Keyword

Revisiting Inconvertible Currencies from the Ground

Modernity in money has been implicitly discussed from a viewpoint of a state-oriented teleology. Such a viewpoint implies that, like water flowing downward, sooner or later, a monetary system will incorporate transactions nationwide into a coherent order under governmental control. Here, the goal towards which a monetary system endogenously evolves as market activities develop within the territory of the state would be the establishment of a set of convertible currencies under one sovereign authority. Colonies should have been no exception. Contemporary researchers in West Africa and Indochina before the Second World War, for example, had no hesitation in depicting as unsophisticated or unreasonable the attitudes of local people towards supporting inconvertible currencies of their own (McPhee 1926, p. 248; Robequain 1939, p. 157).

Convertibility between currencies means there is no transaction costs associated with the exchange of inconvertible currencies within a country. Once this idea is accepted, it is not so difficult to narrate a story of monetary evolution that can be applied universally to any society. Whenever a phenomenon which is not in conformity with a convertible currency under state sovereignty is encountered, it is located among activities associated with pre-modernity, with backwardness in an evolutionistic sense, or with embeddedness within an indigenous culture in a relativistic fashion.Footnote 1 According to this concept, a fluctuating exchange rate between currencies should be fixed to provide the market with certainty for transactions on subsequent occasions. Fixed proportions among different denominations should work to reduce the cost of issuing a large quantity of small denomination currency.

In these terms, a modern history of money should resemble a process of moulding activities using inconvertible currencies into a system of a national convertible currency. Currencies exclusively mediating local transactions should be integrated into a currency available across a nation in order to facilitate interregional exchanges and consequently to enhance the division of labour beyond a locality. Convertibility between small and large denomination currencies should also boost long-distance trade and economize the remittance of official revenue by cutting the transportation cost of coins. The propositions mentioned above fit the expectations of rulers whose administrations depend on taxation paid in currency, and of the authorities for monetary supply who profit through a monopoly in currency issue.Footnote 2

When turning, however, from the perspective of the state to people engaged in business at the ground level, the same phenomenon tells quite a different story. Where peasants made up the majority of inhabitants and most exchanges were made among them locally, such exchanges inevitably experienced a seasonal surge after harvest and a contraction in the slack season. Transactions among peasants were of small amounts, requiring fractional denominations of currency, unless deferred payment was used. The more freely a small-household peasant could negotiate in a marketplace where one-time transactions dominated, the smaller the denomination of currency they needed to reach a mutually agreed price. Fractional currencies mediating transactions in a marketplace were not always in accord with the means of exchange supplied by merchants engaged in long-distance trade. From the viewpoint of providing peasants with a means of exchange on the spot, an unfixed exchange rate between a local currency in a smaller denomination and an interregional currency in a larger denomination could work to supply a flexible means of exchange in transactions between the end-users. In this way, the two currencies work in a complementary, rather than a substitutive, relationship.Footnote 3

Simply speaking, the necessity for a fractional currency changes, and the effect of a floating exchange rate between the means of proximate and distant exchange varies according to the degree of freedom of the ordinary peasant. As far as market activities depend on one-time transactions, local dealers engaging in exchanging currencies functioned to prevent the marketplace from suspending activities due to shortages of currency supply. A society relying on currency actually tends to suffer from chronic shortages of currency, since a currency is easily distributed among end-users while being less easily retrieved from the end-users on demand. A fractional currency in particular tended to remain with the end-user and not return to the marketplace. When a marketplace encountered a serious scarcity of currency, local dealers did not hesitate to generate a means of exchange locally by themselves, even as a temporary device, to meet the demand for transactions among end-users. Thus, from the viewpoint of end-users, locality and/or temporariness of money reflected their demands (Kuroda 2020).

Across the world and throughout history, the most salable goods, such as grains and cloth, often worked as money among ordinary peasants. This is quite an endogenous means of exchange for local transactions, in so far as the grain or cloth are produced locally. In some societies, meanwhile, the authorities provided people with metallic currencies whose values were low enough for the populace to use to make transactions. Since such currencies of low value could not be expected to bring in a high seigniorage, rulers often issued them mainly for military purposes, supplying food for soldiers. In other societies, fractional currencies of a distant origin were imported by traders into the marketplace for peasants. Typically, cowries, shell money from the Maldives, were carried into the inland of southwest China (Yunnan), eastern India (Bengal, Orissa), and West Africa. Rulers sometimes used cowries for their taxation and administration, but it was traders who imported bulky cowries at a distance and distributed them among inland villages. In Africa, some endogenous currencies originally produced manually were substituted by those manufactured by foreign factories in the nineteenth century. A large quantity of cotton cloth from Massachusetts (and later from Bombay) worked as money (Pallaver 2019, pp. 76–77; Prestholdt 2008, pp. 74–85), as did manillas made of brass, manufactured in Birmingham and Nantes (Guyer 2004, p. 72) and glass beads from Venice (Pallaver 2009).

