Chapter One

The Rise and Fall of the Warfare State

In the beginning was war. From the very earliest days of recorded history until the very recent past, war has been the motor of financial change. "War is the father of all things," as Herodotus said; and among those things during the Pelopponesian War was an increase in Athenian expenditure, and consequently a need for higher taxes and other sources of revenue. It was war which, with a powerful symbolism, caused the golden statue of Athena to be melted down and coined.

    It is a truth—almost—universally acknowledged. Nervos belli, pecuniam infinitam: "The sinews of war [are] unlimited money," declared Cicero in his Fifth Philippic, a view echoed by Rabelais in Gargantua: "The strength of a war waged without monetary reserves is as fleeting as a breath." "What Your Majesty needs," Marshal Tribulzio told Louis XII before his invasion of Italy in 1499, "is money, more money, money all the time." The early sixteenth-century writer Robert de Balsac agreed: "Most important of all, success in war depends on having enough money to provide whatever the enterprise needs." "Your majesty is the greatest prince in Christendom," the Emperor Charles V was told by his sister Mary, "but you cannot undertake a war in the name of allChristendom until you have the means to carry it through to certain victory." Writing a century later, Cardinal Richelieu echoed her words: "Gold and money are among the chief and most necessary sources of the state's power ... a poor prince would not be able to undertake glorious action."

    It goes without saying that money at the immediate disposal of the state treasury is usually more limited than the costs of war; and the history of finance is largely the history of attempts to close that gap. Only in the recent past has this relationship between war and finance grown weak. After many centuries during which the cost of warfare was the biggest influence on state budgets, that role was usurped in the second half of the twentieth century by the cost of welfare. No doubt this is a great change for the better: though idleness is no virtue, it is morally preferable to pay men for doing nothing than to pay them for killing one another. But the remarkable extent and novelty of this change are not well understood. It is no exaggeration to speak today of the demilitarization of the West—and, indeed, of large areas of the rest of the world.

    A common error is to suppose that, over the long run, there has been a linear or exponential upward trend in the cost of war. In absolute terms, of course, the price of military hardware and the level of defence budgets have risen more or less inexorably since the beginning of written records. In relative terms, however, the patterns are more complicated. We need to relate military expenditure to the scale and frequency of war; to the size of armies in relation to total populations; to the destructiveness of military technology ("bangs per buck"); and above all to total economic output. Allowing for changes in population, technology, prices and output, the costs of war have in fact fluctuated quite widely throughout history. These fluctuations have been the driving force of financial innovation.

THE INTENSITY OF WAR

It is no part of this chapter to explain why wars happen, though the question will be returned to later. Let us for the moment simply acknowledge that they do, and often. How often is a matter for debate.

    There have been several attempts to quantify the frequency of military conflict, each based on a somewhat different definition of war and covering periods of varying lengths. P. A. Sorokin counted 97 wars in the period 1819-1925, compared with Quincy Wright's total of 112 between 1800 and 1945. Wright confined himself to what he called "wars of modern civilization ... involving members of the family of nations ... which were recognized as states of war in the legal sense or which involved over 50,000 troops;" whereas L. F. Richardson, counting all the "deadly quarrels" he could find, arrived at the much higher figure of 289 for the period 1819-1949. Luard's survey of all "organized large-scale fighting sustained over a significant period and involving at least one sovereign state" arrives at an even higher total of 410 for the period 1815-1984. However, the "Correlates of War" project based at the University of Michigan adopts a narrower definition which excludes most minor colonial wars, as well as wars involving countries with populations of less than 500,000, and wars in which total battle-deaths were less than a thousand per annum. For the period 1816 to 1992, their database lists 210 interstate wars and 151 civil wars. The lowest figure of all for the modern period is Levy's—31—but his survey considers only wars that involved one or more of the great powers.

