The Cash Nexus: Money and Politics in Modern History, 1700-2000

by Niall Ferguson

  In the wake of the Napoleonic Wars, there was a sustained drive in most states to reduce the number of public employees. Reductions in indirect taxation were not only justified on the basis of laissez faire, laissez passer, but also as a means of shrinking the tax-gathering bureaucracy. Although much has been written about the modernization of government in the mid-nineteenth century, the statistics make it clear that for most of the century the ‘night-watchman state’ was a reality. In 1891 total government personnel amounted to less than 2 per cent of the total labour force in Britain. The figures on the continent were higher, but not by much. For Italy in 1871 the equivalent figure was just 2.6 per cent; for Germany in 1881 3.7 per cent. Even the famously elaborate Habsburg bureaucracy was small in relation to the swelling population of the Empire. But from the turn of the century onwards there was a sustained growth in the public sector almost everywhere. By the 1920s public employment exceeded 5 per cent of the workforce in Italy, 6 per cent in Britain and 8 per cent in Germany.

  In his monumental Economy and Society Max Weber portrayed the modern bureaucracy as admirably rational: ‘rules, means, ends, and matter-of-factness dominate its bearing’.64 Yet even as he wrote, disillusionment with bureaucracy was growing, not least in the wake of the enormous expansion of the public sector during the years of war and inflation, a phenomenon more closely associated with proliferating red tape and corruption than with rationality. The reality of modern bureaucracy turned out to be closer to Kafka’s Castle, in which enigmatic files are trundled up and down grey corridors, being allocated apparently at random to faceless pen-pushers behind identical office doors.65 The Beamte – once admired as the epitome of Prussian virtue – became the personification of sloth and self-interest. During a violent political riot in Vienna in 1927, Elias Canetti vividly recollected seeing a distraught official outside the burning Palace of Justice, ‘flailing his arms and moaning over and over again’:

  ‘The files are burning! All the files!’

  ‘Better files than people!’ I told him, but that did not interest him; all he could think of was the files…. He was inconsolable. I found him comical, even in this situation. But I was also annoyed. ‘They’ve been shooting down people!’ I said angrily, ‘and you’re carrying on about the files!’ He looked at me as if I weren’t there and wailed repeatedly: ‘The files are burning! All the files!’66

  The files – die Akten – had become an end in themselves.

  Significantly, bureaucracy was one of Hitler’s bugbears.67 Dining with Himmler in January 1942, he outlined a characteristically infantile scheme to ‘reduce the bureaucracy to a third of its importance’ by simplifying the German tax system:

  As regards direct taxes, the simplest is to take as a basis the amount paid the previous year. The tax-payer is told: ‘You’ll pay the same sum as last year. If this year your earnings are lower, you’ll report the fact. If they’re higher you’ll immediately pay a proportionate supplement. If you forget to announce the increase in your income, you’ll be severely punished.’… Everything could be done by means of an extremely simple piece of apparatus, and the Chinese puzzle of declaring one’s taxes would be done away with….

  If I explain this system to the Ministry of Finance … the reply will be, after an instant’s reflection, ‘My Führer, you’re right.’ But within six months they’ll certainly have forgotten everything.

  Hitler subscribed to the conventional view that bureaucracy is self-perpetuating. ‘The snag is that a tax which is easy to collect doesn’t suit these gentlemen of the administration. What would be the use of having been to a University? Where would one find jobs for the jurists? There’d be no more work for them.’ Yet the way Hitler himself fomented competition between overlapping state and party institutions tended to encourage bureaucratization, as he himself all but admitted:

  One decides to create a group of the Hitler Youth at Salzburg. Suddenly they need a building of five hundred rooms…. I created the Ministry of Propaganda with the idea that it would be at everybody’s service…. Yet there practically doesn’t exist a Ministry today that hasn’t its own press-service…. Göring wanted to get from me a decree conferring powers on Stuckart and Reinhardt [the Ministers of Finance and the Interior] so that they could undertake the reorganization of our administrative services with a view to simplifying them. I refused. Why entrust these men with such a mission when it’s precisely the Ministr[ies] of Finance and Interior … whose administrations are plethorically swollen?68

  Figure 7. Government employment as a percentage of total employment, 1960–1999

  Source: OECD.

