Military historians usually ignore detailed investigations into the economic consequences of wars, concentrating upon investigations into the origins, political effects of wars, or the process involved in waging war. Generalizations about the economic impact of war seem difficult to validate with reference to empirical evidence which might represent a measurable difference that active involvement in armed conflict with rival powers made to the long-term progress of a nation's economy.
Counterfactuals are clearly implicit in the enquiry. What might have occurred without participation in war is a question that arises for every sector of any national economy we begin to consider. War may delay, accelerate, arrest, or totally transform economic development. Occasionally (and this may be the case for ‘smaller’ wars) its influence will turn out to be neither here nor there.
Counterfactual modes of reasoning emerge explicitly in the economists' methods of dealing with connections between war and economic progress. Impatient with history, they cut through detailed investigation by positing the continuation of pre-war trends in rates of growth for national income, industrial production, and consumption per head. They then assert that (1) in the absence of war these trends would have persisted; (2) deviations from trends are imputable to the malign effects of war; (3) observed differences between postulated trends and actual values represent the costs of war; and (4) that the influence of war ceases once national economies are back on their pre-war trends. These methods are used to support the assertion of Arthur Lewis that the world ‘lost' four and a half years' industrial output and five years’ agricultural output from WW I; they led Claudia Goldin and Frank Lewis to suggest that as late as 1913 per capita consumption in the South remained seriously depressed by the effects of the American civil war. Such procedures enabled Simon Kuznets to separate countries that did well out of WW II from those that fared badly.
Historians are unlikely to be convinced that trends in production or consumption can be defined in any ‘scientific way’ or extrapolated forward on the basis of accessible information for runs of so-called ‘typical’ pre-war years. They will not be inclined to accept assumptions that the influence of war ‘ceases’ when an economy is back on some kind of normal growth path. They will insist upon investigating the enduring economic impact of war step by step, sector by sector, institution by institution. Nevertheless macro-statistical exercises performed by economists draw attention to the fact that war usually reduces a country's capacity for steady growth; that deprivation (measured in terms of private consumption foregone) rises to a maximum in wartime but then steadily diminishes as recovery carries economies forward again; and that wars are accompanied by shifts in the relative positions of national economies within the global economy at large. Furthermore choices of time periods for the analysis also predetermine the assessment of historical outcomes. Some attempt at a clear separation should be made between short and enduring effects of war. During and immediately after hostilities, war clearly disrupts ‘normal’ economic activity. By no means everything that occurs during ‘years of conflict and transition’ (peace to war and back to normal again) has significant long-run economic effects upon the growth of national economies; this entry is concerned exclusively with trends and proposes to bypass the ups and downs of economies in wartime.
For the analysis of trends and structural change cost-benefit analysis (as developed by modern economics) provides us with an illuminating taxonomy and vocabulary: what were the ‘costs’ of war, how are they best defined and added up, what were the economic benefits, and how can they be quantified are the most relevant questions to pursue. Most literature on the impact of wars is concerned with their costs. Economic benefits are generally ignored because liberal scholars are reluctant to admit that engagement in war can actually pay. In the mercantilist era positive connections between the successful prosecution of war, political stability, state building, and commercial success were well understood by European statesmen and intellectuals. Latterly, the only material gain that has sustained historical interest is the link between war and technological breakthroughs. Yet not much of that discussion is or could be definitively settled. Spin-offs from military to civilian technology appear to be limited in scope. When set against the redeployment of limited supplies of scientists and engineers and the huge sums of public money allocated to military research and development, it is difficult to make a positive technological case for war without invoking an implausible counterfactual, namely that civilian demand would always be too weak or too diffuse to have imparted a comparable stimulus to technological progress.
Estimates of total costs of war can be found in many books dealing with the question. They operate with commonsensical definitions and offer straightforward balance sheets. Obviously the costs are imputed to those who actually paid the bills; for example, the costs to Italians from their government's decision to participate in WW I. The sums spent by governments on their armed forces are published in their budget accounts and can be refined, moreover, to include only incremental expenditures from engagement in war. Here the easiest assumption to work with is to subtract pre-war expenditures (growing at a certain rate) from actual war and immediate post-war expenditures. The net figures can then be deflated to produce estimates in constant prices for a group of years before a war or discounted from an investor's perspective to produce totals expressed in terms of present or pre-war values. Totals of budgetary expenditures can then be converted to a common currency (dollars or ducats) in order to render total outlays on war comparable across countries.