The long journey of some foreign-origin currency reveals an aspect of what caused locals to actually accept an item as money. It was a shared recognition among dealers along a trade circuit of an item as money, rather than any enforcement by the authorities and/or the intrinsic value of the item, that sustained the acceptability of the item as money in long-distance exchange. In medieval East Asia, copper coins bearing era names of the Song (960–1279) dynasty continued to circulate until the seventeenth century across the China Sea region. In particular, old Chinese coins dominated transactions of high value, such as land sales, in medieval Japan. At that time, the physical appearance of coins, including the use of old era names, mattered to local dealers when shroffing coins between standard coins for accumulating assets and operational coins for local transactions.Footnote 4 A similar phenomenon was associated with the widespread circulation of the Maria Theresa dollar (MTT) in Africa and the Middle East until the early twentieth century. Here silver coins bearing the date of 1780 and minted in Vienna continued to circulate at a premium. Keeping the same Empress’s head and the same date was more decisive in it being accepted by local dealers than any intrinsic value or authorization (Kuroda 2007). In extending those findings more generally, we can see that neither commodity value nor government enforcement is a necessary condition to generate money, though both conditions can enhance people’s acceptance of it. A long circuit of dealers who all recognized the old appearance of the item endorsed its acceptability as money beyond the locality. Such cases demonstrate that money works as a social circuit connecting transaction demand.

A function endogenously mediating exchanges locally made does not exclude other exogenously organizing payments being settled at a distance. These are not always in a substitutive relationship but often in a complementary relationship. A division of labour between copper coins for proximate exchange and silk (later silver ingots) for long-distance exchange existed throughout ancient Chinese history. In India cowries serving transactions in rural marketplaces and silver rupees settling long-distance business and tax payments mutually depended on each other, as did cowries and manillas in West Africa and amole (salt bars) in Ethiopia, which complemented other currencies, including the Maria Theresa dollar (Kuroda 2020). Coupling between the endogenous invention of money and the exogenous introduction of money always generated a variety of customs in the intermediary sphere that interfaced bottom-up monetary establishment and top-down monetary regulation. In the mezzo-scope sphere, neither micro nor macro, implicit or explicit agreements among dealers generated various devices such as imaginary money and currency circuits, for the purpose of maintaining a monetary supply to meet demand in a flexible way.

A global history of money viewed from the ground makes clear the ubiquity of inconvertible currencies that include fractional currencies, petit dealers, and local customs not in accordance with interregional settlements. In other words, such currencies resulted from a common demand for endogenous exchanges. However, ubiquity does not exclude the possibility that each civilization possesses its own characteristics in the monetary system. What is striking about precolonial African money is its openness towards an interregional, including intercontinental, flow of items that worked as money. Cowries, glass beads, manillas, mechanically woven cloth, and the Maria Theresa dollar circulated as currency in Africa, though they were not used as money in their places of origin. Cowries also circulated in India, and the Mexican silver dollar won popularity in China in a similar way to the MTT in Africa. However, monetary supply in precolonial Africa appeared to depend on items from abroad to a greater degree than in Asia. We will consider what the differences mean in this chapter.

Peasants, Marketplace, and Atomic Currencies

Across the world and throughout history there has been a high probability that fractional-value currencies such as cowries were required particularly for transactions made at marketplaces where a number of peasants gathered for exchange.Footnote 5 It often happened that dealers purchasing the products brought by local peasants must have prepared a significant amount of small currencies whose denominations fitted the size of the transactions made by ordinary peasants. Thus, as far as dealers conveyed portable large denomination currencies, they had to exchange large denomination currencies for small denomination ones. Importantly, dealers would agree to a loss in conversion from large to small, since the demand for small currencies on the spot caused them to appreciate.

In other words, there is no demand for fractional-value currencies where there are no marketplaces. For example, the Wolof people in Gambia had few marketplaces and they used cloth as currency, not fractional-value currencies. In the early twentieth century a strip of cloth five inches wide and seventy inches long was valued at threepence and sixteen strips were equivalent to one large goat (Ames 1962, p. 40). In contrast with the use of cowries in other areas of West Africa, the Wolof did not use fractional currencies whose values were smaller than the cloth strips. They were too expensive to use in negotiations in the marketplace. Actually, an observation of a Wolof household by Ames, in 1950–1951, revealed that most exchanges were made among neighbours without going to the marketplace (Ames 1962, pp. 44–45).

An English observer at a marketplace in early nineteenth-century India gives us a vivid example. On the morning of the market day, exchangers brought to the marketplace as many cowries as they could carry on their head or on the back of an ox. They exchanged 5760 cowries for a silver rupee in the morning and gave a rupee for 5920 cowries in the evening. A Japanese research group found in a periodic market in early-1930s China that merchants from the city had to convert silver dollars into copper coins in order to do business there, since local peasants brought small quantities of goods, such as ten to twenty eggs, while a silver dollar was valued at around one hundred eggs. Petit exchangers worked to sell copper coins for silver dollars at a fluctuating rate according to demand/supply (Kuroda 2020, pp. 31–32).

A fluctuating rate between currencies and the importance of dealers were evident in Africa as well. In early twentieth-century Ethiopia the rate of a salt bar to a MTT changed according to, for example, famine or military action. An American observed that in such a situation one of ten locals engaged in exchange to make a profit (Kuroda 2007). Importantly, harvests caused the exchange rate of small currencies to appreciate by season. For example, the exchange rate of manillas in terms of sterling rose during the harvest season in West Africa (United Africa Company 1949, pp. 47–48).

How small was the currency required in transactions among end-users? Robequain observed in early twentieth-century Indochina that locals made a zinc coin, sapèques, small enough to purchase a slice of papaya or a cup of tea (Robequain 1939, p. 155). Hereafter, we will call a single unit currency for indivisibly minute transactions atomic currency.Footnote 6 It might not have been just a coincidence that daily wages for unskilled labour were quoted at around one hundred pieces of currency in different societies, such as 100 cowries in Uganda around 1900 (Pallaver 2015, p. 489), 100 copper coins in Taiyuan, China, around 1900 (Kuroda 2020, p. 133), and 100 copper coins in Ikeda, Japan, late eighteenth century (Nakagawa 2003, p. 305).

