Even in Britain war and the suspension of gold convertibility in 1797 led to inflation, though the scale was much less than on the continent – prices rose by around 80 per cent between 1797 and 1818, and by 1822 had more or less returned to their pre-war level as a result of the return to gold. Unlike on the continent, there was confidence throughout the war that the authorities had the intention and the means to return to gold convertibility after the fighting was over.59
The nineteenth century is usually seen as a time when the spread of the gold standard more or less eliminated the possibility of debt-reduction via currency depreciation. This is not quite true. The American Civil War saw an assignat-style inflation in the states of the Confederacy, and a depreciation of the paper ‘greenback’ in the Union too, with corresponding reductions in the real value of public debts.60 The convertibility of the dollar was suspended from 1862 until 1879, and even after that doubts about the American commitment to gold persisted into the 1890s. War and internal crisis also tended to undermine the attempts to peg the Russian and the Austrian currencies to silver, forcing governments to monetize deficits. Between 1847 and 1853, for example, short-term debt rose from 8 per cent to over 25 per cent of total Austrian debt.61 In the three years 1849–51 high powered money also rose by 25 per cent; while the cost of living peaked in 1854 at 29 per cent above its level seven years before. There were similar problems as a result of the three wars Austria fought between 1859 and 1866.62 Italy too was off gold from 1866 until 1883, and again after February 1894; in 1883 Spain also suspended convertibility. Between 1880 and 1914 Chile, Argentina and Brazil all suffered currency depreciation of between 50 and 80 per cent.63 Even within the gold system, a measure of inflationary debt ‘relief’ was possible. The fact that a large number of European countries saw their debt burdens fall from the 1890s until 1914 has been attributed not only to higher growth rates but also to the global expansion in the gold supply and hence higher inflation (compared with negative rates in the 1870s and 1880s).64
Nevertheless, the significance of this mild inflation was minimal compared with what happened after 1914, when specie payments were suspended by nearly all the First World War combatants (Japan and South Africa excepted) and deficits were financed to varying degrees by resort to the printing press. The extreme case was that of Germany, where wholesale prices rose between 1914 and 1923 by a factor of around 1.3 trillion. Something of the shock this inflicted on ordinary people used to the stability of gold-backed marks can be gauged from Elias Canetti’s memoir of life in Frankfurt in 1923:
It was more than disorder that smashed over people, it was something like daily explosions…. The smallest, the most private, the most personal events always had one and the same cause: the raging plunge of money…. I [had] regarded money as something boring, monotonous … But now I suddenly saw it from a different, an eerie side – a demon with a gigantic whip, lashing at everything and reaching people down to their most private nooks and crannies.65
For some it was too much: the great art historian Aby Warburg suffered a nervous breakdown and was haunted thereafter by visions of cultural devaluation, with art reproductions being churned out like banknotes from the printing press.66 Others, however, remembered their Goethe. The Nordwestdeutsche Zeitung even published a topical parody of Mignon’s Song from Wilhelm Meisters Lehrjahre:
Do you know the land where the currencies bloom,
[Where] in dark night the clip joints shine?
An icy wind blows from the nearby chasm –
Where the Mark stands low and the Dollar high.67
Yet, as Table 4 shows, inflation was an almost universal phenomenon after 1914, affecting even neutral states. In addition to Germany, four countries – Austria, Hungary, Poland and Russia – all suffered something that can be described as hyperinflation.68 The timing of stabilization also varied: in most countries, prices had stopped rising by 1921; but in the more extreme cases, inflation continued into 1922 and 1923. Italy, France and Belgium were unusual in that inflation continued until around 1926, but never boiled over into hyperinflation. The French experience was something of a helter-skelter, with annual inflation peaking at over 50 per cent in 1920, then turning negative in 1921, peaking again at over 40 per cent in 1926, then turning negative again in 1927. France experienced deflation for most of the period between 1930 and 1936; but inflation soared back above 30 per cent in mid-1937.69
The causes of post-war inflations, though complex, were undeniably rooted in the short-term borrowings of governments and their monetary financing by central banks. Inflation only stopped when it was clear that these practices would cease – which (especially in the countries that suffered hyperinflation) necessitated a substantial ‘regime change’, meaning a change of the monetary and fiscal policy regime. In the Italian case there was a change of political regime as well.70 The consequences of high inflation were also in large part fiscal. Above all, the divergent paths of inflation had radically different effects on the real debt burdens of the countries concerned. In Britain and the United States the decision to return to the gold standard at the pre-war exchange rate required deflation. Despite some debt repayment, the combined effect of falling prices and reduced growth caused substantial increases in the real debt burden. Between 1920 and 1931 the nominal value of the British national debt was reduced by around 5 per cent; but the real debt burden, allowing for deflation, rose by a staggering 60 per cent. In the United States in the same period, debt repayments and deflation simply cancelled one another out, leaving the real debt burden unchanged. Yet countries which went down the inflationary road emerged with much, if not all, of their internal war debt gone. In the extreme case, German public debt was reduced to virtually zero in 1923. Although subsequent ‘revaluation’ legislation did something to compensate the holders of pre-war bonds – like the hapless women suffragists – the same treatment was not accorded to war bonds.71 In a parody of the motto on the soldier’s Iron Cross, the German public gave their gold in return for worthless paper.72 Somewhere in between lay countries like France and Belgium. In France the total internal debt rose in nominal terms between 1920 and 1929 by about 37 per cent. But in relation to net national product it fell by almost exactly the same amount.73
Table 4. European price inflation during and after the First World War
* * *
Peak of wholesale prices in terms of paper currency (1914 = 1)
Date
* * *
Switzerland
2
1921
Spain
2
1920
Netherlands
3
1919
Denmark
3
1920
UK
3
1920
Sweden
4
1920
Norway
4
1920
Italy
6
1926
France
7
1926
Belgium
7
1927
Finland
12
1921
Czechoslovakia
14
1921
Austria
14,300
1922
Hungary
23,466
1922
Poland
2,484,296
1924
Russia
4,146,849
1923
Germany
1,261,600,000,000
1923
* * *
Sources: Mitchell, European Historical Statistics; Bresciani-Turroni, Economics of Inflation, pp. 23 f., 161–5; Capie, ‘Conditions in which Very Rapid Inflation has Occurred’, table 6; Sargent, ‘Ends of Four Big Inflations’, tables.
