1 Introduction

Having announced it twice one month earlier, The Ceylon Observer – the most important periodical in this British colony – finally published a long article by Robert Boyd Tytler on 4 October 1852.¹ Despite his young age, many considered Tytler to be one of the founding fathers of Ceylon’s coffee economy. Born in Scotland, he moved to the East Indies in 1837, at the age of eighteen, after doing a three-year internship as a management assistant on coffee plantations in the Blue Mountains, Jamaica. In the following years, along with other planters and managers with previous experience in the West Indies, Tytler would help lay the groundwork for Ceylon’s coffee economy by transmitting and locally adapting the agronomic knowledge that had originally been developed for Caribbean slave coffee plantations.²

Tytler’s 1852 article, however, dealt with a different slave society. The prospects for Ceylon’s coffee culture were still uncertain in the aftermath of the deep crisis of 1847–1848. Many estates founded during the investment boom between 1837 and 1845 had been abandoned, confiscated for debt or sold significantly below the original invested value. Based on a careful assessment of the fixed operating costs and average productivity of Ceylon’s plantations, Tytler pointed out how a fluctuation in coffee prices of ten shillings (up or down) could lead to the recovery of the industry or its definitive demise. The remarkably low prices of 1847–1848 were clearly “the natural result of production exceeding consumption.” The crucial information that coffee investors in Ceylon needed, therefore, depended on the perspectives of the world coffee market, which would, in turn, be determined by the expansion or stagnation of supply in the following years. It was necessary to study rival producers, whose output was still much higher than Ceylon’s at the time. According to Tytler, “with respect to Java perhaps we have little to fear or to hope for,” given the inelasticity of its production under the Cultivation System. The real problem was elsewhere: “The great arbitress … of this vital question, as she is the present overruling mistress of the Coffee Markets of the world, is Brazil with her slave labour.”³

The main objective of Tytler’s article was to start a discussion within the Ceylon commercial and planting community on creating a fund (with quotas of ten pounds each) to finance a voyage to Brazil “with the general purpose of ascertaining the actual state of affairs in that large producing country in respect of its capability of extension of cultivation and the chance it presents of holding out in the competition for supplying the market of the world with Coffee.” This investigation of Brazilian coffee culture would have to concentrate on the problem of slavery. It was not clear for Tytler if “our free labour in Ceylon” would be able to compete “with that of athletic slaves who, under the lash, are made to furnish more than double the daily quota of work.” The main issue was the cost of labour. Wages in Ceylon were low due to the abundant supply of seasonal workers from South India. Brazil could only maintain its position in the world coffee market if slaves remained relatively cheap in the country. News that the Empire of Brazil had ended the transatlantic slave trade two years earlier was reaching Ceylon through London. But, considering the history of the previous decades, when the transatlantic slave trade continued to be illegally carried out despite its formal prohibition, was the enforcement of the 1850 law to be trusted? Would the emperor and the Brazilian government be effectively committed to suppressing the traffic? In Cuba, a large number of slaves were relocated from coffee plantations to sugar plantations: could the same happen in Brazil after the end of the transatlantic slave trade, with an accompanying drop in coffee production?⁴

The seasonality of coffee production ensured that demand for labour was significantly reduced for much of the agricultural year. For this reason, unless coffee planting was combined with the production of other tropical staples, slavery was not particularly suited to coffee in Tytler’s evaluation. Brazilian planters, he argued, were perhaps making the mistake of expanding their coffee plantations too much without having a proportional labour supply. That strategy was driving Brazilian planters to force their slaves to harvest indiscriminately and simultaneously “all sorts [of coffee berries], whether ripe or unripe.” Hence, Brazilian coffee was priced at a much lower level on international markets. Without the transatlantic slave trade and with an estimated annual slave mortality of 10 per cent, planters “must in the end, and at no very remote date, bring themselves to a dead lock.” It was this hypothesis that Tytler wanted to verify on the ground in Brazil.

Tytler’s proposal lost its raison d’être almost immediately. International coffee prices had reached their lowest level in the entire nineteenth century precisely in 1847–1848. Despite some occasional fluctuations, the next thirty years would be marked by a sharp rise in prices,⁵ boosting the “golden age” of Ceylon’s coffee economy. But even though his plan to inspect what was happening with slavery in Brazil was not implemented, the basic premise of Tytler’s article – the existence of structural relations between the coffee economy in Brazil, based on slave labour, and its Ceylon counterpart, based on a peculiar form of wage labour – was clearly correct.