Meanwhile, dependence on fractional currencies apparently brought inconvenience. Atomic currencies are so bulky that dealers cannot carry them without paying a significant cost. The experience of C. H. Robinson, a Briton travelling in West Africa, highlights this point. He failed to sell an unserviceable horse since he would need to hire fifteen porters to carry cowries to complete the transaction; this would make the deal meaningless for him (Vice 1983, p. 13).

The merit of supplying atomic currencies in a marketplace is to support the seller and the buyer, both of whom were engaged in small business, in minutely gradating the price through negotiation and enabling them to finally find a point where both could agree. However, while it is easy to distribute fractional currencies to end-users, it is not easy to reassemble them from the end-users on demand. As a result, a significant quantity of small currencies had less opportunity to return the way they came and stayed within the households of the end-users. Some were piled up in treasure houses; Basden observed heaps of cowries in an Igbo village (Basden 1921, p. 199). Alternatively, they were buried, awaiting the chance to be reused. Sometimes such a chance did not arise, and they were left underground, to be occasionally dug up at a later period by peasants, construction workers, or archaeologists.

A copper coin hoard dating from late nineteenth-century Sichuan, China, shows that the majority of coins were minted in the mid-eighteenth century (Kuroda 2020, pp. 46–47). They stayed underground for a century and a half, waiting to be used. There are similar cases across the world, including in Africa. In 1912, when suffering low cash reserves, the French West African government tried to collect a poll tax from locals. It received a significant amount of 5- and 10-centime coins dated 1855 with the image of Napoleon III, while people in Dahomey tried to keep 1- and 2-centime coins and prevented the government from retiring the smallest coins (Manning 1982, pp. 60, 312). We can only imagine how huge quantities of cowries must have been buried underground, though, unlike coins with dates, shell money does not allow us to tell the year they were collected from the sea.

Stagnancy of currency with end-users meant that while currency proliferated it was rarely assembled. In other words, such currencies made just a one-way trip. In particular, a large quantity of cowries from the Maldives, manillas from Birmingham and Nantes, and glass beads from Venice made a long single journey as far as villages and towns in West and East Africa but never made any return trip, even though some may have been collected by a local bureau through taxation. On the other hand, some currencies working to bridge local marketplaces did travel bilaterally, or made a long circuit to return to a hub city. The case of the Maria Theresa dollar shows a circuit crossing several administrations. The Austrian silver dollars were transported from Aden to Gare in West Ethiopia to purchase coffee, and passed through Gambella and Khartoum to Port Sudan in exchange for cloth, before returning to Aden (Kuroda 2007). However, we should note that, at the continental level, the silver dollars only made a single trip from Vienna to Africa or the Middle East and never returned. This means that other devices facilitating bilateral settlements internationally, such as the pound sterling, were used in collaboration with the MTT. Complementary combination of currencies could organize several layers of exchange, from local to global.

Local currencies mediating mainly proximate exchanges were ubiquitous across the world. In the case of early twentieth-century China, native notes issued by local dealers functioned as a device for supplementing currency shortages. They were available mostly within a walkable distance (Kuroda 2020, p. 180). The notes were usually denominated in copper coin with a square hole in the centre. They are referred to as copper cashes. Copper cashes were atomic currencies issued by Chinese dynasties for two millennia. The use of copper cashes generated units of a larger denomination when dealers tied a certain quantity of the coins together by a string, thus avoiding always counting them piece by piece in transactions. The quantity of copper cashes tied together differed region by region, and sometimes commodity by commodity. The differentiation by locality in the quantity tied together also appeared in the custom of using cowries in Africa. Naturally, in both China and Africa, dealers conducted business in terms of string oftener than in terms of atomic unit. However, it did not mean that they neglected quantities smaller than a string of currency. Foreign observers recorded that local merchants sometimes spent long hours counting copper cashes piece by piece in China (Smith 1899, p. 51) and cowries in West Africa (Law 1997, pp. 199–200). There were even professionals who counted cowries in the markets of Yorubaland (Ofonagoro 1979, p. 639).

Thus, we can confirm the widespread combination of peasants, marketplaces, atomic currencies, and petit dealers in Africa and Asia. The interdependent relationship between the local marketplace and long-distance trade resulted in the differences in monetary functions between local currencies and interregional currencies. There was an argument that the influx of precious metals into Africa did not cause a clear spike of prices, since gold and silver functioned not as money but as ornaments (Hogendorn and Johnson 1986, p. 146). A similar understanding also applied in India and China, that precious metals were absorbed by them not for monetary purposes (Kindleburger 1989, pp. 62–64, 70–72). Unfortunately, these scholars did not understand the multiplicity of market activities and the complementarity between local currencies and interregional ones, which caused precious-metal currencies to disconnect from price-making at the ground level.

However, the way of conjunction between proximate exchanges at the marketplace and at distant exchanges beyond the locality appeared to differ continent by continent. In Africa, a region in which atomic currencies mediated transactions among peasants at the marketplace adjoined another region in which neither minute money nor the market played an important role. For example, as mentioned earlier, the Wolof did not depend on marketplace exchanges and used mainly cloth. In the early twentieth century they used five-franc silver coins in cash transactions (Masaki 2016). Meanwhile, cowries were excavated from many archaeological sites across Senegal and Mali (Haour and Christie 2019). However, cowries were not in circulation as currency but used only for divination in Senegambia where the Wolof dwelled. In contrast, cowries worked importantly as currency in Mali and Songhay (Green 2019, pp. 71, 323–325).

