The Future of the World Monetary System in Historical Perspective
Robert Skidelsky
Social Market Foundation | Tuesday, April 29, 2003

Today most international monetary experts agree that global fixed-exchange rate regimes of the type of the gold standard or Bretton Woods cannot he made to work because of the inevitable politicisation of monetary policy. Any fixed exchange rate system would be subject to a fatal credibility problem. The politicisation of monetary policy, which brought down the gold standard in 1931, was accommodated in the Bretton Woods system by capital controls. But once capital controls broke down, Bretton Woods, too, was doomed. Since 1973 the rule, or rather non-rule, has been generalised floating. This is bound to continue, because of what Obstfeld and Taylor (1998) have called the 'open-economy trilemma'. A country cannot simultaneously maintain fixed exchange rates and an open capital market while pursuing a monetary policy oriented toward domestic goals. Since monetary policy will inevitably remain politicised, and the re-imposition of capital controls is impracticable, the major currencies at least will continue to float.
At the same time it is not doubted that the performance of convertible fixed exchange rate regimes has been superior to that of floating systems, with the Bretton Woods system clocking up the best all-round performance. 'The years 1950 to 1973 were a golden age of unparalleled prosperity' writes Angus Maddison (1995,p.73) Since 1973 OECD economic growth has been half what it was during the ‘golden age’ and the somewhat better performance of the rest is accounted for almost entirely by the East Asian 'miracle'. Moreover, performance has been much more volatile: since the 1970s we have had a 'boom in busts'.(Capdo, 1997, p.80) Flexible rates tend to overshoot wildly and generate equally disruptive movements. The East Asian crisis of the late 1990s cost the region more than $ 100bn. US dollars, 10 per cent of its GDP. Mervyn King, Deputy Governor of the Bank of England recently remarked: 'the sharp reversals of capital flows to emerging markets have caused crises with a frequency and on a scale that threaten support for an open, market-led economy'. (King, 1999) Although the academic case for floating has been severely dented by post-Bretton Woods experience, attempts to limit exchange-rate fluctuations since the 1970s have broken down in face of massive capital flows. Mainstream analysis is understandably sceptical about the future prospects of any future attempts. Mainstream proposals for the reform of the existing financial system concentrate on reforming capital markets and banking systems, not currency systems.
Tonight I want to confront head-on the standard view, represented notably by Barry Eichengreen (1994), that the conditions which made fixed exchange rate systems work in the past have permanently disappeared. I want to argue the exact reverse. that they are reappearing. Of course, no monetary regime is credible for ever, and any regime may be shocked into collapse. However, in the long run we are all dead; and any system which offers a hope of working better than our current non-system is worth some effort to secure. I will end with a proposal. An important subsidiary aim of this lecture is to bring together two literatures which have become separated: the analysis of financial crises and the discussion of currency reform. My contention is that the missing link in recent theories of financial crises is the nature of the exchange rate regime itself.
Before taking up the main thread, let me try to shed some historical light on my theme. The two modern examples of 'working' fixed exchange-rate systems were the classic international gold standard which lasted roughly from 1880 to 1931 and the convertible phase of the Bretton Woods system which lasted from 1959 to 1971. What made them work? And why did they break down? In a way the 'classical' gold standard is more interesting, because it worked without capital controls.
It's now generally accepted that it didn't work in the way economists once believed. Specifically, there was no automatic adjustment mechanism. Adjustment could be blocked at both ends: by the creditor refusing to inflate and by the debtor refusing to deflate. Nevertheless, it worked all the same. Why? Eichengreen's brief answer is because the commitment to currency convertibility was strong, the system was not endangered by domestic political pressures, and because central bank cooperation supported currencies in crises. Credibility, though, was less at the peripheries of the system, particularly Latin America. (Eichengreen, 1992, ch.2; 1994,pp.42-6)
Let me say a word about all four points identified by Eichengreen. Undoubtedly commitment to convertibility was crucially important. To he on the gold standard was the mark of a first-class country; to be off it, or to have to suspend convertibility too often, was to be a 'banana republic'. The gold standard was, in Schumpeter's phrase, 'the ideal to strive for and pray for, in season and out of season'.(Schumpeter, 1954, p. 405) Countries joined the gold standard to signal their creditworthiness; this was the key to borrowing cheaply, and for longer periods. A crucial motive for the globalisation of the gold standard in the late 19th century was the emergence of a world capital (as well as trading) market. The gold standard was a particularly simple and efficient mechanism for providing information about the creditworthiness of sovereign borrowers. I doubt whether we have improved on it, despite the proliferation of data and credit-rating agencies. To stay on gold, a country had to practice monetary discipline, have a balanced budget and be free from the threat of arbitrary regime change.
