The political economy of international currencies: Assembling a framework

The term international currency is often used to describe a currency that is used beyond its borders, whether for commercial transactions or official purposes (for example, as a reserve currency).Footnote ¹¹ Benjamin Cohen speaks for many when he describes an international currency in functional terms as a national currency that exhibits extensive cross-border use.Footnote ¹² For our purposes a dominant international currency is more than a reserve or international currency. It also serves as the main vehicle to denominate international exchange. More important, a dominant currency augments demand at the global level. Our usage is closer to Susan Strange’s identification of a ‘Top Currency’, which is a currency that non-nationals use because of the inherent economic and political advantages of doing so.Footnote ¹³ Nevertheless, we will stick with the more widely-used term ‘international currency’ (occasionally adding the adjective ‘dominant’) to facilitate engagement with the prevailing literature. But it should be clear that for us an international currency is not simply a convenient alternative to barter. Since the Napoleonic Wars, the global political economy has not supported more than one genuine dominant international currency at a time. We thus assume that the current highly globalised and liberal political economy will remain anchored by a single dominant international currency. This is why threats to the dollar by the RMB (or other potential rivals) matter for understanding the future trajectory of the global economy.

Analyses of recent challenges to US dollar hegemony, have, with few exceptions,Footnote ¹⁴ focused almost entirely on the international conditions that enable currency internationalisation.Footnote ¹⁵ They highlight many international factors in what are often multidimensional analyses: the depth of financial markets as a precondition for a currency’s dominance, or the fundamental ability to print money and have it accepted by other official authorities through central bank networks, or for geopolitical concerns.Footnote ¹⁶ We agree that these international foundations matter. But in our view, these foci ignore the role of domestic considerations in currency internationalisation; indeed, they may even derive from those considerations. Rather, we synthesise a more comprehensive framework highlighting those domestic foundations by returning to three early theorists of the political economy of money: Robert Triffin, Charles Kindleberger, and Susan Strange. The role domestic institutions and demand play in currency internationalisation is latent in their analyses.

Triffin was among the first to note the growing contradiction between the role the US dollar played in the early post-1945 global economy and its fixed link to gold.Footnote ¹⁷ Triffin agreed with Keynes that international gold stocks alone could not support adequate growth in global demand. Expansion of the global money supply via the accumulation of dollar balances abroad facilitated the growth of other countries’ reserves and therefore also their ability to support economic growth. Unlike Keynes, Triffin framed this as an issue purely of liquidity rather than demand: did countries have enough dollars to meet ongoing trade obligations?

Triffin’s eponymous dilemma arose from the contradiction between resolving the world’s ‘illiquidity problem’ on the back of a single state’s currency and global confidence in that currency. Triffin predicted that growing foreign dollar balances would eventually exceed US gold holdings, calling into question the fixed rate between dollars and gold.Footnote ¹⁸ The build-up of dollars stemmed from a constant outflow of dollars via investment, unrequited transfers (including military expenditures), and tourism, which boosted global growth and produced inflation (rather than the relatively consistent deflation of a pure gold standard).Footnote ¹⁹ In the absence of such outflows, European economies would have grown much more slowly. Triffin’s solution here was to internationalise the provision of world liquidity, which would in turn internationalise the transmission of global demand.Footnote ²⁰ Yet Triffin missed the domestic side of the confidence problem. The fixed link with gold made him and others focus on gold holdings, yet what surely mattered was whether the United States had an economy capable of validating foreign-held dollar stocks with exports of desirable goods and services. During the 1950s, US growth was powered by continuous export surpluses that in effect validated the build-up of overseas dollar holdings – their value was underpinned by future projections of growth in the American economy. But from the late 1960s forward, continuous net imports subtracted both from US growth and undermined the industrial base for validating overseas dollar and dollar-denominated asset holdings. Only an innovative and productive economy would make it plausible that financial assets sold today could be redeemed tomorrow at their present value. It is the potential for growth, rather than the size of gold stocks, that validates the issuance of financial assets. Ultimately the source of the Triffin dilemma was not a concern with money qua cash, but rather about the future value of dollar-denominated financial assets. This is, for us, the chief value of returning to Triffin.

This aspect of Triffin’s dilemma leads us to Kindleberger’s slightly different framing of the liquidity issue, which looked at New York’s role as a global financial centre.Footnote ²¹ During the Bretton Woods period, New York financial institutions provided intermediation by borrowing short and lending long on a global scale. In doing so, they accommodated what Kindleberger saw as non-American economies’ greater preference for liquidity. Kindleberger’s anxious Europeans parked their funds in US assets. Global liquidity, in other words, was not simply a function of the volume of imports a leading economy consumed (thereby transferring its currency abroad), but could be mediated through the creation and distribution of financial assets.Footnote ²² When Kindleberger first offered this insight (in the 1960s), America’s balance of payments consisted of trade and current account surpluses balanced by the very large capital account deficits noted above. Increased global dollar holdings required intermediation via a globally oriented network of US-based financial institutions.Footnote ²³ Like Triffin, Kindleberger saw a connection between world liquidity provision (that is, monetary stocks) and global demand, but he emphasised the channels through which liquidity flowed.Footnote ²⁴