It was found globally, including Africa and Asia, that rural markets made a periodic rotation circle in the opening days to facilitate peddlers’ business. Although it is necessary to investigate further, Hill has pointed out the lack of hierarchal organization connecting marketplaces in West Africa (Hill 1966, pp. 297–298), especially in comparison with the case of China formulated by Skinner (Skinner 1964, pp. 9–10). Interestingly, the investigation in Katsina Emirate showed no close relationship between the density of population and the number of marketplaces (Hill 1971, pp. 308–309). In contrast, as the density of population increased, the number of marketplaces rose in Hebei, Jiangsu, and Zhejiang, China, in the eighteenth through early twentieth centuries (Ishihara 1987, pp. 111–114, 140–141). Denser networks connecting marketplaces generated a hierarchical structure inserting a reservoir between basic rural marketplaces and cities. The intermediate market worked as a place for peddlers to stock both the necessities sold to peasants and their products collected at basic markets, and to sell some sophisticated goods like stationery that local elites demanded (Skinner 1964, pp. 26–27).

The lesser importance of intermediate markets between basic marketplaces and long-distance trade routes meant that end-users of atomic currencies in marketplaces in West Africa might have more opportunities to directly expose their business to interregional dealers than those in China. The acceptance of foreign-origin currencies to a deeper degree in Africa, mentioned earlier, might have been in tandem with the direct connection of basic marketplaces with interregional trade circuits. The direct connections might explain why cowries, copper wire, and iron bars did not circulate together and formed an exclusive currency circuit respectively at the ground level. As a result, different types of currency circuits appeared side by side, like a mosaic, across the African continent. For example, in early twentieth-century Eastern Nigeria, traders relied on manillas and cowries in the Ngwa region, brass rods and copper wire along the Cross River, and gin and iron bars on the coast (Martin 1988, p. 43). Local currency circuits flourished in early twentieth-century China as well, but they differed one from the other by different kinds of the same item, such as different types of silver dollar much oftener than by different materials, as was the case in Eastern Nigeria mentioned above (Kuroda 2005).

It is impossible to explain what caused the difference in monetary usage between China and West Africa simply as whether or not intermediate markets were buffered between periodic rural markets and ports for long-distance trade. However, it might not be wrong to suppose that until the first decades of the nineteenth century a heavy dependence on the export of slaves from the inland to ports made a midway reservoir less necessary in West Africa than in China, since slaves transported themselves to their destinations on the coast. Even after a serious decrease in the Atlantic slave trade in the second half of the nineteenth century, local rulers used slave labour to farm goods such as palm oil to sell directly at the port. Some former slave ports survived as ports for shipping legitimate products, such as the case of Lagos (Hopkins 1973, p. 145). Importantly, large-scale palm collectors in West Africa may have been able to directly deal with European firms on the coast (Hill 1971, pp. 312–314). Later we will revisit this issue related to the organization of markets in examining the circulation of paper money.

Exporting Peasant Products Circulated New Currencies

It seems that, before the nineteenth century, African economies depended on intercontinental trade more heavily than Asian ones, since in many areas of the continent the slave trade pushed out a significant quantity of the labour force as a commodity from the inland to overseas. However, intercontinental trade did not have such a drastic affect as to transform African monetary structures. Distant trade stimulated the increase in usage of endogenous-type currencies, such as cowries, rather than bringing in currencies which were in circulation in Europe. The abolition of the slave trade in the nineteenth century made African economies depend more on the export of legitimate commodities such as palm oil, groundnuts, and hides. After colonial occupation, governments tried to introduce the currencies supplied by monetary authorities in Europe, but it was difficult to diffuse them among African end-users. Stagnant prices for agricultural commodities in Europe discouraged peasants in Africa from getting more involved in cash crop production through the second half of the nineteenth century, especially in the last quarter of the century, called by contemporaries the Great Depression in Europe (Hopkins 1973, p. 133). Throughout the period, the introduction of steamships and railways reduced the cost of transportation. The innovations decreased the cost of transporting precolonial currencies to the extent that they depreciated against European monies. However, in the last quarter of the nineteenth century, European currencies were not able to substitute for indigenous currencies; rather, they could work only by adapting to local conditions. For example, sterling silver coin tokens were backed by gold in Britain, but they had no specific backing in West Africa and were used with a strong appreciation in the season when export products were sold (Hopkins 1970, pp. 104–105). In this sense, although sterling silver was brought from Britain, it became inconvertible with currencies in the home country soon after it arrived in Africa.

Unlike an economy dependent on plantations, an economy based on numerous peasant households did not allow foreign firms to control producers’ activities at the ground level. Under such an extremely decentralized condition, in collecting their products European authorities had to supply currencies whose value was fractional enough to be equivalent with African atomic currencies. Until the end of the nineteenth century, unlike the early twentieth century as we will see, there was not yet such a strong demand for peasant products in Africa to the extent that European firms would dare to spend a lot on substituting a new type of small currency for existing ones.

During the colonial period, traditional currencies were useful among African peasants for ceremonial occasions as well as for economic activities. In southeast Nigeria, for example, manillas, not British currency, were used for ceremonial purposes such as funerals and the payment of the bride price (Naanen 1993, p. 433). In East Africa, cowries had been indispensable for performing bridal ceremonies (Pallaver 2015, p. 491).