This is precisely the sort of information demanded by investors now. Whereas today, investors must rely on ratings agencies and country analysts to provide them with these indicators, a country’s gold position made them readily available and at low cost to lenders.
Secondly, Eichengreen talks about the insulation of monetary policy from political pressure. He attributes this to imperfect understanding of the connections between monetary policy and employment, as well as limited suffrage and weak trade unions. This is partly misleading. The gold standard actually spread in parallel with the extension of the male suffrage to the working class. In Britain, the Labour Party was its most loyal defender, up to the very moment of its final collapse 1931. More important, I would judge, in keeping monetary policy non-political was the relative lack of strain to which the system was subject. This was partly due to the density of international price linkages. The high ratio of traded to non-traded goods, the commodity structure of international trade, the low share of non-tradeable services in domestic GDPs all tended to promote the 'law of a single price'. The price inertia so characteristic of later economies was much less pronounced. As a result the price levels of trading countries were much less prone to get out of step ex ante and require adjustment post-hoc. (Krugman 1989, p.96; Triffin, 1968, p.14). This is another way of saying that countries on the gold standard were less likely to experience asymmetric shocks. Eichengreen is right to point out that wage flexibility was not remarkable under the classical gold standard. But this was offset by a huge safety valve in the form of trans-continental migration. Between 1881 and 1915, 32 million, or about 15 per cent of its population, emigrated from Europe, 60 per cent to the United States. Population movements were accompanied by investment flows to develop new lands: the investor and trader followed in the footsteps of the colonist. In all these ways the gold standard system was more like an optimal currency area than a collection of national economies. Keynes's quip 'The world as a whole has no 'foreign trade’"was never truer than under the gold standard.
Thirdly, central bank cooperation illuminates the political economy context of the gold standard. The main debate has been between those like Eichengreen who argue that it was a cooperatively managed system, and those like Kindiberger who claim it was a hegemonic system, with Britain as the hegemon. (Kindleberger, 1973, 1981) Eichengreen is undoubtedly right to say that as far as Europe was concerned, the system rested on central bank cooperation, made possible by the relative absence of political conflict.(Despite all the tensions, there was no major war in Europe in the forty-three years between 1871 and 1914.) However, Britain was undoubtedly the hegemonic power for most of the rest of the world, not just because of the wide extent of the British empire, but because of the dominant position of Britain in world imports and the City of London in world finance.
To give just one example, Keynes wrote in 1925: 'To lend vast sums abroad for long periods without any possibility of legal redress when things go wrong is a crazy construction; especially in return for a trifling extra interest'. (Keynes, 1981, p.278) Keynes had Latin America mainly in mind, but for much of the rest of the world it was far from obvious that it was crazy. In the United States, one of the largest capital importers, the rule of law applied. Elsewhere, imperial organisation safeguarded the investor from exchange-rate risk and guaranteed prompt legal redress. By 1870, 70 per cent of British loans went to the British Empire, whose dependent territories were, in effect, on aq sterling standard. Even in Latin America, creditors could enforce their will. Something like the international bankruptcy code now called for by Jeffrey Sachs and others existed in the 19th century precisely because so many foreign borrowers were either not sovereigns or weak sovereigns, susceptible to 'gunboat diplomacy'. It was thus a combination of economics and political domination that sustained the continuous outflow of foreign capital from the surplus areas of Western Europe to the deficit areas beyond; a system which, as Keynes wrote in 1941, 'transferred the onus of adjustment from the debtor to the creditor...' and 'served at the same time to keep the balance of international payments in equilibrium and to develop resources in undeveloped lands'. (JMK, 1980,pp. 21,30)
The limits of informal imperialism were encountered in Latin America. Political turbulence, populist fiscal policies (unbacked, it should be noticed, by universal suffrage), and volatile commodity and food prices made it the most unstable sector of the gold standard world, with frequent defaults, moratoria, capital flights and suspensions of gold convertibility. When the world depression hit the 'new' gold standard of the 1920s, it cracked first in Latin America. It was partly the knock-on effects of the Latin American defaults on Britain which brought the gold standard system crashing down in 1931.