These relations, however, have not been properly investigated yet. The historiography on nineteenth-century Ceylon has focused on the annual migratory flows of Tamil coolies from South India, the nature of their wage labour regime (free or coerced), their living conditions on plantations, the dual or integrated character of the plantation economy in its relations with the local peasant economy, and the ecological origins of the coffee crisis in the 1870s–1880s. Connections between the experience of slave plantations in the West Indies and post-slave plantations in the East Indies have recently attracted the attention of scholars, but research efforts in this field are still rare.⁶ The historiography on slavery in the nineteenth-century Brazilian coffee sector is massive and details its multiple economic, demographic, social and cultural dimensions. Nevertheless, the interconnections between Brazil and the other coffee zones of the world – whether slave-based or not – remain under-examined.⁷

My aim is to analyse how the different types of compulsory labour used in the nineteenth-century capitalist world economy for the production of coffee were related to and conditioned by each other. Thus, my goal is to take up the imperative articulated by Marcel van der Linden, a scholar in the field of global labour history, according to which “we should not study the different kinds of subaltern workers separately, but consider the connections between them as much as possible.”⁸ The present chapter, however, incorporates formal comparisons and connections into a broader integrated approach – which brings a substantive comparison and not just a typological one⁹ – that shows how slavery in Brazil and the kangany system in Ceylon were formed by a double bundle of forces, which united and differentiated them. On the one hand, labour relations in these two spaces were deeply shaped by the global anti-slavery policies promoted by the British Empire after the end of the Napoleonic Wars. On the other hand, although coercive labour was the backbone of coffee economies in both spaces, their distinctive labour arrangements and inscription in the tropical commodity markets of the industrial capitalist world economy ultimately subordinated the fate of Ceylon to that of Brazil.

The chapter is divided into three parts. In the first part, I examine the making of the coffee economy in Brazil and in Ceylon in the first half of the nineteenth century, demonstrating how the joint forces of the industrial world economy and British anti-slavery shaped local processes and were shaped by them. In the second part, I compare the labour arrangements adopted in coffee plantations in Brazil and Ceylon between the 1850s and 1870s, focusing on the mechanisms of landscape and labour management. Finally, in the conclusion I discuss the implementation of Tytler’s plan in 1876, when the Ceylon government sent the journalist G.A. Crüwell on an inspection trip to Brazil. At the same time, they mobilised A. Scott Blacklaw, who had been working as a superintendent on a coffee plantation in Brazil since 1873, to collect additional information. In 1878, the book Brazil as a Coffee – Growing Country was published in Colombo, Ceylon, co-authored by Crüwell and Blacklaw. However, the contexts of Tytler’s 1852 proposal and Crüwell’s 1876 journey were quite different. The observations made by Crüwell and Blacklaw between 1876 and 1878 help us understand the specific nature of the simultaneous crises of slavery (but not of the coffee economy) in Brazil and the coffee economy (but not of the kangany system) in Ceylon.

1 The Making of the Coffee Economy in Brazil and Ceylon, 1815–1850

The first global coffee complex was part of the Ottoman economy and arose at the turn of the seventeenth century, when peasant coffee production in Yemen was connected to the large urban consuming centres of the Ottoman Empire in the eastern Mediterranean. Through a process of emulation prompted by early Orientalism, the consumption of the beverage became increasingly popular in the great urban centres of northwestern Europe at the end of the seventeenth century. The pressures of a growing Western demand on Yemen’s limited supply caused coffee prices to explode in the early eighteenth century. For that reason, the colonial powers of northwestern Europe sought to appropriate the secrets of coffee production, acclimatising the tree to the conditions in their own Caribbean slave colonies. The second global coffee complex took off after the 1720s and was structured along the axis of Caribbean slave colonies/urban consumers in Northern Europe. In the second half of the eighteenth century, the volume of slave-based coffee production financed by European capital grew at an unprecedented pace, with the French colony of Saint-Domingue occupying the most prominent place. Between 1786 and 1790, its average annual coffee exports were around thirty-two thousand metric tons, which was then equivalent to around half of the world’s supply.¹⁰