Two other direct costs are often included in publications dealing with the total costs of war: human casualties and the value of productive assets destroyed and damaged by enemy action. Neither seem easy to define, let alone measure. For example, the published figures on army, navy, air force, and civilian casualties occasioned by war need to be refined in several ways. First, civilian deaths (which turned out to be huge during and immediately after both global wars of the 20th century) must be scaled down to take account of epidemics, famines, and normal death rates among national populations. Furthermore, and since the deaths and wounding of large numbers of soldiers, sailors, and airmen also occurred at some remove from actual combat, figures of dead and wounded personnel as recorded by the armed forces also need to be adjusted for ‘normal’ death and accident rates among the relevant age groups. Finally the gross totals need to take account of ‘birth deficits’ flowing from diminished conception rates, contingent upon the mobilization of young males (and females) into the armed forces. Birth deficits are clearly complex to estimate and if family size was restored (as it was by the baby boom in several belligerent countries immediately after WW II) the allowance for this factor could turn out to be small.
Research and careful calculations might produce rough national totals of the dead and wounded and also of unborn men and women imputable to war. There certainly is considerable interest in comparing these estimates across countries and through time war by war. Just how ‘bloody’ was WW I? Is it true that on a per capita basis more British personnel died in order to contain Napoleon than to defeat Hitler? What was the economic value of all those dead, wounded, and unborn people? The answers rest upon rather simplistic calculations. For example, the arithmetic of such exercises involves (1) converting wounded into dead people by counting two wounded as equal to one dead soldier; (2) estimating the average years of working life potentially available at the time of death, thus a young man killed in the Somme at age 20 hypothetically lost 45 years of employment—assuming full employment for the years after 1918; (3) assessing the value of the skills lost in wartime; (4) multiplying human casualties by appropriate sets of national wage rates, which implies that in 1918 a dead American could be calculated as four to five times the sum imputable to a dead Serb. If these calculations ever became feasible, historians would then be in a position to offer considered estimates of the economic values of ‘lost generations’.
Alas, to estimate how a nation's productive assets (machinery, social overhead capita, housing, transport, equipment, etc.) might have been affected by war seems equally complex. Leaving aside the possibly insurmountable problem of finding usable statistical information, conceptually historians are required to draw up balance sheets of how wars operated to decrease or to increase the pre-war value of capital stocks owned by different societies. On the minus side would be listed the values of destroyed, damaged, and enforced transfers of property to the enemy; reduced rates of repair and maintenance; losses related to lower rates of civilian capital formation in wartime—and that can only be calculated with reference to a posited trend rate of capital formation for the pre-war period. On the positive side the balance sheet should include gains from reparations, from the capture of enemy territory and other assets, from the demise of competition and the enhanced security of capital flowing from victories over enemy powers. For example, success in the battle of Britain, 1940-1, made the world safer for British capital and that should in theory have been reflected in the values of Britain's capital stock from 1942 onwards.
Clearly it looks almost impossible to draw up comprehensive balance sheets of gains and losses on the nation's ‘capital account’. What precisely was the value of Malta acquired by Britain at the Treaty of Vienna in 1815? By how much was the value of the productive wealth of the USA appreciated by the demise of Japanese and German competition in 1945? How far were the transaction costs of protecting commerce and capital reduced from 1815 to 1914 by the British naval supremacy following from Nelson's victory at Trafalgar in 1805? What we can construct (with the aid of data included in reports from reparations commissions and other official sources) are estimates for the costs of damage and destruction to property—particularly to houses, transport equipment, and livestock. Historians appreciate from studies of reconstruction (now available for some modern wars) that stocks of capital could be restored to pre-war levels fairly rapidly after the cessation of hostilities. They are, however, some way from measuring national capital gains and losses imputable to war.
The rough and ready and incomplete ‘guesstimates’ now in print for the total costs of wars (which include estimates for governmental expenditures on war together with some best ‘guesses’ of the net values of human and physical capital destroyed by military conflict) still remain difficult to interpret. For example, there are numbers which suggest that total global costs of WW I came to somewhere between $300 and $400 billion, that Britain spent £600-700 million to defeat Napoleon, and that the American civil war cost something like $10 billion. War might be compared with war, by expressing total costs as multipliers of pre-war national incomes. For example, the American civil war cost 1.2 times the American gross national product for 1861; to defeat France, Britain spent 4 to 5 times its national income for 1801; and WW I used up 5 times Europe's national output for 1913. Historians have not yet calculated the costs of global warfare from 1939-45 but that war was almost certainly ‘more expensive’ than its predecessors. Such costs could be expressed in dollars per head of population country by country. The figures could then be compared with average incomes at the time and reviewed across countries to ascertain, for example, if the British or the French made ‘greater sacrifices’ to defeat the Germans and juxtaposed to see whether Hitler was a more expensive menace to restrain than Napoleon.