In designing the Bretton Woods system during the Second World War, policy makers were impressed by the virtual disappearance of most of the conditions which had made the gold standard possible before the First World War. With the growth in product and labour monopolies, wages and prices had become 'rigid', with deflationary shocks producing heavy and persistent unemployment; democracy had increased the demand for social protection; immigration laws stopped international labour mobility; the ratio of manufactured to commodity trade had gone up and the ratio of traded to non-traded goods had gone down; private property was under challenge at home, and imperialism was disintegrating abroad; and the new fiscal and monetary theories associated with Keynes gave the maintenance of domestic full employment priority over the maintenance of fixed exchange rates.
The main purpose of the Bretton Woods Agreement of 1944 was to re-establish a global fixed exchange rate regime as a precondition for freeing up trade, but one which fitted the new world of politicised money and sluggish adjustments to shocks. The object of an improved system must be to insulate domestic monetary policy not from politics, but from international finance. Hence the three Bretton Woods innovations: adjustable rates, capital controls, and the IMF- all designed to protect national economic autonomy.
Four main sources of exchange-rate difficulties were identified at the time.
1. Domestic shocks. The British, with their powerful trade unions much in mind, emphasised wage inflation. The Americans, with Latin American experience to the fore, emphasised budget deficits.
2. Terms of trade shocks, particularly important for primary commodity exporting countries.
3. Political shocks - revolutions, socialist 'experiments', populist dictatorships.
4. Capital flight. This was regarded as a consequence of such shocks, though aggravating them. Keynes, for example, thought that 'the position of the wealth-owning classes' was likely to be so threatened after the war, that capital flight would be endemic unless steps were taken to prevent it. (Keynes, 1980; p.31)
The remedies provided by the Bretton Wood Agreement were agreed exchange rate adjustments, control of capital movements, and pooled reserves, available through the IMF on certain conditions, to contain short-run balance of payments problems. The architects of Bretton Woods opted for fixed exchange rates, because they wanted to avoid the currency wars of the 1930s, and because most economists at the time doubted that floating would guarantee equilibrium. There was much learned discussion about whether the Marshall-Lerner condition would be satisfied.
Keynes succeeded in inserting into the Bretton Woods agreement a clause stating that a country's domestic policies were not a ground for refusing a request to devalue. This was designed to secure sufficient freedom for domestic monetary policy. One consequence was to undermine credibility of commitment to the fixed rate. This turned out for a long time not to matter. Contrary to Keynes's wishes, the Bretton Woods Agreement imposed only minimal conditions on creditors. All that survived of his original plan was the 'scarce currency clause' allowing countries to impose import controls against a country whose currency was declared to be 'scarce', that is, in excessive demand. This clause never became operative. The reason was that the deflationary 'dollar gap' was eliminated by the US commitment to keep Western Europe and Japan free of Communism. This led America to acquiesce in the large sterling, franc, and deutschmark devaluations against the dollar in 1949; it led to the huge outflow of American dollars on government account, later supplemented by large private ouflows. These events were reflected in the run-down of US and the build-up of European (and Japanese) reserves. The trend in the balance of payments in turn enabled exchange rates to be gradually stabilised and currency convertibility re-established. This promoted trade liberalisation which in turn fuelled economic growth.