    It is possible to take an even longer view, though for extra-European conflicts the evidence becomes more patchy the further back one goes, and even the most ambitious attempts avoid the ancient and medieval periods. On the basis of his relatively broad definition of what constitutes a war, Luard arrives at a total of over a thousand for the period 1400 to 1984. Levy, by contrast, counts just 119 great-power wars in the period 1495 to 1975. Even on the basis of the latter's narrower definition, the perennial nature of war is striking:

The Great Powers have been involved in interstate wars for nearly 75 per cent of the 481 years [from 1495 to 1975] ... On average a new war begins every four years and a Great Power war [i.e. a war involving more than one great power] every seven or eight years.... In the typical [median] year ... slightly over one war involving the Great Powers ... is under way ...

No twenty-five year period since 1495 has been entirely without war.

    It is possible to bring this audit of war up to the present. The Stockholm International Peace Research Institute (SIPRI) estimates that there were 103 "armed conflicts" between 1989 and 1997, of which six were inter-state conflicts. In 1999 there were some 27 major armed conflicts in progress, though only two were between sovereign states (between India and Pakistan and between Eritrea and Ethiopia). Adopting Levy's criteria for wars involving at least one great power, there have been six since Vietnam (the last war considered in his survey): the Sino-Russian War (1969), the Sino-Vietnamese War (1979), the Soviet-Afghan War (1979-89), the Falklands War (1982), the Gulf War (1990-91) and the Kosovo War (1999).

    Has war grown more or less frequent over time? Some would say less so. Counting only wars involving one or more great powers, there was at least one war underway in ninety-five of the years of the sixteenth century and in ninety-four of the years in the seventeenth; but that the figure falls to seventy-eight for the eighteenth and forty for the nineteenth, and rises to barely more than fifty for the twentieth. Put differently, the "average yearly amount of war" was highest in the sixteenth century and lowest in the nineteenth and twentieth centuries? However, using a broader definition of war, Luard lists 281 wars for the period 1400-1559, falling to 162 (1559-1648) and 145 (1648-1789), but then rising to 270 (1789-1917) before returning to 163 between 1917 and 1984. Adding together all the wars covered by the Correlates of War database—including wars that did not involve a major power, as well as civil wars—provides further evidence of modern bellocosity. It is striking that there has not been a single year since 1816 without at least one war going on in the world. Only in Europe has war has grown less frequent since 1945. The percentage of wars that took place in Europe falls steadily from more than 80 per cent in Luard's first sub-period (1400-1559) to just 9 per cent in his last (1917-1984).

    Which of the great powers has been the most belligerent? On the basis of a slightly modified and extended version of Levy's dataset, the answer would appear to be France, which has participated in some 50 of 125 major wars since 1495. Austria is not far behind (47), followed by another former Habsburg realm, Spain (44) and, in fourth place, England (43). According to Luard's larger list of wars, however, the most warlike states in the years 1400 to 1559 were the Habsburg and Ottoman Empires. Between 1559 and 1648 Spain and Sweden led the field, waging war in 83 of those years. France was certainly the prime warmonger from 1648 until 1789 (80 out of 141 years) and again, with respect to European wars, from 1789 until 1917 (32 out of 128 years). However, Britain was more often involved in wars outside Europe between 1815 and 1914 (71 out of 99 years). There were 72 separate British military campaigns in the course of Queen Victoria's reign—more than one for every year of the so-called pax britannica.

    Simply counting raw numbers of wars can only tell us so much, of course. For example, eighteenth-century wars lasted longer and involved more powers than wars in previous or subsequent centuries: in that sense, the average war was, perhaps surprisingly, a bigger affair in the Age of Enlightenment than the average war before or since. Even in terms of "severity" (total battle deaths), the average eighteenth-century war ranks above the average twentieth-century war, to say nothing of the wars of all other centuries. Only in terms of "concentration" (battle deaths per nation-year) was the average twentieth-century war bigger. This reflects the fact that the great-power wars of the twentieth century were more compressed than those of the period before 1815; whereas the periods of peace between the great powers were significantly longer. While the average length of war declined from eight years in the eighteenth century to four and a half in the twentieth, the number of battles in each year of war rose steeply.