Kindleberger argued that financial intermediation was essential for global demand growth in both mature and developing economies, and therefore that the precise structure of financial centres mattered. The organisation and operation of financial institutions and the markets they participate in determines how financial assets denominated in a particular currency acquire a global scale and role. Relying on increases in a national currency to fuel global demand is ultimately neither better nor more efficient than relying on gold; each has an inherent limit (amounts of newly-minted gold versus rates of inflation). The growth of financial assets, however, is much more elastic, especially where the purchasers and holders of these assets have confidence in the future capacity of an economy to redeem them.Footnote ²⁵ As Triffin noted, paper assets, not cash, ultimately provided liquidity in both centuries.Footnote ²⁶ Yet maturity transformation, like net exports, ultimately drains liquidity from the global system. Put simply, borrowing short to lend long inevitably forces the long-term borrower to generate more income – an export surplus with the long-term lender – in order to pay the higher return expected on long term lending.Footnote ²⁷ Pure maturity transformation (intermediation) is simply a slower version of the problem created by using lending to put currency into circulation. It could be argued that risk adjusted returns should equilibrate. But this implies risks for the actor doing maturity transformation, for example in the form of financial crises should specific banking intermediaries fail. And even if a political entity accepted such a risk, it would still have to generate funds to resolve banking crises.

Contra Kindleberger, then, simple intermediation and net lending abroad only boosts liquidity and demand temporarily. Loans can only be repaid, or valorised, if the lender ultimately runs a deficit. Repayment is a critical feedback loop in our framework: in the nineteenth century for example, continued capital outflows from London eventually returned to the UK as imports, the payments for which valorised (or sanctioned) the continued issuing of bills and bonds drawn on City bankers and merchants. The point here is that financial intermediation is itself built upon an economic structure that supports the extension of global demand through transactions that exchange real wealth at some point in the future.Footnote ²⁸ But this points us toward Susan Strange’s analysis of the international and domestic political bases for a dominant currency.

Strange’s work allows us to connect the external demand for dollars, for increased global monetary liquidity, to domestic politics.Footnote ²⁹ Shortly before the demise of the fixed link between gold and the dollar, she noted that the American economy was particularly vulnerable to the effects of international monetary shocks because of the way in which the central position of American capital markets in the global economy amplified those disturbances. She pointed out that a top currency country (in our terms, the country emitting a dominant international currency) was paradoxically more rather than less vulnerable to fluctuations in global demand.Footnote ³⁰ Strange recognised that the central monetary problem of the 1920s and 1930s was the reduction in global demand generated by French and US trade surpluses.Footnote ³¹ This mirrored persistent deflationary pressure on both agricultural and manufactured goods prices in the nineteenth century. The only way to boost liquidity and demand in the global economy in a sustainable manner, she thought, was for a dominant economy eventually to run trade or current account deficits. However, Strange also recognised that these deficits would generate adjustment costs, primarily job losses (due to rising imports) among low and medium-skilled workers, which would in turn lead to a political demand for welfare provision.Footnote ³² She further noted that these groups fared worse in the United States than in other Organization for Economic Cooperation and Development (OECD) countries. As was the case with Britain and sterling, she argued that the United States could only run the world’s top currency successfully if it cushioned marginal or vulnerable domestic groups from the dislocations they would face. This was the key political cost of running an international currency.Footnote ³³

Triffin, Kindleberger, and Strange lead us: (1) to recognise the link between an international currency and growth in global demand; (2) to extend the means for providing global demand to include financial intermediation and the generation of financial assets; and (3) to connect the provision of global demand (and its fluctuation) to rising domestic adjustment costs in the international currency emitter, and to the political need to ameliorate these costs in order to prevent social dislocation. We contribute to this framework by considering how these costs can be met by preserving or increasing purchasing power. By preserving or increasing its population’s purchasing power, a state running an international currency helps to lift global aggregate demand while simultaneously achieving domestic social stability. We highlight below the importance of two historically important policies that have boosted purchasing power and global demand (from among several) because of their magnitude: food policies and land or home-ownership policies. In this we go beyond these three theorists to isolate and explore some of the domestic political arrangements that we believe are central to enabling a state to run an international currency, because these policies produced the assets that expanded global demand in each century. In the nineteenth century, British free trade policies for cheap food lowered the cost of consumer non-durables for low-income groups, thus increasing their purchasing power, and validated assets created by British investment abroad. Similarly, in the twentieth century, US housing finance policy increased access to credit and thus to purchasing power for low- and middle-income groups in America, while also expanding the pool of internationally tradable assets.

The usual factors many identify as the international foundations for the internationalisation of a currency – currency convertibility, deep and sophisticated financial markets, sufficient currency in global circulation and, finally, a state capable of negotiating the complex monetary and political arrangements that are the international infrastructure of an international currency – do matter. But we also insist that this literature must recognise that a state that issues a dominant international currency must also be able to negotiate the domestic adjustment costs associated with running such a currency. These costs will include some combination of increased unemployment associated with import competition and/or an appreciating currency, but also very likely increased inequality resulting from a skill mismatch among its labour force as the finance sector grows in significance to service the demands of currency internationalisation. The claim we advance here is that these costs, which come from many different competitive pressures on firms to realise production efficiencies, need to be negotiated for social stability to be sustained. Although the food and land and home-ownership policies we highlight here worked in different ways, each helped to counter the domestic economic costs (and social dislocation) generated by running the world’s principal international currency. Neither encompasses all the side-payments that make it possible to run current account deficits, but each carried significant weight. Moreover, each connects directly to the production of assets providing liquidity and demand in the global economy.