Interesting exercises, but for purposes of connecting the outcomes of wars to the long-term economic progress of participating nations, the figures remain conceptually flawed, and underspecified. They do not, for example, include the ‘opportunity costs’ (or output forgone or gained) from the manpower and investible funds used to wage war. Manpower in the form of statistics for the numbers of civilian workers recruited into the army and navy can be considered. Such numbers (less mercenaries and foreigners) are normally expressed as proportions of a national workforce. For example, by 1810-11 the British government had mobilized some 4 to 6 per cent of its male workforce. In 1917 that proportion was much larger—30 per cent at least. Such ratios are revealing but what historians wish to establish as another cost of war is the net loss of output that flows from the reallocation of civilians into the armed services. That depended upon how easily normal peacetime output was made up: by recruitment from the ranks of the unemployed and from among those normally classified as outside the labour force (particularly women), by lengthening the work year, by skill dilution, and by increased mobility of labour. Substitutions for men at the front tended to occur only in circumstances of war but they compensated for potential bottlenecks in production. National economies rarely entered wars in states of full employment. War not only operated to take up ‘the slack’, it increased participation rates particularly among women and inculcated skills and attitudes which in several ways enhanced the long-term efficiencies of the post-war workforces. Such conjectures will be difficult to quantify but it is not clear that recruitment into the services during wars fought from 1793 to 1945 seriously depleted the supplies of productive labour available to national economies. For countries the effect of war on labour as a factor of production could have been positive for long-term growth.
Turning to capital, the relevant question is what were the effects of war on the rate at which societies accumulate stocks of assets for long-term civilian production? The effects of damage, destruction, enforced transfers, and changes in the security of property (both malign and benign) still seem unmeasurable. It is also clear that urgent military demands stimulated the formation of productive capacity, some of which is useful only for the specific purposes of combat and becomes surplus, with limited scrap value after the ending of hostilities. Historians do not know how much capital (created for the armed services) was reallocated into civil production at the end of wars, or continued to be used by the forces for the protection of a nation's security and its wealth. Expenditure on war cannot be simply represented as ‘unproductive’ for long-run growth.
Thus assessments of connections between investment and warfare turn largely on the contested issue of what happened to net civilian capital formation during and after major wars. How far was private investment depressed by the demands of states for the funds required for enlarged military expenditures? In some theoretical models of the wartime fiscal process, borrowing by governments crowds out capital formation for the civilian economy. States bid away available supplies of investible funds and private capital formation slumps. By the end of hostilities the economy is left with a depleted stock of capital compared to the level it would (counterfactually) have attained in the absence of government borrowing undertaken to engage in warfare.
Damaging qualifications can be made to the crowding-out hypothesis in its more extreme forms because net capital formation clearly continued in several but not all sectors of wartime economies. Rates of private savings which flowed as loans to the state or continued to support gross domestic investment expenditures tended to rise in wartime and private investment has no clear inverse relationship to government borrowing over the short run. Economists have protected their models with counterfactuals which postulate that (in the absence of war) the share of gross investment to gross national income could have risen but this defence seems difficult to substantiate and the rise in national savings ratios seems too large to be explained away by some latent investment potential creamed off to war finance. Furthermore, after wars mini-booms and the rather rapid recovery of production do not suggest that stocks of productive capital had been ‘severely’ depleted by the reallocation of investible funds into expenditures on the armed forces.