    Almost as remarkable in this long-term perspective was the comparative peacefulness of the century between 1816 and 1913. Although there were around a hundred colonial wars in the period—the majority fought by Britain, France or Russia—the scale of these wars tended to be small because of the technological superiority of the imperial powers. Also on a relatively small scale were the numerous wars of national independence. At the same time, the great powers kept war between themselves to an historical minimum. Apart from the Crimean War, the great power clashes of the period 1854-71 seldom lasted longer than a few weeks. The late twentieth century saw a return to this pattern: the war against Iraq in the Gulf lasted eighty-five days; the war against Serbia over Kosovo a mere seventy-eight. If there has been a discernible trend over the past two or three centuries, then, it has been the increasing concentration or intensity of war.

MEN OF WAR

The dramatic difference between the world wars and the rest of modern history is immediately apparent when we turn to the extent of military mobilization: that is to the say, the proportion of the population employed in the armed forces. In absolute terms, armies reached historically unprecedented sizes in the twentieth century: probably the largest military force in history was that of the Soviet Union in 1945, which numbered around 12.5 million. By comparison, the armies that fought the Hundred Years War seldom exceeded twelve thousand in size. Even today, after some fifteen years of troop reductions, the American services still employ 1.4 million people.

    But such figures tell us little about the relative degrees of mobilization involved. In the eighteenth century the highest recorded percentage of the British population under arms was 2.8 per cent in 1780, when Britain was at war not only with her American colonists, but also with France, Spain and Holland. But in more peaceful years the figure fell below 1 per cent. For France, the proportion of men in the armed forces tended to decline in the eighteenth century, from 1.8 per cent in 1710 to 0.8 per cent in 1790. Austria consistently kept between 1 and 2 per cent of her population under arms throughout the century; but this was a much lower proportion than that of Prussia, which in 1760 had as many as 4.1 per cent of her people in the army. For all countries, the Napoleonic "revolution in war" meant an increase in the proportion of the population that had to be mobilized. In 1810 Britain had more than 5 per cent of her people under arms, Prussia 3.9 per cent, France 3.7 per cent and Austria 2.4 per cent.

    By comparison, the nineteenth century saw relatively low rates of military participation. With the exceptions of Russia during the Crimean War, the United States during the Civil War and France and Prussia during the war of 1870-1, none of the major powers mobilized more than 2 per cent of the population between 1816 and 1913. Apart from the years 1855-6, 1858-63 and 1900-1902, the figure in Britain remained less than 1 per cent until 1912, reaching a low point of 0.5 per cent in 1835. On average, Austria and Piedmont/Italy also had armed forces of less than 1 per cent of the population between 1816 and 1913; and for Prussia, Russia and France, the average proportions were all below 1.3 per cent. Just 0.2 per cent of the population of the United States was in the armed forces during the nineteenth century as a whole. Even in 1913, despite contemporary and historical perceptions of an arms race, only Britain, France and Germany had more than 1 per cent of their populations under arms.

    The First World War saw the highest rates of military participation in all history. At their peaks of wartime mobilization, France and Germany had more than 13 per cent of their populations in the services, Britain more than 9 per cent, Italy more than 8 per cent, Austria-Hungary just over 7 and Russia only slightly less. But immediately after the war, as if in reaction, all the major powers substantially reduced their military participation ratios. On average, only France mobilized more than 1 per cent of her population. In Britain the figure touched a nadir of 0.7 per cent in the mid-1930s; while in the Soviet Union in 1932 it was less than a third of 1 per cent. The United States also reverted to its nineteenth-century level of military unreadiness. Even Nazi Germany took time to raise the share of the population in the army, navy and air force after the enforced reduction that had been a part of the Versailles Treaty of 1919. Not until 1938 did the German armed services exceed 1 per cent of the population. Italy's Abyssinian adventure pushed its armed forces up to above 3 per cent in 1935, but by the eve of the Second World War the figure had sunk back to just over 1 per cent.