International currencies in the modern period: Historical considerations

The historical record of sterling and the dollar shows how the domestic elements we have identified intersect with and support the more widely recognised elements of currency internationalisation. For a state to influence global demand it must eventually run a trade, current account, or balance of payments deficit large enough to expand global liquidity. This deficit necessarily expands global demand. But by definition, these deficits slow economic growth in the issuing country (net imports are a subtraction from GDP for a mature economy). These deficits will generate distributional conflicts among domestic economic sectors and actors that the state must address to preserve social peace.

Our analysis thus begins with domestic considerations. First, we consider how state institutions can mediate and/or ameliorate the losses that balance of payments deficits impose on vulnerable domestic groups. How a state negotiates these costs is critical for the sustainability and stability of an international currency. In both cases, we can identify efforts undertaken by the British and US state which offset the domestic costs of running an international currency.

During the nineteenth century, the simultaneous consolidation of the world’s largest empire and the pre-eminence of British exports of goods, services, and capital combined to make the pound sterling the world’s principal international currency.Footnote ³⁴ At the same time, falling prices for British manufactured goods drove almost a century of continual deflation under a nominal gold standard system. What underwrote both a relatively stable global monetary order and a growing global economy in these conditions? Global liquidity and therefore demand increased through the issuance of foreign public and private sterling securities in London, rather than via growth of the gold stocks notionally backing many currencies.Footnote ³⁵

Although Britain ran a small goods and services surplus until about 1870, this masked an enduring trade deficit throughout the entire nineteenth century. What offset the trade deficit was a large inflow of interest and dividend payments from abroad that in turn generated a current account surplus almost until the outbreak of war in 1914.Footnote ³⁶ From a global demand point of view, however, this surplus returned to the global economy either through the capital account (as foreign investments that recycled interest or dividend payments) or the trade account (as imports, primarily as foodstuffs). In this sense Britain’s goods and services deficit helped to generate global demand for the agricultural products of developing countries and colonial areas, with foods accounting for roughly 45 per cent of British imports by value from 1870 to 1913.Footnote ³⁷ Equally important, the transfer of purchasing power via the City of London reinforced this demand. London was the principal source of long-term capital investment for the world throughout the nineteenth century.Footnote ³⁸ For example, capital was raised in Britain for investments in Argentine ports and railways, and these investments – which also generated British industrial production (of railway components, steel, machinery, etc.) – were repaid by shipping Argentine agricultural products to Britain and other European markets. Sterling denominated financial assets generated by the City not only served to increase global demand; they were valorised by the complex relationship between British industrial capacity and the new productive capacity of developing agro-export economies. Despite periodic financial crises, financial intermediation was thus supported by an exchange of goods and services that manifested itself in an enduring British trade deficit throughout this entire period of international monetary order.

The spectacular growth of food exports to Britain after the repeal of the Corn Laws in 1846 was both the physical manifestation of asset validation and the hinge linking the domestic and international political economy elements of sterling’s internationalisation. The rising and powerful class of British industrial magnates obviously benefited from access to foreign markets. As early as Ricardo this class was conscious of the benefits that might flow from cheaper food for the growing British proletariat, and also of the connection between real wages, demand and deflation.Footnote ³⁹ But some did not fare well. British elites faced social unrest as nineteenth-century economic expansion created an ever more militant working class.Footnote ⁴⁰ Yet this nascent labour movement did not contest free trade and the global centrality of sterling, because falling prices for imported food — the ‘cheap loaf’ — helped to double real wages in the nineteenth century. Britain imported 60 per cent of its calories by 1913.Footnote ⁴¹ Under relentless import pressure, the price of wheat in Britain fell from 8.26 shillings per bushel in 1820 to 3.84 shillings in 1913.Footnote ⁴² Rural land prices consequently fell continuously throughout the nineteenth century.Footnote ⁴³ Agricultural imports and falling food prices thus placed unremitting pressure on rural landowners and their workers. But a gradually enfeebled landed aristocracy exported their bodies to the colonies as an administrative elite (alongside their workers, who came as settlers), and exported their capital into urban real estate, which increased in value as Britain became a manufacturing and services economy.

Expanding food imports both brought a certain amount of domestic social peace and facilitated external compliance to the global role of sterling. Rising volumes of raw materials exported to Britain and Europe generated capital gains for those politically influential landholders in the global periphery who could tap into global flows of capital and labour to develop their land. These peripheral landholders and their bankers in turn parked their earnings in short-term deposits in London, making them short-term creditors on Britain, and thus doubly beholden to sterling. These interlocking domestic and international arrangements were a critical political counterpart to the macroeconomic flows that sustained the international role of sterling during the nineteenth century. They also produced the pattern of short-term borrowing and long-term lending that Kindleberger identified with respect to global capital intermediation through the post-Second World War United States financial system.Footnote ⁴⁴

A similar albeit differently constituted set of interlocking political, economic, and social arrangements provided the foundations for the role of the US dollar in the post-1945 global political economy. The American state’s domestic interventions supporting the dollar’s global role dwarfed what the British state could undertake during the nineteenth century.Footnote ⁴⁵ The US version of the cheap loaf was the New Deal and Great Society welfare state programmes. These both responded to and reinforced significant shifts in electoral political coalitions by extending first welfare benefits and then also civil rights to marginalised and disenfranchised citizens. As with Britain, a few key programmes sustaining domestic growth linked the internal and external arenas. Chief among these were deficit-funded military spending and the restructuring of the financial system to support broad home ownership. Both of these policy directions generated pressures that contributed to American balance of payments deficits.Footnote ⁴⁶