Instead of crowding out there may be a paradox to explain: how did it come about that flows of savings for state and private investment could be maintained at surprisingly high rates both in wartime and over the immediate post-war years? Historians now expect to find elements of an explanation in two general observations. First, public expenditure on rearmament and war carried several economies to full employment. In some industries wartime demand led to more intensive and efficient use of resources and to the more rapid diffusion (not, however, to discovery) of advanced technology in order to cope with pressures on existing capacity. Thus, to some extent funds borrowed for the armed forces came from the extra output that had emerged as a response to intensified military demands. In this Keynesian perspective, war pays for itself and potentially adverse effects of crowding out are mitigated. Secondly, savings rates appear to have been unusually responsive to rather modest upswings in interest rates; although prospects for longer term capital gains also helped—provided governments could maintain confidence in victory or a satisfactory peace treaty. Patriotism (or, as the economists might prefer, the heightened need for the protection of private wealth) operated to increase savings by the rich. Wartime taxes on luxuries and shortages of imports and other goods, normally consumed by the affluent classes, also reinforced propensities to save.
Furthermore, and here the argument ventures into the realms of political economy, wartime taxation is regressive in its social incidence. Thus any accumulation of public debt redistributed income from social groups with lower to those with higher propensities to save and invest. While wartime inflations (which result from failures by states to raise anything but a small proportion of the money required to pay for wars by taxation) tended to be marked by the lag of wages behind prices which again exercised regressive effects on income distribution and positive influences upon investment. Finally financial intermediation often improved in wartime. Banks and other institutions operating within a framework of an unconstrained expansion in the money supplies attracted and channelled savings more efficiently into the coffers of governments without rationing credit to businessmen. Institutional innovation managed to tap latent potential for savings.
To sum up: in time of war a state's overall economic strategy seems simple to represent. Governments endeavour to maximize the growth of national income and to acquire the largest possible share. Consistent with that objective statesmen and their advisers implement war finance policies designed to squeeze consumption and to maintain investment upon which future loans and tax revenues depend. Expressed as a share of gross national expenditure private consumption is normally severely depressed and remains so for several years after the war while governments endeavour to reconstruct economies and clean up the mess left by wartime inflations and the accumulation of national debts. Investment is usually less constrained than consumption, continuing the socially regressive trend.
Economic outcomes flowing over time from participation in wars between states cannot be meaningfully analysed just in terms of the impact on labour, capital formation, and technological progress within particular belligerent countries. Between 1688 and 1939 national economies were no longer autonomous entities. Increasingly they operated within an international economy and became steadily more integrated through flows of trade, capital, labour, and technology across frontiers. The relative significance of these interconnections for particular countries can be estimated in the form of ratios of foreign trade to total outputs.
Wars dislocated international commerce in familiar ways by blockading, attacking, and effectively constraining foreign trade. Armed conflict promoted autarky and the substitution of domestic production for imports. Except on a limited scale and largely for military purposes, flows of capital and labour across frontiers dried up in wartime. Finally the fiscal and financial processes deployed by governments to fund their engagements in armed conflict damaged the international monetary system upon which global commerce depended. After every international war, in circumstances of recrimination, heightened tendencies towards autarky, and radical changes in the structure of trade (often precipitated by conflict), an international economic order needed to be restored. For long-term development the reconstruction of that order can be represented as more urgent in 1945 than it was in 1648 and it was probably of the greatest historical importance in 1918.
To conclude: wars and particularly major wars from 1630 to 1939 represent major discontinuities in history that have exercised profound influences on the political, social, and economic development of nations. The origins and causes of these great power conflicts seem less difficult to analyse than their effects. Their longer term economic outcomes for national economies and the world economy as a whole seem problematical both to conceptualize and to measure. Classical political economy is marked by a laudable bias against armed conflict but liberal economists are prone to exaggerate the malign effects of war on manpower and capital formation. War also clearly disrupted and dislocated international economic relations and pushed national economies away from specialization. Since none of the macro-elements of the connections between wars and long-term economic growth have been or could be properly measured, historians may be inclined to speculate that losses from unrealized but potential gains from trade may have been the most extreme economic outcome of global wars in the 20th century. That same effect could also have been the most malign consequence of the French Revolutionary and Napoleonic wars.
Bibliography
- Goldin, Claudia, and Lewis, Frank, ‘The Economic Cost of the American Civil War’, Journal of Economic History, 35 (1975).
- Kuznets, Simon, Postwar Economic Growth (New Haven, 1954).
- Lewis, Arthur, Economic Survey, 1919-1939 (London, 1953).
- O'Brien, Patrick K., ‘The Impact of the Revolutionary and Napoleonic Wars, 1793-1815, on the Long Run Growth of the British Economy’, Review, 12 (1989).
- Silberner, Edward, The Problem of War in 19th Century Economic Thought (Princeton, 1946)
—