Meanwhile, in the marketplace, westerners observed that Africans were very keen to make profits through exchanges (Basden 1921, pp. 196–197). For example, when Europe suffered from a cotton famine in the 1860s caused by the American Civil War, Senegalese farmers chose to cultivate groundnuts instead of cotton because they knew precisely that the former provided a higher profit. In another example, farmers in the Gold Coast and southern Nigeria began cocoa farming in the 1880s exactly when the export of palm oil and groundnuts declined. African peasants had a strategic sense to flexibly change their crops according to market conditions (Hopkins 1973, pp. 137–139), but circumstances had not been favourable enough for them to make the most of their business sense until the end of the nineteenth century.

The tide of global trade turned around 1900. Intercontinental trade then penetrated marketplaces inland to absorb peasant products for the first time. After the stagnant prices of the 1870s through mid-1890s, the terms of trade for agricultural products kept rising until the First World War. Unlike slaves in the previous period, products by peasants at the ground level became important export commodities. In the case of West Africa, typically, palm oil significantly increased its export to Europe. The strong demand for peasant products caused prices to rise enough for locals to enlarge their production. The boom brought a general rise in prices by 300% between 1904 and 1920 in Southern Nigeria (Ofonagoro 1979, p. 651).Footnote 7 Rapidly rising prices at the ground level attracted a large number of small households to the export economy, besides landlords who had connections with foreign firms in coastal ports. In contrast with plantations, middlemen had to prepare a large quantity of cash to pay the peasants. Unlike the previous period, however, the better terms offered for trade persuaded local peasants to accept a new type of currency of a higher denomination instead of indigenous currencies. Endogenous currencies did not disappear as quickly as the colonial governments had expected, but the export boom, in conjunction with tax payments set in colonial currencies, supported the penetration of these currencies into local markets.Footnote 8

The rising demand for palm oil resulted from the development of the chemical oil industry in Europe. The demand for vegetable oils was spurred by their use as butter substitutes, for the production of soap, for lubricant for machineries, and for making bullets. Apparently, palm oil did not monopolize the supply and the import of other vegetal oil and seeds to Europe increased significantly. The prices of linseed from India and sesame oil from China also rose during the two decades before the First World War (Kuroda 2020, pp. 172–173).

Importantly, parallel stories appeared in societies other than Africa where peasant households dominated the economy. In China, until the end of the nineteenth century, copper cashes continued to mediate transactions in rural markets. Through two millennia of Chinese imperial history, some dynasties tried to issue multiple unit currencies instead of the mono unit currency, that is copper cashes, but no ruler could maintain the issue of multiple unit coins for long, since end-users did not accept them at the face value regulated by the authorities. At the turn of the twentieth century, provincial mints started to issue machine-made copper coins with no hole in the centre. They had multiple units of face value, mainly ten wen. In contrast with previous attempts, ten-wen copper coins were successfully circulated and gradually substituted for copper cashes. The degree of substitution differed from province to province. The successful provinces appeared to have been those with a growing export of peasant products from treaty ports. Not governmental force, but rather increasing prices for their products, persuaded peasants to accept a currency whose face value was larger than traditional currencies. Importantly, the most successful province, Hebei, was able to circulate paper money which was convertible with the new machine-made coins. The combination of paper money (mostly 1000 wen) and the new copper coin (10 wen) was able to respond to growing monetary demands by both dealers and peasants (Kuroda 2005).

Roughly during the two decades before the First World War, traditional currencies that mediated peasants’ transactions began to disappear and machine-made coins gradually proliferated among Asian peasants. In Indochina, under the French colonial government, Chinese style copper coins, dongs or sapèques, gradually decreased in circulation (Robequain 1939, pp. 155–157). In Thailand, under an independent monarchy, the use of cowries was discontinued (Ingram 1955, p. 149). Importantly, the transformation of endogenous atomic currencies into small denomination coins in the modern form accompanied the proliferation of paper money. In Indochina, the annual issue of Banque de l’Indochine (BIC) banknotes in francs doubled in amount between 1903 and 1913 (Gonjo 1993, p. 372). In Thailand, the annual issue of government paper money in baht increased seven times during the same period (Ingram 1955, p. 154). The increasing issue of provincial paper notes in Manchuria, China, in the early twentieth century was in response to the growth of soya bean and soya oil exports from Dalian. The seasonal appreciation of provincial notes in the harvest season indicates that the notes facilitated the absorption of a large quantity of soya beans from the inland (Kaminishi 2013).

Thus, the transition of currencies in Africa in the early twentieth century occurred along a global tide of monetary transformation penetrating to the ground level. Importantly, the change from traditional currencies to colonial ones in West Africa was set in motion by Africans engaging in exporting peasant products rather than European firms clinging to transactions in kind. For example, the Niger River Company did not conduct cash trade until around 1905. Rising prices for local products attracted African producers to accept new currencies in so far as they could expect more profits through the transformation. The profits to be made through exporting cash crops encouraged a large number of peasants to move seasonally to coastal regions, which increased the amount of groundnuts, cocoa, and oil palms planted (Hopkins 1973, pp. 223–225). We should not overlook that this trade boom for peasant products coincided with the emancipation of slaves inside Africa. According to an estimate in 1905 there were two million slaves in French West Africa, 300,000 of whom were freed in the period 1905–1907 and emancipation continued in both British and French West Africa through the early twentieth century. Increasing demand for agricultural labour must have encouraged ex-slaves to establish themselves as free farmers (Hopkins 1973, pp. 226–227).

