Robert Skidelsky
Join our Mailing List
to be notified of any updates

Delivered by FeedBurner

Follow me on Twitter
Bookmark and Share

Exchange Rate System for ‘Mature’ Economies in a Global Economy
Robert Skidelsky
Madrid Convention | Wednesday, May 14, 2003

The main purpose of this paper is to take issue with Bordo's claim that targeting inflation is the efficient 'modern' way to secure stable exchange rates. It argues that the Bretton Woods system of fixed, but adjustable exchange rates, gave the best all-round performance of all the exchange-rate regimes we have had, and that this constitutes a prima facie case for trying to recreate something like it. It further argues that the causes of the collapse of the fixed-exchange rate systems of the last century have been misinterpreted to support the current near consensus in favour of floating. It argues that the advocates of 'autonomous' monetary policy make exaggerated claims on its behalf. The thrust of the paper is that a fixed-exchange rate regime is more in tune with global economic integration than is a system of floating exchange-rates, which comes out of the stable of nationalist economics and carries the seeds of currency wars and protectionism.
i. Bordo and the Trilemma
Most contemporary writers on exchange rates subscribe to 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. Specifically, fixed-exchange rates combined with freedom of capital movement leave inflation and employment goals at the mercy of international finance. Since choices concerning these objects should be made by a country’s own authorities, the choice of exchange rate regime resolves itself into floating with free capital movements or fixing with capital controls. Since control of capital movements is also regarded as unfeasible or undesirable, floating emerges as the only choice for the mature economies (pending the utopian arrival of a single global currency backed by a world government.) Subscribers to the trilemma doctrine tend to be highly sceptical about the currency union established in the EU, mainly on the ground that it abolishes the monetary freedom of the individual country-members, without providing any effective substitutes (eg fiscal transfers from the centre to deal with asymmetric shocks.) Some opponents of the Single Currency are against such transfers in any case, since they presuppose a European government which they oppose.
It is widely agreed that economies which need to import foreign capital to develop may have neither the capacity nor ability to have independent monetary policies, with either fixed or flexible exchange rates. But I leave the question of exchange-rate choice for developing countries to my colleague, Vijay Joshi.
Bordo, as others have before him, tries to overcome the force of the ‘trilemma’ by arguing that floating with inflation-targeting leads to de facto fixing. (Bordo,2003 ) . Common inflation-targeting will achieve exactly what the Bretton Woods system tried to achieve - fixed, but adjustable exchange rates - but without the need for capital controls. (Bordo, 17) His argument is that if the major countries converge on the same rate of inflation, then the nominal exchange-rates of their currencies would exhibit long-run stability, moving only in response to shocks requiring changes in real exchange rates. This was precisely the aim of the Bretton Woods system.
Bordo’s paper argues that the choice of exchange-rate regime should be governed by degree of financial maturity. Mature countries should float; immature ones should give up (or limit) their quest for monetary independence. But as they develop maturity they too can look forward to floating.
His paper reverses the view taken in the past that fixed rates were a sign of financial maturity, and floating rates the sign of financial immaturity. In the late 19th century joining the gold standard was seen as the capping stone of financial reforms which included establishing an independent central bank (independent from financing budget deficits) and a fiscal constitution based on annually balanced budgets. In modern terms, it signalled the achievement of a sound money regime and a commitment to maintain it. ‘Adhering to gold convertibility can be viewed as a commitment mechanism to the pursuit of sound monetary and fiscal policies’. (Bordo,16) Today’s substitute -‘a domestic fiat nominal anchor’ -represents a ‘major technical improvement’, because it achieves credibility of commitment without the high resource costs of the classical gold standard. (Ibid.,17, 24-5). ‘A consequence of this analysis’, Bordo, writes ‘is that logically, the pre-1914 core countries that had developed strong money and financial markets and institutions before World War 1 ought to have floated -something which they did not’ (Ibid.); and so should the core countries today. (Bordo, 24)
If monetary independence turns out to be exaggerated, the force of the trilemma is weakened. There might then be no reason why a system of fixed but adjustable rates should not be adequately credible even with free capital movements, provided a system of mutual support and escape clauses could be agreed between the three main currency blocs.