  In the Soviet Union, of course, there was no other employer than state and party; and that remained true until the 1980s. Yet it was not only in totalitarian regimes that the public sector tended to expand. By 1950, thanks mainly to nationalization, the proportion of public employees in Britain had risen above 10 per cent of the workforce.69 Figure 7 shows that this growth has since continued in more or less every developed country and has been substantially reversed in only one. In twelve out of seventeen OECD countries, government employment has scarcely declined at all since reaching a peak in the mid-1990s. In Sweden, Norway and Denmark the proportion of total employment which is in the public sector is in excess of 30 per cent. In France, Finland and Austria the figure is above 20 per cent. Portugal, Spain, Italy and Germany all saw public sector employment exceed 15 per cent of the workforce in the 1990s; Switzerland and Greece are not far behind them. The exceptions to this pattern are the United States, where government employment peaked in the mid-1970s, and Britain, Ireland, Belgium and Japan, where the peak occurred in the 1980s. But only in Britain has there been a significant decline since that peak. In 1983 government employment reached a post-war high of 22 per cent of total employment; in 1999 the figure had fallen to 13.6 per cent, lower even than the figure for the United States. Among developed economies, only Japan and Greece have smaller public payrolls.

  Whatever the macroeconomic costs of high levels of public employment, the immediate fiscal problem lies in determining the pay of public employees both in the absence of the kind of information that allows productivity to be measured in the private sector, and (usually) in the presence of public sector unions and other pressure groups agitating for pay rises ahead of inflation. The sheer size of the wage-bill is staggering. In 1992 total public sector pay accounted for fully a third of general government expenditure in Britain.70 In the United States the figure is a fifth.71 Small pay rises can therefore have very large fiscal implications. Indeed, one of the paradoxes of modern democracy is the tendency of governments to respond to criticisms of public services by increasing public pay. In fact, to give a British example, pay rises for nurses in the National Health Service may imply real reductions in the amounts of money spent on hospitals, beds, equipment and medicines, and therefore a further deterioration in patient-care – though an improvement, no doubt, in nurses’ living standards.72

  SERVILE STATES

  States have long been able to secure substantial portions of national income by taxation: it is quite wrong to think that high tax burdens are a phenomenon of the twentieth century. The total revenues of the Abbasid caliphate in the late eighth century amounted to between a sixth and a quarter of national product.73 Venetian tax revenues at the end of the sixteenth century amounted to between 14 and 16 per cent of GNP;74 while the total revenues of the United Provinces around 1688 were equivalent to around a quarter of national income.75 According to one calculation, the tax burden in France as a percentage of national income declined from 18 per cent in 1450 to just 10 per cent in 1525; but then rose rapidly in the seventeenth century to 31 per cent in 1683 and reached as much as 38–40 per cent in 1789.76 The great economic historian Alexander Gerschenkron estimated the Russian tax burden at roughly two-thirds of the entire grain harvest in 1710, a level of fiscal extraction not seen again until Stalin’s time.77

  The history of Britain’s rise to great-power sta
tus is also, and not coincidentally, the history of a rising tax burden. Royal revenue in the reign of Elizabeth I never exceeded 2 per cent of national product, or at most 5 per cent if one includes occasional forced loans, charges by officials and local levies.78 Even as late as 1698, according to estimates by the early political economist Charles Davenant, Britons were paying a smaller proportion of their national income in tax than their continental neighbours: whereas the Dutch paid up to a third of national income in tax and the French a fifth, the British proportion was just an eighth.79 However, in the course of the eighteenth century the British tax burden rose rapidly. Total expenditure as a percentage of national income went up from under 4 per cent in the mid-1680s to peaks of between 17 and 20 per cent in the war years of the eighteenth century.80 Even then, the absolute amount of tax revenue raised in Britain was less than in under-taxed France: it is easily forgotten that as late as the 1780s French GNP was more than twice that of Britain. One reason Britain was able to mount such an effective military challenge to her larger neighbour was her higher rate of taxation.81 As a percentage of GNP, total taxes were nearly double what they were in France in 1788 (12.4 per cent compared with 6.8 per cent). If France had only been able to raise more tax her fiscal crisis might have been averted.82

