Co-authored with Vijay Joshi
Published in Rebalancing the Global Economy: A Primer for Policymaking by Stijn Claessens, Simon Evenett and Bernard Hoekman (eds)
This chapter argues that the Keynes Plan of 1941 for dealing with the trade imbalances of his time is highly relevant to the problem of East Asian-US imbalances today. Just as the first Bretton Woods system rested on a “grand bargain” between the US and Britain, a new Bretton Woods would test the statesmanship of the US and China.
The Problem of Global Imbalances
As the world tentatively scrambles out of the worst recession since World War II, the future of the world monetary system remains firmly off the agenda. The global downturn had many interacting causes, but a tenable view is that the accumulation of reserves by a handful of countries in East Asia and the Middle East played a key permissive role in the collapse. Between 2003 and 2009 (measurable) global reserves increased from $2.6 trillion to $6.8 trillion – an average annual rate of increase of about 15%, at a time when global GDP grew at an annual rate of 4.4%. This amounted to a big increase in deflationary pressure. However, the fact that the reserves were held mainly in dollars allowed the US to avoid deflation, and instead run a “Keynesian” domestic policy which set the stage for an unsustainable asset and consumption boom. In short, there was a nexus connecting reserve accumulation by China and expansionary monetary and fiscal policy in the US.
The purpose of this chapter is to how that the Keynes Plan of 1941 for dealing with the trade imbalances of his time is highly relevant to the problem of East Asian-US imbalances today. It proposes two mechanisms for alleviating the current problem of “symmetrical non-adjustment”. The first part of the essay will examine the historical context of the Keynes Plan and the breakdown of the Bretton Woods system; the second will analyse the present problem of non-adjustment and steps which can be taken to overcome it.
The Keynes Plan of 1941
In the 1920s Keynes had come to see deflation as the main cause of British unemployment; and the main source of deflationary pressure as the unbalanced creditor position of the US. In theory, the international gold standard, which was the currency regime of the time provided for automatic and symmetrical adjustment of current account imbalances. Prices would automatically rise in the gold gaining countries and would automatically fall in the gold-losing countries, thus restoring the equilibrium of exports and imports between the two. But Keynes had come to realise, as he put it in 1941, that adjustment was “compulsory for the debtor and voluntary for the creditor”. If the creditor does not choose to make, or allow, his share of the adjustment, he suffers no inconvenience: while a country’s reserve cannot fall below zero, there is no ceiling which sets an upper limit. The same is true if private capital flows are the means of adjustment. “The debtor must borrow; the creditor is under no…compulsion [to lend]”.
During the Great Depression itself, creditor “hoarding” had been aggravated by the flight of capital from deficit to surplus countries. Following the financial crisis of 1931, the gold standard collapsed, the international capital market seized up, and the major countries resorted to tariffs, competitive devaluations, and bilateral trade agreements to balance their accounts. The international payments system created in the 19th century ceased to function.
Keynes’ Clearing Union plan of 1941 was designed to avoid a repetition of this disaster. It would retain the advantages (as he saw them) of a fixed exchange rate system while avoiding the asymmetric costs of adjustment. The essential feature of his plan was that creditor countries would not be allowed to sterilise their surpluses, or charge punitive rates of interest for lending them out; rather these surpluses would be automatically available as cheap overdraft facilities to debtors through the mechanism of an international clearing bank whose depositors were the central banks of the union.
All residual international transactions – those giving rise to surpluses and deficits – were to be settled through “clearing accounts” held by member central banks in an International Clearing Bank (ICB). Member banks could buy foreign currencies and sell their own against debits and credits to their accounts at the ICB (denominated in bank money or “bancor”) up to an “index quota” equal to half the average value of their country’s international trade over the previous five years. Deposits of bank money (credits and debits) would be created by surpluses and deficits and extinguished by their liquidation. Each national currency would have a fixed but adjustable relation to a unit of bank money (bancor) which itself had a fixed relationship to gold. But though bancor could be obtained for gold, it was not convertible into gold. Keynes’ long term aim was to de-monetise gold and make bancor the ultimate reserve asset of the system. By increasing or reducing the total of quotas, the Bank’s managers would be able to vary the supply of bancor contra-cyclically.
