“gold and dollar holdings are an attribute of full national independence. This of course is all poppy-cock; but we cannot yet behave as if sterling balances carry equal glamour”
—U.K. TreasuryFootnote 1
The term “zombie bank” refers to a bank that cannot continue operating without some form of government assistance, such as bailout, credit support, or other guarantees.Footnote 2 I define a “zombie international currency” as a currency that requires the intervention of the issuing country, in the form of exchange controls and diplomatic pressure such as threats or economic sanctions, to continue being used internationally. International holders of zombie currencies are comparable to the creditors of zombie banks: they try to decrease their exposure to the zombie to avoid potential losses. This paper argues that sterling was a zombie international currency during the Bretton Woods era.Footnote 3
Some scholars have claimed that even before the outbreak of WWII, the sterling had already lost its international influence, which benefited the U.S. dollar.Footnote 4 Other scholars, such as Eichengreen, Mehl, and Chiţu (Reference Eichengreen, Mehl and Chiţu2018), Eichengreen (Reference Eichengreen2019), and Schenk (2010), have argued that after 1945, the pound sterling and the U.S. dollar were the two main international currencies of the multipolar monetary system. Eichengreen and Schenk also state that the share of global foreign exchange reserves held in sterling gradually decreased during the 1950s and 1960s as a result of the successful collective management of foreign sterling liabilities. Most of these liabilities were the result of war debts and were concentrated among the countries of the sterling area, which was a monetary zone organized around the pound sterling. The sterling area appeared in 1939, as part of the war effort, when Britain introduced restrictions on international payments around the countries that had been maintaining a fixed relationship with sterling since the 1931 devaluation. Sterling area members were characterized by having a constant exchange rate with the pound sterling and by holding their foreign exchange reserves largely in the form of sterling balances in London (for a survey of the historical literature on the sterling area, see Cohen (Reference Cohen1971), Schenk (2013), or Kennedy (Reference Kennedy2018a)). In exchange, they enjoyed limited capital controls for transactions within the area.
When postwar attempts to restore the convertibility of sterling into dollars failed, Britain formalized the wartime capital controls around the sterling area through the Exchange Control Act of 1947.Footnote 5 At that point, the sterling area covered the countries of the Commonwealth, except Canada, along with the British colonies, protectorates, and additional nations such as Egypt, Sudan, Iraq, Iceland, the Faroe Islands, and Transjordan. British authorities blocked the sterling liabilities of these countries and used exchange controls to limit capital outflows to the non-sterling world. The area disappeared formally in 1979 with the abolition of British exchange controls (de Bromhead et al. Reference de Bromhead, Jordan, Kennedy and Seddon2023).
Interpretations diverge as to why countries remained in the sterling area and kept their reserves in sterling. Shenk states that the sterling area system ”operated to support collective interests of its members in the stability of sterling and freer trade and investment flows, underpinned by carrots and sticks” (Schenk 2018, p. 6) and that members of the system pursued their “perceived national self-interest” (Singleton and Schenk Reference Singleton and Schenk2015, p. 1160; Schenk 1996). Others have argued that the British authorities managed colonial sterling holdings for the benefit of Britain rather than for that of the colonies (Narsey Reference Narsey2016; Nyamunda Reference Nyamunda2017). Hinds (Reference Hinds1991) and Krozewski (Reference Krozewski1996, Reference Krozewski1997, Reference Krozewski2001) have demonstrated that Britain pressured the Gold Coast countries and Ghana to remain in the sterling area after achieving independence, as the British wanted to prevent the liquidation of their sterling holdings, which would affect the dollar convertibility of sterling. By contrast, Schenk (1996, 2010) describes that, from the mid-1950s until the 1967 sterling devaluation, foreign sterling liabilities were simply “a niggling potential discomfort” for British monetary authorities (Schenk 1996, p. 872). Some case studies and surveys mention the use of “capital controls, moral suasion and geopolitical influence” by Britain (Eichengreen, Mehl, and Chiţu Reference Eichengreen, Mehl and Chiţu2018). However, there is a lack of analysis within the literature of the systematic coercion exercised by British authorities on sterling area countries in light of the weakness of the British economy. I provide a comprehensive analysis of the British threats and sanctions aimed at deterring the liquidation of sterling holdings by members of the sterling area.
Using both quantitative analysis and archival research of recently declassified documentation, I offer a new narrative on the decline of the pound sterling: From 1945 onward, sterling survived as an international reserve currency only in the captive market of the sterling area. European countries rapidly reduced the share of sterling within their foreign reserve portfolios after the war. To sustain the international status of sterling, the British authorities prevented the liquidation of sterling debts held by countries in the sterling area. Sterling area countries could not freely diversify their foreign exchange reserves as British authorities systematically threatened to apply commercial and exchange controls, sanctions, and freeze the assets of those who attempted to do so without approval from London. British authorities distorted the international distribution of sterling through international blackmail, propaganda, and economic sanctions to limit the decline of sterling. I demonstrate that, when a country managed to leave the sterling area, it rapidly rebalanced its reserve portfolio outside of sterling while the remaining countries faced lower returns on their reserve portfolios. Thus, I present an alternative to the carrot-and-stick metaphor: After 1945, the carrot ceased to attract countries in the sterling area, but the stick became a sledgehammer.
This challenges the view of the Bretton Woods era as a multipolar monetary world (Eichengreen, Mehl, and Chiţu Reference Eichengreen, Mehl and Chiţu2018; Eichengreen Reference Eichengreen2019). Sterling’s key currency role ended during the interwar period. After WWII, sterling could not compete with the dollar on the international stage, as its presence in global foreign exchange reserves was secured only by artificial barriers built by British monetary authorities around the sterling area. In countries free from British influence, the dollar was the sole key international reserve currency.
METHOD AND SOURCES
In this paper, I focus on the reserve role of sterling and study the decision of central banks to hold their reserves in the form of pounds sterling. This approach is standard in the historical literature on international currencies, but I am the first to apply it to the Bretton Woods era using a large sample of country-level data (Chinn and Frankel 2008; Eichengreen and Flandreau Reference Eichengreen and Flandreau2009).
During the Bretton Woods era, countries’ foreign exchange reserves primarily comprised foreign deposits and first-class government securities, similar to previous periods. Due to the sensitivity of this data, neither central banks nor national governments have published the composition of their foreign exchange reserves. They conveyed the total value of their gold and foreign currency reserves to the International Monetary Fund (IMF), which then published the data.Footnote 6 Relying on IMF data, Singleton and Schenk (Reference Singleton and Schenk2015) and Eichengreen, Mehl, and Chiţu (Reference Eichengreen, Mehl and Chiţu2018) described a progressive decline of the pound sterling at the world level but were unable to measure the individual choices of countries regarding the currency composition of their reserves.
To analyze the international use of sterling as a reserve currency, I compare the share of sterling in the reserves of European and sterling area central banks. I collected data for nine Western European countries using local central bank archives.Footnote 7 I utilized archival documents from the Bank of International Settlement (BIS), the Bank of England, and Her Majesty’s Treasury to reconstruct the historical proportion of sterling in the reserves of the majority of the countries in the sterling area.Footnote 8 My sample consists of 22 sterling-area countries whose reserves represented, on average, 73 percent in all sterling holdings of the sterling area.Footnote 9
THE USE OF STERLING AS A RESERVE CURRENCY
A New Perspective from Country-Level Data
Figure 1 displays the share of sterling within official reserves of the sterling area countries and European countries.