One point is worth remembering in what follows. Any monetary regime which is based on fiat money requires a nominal anchor, be it an inflation target or an exchange-rate target. Both equally require money and politicians to be exogenous. But neither money nor politicians are wholly exogenous. Money is pushed and pulled by the structure of the economy; politicians by the problem of maintaining power and consent. So any monetary regime is doomed to imperfection and ,probably, to eventual collapse. The question is which system promises better results. Two arguments may be adduced in favour of fixed-exchange rate over inflation targeting. First,while the maintenanceof both commitments rests with the politicians, the first is a multilateral system, whereas the second is unilateral. That is to say, it engages the forces of international cooperation on its behalf in a way that the system of inflation targetting, being wholly national,does not. As such it can be regarded as an institutional building block of a globalizing economy. The second advantage flows directly from the first. Because there is an agreed contractual element, implicit or explicit, to any fixed exchange rate system ('rules of the game'), the commitment to maintain the system as a whole is likely to be stronger than that for an individual country to maintain any particular inflation rate; and therefore it may be more credible under pressure. This tallies with our experience of these systems in the past.
Two sections follow.The first tries to shed some light on the reasons for the collapse of the two major fixed exchange rate systems of modern times: the gold standard and the Bretton Woods gold-exchange (or dollar) standard which succeeded it. The second raises questions about the extent of the benefits in terms of domestic policy independence which floating is supposed to bring.
ii. Exchange Rate Systems and the Trilemma
The theory of the trilemma may be seen as an attempt to summarize our historical understanding of why it proved so difficult to maintain a fixed exchange rate system in the 20th century. Two main conclusions have emerged: the gold standard was wrecked by the rise of democracy and the demand for political control over the business cycle; the Bretton Woods system was destroyed by the breakdown of barriers to the free movement of capital. Both can be questioned. But first a brief history.
The 'classical' gold standard finally broke down in 1931. Interwar experience of floating currencies created a consensus for ‘fixed but adjustable’ rates after the second world war, Nurkse arguing in an influential paper (1944) that floating rates added major sources of uncertainty to trade and capital formation. The ‘fear of floating’ gave rise to the Bretton Woods system of fixed but adjustable rates, supported by the International Monetary Fund. After the collapse of the Bretton Woods system in the early 1970s ‘fear of floating’ gave way to ‘fear of fixing’ in the Anglo-American world (and among Anglo-American economists and policy-makers) but not in the European Union, where a ‘mini-Bretton Woods system’ was created as a step towards a single European currency.
Historically, periods when exchange-rates were fixed are associated with better macro-performance than when floating is dominant. The best perfomance in overall results -growth, employment, prices, stability -was achieved under the Bretton Woods System of fixed but adjustable exchange rates (c.1950 -1971/3) Quote Davidson,if appropriate. There are sound logical reasons for this. Provided that certain preconditions are met -that the system is credibly supported, and that not too much is expected of its adjustment mechanisms -then it will, on balance reduce uncertainty and the waste associated with uncertainty. This provides the best background for the growth of trade and investment.
According to currently received wisdom, the credibility of the gold standard rested on the ‘incompletely understood’ connection between monetary policy and the domestic economy, and the restricted nature of the franchise; it broke down in face of better economic understanding and the political demands for full employment from newly empowered voters. (Eichengreen, 1994, pp.43, 47-49).
What wrecked Bretton Woods was the emergence of capital mobility. (Obstfeld and Taylor, 1998, cited Bordo, 2003, p.5) There is much dubious history in all this, which leads to wrong conclusions for policy.
In fact, the gold standard was doomed by the first world war, not by a full employment commitment which only came after the second world war; in Britain and Germany working-class parties were the most robust defenders of the gold standard. What the first world war did was to destroy the system of international politics and finance on which the adjustment mechanisms of the pre-war gold standard had depended (including migration and the pivotal role of the City of London).The partly-restored gold standard of the 1920s not only started off with severely misaligned currencies, but was subject to shocks which no system of fixed exchange rates could have survived. (See Skidelsky,1999)
The collapse of the Bretton Woods system has been attributed to growing mobility of capital. Hence the birth of the dogma that a fixed exchange rate system can only be maintained with capital controls. In fact, growth of capital mobility had little to do with the breakdown of the Bretton Woods (gold-exchange) system. The collapse of the central relationship between the major currencies - the dollar, franc, DM, sterling - was brought about by the inflationary financing of the Vietnam war by the United States in the late 1960s, which took priority over the defence of the $35 gold price. This deprived the system of its anti-inflationary anchor; its disintegration was an inevitable response to American political decisions -which under different Administrations might have gone a different way.