But it was American housing policy that in our view played the same role that free trade in agricultural products did in mid-nineteenth-century Britain, particularly after the dollar went off gold in 1971. Through the 1960s and 1970s home ownership spread throughout the American middle class, and inflation amplified the wealth factor of home ownership by reducing the real value of housing debt in the United States (that is, nominal real estate asset values grew faster than nominal mortgage debt). This helped to increase the purchasing power of American homeowners by reducing their debt burden. After the mid-1980s, however, global goods and services disinflation took hold and flowed through the US housing finance system as debt refinancing. Now it was inexpensive housing-related debt that helped to increase middle-class purchasing power.Footnote ⁴⁷ This increase in purchasing power was in turn amplified by the emergence of asset-inflation – including house prices – in a low-interest rate environment. Growing purchasing power on the part of middle-class Americans was reflected in an ever-larger array of imports, which were physically paid for by generating and exporting dollar-denominated assets, including and perhaps especially mortgage-backed securities (MBS).

By 2008, US MBS constituted 9 per cent of the global public and private bond market, which is one reason why an ostensibly domestic market-oriented, non-traded sector could influence the global financial system. The creation of these MBSs, many of which were bought by foreigners, also served to support ever-larger US trade deficits.Footnote ⁴⁸ At its peak, in 2008, foreign holdings of US mortgage related debt amounted to 15 per cent of all foreign portfolio investment in the United States, and foreign entities held 20 per cent of US Agency (for example, Fannie Mae) debt.Footnote ⁴⁹ These assets were the financial counterpart to the export surpluses that powered growth in key US allies such as Japan and Germany, in emerging market economies such as China, and in many states within the Organization for Petroleum Exporting Countries. During the 2000s, the United States accounted for between 50 and 80 per cent of global trade deficits in any given year.Footnote ⁵⁰ Meanwhile, those assets owed their existence to enormous US state intervention in asset markets. In 2014, for example, roughly 80 per cent, or US $7.7 trillion of the total stock of US $9.4 trillion in securitised debt in US financial markets, had the full faith and credit of the US federal government behind it, and most of that debt was mortgage backed securities.Footnote ⁵¹

As with Britain in the nineteenth century, the period in which the United States ran trade surpluses, recycled purchasing power via lending, and positioned the US dollar as a dominant international currency, was much shorter than the period of trade deficits and full-scale dominance. As Barry Eichengreen and Marc Flandreau point out, the dollar was not dominant in the interwar period, when the United States ran large and destabilising trade surpluses.Footnote ⁵² Those surpluses were recycled as overseas lending, creating conditions for the global financial collapse of the early 1930s.Footnote ⁵³ During the relatively brief Bretton Woods period – at most the twenty years from 1950 to 1971 and at worst the decade from 1959 to 1971 – US postwar export surpluses were recycled as aid and direct investment rather than loans, obviating repayment pressures. But even then, US import demand provided a significant growth impulse for the European and Japanese economies. The theoretical possibility of expanding liquidity primarily via lending corresponds to a relatively limited period of time by contrast to the extensive liquidity and demand provided by substantial current account deficits. During the (roughly) 45 years post-Bretton Woods, cheap imports and easy credit – about three-fourths of US consumer debt is housing related – helped to support groups disadvantaged by the dollar’s central role.

This schematic examination of the British and American cases illustrates the critical connections between the international and domestic foundations for an international currency. In the nineteenth century, the British decision to open its market to food imports challenged agricultural interests, but it increased the purchasing power of the growing working classes being forced into factory jobs in the new textile and manufacturing sectors. This single policy stimulated aggregate demand in the global periphery, as countries such as Canada and Argentina expanded agricultural production for British and European markets, while at the same time placing more purchasing power in the hands of British working classes by dramatically reducing their food bills. To be clear, we do not claim here that this was the intended outcome of these policies, only that these were the effects on the operation of international monetary order as it evolved over the nineteenth century.

Similarly, after the Second World War, the way in which the US balance of payments worked to stimulate aggregate global demand shifted from foreign aid and military expenditures (that is, the Marshall Plan and spending on the Korean and Vietnam wars) to capital outflows and then, after 1980, to the creation and take-up abroad of a wide array of domestically generated financial assets.Footnote ⁵⁴ In the 1990s, a large part of these assets consisted of mortgage-related products linked to the operations of Fannie Mae and Freddie Mac, the government-sponsored entities mandated to make home ownership as widely available as possible to Americans. These policies decreased housing costs, and thus had the effect of increasing the purchasing power of lower- and middle-class Americans. These policies were further supported by the home mortgage interest federal tax deduction and by federally sponsored mortgage insurance.Footnote ⁵⁵ Throughout most of the post-1945 period, US housing costs largely tracked general price inflation, unlike in many other societies. And while stagnant real wages after 1970 stemmed in part from the erosion of manufacturing capacity in the face of growing imports from Asia — in other words, from the cost of supplying an international currency — those imports also lowered the cost of consumer non-durables. US prices for consumer non-durables net of food and beverages rose 12 per cent more slowly than all prices, 1990 to 2014, and prices for apparel – virtually all of which is imported, and which constitutes a large portion of low-income groups’ consumption – barely rose at all over those 25 years.Footnote ⁵⁶ Even so, the domestic costs were high enough to propel Donald Trump to victory in the 2016 presidential election.

In both the British and US cases, therefore, the domestic political costs of running an international currency were offset by bargains compensating vulnerable parts of the population. In the British case, food policy and sterling’s global predominance were intimately intertwined; in the US case, while housing policy was almost entirely a domestic concern, the foreign take-up of MBSs before and especially after 2000 clearly entangled it in the global role of the dollar. Yet as we shall see in the following section, these kinds of domestic political bargains have yet to receive attention in the growing literature on the internationalisation of the RMB.