SHARE OF STERLING IN RESERVES OF CENTRAL BANKS (GOLD + FOREIGN EXCHANGE)
Notes: The numerator is the volume of sterling holdings, and the denominator is the sum of all gold and foreign exchange reserves. I aggregate reserve portfolios by group of countries and calculate the sterling share for each group. The line for the sterling area represents the average share of sterling for all members of the sterling area. This group is also divided among the British colonies and the independent sterling area members.
Sources: Author’s dataset, see text.
From 1952 onward, sterling accounted for less than 10 percent of the reserves of Western European countries; however, in sterling area countries, it accounted for more than 60 percent until 1967. This picture contrasts with the global decline of sterling’s relative position in the 1950s and 1960s. Sterling was the main reserve currency of the sterling area throughout these decades. In Europe, by the early 1950s, the shift away from sterling was complete, and sterling did not re-emerge as an attractive asset in later years. A comparison of sovereign nations in the sterling area and British colonies indicates that starting in the mid-1960s, countries began to diversify their reserves away from sterling. This suggests that colonies would have divested from sterling, had they been afforded the freedom to do so. Sterling area countries diversified their portfolios by accumulating new reserves and not by converting their sterling holdings into gold or other reserve currencies.Footnote 10
This picture is consistent with the fact that, according to the Bretton Woods agreements of 1944, sterling no longer played a pivotal reserve role in the international monetary system. The U.S. dollar was the key currency of the system, convertible in gold at a fixed parity, while currencies of other members guaranteed the convertibility of their currency in dollars only. This pyramidal system, with the dollar and gold at the top, left sterling with little opportunity to function as an international currency.
A Zombie International Currency
The existing literature on the theory of demand for reserve currencies reveals several potential causes for the heterogeneity of the distribution of foreign holdings of sterling (see Frankel (Reference Frankel2012) and Eichengreen, Mehl, and Chiţu (Reference Eichengreen, Mehl and Chiţu2018) for recent surveys of the literature). Some studies point to the specific characteristics of Britain, such as the credibility of its monetary policies, the size of its economy, and its financial depth. Other potential drivers include trade relations and military alliances between Britain and sterling-holding countries.
Previous macroeconomic and historical studies, such as those by Bean and Crafts (1995) and Broadberry and Crafts (1996), highlight the hardships plaguing the U.K. economy during the Bretton Woods era. While the United Kingdom boasted strong macroeconomic fundamentals in the nineteenth century, the two world wars transformed Britain from the world’s largest creditor into the world’s largest international debtor (May Reference May2013, p. 30). At the end of 1945, the sterling liabilities of the United Kingdom totaled £3.7 billion, while British gold and dollar reserves stood at only £620 million.Footnote 11 Only in 1954 were the war restrictions on transfers of sterling for current or capital purposes lifted for residents of 43 countries outside the sterling and dollar areas (International Monetary Fund 1954; Schenk Reference Schenk1994a). Convertibility was fully restored in 1961 for non-sterling area countries and in 1972 for sterling area countries.Footnote 12 Despite these heavy restrictions, the money markets had little faith in the pound sterling, forcing the Bank of England to devalue the currency twice in response to market pressure, first in 1949 and then again in 1967.
The United Kingdom did not display the economic strength expected of an international currency issuer, and by the mid-1950s, it had lost its leading role in Western European economies.Footnote 13 Between 1950 and 1970, its GDP per capita grew by an average of 7 percent, which was slower than the average growth of 11 percent experienced by most Western countries.Footnote 14 Britain was also a declining trade power during this period: Even though its exports increased during this period, its share in world trade decreased steadily from more than 10 percent in 1950 to 6.2 percent in 1970. Britain suffered from payment deficits due to trade and current account imbalances. Between 1945 and 1971, current account deficits occurred in more than half of all years, and trade deficits were recorded almost every year.Footnote 15
The postwar period was also characterized by the breakdown of the British Empire. Twenty-nine countries in the sterling area gained independence between 1945 and 1971. The 1956 Suez crisis, along with the withdrawal from Greece in 1947, brought to the forefront the dwindling military power of Britain. To defend the pound sterling after the Suez crisis, Prime Minister Harold Macmillan had to cut defense expenditures and reconsider Britain’s position as a world power (Cain and Hopkins Reference Cain and Hopkins2014, p. 677). The lasting difficulties faced by Britain, including recurring public deficits, were reflected in financial markets. This was evident in the significant increase in yields on British government bonds from 3 to 9 percent between 1950 and 1970.Footnote 16 Interest rates on British long-term government bonds were on average 40 percent higher than those of the United States for the period 1950–1971.Footnote 17
The City of London, Britain’s financial center, began facing increasing competition from New York, starting in the interwar period (Eichengreen and Flandreau Reference Eichengreen and Flandreau2012). Throughout the 1930s, the new capital issues for British colonies declined.Footnote 18 The implicit imperial guarantee that colonies would benefit from choosing to issue debt in London became less valuable after Britain’s departure from the gold standard in 1931 (Degive and Oosterlinck Reference Degive and Oosterlinck2019). After WWII, New York became the world financial center, while in London, the domestic money market was tight and tensions arose between the City and the Treasury, which imposed capital controls on international transactions (Schenk 1998; Atkin Reference Atkin2004; Davies Reference Davies2017). The emergence of the London Eurodollar market increased the significance of the U.S. dollar in international finance, but it did not restore the international prominence of the pound sterling.Footnote 19
These macroeconomic indicators pointed toward the high risks of holding sterling as a reserve currency. Weak performances in terms of macroeconomic strength, monetary stability, and international monetary transaction imbalances suggested an impending devaluation.Footnote 20 As the issuer of the pound sterling, the United Kingdom lacked the necessary credibility, size, and military power to fully support its currency as an international reserve currency, which may explain the low share of sterling in European countries’ reserves. However, these factors cannot account for the continued importance of sterling within the sterling area.
Bilateral Drivers of the Demand for Sterling: An Empirical Investigation
Bilateral links between Britain and sterling holders constitute alternative drivers of the persistence of sterling in the reserves of some countries. Following the literature, I examine how trade relations impacted the share of sterling in reserve portfolios (for a survey, see Reference Eichengreen, Mehl and ChiţuEichengreen, Mehl, and Chiţu (2018)). Countries with stronger commercial ties with the British market should hold a larger share of sterling in their reserves to help their importers settle transactions. I also investigate the role of the relative size of the economies of the holding countries in comparison to Britain. Richer economies have more agency and opportunities to orientate the composition of their reserves toward the assets that they deem safest and most useful for their international settlements. I study whether membership in the sterling area influences the composition of a country’s reserves and the elasticity of its reserve portfolio in response to changes in its relative economic size and trade relations with Britain. If sterling area countries had the freedom to manage their reserve portfolios without interference, they would have a higher share of sterling, on average, than European countries, and their reserves would be equally responsive to changes in their trade relations with the United Kingdom.