The Age of Revolutions completely changed the coffee commodity chain. Despite the efforts of Touissant Louverture to recover the plantation economy after the slave revolution of 1791 and the abolition of slavery in 1794, in 1801 Saint-Domingue’s coffee exports fell to about two-thirds of what they had been in 1790.¹¹ The Napoleonic invasion in 1802 and the Haitian War of Independence fought over the following two years created an even greater vacuum in the world coffee supply by destroying much of the coffee infrastructure and disrupting the island’s connections with the North American merchants that re-exported the product to Europe. The opportunities opened by the disruption of Saint-Domingue were immediately seized upon by rival Caribbean slave colonies. Jamaica had partially occupied the space opened by the Saint-Domingue Revolution since the 1790s; after 1804, it was accompanied by Cuba, which was also dependent on slave labour.¹² In 1815, Cuba and Jamaica sent a total of around twenty thousand tons of coffee to the world market, which indicates that they had still not been able to fully fill the vacuum created by revolutionary processes in the French Caribbean.¹³ The dysfunction in and transformations of the world coffee market between the 1790s and the 1800s became evident with the return to peace in 1815, when coffee prices rose to a new high.¹⁴ In the following decades, coffee consumption in the urban centres of the North Atlantic would gain an additional and decisive impulse with the English Industrial Revolution and economic growth in the United States and Continental Europe.¹⁵

Britain’s undisputed victory at the end of the Napoleonic Wars, with the complete dominance of the seas, placed it in a strong position to reshape the European role in overseas territories. The Dutch possessions are a case in point. In 1795, the new Batavian Republic became an ally of the Republic of France. As a consequence, the British conquered old voc (Vereenigde Oostindische Compagnie – the Dutch East India Company) domains in Asia, such as the Cape Colony and Ceylon in 1796, and the island of Java in 1811. Demerara, Essequibo, Berbice and Surinam were, in turn, conquered in 1804 in the West Indies. At the Congress of Vienna, the Netherlands was seen by Britain as a key actor when it came to containing France’s future revolutionary power; for this reason, the British representatives made sure to partially restore the overseas domains of the new United Kingdom of the Netherlands (Holland plus Belgium), returning Java and Surinam to it. Britain, however, retained Cape Colony and Ceylon, converting the second into a Crown Colony, and forcing the Dutch to ban the African slave trade in their colonial possessions. In Ceylon – the colonial space that interests us in this chapter – the years following the confirmation of British rule over its coast would be marked by the conquest of the Kingdom of Kandy in the interior of the island (1815–1818).¹⁶

Britain also played a major role in the redesign of the Portuguese Empire. In November 1807, after the Napoleonic invasion of the Iberian Peninsula, the Portuguese royal family fled to Brazil escorted by a British fleet. One of the first measures taken by the Portuguese Crown when settling in the New World was to open Brazilian ports to the so-called “Friendly Nations,” which, at the time, meant Great Britain. In 1808, Brazilian slaveholders got free access to the world market, without the intermediation of Portugal’s previous mercantilist policy. The 1810 Trade Treaty, which granted a series of benefits to British traders who operated in Brazilian ports, consolidated the new framework of close association between the Brazilian slave economy and the British mercantile power, crucial for its exports. On the issue of the transatlantic slave trade, however, Great Britain did not achieve the same success with the Portuguese Crown as it did with the Dutch. British manoeuvres taken at the Congress of Vienna to forbid the transatlantic slave trade to Brazil failed and, as an expression of the strength of slavery within the Portuguese Empire, Brazil was turned into a United Kingdom with Portugal and Algarve in 1816, with its headquarters located in Rio de Janeiro instead of Lisbon.¹⁷

Although spatially separated and very distinct from each other, these events in Ceylon and Brazil were part of a unified process of transformation of the European imperial structures during the Age of Revolutions. At the end of the Napoleonic Wars, Britain had consolidated itself as the hegemonic power of the capitalist world economy. And, for reasons that are beyond the scope of the present chapter, Britain started presenting an anti-slavery ideology – as a constitutive part of its imperial strategy – that implied an endless fight against the transatlantic slave trade at the international level and the abolition of slavery in its own West Indian possessions, followed by a looser policy of gradual abolition in its East Indian possessions.¹⁸