In 1960 J.P.Morgan partner Russell Leffingwell wrote to Waiter Lippmann: "Wisely we undertook to set the world to rights. We gave money and know-how to our foreign friends, we made fixed foreign investments, and we policed the world against the Russians and Communist Chinese with foreign bases and foreign based troops and ships and planes. All this involved spending immense sums of dollars abroad. We and our friends abroad had been so obsessed by the thought of the...dollar gap... that until recently few noticed that the dollar shortage had disappeared and a dollar glut had taken its place. our foreign aid has been successful beyond our dreams. Western Europe and Japan had full recovered and were in hot competition with us here and abroad. So our favorable trade balance had dwindled to little or nothing. We are still spending abroad billions more than our income from abroad, and the resulting deficit is reflected in our loss of gold and increased short teen debt to abroad ... For the first time in more than a quarter of a century we are being subjected, willy-nilly, to the discipline of the gold standard’. (Lippmann Papers)
This development was completely unanticipated. Harry Dexter White, the main architect of the Bretton Woods plan, assumed that the main countries would gradually return to a modified gold standard. Gennany, Japan, and France could certainly have done so in the 1960s. De Gaulle and his adviser Jacques Rueff wanted to. But memories of the old gold standard were too vivid. The fact that the main gold producers were South Africa and the Soviet Union was an insuperable political problem.
The result was that the US dollar alone remained convertible into gold; every other currency pegged to the dollar. With sterling rapidly fading out as a reserve and transactions currency, the Bretton Woods world, as Professor Mundell noted in his introduction to this conference, became a 'vast dollar area'. The one thing which America did not, or perhaps could not, do was to stick to the disciplines of the gold standard. Not only was the dollar the world's chief reserve asset, but the existing arrangement, whereby countries accumulated dollar liabilities, suited both the United States and its Cold war partners. The denouement of this 'Triffin Paradox' might have been postponed had the United States pursued more conservative fiscal and monetary policies in the 1960s. Instead it ignored the 35 dollar gold price constraint and inflated to fund the Vietnam War at the same time as its 'Great Society' programme, using the dollar's seigniorage to export inflation. Once its short-debt came to exceed its gold reserves, it was forced to suspend dollar convertibility into gold and float its currency. This in turn led to generalised floating in the 1970s, as there was no suitable alternative currency on which to peg.
Perhaps the most important thing to remember about Bretton Woods was that it was a deal between the United States and Britain. The problems it set out to solve were their problems. The features of the system that most strongly differentiated it from the old gold standard -adjustable exchange-rates, the loan facility, and control on capital movements -were designed to protect a weakened Britain, which felt vulnerable to domestic and external shocks. Today it is not the 'core' countries of the system which are prone to banking and currency crises but the 'emerging market' economics. What globalisation has done is to expose the financial inequality of nations. These are no longer hidden from the scrutiny of the vulgar by quasi-autarkic policies. In thinking about a new global financial architecture, therefore, it is essential to keep in mind the distinction between how the 'classical' gold standard worked at its core and how it worked at its peripheries. I will take up this point in the next part of my lecture.
Today there is no global fixed exchange-rate system. However, if we took hard enough, we can see the shadowy outlines of a 'system' emerging from the 'non-system'. Regionalisation, furthest advanced in Europe, has whittled down the ‘core’ currencies to three: the dollar, the euro, and the yen. Peripheral currencies increasingly tend to want to fix to the core currencies, both to import credibility and to safeguard trading connections. The monetary conditions for most of the global economy are close to being set by three central banks. With globalisation, some of the structural features of the pre- 1914 world are reasserting themselves. I would argue that the consensus of economists has been slow to adjust to these developments. The major agreed intellectual axiom remains that in the presence of capital mobility, only permanently fixed rates (as in a monetary union) or freely floating rates are sustainable in the long run. Fixed but adjustable rates -the Bretton
Woods formula -are not credible, for reasons which Milton Friedman explained as long ago as 1953. The advantages of free capital flows are also stressed. In addition to the usual efficiency arguments, punishment by bond markets is regarded as a much more credible sanction against inflation-prone governments than commitment to an exchange rate target. The great reduction in world inflation rates since the 1980s is thought to confirm this, though there may be other explanations.
The post-Bretton Woods currency crises have given rise to a rich interpretative literature. I briefly consider three approaches. It should be noticed that all of them assume fixed-exchange rates, the implication being that, under floating, they would not have happened.