My sample consists of 9 European countries and 22 sterling area countries. I run the model over the period 1954–1971, as exchange controls on sterling holdings started to be lifted for non-sterling area countries from 1954 onward. I estimate Equation (1) in line with the existing literature on the determinants of the composition of foreign exchange reserves (Chinn and Frankel 2008; Eichengreen, Mehl, and Chiţu Reference Eichengreen, Mehl and Chiţu2018, 2019).
$Shar{e_{i,t}} = {X_{i,t}}\beta + S\_Are{a_{i,t}}\gamma + ({X_{i,t}} \times S\_Are{a_{i,t}})\eta + {Z_{it}} + colonie{s_{i,t}} + yea{r_t} + {u_{i,t}}$
Share i,t is the dependent variable and denotes the share of sterling holdings in the reserve portfolio of country i for year t.Footnote 21 X i,t is a vector of bilateral explanatory variables, including the intensity of bilateral trade with the United Kingdom, measured by an index of trade intensity (see Online Appendix 1), and the relative GDP of the holding country i compared with the United Kingdom’s GDP. S_Area i,t is a dummy interacted with the explanatory variables, coding 1 for members of the sterling area.Footnote 22 By introducing the interaction, it becomes possible to observe whether the reserve portfolios of sterling area countries were equally responsive to the factors driving the demand for reserve currencies as those of European countries. Z i,t is a gravity control that uses distance to the United Kingdom weighted by population size. Colonies i,t is a dummy controlling whether country i was a British colony in year t. Specific controls related to Britain, such as the credibility of its monetary policy or its financial depth, are not included as they are captured by the year fixed effects, denoted here by year t. All errors are clustered at the country level.
To estimate this model, I matched my data on foreign exchange holdings with the Historical Bilateral Trade and Gravity Dataset (TRADHIST), which gathers bilateral nominal trade flows, country-level aggregated nominal exports and imports, nominal GDPs, and gravity controls (Fouquin and Hugot Reference Fouquin and Hugot2016). Table 1 presents the results.Footnote 23 Column (1) reports the baseline without the interactions, which are added in Column (2). Column (3) includes country fixed effects. In Columns (4) and (5), I use a fractional logit model fitted by the method of quasi-maximum likelihood, as my dependent variable is continuous but bounded between 0 and 1. Other commonly used variables in similar settings are inertia, the credibility of the currency issuer, and its financial depth. As reserve portfolios are strongly correlated over time, I include in Column (5) a variable “inertia” measuring the sterling share of the reserves in 1953. This variable is used to account for the difference in the composition of the reserve portfolios at the beginning of the period studied.
PANEL REGRESSIONS
Notes: The dependent variable is the share of sterling in reserves. All errors are clustered at country level. All variables are winsorized at level 1 and 99 percent levels. To report marginal effect in Columns (1)–(3), I standardized the variables by rescaling them using the z-score method. Columns (4) and (5) report odds ratios, a coefficient smaller than 1 indicates lower odds of association between the explanatory variable and the share of sterling, while a coefficient greater than 1 indicates greater odds of association. P-values are in parenthesis. *p<0.1, **p<0.05, ***p<0.01.
Source: See data sources in the section “Method and Sources.”
The results indicate that trade relations with Britain explain part of the persistence of the international role of sterling. Countries trading more with the United Kingdom tended to have a larger share of sterling within their portfolios. However, the impact of trade for sterling area countries is consistently smaller than those for European countries. This result is consistent across specifications; however, when country fixed effects are included (Column (3)), they absorb the effect of the trade relationship with the United Kingdom, and the trade coefficients are not significant for either European or sterling area countries. The coefficients for relative GDP are negative, indicating that when countries grew faster than the United Kingdom, they tended to rebalance their portfolio away from sterling. Footnote 24 This result suggests that sterling is not an attractive reserve currency. The coefficients are larger for sterling area countries than for European countries, indicating that this effect was stronger for the former. This result is robust against the inclusion of country fixed effects and the use of fractional logit model with inertia.
Reserve portfolios were strongly associated with their past value, which is in line with the literature. When excluding the period of limited convertibility of European currencies from the sample (see Table 1, Column (1) in Online Appendix 3), the effect of inertia is more pronounced in sterling area countries compared to European countries. This finding indicates that European countries rebalanced their portfolios more quickly than sterling area countries. Finally, the coefficients for membership of the sterling area are positive and significant. Being a member of the sterling area was associated with an average increase of 50 percentage points in the share of sterling. Membership also affected the elasticity of the reserves as sterling area countries and European countries reacted differently to the drivers of sterling holdings.
These results withstand a series of robustness checks, such as excluding colonies and replacing the index of bilateral trade intensity with alternative measures of trade relationships, such as the ratio of exports to the United Kingdom on total exports and the ratio of exports to all members of the sterling area on total exports. Tables 1 and 2 in Online Appendix 3 report the estimates of these robustness checks. Only when the trade relationship is proxied by imports from the sterling area are the odds of association between trade and the share of sterling the same for European countries and sterling area countries. This suggests when sterling area countries increased their imports from the area, they rebalanced their portfolios toward sterling at a similar pace to that of European countries. But members of the sterling area were slower to rebalance their portfolios away from sterling when diversifying their export destinations. The results are also robust to alternative specifications, such as tobit regressions and an Arellano-Bond linear dynamic panel-data estimation.
These estimates contradict the characterization of sterling area monetary authorities as “free portfolio managers.” Instead, they reveal the existence of a “sterling area effect” in the international distribution of sterling holdings, which creates a distortion similar to the “empire effect” presented by Accominotti, Flandreau, and Rezzik (Reference Accominotti, Flandreau and Rezzik2011). The “empire effect” describes the biased relationship between macroeconomic fundamentals and borrowing costs for countries that belonged to the British Empire. I argue that exchange controls surrounding the sterling area distorted their reserve portfolios. The sterling area was not a free market that one could enter and leave at will because Britain threatened potential defectors with economic sanctions and asset freezes. Because of switching costs, sterling area countries were restricted in their ability to rebalance their portfolio away from sterling.
THE STERLING AREA AS A CAPTIVE MARKET
British authorities sought to maintain the existence of the sterling area to ensure that its members, which collectively held 65 percent of net U.K. liabilities in 1945, did not sell their sterling holdings more quickly than what the Bank of England could tolerate. To achieve this goal, British authorities enforced a system of controls, economic inducements, and sanctions on the sterling area members. The British Treasury imposed strict rules on the currency composition of members’ official reserves in exchange for allowing them to participate in the sterling area. Sterling area members were not allowed to build up independent reserves or diversify their assets toward the dollar or gold. Rather, they were required to pool their gold and dollars earned from capital and current account transactions at the Bank of England.Footnote 25 British authorities controlled any withdrawal from the central gold and dollar reserves when making payments to a country outside of the sterling area to minimize drains on the Bank of England’s reserves. Failing to comply with this system of controls meant expulsion from the sterling area. In the words of British policymakers, the sterling area worked similarly to Bentham’s panopticon, whereby a central authority disciplined members who surrendered their earnings (Bentham Reference Bentham, Bozovic and London1995):
At the end of the war therefore, the sterling area consisted of a named list of countries, with a strong exchange control fence around them, who surrendered their currency earnings, pooled their reserves in sterling, had complete freedom for all payments within the area and limited convertibility outside; the whole system subject to control at the center.Footnote 26
One dimension of this picture changed in December 1958: for persons outside of the sterling area, sterling became freely transferable anywhere (Naef Reference Naef2022). But controls over the United Kingdom and sterling area members remained.