    Surprisingly, no country mobilized as large a percentage of the population into its armed forces between 1939 and 1945 as France managed in 1940 (just short of 12 per cent). The peak figure for Germany was 8.3 per cent in 1941, rather less than Britain managed in 1945 (10.4 per cent). It is also noteworthy that the Soviet proportion in that year (7.4 per cent) was less than the American (8.6 per cent). In the First World War, Germany had almost certainly committed too many men to the army at the expense of the industrial workforce. The Second World War apparently saw a more balanced allocation of labour.

    By comparison with the previous two post-war eras after 1815 and 1918, the years after 1945 did not witness such a rapid and sustained demobilization. In the Soviet case, the armed forces jumped back up from 1.5 per cent of the population in 1946 to 3.1 per cent in 1952; while American military participation rose from 0.9 per cent in 1948 to a post-war high of 2.2 per cent in 1952. Britain too experienced a slight rise associated with the Korean War. The French figure rose to a peak of 2.2 per cent in 1960 as a result of conflicts associated with decolonization.

    Nevertheless, during the Cold War period as a whole there was a steady fall in military participation ratios in many major countries. The average rate of mobilization in Germany, Italy and Austria was lower in the period 1947-85 than it had been between 1816 and 1913. Even for Russia the figure was below 2 per cent. Moreover, the break-up of the Warsaw Pact and the collapse of the Soviet Union has allowed military participation to fall back to inter-war levels and in some cases even lower. In 1997 just 0.37 per cent of the British population was serving in the armed forces: the lowest figure since 1816. The present French proportion (0.65 per cent) is the lowest since 1821.

    Rates of military mobilization, then, have been subject to sharp fluctuations above a relatively stable (and perhaps over the very long run even declining) base line. The major wars of the modern period, and particularly the world wars, have necessitated large but not sustained increases in military participation. Indeed, it is precisely because of its discontinuous, noncyclical character that warfare has exerted such a decisive influence over the development of financial and political institutions.

BANGS PER BUCK

Sudden increases in the proportion of men under arms are not the principal source of pressure on military budgets, however. Changes in military technology matter more. From the fourteenth-century gunpowder revolution onwards, artillery has periodically increased its range, accuracy and destructive power The development of the cast-iron cannon, with its iron ball, "corns" of powder and wheel base, necessitated a parallel improvement in fortifications like the trace italienne. Indeed, it was partly the rising cost of fortifications that put the finances of continental powers under strain in the sixteenth century. Likewise, the standardization and improvement of handguns in the early eighteenth century enhanced the firepower and raised the cost of equipping the individual infantry man? The eighteenth century saw further improvements in the manufacture of artillery, notably the bored barrel introduced to France by the Swiss engineer Jean Maritz, which set the standard until the advent of the breech-loading gun in the 1850s. The parallel development in Britain was in maritime technology: copper-sheathed bottoms for ships, short-barrelled, large-calibre carronades and steering wheels for ships.

    Moreover, the pace of technological advance quickened in the course of the nineteenth century: at sea, the application of steam power, Henri Paxihans' large-calibre shell-firing gun and iron cladding, followed by the torpedo, the submarine, Nordenfeldt's and Vavasseur's naval guns, the tube-boiler and the turbine; on land, the new rifles of Minié, Dreyse and Colt and the improved breech-loading artillery pieces of Krupp, Armstrong and Whitworth—to say nothing of brass cartridges (1867), steel artillery (1883), the Maxim Gun (1884), magazine rifles (1888) and the Schneider-Creusot quick-firing field gun (1893). The cauldron of the First World War brought forth new instruments of destruction, barely imagined before 1914: among them the tank, the aerial bomber and the fighter plane, as well as the hand grenade, the trench mortar and poison gas. Despite all talk of war-weariness, the process did not halt in the 1920s and 1930s: one need only compare the aircraft and tanks of 1938 with those of 1918 to see that. But the pace of change accelerated dramatically during the Second World War as the major combatants sought to out-innovate as well as out-produce one another, increasing the speed, range, accuracy and armour-plating of nearly all the machines of mid-century warfare. The British Spitfire—to give one example—was modified 1,000 times between 1938 and 1945, adding 100 mph to its top speed. At the same time, advances in radio technology ushered in a revolution in battlefield communications (wireless communication, radar detection), while a host of new inventions arrived in time for use in the final phase of the conflict: jet engines, amphibious vehicles, guided missiles, rockets and, of course, atomic bombs. This technological race continued in the Cold War, as A-bombs gave way to hydrogen and neutron bombs and the arms race became simultaneously a space race between rockets and satellites (with astronauts and cosmonauts thrown in to sustain public interest).