Keynes sought to secure creditor adjustment without renouncing debtor discipline. To this end his scheme aimed to bring a simultaneous pressure on both surplus and deficit countries to “clear” their accounts. Persistent creditor countries would be allowed or required to revalue their currencies, unblock any foreign-owned investments, and be charged rising rates of interest (up to 10 per cent) on credits running above a quarter of their quota. Any credit balances exceeding quotas at the end of a year would be confiscated and transferred to a Reserve Fund. Persistent deficit countries would be allowed or required to depreciate their currencies, to sell the ICB any free gold, and prohibit capital exports. They would also be charged interest on excessive debits. If all countries were in perfect balance at the year’s end, the sum of bancor balances would be exactly zero.
The Keynes plan was vetoed by the US, which was not prepared to allow its “hard earned” surpluses to be automatically at the disposal of “profligate” debtor countries. Instead the Bretton Woods Agreement of 1944 set up an International Monetary Fund to provide short-term financial assistance for countries in temporary balance of payments difficulties. The IMF was a fund, not a bank, into which members would pay contributions or quotas made up of gold and domestic currencies. (The total resources of the Fund were set at $8bn, as opposed to $25bn. for Keynes’ ICB). The Fund would supply foreign currencies to members up to the limit of their quotas, provided they corrected their domestic policies. Par values of currencies would be fixed in terms of gold, which could be altered only to correct a “fundamental disequilibrium”. Both the Keynes Plan and the IMF system relied on capital controls to prevent the destabilising flows of “hot money”.
The crucial point was that, while accepting the idea of fixed, but adjustable exchange rates, the Fund provided no mechanism to stop persistent reserve accumulation. It upheld, that is, the orthodox doctrine of debtor adjustment, and to that extent failed to solve the Keynes problem of persistent creditor hoarding. The contrast between the two plans was deliberate. For Keynes and the British, the problem which brought down the gold standard in 1931 had arisen from the refusal of the surplus countries to spend their surpluses; for the Americans it had arisen from the monetary indiscipline of the deficit countries.
The Bretton Woods system in practice
That the Bretton Woods fixed exchange rate system, which lasted from 1949 to 1971, did not reproduce the deflationary character of the inter-war system, was due to the general commitment of governments to full employment policies backed by the “dishoarding” policies of the US. America flooded the “free” world with dollars, to such an extent that by the late 1960s it was starting to run a balance of trade deficit itself. The boot was now on the other foot, but the need for the deficit country (now the USA) to deflate was circumvented by the role of the dollar as the world’s main reserve asset. As its trade deficit widened, the USA printed an increasing quantity of dollars to cover its unrequited imports. The surplus countries accumulated American dollar liabilities which they invested in US Treasury bonds. The US did not have to restrict domestic credit by raising interest rates since the dollars it printed came back to it. In the absence of what would have been a major deflationary force, the world economy boomed for twenty years.
The fl aw in the system, as pointed out by Professor Triffin of Yale University, was that the increase in the liabilities of the key-currency country was bound to raise doubts about its ability to redeem these liabilities in gold. At the end of the 1960s, the French started converting their dollar reserves into gold. This brought about the predicted collapse of the gold-exchange standard in 1971. The dollar became inconvertible. A new supplementary international reserve currency, Special Drawing Rights (SDRs), had been set up, but since there was no mechanism for converting dollar balances into SDRs, the dollar continued to be the world’s main reserve asset in a mixed world of floating, fixed, and managed exchange rates.
In theory, floating exchange rates remove the need for any reserves at all, since balance of payments deficits and surpluses would not arise. But the need for reserves unexpectedly survived, mainly to guard against speculative movements of hot money which could drive exchange rates away from their equilibrium values. This happened throughout the 1980s. Starting in the late 1990s, after the East Asian Crisis, East Asian governments unilaterally erected a “Bretton Woods II”, linking their currencies to the dollar, and holding their reserves in dollars. This reproduced the expansionary benefits of Bretton Woods I, but at the cost of an increasingly unbalanced reserve position, as the dollar became progressively overvalued against the super-competitive renminbi.