The first pillar of the system was exchange controls on transactions between the sterling area and the rest of the world. These controls dated from the outbreak of WWII in 1939. They were institutionalized by the Exchange Control Act of 1947 and remained in place until 1979. Any capital or current account transactions that might have affected the reserves of the Bank of England required the consent of the British Treasury.Footnote 27 Individuals and businesses located outside the sterling area needed permissions from the Treasury to engage in commercial transactions with the United Kingdom in the form of international payments, securities transactions, coupons, or gold and foreign exchanges. Company ownership and international lending were also regulated. Violating the Exchange Control Act was punishable by imprisonment and the freezing or seizure of the concerned funds. These controls also applied to transactions between residents of the sterling area and entities outside the area, but not between Britain and sterling area members. If a country decided to exit the sterling area, its residents would be required to seek permissions from the British Treasury for capital and current account transactions with the United Kingdom and remaining members. The July 1958 Iraqi revolution provides an illustration: the British Treasury was anxious that the new regime would disregard the rules of the area and liquidate Iraq’s sterling liabilities. The Bank of England responded by announcing that it had the power to block any payments in sterling from Iraq to countries outside of the sterling area.Footnote 28 In addition, British officials “scrutinised payments, whether from the Bank of England or from commercial banks, on behalf of Iraq to addresses outside the Sterling Area.”Footnote 29
The second pillar of the British system of controls on the sterling area was threats of commercial sanctions. The majority of the sterling area countries were also members of the Commonwealth, which was granted preferential treatment by the United Kingdom.Footnote 30 The Commonwealth was exempt from the general 10 percentFootnote 31 tariff imposed on imports to the United Kingdom since the Import Duties Act of 1932 (de Bromhead et al. Reference de Bromhead, Fernihough, Lampe and Hjortshøj O’Rourke2018). The postwar negotiations and the inception of the General Agreement on Tariffs and Trade prohibited granting the Commonwealth new trade preferences, but existing preferences remained in place until 1973, when Britain joined the European Economic Community (EEC) (Cain and Hopkins Reference Cain and Hopkins2014, p. 684). In addition, from 1945 to the early 1960s, the United Kingdom relied on a system of quantitative controls on imports to limit payments in foreign currencies and protect the reserves of the Bank of England (Hemming, Miles, and Ray Reference Hemming, Miles and Ray1959; Brennan and Milward Reference Brennan and Milward1996; Schenk Reference Schenk1994a). Imports from the sterling area were consistently the most favored throughout the 1950s, while imports from the dollar area were the most restricted. Leaving the sterling area threatened participation in the Commonwealth. Departing countries would suffer from higher tariffs on trade with the United Kingdom as well as increased quantitative import controls.Footnote 32
The third pillar of British control over the sterling area was the threat of losing access to the London capital market. The British Treasury advertised for sterling area members free transit of private capital from the United Kingdom as well as access to the London market for private and public purposes. British authorities promoted the sterling area as an international payment system, which simplified international transactions for its members.Footnote 33 Only Commonwealth members could float government loans in the City, and enjoy the favorable borrowing terms advertised by British authorities. Banks and other financial institutions in sterling area countries could freely access the London money market to meet their short-term liquidity needs.Footnote 34
When several British colonies gained independence and considered leaving the sterling area, the British Treasury outlined the negative economic consequences that would result from their exit due to the measures in place. For example, consider the discussion on the case of Ghana in 1957:
The effect of Ghana of leaving the Sterling Area was worked out a few months’ ago when Dr. Krumah threatened to do unless he was guaranteed certain financial assistance. The disadvantages to Ghana […] included: Handicaps to the free flow of private capital to Ghana, Imposition of exchange control, Adverse reactions on trading relations, Injury to credit and confidenceFootnote 35
They also threatened that “it is […] doubtful whether Commonwealth preference would survive […] if the sterling area did not exist” and stated that terminating the sterling area would “be a major disruption of world trade.”Footnote 36
EXIT, VOICE, LOYALTY: TROUBLES IN THE STERLING AREA, 1945–1967
The Exiters: Trading Diversification of Reserves against Exchange Controls
The departure or exclusion of Egypt (1947), Iraq (1959), and Burma (1966) illustrate how British authorities treated countries seeking autonomy in the management of their reserves.Footnote 37
THE EXCLUSION OF EGYPT
As of 1946, Egypt was the second-largest holder of sterling liabilities, after India. It held £440 million, of which £345 million was in the reserves of the public authorities. Moreover, £400 million had come from British military expenditures during WWII.Footnote 38 In 1947, Egypt requested permission to accumulate an independent gold reserve equal to 25 percent of its currency or to release some of its blocked sterling liabilities. The British authorities, who were negotiating for a partial cancellation of their war debts with Egypt, deemed these requests “completely unacceptable”Footnote 39 and instead offered to accept a gradual release of the blocked sterling of £10 million per year.Footnote 40 While Egypt was a member of the sterling area, it had the right to request access to the Bank of England reserves in order to settle its international transactions denominated in dollars. Within just a few months of the negotiations, the British authorities began to contemplate expelling Egypt from the sterling area to block Egyptian liabilities. As the pound was not convertible outside of the sterling area, expelling Egypt from the area rendered Egyptian sterling fully unconvertible. In addition, Egypt would lose its right to access the reserves of the Bank of England.