    In absolute terms, expenditure on military hardware has therefore risen inexorably in the long run. By 1982 a critic of the arms race could lament: "Bombers cost two hundred times as much as they did in World War II. Fighters cost one hundred times or more than they did in World War II. Aircraft carriers are twenty times as expensive and battle tanks are fifteen times as expensive as in World War II." Writing four years later, Paul Kennedy enlarged on this point:

Edwardian statesmen, appalled that a pre-1914 battleship cost £2.5 million, would be staggered that it now costs the British Admiralty £120 million and more for a replacement frigate! ... The new [American] B-1 bomber ... will cost over $200 billion for a mere one hundred planes ... Cynics [forecast] that the entire Pentagon budget may be swallowed up by one aircraft by the year 2020.

According to Kennedy, weapon prices in the 1980s were "rising 6 to 10 per cent faster than inflation, and ... every new weapon system is three to five times costlier than that which it is intended to replace." Despite a "near trebling of the American defence budget since the late 1970s," there had occurred by the late 1980s "a mere 5 per cent increase in the numerical size of the armed forces on active duty." To Kennedy, warnings were not misplaced of an impending "militarization of the world economy."

    Even allowing for inflation and relating expenditure to the size of armed forces, military expenditure has tended to rise. In 1850 Britain spent just under £2,700 per man on her armed forces (in 1998 prices); by 1900 the figure had risen to £12,900, and by 1950 £22,000. In 1998 the figure was close to £105,500. The United States spent $30,000 per serviceman in 1900 (again in 1998 prices); $71,900 in 1950; and $192,5000 in 1998 (see Figure 3). Nearly all the increase has been due to increased quantity and quality of military hardware (as opposed to improvements in soldiers' pay and living conditions). It is not too much to say that the increase in the military capital/labour ratio in the course of the twentieth century has been exponential.

    Yet in assessing the growing sophistication of military technology there are a number of things we should not lose sight of: in particular, its increasing destructiveness. For in the purchase of a new weapon, it is not only the price that matters; it is also its capacity, compared with the weapon it is intended to replace, to mete out murder.

    The death toll of the War of the Spanish Succession (1701-13) was 1.2 million. A century later, the Napoleonic Wars killed 1.9 million men. And a century after that, the First World War cost more than 9 million servicemen their lives. Perhaps as many as 8 million people died in the maelstrom of the Russian Civil War of 1918-21 (though most of these were the victims of the famine and pestilence unleashed by the conflict). But even this figure pales into insignificance alongside the total mortality caused by the Second World War. For military personnel, the total body count was roughly twice the figure for the First World War. But this figure excludes civilian casualties. According to the best available estimates, total civilian deaths in the Second World War amounted to 37.8 million, bringing the total death toll to nearly 57 million people. In other words, the majority of deaths in the Second World War were due to deliberate targeting—by all sides—of civilians on land and sea and from the air. Including all the minor colonial wars like the Boer War and all the civil wars like the one that raged in India after independence, the total figure for war deaths between 1900 and 1950 approaches 80 million.

    The increase in the destructiveness of war becomes even more striking when the relative brevity of the world wars is taken into account. Though it lasted five times as long, the Thirty Years War caused only a ninth of the battlefield mortality inflicted during the Second World War, and an even smaller fraction of the civilian mortality. The First World War caused five times as many deaths in four and a quarter years as the entire Napoleonic Wars in the space of twelve. Another way of expressing this is to calculate the approximate annual death rate during the various wars. This rose from above 69,000 in the Thirty Years War to some 104,000 in the War of the Spanish Succession, 124,000 in the Seven Years War, 155,000 in the Napoleonic Wars and for the world wars, respectively, 2.2 and 3.2 million—or 9.5 million if civilian deaths in the Second World War are included. In short, between the seventeenth and the twentieth century, the capacity of war to kill rose by a factor of roughly 800. From the time of Napoleon to the time of Hitler—born a mere 120 years apart—the increase was more than 300-fold (see Appendix, table A).