The first, and most traditional, starts with a structural budget deficit which is monetised or leads to the build up of unsustainable debt. Ether can produce a loss of reserves and capital flight out of domestic into foreign currencies, leading to a collapse of the peg. This was the 'canonical' explanation of the Latin American crises of the 1980s and
The theory of self-fulfilling speculative attacks was developed to explain the ERM crisis of 1992. This postulates multiple equilibria, a benign one in which agents have confidence in a country's commitment to the peg, and no speculative attacks take place, and a malign one in which they have no confidence. The second triggers off a speculative attack, which validates itself by bringing the peg down. The basic idea is that attacks of the multiple equilibria model type can occur when traders know that the fundamentals are sound but the soundness of fundamentals is not common knowledge. Adding transparent information about market fundamentals guarantees a unique equilibrium. (See S Morris and H.S. Shin, 1998)
The East Asian troubles of the 1990s produced a new crop of models, in which moral hazard featured prominently. These shifted the analytical spotlight onto the private sector, particularly capital flight from the banking sector. Developing countries with high growth rates and sound budgets attract investors like bees. Banks all around the world were falling over themselves to lend dollars to the East Asian tigers. Local banks used short-term dollar denominated debt to make long term investments. When doubts about the soundness of these investments arose, many banks found themselves insolvent. Here it was a maturity rather than a currency mismatch which triggered off the crisis. Why were the NICs only able to get overnight or other very short maturity loans? Probably because so few of their banks had a C+ rating or higher.
But th is not the whole story. Government guarantees of the banking system had created a stock of contingent implicit government liabilities -expenditure commitments not explicitly laid out by law which the state would have to honour if the banks failed. Contingent budget deficits may he more alarming than actual or expected ones. Investors may suspect that governments will be tempted to monetise the debt. Secondly, it was not obvious that fiscal positions were sound given the rapid build up of private sector debt. It may be that governments should have run. even larger budget surpluses. Finally, the balance sheet mismatch was exacerbated by 'crony capitalism' -the allocation of capital to political projects which could not expect to cover interest and debt repayment charges.
The main conclusion drawn from these episodes is not that capital flows should be curtailed, but that currencies should be allowed to float and that banks needed to be more tightly regulated, or even foreign-owned. Indeed, the attempt to maintain nominal exchange rates in face of capital flight is widely regarded as an important contributory cause of the East Asian crisis.
There is no shortage of ideas for the prevention and resolution/containment of crises. In
his lecture, 'Reforming the International Financial System: The Middle Way', Mervyn King, Deputy Governor of the Bank of F-ngland, lists the main ones. On the side of prevention/containment come:
• 'Do it Yourself-Lender of Last Resort' aimed at providing self-insurance against liquidity crises, either by building up large foreign currency reserves on a national or regional basis, or creating contingent credit or collaterised loan facilities;
• Better management of national balance sheets;
• Encouragement of equity rather than debt finance by creating credible legal and institutional infrastructures;
• Better design of debt contracts to provide a framework for negotiation between creditors and debtors when financing difficulties arise;
• Better banking supervision in emerging markets;
• Reduction of moral hazard in lending countries which encourages short-term debt flows.
Ideas for crisis resolution include:
• An international bankruptcy code;
• Temporary capital controls.
• Greater transparency in public and banking sector accounts, including information
about the degree of transparency, can help both reduce the frequency of crises and
their severity.
Dr. King rejects two 'purist' solutions - an international tender of last resort and permanent capital controls. The first would create unacceptable problems of moral hazard, as well as being financially and politically infeasible; the second goes against the grain of promoting market reforms and good governance. (King, 1999)
Two comments may be made about this class of remedies. First, they ignore the possible contribution of the exchange-rate regime itself to the financial problems of the emerging markets. As Fred Bergsten, Paut Krugman and Marcus Miller have written 'it is a debate with a hole in its heart'. (Bergsten, Krugman, Miller, 1999) The dollar rose 80 per cent against the yen from early 1995 to mid-] 998. As a result the Asia dollar pegs led to substantial overvaluations and large trade deficits. This became inconsistent with the scale of borrowing and would have led to investor doubts, irrespective of the suitability of the loans being made by the borrowing banks. Most analysts have regarded this as an irrefutable argument for floating. But it might just as well be stated that had the real dollar exchange not gyrated so heavily the excesses of the boom and bust would have both been avoided. (As does Hamada, 1998)
My second comment is that all current reform suggestions are rightly targeted on the problems of emerging markets. With the exception of the ERM crisis of 1992 - surely a special case, connected with the birthpangs of the Euro- all the post-Bretton Woods financial crises have been Latin American or East Asian. The remedies proposed in current thinking have -except for those pertaining to lending bank behaviour - little relevance to the 'core' countries of the world economy, because these are no longer 'emerging markets'. They already apply the domestic rules which the international institutions recommend for the rest.