If the negotiations break down, […] we must block the whole account, i.e. not only National Bank holdings, but those of commercial banks and private persons, to bring the whole of Egypt’s external trade to a standstill and of course affect confidence in their currency. […] To make blocking effective we should probably have to put Egypt outside the Scheduled territories.Footnote 41
On 4 June 1947, the British authorities concluded that the exclusion of Egypt from the sterling area was necessary to maintain full control over the pace of the liquidation of Egyptian liabilities. A top-secret memorandum written at the time stated,
A major British interest in the forthcoming Sterling Balance negotiations with Egypt will be to secure adequate control to prevent the Egyptians drawing down their balances or realising their securities faster than the agreed rate. We cannot be content to rely on administrative action by the Egyptians since we have not sufficient confidence in their machine […] as a long-term control, to operate for the duration of the agreement, only the exclusion of Egypt from the Sterling Area (in the Exchange Control meaning of the phrase) will suffice.Footnote 42
Excluding Egypt from the area would thus protect the reserves of the Bank of England. Next, British authorities stated that Egypt must choose between either leaving the area with an agreement on the partial release of their holdings or being “push[ed …] out to make a block effective.”Footnote 43
The Egyptian authorities opted to “go outside the sterling area by agreement” to avoid a complete block and secure an immediate partial release of £8 million with a promise of an additional gradual release of £12 million. Egypt was officially excluded from the sterling area in July 1947.Footnote 44 However, due to the decline in British dollar reserves during the 1947 convertibility crisis, British authorities decided to limit the amount of sterling they would convert to only £1.5 million, despite having agreed to £12 million a few weeks earlier.Footnote 45 Egypt now faced a dollar shortage, yet the British authorities asserted:
Egypt left the sterling area at her own request with effect from 15th of July, 1947. There is therefore no obligation on the United Kingdom to assist her in her dollar difficulties. […] It is clearly out of the question that we should make up in full the Egyptian dollar deficit. Egypt must be asked to accept some further degree of dollar austerity.Footnote 46
Egypt reached its maximum dollar withdrawal amount on 20 October, and was prevented from withdrawing more until the end of the year, which resulted in an exchange crisis.Footnote 47 A Treasury official later commented that “the chance of getting the precedent of a voluntary blocking established for a Sterling Area country was too good to be missed.”Footnote 48
The Bank of England used exchange controls to maintain a full freeze on Egyptian sterling liabilities until the next calendar year, when another short-term limited release of sterling was authorized. In 1951, a long-term agreement on the settlement of sterling liabilities was finally reached. British authorities permitted the conversion of £20 million per year.Footnote 49
Due to concerns that news of the exclusion may negatively impact ongoing negotiations with other independent countries in the sterling area, the British Treasury propagated a narrative that Egypt had chosen to leave the sterling area due to technicalities in the exchange controls.Footnote 50 In July 1947, The Economist reported on the narrative of Egypt’s voluntary departure from the sterling area,
Egypt’s decision to leave the sterling area is a product of circumstances which are peculiar to her particular case. […] Nor should the formal step of Egypt’s withdrawal from the sterling area be regarded as anything more than a technical change. […] the whole of Egypt’s external reserve will still be held in sterling – albeit unavailable sterlingFootnote 51
THE DEPARTURE OF IRAQ
The second major case of departure from the sterling area was Iraq. In 1955, the Iraqis began asking to be allowed to diversify the currency cover of the Iraqi dinar. However, their request was frowned upon by the British authorities:
It is my impression that when Iraqis speak of diversifying their currency cover they are thinking of gold as well as of other currencies. Whether this is because of the innate Oriental love of gold or not I cannot say but there is undoubtedly a feeling that prestige is enhanced if part of the national currency cover is held in gold.Footnote 52
The British Treasury allowed Iraq to convert just £5 million of its liabilities into gold between 1955 and 1957. In 1957, Iraq still held around £127 million in liabilities and demanded more conversion of sterling.Footnote 53 The Chancellor of the Exchequer answered that it would go against Iraq’s best interests to “make the switch at a bad time” as the Suez Crisis was already draining the United Kingdom’s reserves.Footnote 54
In September 1958, the new Iraqi government announced that they planned to leave the sterling area. British authorities could not prevent Iraq from leaving the sterling area but they considered blocking Iraq’s holdings and releasing them gradually, as they had done with Egypt. Eventually, British authorities abandoned this strategy to safeguard confidence in sterling. In the aftermath of the Suez Crisis, they were apprehensive about signaling to other Middle East countries, such as Jordan or Libya, that their sterling liabilities could be blocked.Footnote 55 They chose to allow a limited conversion of sterling for the purpose of current payments.Footnote 56
In June 1959, amidst ongoing financial negotiations, the Iraqis asked for a gold guarantee or, at least, a convertibility guarantee of their sterling holdings, which British authorities refused.Footnote 57 In response, Iraq decided to leave the sterling area. They held about £100 million in sterling and £20 million worth of gold and other foreign currencies. The following statement by the Iraqi Minister of Finance highlights the economic costs of staying within the sterling area and the impossibility of establishing a fully independent monetary policy:
Iraq was unable to acquire what she needed of currencies unless through the Sterling Area. The amount of foreign currencies at Iraq’s disposal were subjected to negotiations carried out at intervals. These used to depend on the position and strength of the Sterling Pound. […] It was not possible to acquire varied reserves except during the past few years and at a very meagre level at that. That situation also led to the accumulation of the Sterling balances in England. It was not possible to dispose of these balances except within certain limits.Footnote 58
British authorities publicized Iraq’s departure as primarily driven by the specific political context of Iraq rather than by the costs of staying within the sterling area.Footnote 59 Upon departure, Iraq lost the most-favored-nation status for imports to the United Kingdom and faced new import controls. By the end of 1960, the share of exports to the United Kingdom in Iraq’s total exports faced a cumulative decrease of 12 percent compared to the year before the departure.Footnote 60 Iraq also became subject to the exchange controls applicable to countries outside the area, as described earlier.Footnote 61 However, since 1958, sterling had been formally convertible outside of the sterling area; therefore, Iraq managed to convert some of its sterling holdings. At the end of 1959, Iraq’s reserves totaled £106 million, of which 37 percent was in sterling, compared with 82 percent a year earlier, a decrease of £60 million. By comparison, only £20 million had been converted from sterling to gold or dollars from 1955 to 1958.Footnote 62
THE REQUESTED DEPARTURE OF BURMA
In 1966, Burma’s departure from the sterling area represented the third significant instance of exit. In 1962, the Burmese, which held £35 million, negotiated a diversification of its reserves with the British authorities and built an independent gold holding of £15 million.Footnote 63 Between 1964 and 1965, an increase in imports due to a domestic shortage resulted in an annual trade deficit of £10 million.Footnote 64 Therefore, during this period, the Burmese spent much of their sterling holdings, keeping only 7 percent of their reserves in sterling.Footnote 65 They also started selling their forward sterling accruals against foreign currencies.Footnote 66 In addition, they introduced exchange guarantees in their commercial contracts in the form of gold clauses to hedge against a potential sterling devaluation. The clauses stipulated that if the gold content of sterling changed, then all upcoming payments would be corrected so that the original price expressed in gold would still be paid to the seller. Options to terminate a contract in case of sterling devaluation were also used.
British authorities condemned such clauses for sending a negative signal about the general confidence in the strength of sterling.Footnote 67 The Foreign Office sent a warning to Burmese authorities in October 1966, advising against these practices. In a later meeting, they repeated that the Burmese should remove the “offending” gold clauses and renegotiate the proportion of Burma’s reserves that should be held in sterling.Footnote 68 They warned that, if these steps were not taken, Burma would be expelled from the sterling area by the end of the month.
Burma moved ahead without waiting for an official expulsion. The 17 October 1966, they publicly announced their withdrawal from the sterling area, without informing the British authorities in advance. In the local press, they stated that this move was motivated by a desire to “secure freedom of action to take the necessary protective measures such as purchase of gold and investment in hard currencies, in the public interest, for the conservation of country’s exchanges reserves obtained from exports of goods and services.”Footnote 69 Following the departure, transactions between residents of the sterling area and Burma became subject to exchange controls. By the end of 1967, the share of Burmese exports to the United Kingdom among all Burmese exports dropped by 14 percent compared to the year before the departure.
The Commonwealth office requested that no mention of the fragility of sterling be made in association with the Burmese departure and stressed that there were “rules of the club which each country is expected to observe.”Footnote 70 Four days later, The Economist commented on the departure stating that “Repeated warnings from London that Burma could not reasonably expect to enjoy the privileges of membership in the sterling club if it did not observe the rules […] So, pushed, Burma opted out. It was this or expulsion.”Footnote 71
Voicing Concerns and Challenging Sterling Area Membership
Newly independent countries in the sterling area were also eager to diversify their reserves. Figure 2 shows the evolution of the share of sterling among the reserves of large independent members of the sterling area. It reveals that some countries managed to diversify more than others. I investigate their different situations in the following sections.