  As Chapter 1 demonstrated, the principal cause of increases in the level of state expenditure and hence of taxation has, for most of history, been war. In peacetime, expenditure and taxation tended to fall substantially. This was one reason why the nineteenth century was also a time when tax burdens fell to historic lows in most countries. In the twentieth century, by contrast, there was a ratchet effect. In the aftermath of both the world wars, public spending failed to revert to its pre-war level, whether in absolute terms, in inflation-adjusted terms, in per capita terms or in relation to GDP. In 1990 prices, total public expenditure in Britain was £15.3 billion in 1913;£27.5 billion ten years later; and it never fell below £60 billion after the Second World War.83 Moreover, in both real and per capita terms, public spending since 1945 has continued to rise decade by decade, despite the absence of a major war. Even as a proportion of GDP, the trend was upwards until the 1980s and 1990s, when a plateau was reached. In the same way, federal government outlays as a percentage of US GDP were just 16 per cent in 1950, 18 per cent in 1960, 19 per cent in 1970 and 22 per cent in 1980 and 1990.84 The peak for total US public sector outlays came in 1992 (36.6 per cent); they were projected to fall to 32 per cent in 2000. Spending ratios rose even higher in Europe: to peaks of 45 per cent in Britain (1993), 50 per cent in Germany (1995), 55 per cent in France (1996) and 57 per cent in Italy (1993). The highest spending ratios in the developed world were in Scandinavia: the Swedish figure peaked at 71 per cent in 1993, while in Denmark the figure was 60 per cent.85 This was quite different from the experience of previous centuries, when such levels of public expenditure were seen only in time of war.

  The familiar explanation for this almost universal phenomenon is the rise of the ‘welfare state’. But what exactly is the meaning of this well-worn phrase, first used in English by the Anglican bishop of Manchester, William Temple, in 1928? If by the welfare state we mean public spending designed to reduce income inequalities – whether by direct supplements to those on low pay or the provision of services to the poor at below market prices – then that too is not a wholly modern invention. Nearly half the adult male population of fifth-century BC Athens received some form of payment from the state. Around 10 per cent of state spending in Augustan Rome went on ‘doles’ to the Roman plebs. However, most transfers in the medieval and early modern periods were from taxpayers to relatively well-off groups: lawyers, soldiers, arms suppliers and financiers.86 In Mughal India, the emperor and 122 nobles – a tiny fraction of a total population of 110 million – received roughly an eighth of the total national product.87 However, attitudes to poverty in Western societies oscillated, with public relief sometimes being made available in a crisis, but more often the responsibility for care of the needy being left to charity and self-help, leaving the state to perform a more disciplinary role towards marginal groups.

  The extent of what could be done to relieve poverty and sickness without the state’s intervention is often forgotten. In Victorian Britain, ‘Friendly Societies’ were responsible for an astonishing volume of prudential saving. Until just before 1914, spending by registered and unregistered charities, friendly societies, trade unions and other benevolent and self-help institutions was more than the annual budget of the poor law and dwarfed the central government’s expenditure on social welfare.88 As early as 1803 there had been over nine thousand mutual or ‘friendly’ societies, with more than 700,000 members. By 1877, total membership of registered friendly societies had risen to 2.75 million, and less than forty years later it stood at 6.6 million. In addition, more than two million people were members of unregistered societies.89 When national insurance was introduced in Britain in 1911, over three-quarters of those covered by the new scheme were already members of friendly societies. Even thereafter, private insurance grew more rapidly than national insurance: by the eve of the Second World War, premiums on private insurance policies exceeded the total contributions to the state schemes for health, unemployment and pensions.90 Self-help, in short, was more than a pious Victorian aspiration; for a substantial proportion of the working population it was a reality. And its corollary was often a deep suspicion of the interference of the state: in particular (to quote a Longton miner interviewed by the Fabian R. H. Tawney in 1912), irritation at outsiders ‘mak[ing] us ignorant people live in the way they think we ought’.91 It was not only libertarians like Hilaire Belloc who were hostile to ‘the servile state’.92

  Moreover, help for those who could not help themselves was also forthcoming, unprompted by the state. Donations to charities amounted to a substantial ‘voluntary tax’ funding a myriad of good causes, principally in education and health care. In Britain the total income of the registered charities was £13 million in 1910, more than total local authority expenditure on poor relief (£12.3 million); and this figure excludes smaller charities and sums raised informally and distributed by parish churches and Nonconformist chapels. Samples of wills suggest that an average of 13 per cent of wealth was being bequeathed to charities in the years before 1899.93

  Nevertheless, the political arguments for more state activity became irresistible around the turn of the century, thanks to a combination of socialist theory, ‘New’ Liberal repudiation of laissez faire doctrine and conservative fears of the declining ‘national efficiency’ exposed by the Boer War. The Right was as responsible as the Left for the rise in public spending before 1914. In Britain free elementary education and subsidies to Irish peasants – ‘the price we have to pay for the Union’ – were both introduced by Salisbury’s Conservative government. But the real watershed was the Liberal government of 1905–15. The Liberals introduced school meals and compulsory school medical inspections. Adapting a system that had originated in Bismarckian Germany, they conferred a non-contributory old-age pension as an entitlement from the age of 70.94 And for those on lower incomes, they brought in a system of compulsory national insurance against both ill health and unemployment, with the state supplementing employers’ contributions.

  Like many continental systems, national insurance built on existing networks of friendly societies and insurance companies. Nor can it be regarded as a failure. True, dependants were excluded from the scheme, and the Treasury maintained a tight control on payments made and benefits available.95 On the other hand, between 1912 and 1938 the number of people covered by the scheme rose by a factor of more than four. Henceforth, transfers to the old, the sick and the poor were an integral and growing part of total public expenditure.

  The increase in spending was also marked at the local level. In Britain local government expenditure had been held in check for much of the nineteenth century by the New Poor Law of 1834, which effectively deterred all but the desperately poor from clai
ming the austere relief of the workhouse. That began to change in the 1880s. In 1885 a Royal Commission recommended that the London county council be empowered to improve housing in the capital. In 1902 the county councils also acquired responsibility for education. Between 1870 and 1913 local spending increased by a factor of five.96 In Germany the federal system gave even more latitude to state and local governments: their expenditure on education, welfare, health and housing rose steadily, so that all these items accounted for nearly half all public sector spending in 1913.97

  As is well known, the First World War not only increased expenditure on defence, but also expanded significantly the range of non-military governmental activities. In Britain, there were new ministries not only for Munitions and Air, but also for Food (1916), Labour (1916) and Health (1919), not to mention the short-lived departments for National Service and Reconstruction. Although the ambitious schemes for post-war expenditures on ‘homes fit for heroes’ foundered on the rocks of retrenchment, it proved impossible to roll back the state to its pre-war position. In terms of new dwellings completed, the public sector outbuilt the private sector in 1921 and 1922, fell back thereafter but then forged ahead in every year from 1941 until 1959.98 Indeed, the unprecedented unemployment of the inter-war years forced governments everywhere to spend more money, no matter how they strove to avoid it. The pre-war schemes for compulsory insurance could not cope with such high and sustained unemployment (and in countries that had experienced hyperinflation, their funds were largely wiped out). Governments found themselves having to pay doles to the unemployed or using public money to give them work, which seemed the more expensive option. Much used to be made of the power of Treasury orthodoxy in resisting pressure for higher public spending during the Slump. But in terms of public expenditure on transfers and public works of various sorts, the Treasury yielded much ground before 1939.