In conclusion, I want to return to my introductory theme. The current view is that freedom of capital movements goes with floating and that control of capital movements goes with fixing. Yet both efficiency and disciplinary arguments for combining fixing with free capital movements are strong, provided the combination can be made to work.
The general case for fixed exchange rates is that they reduce the risks and uncertainties of international transactions and therefore encourage trade expansion and foreign investment. But they have also functioned as a disciplinary device -as a constraint on 'wicked' Finance Ministers. This was a major rationale for the old gold standard.
The general case for free capital flows is that they increase the supply of investment funds, and make their allocation more efficient. But bond markets also punish 'wicked' Finance Ministers and reward good ones, thus creating a pay-off for good governance which might not exist otherwise. What I want to argue is that the core countries of the global economy, by which I mean the United States, the European Union and more doubtfully Japan, are reaching a point when they can contemplate stable exchange rates and free capital flows between themselves with equanimity for efficiency reasons alone, because their fundamentals are sound, and they have well-established habits of cooperation. Their pegs will be credible and they will be able to take full advantage of the international division of labour which stable exchange rates and free capital flows bring.
On the other hand, many other countries will find it suits them to attach themselves to the main monetary centres to get the same benefits. History as well as contemporary practice reveals many kinds of subordinate monetary systems designed to enhance credibility - such as currency boards and 'dollarisation'. In short, I reverse the ruling paradigm: fixed exchange rates and free capital mobility should coexist. Can they be made to?
Let me recall the conditions which underpinned the classical gold standard and compare them with what we have today. They are certainly not what they were before the First World War. But neither are they what they were in the fifty years or so up to the 1970s. We no longer live in the kind of world envisaged by the fathers of the Bretton Woods system, a world whose susconscious influence explains much of the pessimism which still attaches to projects to 're-fix' the exchange rates.
Here is a quick check-list.
(1) In the gold standard era, labour was domestically immobile, but internationally highly mobile.
(2) In the gold standard world, trade unions were weak.
(3) The structure of the pre-1914 economy made for direct international price linkages, tending to promote the 'law of a single price'. As a result the price levels of trading countries were less prone to get out of step ex ante and require correction ex post.
Today we are returning to this kind of world. Mass transcontinental labour migration has disappeared, but within large areas, such as North America, and Europe, cross-border migration is growing and will probably continue to grow, not least in response to the ageing of resident populations. Today trade unions are much weaker. Finally, the ‘new economy’ created by the Internet is a ‘single price’ economy.
Consider three factors which surely dominated from the 1920s to the 1970s in most developed countries: (a) wages were union-determined, (b) labour markets were heavily regulated, and (c) industry was increasingly concentrated. Galbraith drew from all this the conclusion that prices were set by monopolistic bargaining, not by competitive markets. This situation, which surely captured a moment in history and a state of technology, has started to unravel. Changes in the structure of the economy have brought about much more flexible product and labour markets. The most important influences are the huge reductions in communications' and transactions' costs caused by microelectronics, the rise in self-employment, the growing openness of economies, the rising ratio of traded to non-traded goods and the growing share of services and consumer goods in the trade basket. These developments, which have led to the decline in classic trade unionism, both cut away at the roots of domestic rigidities and raise the benefits of exchange rate stability.
In short, the supply side characteristics of developed economies are edging back to the greater flexibility of earlier conditions. This puts less pressure on macro-policy. Some consequences:
(4) Governments, or more accurately electorates, are more tolerant of unemployment than they were, partly because the long experiment with (hubristic) Keynesian macroeconomic policy ran into a dead end and partly because it is no longer associated with the scale and duration of economic downturns like the Great Depression: we are back, that is, in a business cycle world rather than one of permanent demand deficiency.