The thrust of this argument is that, contrary to received wisdom, there was nothing ineluctable about the demise of fixed-exchange rate systems in the last century. Without the two world wars and the intervening Great Depression, it is perfectly conceivable that the world would have evolved a modified version of the gold standard system to which Keynes looked forward before the first world war: one in which gold occupied much the same position as the monarch does in the British constitution. As time erodes the effects of these disturbances on the world’s political and economic structures, re-fixing of major currencies may well be seen as part of the evolving process of reconnecting our own world with the world which existed before 1914.
iii The Theory of Monetary Independence
A monetary authority is said to be independent if it can fix the rate of inflation or level of unemployment at whatever it wants it to be. This is said to require either a floating currency or control over capital movements. Friedman put the case for floating: (1953, 210): ‘Flexible rates would allow each country to pursue the mixture of unemployment and price trend objectives it prefers….’ Keynes stated the case for fixing with capital controls: 'In my view the whole managment of the domestic economy depends on being free to have the appropriate rate of interest without reference to rates prevailing elsewhere in the world. Capital control is the corollary of this'. (1942)
In the history of economic thought, the doctrine of monetary independence first emerged, in the writings of Fisher (1911), Wicksell (1936,1898), and Keynes (1923), as the ‘modern’ method of achieving price stability. It was put forward as an alternative to, or modification of, the gold standard, under which the internal price-level was pushed around by the vagaries of gold production. The monetary reformers of that day were so keen on price stability because they understood that there was a connection -though how this worked was less clear -between price fluctuations and fluctuations in the real economy - it was the latter which it was their real object to prevent. But in order to establish the possibility of deliberate ‘inflation targeting’ they had to rely on a rigid version of the Quantity Theory of Money. They also had to argue that monetary policy could be made ‘politician proof’.Both sets of arguments have been taken over by modern advocates of monetary independence.
The feasibility of inflation targeting presupposes exogenous money; the credibility of inflation targeting presupposes exogenous politicians. Belief by agents in both is a necessary condition for full credibility of inflation targeting. Since the management of expectations is now a key part of the management of money, any failure of credibility arising from one or other of these sources undermines the value of the commitment.
The model of the macro-economy underlying all nominal targeting (whether the target is the inflation rate or the exchange rate) is the QTM. If the central bank can fix the rate of growth of the money stock it can control the price level. In the early days of monetarist enthusiasm, it was thought that the best policy for low inflation was to combine ‘a money supply rule with floating exchange rates’. (Congdon, 1992, 20) Today, central banks attempt to control prices by using interest rates to regulate the volume of bank credit ( deposits). The Bank of England’s current inflation target is 2.5 per cent a year. To achieve this the Bank (more exactly, its Monetary Committee) raises the interest rate when inflation is expected to go above target, and lowers it when it is expected to go below. The Bank thus enjoys a ‘constrained discretion’, allowing monetary policy to be expansionary in response to an anticipated output decline. In its attention to both prices and output, and its rejection of a simple link between money and prices, monetary policy today is broadly Keynesian. (Skidelsky, FT, 16 August 2001 ) The main difference from the old days is that fiscal policy is now treated as a support for monetary policy, rather than the other way round. ‘Inflation targeting’ is based on forecasts of money growth, prices and output, as well as of ‘inflation’ and ‘output’ gaps. All these are subject to a high level of uncertainty. (By contrast, the only indicator for monetary policy under a fixed-exchange rate system is trends in the movement of reserves.)
Current arrangements for the conduct of monetary policy have sought to insulate it from political interference-from what Keynes called the ‘wicked Chancellor problem’. Central banks have been given varying degrees of independence from political control. But the problem of the endogenous politician remains. Typically,the inflation targets themselves are decided by governments, and are subject to change in the light of circumstances. For example, the Reserve Bank of New Zealand Act of 1989 specifies that price stability is to be defined by agreement between the government and the governor of the of the central bank and is subject to revision.The view that politicians are thus ‘exogenous’ to the conduct of monetary policy is illusory -probably even in the case of the ECB which is mandated to price stability by the Maastricht Treaty. G3 countries are reluctant to commit themselves to long-term inflation targets. ‘Everyone knows that the Federal Reserve operates some form of inflation targeting regime, but the exact policy objective is vague’. (FT, 24 April 03) In all countries with flexible currencies, governments retain unfettered discretion to adjust the inflation rate in the light of unilaterally-determined circumstances. Contrast this with the commitment under Bretton Woods to alter exchange-rates only with IMF agreement.