Increasing export of peasant products brought European currencies into circulation with which Africans could purchase imported commodities and pay taxes. In British West Africa the annual circulation of British sterling silver increased from £24,426 on average between 1886 and 1890 to £666,190 between 1906 and 1910 (McPhee 1926, p. 236). In addition, in 1908, subsidiary nickel and aluminium coins with a face value of one penny and one-tenth of a penny were introduced to substitute for cowries and manillas in British West Africa (McPhee 1926, p. 238; Hopkins 1970, p. 124).

What happened across the world in the two decades before the First World War was that rising prices for peasant products, including those made in Asia and Africa, bridged denominations representing monetary usage in industrial countries and those reflecting monetary usage in peasant economies. Higher denominations in the former and lower denominations in the latter did not always reflect the differences in wealth between nations. Though it is difficult to measure, according to one study, the daily wage for unskilled labour in late nineteenth-century England was around thirty pence (Burnett 1969, pp. 250–253). If we can take the daily wage in Uganda of 100 cowries, as mentioned above, and roughly assume that 100 cowries were equivalent to one penny, the money earned by a day’s work in England could purchase what would take thirty days labour in Uganda. The exchange rate of a European currency against an indigenous one was far higher than the real value, reflecting that human labour lies at the base of the acceptability of currencies. One silver shilling worked as a fraction of numerous sterling pound currencies circulating across the British Empire, while shell money fulfilled its function within a marketplace zone after being dispersed among end-users. The difference of scope in circulation between the silver shilling and cowries rather than the difference in real purchasing capability by people in Britain and in West Africa caused the former to be overvalued against the latter when the two currencies were exchanged. Increasing prices of peasant products in terms of European money in the early twentieth century acted to bury the denomination gap between the currencies in metropolitan countries and those in other regions including colonies through the circulation of convertible currencies among peasants engaged in producing export products.

Eventually, however, convertible currencies newly introduced could not sufficiently supply what end-users in peasant economies demanded. For example, in spite of the continual import of French currencies, Dahomey always suffered from a chronic shortage of currency through the early twentieth century. Copper and lead coins with a face value of 10 cents, 5 cents, and 2 cents in particular were in extremely short supply. Thus in Ouidah and Abomey 50 centime silver coins were discounted by 5 and 10% in exchange for 10 centime and 5 centime bronze coins (Manning 1982, p. 160). According to the official estimate in 1909, 1,100,000 francs worth of 50 centime pieces and 500,000 francs worth of 5 and 10 centime pieces were in the hands of the public (Manning 1982, p. 312). The dominance of the smallest denomination currency also appeared in Kenya and Uganda where the one-cent rupee coin was nearly half in value of colonial coins in circulation in 1921 (Pallaver 2018, p. 320).

Inevitably indigenous currencies still supplied what African societies demanded. Typically, cowries remained as the main currency in some areas such as Aplahoué in Benin. In 1908, to make tax payments in francs possible, the French administration had to give locals a few extra weeks to go to towns outside the region to exchange cowries for francs (Manning 1982, p. 158). The prohibition of importing manillas to eastern Nigeria in 1902 worked opposite to the official intention of abolishing manillas by stabilizing their value among end-users and leaving them still circulating until the Second World War (Naanen 1993).

Keeping in mind the shared trends examined in this section, we should note that there was an important difference between the convertible currencies of Asia and Africa. In Africa, a certain portion of paper money already circulated in the early twentieth century. For example, the Banque de l’Afrique Occidentale (BAO) was estimated to have provided 2.4 million francs in banknotes in 1909. Manning thought that the amount accounted for some 15% of the entire money supply in French Africa (Manning 1982, p. 161). The introduction of paper money in Africa was not later than in Asia, except for China and surrounding countries which had the earliest issue of paper money in the world. Thirty years later, however, the proportion did not necessarily rise significantly in Africa. This contrasts with the situation in Asia, as we will see in the next section.

Global Depression Revitalized Monetary Locality

During and after the First World War, the economies of exporting peasant products in Asia and Africa had a similar outlook. The loss of the European market for the products during the war caused peasant households in Asia and Africa to suspend the cultivation of crops for export and switch to other crops according to profitability. Meanwhile, the contraction of imports from Europe encouraged local entrepreneurs to try a substitute industrialization across Asia and Africa. The boom for peasant products soon after the resumption of European trade was followed by a slump at the beginning of the 1920s. Towards the late 1920s, however, international trade in peasant products from Asia and Africa continued to grow.

Throughout the 1920s, in both Asia and Africa, there was a surge in credit supply to businesses exporting peasant products. For example, in Cochin-China, the Banque de l’Indochine increased lending to mutual credit associations from 99,642 piasters in 1913 to 15,108,778 piasters in 1930. Credit supply supported landowners to produce rice for export (Robequain 1939, pp. 193–195). In Ghana, the Kumasi branch of the Bank of British West Africa increased advances to European cocoa exporters from £106,187 in 1920–1921 to £411,927 in 1926–1927. European merchants gave advances to local cocoa brokers who collected cocoa from peasants (Austin 2005, pp. 283–284). However, the rapid growth in lending resulted in a series of defaults and credit crunches across the world, which became a part of the buildup to the Great Depression of 1929.