(5) Under the gold standard a limited, or more accurately, unmobilised electorate, allowed budgets to be balanced at a low level of public spending and taxation. Today we are much more democratised than then. But the demand for social protection has waned, partly because private assets are much more widely distributed. The majority of 'have nots' have shrunk to a minority of 'excluded'. There has been a revival of belief in limited government. Neo-Victorian fiscal rules have been reinstated. The era of rising public spending and large budget deficits seems to have come to an end.
(6) Under the gold standard, monetary policy, insofar as it existed, was discretionary but non-political, i.e., made by independent central banks. From the 1930s to the 1970s, monetary policy was under the direct control of governments. The result was an almost continuous rise in inflation. Few monetary authorities now claim the right to 'choose' their own inflation rates. They are all committed to low inflation targets. Belief in low inflation is institutionally reinforced by the return to the apolitical tradition of central banking. Since the mid-1980s, the inflation rate has been falling and is now not much greater than it was under the gold standard. As I have said, this is as much due to changes in the structure of the economy as to changes in policy. However, one should not be blind to the potential conflict between the growing demand for' democratic' accountability and the renaissance of the international financial market. For much of the twentieth century 'international finance' was kept under strict national control. Now it has been liberated again. Will there be a similar reaction back to economic nationalism? The answer surely depends on whether financial markets are benign or malign. In the 1920s they were associated with unacceptable levels of economic volatility. I have suggested that this may he ceasing to be a problem in developed countries. But it remains possible, and likely, in the emerging market sector of the world economy.
(7) The political economy of the gold standard world supported the maintenance of robust monetary rules. I singled out the dominance of private property, the long period of peace and imperialism. In 1941, Keynes gave as a prime reason for capital flight uncertainty about the position of the wealth-owning classes in the social system of the future. Today with the collapse of communism, and the end of old-fashioned socialism, the position of wealth-holders in the developed world has never been so secure. Major inter-state wars are also much less likely than they were at any time since 1914. In short, the developed world is less prone to major political shocks than at any time this century. Despite the end of formal imperialism, a new Western hegemony or informal imperialism has emerged (the 'Washington Consensus') exercised through fig-leaf international institutions like the World Trade Organisation and the IMF, and subordinate monetary systems typical of the gold standard era. This is a source of is conflict as well as coordination. How countries like China and India will fall in is yet unclear.
What am suggesting is that we are returning to the conditions when it will be possible to stabilise the exchange-rates of the three main currencies, the dollar, the euro, and the yen. However, there is a powerful counter-argument, which may be summed up in the quip: ‘Who needs gold if we have Greenspan? More generally, if de-politicised monetary policy has become possible, for the first time since 1914, why do we need exchange-rate targets? Why not just have money or inflation targets, letting the exchange rate take any residual strain?
However, there are a number of large chinks in this argument.
First, there exists no general agreement on the transmission mechanism between gold and prices. Central banks may congratulate themselves on having provided sound money in the last ten years or so, but is the money supply exogenous or endogenous? This academic debate which flourished in the 1970s and 1980s is not resolved.
Secondly,present anti-inflationary policy rests partly on memories of the 1970s. Memories fade, and central bank determination to keep money sound may weaken in face of modest shocks, without the discipline of fixed-exchange rates.
Thirdly, no central bank is truly independent; there is always a political override. Perhaps the European Central Bank is an exception. In this case, there may well be a trade-off between independence and political legitimacy.
Given the other advantages of fixing described above, it is not absurd to predict that, after a few more financial shocks, the main currencies will be fixed.
A historical reminiscence of how this might come about is provided by the Tripartite Monetary Agreement of 1936. This was triggered off by France's decision to devalue the franc. An Agreement was signed, on 25 September 1936, by which the United States and Britain accepted a30 per cent franc devaluation, and the three countries agreed to use their gold reserves to support each others' currencies in the foreign exchange markets within agreed bands. These monetary arrangements soon embraced the core Western democracies. In July 1937, the United States negotiated a series of currency stabilization agreements with Latin American countries.