But surely, it will be argued, the low inflation rates achieved in the 'mature' economies since the 1980s is proof that 'inflation targeting' works. There is an alternative view which regards the inflation rate as the outcome of the ‘structure of the economy’. This issue goes all the way back to the Currency versus Banking school debate of the 1840s. In essence, the Banking school contended that there could never be any over-issue of notes: the supply of money was always exactly proportioned to the demand for money (loosely, the needs of trade). One of the most prominent modern exponents of the Banking view was the late Nicholas Kaldor: ‘an excess in the supply of money cannot come into existence; and if it did, it would be automatically extinguished through the repayment of bank indebtedness (or its equivalent)…’ (Kaldor, 1985, 8) On this view, a monetary authority which has regard for either the unemployment rate or the solvency of the banking system is bound to accommodate the supply of money to any demand for it.
Kaldor did not claim that monetary policy was impotent. But it works on prices indirectly through its impact on the the level of output and employment. A deflationary policy hits output and employment; the reduction in activity brings down prices in its train. On this view, it was the reduction in the ‘demand for money’ (deposits) in the 1980s that drove inflation to the low levels which have persisted since.
The heretical view that real structural forces influence nominal variables is also the basis of Roger Bootle’s explanation of current price behaviour. (Bootle,1997). The reason for inflationary pressures in the ‘golden age’ of the 1950s and 1960s lay in the institutional structures of the economy of that period which diminished price competition and inhibited market forces: managerial capitalism and producer power, at the root of which lay economies of large-scale production. Bootle argues that globalization has led to a systematic decline in inflationary pressure. Since the 1980s, changes in technology, weakened unions, a more competitive climate, both domestically and internationally, and the changing composition of the labour force, have eased the pressure on central bankers.
On this view it is not low inflation targets which have maintained low inflation; it is low inflation which has made possible low inflation targets.
Keynesians support ‘monetary independence’ out of concern for employment rather than for price stability The ‘Keynesian’ argument for monetary independence is straightforward. A fixed exchange rate places the burden of adjustment to economic disturbances on variations in money wages and prices. If these are inflexible, then unemployment results, unless the exchange rate can be used as a shock absorber. A closely related point is that in a fixed exchange regime, monetary policy is powerless, if capital is mobile. The home interest rates cannot differ from the foreign interest rate because any difference would be eliminated by capital movements. As a result, the country loses one of its principal weapons in fighting slumps.Given the current taboo on deliberately budgeting for a deficit, interest rate-policy has moved into pole position in the fight against unemployment. This is the basis of the charge that the 'euro' has locked the main European economies into a sub-optimal level of activity.
Prima facie, this argument is compelling. Few economists doubt that Germany would have benefited from a lower interest and exchange rate than those produced by the policy of the ECB. However, the prolonged stagnation of Japan, which has enjoyed monetary independence, casts some doubt on the validity of this view. Despite pushing nominal interest rates to near zero, the Bank of Japan has been unable to prevent prices from falling and thus the appreciation of the real interest rate. A positive real rate of return on money can make holding it more attractive than buying investments, or goods and services. In this case, monetary policy -the expansion of some monetary aggregate - is ineffective in stimulating the economy. Tim Congdon (2003) calls this the ‘narrow liquidity trap’ -all that massive bond purchases by the central bank do is to swell the commercial banks’ cash ratios. This situation arises when loans to the private sector are expected to be relatively less profitable than holding cash or when yields on bonds have fallen so low that the most probable next move is an upward yield movement which will reduce their capital value. Congdon (2003,8) cites figures which show that in 2001-2 massive bond purchases by the Bank of Japan resulted in a trebling of banks’ cash reserves. But their loan portfolios declined by almost 10%.
In theory, purchases by the Bank of Japan of foreign securities could stimulate the economy by driving down the exchange rate of the yen. However, perverse exchange-rate expectations can prevent this effect. If Japanese savers know that the Bank of Japan is committed to a strong currency, they would expect any currency depreciation to be reversed. If the Bank bought large quantities of foreign currency to drive down the exchange-rate, it would pay for them by creating extra local currency deposits. But expectations of bounceback would cause local currency to accumulate rather than be spent. If the demand for local currency becomes infinitely elastic with respect to the exchange rate at the same time as the demand for money to hold is infinitely elastic with respect to the interest rate, monetary policy is completely impotent.
