Which of the following was created to regulate fixed exchange rates and enforce the rules of the international monetary system?

1.1 Introduction

The 1944 Bretton Woods Conference, which created the International Monetary Fund (IMF) and the International Bank for Reconstruction and Development (World Bank), and the San Francisco Conference, which created the United Nations one year later, were major landmarks in international cooperation—true ‘acts of creation’, to use the title of one of the best-known books on the founding of the United Nations (Schlesinger 2004). These success stories were particularly remarkable in the light of the failures of international political and economic cooperation in the 1930s. There were, of course, disappointments, particularly the incapacity to launch an additional leg of the system of economic cooperation, the International Trade Organization agreed to in Havana in 1948, as the US Congress failed to ratify the agreement; only one of its components, the General Agreement on Tariffs and Trade, approved one year earlier, was put in place. Almost half a century later, the World Trade Organization was created. In any case, there has never been another moment in the history of international cooperation that matches the end of the Second World War and the early post-war years.

In the economic area, the success of cooperation was reflected, in particular, in the rapid reconstruction of Western Europe and Japan, which led to the period of the fastest economic growth and, particularly, the fastest growth of international trade in world history. There were also disappointments, notably the inherent design problems, which in the case of the international monetary system—the subject of this book—finally led in the early 1970s to the collapse of the Bretton Woods arrangements, the failure to agree on an alternative system, and the de facto rise of the ‘non-system’ that has survived until the present. To this we can add the incapacity to line up the Communist countries as members of the Bretton Woods institutions, until 1980 in the case of China (when it took from Taiwan the membership in the IMF) and after the fall of the Berlin Wall in the case of the Soviet bloc.

This chapter focuses on the international monetary system, as an introduction to the issues that are analysed in detail in the rest of the volume. The histories of the international monetary system and of the IMF in particular have, of course, been the subject of significant attention. This includes old and new histories of US–UK negotiations in the late war years and the agreement finally reached at Bretton Woods (Conway 2015; Gardner 1969; Steil 2013), as well as more recent analyses of the role of developing countries in those negotiations (Helleiner 2014). It also includes academic histories of the international monetary system (Eichengreen 2008; Helleiner 1994; Yago, Asai and Itoh 2015), the views of protagonists of that history (Solomon 1982), and the official and semi-official histories of the IMF (de Vries 1976, 1985, 1987; Horsefield 1969; James 1996). The chapter does not, therefore, aim to make a detailed reconstruction of the history of the system as such, but rather to serve as a historical background to the contemporary issues that are analysed in the rest of this volume: the genesis of the problems faced by the system at different times, their similarities and differences, and the role of emerging and developing countries in the system.

The chapter is divided into six sections, the first of which is this introduction. The second looks at the background of the debates and the design of the Bretton Woods system. The third analyses the tensions that the Bretton Woods monetary system faced from the 1960s until its collapse in the early 1970s. The fourth looks at the management of the collapse, the failure to reach an agreement on a new system, and the resulting non-system or ad hoc arrangements that followed. The fifth considers the following quarter century or so, during which these arrangements matured. The last section looks at current issues, which may be seen as the construction of a broader global ‘financial safety net’, to use a term that has become fashionable. This process started before the North Atlantic financial crisis1 but developed fully after the outbreak of that crisis.

1.2 The Bretton Woods Monetary System

The major objectives of international monetary cooperation as agreed at Bretton Woods are best captured in Article I.ii of the IMF Articles of Agreement, which states that the purpose of the IMF is: ‘To facilitate the expansion and balanced growth of international trade, and to contribute thereby to the promotion and maintenance of high levels of employment and real income and to the development of the productive resources of all members as primary objectives of economic policy’. This objective reflects the central relation that, in the conception of the Bretton Woods architects, the new international monetary system had with the reconstruction of world trade after its collapse during the Great Depression and the Second World War. It also shows the centrality of the new economic ideas that came from the Keynesian revolution, which placed employment as the central objective of macroeconomic policy. In a more indirect way, the growth objectives of developing countries were captured in the reference to the ‘development of the productive resources’ of its members.2

The rest of Article I can be read as instruments to achieve this major objective: (i) to create a permanent institution ‘to promote international monetary cooperation’ (Article I.i); (ii) ‘to promote exchange stability, to maintain orderly exchange arrangements among members, and to avoid competitive exchange depreciation’ (Article I.iii), which was deemed as essential to reconstruct international trade; (iii) to establish ‘a multilateral system of payments in respect of current transactions’ that would eliminate the ‘foreign exchange restrictions which hamper the growth of world trade’ (Article I.iv); and (iv) to provide IMF financing ‘to correct maladjustments in their balance of payments without resorting to measures destructive of national or international prosperity’ (Article I.v), meaning in this regard policies that could negatively affect employment in the country adopting them or generate negative spillovers that would have that effect on other countries.

Each of these provisions reflected the existing consensus on the problems that the previous system faced, particularly in three areas. The first was the view that the ‘rules of the game’ of the gold standard had been pro-cyclical and therefore guaranteed exchange rate (and monetary) stability at the cost of the employment objectives. Rather, the view was now enshrined that countries should undertake balance-of-payments adjustment but maintaining enough policy space to pursue the employment objectives of macroeconomic policies—in particular the capacity to adopt counter-cyclical policies during crises, as well as the growth objectives of developing countries—the ‘development of the productive resources of all members’. This required the creation of new policy instruments to pursue the consistency between internal and external balances, particularly balance-of-payments financing and the possibility of adjusting the exchange rate to guarantee such balance.

The second consensus was related to the chaos in exchange rate movements and international payments generated by the Great Depression and the final collapse of the gold standard, particularly after its progenitor, the United Kingdom, finally abandoned it in September 1931; it had already been abandoned during the First World War but was restored in 1925 under conditions that are generally perceived to have been unsustainable. Some countries had already abandoned the gold standard before Britain and others were forced to follow, sometimes with a lag. In the case of the United States, it was abandoned after a series of decisions adopted between April and June 1933 when gold coins and certificates were taken out of circulation, and convertibility of dollar3 bills for gold and the right of creditors to demand payments in gold were abrogated, and in January 1934 when the Gold Reserve Act forced banks to hand their gold to the Treasury and allowed the President to change the official price of gold, which was immediately increased from US$20.67 to US$35 per troy ounce. Following a pattern that several countries had already faced when crises hit during the gold standard years, some developing countries had also abandoned convertibility before the United Kingdom, and others soon followed. A few countries retained the gold standard for a few more years, but the most important of them, France, finally abandoned it in 1936. However, this process led to competitive devaluations as well as the use of foreign exchange controls by many countries that hampered the international system of trade and payments. To facilitate trade and avoid competitive devaluations, it was then agreed at Bretton Woods that exchange rates should be normally fixed but could be adjusted ‘to correct a fundamental disequilibrium’ (Article IV, Section 5). This concept, like some others in the agreement, was never clearly defined, but its meaning was perhaps self-evident, as reflected in the severity of the balance-of-payments crises faced by specific countries at different times. The Great Depression had also led to broad-based protectionism and, under the leadership of Germany, to a myriad of bilateral trade agreements that ended up destroying multilateralism in trade and payments.

The third consensus was that capital flows had also had a large speculative component during the 1930s that had hampered rather than supported exchange rate stability. For this reason, the commitment to eliminate restrictions on current transactions that was adopted at Bretton Woods was not extended to capital flows. So, the freedom to regulate capital flows that did not hamper current transactions was adopted as an important element of the system. As the Articles of Agreement still read: ‘Members may exercise such controls as are necessary to regulate international capital movements, but no member may exercise these controls in a manner which will restrict payments for current transactions’ (Article VI, Section 3).

The proposals that the two fathers of the IMF, John Maynard Keynes and Harry Dexter White, placed on the table before the negotiations, shared these points of consensus and had other common elements, but also had disagreements in other areas. Notable among them was the design of the global reserve system.4 In particular, Keynes (1942–3) held strongly the view that the major problem of all international monetary systems had been that they forced asymmetric balance-of-payments adjustment on deficit versus surplus countries: the former were forced to adjust, as they generally lacked adequate external financing or adequate reserves to manage crises, whereas surplus countries did not face similar pressures. Keynes’ obsession with this issue was, of course, related to the fact that this asymmetry generates a global contractionary bias during crises.

Keynes’ proposal to create an International Clearing Union was aimed at correcting this basic asymmetry by forcing surplus countries automatically to finance those in deficit—obviously within certain limits. Under the circumstances that characterized the world after the Second World War, it was clear, however, that this would imply that the United States was bound to become the major surplus country and would therefore have to provide large amounts of automatic financing. Therefore, it was unacceptable to the host of the Bretton Woods negotiations. White offered in return what came to be known as the ‘scarce currency clause’, under which other countries were authorized ‘after consultation with the Fund, temporarily to impose limitations on the freedom of exchange operations in the scarce currency’ (Article VII, Section 3b). However, this provision, which curiously is still in place, was never legally used5 and the ‘dollar shortage’ that characterized the early post-Second World War period was managed in a different way—in the case of Western Europe and Japan, as we shall see, through US aid and specific arrangements for intra-European payments. The asymmetric adjustment problem, therefore, continued to be a feature of the system designed at Bretton Woods as well as of the non-system that succeeded it.

Of the two currencies that had served the role of major international monies since the First World War, it was also clear to negotiators that the dollar was the only one that could be placed at the centre of the system. Sterling could not play that role as the United Kingdom would inherit from the Second World War major debt obligations with the United States as well as large sterling balances in the hands of other central banks, primarily but not only of the sterling area—an arrangement that had de facto been in place since September 1931 and was formalized in 1939 (Schenk 2010). Indeed, the tough negotiations between the United States and the United Kingdom to clean up the obligations that the latter had assumed during the war under the ‘lend–lease’ arrangement (which added up to US$22 billion) led to an agreement on a fifty-year loan of US$3.75 billion at 2 per cent interest and a thirty-year loan of US$650 billion at 2.375 per cent, but conditional on convertibility for current sterling-area operations that would have to be adopted within fifteen months. Fulfilling this commitment, the United Kingdom restored convertibility on 15 July 1947, but the rapid reduction of reserves forced it to suspend it again slightly over a month later, on 20 August (United Nations 1948: 18–19; Steil 2013: ch. 9 and 10). This early sterling crisis was the background for the flexibilities in the restoration of current account convertibility that were adopted from the late 1940s and which continued to prevail through the 1950s with European members and Japan, beyond the initial transition that was agreed (which formally came to an end in early 1952) and much longer with developing countries.

As much as it was clear that sterling could not be restored to the central role it had held in the past, and that it could not even be convertible, it was also clear to White that placing the dollar at the centre of the system was possible only if the dollar was backed by gold vis-à-vis other central banks. A peculiar system was then agreed in which dollar reserves of central banks were convertible into gold at the price fixed by the United States in January 1934, although maintaining the inconvertibility of dollars in circulation for gold established in 1933 for private agents. In fact, transactions among central banks in gold had continued after the abandonment of the gold standard in the 1930s—and in the case of the US, at that official price.

The system that then evolved, therefore, had four distinctive features:

•

A global reserve system based on a dual gold–dollar standard, but with dollar reserves being the most dynamic element in reserve accumulation.

•

A system of fixed exchange rates, but adjustable under ‘fundamental disequilibrium’. In the Articles of Agreement, it was called a system of ‘par values’, but came to be known as ‘adjustable pegs’ in the economic literature. Under the original Article IV, countries were supposed to consult the IMF before modifying their exchange rates, a process that was supposed to be fairly automatic if the variation was under 10 per cent, but the Fund could object if it was larger. However, this possible objection was rarely used6 and, if anything, it rather became a frequent practice for the Fund to pressure countries (particularly developing countries) faced with balance-of-payments deficits to depreciate their currencies; there was also pressure on surplus countries to appreciate, but in this case, the IMF’s leverage was very limited. Countries were also prohibited from engaging in ‘any discriminatory currency arrangements or multiple currency practices except as authorized under this Agreement or approved by the Fund’ and, if they had them in place before the agreement, to agree with the Fund on ‘their progressive removal’ (Article VIII, Section 3).

•

Convertibility for current account transactions, which would be achieved in a gradual way in countries that lacked it—initially the broad majority of IMF members. In contrast to that, countries maintained the capacity to manage capital flows to insulate them from speculative capital flows and, therefore, there was no commitment to capital account convertibility (i.e., in current terminology, no obligation to liberalize the capital account).

•

Official balance-of-payments support, financed by quotas, but limited in size, as they were supposed to finance only current account deficits. As country quotas were set one-quarter in gold and the rest in national currencies, and countries could initially access their full quota, this essentially meant that they could temporarily use their national monies to buy international currencies—a system that may thus be understood as a generalized swap arrangement. No conditionality was initially associated with such support, but it was introduced in 1952, when the stand-by agreements were created, which allowed automatic drawing on the gold tranche plus conditional access to four additional tranches of 25 per cent, thus increasing access to 125 per cent of quota. Conditionality included putting in place macroeconomic policies to achieve balance-of-payments equilibrium and encouraging current account convertibility.

Whether this was a coherent system or not was a major subject of debate in the 1960s, when its flaws were identified (see Section 1.3). In the early years, the massive dollar shortage that characterized the world economy led to a significant exercise of flexibility in the fulfilment of the rules and to the adoption of complementary policies.

Flexibility was exercised, in particular, in relation to the commitment to current account convertibility and the elimination of multiple currency practices. So, exchange controls were maintained for much longer than originally envisioned (de Vries 1987: ch. 1). Also, many countries maintained multiple currency practices, including a parallel flexible exchange rate market for capital flows, which effectively meant that the exchange rate system had flexibilities that were not initially envisioned (Reinhart and Rogoff 2004). In contrast, countries were reluctant to modify their core exchange rates. A major depreciation of European currencies took place in 1949, which may be understood as the reflection of the fundamental disequilibrium created by the dollar shortage, but also by the strong balance-of-payments effects of the 1948–9 US recession on Europe and the rest of the world, with the United Kingdom being again strongly hit, including through capital flight (United Nations 1951: ch. 8). After that major realignment, changes in parities by advanced countries were rare. This may be seen as evidence of a commitment to the rules of the Articles of Agreement, but it could also be interpreted—as, in fact, it came to be interpreted—as evidence of the lack of adequate adjustment mechanisms in the Bretton Woods arrangements. Canada was, in turn, the pioneer in using flexible exchange rates (with some level of management), which it first put in place between 1950 and 1962.

The complementary policies were the Marshall Plan (European Recovery Act) and the European Payments Union (EPU). The former provided resources for the European reconstruction in amounts that neither the World Bank nor the IMF could have provided (Eichengreen 2008: ch. 4). Parallel support was also provided by the United States for the reconstruction of Japan. According to US balance-of-payments statistics, unilateral current transfers were massive in the decade 1946–55, reaching an average of 1.5 per cent of gross domestic product (GDP). In turn, the EPU became an excellent mechanism to multilateralize the myriad bilateral payments agreements that had been in place among its members for some time, but more importantly to reconstruct intra-European trade and payments while saving on the use of dollars for that purpose (Triffin 1957). The effects of the reconstruction of production capacities in Europe and Japan and the reconstruction of intra-European trade were already evident in the much weaker effects that the 1954 US recession had when compared to that of 1948–9 (United Nations 1955: ch. 5). The final success was the return of current account convertibility by fourteen Western European countries on 29 December 1958, which were joined by Greece five months later. This decision may be seen as the end of the most critical period of the dollar shortage. In 1961, nine European countries accepted the obligations of Article VIII of the Articles of Agreement on the elimination of multiple currency practices. Japan did so in 1964.

The transition took much longer in the developing world. Many of the countries maintained strong exchange controls and multiple exchange rates for an extended period. Indeed, as the situation of developed countries improved, that of developing countries became more difficult after the mid-1950s due to a cyclical downswing of commodity prices that lasted through most of the 1960s (Erten and Ocampo 2013a). So, as the first peak in the demand for Fund resources by developed countries eased in 1957–8, developing countries started their first period of large demand for balance-of-payments support in the late 1950s (Figure 1.1). As already indicated, an interesting element of the adjustment programmes that were put in place in developing countries was the active promotion by the IMF of changes in exchange rate parities.

Figure 1.1

IMF loans as % of world GDP

Source: Author’s estimates based on IMF database. World GDP according to the World Bank.

In terms of long-term adjustment of the global reserve system, a significant trend that was visible during these years was an increase in the gold reserves in the hands of Western European countries. This meant that the United States, which held about three-quarters of world gold reserves at the end of the Second World War (excluding the Communist countries), held only about half by the end of the 1950s. However, this was not associated with a reduction of US gold reserves as such, but was rather due to the increase in world gold reserves concentrated in European countries (Figure 1.2). The big winner was Germany, which by 1957 had surpassed Switzerland and would soon surpass the United Kingdom as the second-largest holder of gold reserves in the non-Communist world.

Figure 1.2

World gold reserves

Source: Author’s estimates based on IMF database.

Foreign exchange reserves, particularly dollar reserves, started to increase, although gradually, and continued to represent a minority share in overall reserves (see Figures 1.3 and 1.6). In contrast to initial predictions, the demise of sterling as the second reserve currency turned out to be a gradual process: the share of sterling in foreign exchange reserves fell from over 50 per cent in the early post-Second World War years, but it was still close to 30 per cent in the 1960s (Schenk 2010). This was thanks to the persistence of the sterling area, although with significant geographical changes in the demand for sterling assets, and the fact that a not insignificant part of world trade continued to take place in sterling, supported initially by dollar scarcity.7

Figure 1.3

Foreign currency holdings, 1948–90 (million SDRs)

Source: Author’s estimates based on IMF database.

Figure 1.6

Composition of reserves

Source: Author’s estimates based on IMF database.

1.3 Increasing Tensions and Collapse of the Bretton Woods Arrangements

As the dollar shortage came to an end, the international monetary system was shocked by the problems faced by the two major reserve-issuing countries. The United Kingdom experienced deep crises in 1961 and, particularly, in 1964–8, which led to a devaluation in November 1967. As a result, it became one of the major regular Fund borrowers, and was also supported by swap arrangements through the Bank for International Settlements (Borio and Toniolo 2006; Toniolo 2005: ch. 11). Such international support helped to smooth the decline of sterling as a global reserve, but the 1967 devaluation also forced the United Kingdom to sign the 1968 Sterling Agreements, by which it offered a guarantee of the dollar value of sterling balances as a counterpart to the commitment by sterling holders to limit diversification (Schenk 2010).

In the case of the United States, the strong trade and current account surpluses that had characterized the early post-war years came to an end in the early 1950s and experienced a further decline in the second half of the 1960s, though the trade and current balances generally remained in positive territory (see Figure 1.7). The weakening in the current account was reinforced by foreign investment by US firms and the adverse speculative attacks that it started to face. The United States managed this problem by regulating capital flows, with the introduction of the interest equalization tax in 1963 and restrictions on US investments abroad in 1965 (Solomon 1982: 47–9).

Figure 1.7

US imbalances

Source: Author’s estimates based on data from the US Department of Commerce, Bureau of Economic Analysis, except Figure B, series A, which is estimated from the Historical Statistics of the United States.

One of the results of the weakening of the US balance of payments and the decision of some European central banks to convert their dollar holdings into gold was the sharp reduction in the gold reserves of the major reserve-issuing country. US gold reserves fell by more than half between 1957 and 1968 and its share in world gold reserves dropped from 62 to 28 per cent (see Figure 1.2). The big ‘winners’ were the Continental European countries, particularly France, Italy, and Germany, in that order.

To manage the potential effects of this process on the international monetary system, the ‘gold pool’ was created in 1961 by eight countries (Belgium, France, Germany, Italy, the Netherlands, Switzerland, the United Kingdom, and the United States) to regulate the London price of gold and share responsibility for stabilizing the market. The Bank for International Settlements served as the secretariat for this new instrument, with the Bank of England as the operator (Toniolo 2005: ch. 11). The United States was obviously interested in slowing down its loss of gold reserves, but other countries also had a collective interest in the stability of the dollar in order to avoid potential capital losses on their dollar reserves. The mechanism worked smoothly in a first phase, between 1962 and 1965, thanks in part to incremental supply from South Africa, but faced increasing strains after 1966 due to growing doubts about the sustainability of the gold–dollar parity and declining sales by South Africa and the Soviet Union. France left the pool in the summer of 1967, and the very strong demand pressures led to its collapse in March 1968, leading to the decision of gold pool countries (without France) to adopt, on 18 March 1968, a two-tier market. The market shot up initially before stabilizing in 1969 (Eichengreen 2007: ch. 2).

The fall in US gold reserves was contained by the mix of the dual gold market and the implicit agreement by major partners to avoid converting their dollar reserves into gold. The stability in the gold market was also supported after the collapse of the gold pool by the balance-of-payments crisis that followed the 1968 political turbulence in France, which led to a loss of about a third of its gold reserves. Although the effective limits on the convertibility of dollar reserves into gold meant that the world had in a sense moved into a ‘reluctant dollar standard’ (Williamson 1977: 41), the growth of foreign exchange reserves ran at moderate rates until 1969: 5.7 per cent a year in 1957–69, similar to the 5.6 per cent rate of growth in 1949–57. It was only in 1970 and 1971 that, owing to the major problems faced by the US balance of payments, largely associated with the Vietnam War and speculative capital flows, the supply of foreign exchange reserves exploded (Figure 1.3). In short, the dollar shortage had been replaced, with a lag, by a dollar glut. Under these conditions, the global reserve system agreed at Bretton Woods was clearly unsustainable, and the United States finally unilaterally abandoned it on 15 August 1971 (see the discussion later in this section).

The 1960s were times of major debates on the international monetary system. Indeed, foreshadowing the views that became common after the evolution of international monetary arrangements in the 1970s, the famous Bellagio Group of thirty-two economists underscored the major inconsistencies in the existing arrangements by stating in 1964 that:

The present international monetary mechanism is not a simple and logical ‘system’. Rather, it is a set of arrangements which is the composite result of agreements, compromises among conflicting interests and opinions, adaptations to unforeseen developments in the evolution of world trade and finance, and precedents that grew out of ad hoc arrangements or individual policy decisions.(Machlup and Markiel 1964: 66–7)

The discussion on the nature of and the solutions to the problems that the system faced related to two central issues: those that characterized the global reserve system itself, and the lack of adequate adjustment mechanisms to correct payment imbalances,8 with a common underlying factor behind both of these problems being growing capital mobility.

In the early formulation by Triffin (1961, 1968), the essential problem was the instability of the global reserve system related to the use of a national currency as an international currency. The essence of the problem, in his view, was that the only way to provide increasing global liquidity was for the issuer of that currency (i.e. the United States) to run balance-of-payments deficits. Such deficits generated, in turn, the risk of loss in the confidence in that currency. The loss of US gold reserves was the major manifestation of that problem in the 1960s; other manifestations came after the original reserve system had been abandoned.

The formulation of the problems of the global reserve system by the Bellagio Group focused on the stability of the dual gold–dollar standard. Despite the very different opinions among its members on how to reform the system,9 one common view, which was largely in line with Triffin’s formulation, was that the international liquidity-creating mechanism was inefficient, as it ‘relies on deficits of the reserve-currency countries, which increases the ratio of liquid liabilities of these countries relative to their gold holdings, and thus [is] a growing threat to the value of the reserve holdings of other countries that undermines the confidence in the stability of the system’ (Machlup and Markiel 1964: 74–5, 81, 89, and 94).

A closely associated feature of the Bretton Woods agreement related to the asymmetries in the correction of the imbalances among major economies. However, the United States and other developed countries held significantly different views on the nature of this problem. This became the central issue of debate as the crisis of the system advanced, finally leading to its breakdown. For the United States, the major problem was the asymmetries in the adjustment of surplus versus deficit countries. This was, interestingly, the same issue that Keynes had raised during the discussions leading to the Bretton Woods agreement, but reflected the fact that the United States was now in a deficit position—much as Britain was and was expected to be after the Second World War. This was furthermore complicated by the fact that the United States had great difficulty in initiating adjustment, which basically implied that, given the central role in the system and the fixed gold–dollar parity, the depreciation of the US dollar relative to other currencies to correct its growing fundamental disequilibrium depended on the willingness of surplus countries to appreciate their exchange rates. In the view of US authorities, this made the country vulnerable to the export-led or even mercantilist policies of other countries.

For other developed countries, the main issue was the major advantage that the United States had in terms of financing its deficits by issuing its own currency, thus avoiding the need that deficit countries had to adopt contractionary macroeconomic policies to correct their payment imbalances. This was the ‘exorbitant privilege’, a term coined by Valerie Giscard d’Estaing when France’s finance minister but widely attributed to Charles de Gaulle. They also resented the loss of monetary policy autonomy that they increasingly faced in a world of growing capital mobility, and therefore the dependence on the monetary policy of the United States. When inflationary conditions started to build up towards the end of the 1960s, this was reflected, in their view, in imported inflation, an issue that was regarded as particularly problematic by countries with strong anti-inflationary motivations, notably Germany.

In relation to the second problematic feature of the system, the major problems were the lack of a reliable adjustment mechanism to manage payment imbalances as well as the persistent asymmetries in the pressures that surplus and deficit countries faced to adjust (of course, in a broader sense the issue of the reserve-issuing versus other countries). Although a few called for a return to the harsh adjustment mechanisms that had characterized the gold standard, the majority view was that the system required greater exchange rate flexibility. Although the defence of a system of flexible exchange rates had been put forward by Friedman since the 1950s (see Friedman 1953), what many analysts meant was a more frequent use of changes in exchange rate parities. The old asymmetry related to the greater room to manoeuvre that surplus countries enjoyed in their balance-of-payment adjustment processes versus the strong pressures that deficit countries faced to adjust. There were major disagreements on the use of demand policies to correct balance-of-payments imbalances, particularly in the case of deficit countries that also faced domestic unemployment problems, as well as on the possible success of stronger coordination of national policies among major economies (Fellner, Machlup, and Triffin 1966).

There was, of course, a recognition that balance-of-payments disturbances differed substantially in source and duration and called for differentiated responses. It was agreed that enduring imbalances should give way to prompt corrective action, including changes in exchange rate parities, but temporary imbalances should be financed with reserves and balance-of-payments support by deficit countries, and by reserve accumulation or debt reduction by surplus countries. However, the prompt identification of how permanent or temporary the imbalance was in a particular country at a specific time, and therefore what type of adjustment was required, was difficult and sometimes impossible (Machlup and Markiel 1964: ch. 5). In terms of financial support to deficit countries, the duration of the support beyond the one-year stand-by programmes was a critical issue in the debate, as was the question of whether programmes should take into account if the source of the imbalance was an external shock rather than excess domestic demand. These issues were particularly critical for developing countries, in particular after the collapse of commodity prices in the mid-1950s.

An underlying factor behind both of these problems was the greater capital mobility that developed countries faced with the reconstruction of global capital markets, particularly in the form of the London-centred Eurodollar market that started to evolve in the late 1950s and the expansion and creation of offshore financial centres.10 Although capital mobility was still moderate relative to the gold standard years and to what would be the typical pattern later on (Obstfeld and Taylor 2004: part two), it became a growing threat to the par value system (Eichengreen 2008: ch. 4; James 1996: ch. 7; Obstfeld and Taylor 2017). A major reflection of that was the much larger magnitudes of intervention in foreign exchange markets that major developed countries had to do to maintain their exchange rate parities, as reflected in the significant increase in United Kingdom and German central bank interventions from the mid, but particularly the late, 1960s—a ten- to twenty-fold increase relative to the earlier post-Second World War period (Williamson 1977: 47–51). Speculating against an exchange rate peg could be quite a profitable one-way bet when exchange rate misalignment was evident to market agents. Here the asymmetries faced by surplus versus deficit countries were also evident, and in both cases troublesome. Surplus countries would face capital inflows that would expand the domestic money supply and generate inflationary pressures. Deficit countries would face stronger pressure on their reserves and might be forced to adopt contractionary macroeconomic policies, and would otherwise require much greater financing than under the Bretton Woods principle that the focus of international support should be on financing current account deficits.

There was a significant diversity of proposed solutions to these problems. Common features were the focus of authorities on the first set of problems and, therefore, on how better to generate international liquidity and the rather limited attention given to the adjustment problem. The latter was in open contrast to the academic consensus on the need for more exchange rate flexibility. Since the devaluation of sterling in November 1967, the uncertainties of France after the 1968 political turmoil, and the surpluses of Germany and Japan, authorities became interested in exchange rate flexibility, but more in internal discussions than in public debates.

An additional feature of the international debate was the forum for major policy debates. With the transformation in 1961 of the Organization for European Economic Co-operation (OEEC), created to manage the Marshall Plan, into the Organization for Economic Co-operation and Development (OECD), its Economic Policy Committee (Working Party 3), made up of the ten largest countries (which thus came to be called the ‘Group of Ten’, G10), became the main policy forum for major developed countries. This generated resistance not only from the IMF and developing countries, which saw themselves excluded from the debates, but also from developed countries that were excluded from the G10 (notably Australia). This confrontation between a club of rich countries (later also the G5 and the G7) and a more representative treaty-based organization has become a central issue of the institutional arrangements for global monetary debates since then. It was mitigated in the 1960s and into the 1970s by the relative isolation that the United States faced within the G10, and therefore its willingness to bring the issues to the IMF—interestingly, to look for the alliance of developing countries. There was a diversity of views among European countries, as well as Australia, Canada, and Japan, which enriched the debate.

Proposed solutions included quite a diverse menu: (i) a return to some form of the gold standard (a position that de Gaulle favoured at one point), which would include an initial increase in the price of gold; (ii) the evolution into a fully flexible exchange rate system, in which no reserves would be required; (iii) the centralization of international reserves, making the IMF an effective world central bank; and (iv) even the evolution into a full dollar standard, in which the United States would share its monetary autonomy. However, the solution that became central in the discussion among authorities was the creation of a new global reserve currency.11

This proposal came in different versions. The essential proposal by the French was to create a collective reserve unit (CRU) tied to gold and outside the IMF. In turn, the suggestion of the United Kingdom in 1962 was the creation of a mutual currency account into which surplus countries could deposit their accumulating reserves in exchange for another asset—the predecessor of the substitution account that played a central role in the debates of the 1970s. This proposal served as a catalyst for the United States to propose in 1963 a discussion of international monetary reform in the G10. A major contribution to this debate was the 1965 report to the G10 by the Study Group on the Creation of Reserve Assets chaired by Rinaldo Ossola, then vice-chairman and later chairman of the Bank of Italy (see G10 1965). This report discussed several alternatives: (i) a reserve asset outside the Fund (the CRU proposal in some way), (ii) the creation of new assets or drawing rights in the IMF, and (iii) reserve assets in the Fund in exchange for countries’ holding of currencies (the United Kingdom’s mutual currency account proposal). There was no support for the first proposal in Europe outside France, and the United States favoured the second alternative. The group discussed the links between the new reserve asset and gold, the possible role of the IMF in reserve creation, and membership and rules of decision-making for reserve creation. A significant difference between France’s CRU proposal and the other proposals for creating an international reserve asset was that the former was seen as a substitute for dollars, whereas others were looking for a complement or a substitute for gold; both proposals aimed at controlling the expansion of the dollar as a reserve currency.

Reflecting US interest in the support from developing countries, the discussion was brought into the IMF, where negotiations took place in 1966–7, in parallel to those that continued in the G10. The IMF discussions dealt with whether it would be a drawing right or an asset, with the initial agreement leading to the concept of a ‘reserve drawing right’, but the French proposed the word ‘special’ to satisfy its view that it was a drawing right like any other that Fund members already had under the Articles of Agreement. The final agreement, according to the US negotiator, Robert Solomon, was that it was ‘a drawing right in name and a unit in substance’, as it was directly transferable (Solomon 1982: 142). Indeed, thanks to its transferability, it was a true reserve asset, but IMF members would have to make payments to the Fund on the net use of Special Drawing Rights (SDRs). It was also agreed that they could be used for payments to the IMF but not for new quota subscriptions and that a selected set of international organizations could also hold them. A group of experts convened by the United Nations Conference on Trade and Development (UNCTAD) also proposed the creation of a ‘development link’ in the associated allocations (UNCTAD 1965). This was not included in the agreement as allocations were made proportional to quotas, a matter discussed again as a major issue in the 1970s.

The creation of SDRs was agreed in the IMF meetings in Rio de Janeiro in September 1967, with the Board being given the task of preparing the draft changes in the Articles of Agreement. They were ready in mid-April 1968 and by July 1969 had received enough votes for ratification. Although opting out of this arrangement had been agreed as a possibility, basically to accommodate France, this country finally became a part of the agreement. The legacy of these debates on membership, however, led the division of the IMF accounts into ‘general resources’ and the ‘SDR account’ which limited the use of SDR allocations by countries and made it impossible to use them to finance IMF lending. Activation was agreed thanks to US surpluses in 1968–9 (which would soon disappear): 9.5 billion in three years, the first allocation being made in January 1970.

As indicated, the issue of more exchange rate flexibility remained outside the open policy debate, but there were advances in other areas. Notably, the Compensatory Financing Facility was created in February 1963 for members with shortfalls in export earnings that were temporary and out of countries’ control; it had low conditionality as it did not require a stabilization programme. It was not limited to developing countries but it was essentially aimed at their call for a financing mechanism to manage crises that originated in external shocks. This was complemented by the agreement in September 1968 that programmes not going beyond the first credit tranche would not contain performance criteria and by the creation of the Buffer Stock Facility in 1969 to help finance the contributions to commodity stabilization funds by countries that participated in such arrangements. Also, to increase the resources available to the Fund, the first increase in quotas was approved in 1959, followed in 1962 by the General Arrangements to Borrow, which provided additional financing from G10 countries in times of excess demand for Fund resources. New quota increases were agreed in 1965 and 1970. These increases kept IMF quotas around 1 per cent of world GDP after subscriptions were made, but they started to lag, in particular relative to world trade (Figure 1.4).

Figure 1.4

IMF quotas of % of world GDP and exports

Source: Author’s estimates based on IMF database.

Note: Exports of goods according to UNCTAD and world GDP according to the World Bank. For 2016 the new quota is compared to world GDP and exports in 2015.

After the collapse of the gold pool, the United States pressed other countries not to convert their dollar reserves into gold, implying that the world had moved into a ‘reluctant dollar standard’—to use again John Williamson’s (1977) term. This system faced a myriad of problems due to US balance-of-payments difficulties in 1970 and 1971, the major effect of which was the explosion of dollar reserves in the hands of other countries. World non-gold reserves, which had been growing at a moderate pace until 1969 (5.7 per cent a year in 1957–69, as we have seen), exploded with a 41.9 per cent growth in 1970 and 55.1 per cent growth in 1971 (see Figure 1.3). This led Germany and the Netherlands to float and Austria and Switzerland to revalue in May 1971. In turn, Canada had returned to a system of floating in 1970. Despite convergence in interest rates between the United States and Europe, speculative movements accelerated in June and July 1971, and major foreign central banks had to purchase record levels of dollars on August 12 and 13 (Solomon 1982: ch. 11). The dollar crisis led to the decisions taken by the US government at Camp David, and announced on 15 August 1971, to temporarily suspend convertibility of dollars into gold for other central banks—a decision that would prove to be permanent—as well as to impose a 10 per cent import surcharge and, on the domestic front, to decree a ninety-day wage and price freeze.

1.4 The Collapse of the System, the Failure to Reform, and the Resulting Non-System

The decisions of 15 August 1971 were the beginning of the collapse of the Bretton Woods arrangements. Their announcement was followed by turmoil in foreign exchange markets that led to the reluctant acceptance by the major developed countries of the need to move to a system of flexible exchange rates in March 1973, and to recurrent episodes of crisis during the rest of the 1970s and the first half of the 1980s. This was part of a broader set of events that shocked the global economy during the 1970s and early 1980s. The collapse of the Bretton Woods system was followed by the oil embargo of October 1973 and quadrupling of oil prices in December 1973, which generated global imbalances of a magnitude that had not been known since the dollar shortage period but which at the time had been repressed by current and capital account inconvertibility. The oil shock fuelled the inflationary pressures that had been evident since 1972 and, at the same time, led to the worst downturn of economic activity in the post-Second World War period. The growth recovery was insufficient and inflation was not entirely brought under control, and so the first oil shock marked the end of the post-war golden age of rapid economic growth, particularly for Western Europe, and the beginning of a period of slower world growth with higher rates of inflation, a mix that came to be known as ‘stagflation’. Individual developed economies faced great difficulties, particularly the United Kingdom and Italy. Developing countries also slowed down but their growth remained faster (see Figure 1.8). However, some regions and countries faced unprecedented external deficits financed by the recycling of petro-dollars, building up major debt burdens for the first time since the 1920s. New shocks came in the early 1980s, particularly massive global imbalances and a new dollar crisis, as well as a major debt crisis in the developing world, particularly in Latin America; these developments are reviewed in Section 1.5.

Figure 1.8

World economic growth

Source: United Nations; estimates at market exchange rates.

Abandonment of the convertibility of dollars into gold led to two parallel processes to reform the international monetary system.12 The first was the attempt to reconstruct a system of exchange rate parities among major currencies. The second was comprehensive negotiations to design a new international monetary system. The first took place in the context of the G10 and the second in the context of the IMF, in the Committee of Twenty (C20) that was created for that purpose. Both failed, and the result was a de facto transition to what can be adequately characterized as an ‘ad hoc non-system’ (Williamson 1977: xiii).

After the Camp David decisions, most developed countries floated (Japan with a lag) and France adopted a two-tier exchange rate system. Currency appreciation was limited by strong interventions in foreign exchange markets and capital controls. The G10 meetings that followed made evident the conflicts between the desire of the United States to guarantee a large current account improvement and the unwillingness of other countries to appreciate their currencies. The Smithsonian Agreement of 18–19 December 1971 generated a 7.9 per cent effective devaluation of the dollar, achieved with a mix of strong appreciation of Germany, Japan, the Netherlands, Belgium, and Switzerland, a small one for Italy and Sweden, no change in parities for the United Kingdom and France, and continued flotation of Canada (Williamson 1977: 60). Margins on both sides of the new parities were also increased from 1 to 2.25 per cent. Unrealistically, given that gold had already been priced in the market above the old parity for several years, and further increased as a result of the ongoing turmoil, the United States continued to hold that the official gold price would not be changed.

Speculation returned shortly after the agreement. As a result, in February 1972, Germany introduced a 40 per cent deposit requirement on borrowing from abroad, and the next month there were heavy purchases by central banks. In March 1972, the major European countries decided that they would try to maintain their exchange rate within a band of 2.25 per cent of each other—a system that came to be known as the ‘snake’. However, pressures continued to build up, and in mid-1972 the British pound was allowed to float, Denmark withdrew from the snake, Italy threatened to do so, and Germany adopted even stricter capital controls. The market stabilized after the US Federal Reserve reactivated in July 1972 the swap network that had been suspended on 15 August 1971.

Crisis conditions returned in early 1973. In January, the United States ended wage and price controls, Italy adopted a two-tier foreign exchange market, and Switzerland floated. In February, Germany adopted stronger capital controls and the United States devalued the dollar vis-à-vis gold by 10 per cent (to US$42.22) and announced a willingness to eliminate capital controls by December 1974. Gold was soon priced in the market about double the new official price. In early March, European central banks closed their interventions and met to discuss a joint float, which they soon adopted, and the United States and Japan agreed to do so on 16 March 1973.

The move to floating in March 1973 was widely regarded as temporary, as the C20 was discussing a return to the par value system, but the first oil shock made the return to parities unrealistic. Several short exchange-rate cycles followed, involving an initial depreciation of the dollar and appreciation of the West German mark and some other currencies (in some cases, including an appreciation of the mark within the snake), large interventions in foreign exchange markets (some of them coordinated among central banks), and complaints of chaos at critical moments. As discussed later in this chapter, floating was finally accepted as a de facto reality in 1976, after the failure of negotiations in the C20.

The C20 negotiations started with quite an ambitious agenda but failed on the two central issues that had been widely discussed in the 1960s: (i) the design of a new global reserve system and (ii) the design of proper adjustment mechanisms to correct payment imbalances. On some issues, there was broad consensus among members. On the reserve system, there was the widely shared aspiration to place the SDRs as the principal reserve asset and to reduce the role of gold—although with conflicting views on the latter issue. There was also agreement on the need to control global liquidity, but significantly different views on how to do so. With the move to floating among major currencies in March 1973, there was also agreement on new rules of valuation for the SDRs, which were agreed on 1 July 1974 on the basis of the ‘standard basket’ of currencies. A further agreement, as already indicated, was on the desirability of returning to the par value system.

In contrast, there were major disagreements on how to correct the asymmetries of the global reserve system, along the lines that had been on the table since the 1960s and mirrored the diverging interests of the main reserve-issuing country versus those of other major economies. The main US proposal was a reserve indicator system under which countries would be forced to reduce their reserves when they crossed a certain threshold. In particular, countries would not be able to convert their foreign exchange reserves into primary reserve assets (SDRs) if reserves exceeded primary asset holding limits (PAHLs). This was meant to be a presumptive rather than automatic signal, which could thus be overridden by comprehensive assessments of conditions of individual countries. If countries crossed the thresholds, they should have to take action and, if not, be subject to graduated pressures, which according to the discussion could take the form of a loss of SDR allocations, negative interest rates, ineligibility to borrow, or discriminatory exchange controls. As several researchers have indicated, there was a striking resemblance between this proposal and that of Keynes before the Bretton Woods negotiations. Preference for reserves rather than current account indicators reflected the view that the rules should allow for freer capital movements. The United States was also of the view that holding of foreign exchange should not be banned and, therefore, that there should not be strong rules on reserve composition. It also indicated that it was willing to restore dollar convertibility once the US position was strong enough.

The main objective of the United States, as in the 1960s, was to achieve greater symmetry between deficit and surplus countries, whereas for Europeans the objective was to achieve symmetry between reserve-issuing and other countries. Their main proposal was to place asset settlement at the centre of the system—that is, convertibility of foreign exchange into primary reserve assets, which again was meant to be SDRs rather than gold. Of course, this was not an immediate demand on the United States, as they were conscious that there was a dollar overhang. The creation of a substitution account in the IMF would also be a useful instrument to manage existing or future overhangs of foreign exchange, an idea that the United States shared. However, Europeans rejected the indicator system, and particularly the PAHLs, because, in their view, it would enable the United States to escape adjustment when it ran deficits. Other problems with a reserve indicator related to the width of the reserve band (if too narrow, it would curtail the stabilizing role that reserves played), the possible speculation it could trigger, and the need to define it on net rather than gross reserves. In turn, the United States opposed asset settlement, basically because it claimed that it would deprive the system of elasticity. Europeans argued that elasticity would be better provided by lines of credit.

There were variants to these views. Italy proposed the use of cyclically adjusted basic balances rather than the level of reserves as the key indicator, as well as using the dollars in the substitution account as a buffer stock. Germany placed the control over global liquidity at the centre of any reform, which meant asset settlement but also restricting freedom of asset holders to switch into non-traditional reserve assets: this required controlling the potential multiple credit expansion generated by the Eurocurrency market, which meant restrictions on portfolio choice. France was, as usual, the major country supporting a role for gold but agreed on controls on surplus countries, and initiated the compromise proposal of a reserve ceiling beyond which countries would have to deposit a specific amount in the Fund, where it would carry a negative interest rate. Specific support by the United Kingdom to the substitution account was associated with the fact that it would have allowed it to (finally) dispose of its reserve-issuing role.

Developing countries were strong supporters of a development link in SDR allocations, which meant that they would be allocated a greater proportion of those rights than their share in quotas, with least developed countries given an even more generous allocation. As we have seen, this idea went back to proposals by a group of experts convened by UNCTAD in the 1960s. The United States opposed the link as it thought that it would undermine the confidence in the SDRs, and Germany opposed it because of potential inflationary implications if they were allocated to countries likely to spend them; other European countries were willing to concede the link. Developing countries also pressed for a broader agenda of access to resources and a greater role of development issues in the IMF. They succeed in the creation of the Extended Fund Facility and the Development Committee. They favoured freedom in reserve composition—a view that they shared with the United States—and, therefore, opposed restrictions on placement of reserves in the Eurocurrency markets.

Among other countries, Japan played a minor role in the negotiations, and was a strong opponent of exchange rate flexibility and the reserve indicator. Canada was, as it had always been, a strong supporter of floating, and Australia of freedom to act. Given its role as a major gold producer, South Africa was a backer of a role for gold. Oil exporters called for reserve assets with attractive yields and the United States was conscious of the need to exempt them from the indicator system.

Success, therefore, was modest. It included the creation of a permanent IMF Council and the Development Committee, the adoption of the Extended Fund Facility under which developing countries could receive longer-term finance, and approval of the new valuation rules for the SDRs. By a paradox of history, only an Interim Committee rather than the Council was put in place in 1974. It only became a permanent arrangement, the International Monetary and Financial Committee, in 1999. However, there was no agreement on adoption of the SDR standard, correcting the asymmetries of the system, introducing the substitution account, or adopting a viable adjustment mechanism. Meanwhile, the collapse of the Smithsonian Agreement and the first oil shock served as a justification to abandon the quest for comprehensive reform.

Major discussion continued after the failure to reform. As the C20 was finishing its activities, the Executive Board agreed in June 1974 on some guidelines on floating, conceived as an interim measure (IMF 1974: 51, 112–16). Guidelines 1 and 2 indicated that countries should lean against the wind (defensive intervention). If they wanted to intervene other than defensively, guideline 3A indicated that they have to agree on a target zone with the IMF, and guideline 3B that the IMF could encourage countries to move in a certain direction. The other three guidelines concerned the application of a type of reserve indicator (guideline 4), no use of current account controls (guideline 5), and the need to take into account the interest of other countries (guideline 6).

The United States soon moved from the restoration of adjustable pegs to advocacy for floating. France continued to be concerned by the disorder introduced by floating, but finally accepted limited flexibility. A crucial step forward was the US–French agreement in November 1974 to reform the Articles of Agreement to promote a ‘stable system of exchange rates’. The annual meetings in Jamaica in 1976 served to amend the IMF articles to legalize this agreement and, more generally, existing practices. The new Article IV allowed freedom to choose an exchange rate regime within the commitment ‘to assure orderly exchange arrangements and to promote a stable system of exchange rates’ (Article IV, Section 1), which implied that countries should ‘avoid manipulating exchange rates or the international monetary system in order to prevent effective balance-of-payments adjustment or to gain an unfair competitive advantage over other members’ (Article IV, Section 1, iii), and are subject to IMF surveillance over their exchange rate arrangements (Article IV, Section 3). Beyond these very general guidelines, the precise meaning of ‘manipulating’ the exchange rate was never clearly defined, much as had been true of the concept of ‘fundamental disequilibrium’, but in this case it became a problematic issue. As a concession to France, but also as an indication of the strong reluctance to abandon the par value system that had been evident in the early 1970s, the new Article IV also maintained the possibility of returning to this system if approved by an 85 per cent vote—which, of course, gave veto power to the United States.

The system that evolved had, in fact, diverse arrangements. It included flexible rates among major currencies, but also a type of ‘regional Bretton Woods’ among European countries, which reflected the strong preference for more stable exchange rates among them to deepen intra-regional trade. For developing countries, and particularly for middle-income countries, there was much less of a sharp change relative to the past, as they had been using other forms of flexibility, including the crawling peg and managed floats (Reinhart and Rogoff 2004; see also Chapter 3 in this volume).

In the case of Europe, the snake was followed by the approval of the European Monetary System (EMS) in March 1979, which maintained limited exchange rate flexibility among its members and created the European Currency Unit (ECU) and credit facilities. Countries transferred 20 per cent of their gold and dollar reserves to the European Monetary Cooperation Fund (through renewable swap arrangements) and received ECUs in exchange. This followed the earlier goal, set in 1969–70, of forming a monetary union by 1980, which would only materialize in 1999 among a majority of members of the European Union after a crisis of the EMS in the early 1990s (see Section 1.5).

Gold also continued to be a subject of debate, on the basis of the consensus that its role would be reduced. Given the high concentration of gold reserves in the hands of developed countries, developing countries were particularly opposed to giving back a major role to gold, and argued that official gold revaluation would generate an arbitrary distribution of new liquidity and delay SDR allocations. In August 1975, consensus was reached on enhancing the role of SDRs, abolishing the official gold price and its monetary use in the IMF (particularly the obligation to use it in payments between the Fund and members), giving back to countries a sixth of the gold reserves and selling an additional sixth of Fund gold to constitute a trust fund for the benefit of developing countries, but allowing inter-central bank transactions at a mutually accepted price after a two-year transition.

The word ‘demonetization’ started to be used in 1975 in relation to gold. By then, the metal had effectively become a speculative asset, as the volatility of real gold prices since indicates, with peaks associated with major disturbances in foreign exchange markets—1980 and the aftermath of the 2007–9 North Atlantic financial crisis (Figure 1.5). In fact, after falling in the late 1960s, the quantity of gold reserves (in troy ounces) showed a very small downward trend through the 1970s and 1980s, and fell more sharply in the 1990s and 2000s, reflecting in particular the reduction of European gold reserves (see Figure 1.2). It is only in the aftermath of the North Atlantic crisis that there has been a small revival of gold reserves (see Section 1.6). Furthermore, as a result of the explosion of global liquidity that took place during the breakdown of the system and through the rest of the 1970s (see Figure 1.3), the share of gold in global reserves, calculated at the price of SDR 35 per troy ounce, fell to a small proportion by the end of the decade (Figure 1.6a). If estimated at market prices, it remained high, though volatile, in the 1970s, before a sharp decline in its share in global reserves in the 1980s and 1990s (Figure 1.6b).

Figure 1.5

Gold price (1980 dollars per troy oz)

Source: Gold price according to UNCTAD since 1970. Before according to official US prices.

Note: Series deflated by the Manufacturing Unit Value (MUV) in world trade as estimated by Erten and Ocampo (2013a).

The explosion of liquidity killed the expectation of ‘making the special drawing right the principle reserve asset in the international monetary system’, as it still reads in the Articles of Agreement (Article VIII, Section 7; Article XXII). After the initial allocation of SDRs, the managing director of the IMF recommended new allocations in 1978, at a time of severe pressure on the dollar. These allocations were made in 1979–81. However, at their peak, in 1972, they represented 8.4 per cent of non-gold reserves and 6.1 per cent of total reserves (with gold estimated at US$35 per oz.; 5.1 per cent if estimated at market price), and reached a somewhat lower new peak after the second allocation, of 6.5 per cent of non-gold reserves and 5.8 per cent of total reserves in 1981–2 (3.0 per cent in 1981 with gold estimated at market price) and one-thirteenth of foreign exchange reserves, before becoming a marginal reserve asset in the following decades. An important decision, however, was the elimination of the reconstitution requirement in April 1981, which allowed a more active use of SDRs by countries, a move that benefited developing countries.

Lending facilities were also improved, deepening the trend that had been evident since the 1960s to generate larger and multi-year facilities, and with low conditionality when it involved managing external shocks. This required additional funding through new quota increases in 1976 and 1978, the now old General Arrangements to Borrow, and a credit agreement with Saudi Arabia. Following the IMF’s view that the oil shock should be managed with financing rather than adjustment to avoid a major depression, an oil facility was created in June 1974 and extended in 1975, finally expiring in May 1976. It also had low conditionality—though slightly greater than the Compensatory Financing Facility—and counted with a subsidy account to finance developing countries. As already mentioned, the Extended Fund Facility was approved in 1974 to avoid recurrent one-year programmes: countries could borrow for up to three years, and repayment took place over four to eight years. The Fund also liberalized the Compensatory Financing Facility at the end of 1975, leading to a large increase in drawings, and allowed for a temporary enlargement of credit tranches in January 1977 from 25 to 37.5 per cent of quotas each. The Supplementary Financing Facility, which allowed lending substantially above quota, was also created in 1977 and became operational in February 1979. Finally, the repayment period for the Extended Fund Facility was extended for up to ten years in December 1979, and in fact this Facility was only used extensively starting in that year. All of these initiatives led to a peak in IMF lending in 1976–7 that, as a proportion of world GDP, surpassed the previous peak reached in the second half of the 1960s (see Figure 1.1). This included a large number of developed countries but also a few high-income ones—the United Kingdom, Italy, and Spain—that absorbed a significant share of IMF financing at its peak. In the case of the United Kingdom, the crisis faced during these years brought to an end the long transition away from its role as a major reserve-issuing country. In any case, most of the financing came, not so much from official sources (including the IMF) but from recycling of oil surpluses through the private financial system at low or even negative real interest rates (de Vries 1987: ch. 6).

The ad hoc or de facto system that evolved, therefore, had the following features:

•

A global reserve system essentially based on an inconvertible (fiduciary) dollar—a ‘fiduciary dollar standard’, as I will call it in this volume—but open in principle to competitive reserve currencies. This was complemented by sporadic issues of SDRs, which nonetheless came to play a secondary role after a good start, despite the formal commitment to make the SDRs the principal reserve asset of the system. Gold was largely demonetized but kept a role as a speculative reserve asset.

•

Freedom for each country to choose its exchange rate regime, as long as they avoided ‘manipulating’ their exchange rates, a term that has never been clearly defined.

•

Persistent commitment to effective convertibility of the current account, which developing countries finally came to abide by, but continued freedom to regulate capital flows—although, as we shall see, with growing capital account liberalization.

•

Step-by-step increase in the size of official balance-of-payments support, and design of multi-year programmes. Larger financing was particularly important to manage volatile capital flows. Low conditionality to manage external shocks continued to be a characteristic of the system, but this feature soon disappeared. The additional resources needed to finance larger programmes came from a mix of quota increases and borrowing arrangements.

•

Stronger surveillance, which has been largely ineffective vis-à-vis developed countries, and limited macroeconomic policy coordination, which essentially took place outside the IMF.

In the words of Williamson: ‘What emerged after the C20 cannot be described as an international monetary “system”, in so far as the word system implies a well-defined set of rights and obligations. Countries are free to do in large measure as they please’ (1977: 74–5). It is true, as Solomon (1982: 363) claimed, criticizing Williamson’s view, and as it had been argued by the Bellagio Group before, that Bretton Woods was hardly a fully coherent system, as in particular it lacked criteria to govern changes in par values, and had no systematic means for increasing reserves. However, the ad hoc system that came about as a result of the pressure of market forces and the failure to reform was even more distant from any coherent design. Furthermore, what emerged did not have the features of any of the alternatives discussed in the 1960s or the SDR-based system that the C20 aimed at. Rather, the fiduciary dollar standard that emerged had more in common with the ‘reluctant dollar standard’ that had evolved at the end of the 1960s. Among its major problems, and in the light of the objectives set in the initiatives to reform the system in the first half of the 1970s, we could underscore the persistence of the major asymmetries between surplus and deficit countries and between the main reserve-issuing and other countries, the maldistribution of seigniorage and lack of control over the volume of international liquidity, and the high level of exchange rate volatility, particularly at critical times.

In any case, although the failure to reform did not lead to monetary collapse, strong disturbances became common. The next wave of problems came as a result of the 1979 Iranian revolution and the second oil shock, but particularly as a result of the decision of the United States to place inflation control at the top of its policy agenda. The decisions taken by the US Federal Reserve policy after Paul Volcker was named its chairman in 1979 made evident how the new arrangements had made the world come to depend even more on the monetary policy decisions of the major reserve-issuing country.

1.5 Maturing of the New Arrangements

The major shocks and the failure to reform in the 1970s were only the beginning of other major transformations of the international monetary system, some of them strongly related to changes in the international financial system. Three closely interrelated trends were particularly important. The first was the major current account imbalances of the major reserve-issuing country, as part of the larger global payment imbalances characteristic of the fiduciary dollar standard relative to the Bretton Woods years. This also implied that the net investment position of the United States deteriorated, turning negative from the late 1980s. The second was increased capital mobility, enhanced by the move towards flexible exchange rates but also by the gradual liberalization of the capital accounts. One of its major effects has been persistent exchange rate volatility. The third was the series of major crises in emerging and developing countries that was closely associated with strongly pro-cyclical external financing and was even more difficult to manage due to the lack of an adequate financial safety net and an appropriate sovereign debt restructuring mechanism.

International cooperation to manage these transformations was highly imperfect. In institutional terms, there was an accentuation of the tendency of developed countries to manage their coordination efforts through groupings of countries with limited membership—now not so much the G10 but the G7 and its predecessor, the G5. This tendency, plus the fact that industrial countries largely ceased to use IMF resources after the late 1970s, implied that IMF activities focused increasingly on emerging and developing countries. The divorce between the handling of cooperation among developed countries outside the Fund and the management of the crises faced by emerging and developing countries from the late 1970s to the early 2000s within the Fund implied that, during this period, the IMF became a very controversial North–South institution.

The fairly permanent current account deficits of the United States have indeed been a striking feature of the global economy since the 1980s. Indeed, the current account of the major reserve-issuing country turned persistently negative from 1982, with only one year of surplus (1991); the trade balance deteriorated too, with persistent deficits from the mid-1970s (Figure 1.7a). Furthermore, both the magnitude of the deterioration of the US current account in the 1960s as well as the deficits it ran in 1971–2 and 1976–7, all of which had led to major controversies at the time, look quite modest relative to those experienced since the 1980s. This is part of a broader pattern of larger global payment imbalances that has characterized the world economy since the 1980s, in which the US imbalances have played a crucial role, generally matched by surpluses in Germany, Japan, China, and the oil-exporting countries at different times. The major precedent for this was the global imbalances generated by the oil shocks of 1973 and 1979, but those shocks turned out to be much more temporary than those that have followed since the 1980s (see Chapter 3).

An important effect of this trend was the transformation of the net investment position of the major reserve-issuing country from positive to negative in the late 1980s (Figure 1.7b). The sharply negative trend of the US net investment position since then has been interrupted by periods in which the current account balance improves and/or there are changes in asset valuations associated, among other factors, with exchange rate variations. A breakdown of the US net investment position (not shown here) indicates that net foreign direct investment continued to be positive (it has actually improved since the early twenty-first century) and it is, therefore, the net portfolio position that has become increasingly negative. This pattern implies that there is a sharp difference not only between current conditions in the United States and those that characterized the United Kingdom when it was at the helm of the international monetary system during the gold standard, but also between current conditions in the United States and those that prevailed during the Bretton Woods years.

There are several explanations for this pattern, all of which are associated with the fiduciary dollar standard. This standard eliminated restrictions on the accumulation of external liabilities by the United States, except of course the possible lack of confidence in the dollar that they may generate and the associated pressures on the dollar to depreciate. Indeed, these problems and the cycles in the real exchange rate of the US dollar are the manifestation of the Triffin dilemma under the new world monetary arrangements (see Chapter 2). However, the arrangements do contain a stabilizing element: the character of dollar assets as ‘safe assets’ and the United States as a ‘safe financial haven’ during periods of global turbulence. This is notably reflected in the demand for dollars as foreign exchange reserves, which has boomed in the developing world since the Latin American debt crisis of the 1980s and, particularly, the crisis in several emerging economies of the late twentieth and early twenty-first centuries, as a form of ‘self-insurance’ against global financial volatility and the lack of an adequate global financial safety net to manage it (see Chapter 2). It was also reflected in the strong role played by the United States as a safe financial haven after the North Atlantic financial crisis, despite the fact that it was one of the epicentres of the crisis, and also despite the expectations that the very large imbalances accumulated since the early 2000s would generate a run on the dollar.

Cyclical swings in these trends have also been important. Notably, all corrections of the US current account deficit have been associated with a slowdown in world economic growth: the early 1980s, the early 1990s, and the late 2000s. This implies that the correction of the US current account deficit generates a recessionary effect on the global economy. In contrast, the only case in which a global slowdown was accompanied by a rising US current account deficit was in the early 2000s, when the rising US deficit served, in particular, to compensate the adverse effects on global demand generated by the massive crisis in the emerging and developing world—a pattern that some have referred to as the role of the United States as the ‘consumer of last resort’ during those years.

The major destabilizing role of US current account deficits took place in the early 1980s, when the strongly contractionary monetary policy adopted to fight inflation since 1979 succeeded in taming inflation but generated massive international monetary problems associated with the sharp increase in US interest rates—also affected by the larger fiscal deficits generated by the policies of the Ronald Reagan administration that took over in 1980—and the strong appreciation of the dollar. The international spillover included not only a global slowdown in 1980–2, which was stronger than that of 1974–5 (see Figure 1.8), but even larger payment imbalances and peak turmoil in the foreign exchange markets of developed countries, which can be understood as a new dollar crisis. It also sparked the debt crisis that shocked several developing countries and, particularly, Latin America, generating this region’s ‘lost decade’—a term coined at the time by the United Nations Economic Commission for Latin America and the Caribbean, but which has been used later in relation to other countries and regions of the world.

The negotiation of the enlarged global imbalances took place in the context of a new ad hoc elite club, the G5 (France, West Germany, Japan, United States, and United Kingdom), who agreed to intervene in currency markets to depreciate the US dollar in relation to the Japanese yen and the West German mark. This agreement was ratified in the Plaza (Hotel) Accord of September 1985, and was followed by the Louvre Accord of February 1987 among the larger G7, which also included Canada and Italy (although the latter declined to support the final agreement) to stabilize the exchange rates after the major realignment that had taken place in previous years. Massive adjustment of the yen, which was a major element of this agreement, fed into the speculative bubble of the late 1980s in that country that ended up in a domestic financial crisis and Japan’s own lost decade.

Exchange rate volatility among major currencies peaked with these events. Even though it declined later, it continued to be an intrinsic feature of the new regime. This was, of course, closely interrelated with capital mobility, which under the flexible exchange rate system gradually reached and, according to some indicators, surpassed the levels achieved during the gold standard years (Obstfeld and Taylor 2004: part two). The two phenomena were closely interrelated, as the flexible exchange rate system fuelled speculative flows that were not necessarily stabilizing, while capital mobility eroded the effectiveness of the capital account regulations that had been widely used during the Bretton Woods years to manage volatile capital flows. The basic reason was that flexible exchange rates generated a ‘privatization of risk’, to use Eatwell and Taylor’s (2000) terminology, which induced capital flows associated with different perceptions of risk by market agents.

The shift towards liberalizing capital flows started with the United States in 1974 but then spread to the rest of the developed world in the second half of the 1970s and through the 1980s, and was essentially completed by these countries in the early 1990s. A major step in that direction was the decision by the European Union to end the liberalization of capital flows (adopt ‘capital account convertibility’) in 1990. Emerging and developing countries liberalized in a more gradual and limited way, and with significant regional differences, Latin America leading the way, although with a temporary reversal of the liberalization process during its debt crisis. This worldwide trend was reinforced by the multiplication and expansion of offshore financial centres.13 In any case, IMF rules continue to allow countries to regulate capital flows. The attempt by the managing director of the IMF, with US support (and pressure), to change the Articles of Agreement in 1997 to impose the obligation of capital account convertibility on Fund members was defeated. Major constraints on capital account regulation came with free trade agreements, notably those with the United States (Erten and Ocampo 2016; see also Chapter 4 in this book).

Capital account liberalization, as financial liberalization in general, generated its own problems, particularly the frequency of both domestic financial crises and also twin crises (joint domestic and external financial crises) that followed liberalization episodes. One of the most important international episodes in this regard was the crisis generated in the EMS by the liberalization of the capital account in 1990, which led to massive turmoil in several European countries that were forced out of the agreed exchange rate parities. More broadly, liberalization made economies subject to the boom–bust cycles typical of financial markets and to the contagion of both optimism and pessimism among countries and agents in different phases of the cycles. The major episodes in this regard have been the series of crises in emerging and developing countries generated by boom–bust cycles of external financing, notably the Latin American debt crisis of the 1980s, the smaller Mexican financial crisis of 1994, and the sequence of financial crises in several emerging countries that began in East Asia in 1997, spread to Russia in 1998, and then affected Brazil and Argentina, and had strong effects on many developing countries. The spread of the North Atlantic financial crisis also belongs to this family, and affected in particular the European periphery (Cyprus, Greece, Ireland, Iceland, Portugal, and Spain), which in a sense behaved as emerging economies had done in the past.

The major problem of emerging and developing countries is the strong pro-cyclical swings in external financing: strong booms followed by ‘sudden stops’ of external financing—to use a term en vogue since the mid-1990s.14 Emerging and developing countries are also plagued by other forms of external shocks and by inadequate financing during crises. Notably among the former are the terms of trade shocks in commodity-exporting economies and, with the opposite sign, those of increases in oil prices in energy-dependent economies. As we have seen throughout this chapter, these issues received significant attention by the Fund in the second half of the 1950s and through the 1960s, and during the oil shocks of the 1970s, and returned to centre stage during the new downswing of commodity prices from the 1980s (mid-1980s in the case of oil) to the early 2000s. On many occasions, the difficulties experienced during crises are associated with the tendency to spend and even overspend the terms of trade and financing booms that precede them.

The major response of the international community was increased financing. This included new credit facilities, including the active use of those facilities that been created since the mid-1970s. Furthermore, in the face of the crisis of several emerging economies that started in East Asia, the IMF created the Supplemental Reserve Facility in December 1997, which served as the basic framework for the largest loans made to emerging and developing countries during this crisis. This facility came with short maturities, which were later extended, and penalty interest rates. For the poorest countries, the structural adjustment lines were created in the mid-1980s and transformed in 1999 into the Poverty Reduction and Growth Facility. IMF lending reached its historical peak relative to world GDP in the mid-1980s, and a slightly lower peak in the late 1990s and early 2000s, as the result of major demands by emerging and developing countries (see Figure 1.1). In the intermediate period, demand for Fund resources remained at relatively high levels by historical standards owing to demands from Central and Eastern European countries of the Soviet bloc that joined the IMF after the fall of the Berlin Wall.

Quota increases agreed in 1983, 1990, and 1998 allowed this basic source of IMF financing to periodically catch up with the growth in world GDP, although at slightly lower ratios than had been typical up to the 1960s (see Figure 1.4). The share of quotas in world trade experienced a stronger reduction, particularly in the 1970s. However, the possibility of individual countries borrowing above quota, which became a standard feature with the changes in lending practices after the 1970s, allowed IMF lending to increase relative to world GDP even when overall quotas decreased as a proportion of that aggregate (Figure 1.1). Quotas were also complemented by borrowing mechanisms: the old General Arrangements to Borrow, which was expanded in 1983 to respond to the Latin American debt crisis, and the New Arrangements to Borrow with a larger group of countries (now thirty-eight), which became effective in 1998 and was a response to the massive crisis of the emerging economies.

The rise in lending to emerging and developing countries from the 1980s to the early 2000s came, however, with a major cost: increased conditionality. In contrast to the trends of the 1960s and 1970s, when conditionality had been eased as countries faced shocks of an external origin, conditionality was ratcheted up, now going clearly beyond the macroeconomic focus that was its typical feature into ‘structural conditionality’—the name market reforms were given, with the World Bank as the leader. This process came in three phases, through which it became increasingly more intense: the Latin American debt crisis, the transition to capitalism of the former Soviet bloc, and the crisis of the emerging economies of the late twentieth and early twenty-first centuries (see Chapter 5). Even the oldest low-conditionality credit line, the Compensatory Financing Facility, languished under excessive conditionality and ceased to be used after the turn of the century.

Although lending to industrial countries came under tougher conditions in the late 1970s, no doubt the fact that the IMF concentrated its support on the emerging and developing countries served to consolidate conditionality. This trend was further reinforced by the victory of the market-reform agenda after the electoral successes of Margaret Thatcher and Ronald Reagan in 1979–80 in the United Kingdom and the United States, respectively. However, it generated the perception of the IMF as a controversial North–South institution. This problem reached its peak during the crisis of the late twentieth century, when it was perceived that some powerful countries were using IMF lending to push for reforms in emerging economies (e.g. pressure on East Asian countries to open their domestic financial systems to foreign investment) that they had failed to get otherwise. As a result of the strong opposition to rising conditionality from the emerging and developing countries, new guidelines on conditionality were approved in 2002, underscoring the principle of ownership of policies by countries, the fact that structural conditions must be ‘macro-relevant’ and must be core competencies of the Fund (monetary, fiscal, and exchange rate policies, as well as financial system issues), and that conditions must be critical to achieve the programme goals (IMF 2002b). Despite the fact that conditionality was reduced, the pro-cyclical effects of some of its programmes were not eliminated, and it left a clear stigma to borrowing from the IMF that has not been entirely overcome.

The controversies surrounding IMF interventions in East Asia were the background for the Japanese proposal to create an Asian Monetary Fund. This idea was strongly opposed by the United States, but led to the 2000 Chiang Mai Initiative, which created a system of bilateral swap arrangements among the central banks of the member countries of the Association of Southeast Asian Nations (ASEAN), China, Japan, and the Republic of Korea (ASEAN+3), engulfing the former ASEAN swap arrangement. At the global level, and following Canadian proposals, the Financial Stability Forum and the Group of Twenty (G20) were created, including in the latter case major emerging economies in the follow-up to global macroeconomic and financial developments.

An additional problem was the lack of a framework to handle sovereign debt crises. This issue became particularly problematic during the Latin American debt crisis of the 1980s, when the region was unable, for the first time in history, to use default to handle the difficult conditions generated by over-indebtedness mixed with a collapse of commodity prices.15 Furthermore, the debt crisis was managed so as to avoid major losses by creditors, the most important of which were US banks. So in a very significant sense Latin American countries were pressured to service the debt to avoid a US banking crisis. The fact that IMF programmes were conditional on servicing private obligations made the institutions part of the instruments to force debt payments and avoid defaults. Additional financing and better rescheduling terms came after the launch of the 1985 and 1987 Baker Plans, but a debt relief initiative came only with the 1989 Brady Plan, seven years after the outbreak of the crisis, which was in any case modest relative to previous debt relief initiatives, notably that of the 1940s and 1950s to renegotiate the debts that had been accumulated by Latin America up to the Great Depression. After the launch of the Brady Plan, the IMF changed its policy in favour of the principle of ‘lending into arrears’, which was adopted in 1989 in relation to commercial debt and extended in 1998–9 to include bonds and a less stringent interpretation of what negotiating in good faith means.

It was only in the 1990s, after the 1994 Mexican crisis, that the discussion of better mechanisms to better handle sovereign debt crises entered into the international debate. The major initial initiative came from a G10 working party, which proposed introducing collective action clauses (CACs) into debt contracts (see G10 1996). This was followed by 2001–3 negotiations of a statutory mechanism—a sovereign debt restructuring mechanism. It failed because of the opposition of the United States (which had initially unleashed the negotiations) and also of various developing countries (notably Brazil and Mexico) that feared that a mechanism of this nature would impair their access to private markets. However, this led to the initiative by Mexico to introduce CACs into a March 2003 New York bond issue, which thereafter became regular practice, as was already the case with London bond issues. This is why individual voluntary renegotiations continued to be the norm.

The response of emerging and developing countries to increased conditionality, the perception of unfairness in the management of sovereign debt crises and, more generally, the deficiencies of the international financial safety net, was self-insurance in the form of accumulation of foreign exchange reserves. After the Latin American debt crisis, and particularly after the crisis of several emerging economies of the late twentieth and early twenty-first centuries, the accumulation of foreign exchange reserves in the hands of emerging and developing countries became massive (see Chapter 2). This was facilitated by the financial boom that took place from 2003 up to the North Atlantic financial crisis. In some cases, the degree of accumulation of reserves might have been excessive, and it had costs. An interesting implication is that this response also provided demand for safe assets from developed countries, particularly the United States, reinforcing the role of the dollar in the fiduciary dollar standard. In fact, it amounted to a transfer of resources to the major reserve-issuing country, making it clear that self-insurance generates an inherent inequity in the current international monetary system.

1.6 The North Atlantic Crisis and its Aftermath

The North Atlantic financial crisis threw the developed world into the first open recession since the Second World War (Figure 1.8) and revived fears of a new Great Depression. Although that outcome was avoided, the developed countries did experience as a group a longer-term slowdown in terms of economic growth, much as they had after the first oil shock, but now with strong deflationary rather than inflationary trends—in a sense as a ‘stag-deflation’ rather than stagflation. The developing world was less affected: it also experienced a strong slowdown but it was less intense and shorter than those it had experienced during the Latin American debt crisis and the crisis of several emerging economies in the late twentieth and early twenty-first centuries. The policy response introduced some changes in the international monetary system, but it did not reach the character of a ‘Bretton Woods moment’, nor was there any ambitious reform agenda similar to that which was on the table after the breakdown of the gold–dollar parities in the early 1970s. It did, however, lead to the strengthening of macroeconomic cooperation, as well as of the IMF and other mechanisms to provide international liquidity. The most important efforts took place in the G20, now upgraded to a heads of state grouping, which designated itself in Pittsburgh in September 2009 ‘the premier forum for our international economic cooperation’ (see G20 2009a). This reproduced, in a new form, the preference of developed countries—and now, perhaps, some emerging countries—for elite clubs over treaty-based organizations.

In the first three meetings of the upgraded G20 in 2008–9, cooperation was enhanced in two critical areas of the international monetary system: agreement to undertake coordinated expansionary policies and to create new surveillance mechanisms for the macroeconomic policies of G20 members, and the provision of international liquidity. In 2010, there was also agreement on an IMF quota reform—which unfortunately took more than five years to become effective due to the lag of final approval by the US Congress. In a complementary manner, members also agreed to strengthening financial regulation and supervision through the transformation of the Financial Stability Forum into the Financial Stability Board (which now included all G20 members), increasing international tax cooperation under the leadership of OECD, and avoiding protectionist policies, which in the latter case had deepened the global crisis during the Great Depression. However, reform efforts have been absent in other areas, notably in the global reserve system and the creation of an international debt workout mechanism.

The coordinated response to the crisis, first in an informal way through the major central banks and then in a formal way through the coordinated fiscal expansion agreed by the G20, can perhaps be considered the greatest success in terms of macroeconomic cooperation in history. In relation to the role of fiscal policies, the consensus broke down very soon, in the June 2010 G20 Toronto meeting, when several countries placed priority on public sector debt sustainability. Expansionary monetary policies of central banks continued to be the rule, except temporarily for the European Central Bank in 2011 and for the very gradual dismantling of the expansionary policies of the US Federal Reserve since 2013—reflecting in the latter case the better relative performance of the US economy. A major issue was the incapacity to fight very low inflation or open deflation in major economies, and the unprecedented spread of negative interest rates in several developed countries, the implications of which (and, particularly, their partial ineffectiveness) have been the subject of a heated debate. These effects are only starting to be overcome in 2017, one decade after the beginning of the crisis. Another effect of the crisis was to spark a new boom of financing towards emerging and developing economies, the risks of which were made evident during the downswing that these economies have experienced in recent years.

The G20 also launched during this crisis its own macroeconomic peer review framework, the Mutual Assessment Process. The relevant indicators were agreed by the G20 finance ministers and central bank governors, and the IMF was given the task of assessing the submissions of individual countries, aggregating them to assess their mutual consistency, and making policy recommendations (IMF 2011e). This has been combined with a proper IMF activity, which has been the strengthening of its multilateral and bilateral surveillance. In contrast to the success of the initial expansionary policy, these mechanisms have had very limited effectiveness, as reflected in the slow growth of the developed economies and a long lag in combating deflation—i.e. in the long period of stag-deflation that took place—as well as in the persistence of significant global imbalances.

In terms of financing, in a series of decisions taken throughout 2009, the IMF approved perhaps the most ambitious reform of Fund lending in history (IMF 2009c, 2009e). First, it created the first successful contingency credit facility, the Flexible Credit Line (FCL), which was soon demanded by three countries. This followed three unsuccessful attempts to create such a facility in the early twenty-first century. This was complemented in 2010 by the creation of the Precautionary Credit Line (later called Precautionary and Liquidity Line), for countries with good policies but that do not meet the criteria for access to the FCL. Second, other credit lines were doubled in March 2009, and countries were allowed to use stand-by agreements for preventive purposes. This was accompanied by a reduction of conditionality, basically the elimination of the link between IMF disbursements and structural conditionality (the structural benchmarks), though maintaining some of the pro-cyclical features of macroeconomic conditionality, which have been particularly noticeable in the European programmes. Third, the concessional credit lines were also doubled and structured along three facilities: (i) Extended Credit Facility, which replaced the Poverty Reduction and Growth Facility, (ii) the stand-by lines, which can now be used for dealing with external shocks, and (iii) a rapid credit facility. Later, the IMF further reformed its concessional loan lines from a single design to a menu of options, which takes into account the vulnerability of countries to debt and their macroeconomic and public finance management capacity. All these reforms were accompanied by the elimination of several existing credit lines, including the Compensatory Finance Facility, which had ceased to be used owing to its increased conditionality (see Chapter 5).

This was accompanied by other forms of liquidity provision in the context of a broader global financial safety net. Notable in this regard was the massive provision of dollar liquidity to central banks of other developed countries through the US Federal Reserve swap facilities. A few emerging economies (Brazil, Republic of Korea, Mexico, and Singapore) were allowed access to this mechanism in 2008–9, but it remained essentially an instrument of cooperation among developed countries. European cooperation was enhanced by the creation of the temporary European Financial Stability Facility put in place in 2010, and the permanent European stability mechanism inaugurated in October 2012. The Chiang Mai mechanism was significantly expanded and multilateralized, and a monitoring unit to support it was put in place, but has so far not been used, possibly because of its link to IMF financing beyond a certain level. The BRICS (Brazil, Russia, India, China, and South Africa) also launched in 2014 a Contingency Reserve Arrangement and other initiatives of a smaller scale have been adopted in other parts of the world (see chapters 5 and 6).

The other important decision in terms of international liquidity provision was the largest issue of SDRs in history, also agreed to in 2009: SDR 161.2 billion, equivalent to US$250 billion. It was made together with an allocation for SDR 21.4 billion, which had been approved in 1997 but only became effective when the related changes in the Articles of Agreement were approved by the US Congress in June 2009. Together, they revived this dormant mechanism of international cooperation, temporarily increasing their share in total global reserves to 3.9 per cent in 2009 (3.3 if gold is valued at market prices), below the peaks reached after the first two allocations. However, no agreement was reached on how to enhance the role of SDRs in the international monetary system.

Interestingly, the North Atlantic nations also revived somewhat the demand for gold as a monetary reserve (see Figure 1.2). The additional demand came from several countries, in particular China, the Russian Federation, Turkey, and India, in that order. However, the continuous reduction in gold reserves by some European countries, together with the sale of such reserves by a few countries experiencing deep balance-of-payments crises (notably Libya and Venezuela), generated a very moderate global increase of world gold reserves. Furthermore, and despite a new speculative boom in gold prices (see Figure 1.5), the share of gold in total reserves actually declined, even if gold is measured at market prices (see Figure 1.6b).

In terms of governance, the G20 supported and the IMF later agreed in 2010 to the doubling of IMF quotas and a redistribution of quotas and voting power to increase the share of developing countries in decision-making. Relative to the situation before the Singapore 2006 annual IMF meeting, when this issue was first revised, the quotas of developing and transition economies were increased by 3.9 percentage points, and their voting power by 5.3 percentage points—certainly much less than these countries expected (see Chapter 6: Figure 6.2). The voting power of the poorest countries was protected by increasing the basic votes of members. This decision was significantly delayed by lack of approval by the US Congress and only became effective in 2016. It reversed the very strong negative trend in the share of IMF quotas in world GDP, which had reached its lowest historical level in 2015 (see Figure 1.4). To provide resources to the IMF while the quota increase became effective, the New Arrangements to Borrow were increased in 2011, and new bilateral lending by individual countries was designed, also involving some emerging economies.

One area in which there was only very limited advance was in the mechanisms for the resolution of sovereign debt crises. Debt renegotiations between debtors and creditors continued to be the rule, the most important being the Argentine debt renegotiations of 2005 and 2010, and that of Greece in 2012. The unsuccessful litigation by Argentina in 2013–14 concerning the interpretation by US courts of the pari passu clause in bond contracts16 prohibited the country from making payments on its restructured debts if it did not pay the holdouts in full, significantly reducing the incentives of any creditor to participate in a restructuring process. This finally forced Argentina to reach an agreement with the holdouts in 2016. On the other hand, there was broad agreement on the need to aggregate different claims by including aggregation clauses in bond contracts guaranteeing inter-creditor equity. Based on these precedents, the IMF (2014b) proposed the inclusion of aggregation clauses in debt contracts as well as a revision of the pari passu clause. Mexico led the way again, by including the new clauses in a November 2014 debt issue in New York. Aggregation was included as an element to be covered in eurozone bonds from 2013, and some other countries had included them in their bond issues since 2003.

Viewed overall, the ad hoc international monetary non-system that arose out of the crisis of the early 1970s has proved to be fairly resilient. However, and despite some advances since then, some of its major gaps continue to have negative effects on the global economy. They relate to the deficiencies of the global reserve system, the weakness of global macroeconomic cooperation, the lack of an exchange rate system, the instability generated by pro-cyclical capital flows particularly for emerging economies, and the absence of an adequate sovereign debt restructuring mechanism. Its governance structure continues to be deficient, in particular in terms of the predominance of elite clubs and the inadequate participation of emerging and developing countries in decision-making. These are the areas that must be the focus of future reforms. The remaining chapters of this volume explore in greater detail the nature of these problems and make proposals to overcome them.

Notes

1

As pointed out in the Preface, I follow in this book the use of this term, particularly by Mohan and Kapur (2014) but also by Buiter (2008), rather than that of ‘global financial crisis’ because, although the crisis had global effects, its epicentres were the United States and Western Europe.

2

This was indeed the agreed language to proposals by India, with the backing of other countries, to incorporate the ‘development’ objectives of the Fund (see Helleiner 2014: ch. 9).

3

Please note that, although the dollar is also a currency of other countries, I generally use the term throughout this book to refer to the US dollar.

4

See a fascinating account of these debates in Skidelsky (2000: part 2).

5

For the first long decade after the launch of the IMF, the widespread use of foreign exchange restrictions for current account transactions by the European countries had, nonetheless, discriminatory effects against trading with the US and in favour of intra-regional trade. The European Payments Union, to which I refer later, may have had similar effects. The persistence of current account restrictions in the developing world for a much longer period could have had similar effects.

6

There was only one important case: the objection to a French devaluation in 1948, but because it also included multiple-currency practices (Horsefield 1969: 202).

7

According to Schenk (2010), when the demand for sterling assets declined in the Indian subcontinent in the early post-war years, it was replaced by Australia, New Zealand, and the colonies of the Far East and Africa. In turn, when Australia, New Zealand, and South Africa started to diversify in favour of dollar reserves in the early 1950s, the official demand for sterling increased in Hong Kong and the Middle East. Hong Kong became the major official holder of sterling assets in the late 1960s.

8

In the formulation of the Bellagio Group, the system had three different problems: (i) the inadequate functioning of the mechanism of adjustment to correct persistent payments imbalances, (ii) the provision of adequate international liquidity, and (iii) confidence in the reserve media (Machlup and Markiel 1964: 24). We subsume the latter two into the problems of the global reserve system.

9

There were four proposals among the Bellagio Group members: (i) a semi-automatic gold standard, (ii) centralization of international reserves, (iii) moving to multiple reserve currencies, and (iv) fully flexible exchange rates.

10

The development of ‘offshore finance’ was also associated with tax elusion and evasion. The most important centre, Switzerland, had its origins in the 1920s and, after a pause during the Second World War and the early post-war period, experienced a boom from the 1950s to the 1970s, and would be joined by new offshore centres in several parts of the world, and notably so since the 1980s (Zucman 2015).

11

See, among many, a summary of these debates in Solomon (1982: ch. 4 and 8). This set of proposals matched, to a significant extent, that of the Bellagio Group members (see footnote 9).

12

See, among the many other analysts of these processes, Eichengreen (2008: ch. 5), James (1996: ch. 9–10), Solomon (1982: ch. 11–13), and Williamson (1977), whose contributions I summarize in this section.

13

See in this regard footnote 10.

14

The term was coined by Rüdiger Dornbusch in a paper on the 1994 Mexican crisis (Dornbusch and Werner 1994), in which he argued that ‘it is not speed that kills, it is the sudden stop’, but its popularization is due equally to the work of Guillermo Calvo (for whose early work on the subject, see Calvo 1998).

15

For an analysis of the handling of the Latin American debt crisis, see Devlin (1989), Bértola and Ocampo (2012: ch. 5) and Ocampo (2015a).

16

This clause has been generally interpreted as equal ranking but it was interpreted by New York courts in this case as equal ‘ratable payments’.

Which of the following was originally created to regulate fixed exchange rate and enforce the rules of the international monetary system?

Who regulates fixed exchange rates?

What is IMF and its function?

Why was the IMF created?