The making of the Brazilian slave-based coffee economy took advantage of the circumstances created by these revolutionary and inter-imperial conflicts at the turn of the nineteenth century. The establishment of the Portuguese royal family in the city of Rio de Janeiro played a key role in this. The Paraíba do Sul River Basin (the so-called Paraíba Valley), whose course runs from east to west paralleling to the Atlantic coast, was located in an arc within a maximum of two hundred kilometres distance from Rio de Janeiro’s port. Situated between the foothills of the Serra do Mar and the Serra da Mantiqueira, the lands of Vale do Paraíba were, at the beginning of the nineteenth century, entirely covered by the tropical Atlantic Forest and with ideal altitude and climatic conditions for growing coffee.¹⁹ At the time, Rio de Janeiro was one of the most thriving Atlantic commercial ports, with a commercial infrastructure that had been established in previous centuries for sugar and gold exports, among other products. Most importantly, with its transformation into the capital of the Portuguese Empire, Rio de Janeiro became the largest port of disembarkation for enslaved Africans in the Americas. The ample supply of local capital, merchant facilities and cheap slave labour could easily be mobilised to take advantage of the favourable post-1815 climate for coffee production, with the rapid establishment of plantations in the virgin soils of the Paraíba Valley. The growth of coffee exports from Rio de Janeiro was overwhelming. In the first half of the 1810s, the export of coffee amounted to 1,500 tons per year; in 1821, on the eve of Brazilian independence, these exports were already equivalent to those of Cuba: thirteen thousand tons per year.²⁰

In the estimates of one economist, “between 1821 and 1850 the trend rate of growth of coffee exports was 8.47 per cent per annum.”²¹ In the early 1830s, the Empire of Brazil alone would reach the total volume of world coffee exports of 1790, when the Saint-Domingue Revolution started. In the last five years of the transatlantic slave trade (1845–1850), coffee exports from Brazil amounted to c. 125,000 tons per year, which made the country responsible for half of the world’s supply. The Brazilian coffee economy, moreover, was now closely linked to a new consumer market, in addition to the traditional urban markets of continental Europe that continued to grow. Since the 1790s, as a neutral power in the imperial conflicts of the period, the United States had established itself as an important player in the re-exports of coffee from the Caribbean, which simultaneously allowed the creation of a solid base of coffee consumers in the country itself. However, the decisive event for the popularisation of coffee in the US was the abolition of import tariffs for the product in 1833. In the middle of the nineteenth century, driven by population growth and the so-called “Market Revolution,” the US was responsible, by itself, for about a quarter of the world’s coffee imports, 90 per cent of which came from a single supplier, Brazil.²²

Perhaps the most important piece of data to understand what happened to the world coffee market in the first half of the nineteenth century is the downward trend in international prices between 1823 and 1848. This downward trend was not uniform: concerning the New York prices, for instance, there was a very strong fall between 1821 and 1830; a short recovery between 1832 and 1836; a soft fall between 1837 and 1840; and another sharp fall between 1841 and 1847.²³ These movements suggest how slave production in the Paraíba Valley was transforming the world coffee market by allowing the popularisation of coffee consumption through oversupply. The spread of coffee consumption among urban and rural workers in the United States and the growing urban proletariat of continental Europe reflected a broader pattern, characteristic of the third global coffee complex. The domestic consumption of coffee as part of the “invention of the breakfast,” when the luxury of the eighteenth-century poor became the basic necessity of the nineteenth-century proletarian, transformed the product into a “wage-food,” that is, an agricultural good, produced on industrial scale, that became part of the reproduction of the wage labour force.²⁴

The growth of coffee production in Brazil between the 1820s and 1840s took place against British anti-slavery pressure. After 1822, Great Britain established two conditions for recognising the newly independent country: first, that Brazil would not incorporate the Portuguese colony of Angola as a province of its new empire; second, that it would ban the transatlantic slave trade within three years. When ratifying the anti-slave trade treaty in 1827, Emperor D. Pedro I lost political support among Brazilian slaveholders, which greatly contributed to his fall in 1831. That same year, the Brazilian Parliament passed its own national law banning the transatlantic slave trade. As a result, the number of disembarked slaves significantly decreased between 1831 and 1834. However, due to the strong demand for labour in the new coffee plantations (a direct result of the US free tax measure of 1833), from 1835 onwards the illegal transatlantic slave trade recovered with an enormous strength. Despite all British naval pressure, between 1835 and 1850 almost 550,000 enslaved Africans were disembarked at the coastal ports that served the coffee production in the Paraíba Valley. This was only possible because the Paraíba Valley coffee planters were crucial actors for the institutional construction of the Brazilian Second Empire (1841–1889), which gave them all the necessary support to confront British diplomatic and naval pressure and to keep untouched their illegal slave property.²⁵