Krugman (1999, 73) has argued that the solution to the Japanese recession is to have an inflation target of say 3 per cent. If believed, this would stimulate spending through various channels. But why should agents believe that the Bank of Japan can achieve a 3 per cent inflation rate? It could directly buy securities or corporate bonds.But such a cash injection would not stimulate the economy if the Bank is simply seen to be bailing out unprofitable enterprises. The government could spend the money itself (fiscal policy), but with a debt/GDP ratio already well over 100 per cent, this may well cause real interest rates to rise even further. Martin Feldstein ( ) has proposed an ingenious ‘last resort’ measure:the authorities should cut sales tax to zero and then announce that it will be raised to 10% at a definite future date. The idea dates back to the stamped money plan of Silvio Gesell. The aim is to create the definite expectation that money will lose value if it is not spent now. But this is unorthodox fiscal policy, not monetary policy.
Bordo argues that with a common inflation rate, nominal exchange rates among the main countries would tend to remain stable, with changes in nominal rates tracking changes in real exchange rates only. This is not borne out by experience. Volatility has been much greater than can be explained by inflation differentials,and there have been prolonged misalignments. For example, gyrations of the dollar-yen rate between 1995-9 can’t be explained by inflation or productivity-growth differentials. Speculator-led changes in the relative prices of the two currencies imposed large movements in real exchange rates (competitiveness) on the two economies: that is, financial movements resulted in large and unnecessary resource movements. Against the view that nominal exchange rate changes are needed to absorb real shocks, it may be argued that many real shocks are produced by nominal exchange rate changes. Richard Cooper writes: ‘In short, movements of exchange rates, while providing a useful shock absorber for real disturbances to the world economy, are also a substantial source of uncertainty for trade and capital formation, the wellsprings of economic progress’. (1999, 115)
The conclusion to which these observations point is the following. If underlying conditions are making for the stability of some aggregate variable, be it prices or output, monetary targeting can marginally increase the stability of the variable. The fact that policy works partly through expectations adds power to it. But if shocks are large, there is little that monetary policy can do on its own. It is surely worth questioning whether the relatively minor advantages of unfettered monetary independence are worth the major costs of exchange rate instability.
iv. Conclusion
The issue posed by Bordo's paper is whether the exchange-rate outcome he favours -which is a system of stable, but adjustable rates between the mature economies -is best achieved indirectly by low inflation targeting or by directly targeting exchange-rates. Most of the arguments in favour of the first emphasise the lack of credibility of the second. I have argued that there are credibility problems with both, but that targeting the exchange rate is a stronger commitment than targeting the inflation rate. Furthermore, evidence suggests that even if low inflation prevailed across the 'mature' economies 'undefended' exchange rates would remain volatile. The only secure way of achieving stable, but adjustable exchange-rates is to commit to a system which has that as its object. I agree with Richard Cooper that a ‘cost-benefit calculation for flexible versus fixed exchange rates will gradually alter the balance against flexibility, even for large countries’. (117).
Bordo, M.D. (2003) ‘Exchange Rate Regime Choice in Historical Perspective’, Henry Thornton Lecture at the Cass Business School, City University, London, England, March 26 2003. (Mimeo).
Congdon, Tim (1992) Reflections on Monetarism.
Congdon, Tim (2003), ‘Debt Management and Deflation’, Lombard Street Research Limited, April 4, 2003
Cooper, Richard N. (2000), ‘Exchange Rate Choices’, in Jane Sneddon Little and Giovanni P. Olivei, ‘Rethinking the International Monetary System.
Eichengreen, B. (1994), International Monetary Arrangements for the 21st Century
Feldstein, M.
Fisher, Irving (1911), The Purchasing Power of Money.
Friedman, M. (1953) ‘The Case for Flexible Exchange Rates’, in Essays in Positive Economics.
Kaldor, N.(1985), ‘How Monetarism Failed’, Challenge, May-June 1985
Keynes, J.M. (1923), A Tract on Monetary Reform
Krugman, P. (1999), The Return of Depression Economics
Nurkse, R. International Currency Experience, League of Nations.
Obstfeld, M. and Alan M. Taylor (1998) ‘The Great Depression as a Watershed: International Capital Mobility over the Long-Run’ in Bordo, M.D., Claudio Goldin, and Eugene N. White eds. The Defining Moment: The Great Depression and the American Economy in the Twentieth Century.
Skidelsky, R. (1999), ‘Historical; Reflections on Capital Movements’, in Capital Regulation: For and Against, Social Market Foundation, London
Wicksell, K. (1936), Interest and Prices: A Study of the Causes Regulating the Value of Money (published in German, 1898)
Bookmark and Share


There are no comments for this entry...

Add comment: