nuclear military technology, this would be foolishness on a scale that even medieval knights might find implausible.

It was to avert such dangers that Congress in 1993 passed the Spratt-Furst Amendment to the 1994 National Defense Authorization Act. This amendment stated that it was not the policy of the United States to develop or use nuclear weapons of low yield, defined as less than the equivalent of five thousand tons of TNT. There has been an argument within Congress and the Department of Energy about whether the Spratt-Furst Amendment bans research as well as development of such weapons, but this argument may become moot. In a message from the Office of Management and Budget to the House Armed Services Committee in May of this year, the administration called for a complete repeal of the Spratt-Furst Amendment. The Senate has voted to repeal the amendment, but at the time of writing the House has not. The issue will presumably be settled in a House-Senate conference committee this autumn.

National missile defense has for some the allure of a good science fiction movie. A fleet of American satellites detects the launch of a missile by some evil regime in Asia. Powerful radars in Alaska track the missile’s warhead, and American missiles are launched to meet it above the earth’s atmosphere. Our missiles then release exo-atmospheric kill vehicles (EKVs), which guide themselves to collide with the enemy warhead, and destroy the bomb it contains. Seattle is saved.

Unfortunately, missile defense is not likely to work so neatly.12 So far, tests of the EKV have alternated between successes and failures, and now the tests have been suspended while work continues on the booster that would carry the EKVs into space. But even if the tests had an unbroken series of successes, they would still have the same quality of phoniness as the destruction of the Ostfriesland by bombers in 1921. It is not that tests like this have no value, but rather that they can’t be relied on to tell us whether the system being tested would be really effective. Just as the Ostfriesland in Mitchell’s test was neither maneuvering nor shooting at the bombers attacking it, none of the warheads targeted in the tests of the EKV has been accompanied by the simple realistic decoys that could defeat the EKVs. Nor have other possible countermeasures against EKVs been addressed in any of the tests so far. In the real world B17s could not hit Japanese warships that were maneuvering in unpredictable ways, and similarly our EKVs will not be able to destroy incoming warheads if the enemy uses countermeasures that we cannot know in advance.

The proponents of missile defense to whom I have talked generally acknowledge these problems, but they argue that we must at least make a start, and then learn as we go along. As Secretary Rumsfeld has put it, the proposed national missile defense system is at least better than nothing. I think a strong case can be made that it is worse than nothing. It will hurt relations with our allies, discourage Russia from taking its missiles off ready alert, and encourage China to increase its missile forces more than had been planned.

A “boost-phase intercept” system, which relies on attacking missile boosters soon after launch, would not be subject to these criticisms if it defended our allies as well as ourselves, and if it were located so that it could not reach missiles launched from Russia or China. But a recent study by a panel of the American Physical Society indicates that feasible interceptors would not be able to reach missiles launched from North Korea or Iran before warheads separate from the booster.13

Granted, one cannot be sure about how other countries will respond to an American missile defense system. But there is no question about the enormous cost of missile defense. We are currently spending about nine billion dollars a year just for research and development, and a deployed system covering the entire United States would surely cost several hundred billion dollars, all to ineffectively counter a highly implausible threat. Any missile attack on the US would immediately reveal which country launched it and expose it to devastating retaliation. Such deterrence protected the US from nuclear attack for almost half a century of cold war. It is true that without missile defense the US might be deterred from trying to overthrow a “rogue state” that possessed nuclear-armed intercontinental missiles. But for the US to take such an action without fear of nuclear retaliation, a missile defense system would have to be nearly perfect, not merely better than nothing.

Even those for whom national defense is the one clearly legitimate reason for government spending ought to consider whether the enormous sums required for missile defense would not be better spent on defense of other sorts. There are many ways to attack the United States with nuclear or biological weapons that, unlike ballistic missiles, do not immediately reveal the source of the attack. Over the past year or so I have served on two panels of the Council on Foreign Relations, the Hart-Rudman Independent Task Force on Homeland Security Imperatives and the Rudman Independent Task Force on Emergency Responders. It has been painful to learn how much the lack of funds has limited our ability to defend the country from terrorists. For instance, the cost of adequate physical security at our commercial seaports would be about $2 billion, but only $92.3 million in federal grants has been authorized and approved. The US may be spending one third of what is required to adequately provide for those who would have to respond to emergencies. American cities have fewer policemen and firemen now than they did before September 11, 2001. Last October the Hart-Rudman panel concluded that “a year after September 11, 2001, America remains dangerously unprepared to prevent and respond to a catastrophic attack on US soil.” This remains true.

We face another danger, even greater than that from terrorists. For all the good relations between Russia and the US today, Russian nuclear forces are frozen in a cold-war configuration, one designed to respond to warning of an American attack within ten minutes by a massive nuclear counterattack, before a single nuclear weapon can reach Russia’s land-based missiles or control centers. This puts not just a city or two but the entire US in danger of irreversible destruction by mistake, a danger that will increase as Russia’s capacities to detect an attack become more and more degraded. Much could be done to lessen this danger, such as seriously reducing nuclear arsenals on both sides and sharing information about missile launches, but this has not been a priority of any US administration. Reducing the threat of nuclear attack by mistake, improving port security, providing for emergency responses, bringing effective government to Iraq and Afghanistan—in none of these do our leaders find sufficient glory.

—October 7, 2003

This Issue

November 6, 2003