China and the RMB: the intersection of international and domestic considerations

The small but growing literature on RMB internationalisation parallels the treatment of currency internationalisation more generally, with a focus on the external economic constraints and functions performed by such a currency. Paul Jenkins and John Zelenbaba, for example, follow Cohen’s early taxonomy of roles played by any international currency.Footnote ⁵⁷ Paul Bowles and Baotai Wang provide a thorough but similarly categorised assessment of these policies, which they see originating in China’s response to the vulnerability of its exports to fluctuations in the value of the US dollar. They see RMB internationalisation as a protective policy that will be successful to the extent that a sustainable, long-term path toward liberalisation of the capital account is achieved. Domestic considerations only surface when they note that China’s domestic capital markets are not yet deep or liquid enough to sustain full liberalisation.Footnote ⁵⁸

Analyses of RMB internationalisation focusing on domestic political considerations do so in one of two ways. On one hand, questions connected to economic growth can foreshadow the kinds of distributional questions we highlight because the size of China’s economy is directly related to the prospect of the RMB becoming an international currency. For example, Takatoshi Ito considers the demographic determinants of China’s future economic growth, which in turn will affect RMB internationalisation. There are, however, no efforts to connect these demographic trends to China’s ability to affect global aggregate demand.Footnote ⁵⁹ Similarly, both Hongying Wang and David Steinberg touch on domestic considerations in a recently published volume on the question of the politics of RMB internationalisation.Footnote ⁶⁰ Yet both stop short of exploring the social dimension of China’s domestic politics, preferring instead to focus more narrowly on how domestic economic considerations intersect with China’s global monetary ambitions.Footnote ⁶¹ On the other hand, the question of regime stability often enters the discussion, as when Cohen concludes his analysis of the RMB’s future by noting that political rigidity in China erects important barriers to successful internationalisation (however without specifying how this rigidity may operate).Footnote ⁶² In effect, most of the literature on RMB internationalisation mirrors that on the euro, where the primary focus concerns the international arrangements that characterize currency internationalisation.Footnote ⁶³

In contrast, as our discussion above suggests, the ability to project a domestic currency into the global economy as a principal international currency rests on stable domestic political institutions that can generate new demand in global markets via trade, current account, and/or balance of payments deficits, along with the capacity to ameliorate the costs that arise from those deficits. China currently faces significant institutional weakness on both accounts. Three pieces of evidence are central to our understanding of China’s structural problems as well as the weaknesses of official efforts to respond to China’s reliance on the US market for growth. First, China averaged trade surpluses of 4.7 per cent of GDP from 1999 to 2008; these surpluses are now shrinking, but still averaged 2.8 per cent from 2009 to 2014.Footnote ⁶⁴ Second, though the wage share of GDP is difficult to calculate precisely, the World Bank estimates that Chinese household final consumption fell from 46 per cent of GDP in 2000 to 37 per cent in 2014 – a level well below that of every other Asian developing economy except Singapore (which however has systematic forced saving through its Central Provident Fund).Footnote ⁶⁵ We of course recognise that in absolute terms real household consumption nearly tripled from 1999 to 2013, and real GDP per capita nearly doubled from 1999 to 2008. But consumption growth clearly fell short of supply growth or the consumption share would have risen. Instead, much of that gain went into infrastructure investment, real estate and the creation of export capacity. Third, exports jumped from roughly 20 per cent to 39 per cent of China’s GDP from 1998 to 2008, far ahead of the rise in imports, and net exports accounted directly for one-third of growth before the 2008 crisis.Footnote ⁶⁶ Despite its considerable appetite for raw materials, the Chinese economy has not been adding to net aggregate global demand in line with the growth in its economy.

These conditions reflect the core institutional features of the Chinese political economy. China has a largely repressed financial system that captures and channels household savings toward large state-owned or quasi state-owned firms.Footnote ⁶⁷ Until recently, savers had almost no choice but to deposit money into the state-controlled banking system at essentially negative interest rates. Slightly better off savers could speculate in real estate, although this also tended to reinforce the economy’s overreliance on investment for growth and more recently has generated a potential real estate bubble.Footnote ⁶⁸ State-owned firms (SOEs) and localities borrowed these savings at low or negative real interest rates to fund the productive and infrastructure investment that has powered Chinese economic growth until 2009, and the real estate boom that powered growth afterward.

These state-owned firms possess varying degrees of monopoly power, allowing them to systematically capture profits. The spectacular increase in Chinese savings (and thus current account surplus) from 37 per cent of GDP in 2000 to over 50 per cent in 2007 is largely explained by an increase in the state share of savings from 2.6 per cent of GDP in 1999 to 21 per cent in 2008.Footnote ⁶⁹ Much of this is a function of retained earnings by SOEs. Rather than contributing to expanded domestic consumption, SOEs channelled those profits away from mass consumption and into expanded productive capacity and then real estate speculation.Footnote ⁷⁰ Close ties with the Chinese Communist Party assured that these firms were not punished for this mal-investment, perhaps because party elites and their families were skimming SOE profits.Footnote ⁷¹ Some estimates suggest that insiders skimmed as much as US $1 trillion since 2009.Footnote ⁷² Moreover, families in the top income decile accounted for a disproportionate share – almost 40 per cent – of unoccupied housing held for speculative purposes.Footnote ⁷³ In any case, SOEs’ privileged access to bank credit comes at the expense of the rest of the economy. While SOEs provided only 13 per cent of employment in the late 2000s, they accounted for over half of non-agricultural fixed investment.Footnote ⁷⁴