    Even allowing for the accelerating growth in the world's population, then, the world wars were the most destructive in history. Somewhere in the region of 2.4 per cent of the world's entire population was killed in the Second World War and 0.5 per cent in the First, compared with roughly 0.4 per cent in the Thirty Years War and 0.2 per cent in the Napoleonic Wars and the War of the Spanish Succession. The total death toll in the First World War amounted to something like 1 per cent of the pre-war population of all fourteen combatant countries, 4 per cent of all males between 15 and 49 and 13 per cent of all those mobilized. For Turkey the equivalent figures were 4 per cent of the population, 15 per cent of males between 15 and 49 and almost 27 per cent of all those mobilized. Even worse affected was Serbia, which lost 6 per cent of the population, nearly a quarter of all men of fighting age and over a third of all those mobilized. In the Second World War roughly 3 per cent of the entire pre-war population of all combatant countries died as a result of the war. For Germany, Austria and Hungary the figure was around 8 per cent, for Yugoslavia and the Soviet Union 11 per cent and for Poland—of all countries the worst affected by the war—nearly 19 per cent: almost a fifth of the entire pre-war population. The armies of some countries were almost wholly annihilated. Military deaths as a proportion of all troops mobilized were in the region of 85 per cent for both Poland and Romania. Forty-five per cent of the troops mobilized in Yugoslavia were killed. For the Soviet Union and Germany, locked for four years in the most bloody conflict of all time, the equivalent figures were, respectively, 25 and 29 per cent. Around a quarter of Japanese and Chinese troops were killed in the war in Asia and the Pacific.

    To be sure, casualties as a proportion of troops engaged were sometimes very high in previous wars. Though the statistics are far from reliable for medieval battles, it is nevertheless plausible that the proportions (including wounded and prisoners) were between a quarter and a third of combatants at the battles of Hastings (1066), Crécy (1346), Agincourt (1415), Breitenreid (1631), L£tzen (1632), Naseby (1645), Austerlitz (1805), Waterloo (1815) and Gettysburg (1863). At Blenheim (1704) the figure may have been as high as 43 per cent. These figures bear comparison with some First and Second World War battles: for instance, El Alamein (c. 14 per cent), though not Stalingrad, where, in the space of six and a half months, the Red Army alone suffered 1.1 million casualties and the Wehrmacht as many, if not more. Yet these proportions need to be seen in the context of substantial increases in the numbers of troops committed to battle. Perhaps 14,000 men fought at Hastings; perhaps 39,000 at Crécy. But 68,000 fought at Breitenreid and 108,000 at Blenheim, while more than double the number who fought at Breitenfeld were deployed at Austerlitz. The Battle of Waterloo saw 218,000 men in the field; but even it was dwarfed by El Alamein (300,000) and Stalingrad, where millions fought. Just as military technology had magnified the destructive power of the individual, innovations in drill, discipline, communications and logistics had allowed armies to get ever larger, battles longer.

    Why then have the casualties suffered by Western forces in wars since 1945 tended to fall? The number of US servicemen who died in the Vietnam War was "only" 57,939; the number killed in Korea 37,904. And the death toll has continued to decline. In the Gulf War there were 148 American combat deaths, excluding victims of accidents and "friendly fire": a tiny proportion of a total force numbering 665,000. In the 1999 war against Serbia the figure was precisely zero. Compare those figures with the body counts in the two world wars: 114,000 American servicemen in the First World War and 292,100 in the Second. The drop in military casualties is even more marked in the case of Britain: 720,000 Britons lost their lives in the First World War; over 270,000 in the Second; yet in the Korean War just 537 British soldiers were killed. All told, 719 British soldiers have been killed in Northern Ireland since "the Troubles" began in 1969, along with 302 members of the Royal Ulster Constabulary. Just 24 UK servicemen were killed in the Gulf War, not including 9 killed accidentally by their own side.