In many ways, this story repeated itself during and after the Second World War. In Germany there was an even steeper increase in both public debt and pa
per currency, and only strict price controls prevented an inflationary explosion during the last two years of the war. When the regime collapsed in 1945, the reichsmark went with it almost immediately, and was followed with astonishing rapidity by the occupation currency printed by the Americans and (in excessive quantities) the Soviets, forcing victors and vanquished alike to improvise with cigarette money and other substitutes until the currency reform of June 1948. Other countries which experienced very high post-war inflation were Greece, China and Hungary; in two of these cases civil war was a primary cause of the problem.74 By contrast, Britain succeeded in keeping monetary expansion and inflation below the First World War levels: prices rose by just over 50 per cent relative to 1938.75
Between 1914 and 1945 the world veered between inflation and deflation. With only a few exceptions – American consumer prices fell by small amounts in 1949 and 1955, for example, and Japan experienced slight deflation of less than half of one per cent in 1980, 1995 and 1999 – the world since 1945 has been inflationary, though with distinct phases of low and high inflation. In the 1950s and 1960s most economies experienced mild inflation under the gentle strictures of the Bretton Woods system (see Chapter 11). In the 1970s and 1980s, however, the breakdown of that system led to a more or less global adoption of paper money. The consequence was a general increase in inflation, though there was considerable variation between countries, depending on the way fiscal and monetary authorities reacted to the higher oil prices imposed by the OPEC cartel in 1973 and 1979. (To give an impression of this variation, compare the average inflation rates for Germany, the United States, Britain and Portugal between 1961 and 1999, which were respectively 3.3, 4.6, 7.1 and 12.0 per cent.) Since the late 1980s, however, there has been a marked decline in inflation rates in most countries. Portuguese inflation, which exceeded 50 per cent in May 1977, fell below 3 per cent in 1999. French inflation, which reached 14 per cent in November 1981, fell to just 0.2 per cent. A few bold commentators have even ventured to speak of the ‘death of inflation’.
Figure 10. British money supply and inflation (annual growth rates), 1871–1997
Source: Goodhart, ‘Monetary Policy’; Capie and Webber, Monetary History.
RULES AND DISCRETION
The great variations in inflation over time and between countries are perhaps as well explained by institutional changes as by universal economic laws like the quantity theory of money or its derivatives.76 Figure 10 presents long-run evidence on British consumer price inflation since 1871, showing that there are indeed rough correlations between the inflation rate and monetary growth rates (that is, the rate of growth of the money supply, whether defined narrowly to include just notes and coins in circulation, or broadly to include bank deposits).77 But the relationships have clearly changed as the nature of money and the institutions that generate it have evolved. A good example of the difficulties that confront a narrowly monetarist interpretation of inflation is the divergence between broad money and inflation in the mid-1980s, a period when, ironically, government policy was avowedly monetarist.
An institutional approach emphasizes the changing role played by central banks, in particular the fundamental difference between ‘rules’ and ‘discretion’. In the first instance, as we have seen, most note-issuing public banks existed to help governments finance their mainly war-induced deficits. However, the gold standard evolved as a system designed to limit the discretion of central banks to lend too freely in peacetime. Only gradually did the idea evolve that the central bank should be responsible for the management of the currency and the stability of the banking system as a whole.78
In the theory developed by classical economists, price stability was not the raison d’être of the gold standard. Rather, the appeal of maintaining a fixed rate between gold and the currency was that it automatically kept the international and domestic economy in equilibrium by relating the domestic money supply to the external balance of payments. According to the ‘price-specie-flow’ theory first propounded by David Hume, an outflow of gold would act on the domestic price level, causing it to fall, while at the same time raising the external price level, leading to an increase in exports, a reduction in imports and a reflux of gold.79 Under the nascent ‘rules of the game’ (the phrase was not in fact coined until 1930), the Bank of England was supposed to respond to such an outflow of gold by raising its discount rate,80 thereby restricting credit, so as to maintain the ratio between notes and gold. The resulting monetary tightening would, in theory, reduce prices in Britain relative to the rest of the world and therefore increase the competitiveness of British exports, while at the same time depressing domestic demand for imports. This was the underlying rationale behind Sir Robert Peel’s Bank Charter Act of 1844, which separated the Bank’s note-issue department from its commercial banking operations and imposed a fixed one-to-one ratio between gold and the note issue beyond a fixed quota (initially £14 million).81
It is important to distinguish between the formal statutory rules governing the gold reserve and note issue and the unwritten ‘rules of the game’. It is often assumed that the rules were simply that the bank should raise its discount rate when the gold reserve diminished and lower it when it increased. This was not always the case. As far as the Bank was concerned, ‘the rate of discount charged … [was] regulated more by the proportion of the reserve82 to liabilities than by any other consideration’.83 Changes in this proportion were monitored on a daily basis, though Bank rate was announced weekly when the Court of Directors met. In addition the Governor could order an increase (or decrease) in the rate at any time on his own authority, as happened in the 1907 crisis. Modern research has confirmed that changes in the gold reserve were indeed the principal determinant of changes in Bank rate.84 However, the Bank’s reaction to changes in its reserve was not perfectly symmetrical. As a spokesman put it in a statement to the American National Monetary Commission in 1909: ‘The Bank rate is raised with the object either of preventing gold from leaving the country, or of attracting gold to the country, and lowered when it is completely out of touch with the market rates and circumstances do not render it necessary to induce the import of gold.’85 The Bank Directors also took into account the movements of foreign (mainly European) exchange rates, on the ground that these acted as an indicator of impending reserve changes.86
Nor should it be assumed that the Bank was ‘setting’ short-term interest rates for the money market as a whole. In his classic account, Lombard Street (1873), Walter Bagehot questioned the extent of the Bank’s influence over the market:
The value of money is settled, like that of other commodities, by supply and demand … A very considerable holder of an article may, for a time, vitally affect its value if he lay down the minimum price which he will take, and obstinately adhere to it. This is the way in which the value of money in Lombard Street is settled. The Bank of England … lays down the least price at which it will dispose of its stock, and this, for the most part, enables other dealers to obtain that price, or something near it…. The notion that the Bank of England has control over the Money Market, and can fix the rate of discount as it likes, has survived from the old days before [the Bank Charter Act of] 1844… But even then the notion was a mistake.87
Bagehot thought it desirable that the Bank should increase its control over the money market. But for much of the period before 1914 it clearly struggled to make its rate ‘effective’.88
The most common contemporary explanation for this was the decline in the Bank’s size relative to the rest of the financial sector, particularly joint-stock banks. Between 1826 and 1858 the Bank’s original monopoly as the country’s only joint-stock bank was whittled away, allowing the growth of large commercial banks (which together developed the clearing system) and discount houses (which worked in the market for commercial bills).89 In theory the Bank of England still had ‘the largest paid-up capital of any bank in the world’ even after the turn of the century (£14.5 million, plus a further £3 million
of ‘accumulated and undivided profit’). But this was not vastly greater than the biggest of the City’s merchant banks, N. M. Rothschild & Sons, which had total capital of £8.4 million in 1905. Indeed, the Bank of England was smaller than the Rothschild bank if one adds together the Rothschilds’ London, Paris and Vienna houses, which formed a united partnership with around £37 million capital until that date.90 Moreover, the growth of joint-stock commercial banks, which seldom borrowed from the Bank, further reduced its leverage.91 For the years 1894 to 1901 the Bank’s reserve averaged just over 3 per cent of the deposits, current accounts and note circulation of all UK banks.92 This alarmed contemporaries. Palgrave was only one of many critics who urged ‘the attainment of really sufficient reserve’. In vain: it remained a ‘thin film of gold’.93 In addition to Bank rate changes, the Bank therefore had to evolve a variety of supplementary devices designed to make its rate ‘effective’: prototype open market operations (mopping up excess cash in the money market by selling consols ‘spot’ and repurchasing them forward); borrowing from major customers like the India Office, the Bank of Japan, or even (as in 1905–6) from the clearing banks; curtailing its loan and rediscount facilities to the market; and manipulating its buying and selling prices for foreign gold (bar and coin).94 There is even some evidence that the Bank occasionally reacted counter-cyclically, cutting rates to mitigate commercial downturns.95 Indeed, it sometimes reduced its reserve as interest rates went up, the very opposite of the sequence required by the rules of the game.96 In all this, long-run price stability was a mere by-product of monetary policy. Indeed, short-run instability was a corollary of pre-1914 monetary policy (a point we shall return to in Chapter 11).
The Cash Nexus: Money and Politics in Modern History, 1700-2000 Page 20