Two alternatives suggest themselves, both of which merit historical assessment. Should the world attempt to return – with or without gold – to a system of fixed exchange rates? Or does the future lie not with fixed exchange rates but with outright monetary unions like the one established by eleven European countries in 1999?
THE YELLOW BRICK ROAD
In The Wizard of Oz, Dorothy, the Scarecrow, the Tin Man and the Lion have to follow a winding and hazardous ‘yellow brick road’ in order to reach their destination. Few devotees of the classic Judy Garland movie are aware that the original 1900 book by Frank Baum was a satire on America’s entry into the gold standard.23
The road to an international gold standard was indeed more tortuous than is often acknowledged. For most of the nineteenth century two of the five great powers – Austria and Russia – had widely fluctuating exchange rates.24 The United States suspended convertibility as a result of the Civil War in 1862 and remained on a paper standard until 1879. Although the redemption of US bonds was restored to a gold basis in 1869 and the silver dollar was dropped from the coinage in the so-called ‘crime of 1873’, there was a sustained political campaign against gold lasting into the 1890s.25 (In the eyes of the Populists, the gold standard was a British and/or Jewish racket to depress mid-western farm prices and enrich the financiers of Wall Street.) In 1868 only Britain and a number of its economic dependencies – Portugal, Egypt, Canada, Chile and Australia – were on the gold standard. France and the other members of the Latin Monetary Union, as well as Russia, Persia and some Latin American states were on the bimetallic system; while most of the rest of the world was on the silver standard. Not until 1900 was the transition to gold more or less complete. By 1908 only China, Persia and a handful of Central American countries were still on silver. The gold standard had become, in effect, the global monetary system, though in practice a number of Asian economies had a gold exchange standard (with local currencies convertible into sterling rather than actual gold) and a number of ‘Latin’ economies in Europe and America did not technically maintain convertibility of notes into gold.26
How exactly the gold standard worked has been debated ad nauseam by economists. As an international system, its primary function was obviously to fix exchange rates: or, to be precise, to narrow the band of fluctuation down to the so-called ‘gold points’, the rates at which it became profitable to import or export gold.27 As we saw in Chapter 5, the classical model of the system derived from the ‘price-specie-flow’ mechanism, which was supposed to adjust trade imbalances through the effects on relative prices of gold flows. Later theorists discerned a more swiftly adjusting mechanism at work, whereby short-term capital flows responded more or less instantaneously to increases or decreases in discount rates. Keynesians concentrated on the effects of the gold standard on incomes and demand in peripheral countries such as Argentina.28 Monetarists, meanwhile, have sought to show that the gold standard did not operate through changes in the terms of trade because arbitrage kept the prices of all internationally traded goods the same around the world.29
According to the ‘rules of the game’, central banks were supposed to mediate between the international and domestic economy by varying interest rates in response to gold flows. However, there is considerable evidence of outright breaches of these rules, for example by the French and Belgian central banks.30 Some central banks were more likely to increase rates in response to an outflow than to decrease them in response to an inflow; some manipulated the price of gold to change the gold points.31
Table 19. Exchange rate regimes and inflation
Source: Bordo, ‘Gold as a Commitment Mechanism’, pp. 32 f. Inflation defined as the annual mean of the GDP deflator.
What were the benefits of the gold standard? Clearly, fixed exchange rates remove an element of uncertainty from international trade. But there is little reason to think that this was the reason for the increased volume of trade between 1870 and 1914, which might just as easily have occurred without fixed exchange rates. What the gold standard is usually said to have delivered is long-run price stability. The average annual change in wholesale prices between 1870 and 1913 was -0.7 per cent in Britain and 0.1 per cent in the United States. Table 19 shows average inflation rates for twenty-one countries since 1881, subdivided according to regions and exchange regimes. Though it would not be true to say that the pre-1914 gold standard delivered the lowest inflation for all twenty-one countries in the sample – that accolade belongs to the deflationary inter-war period – it is clear that prices were most stable in the industrialized economies when they were on gold.