The global depression quickly affected the prices of peasant products from Asia and Africa in metropolitan countries. Table 11.1 shows that the price of rice from Saigon and the price of palm kernels from West Africa in London both decreased from 1930, falling by around 75% between 1929 and 1934. A common strategy in response to the falling prices of export products by those who involved in the export economy across Asia and Africa was to increase production. The export of palm kernels from French West Africa fell greatly in value from 64,527 thousand francs in 1930 to 27,884 thousand francs in 1935, while exports in volume slightly increased from 51,701 tons to 61,123 tons during the same period (Manning 1982, p. 382). In Southern Central Cameroon, cocoa planting had a peak in both total area and the number of plots between 1932 and 1935 (Guyer 1981, p. 76). There was a very similar trend in the Asian export peasant economy. The export of rice from French Indochina significantly decreased in value from 1198 million francs in 1930 to 665 million francs in 1935, while the volume of exports rose from 1121 thousand tons to 1765 thousand tons in 1935 (Touzet 1939, p. 244).

Table 11.1 Prices of Rice from Saigon and Palm Kernels from West Africa in London 1927–1935 (average in July of each year)
Full size table

The decrease in the value of exports meant a contraction of the monetary flow into the countries relying on exporting products. Maintaining the volume of exports implies that, in spite of unfavourable conditions, peasants did not switch their crops to others not aimed at foreign markets, but continued to produce export crops. Those statistics suggest that the export economy had already taken root very deeply in the colonies.

The shrinkage of the trade surplus due to the fall in the value of exports directly affected the issue of currency for colonies. The issue of banknotes greatly decreased after 1930 both in French West Africa and French Indochina, as shown in Table 11.2. Together with Table 11.1, we can see that the amount of banknotes put into circulation by Banque de l’Afrique Occidentale followed the price of palm kernels and that by Banque de l’Indochine chased the price of rice. Confirming a common trend, however, we should note that the slump in the circulation of banknotes by Banque de l’Afrique Occidentale was far greater than that by Banque de l’Indochine during 1932 and 1935.

Table 11.2 The amount of banknotes in circulation: French West Africa and French Indochina
Full size table

Conditions in Africa may have been different from Asia. The fall in maritime trade brought European colonial administrators in Africa a serious decrease in revenue from maritime customs. The contraction of customs duties forced colonial governments in Africa to depend more on direct taxes imposed on peasants, and consequently more currency was retrieved from households by the authorities. Manning understood that the decrease of official currencies through direct taxation caused a significant shortage of money in 1933–1936, the only period under colonization that direct taxation exceeded customs duties (Manning 1982, p. 254). However, we should rather pay attention to the unpopularity of banknotes in Africa rather than to the decrease of money in circulation in general.

Before the First World War, across the world, banknotes could basically only be issued with a full reserve of hard currency plus sound negotiable securities. The issue of paper monies in Asia and Africa could not deviate from this principle at that time. However, after the First World War, the issue of paper money in some Asian countries increased without a sufficient reserve. In the case of Banque de l’Indochine the rate of note reserve fell from 70% in 1910 to 9% in 1930 (Gonjo 1993, pp. 374–375). The proportion of coins against notes in circulation decreased with the increased issue of banknotes. In 1930, the French Indochina authority possessed 12 million piasters of silver, an amount thought to be equivalent to 10% of the amount of BIC banknotes in circulation (Touzet 1939, p. 118).

The proportion of coins against notes in colonial West Africa was quite different from that in Asia. A survey on 30 June, 1935, showed the breakdown of currency circulation in Nigeria as follows: West African silver coins £1,348,318; West African alloy coins £7,276,567; West African nickel-bronze coins £653,064; West African currency notes £717,295 (Nigeria. Chief Secretary's Office 1936, p. 73). Considering the large denomination gap, we can see how large quantity of small currencies was in circulation and what a small part paper money played in the entire monetary supply. According to another study, the proportion of notes in the estimated entire currency circulation in Nigeria changed as follows: 15% in 1926, 11% in 1936, and 18% in 1946 (Hawkins 1958, p. 347).

Not only was the proportion of notes against the amount of the entire monetary supply small in West Africa, but notes were often discounted against their face value in inland trade. It is considered that one reason paper money was not popular in West Africa was that the Currency Board failed to guarantee the convertibility of notes into coins on demand (Hawkins 1958, p. 346). However, whether a note is convertible is one thing, but whether the note is acceptable is another. For example, through the early twentieth century, banknotes proliferated in China often with discounts against their face values. In 1920, Banque de l’Indochine issued notes whose face value was 10 cents, 20 cents, and 50 cents. However, they were so unpopular that they were discounted against metallic currencies. In 1921–1922 French Indochina authorities minted a large number of 10 cent and 20 cent silver coins, withdrawing the small denomination banknotes (Touzet 1939, pp. 63–64).

Importantly, the increasing popularity of paper money in Asia did not mean that paper money could substitute for small denomination currency in transactions among end-users. Falling prices during the Great Depression revealed a strong demand for indigenous fractional currencies. In Tonkin, two kinds of copper coin called sapèques, equivalent to a half cent and a sixth cent in French francs, circulated even in the 1930s. One piaster was equivalent to 195 sapèques according to the order of the Annan emperor in 1934. That means that one indigenous copper coin was quoted at almost half a cent. That year, the French Indochina authority finally began to issue half-cent coins (Touzet 1939, p. 3).