For the first time in the history of the world coffee economy, coffee producers were commanding a national state to serve all of their interests. During this period, the only coffee zone that managed to compete with slavery in Brazil was the Dutch colony of Java, whose exports were mounting to sixty thousand tons per year in the late 1840s. However, Dutch colonial authorities had to recreate and amplify the voc’s past modalities of forced crops, the so-called Cultivation System (see Elise van Nederveen Meerkerk’s chapter in this volume), in order to compete with Brazil. Cuba was well supplied by the transatlantic slave trade and had ample reserves of virgin land, but lost ground to Brazil in the global coffee markets and – as Tytler pointed out in his 1852 article – was compelled to abandon coffee production in favour of sugar. The abolition of slavery in the British Empire led to the collapse of coffee production in Jamaica. It was in this context that Ceylon started to build its coffee plantation economy at the turn of the 1840s.

Ceylon was separated from the British East India Company (eic) possessions early in the nineteenth century and, therefore, lost access to its vast financial resources. Established as a Crown Colony, Ceylon was of crucial strategic importance for the British Empire in Asia, but remained in deficit in its first decades. The economy of cinnamon, which had been previously explored by the Portuguese and the Dutch, was no longer profitable. In 1832–1833, as slavery was being abolished in the West Indies, London initiated a series of administrative and judicial reforms to make Ceylon a self-financing colony and, if possible, a lucrative one. The reformist impulse deepened in the following years, always obeying the objective of converting the island into a prosperous plantation colony along the lines of the West Indies and Mauritius, but without slavery.²⁶ From the beginning, coffee was conceived as a possible solution to the impasses of Ceylon’s colonial economy. The native Sinhalese population had long cultivated the shrub on a small, peasant scale for commercial purposes, and the post-1815 commercial stimuli led this peasant production to quadruple in the next twenty years. However, since it was not controlled by British capital and was relatively inelastic (in 1838, its amount was still only 1,860 tons), the “peasant way” for coffee production in Ceylon was not a viable option in the eyes of the colonial power.²⁷

The first issue to be tackled was the equalisation of the imperial tariffs that regulated the British market. Until 1835, coffee shipped from the East Indies paid 50 per cent more duty (9d. per pound weight²⁸) than the coffee coming from the West Indies (6d.). From 1836, as part of the extensive imperial reforms introduced in the midst of the abolition of African slavery in the Caribbean, these tariffs were equalised at 6d. Foreign coffee was now to be surcharged at a rate that was two-and-a-half times higher (i.e., 1s.3d.) than the British coffee imported from the West Indies or East Indies/Ceylon.²⁹ The second issue to be tackled was land. The highlands of the ancient Kingdom of Kandy, which until then had been exploited secondarily by the Sri Lankan peasant communities as a complement to the rice paddy cultivation in its lower zones, saw a first wave of privatisation promoted by British colonialism. Ceylon became one of the colonial laboratories for Wakefield’s thesis on the closure of the frontier – by suspending free land concessions – as a condition for the creation of commercial agriculture in open resource zones. Between 1838 and 1843, as an immediate response to the intra-imperial tariff equalisation, the abolition of slavery in the West Indies, and the rise in coffee prices within the protected metropolitan market, nearly 250,000 acres of land were sold by the British government to set up coffee plantations or for financial speculation by British investors.³⁰