China’s repressed financial system reduces households’ ability to consume, and thus to import. Roughly speaking, savers faced a 3 per cent per year negative interest rate through the 2000s.Footnote ⁷⁵ At the same time, the dismantling of the ‘iron rice bowl’ – the old system of employment-based social welfare – and the enforcement of hukou regulations that link eligibility for the new local welfare state to one’s registered domicile, put pressure on Chinese households to self-insure by saving large proportions of their income. The hukou system means that the hundreds of millions of internal migrants in China lack access to the welfare state in the place where they currently work. Chinese households are thus ‘target savers’, who increase their savings in response to low returns, rather than dissaving when interest rates fall. Target saving creates the funds that China’s banking system on-lends to the SOEs mentioned above. This is one reason China has been able to mobilise so much savings for investment, and thus attain historically unprecedented rates of growth. Target saving drastically reduces Chinese households’ ability to consume, and by implication their ability to import. Recent reforms to the hukou system have not substantially changed these flows.

Local government behaviour reinforces these tendencies toward over-saving and under-consumption. Local government does about 80 per cent of state spending in China, but independently raises only about 30 per cent of tax revenue. It thus relies on centrally raised and redistributed revenues. More than half of local government revenues come from taxes on local businesses, with another quarter coming from local governments’ share of the national value-added tax.Footnote ⁷⁶ Alone, this reliance on local businesses for revenue would incline local government to be solicitous of local firms’ profitability. But three other factors magnify this orientation. First, perhaps most important, is the orientation to ‘GDP-ism’ and ‘FDI-ism’ – high rates of local growth were a major metric for promotion in the Communist Party until quite recently, and similarly, FDI represented essentially cost-free resources for local officials. Second, local government is relatively corrupt, encouraging collusion among local developers and local governments seeking revenues. And third, land sale and lease revenue contributes another 10 per cent of locally controlled funds, which creates discretionary funds for these same governments. These three factors motivate local government to over-borrow and over-develop local real estate, with local firms using access to local government-controlled banks to fund this development. This dovetails with the more general tendency toward over-investment and under-consumption noted above.

China’s political economy is thus structurally biased toward the creation of overcapacity and excess exports, and we agree with those who suggest that this will not be easily unwound.Footnote ⁷⁷ While the shift towards real estate speculation after 2009 shrank China’s current account surplus as a share of GDP, it did little to lower its absolute size. China will have difficulty transforming the RMB into an international currency because China cannot currently supply the world with enough RMB to expand global demand sustainably. While China could lend to the world, this would most likely increase China’s overall balance of payments surplus as those RMB return to China as purchases of Chinese-made goods. As noted in the preceding section, more liquidity can induce some investment. But when Britain and the United States provided intermediation, they borrowed short from the rest of the world and lent back to it long term. It is difficult to see how this could make sense in reverse for rich country investors and firms. Foreign firms, for example, would have to be sure of RMB cash flow in order to repay those loans. And that cash flow can only come from a Chinese current account deficit generated by Chinese demand for foreign goods and services. Furthermore, China limits foreign capital inflows that might accumulate RMB denominated assets. RMB loans to foreigners might combine with Chinese firms’ unrepatriated RMB-denominated earnings to create a euro-RMB market akin to the euro-dollar market. But while this may be a necessary condition for a dominant currency, it is not sufficient. First, extant euro-currency markets have not displaced the dollar. Second, RMB denominated debts have to be paid off through RMB earnings, which ultimately means net exports to China.

All other things being equal, China’s trade and current account surpluses subtract from global demand, which is exactly the opposite of what the pound and dollar systems did and/or continue to do.Footnote ⁷⁸ While China’s growth has certainly increased the demand for raw material and component parts imports to China, the corresponding outward flood of cheap Chinese exports depresses goods production in both rich and emerging markets. China’s trade surpluses reflect a lack of domestic demand that would otherwise absorb local production. If Chinese consumers’ purchasing power were higher, China’s export surplus would fall as those consumers imported more tourism, health care, education, environmental protection, and so on. Externally, Chinese competition displaces manufacturing jobs that almost always pay higher wages than do service sector jobs. This decreases purchasing power and thus global demand.Footnote ⁷⁹ As with US trade surpluses in the 1920s, China’s efforts to recycle that demand – to recycle its trade surpluses – take the form of lending to the United States via the purchase of Treasury and Agency debt. Rather than internationalising the RMB, this lending merely reinforces the US dollar’s pre-eminent position. Even those who argue that the US dollar is in decline acknowledge that China is in something of a dollar trap on account of its large dollar-denominated asset holdings.Footnote ⁸⁰

The financial repression and export orientation at the heart of China’s current political economy stimulated growth for two decades, but going forward it seems more likely to act as a drag on growth. High levels of investment validated by external demand led to extraordinarily high rates of total factor productivity growth, as Chinese and foreign firms matched rural workers to modern machinery.Footnote ⁸¹ Increased productivity accounted for more growth than increased capital inputs from 1980 through to the late 1990s; both dwarfed increased labour inputs. But in the 2000s, China’s incremental capital output ratio (ICOR) began rising – physical capital inputs eclipsed productivity gains as a share of total growth. Where a capital investment of US $2 sufficed to produce a US $1 increase in China’s GDP in the 1980s and 1990s, by the 2000s a similar US $1 increase in GDP required more than US $4 of new investment.Footnote ⁸² Subsidised finance made these physical investments profitable, but also assured the diversion of funds into projects that could not pay a market rate of return. Indeed, Jinghai Zheng, Arne Bigsten and Angang Hu suggest that capital productivity actually declined between 1995 and 2005, and these trends only worsened as investment shifted from export production to real estate construction after 2010. Insofar as one precondition for the RMB to challenge the dollar is a Chinese economy whose faster rate of growth pushes absolute GDP to a multiple of US GDP and Chinese per capita GDP to within shouting distance of US per capita GDP, China’s recent growth slowdown and its 2015 mini-devaluation suggests that the RMB will struggle simply to become a regional currency. The data on trade invoicing confirm this. Although the RMB is now the second largest currency for denominating letters of credit in transactions through the global SWIFT (Society for Worldwide Interbank Financial Telecommunication) network, its 2015 share was roughly 9 per cent versus 90 per cent for the US dollar, and in 2016 its share actually declined slightly. Moreover, fully 80 per cent of those transactions occurred among Chinese and Hong Kong entities. The RMB thus barely functions as a regional currency for China’s periphery.Footnote ⁸³

In this context, the trade and current account deficits that would have to accompany an effort to internationalise the RMB, even just as a reserve currency, would cause considerable domestic financial distress and thus encounter considerable political opposition. On one hand, shifting resources to consumption would reduce both the credit available to overleveraged firms and the demand for heavy industrial and construction inputs. Shifting resources to households also implies reduced retained earnings for SOEs. Reduced demand and earnings would put even more pressure on firms that need to roll over large existing loans. At the same time, SOEs are engaging in a large cross-border carry trade with the US dollar, indicating a lack of profitable investment opportunities and a search for greater yield.Footnote ⁸⁴ On the other hand, as Michael Pettis has long argued, shifting income towards lower income groups in China necessarily involves shifting income away from elites and the enterprises they control.Footnote ⁸⁵ Even reforms that merely raise the cost of capital to privileged firms reduce their income, given the degree of subsidy built into current lending and deposit rates.Footnote ⁸⁶ After all, in one important respect China’s repressed financial system is designed to ensure that savers ‘contribute’ to national investment through low or negative returns; why then would foreign investors take up RMB assets in the absence of massive reforms (which strike at the heart of the existing Chinese growth model)? Indeed, the first half of 2015 saw capital outflows of about $500 billion, which may well be a vote of non-confidence in the sustainability of current efforts to alter China’s growth model.Footnote ⁸⁷ And although capital outflows abated in 2016, they have not reversed. Our point is not that a shift towards a more consumption-oriented economy is impossible, but rather that it faces considerable political opposition. Otherwise President Xi Jinping would not have to expend so much political capital trying to shift the system.

Conclusion: ‘Outside-the-box’ thinking about the political implications of currency internationalisation

We have argued that a critical feature of an international currency’s political economy is constituted via the domestic institutions and policies a state uses to negotiate sustainable social trade-offs around the costs of supplying global demand. An international currency emitter adds to global demand by putting its currency into circulation outside its borders. The resulting balance of payments deficits necessarily impose costs on domestic groups losing jobs and income, which states must accommodate through some combination of polices. Before considering the political implications of this argument for China and RMB internationalisation, we need to address two important objections to our framework and argument. The first objection is that, by following Triffin, we have mistakenly equated increases in currency circulation – liquidity – with demand, just as Triffin was mistaken to equate increases in foreign dollar balances with growth in global demand. However, we do not argue that liquidity and demand are identical. Rather, we argue that a real, material set of economic exchanges is necessary to validate increases in world liquidity as growth in global demand. Otherwise increased liquidity just flows into Keynes’s liquidity trap.

To return to the nineteenth-century example, capital raised in the City for investments in other parts of the world not only funded these investments but they acted as collateral for financial institutions both in the City (as assets) and in other parts of the world (as funds deposited in accounts and used in turn by those financial institutions to lend on to others). Because of the global position of sterling, its circulation had a double effect in expanding both global liquidity and global credit. But this effect was supported by the real economic exchange that those investments enabled once they came on stream. If such investments had failed to generate new productive capacity, then increases in world liquidity would have been illusory or inflationary. The fact that the global circulation of sterling not only generated new production abroad but also facilitated the return to Britain of the fruits of that production (in the form of merchandise imports of foodstuffs primarily) transformed liquidity increases into demand growth. This material process allows us to connect liquidity and demand on a world scale.

The second objection is that we have mistakenly confused the means by which an international currency attains a global circulation. Our emphasis on the role of balance of payments deficits obscures, for some, the real source of currency circulation: financial intermediation. This, as Benjamin Cohen points out recently, is the main way that sterling and the dollar attained their position atop the ‘Currency Pyramid’. In his view there is no necessary connection between the trade and current account deficits both economies eventually ran and the sustainability of sterling and the dollar as principal international currencies in their respective eras.Footnote ⁸⁸ The problem with this understanding, in our view, is that it obscures how adjustment mechanisms operate in the real economy. The trade and current account surpluses of the United States (in the interwar period) and Britain (until the 1870s) were complicit with and even constitutive of global deflation. American and British lending ameliorated but did not eliminate financial stress on borrowers as they sought to make payments in the key currency. Similarly, China’s very low level of household consumption generates large trade surpluses; domestic consumption as a share of GDP must rise to eliminate those surpluses, as it did in Britain during the last quarter of the nineteenth century and in the US post-1945.

Our framework also easily incorporates financial intermediation as a key mechanism of a currency’s global circulation – Kindleberger is, after all, our guide on this. But we also recognise that the economic exchange encouraged by such global circulation, which historically has manifested itself as a trade and/or current account deficit, generates real adjustment costs for the states that emit an international currency. We think it is theoretically compelling to include the domestic political consequences of these real economic processes as part of our framework. Unlike Cohen, therefore, we see these deficits as a necessary part of the structure of economic exchange that a state must negotiate if it is to run an international currency in a sustainable manner. We believe they are, in social science terms, necessary (although not sufficient) conditions for running an international currency.

The priority that we place on the negotiation of domestic trade-offs to meet the costs of running an international currency raises for us the thorny question of the link between democracy and currency internationalisation. Our argument suggests that this relationship must surely be a fit subject for future research, given the striking connection between the costs we have identified with running an international currency and the political trade-offs required to meet those costs.Footnote ⁸⁹ Of course, neither Britain in the nineteenth century nor the United States in the middle years of the twentieth century had full suffrage. But a large part of the population could vote and was mobilised by elites contending (at least in part) over trade and monetary policy. Elites in both countries offered concessions to lower income groups and to some losing business sectors. These concessions were dual use – they brought both mass compliance and secured what we have argued is an essential foundation for running an international currency. In the British case, free trade around food imports brought rising real wages to the working class while allowing Britain’s foreign debtors to service their loans in sterling with proceeds of their exports. In the US case, suburbanisation helped expand the domestic economy while generating financial assets (MBS) that could be sold offshore. This gave the US middle class rising living standards, plugged the gap in the US balance of payments and helped the global economy grow.Footnote ⁹⁰

Can China and its elites generate similar compromises? Party elites, state-owned enterprises, and local government elites currently dominate China’s political economy. All stand to lose from any relaxation of what has been identified as a ‘state-permeated market economy’ and its attendant repressed financial system, which currently subsidises the investments whose profits they harvest.Footnote ⁹¹ Indeed, the cynical and bottom-up phrase guo jin min tui – roughly ‘state advances, private sector retreats’ – has replaced the older top-down policy slogan of ‘grasp the large, let go of the small’. The shortfall in domestic demand that arose from Party and SOE control over profits produced large trade surpluses that helped China become internationally competitive, but this narrow interest also means that China’s relatively poor citizens subsidise not only their own domestic elites, but relatively rich Americans as well. While a political base for reversing these perverse subsidies exists, it requires activating and mobilising a mass public. Quite how entrenched elites will voluntarily give up their rents in the absence of countervailing political forces, and substantially expand political participation that will challenge their own privileged position, is for us a critical question to ask about RMB internationalisation.

Historical parallels in the rest of developing Asia offer some guide to this delicate transition. Both Korea and Taiwan had state-owned banking sectors that provided subsidised loans to industrial firms at the expense of savers. Like China, these countries had demand shortfalls, although Korea periodically ran trade deficits in high-growth periods. For both, the shift away from a wholly export-oriented economy occurred in parallel with the end of martial law or authoritarian government and a shift toward higher consumption. The subsequent democratic political regimes of the 1990s reinforced these trends, with Korea experiencing a credit-driven consumer economy rather than an investment economy after 2000. As in China, this shift threatened significant entrenched interests. A combination of external and internal pressures – unionised assembly line workers in Korea, the multitude of ethnic Taiwanese small business owners in Taiwan – partially overcame the residual interest of big firms for subsidised credit. Even so, Taiwan and Korea both remain substantial net exporters. Currently the consumption deficit in China is larger than was the case in either Korea or Taiwan, suggesting that even if elite interests in China succumb to pressure for the financial reforms needed for comprehensive internationalisation of the RMB, China will still be a net drag on global demand via its current account surpluses.

Thus, in the absence of something like the depression and war that induced change in America, or the unprecedented conditions that permitted British imperial expansion, we do not expect the internationalisation of the RMB to proceed to the point where it genuinely rivals the US dollar as the world’s pre-eminent international currency. Even with China’s enormous economic heft – on a purchasing power parity basis it is almost as large as the US economy – the RMB has barely attained the status of a regional currency. As noted above, the RMB mostly functions as a regional currency for China’s periphery.Footnote ⁹² Unless Party elites can find a way to discipline their own membership and shift them toward domestically oriented production and a smaller share of what will still be a growing economy, the Communist Party and indeed the Chinese state more generally are unlikely to follow their Asian neighbours’ precedents; indeed, a reversal of openness cannot even be ruled out.Footnote ⁹³ The most important limits to RMB internationalisation, in this sense, are the domestic political foundations of China’s economy.

We close with one final theoretical reflection. Although we have refrained from identifying the broader theoretical tradition of political economy within which we work, it would be a mistake to equate it with a version of open economy politics, despite our concern with the distributional effects of running an international currency.Footnote ⁹⁴ We do not believe any of our theoretical inspirations would agree with the methodological principles underpinning open economy politics; they, like ourselves, would be much more sympathetic to the historical and critical sensibilities pioneered by Strange, Kindleberger, Keynes and Schumpeter (and, to a more limited extent, Triffin), who anchored their understanding of current possibilities and trajectories in the historical record and its associated historical narrative. Here we would issue a call for more historically-informed research to be conducted into international monetary order, including the place and role of international currencies in the contemporary global political economy.