STERLING SHARE OF RESERVES IN MAJOR INDEPENDENT MEMBERS OF THE STERLING AREA
Sources: Author’s dataset, see text.
THE DIVERSIFICATION OF INDIA
In 1945, Sterling liabilities held in India amounted to £1.3 billion (Abreu Reference Abreu2017, p. 586). After gaining independence in 1947, India gradually liquidated its reserves through the partition, the payments of British imports, and pension capitalization (Abreu Reference Abreu2017, p. 596). Following independence, India faced expulsion from the sterling area on several occasions. Financial negotiations in 1947 and 1948 were tense as India was pressing for the conversion of some blocked sterling into dollars, and British authorities were considering excluding India from the sterling area to prevent such conversions. Exclusion was eventually ruled out, as Indian exports were essential to Britain and sterling area countries. The British Treasury used the Egyptian precedent to pressure India to accept an agreement, ultimately reached in 1948. While India had asked for a release of £200 million in three years, of which half was to be convertible, the United Kingdom allowed conversions amounting to a total of £80 million in equal installments from 1948 to 1951. In first year, only £15 million of it could be converted into dollars (Abreu Reference Abreu2017, p. 594). The releases were aimed at facilitating British exports to India at a time when the U.K. government was anxious to support British industry (Tomlinson Reference Tomlinson1985, p. 155).
In the mid-1950s, India’s trade and payments liabilities deteriorated, mostly due to the material and equipment purchased in connection with the Second Five Year Plan, a development program that came into operation in 1956.Footnote 72 India’s distrust toward sterling grew in 1956 when the United Kingdom blocked the Egyptian sterling liabilities following the Suez Crisis.Footnote 73 However, British authorities were opposed to an acceleration of the release of Indian sterling liabilities, which were then fixed at £35 million per year. They considered again expelling India from the sterling area, but because sterling had been de facto convertible since 1954 for countries outside the area, such a move would render India’s £400 million convertible into dollars. The Bank of England’s dollar reserves could not cover this amount. The British authorities would have to float sterling or block Indian sterling holdings. The first scenario was opposed by the United States. The second was no longer an option after the Suez Crisis, as many countries, especially those in the Persian Gulf, saw their holding of sterling as a political weapon in the hands of the United Kingdom. Further use of the blocking “could only be regarded as the end of sterling as an international currency and would be suicidal.”Footnote 74 The British authorities eventually decided to negotiate that India would draw $127.5 million from the IMF in March 1957 and another $72.5 million in June of the same year. As India was not permitted to accumulate a large dollar reserve, these dollars were then sold to the Bank of England.Footnote 75
Despite the resistance of British authorities limiting the release of India’s sterling liabilities, the Indian authorities gradually reduced their sterling holdings by running trade deficits with the United Kingdom. While India’s gold and dollar holdings remained stable between 1957 and 1965, its sterling holdings were divided by 5, falling from £417 million in 1956 to £85 million in 1965.
THE SOUTH AFRICAN EXCEPTION
On several occasions, South Africa either considered exiting the sterling area or was threatened with expulsion (Henshaw Reference Henshaw1996). In 1947, British authorities contemplated expelling South Africa because of large capital flows to South Africa. But because “[British] paramount interest in the gold mining industry must be protected,” they instead negotiated an agreement whereby South Africa would directly cover hard currency drawn from the Bank of England reserves through sales of an equivalent amount of gold (Henshaw Reference Henshaw1996, p. 210). This created a unique case in the sterling area where South Africa was permitted to quickly diversify its reserves: by 1955, sterling represented only 25 percent of its reserves, and by 1967, this percentage had fallen to 3 percent.Footnote 76 In the mid-1950s, South Africa demonstrated a willingness to exit the area to signal its economic independence. However, the British authorities promptly threatened to restrict South Africa’s access to the British markets for its exports in retaliation. Because Britain was the largest market for South Africa’s exports and because Afrikaners could not afford to lose access to the London capital market while apartheid policies were turning away prospective investors and creditors, South Africa chose to remain in the sterling area and send much-needed gold to the Bank of England (Henshaw Reference Henshaw1996, pp. 216–17).
For the Loyals: A “Sterling Trap”
Australia and Ireland, which were major independent players in the sterling area, appeared to remain loyal to sterling. However, internal debates occurred among local officials, leading to the adoption of undisclosed measures aimed at reducing their exposure to sterling. In 1966, Australian sterling holdings represented 13 percent of the official sterling holdings of the sterling area and 32 percent of the Bank of England’s gold and foreign exchange reserves. These same percentages were 5 and 13 percent, respectively, for Ireland.
Both Australian and Irish officials anticipated the 1967 devaluation of sterling, but refrained from liquidating a significant share of their sterling reserves out of fear of sanctions. Additionally, such liquidations would have spurred speculation, as they were major players in the market. They found themselves in a “sterling trap” similar to the one France experienced in 1931 (Accominotti Reference Accominotti2009).
The Reserve Bank of Australia (RBA) repeatedly voiced concerns to the Australian Treasury over the stability of sterling between 1962 and 1968, calling for greater diversification of reserves, but the Treasury declined to act (Kennedy 2018b). In July 1965, the RBA asked British authorities for a forward cover of around half of their sterling holdings, which was denied.Footnote 77 The RBA subsequently called for a reduction of sterling risk “without attracting attention,”Footnote 78 indicating that they understood that British authorities would refuse a deviation from sterling area rules. Egypt and Iraq constituted a clear precedent. Tensions rose between the RBA board in Sydney and the Treasury in Canberra, with the former adopting a more aggressive stance on the question of diversification away from sterling due to the weakness of the British economy. The RBA noted that “one’s currency only stays in demand as a reserve currency when one is a dominant trader.”Footnote 79 However, the Australian Treasury prioritized maintaining access to the London capital market for government borrowing, which was dependent on adhering to the rules of the sterling area (Singleton and Schenk Reference Singleton and Schenk2015, p. 1168). Only a few weeks before the 1967 sterling devaluation, RBA officials wrote,
On pragmatic grounds an attempt by Australia to make a very large switch [away from sterling] quickly would at once become common knowledge, and would be likely to start a flood of speculation against sterling.Footnote 80
In July 1968, the RBA Research Department regarded the 2 percent interest premium on sterling as insufficient compensation for the risks.Footnote 81 Limited action was taken by the RBA to accumulate IMF liquidity, the “gold tranche,” to diversify their reserves without changing sterling liabilities held in London.
A similar situation developed in Ireland. In July 1966, the governor of the Central Bank considered writing to the Bank of England to express a willingness to increase the proportion of Irish reserves held other than in sterling up to £25 million by drawing from the IMF liquidity and by purchasing foreign currencies accruing from Irish commercial banks.Footnote 82 This strategy would have allowed Ireland to diversify its reserves without drawing on the Bank of England’s gold and foreign exchange reserves, even if these reserves were assertedly available to sterling area members. The Irish authorities knew that any move against the sterling area principles would antagonize the Bank of England. The head of the Department of Finance, T.K. Whitaker, replied to the governor of the Central Bank of Ireland with the following message:
The events of the past few days, while they show how precarious sterling is, also portend an unwelcome reception for any signal of waning faith on our part. It would, perhaps, be politic not to write anything that might be so interpreted […].