West Africa also suffered from a chronic scarcity of small denomination currencies. Locals had no hesitation about reviving unearthed cowries as active currencies in the 1930s (Manning 1982, p. 254). In 1933 in the Idere area, 150 cowries were used in exchange for one penny (Hodder and Ukwu 1969, p. 47). Even in the early 1940s, the British authorities put a large quantity of tenth-pence nickel coins into the district where manillas were used in order to drive them out. The demand for the lowest denomination nickel coin was so strong that sometimes 11 pence of nickel coins were exchanged for the alloy shilling of 12 pence (Naanen 1993, p. 433). Thus, independent circulation at the ground level of fractional currencies, including traditional ones such as sapèques and cowries, occurred both in Indochina and West Africa, although the demand for infinitesimal-scale currency was stronger in West Africa than in Indochina.

Let us consider other traditional currencies that kept their popularity in West Africa. When, in 1948–1949, the British colonial government redeemed the manillas, 19% of the manillas stock came from the Ndoki district, which had been a centre for collecting palm oil. The positive relationship between the proliferation of manillas and the export of palm oil is quite apparent (Naanen 1993, p. 432). The redemption of manillas in 1940–1941 cost Britain a total of £401,135. Britain only got back £135,000 when they were melted (Naanen 1993, p. 445). We can see how highly the brass items, manillas, were overvalued when functioning as currency among local end-users.

However, manillas did not always function as fractional currency in the 1930s and 1940s. The Okpoho manillas, the commonest manillas in circulation, was redeemed at an exchange rate of 80 to £1 in 1948 (United Africa Company 1949, p. 54). One Okpoho manillalas was valued at three pence, not a small amount for locals. It might not have been just a coincidence that in the region in Gambia that used cloth currencies, the basic strip, wala wala, was valued at three pence as well. The two units of cloth currency most frequently used among the Wolof were valued at 6 pence and four shillings (Ames 1962, p. 38). Importantly, the manillas and cloth currencies worked as a unit of account in expensive transactions among locals. The British observed that people preferred the payment of dowry by manillas in the 1930s through the 1940s (Guyer 2004, p. 58). Colonial governments in West Africa could not substitute paper money for endogenous currencies that functioned as a device for accumulating assets to the extent that the government in Indochina did.

The decline of exports of peasant products in value during the Depression must have worked to reduce intra-African trade of such commodities as cattle and kola nuts (Hopkins 1973, p. 253). The contraction reduced the purchase of African goods through inland trade within Africa. Though it is difficult to measure the size of inland trade, there is a possibility that the loss of income from exports caused inland transactions at a distance to be in a slump and consequently decreased the demand for convertible currencies, such as banknotes by the BAO, which were available to settle long-distance trade in West Africa. From the viewpoint of complementarity among monies, the decrease in the supply of convertible currencies available for interregional trade as well local exchanges gave indigenous currencies the opportunity to revive in local exchange to the extent that they worked as a measure of accumulating assets and a unit of account.

Conclusion

Economic and cultural frameworks are indeed inextricably connected (Green 2019, p. 470). Among the Wolof in Gambia in the early twentieth century, cloth currencies were indispensable for fulfilling Muslim rituals (Ames 1962, p. 38). In the Kingdom of Kongo in the late eighteenth century, cloth currencies were used to wrap the corpses of the elite at their funerals. Among the Igbo in Nigeria in the early twentieth century, 900 manillas were required to join a powerful secret society (Green 2019, pp. 19–21). It is easy to imagine that banknotes could not substitute for manillas for those purposes, even though both the notes and brass items were actually made in Europe.

However, a variety of money does not always result from differences in culture. It is also important to examine if exchanges mediated by different currencies are fungible activities. Whether a particular money is indispensable for a local custom is one thing, but whether the money is available only within a locality is another. Keeping in mind the relationship between exchanges and social actions, we should note that exchanges have distinctions of themselves. We should pay especial attention to whether sellers and buyers can directly and multilaterally negotiate, which differentiates proximate exchange from long-distance exchange, and also whether sellers and buyers can anonymously make deals, which distinguishes cash transactions from deferred payments. Global history shows us that a variety of exchanges resulted in the complementarity among monies (Kuroda 2020).

A triangle dependence among small-size peasant households, proliferation of the marketplace, and preference for indivisible fractional-value currencies was shared by Africa and Asia. The investigations of peasant households in early twentieth-century China revealed that peasants changed crops according to the price movement at the marketplace where they could directly negotiate prices, and that monetary demand/supply at the marketplace remained independent from that for long-distance trade (Kuroda 2018). Peasants and petit dealers in West Africa appeared to conduct business in the same way. The incorporation of the peasant economy into international trade stimulated peasant households to accept new currencies, transferring profits to small holders in the early twentieth century. However, unlike the case of plantations, the export economy depending on metropolitan countries did not succeed in controlling peasant households at the ground level either in Africa or Asia. That is why cowries and manillas revived or survived in West Africa and local notes supplemented the shortage of currencies in China during the Great Depression (Kuroda 2020, pp. 189–190).

On the other hand, the different degree in the proportion of paper money in circulation between Africa and Asia gives an insight relating to the dissimilarity in the pattern of complementary relationship among monies between Africa and Asia. Though not yet examined sufficiently, how networks were organized for interregional exchange inland might have been different between the two. As far as the spread of banknotes outside trade ports in 1920s China was concerned, the acceptance of banknotes by native financiers in local towns increased the importance of banknotes in transactions (Kuroda 2020, pp. 178–179). A lower circulation of banknotes in West Africa in the early twentieth century may reflect the lesser importance of intermediary markets functioning as a bridge between trade ports and rural marketplaces.

The dissimilarity in money on the surface might have resulted from differences in the organization of exchanges at the lower level of markets. We should understand how differently exchanges are organized according to society and period before reducing the variety of money within the scope of embedded economy with culture.