Today’s problem of current account imbalances reproduces the problems which brought down both the old gold standard and its successor Bretton Woods system. The gold standard failed to provide for the symmetric adjustment of surpluses and deficits. The Keynes Plan was designed to replace asymmetric adjustment which brought deflation to the deficit countries by deliberate provision for symmetric adjustment through the International Clearing Union. The Bretton Woods system did not solve the Keynes problem. It upheld the orthodox doctrine of debtor adjustment, but, through the IMF, gave debtor countries time to “put their houses in order”. The deflationary pressure against which the Keynes plan was directed was solved not by the mechanisms he had envisaged, but by the voluntary “dishoarding” of its surpluses by the US. But this called into question the credibility of its promise to redeem dollars for gold. Today’s system can be characterised as one of symmetric non-adjustment: as long as the surplus-earning countries are content to hold their accumulating reserves in dollars, neither side is under any pressure to adjust.
However, the main issue today is no longer the “sustainability of the deficit”, but its effect on the economies of both surplus and deficit countries. A sequence of financial booms and busts is built into a system which brings no pressure for adjustment to bear on either creditors or the principal debtor (the U.S.). This is both irrational and costly. Unless steps are taken to re-balance global current accounts, we will be walking into the next crisis. To secure the automatic adjustment of current account surpluses and deficits was the object of the Keynes Plan of 1941. This should be the starting point of contemporary efforts to rebalance the world’s money.
The present non-system
As we have suggested, global imbalances played a part in causing the severe credit crunch of 2008-9. But they are also dangerous per se. They can lead to disorderly reversals triggered by large capital movements; and they can also provoke trade restrictions. It is a fair bet that a continuation of the global imbalances of 2006 would have led to a dollar crisis or a protectionist frenzy if the credit bubble had not imploded first. The imbalances have now decreased but could open up again when the world economy recovers. They thus continue to be a serious potential problem.
Today’s circumstances are different from Keynes’ day. Capital mobility is now much greater. We do not have adjustably- pegged exchange rates. Indeed, there is now nothing that can be called an exchange rate “system”. There is instead a wide diversity of exchange rate regimes. The reserve system is also different. It is centred on the dollar, not on gold. Today the dollar is the principal reserve currency, with the euro in second place but far behind. There also exists a fiduciary international central-bank money, the SDR, created in the late 1960s, but so far of minor quantitative importance. But despite the changed environment, Keynes’ insights, encapsulated in his Clearing Union proposals, are still highly relevant to avoiding future imbalances.
A necessary requirement of smooth international adjustment is a well-functioning mechanism for changing real exchange rates. That does not currently exist. Many major countries are floating but some (notably China) are not. Nonfloaters who run balance of payments surpluses are able to block real exchange rate changes by sterilising their reserves. International adjustment also ideally requires some international coordination of macroeconomic policies, at least sporadically. This requirement is conspicuous by its absence. (The cooperative demand stimulus in 2008/9 was an exception.) The absence of a satisfactory adjustment mechanism has resulted in the revival of the asymmetry strongly emphasised by Keynes. Adjustment pressures are concentrated on the deficit countries (unless the deficit country is a reserve-issuer like the US); the countries in surplus can get away without adjustment. A case in point is that of the emerging countries that have discovered the advantages of export-led growth. This strategy has yielded many benefits for these countries but it suffers from a fallacy of composition; the export surpluses must have counterpart deficits elsewhere. In other words, they can generate global imbalances.
The current reserve system is equally unsatisfactory. It is notable that greater capital mobility has increased, not reduced, the demand for owned reserves. Many countries have a rational fear of floating, as well as a rational fear of unstable capital flows; and reserves obtained by short-term borrowing can evaporate in a crisis. The sure-fire way of accumulating owned reserves is to run current account surpluses. East Asian countries were taught the value of owned reserves by the bitter experience of 1997 and the recent crisis has only confirmed this lesson. But substantial reserve accumulation, though rational for an individual country, is systemic nonsense if practised by many countries. Indeed it leads precisely to the global imbalances under discussion: reserve-hoarders run large surpluses while America, the main reserve issuer, runs large deficits. The connection between reserve accumulation and global imbalances is not a logical necessity. It is possible in principle for current accounts to be balanced and reserves to be generated by purely capital-account transactions, say long-term foreign direct investment by the US matched by short-term deposits by the recipient countries in US Treasuries. But as a matter of fact, the recent growth of the reserve mountain did predominantly have US current account deficits as its counterpart. And the underlying behavioural connection has psychological and empirical plausibility: the “exorbitant privilege” conferred by the power to issue reserves weakens balance of payments discipline on the issuer country and sooner or later tempts it to overspend.
The argument that current account surpluses are deflationary for the world is only partially correct. Certainly they are deflationary in the first instance for the deficit countries. But the deficit countries are now very likely to be committed to full employment. So their probable and understandable reaction is expansion; large imbalances are the consequential by-product. This is certainly a plausible description of events in the middle of this decade: the “glut” of savings in parts of the world evoked a Keynesian expansionary response in the US, which widened global imbalances. Of course the day of reckoning has to come in the end and has the potential to be strongly deflationary for the world since the burden of adjustment would fall on the deficit countries.
Reforming the non-system
What can be done to prevent global imbalances? Complete overhaul of the system is not going to happen. But major improvements are possible that would ameliorate the imbalances problem and enable an evolutionary development of the world monetary system in a desirable long-term direction. Here Keynes’ Clearing Union proposals are very relevant.
We begin with the adjustment mechanism. This needs to be improved whatever reserve system is adopted. Ideally, as seen above, we need better macroeconomic coordination as well as a better Exchange rate system. But these are subjects of great contention which are not going to be resolved easily or soon. A second best but major step forward would be to adapt Keynes’ idea of penalties on bancor imbalances; the contemporary equivalent would be to tax persistent and excessive current account surpluses. The numerical specification of “excessive” and “persistent” would have to be agreed. So would the rate of tax to be paid to the IMF: it must obviously be big enough to affect behaviour. Taxing excessive reserve accumulation would be an inferior alternative because the connection between reserves and the current account is loose; moreover, reserves can be hidden in various ways.
We now turn to the reserve system. Keynes presciently wanted to abolish altogether the use of national currencies as international reserves and substitute “bancor” in their stead. Such a change would strike at the root of the self-insurance demand for dollars. And it would do so by enabling countries to acquire fiduciary reserves which they own but do not have to earn since they would be supplied by an international central bank. This is a bridge too far to reach quickly. But it would be possible and desirable to take immediate steps to raise the SDR share of world reserves, within the context of a dollar-based reserve system. What is needed is an amendment of the IMF Articles to enable SDRs to be more flexibly created by substantial regular emissions, as well as by occasional issuance to meet world liquidity crises (to be followed by withdrawal when the crisis is over.) Moving decisively towards promoting the use of SDRs would reduce the need for countries to run current account surpluses to accumulate dollar reserves. Concomitantly, it would also help to reduce the “exorbitant privilege” of reserve-issuers and distribute the seignorage from reserve creation more equitably, promote a more symmetric adjustment mechanism, make the IMF a more genuine lender of last resort and reduce the risk of instability caused by switches between reserve currencies. Of course the appeal of the SDR would be materially enhanced if it were transformed into an asset that can be held by the private sector, not central banks alone. There are many ideas on the table for achieving that goal over the long haul. But promoting central-bank use of SDRs, as proposed above, need not wait on such schemes, and would bring about a marked improvement in the functioning of the world monetary system, even in their absence.
The reforms above should be accompanied by setting up a “substitution account”, lodged in the IMF, to enable countries to convert their reserve holdings into SDRs that are by their nature more stable in value than any single reserve currency. The main bone of contention would obviously be: who takes over the exchange risk that countries making use of the facility will want to shed? The advantages of this scheme would be two-fold. It would enable an increase in the volume of SDRs and a reduction in the quantity of foreign currency reserves. And it would open the road to a bargain with China. If the terms of the substitution account were such as to give China an incentive to convert its dollar reserves into SDRs, China may in return agree to penalties on excessive current account surpluses.
Just as the first Bretton Woods system rested on a “grand bargain” between the US and Britain, so a new Bretton Woods would require an agreement between the leading surplus and the leading deficit country. The challenge to the statesmanship of the US and China is to strike one.