True, such figures conceal considerable short-run fluctuations and significant movements over shorter periods. Between 1849 and 1873 British prices rose by 51 per cent; then fell by 45 per cent between 1873 and 1896, only to rise by 39 per cent between 1896 and 1913. American prices followed a similar path.32 Nevertheless, statistical analysis reveals these movements to have been a ‘random walk’ or ‘white noise’ process, meaning that they showed no tendency to persist.33 No matter what the short- or medium-term movements of the price level, people could be confident that prices would ultimately revert to their historic mean.34
Was this economically beneficial? In theory (or at least in some theories) price stability always is. And there is some empirical evidence to suggest that the gold standard was. In Britain and the United States, real per capita income was less variable between 1870 and 1913 than it was thereafter. Unemployment was lower than between the wars.35 So too were long-term interest rates, though not necessarily real rates.36 However, it is not certain that these differences can be attributed solely or even partly to the presence or absence of gold. It has been argued persuasively that bimetallism would have been preferable to monometallism: the real ‘crime of 1873’ was the French decision to abandon silver, and hence to give up Paris’s role as the key centre of international monetary arbitrage.37 Moreover, the greater short-run variation of inflation and output under gold may have been more distressing to contemporaries than the lower long-run inflation emphasized by gold standard enthusiasts.38 Keynes’s famous remark – ‘In the long run we are all dead’ – bears repeating: people are generally far more conscious of short-run ups and downs of the economy than of secular movements of prices and output. From the point of view of achieving low and stable inflation and high and stable growth, the Bretton Woods system of which Keynes was a principal architect was superior – though other factors almost certainly contributed more to post-war success than the exchange rate regime.39
The fluctuations in prices mentioned above were partly a result of changes in the global stock of gold. The worldwide shift to gold might have had disastrously deflationary consequences had the supply of gold not proved relatively elastic. In the 1840s world gold production averaged 42 tonnes a year, of which more than half came from Russia. By the 1850s total production had risen to 965 tonnes, with around half the increase from California and half from Australia.40 Thanks to the development of the Rand goldfields in South Africa and the Kalgoorlie field in Western Australia in the 1890s, as well as discoveries in Colorado, the Klondike and Siberia, world gold stocks more than trebled between the 1850s and 1900s.41 As Figure 30 shows, the additions to the world gold stock were far from gradual, ranging from 88 per cent in the 1850s to just 11 per cent in the 1880s.42 Changes in the technology of gold processing such as the discovery of the cyanide process also had an influence.
Figure 30. World gold production, five-yearly totals, 1835–1989 (metric tonnes)
Source: Green, World of Gold, appendix I, pp. 364 f., and ‘Central Bank Gold Reserves’.
At the same time, it is important to recognize that the gold standard was not rigidly bound to the gold supply. Financial innovation, as the use of paper money and cheques became more widespread, loosened the golden fetters. The greater part of the monetary expansion that occurred between 1885 and 1913 was due to the fivefold growth of demand deposits; by comparison gold r
eserves grew only three and a half times.43 Remarkably, given its much-vaunted ‘hegemonic’ position, the UK accounted for just 3.6 per cent of all gold held by central banks and Treasuries in 1913.44 The Bank of England was like a ‘man with little flesh on his bones … [with] only a slim gold reserve surrounding a vulnerable gold standard frame’.45 The disadvantage was that there was a clear trade-off between the size of the gold reserve and the volatility of short-term rates: countries with bigger reserves (such as France) needed to change the discount rate less frequently.46
A rather different justification of the gold standard is that, by removing exchange rate risk and affirming a country’s commitment to ‘sound’ fiscal and monetary policies, it reduced the cost of international borrowing for countries that joined. It was a ‘commitment mechanism’: going onto gold was a way of forswearing ‘time-inconsistent’ fiscal and monetary policies such as printing money to collect seigniorage or defaulting on debt.47 However, so the argument runs, gold convertibility was ‘a contingent rule, or a rule with escape clauses’: it could be suspended ‘in the event of a well understood, exogenously produced emergency, such as a war, on the understanding that after the emergency had safely passed convertibility would be restored at the original parity’.48 In some cases there was a second legitimate exception to the rule: in banking crises (such as 1847, 1857 and 1866 in Britain) the authorities could temporarily suspend the golden rule to act as lender of last resort.49
As has already been suggested, the idea of wars or financial crises as ‘well understood emergencies’ is the weak link in this argument. Its proponents merely infer from statistical data that such exceptions to the gold convertibility rule were ‘well understood’; they provide almost no historical evidence that contemporaries regarded them as such. Nor is it entirely clear why some emergencies were regarded as legitimate reasons for departing from the rule, while some were not.50 It is certainly true that countries bearing the ‘Good Housekeeping seal of approval’ conferred by gold standard membership could borrow money on better terms than those which did not. However, the premiums involved were surprisingly low in view of the very substantial depreciations experienced by Argentina, Brazil and Chile between 1880 and 1914. Already by 1895 the currencies of all three had depreciated by around 60 per cent against sterling. Yet the premiums they had to pay on gold-denominated bonds amounted to no more than two percentage points relative to gold-adhering borrowers like Canada and Australia. The average yield on Chilean paper-denominated bonds over the period as a whole was just 7 per cent, compared with a figure for the United States of just over 4 per cent. But by 1914 the Chilean currency was down to 20 per cent of its 1870 exchange rate; whereas the dollar was at par.51 The limits of this approach may well be revealed by the fact that the same calculations for the inter-war period produce similar results.52 Being on gold was worth between 100 and 200 basis points on a country’s yields. But what were the costs in terms of output of being on gold?
It is also worth noting that a commitment mechanism might be possible without fixed exchange rates. For example, the current fashion for inflation targeting or the adoption of some other form of explicit rule by independent national central banks could amount to a form of credible commitment. However, it remains to be seen how far heterogeneous national targets could ever constitute an international commitment mechanism comparable with the gold standard.53 Given the well-established tendency for monetary targets to drift in the 1970s and 1980s, it seems doubtful that inflation targets will ever attain the credibility associated with a gold convertibility rate.54
Still, with all these qualifications, the combination of long-run price stability, a commitment to time-consistent fiscal and monetary policies and lower interest payments on foreign debt makes the gold standard sound quite attractive. Indeed, there was a period (in the aftermath of the collapse of the Bretton Woods system) when many American economists – including the young Alan Greenspan – argued seriously for a return to gold for precisely these reasons. This view is by no means defunct. Pointing to the high costs of the enormous number of foreign exchange transactions carried out every year in the absence of fixed rates, the economist Robert Mundell has argued for ‘the introduction of [an] international composite currency’ with, at least to begin with, some kind of ‘confidence-building identification with gold’.55 In May 1999 Greenspan himself reassured the House Banking Committee that gold still represented ‘the ultimate form of payment in the world’; indeed, he had told a Bank of England symposium five years before that ‘the pressure towards … focusing central bank activity to the equivalent of the gold standard [would] become increasingly evident’.56
The reason most economists are sceptical of such arguments is perhaps best summed up by the idea of a policy ‘trilemma’. In essence, the trilemma is that a country can have at most two out of three economic policy objectives: a fixed exchange rate; free capital movements; and an independent monetary policy.57 Members of the gold standard or any successor system generally had the first two, but could not have the third. The possibility therefore existed that the monetary policy required to maintain exchange rate stability in the context of free capital movements might be inappropriate from the viewpoint of the domestic economy. This had been Keynes’s point in the Tract on Monetary Reform:
In truth, the gold standard is already a barbarous relic. All of us, from the Governor of the Bank of England downwards, are now primarily interested in preserving the stability of business, prices, and employment, and are not likely, when the choice is forced on us, deliberately to sacrifice these to outworn dogma, which had its value once, of £317s 10½d per ounce. Advocates of the ancient standard do not observe how remote it now is from the spirit and the requirements of the age.58
For this reason, the price of gold standard membership was not necessarily as low as some calculations might seem to imply. A government that committed itself to a fixed exchange rate might find the domestic costs outweighing the benefits. The credibility of the exchange rate would weaken as the domestic costs rose but, in a vicious circle, attempts to reaffirm credibility by raising interest rates would merely increase the pain for domestic borrowers. Beyond a certain critical point, the speculators would begin to circle overhead and, in the absence of effective external assistance, the central bank’s ability to intervene (to buy the currency as fast as others were dumping it) could quite quickly be exhausted.