    The answer lies in the nature of the wars fought since 1945—which have invariably been against far less well-equipped opposition. These death rates do not, however, signify a decline in the destructiveness of modern weaponry. As we have already seen, there was no shortage of wars in the rest of the world in the second half of the twentieth century. Indeed, according to one estimate, the total war-induced death toll for 1945-99 lies somewhere between 15 and 20 million. The world has not become that much more peaceful. It is just that the overwhelming majority of the victims of war have been Asians and Africans.

    Moreover, the wars that have been fought since 1945 have given barely a glimpse of the colossal increase in destructiveness achieved in the past half-century. A simple calculation suffices to give an illustration of the potential for military catastrophe that still existed shortly after the end of the Cold War. In January 1992 the deployed strategic nuclear forces of the two superpowers had a combined "yield" of at least 5,229 megatons; and this was after a 22 per cent reduction in the total number of superpower warheads since the peak in 1987, and excludes non-strategic nuclear warheads. Since the 12-15 kiloton bomb dropped on Hiroshima in 1945 killed around 100,000 people instantly and a further 100,000 subsequently through radiation sickness, the superpowers in 1992 had the notional capacity to destroy (with their strategic forces alone) 387,302. Hiroshimas or 77.5 billion people. To put it another way, given that the Hiroshima bomb destroyed around 4.7 square miles, the superpowers had the capability to lay waste to 1.8 million square miles, an area rather larger than the state of India. It is scant consolation to reflect that this amounts to just 3 per cent of the planet's land surface, since the contamination after such a conflagration would spread much further. Given that the population of the world in 1992 was approximately 5 billion, nuclear weapons gave the superpowers the notional ability to destroy the entire human race fifteen times over. Any assessment of the changing cost of defence needs to take account of this astonishing increase in the destructiveness of weaponry.

    Also relevant to such an assessment is the way techniques of mass production have tended to lower the unit cost of almost any new piece of hardware. Because of the relative lack of competition in the arms market—with governments the biggest buyers and a small number of huge producers enjoying more or less privileged positions in their home markets—the defence industry has acquired a reputation for excessive pricing. This reputation was certainly merited in the United States and Britain in the 1980s, when public attention was drawn to such puzzling phenomena as "cost-plus contracts" and gold-plated taps in admirals" baths. But over the long run, and considering all levels of armament, the theory that the price of arms tends to rise above the price of consumer goods looks unsustainable. The Second World War in particular showed how techniques of mass production could dramatically reduce the unit-cost of guns, tanks, planes and even naval vessels. High prices for new aircraft and submarines in the late Cold War period merely reflected the very low quantities being ordered; where there has continued to be a significant demand for defence industry wares, prices do not seem to have been subject to above-average inflation.

    Moreover, the Soviet practice of systematically under-pricing defence goods has left an enduring legacy of cheap weaponry, the main beneficiaries of which have been and remain the guerrilla armies of sub-Saharan Africa, the terrorist groups of Western Europe and the drug gangs of the Americas. At the time of writing, a used AK-47 assault rifle could be purchased in the United States for $700; a new one for $1,395: almost exactly as much as the cost of the portable computer on which this book was written. For around $160 billion—just over half the current US defence budget—every American male between the ages of 15 and 65 could be issued with a new Kalashnikov (or, for that matter, two second-hand ones). And of course the prices for such weapons are substantially lower in the developing world. In the same way, the real cost of a nuclear warhead—and certainly the real cost of a kiloton of nuclear yield—is almost certainly lower today than at any time since the Manhattan Project achieved its goal at a cost of $2 billion 1945 dollars. Converted into prices of 1993, that figure rises tenfold: enough to buy 400 Trident II missiles. The fact that France could almost double its nuclear arsenal from 222 warheads in 1985 to 436 in 1991 while increasing its defence budget by less than 7 per cent in real terms speaks for itself. In terms of "bangs per buck"—destructive capability in relation to expenditure—military technology has never been cheaper.

(Continues...)