As we both fully understand, what we can (or need, in reason) do to protect ourselves against the ill-effects of a devaluation is marginal.Footnote 83
THE ZOMBIE TWILIGHT
In the late 1960s, countries such as Malaysia realized that a devaluation was imminent and pressed the Bank of England for guarantees, but the Bank of England assuaged fears of any devaluation. The 14.3 percent devaluation of the pound sterling in 1967 was not announced in advance to sterling area countries, and many of them felt betrayed by the Bank after this episode.Footnote 84 The devaluation caused heavy losses for several sterling holders, as seen in Table 3 in Online Appendix 4. Kuwait, which suffered losses amounting to 5.89 percent of its GDP, had also requested in 1964 a guarantee on the value of its sterling holdings. The request had been denied on the promise that there would be no devaluation.Footnote 85 In response, Kuwait decreased its sterling exposure by limiting its pooling of gold and dollar earnings in London. This allowed it to keep its sterling holdings in London untouched while decreasing the share of sterling in its reserves from 80 to 62 percent between 1964 and 1966.
After the second British application to the EEC and the 1967 devaluation, there was little hope that the sterling area could continue. Its members tried to diversify their reserves away from sterling by buying gold and U.S. dollars from local banks, on the Euromarkets, and by reducing the pooling of their gold and dollar reserves. The Irish head of the Department of Finance notably stated that sterling had become “less valuable as an international currency.”Footnote 86 They rapidly decreased their sterling holdings from £123 million in April 1968 to £85 million at the end of August 1968, investing mostly in gold and, to a lesser extent, in dollars; the share of sterling in their portfolio thus decreased from 77.3 to 60 percent.Footnote 87 As these shifts occurred in most sterling area countries, the Bank of England considered possible responses, such as threatening exclusion, enacting stronger exchange controls, freezing assets, providing an exchange guarantee, or asking for liquidity support from the Group of 10 (G10).
Traders and bankers are reluctant to continue holding sterling […] we must be prepared to use all our powers of persuasion, […] to discourage them. In some cases, it may be necessary to consider […] a reduction in economic aid. Threatening to exclude offending countries from the Sterling Area would be unproductive; it would probably suit them very well and lead to other application to withdraw. Imposing Exchange Control […] would precipitate such applications. Blocking would be equally dangerous unless it were universal and amounted to a moratorium on our debts. […].Footnote 88
Eventually, United Kingdom officials asked for international support from the G10 and the IMF in September 1968 (Schenk 2010). The central banks of the G10 agreed to provide a $2 billion line of credit, which the Bank of England could draw on to offset declines in its reserves caused by the diversification of sterling area reserves. In exchange, they insisted that the United Kingdom negotiate bilateral agreements with sterling area countries, in which the latter would commit to keeping a minimum proportion of their reserves in sterling. In exchange, the Bank of England would guarantee the U.S. dollar value of 90 percent of the sterling reserves held by these countries. If a country were to break the agreement and lower the sterling proportion of its reserves, it would lose the dollar exchange guarantee (Schenk 2010, p. 273). The agreements also included a guarantee to maintain, at least to a degree, the sterling area’s privileged access to capital exports from Britain (Cohen Reference Cohen1971, p. 85).
Minimum Proportions of Sterling agreements (MSP) were negotiated bilaterally. For example, New Zealand was offered an MSP of 80 percent. The local authorities deemed this proposal unacceptable as it would penalize New Zealand for having “played by the rules” while other sterling area countries had been diversifying their reserves as quickly as they could.Footnote 89 New Zealand eventually secured an MSP of 70 percent, while Australia negotiated 40 percent, and Ireland, which had more than 75 percent of its reserve in sterling in early 1968, obtained 55 percent. The outcomes of the negotiations were uneven, with colonies and recent newly independent countries receiving the highest MSP, while developed sterling area countries were allowed to diversify more.
The MSP agreement succeeded in stopping the 1968 run on the pound by sterling area countries. Cohen (Reference Cohen1971) argues that the MSP agreements were “a kind of ransom paid by Britain to keep the sterling system going” (Cohen Reference Cohen1971, p. 85), while Singleton and Schenk (Reference Singleton and Schenk2015) state that sterling holders “were eventually rewarded with a dollar value guarantee for their official sterling reserves” (Singleton and Schenk Reference Singleton and Schenk2015, p. 1166). I see the MSP agreements as a form of “acceptable freeze” on sterling liabilities to allow for a continuation of the sterling area. These agreements constituted another initiative meant to limit the diversification of the reserves of sterling area countries. The British had jeopardized confidence in the value of the pound sterling with the 1967 devaluation. The “word of mouth” agreement that instituted the good practice of reserve pooling in the sterling area was gone. Gaps in the exchange control fence through the dollar markets of Hong Kong and Kuwait prevented the United Kingdom from stopping the 1968 run on sterling.Footnote 90 The dollar guarantee was needed to convince sterling area countries to collectively give up on diversification. Without an agreement, the sterling area countries would have kept slimming down their holdings, compelling the United Kingdom to devalue its currency once again.
GAINS AND LOSSES IN THE LONG LIFE OF THE ZOMBIE
The Sedation of Sterling Holders
British policymakers were not caught by surprise by the difficulties that arose after the 1967 devaluation. By 1966, they already knew that “the sterling area [was] a bank with insufficient assets to meet its deposit liabilities,” and therefore they would soon face a sterling crisis due to the low level of United Kingdom reserves compared to sterling liabilities.Footnote 91 In 1965, they considered offering a guarantee to sterling holders but decided that British interest rates qualified as a form of compensation for the risks of devaluation. Moreover, a guarantee to all sterling area holdings would be too costly in the event of a devaluation.Footnote 92 They chose “to slow down the erosion [of the sterling area] to a manageable pace,” that is, they pursued “the sedation of holders of sterling.”Footnote 93
To assess the losses of the limited diversification outside sterling induced by this strategy of “sedation of sterling holders,” I conduct a counterfactual analysis.Footnote 94 I compare the return of observed sterling holdings with a theoretical portfolio in which sterling shares are set at a working level of 20 percent of reserves and all sterling accumulated above that 20 percent threshold is converted in dollars, and kept as dollar investments. Figure 3 displays the spread of portfolio returns between sterling holdings and the theoretical mixed dollar-sterling portfolio, expressed as a percent of the end of the period GDP. If Ireland could have reduced its sterling share to 20 percent of its reserves, it would have made real gains cumulatively, representing 1.3 percent of its GDP by 1971. This assessment demonstrates the existence of a portfolio loss for sterling area countries. Despite the relatively higher interest rates in the United Kingdom compared to the United States, sterling was not a profitable choice because of devaluations and higher inflation. If sterling had been convertible after WWII, sterling area countries would have made more profitable investments by converting the majority of their sterling into U.S. dollars.
SPREAD OF PORTFOLIO RETURNS BETWEEN A MIXED STERLING AND DOLLAR RESERVE PORTFOLIO AND A STERLING PORTFOLIO
Note: Cumulative difference in returns between observed portfolios and counterfactual portfolios, 1955–1971, expressed as percent of end of the period GDP.
Sources: Author’s calculation. See Online Appendix 2 for details and sources.
British and U.S. Gains: Sterling Over-Valuation and the International Monetary System
The persistence of the sterling area after the immediate postwar years mostly benefited Britain and the City. The sterling area worked as a mechanism to restrict sterling conversion into dollars and gold when British reserves were low, firstly after the sterling crisis of 1931 and then after WWII. Most of the external sterling liabilities were held in the sterling area and formed “an important part of the inherent weakness of sterling.”Footnote 95 In the late 1930s, the United Kingdom had managed to accumulate enough gold and foreign exchange to cover 100 percent of its liabilities.Footnote 96 However, the British Treasury was never able to durably replenish British reserves after the war and struggled to maintain credible coverage of its liabilities. Except in the immediate post-devaluation period, the Bank of England reserves represented less than 50 percent of United Kingdom foreign liabilities.Footnote 97 On the contrary, the large western economies had a large coverage of their foreign liabilities by their reserves.Footnote 98
As monetary authorities from both the United Kingdom and the United States opposed a float of the pound sterling, the Bank of England also resorted to window dressing for its foreign exchange reserves in the 1960s to hide its difficulties and avoid a confidence crisis (Naef Reference Naef2020). It organized short-term swaps with the U.S. Federal Reserve (Fed) to artificially inflate its reserves just before publishing the level of the reserves in the press and in its quarterly bulletins. The Bank borrowed over £460 million from the Fed to the Treasury in May 1968 to publish a reserve level of £500 million, despite its real reserves amounting to just approximately £30 million (Naef Reference Naef2020). This was at a time when the United Kingdom’s foreign liabilities were higher than £5 billion.
The Bank of England relied on numerous international liquidity support programs throughout the period to resist the drain on its gold and dollar reserves. A $5 billion loan was first negotiated with the United States and Canada after WWII, followed by the $89 million Marshall Aid in 1948 (Cairncross and Eichengreen Reference Cairncross and Eichengreen1983, p. 114). The Suez Crisis precipitated a $650 million drain on British reserves, forcing the British to negotiate a $1.8 billion stand-by agreement with the IMF and the United States to reassure markets. Further agreements were negotiated during the 1960s with the IMF, the BIS, and Western European central banks. From 1965 onward, the Bank of England had to draw regularly on international liquidity, as seen in Figure 4.
INTERNATIONAL LIQUIDITY ASSISTANCE USED BY THE BANK OF ENGLAND
Note: This figure reports the use of the international facilities made available to the Bank of England since the time of the first Basle agreement of June 1961.
Source: Archives of the Bank of International Settlement, LAR2 F02.
The existence of the sterling area and the authoritative enforcement of its principles enabled British authorities to maintain an international presence of sterling within the Bretton Woods international monetary system. Had foreign sterling liabilities been freed earlier, the Bank of England would not have been able to cope with the inflow of sterling and would have been forced to devalue. In the words of United Kingdom officials, capital and exchange controls aimed at supporting the international use of sterling to “give [the United Kingdom] command of resources” and help them “remain a first-class power.”Footnote 99 The pooled gold and dollar reserves of sterling area countries at the Bank of England also helped to finance the United Kingdom’s own deficits and the expansion of its expenditure.Footnote 100
By being forced to hold sterling as their primary reserve, sterling area countries also contributed to the prosperity of the City of London (Krozewski Reference Krozewski1993; Schenk 2010, pp. 212–19). British banks and insurers benefited from the fact that sterling was used in invoicing 25 to 30 percent of international trade.Footnote 101 The Bank of England considered that relieving the United Kingdom of “the burden of an international currency” would be “at the expense of destroying the financial mechanism of the City. […] Obviously this could not be the Bank’s answer. The U.K. economy needs the City’s financial and commercial acumen […]. Trade still follows the flag (or the £).”Footnote 102 However, the reality was that the flag was mostly gone, trade was mostly gone, but foreign sterling liabilities persisted.
After WWII, the United States permitted the continuation of the sterling area, perceiving several advantages, even if exchange controls and commercial preferences violated Bretton Woods principles. In the late 1940s, the United States feared that the termination of the sterling area would weaken the link between some of its countries and the Western world. In the context of the developing Cold War, they regarded the currency area as a mechanism to deter its nations from aligning with the Eastern bloc (Cairncross and Eichengreen Reference Cairncross and Eichengreen1983). From the early 1960s onward, they started to object to the conversion of foreign sterling liabilities into dollars, as such operations would increase the pressure on the reserves of the Fed, which were already frequently under attack from speculators.Footnote 103 Moreover, U.S. authorities were committed to an international monetary system based on fixed exchange rate stability; sterling was seen as the first line of defense for the dollar (Bordo, Monnet, and Naef 2019). Sterling’s devaluation in 1967 renewed speculation on the dollar’s parity with gold, resulting in the closure of the U.S. gold window in 1969.
CONCLUSION
This paper examines the international role of sterling during the Bretton Woods era and characterizes it as a zombie international currency, that is, a currency that would not have survived as a reserve currency without British authorities’ interventions in the forms of exchange controls, threats, and economic sanctions. Similar mechanisms were actively put in place by Russia in the wake of the 2022 Ukrainian War. Therefore, studying sterling’s history provides an example of the potential gains and losses associated with such interventions.
After WWII, the United Kingdom did not have the economic fundamentals of an issuer of an international currency. Countries that could access alternative foreign exchange reserves, such as Western European countries or Iraq, chose not to hold sterling. I show that trade relations with the United Kingdom had a low impact on the composition of sterling area countries reserves compared to European countries’ reserves. In the 1930s, the sterling system was based on the carrot of a strong, highly desired currency and on the stick of imperial power and colonial government. After 1945, as war debts crippled the Bank of England, sterling was no longer desired, but the stick remained. The sterling area constituted a captive market in which countries were dragooned into keeping their reserves in sterling. The area was designed to prevent the liquidation of foreign sterling liabilities and protect the fragile reserves of the Bank of England. British authorities used threats, propaganda, and sanctions to curtail the divestment of sterling assets. These manufactured high switching out costs explain most of the permanence of these foreign sterling holdings. The expulsion of Egypt from the area and the departure of Iraq and Burma resulted from British opposition to economically rational motives: diversification and insurance against valuation risk. The countries that remained in the area faced portfolio losses. When commercial and exchange control sanctions became less credible due to policies of trade liberalization and the development of Euromarkets, sterling area members increasingly free-rode the rules of the area to decrease their exposure to sterling risk, which led to an erosion of the reserve role of sterling. Only a G10 intervention compelled British authorities to provide a guarantee in exchange for the limitation of the divestment out of sterling.
This research illustrates the unsustainability of coercive diplomacy to maintain zombie currency. Investors seek a liquid and easily convertible international currency and avoid exchange controls and asset freezes. Imposing exchange controls to prevent asset liquidations can only delay the problem of investors trying to run the zombie currency. Coercion can only last for so long. Like a zombie bank, the zombie currency issuer will eventually need to negotiate with its creditors if it wishes to resume normal operations and reopen its capital markets.
Open access