Currency crises like these have been fairly regular events because, even in the absence of an international system of fixed rates, many developing countries with external debts denominated in foreign currency will tend to peg their exchange rates rather than risk a market-led depreciation. One way of assessing the costs of fixed exchange rates is to compare financial crises59 before 1914 with financial crises since the collapse of Bretton Woods. There is in fact some evidence that crises in the gold standard era were, if not less severe – the crisis in Argentina in 1890 was worse even than Thailand’s in 1997 – then at least shorter in duration than equivalent crises in the modern world of more or less freely floating rates. This has been cited as evidence in favour of the ‘commitment mechanism’ view of the gold standard: because countries were seen to be committed to convertibility, their breach of the rules in a crisis was seen as only temporary, encouraging investors to return.60 Yet the differences are fairly marginal, and firm conclusions are hazardous given the difficulty of finding data from a strictly comparable sample group of countries.61 Moreover, if there was greater rapidity of adjustment in the pre-1914 period, it may have been a function of quite different factors:62 greater flexibility of prices and wages, greater mobility of labour in the absence of restrictions on cross-border migration, limited political representation of the social groups hardest hit by such crises and perhaps also the widespread adoption in colonial economies of British legal and accounting standards.63 On the other hand, emerging markets before 1914 tended to lack effective lenders of last resort, so that recovery from banking crises ought
to have been slower than in the 1990s.64
The experience of the inter-war period also suggests that fixed exchange rates run the risk of exporting financial crises: the phenomenon known as ‘contagion’. Few historians would now dispute that disastrous errors were made by the American monetary authorities in the 1930s.65 But the sterilization of gold inflows in the 1920s,66 the over-reaction to gold outflows in September 1931 and the failure to continue open market operations in 1932 had catastrophic effects not only on the United States but on all economies with currencies pegged to the dollar. That meant most of the world, since, at the system’s high noon in 1929, no fewer than forty-six nations were on the gold exchange standard.67 It was only when countries abandoned the gold standard in the 1930s that they began to experience recovery.68
By the same token, the system of ‘fixed but adjustable’ exchange rates established at Bretton Woods ultimately subordinated the rest of the world to American monetary policy.69 Although it coincided with a period of rapid economic growth, it is important to remember how unstable the Bretton Woods system was as a system of fixed exchange rates. Almost from the outset, there was doubt about the sustainability of the dollar– sterling rate, in view of British balance of payments problems. After a failed attempt to restore convertibility in 1947, sterling was devalued in September 1949 and again in November 1967. The franc was also devalued three times. Meanwhile, the deutschmark was revalued twice. Until 1959 only the dollar was fully convertible; capital and exchange controls were retained elsewhere because of the shortage of international reserves outside the United States. Thereafter, pressure on the dollar increasingly required international intervention (the ‘gold pool’) and other devices to limit conversion of dollars into gold. From March 1968, when the pool was suspended, the system was in terminal decline, as dollar reserves accumulated in Germany and Japan, both of which were running large current account surpluses with the United States.70 The fundamental reason for the breakdown of the system was the reluctance of the other members – France in particular – to import rising American inflation.71
The Cash Nexus: Money and Politics in Modern History, 1700-2000 Page 39