Any such agreement today would have profound implications for the emerging and transitional economies. The more stable the central rates are, the easier it will he for peripheral countries to avoid the mistake of pegging to the 'wrong currency' as a number of East Asian countries did in the 1990s.
However, for such countries there is still likely to be a conflict between fixing and free capital movements. The best way to make the 'fix' credible is through a currency board system, or by adopting the dollar or euro as domestic currencies. This too echoes pre-1914 arrangements.
We then need to answer a final question. Can a fixed-exchange rate system coexist with capital mobility without some link to gold? It would be the first time in human history that such an experiment is tried. It should be noticed here that fixing exchange- rates between three ‘core’ currencies does not avoid the old choice between sound money and politicians. It just reduces the number of politicians involved. The arguments for re-establishing a gold, or more generally, a commodity basis for money is far from negligible.
First, it may be rational for governments to buy votes by printing money. Secondly, a link to gold can be a commitment device which avoids the time-inconsistency problem. Thirdly, there may be welfare gains in moving from a policy of targetting low inflation to a gold-based standard which is equivalent to zero inflation.
I doubt if any of these considerations will suffice on their own to restore a monetary use for gold. Events, rather than arguments, will decide whether the monetary systems of the future will need this extra prop. Professor Mundell’s prediction that ‘gold will be part of the monetary system of the 21st century’ rests, essentially, on a conservative, even Christian, view of humankind, a scepticism about human reason, a belief in the fallibility of all rational designs. Yet this might be the ultimate reaction to the hubris of paper money. Even Keynes, who is credited by Roy Harrod with having destroyed the gold standard ‘almost single-handedly’, wanted to retain a ‘constitutional’ role for gold. Stranger things have come to pass.
Bergsten, C.Fred, Paul Krugman and Marcus Miller (1999) 'Global Financial Stability: The Missing Link?', Mimeo, 112199
Caprio, G. (1997) 'Safe and Sound Banking in Developing Countries: We're not in Kansas Anymore', Research in Financial Services.. Private and Public Policy
Eichengreen, B. (1992) Golden Fetters: The Gold Standard aiid the Great Depression 1919-1939, OUP
Eichengreen, B. (1994) Intemational Monetary Arrangements for the 2 Ist Century, Broookings Institution
Hamada, K. 'Economic Consequences of Pegging to the Dollar in a Multicurrency World, 20 November 1998, Working Paper
Keynes, J.M.(1980),'Post-War Currency Policy, 8 September 1941, reprinted in The Collected Writings of John Maynard Keynes, vol.XXV, Macmillan, CUP for the Royal Economic Society
Keynes, J.M. (1981)'Foreign Investment and National Advantage', 9 August 1924, reprinted in ibid, vol. XIX
Kindleberger, C.P. (1973) The World in Depression 1929-1939, London
Kindleberger, C.P. (1981) 'Dominance and Leadership in the International Economy', International Studies Quarterly, vol.25, no.2, June 1981, p.242-254
King, Mervyn (1999)'Reforming the International Financial System: The Middle Way', Bank of England, Speech at the Federal Reserve Bank of New York on 9 September 1999.
Krugman, P. (1989) Adjustment in the world economy
Lippmann Papers, Yale University, File 1316, R. Leffingwell to W.Lippmann, 2 March 1960
Maddison, Angus (1995) Monitoring the World Economy OECD.
Morris, S. and H.S. Shin (1998) 'Unique Equilibrium in a Model of Self-Fulfilling Currency Attacks', American Economic Review, 88, pp.587-599
Obstfeld, M. and Alan M. Taylor (1998) 'The Great Depression as a Watershed: International Capital Mobility over the Long-Run' in Bordo, M.D., Claudia D. Goldin, and Eugene N. White eds. The Defining Moment: The Great Depression and the American Economy in the Twentieth Century, 1998.
See also: Obstfeld, M. The Global Capital Market: Benefactor or Menace?', mimeo, 27 April 1998.
Schumpeter, J.A.(19-54) History of Economic Analysis London
Triffin, R.(1968) Our International Monetary System: Yesterday, Today and Tomorrow, New York