Most of these investments, however, were made without specific technical expertise on how to produce plantation coffee. The geo-ecological and climatic conditions of the Central Province highlands, although adequate for growing coffee, were different from those of the Caribbean, especially regarding the rainfall regime. Even experienced coffee growers, such as Tytler, were still learning how to deal with these local conditions at the turn of the 1840s, at the same time that forests were being cleared for large-scale plantations. It was also necessary to create specific labour arrangements in view of the refusal of the native Sinhalese peasantry to engage in wage labour on the new plantations.³¹ The solution was finally found in the lower Tamil castes of Southern India, who had long sought temporary jobs in the coastal areas of Ceylon. Attracted by the wages offered in the Central Province highlands during the first wave of coffee investments (which were comparatively low compared to other plantation zones in the British Empire, but much higher than present-day payments in India), these coolie workers, organised in gangs coming from the same villages, appointed a foreman from their own ranks – the kangany, a Tamil word that refers to an agricultural supervisor – to negotiate the specific labour arrangements for harvesting the coffee with the British planters or their superintendents, all of them (the coolies and their kangany) going back to South India at the end of the picking season.³²

This vast mobilisation of land, labour and capital took place in less than a decade (1837–1845). It should be remembered, however, that newly planted coffee trees take from three to five years to enter full production. Precisely when the viability of these investments was to be tested, customs duties were changed in Britain. In 1842, British taxes for its colonial coffee were reduced from 6d. to 4d. per pound weight, but those of foreign coffee were reduced from 1s.3d. to 8d. – that is, practically a cut by half. The worst was yet to come: in 1844, tariffs on foreign coffee dropped to 6d., but the tariff for colonial coffee was kept at 4d.³³ This progressive tariff equalisation was a direct development of the broader free trade policy that Britain was adopting in those years. Mainly focused on the sugar and wheat markets, this policy also ended up affecting coffee, a good with a relatively small market in the metropole due to the dominance of tea. What was happening with coffee had little impact on the metropolitan economy, but its impact on Ceylon was enormous. As the British market was no longer protected, imperial and international prices tended to converge³⁴ right when the Brazilian oversupply was imposing a huge drop in world coffee prices (1841–1847).

Only a decade old, the coffee plantation economy in Ceylon was already at a crossroads. The problem was not only the fall in world prices, but also how to recruit labourers. In 1847 and 1848, two official inquiries – one local, the other imperial – carefully discussed the impasses of Ceylon’s coffee culture.

The first one looked at the conditions of the annual migration and the work performed by the Tamil coolies in the highlands of Ceylon. The vast inquiry carried out in the early months of 1847 by James Emerson Tennent, Ceylon’s governor, sought to understand why the flow of coolies had suddenly and substantially decreased in 1846. Despite the almost unanimous denial of ill treatment of the coolies by planters and their superintendents, what emerged from the testimonies was an abundant documentation of practices such as whipping, arbitrary arrests within the plantations, and the suspension of payments – or even non-payment – to workers. Coolies staying in India were voting with their feet by refusing to come back to Ceylon for the next harvest. In a few testimonies, the metric for judging what was happening to the coolies in the Ceylon coffee plantations was given by the recently abolished black slavery in the West Indies.³⁵ The inquiry, however, remained unfinished, most likely because the coffee crisis in those years decreased the demand for labour, thus minimising the supply restriction. In the end, coffee planters (who, at one point, asked for firmer action from the colonial government to guarantee the continuous flow of coolies in the future) and authorities (who replied that Ceylon did not have a plantation slavery past like the West Indies and Mauritius did, and therefore, London owed them nothing) seemed to agree that the regulation of the annual Tamil migration as well as discipline within the plantations should be exclusively left to market forces. Nor did metropolitan abolitionists worry about what was going on in Ceylon: since it was not indentured labour, the work of coolies on the coffee plantations never came to their attention.³⁶

The anti-slavery movement was, in turn, crucial for the debates surrounding the imperial inquiry. The free trade policy adopted by London in the 1840s lowered sugar and coffee prices in the British market, as it was opened to the slave product of Cuba and Brazil. Amid the so-called “Mighty Experiment” of British emancipation, the former slave colonies of the Empire were now in the midst of a deep economic crisis. In the early months of 1848, a Select Committee of the British Parliament thoroughly investigated the problem. The free trade advocates – anti-slavery or not – saw the equalisation of tariffs as the final test for the success of the abolition experiment: under the rules of free trade, would free labour be able to prove its superiority over slave labour, as the abolitionists thought it would? Or would abolition finally prove that it had been a misconceived experiment from its very beginnings? The planter and merchant interests in the former slave colonies dominated the depositions to the Committee. Throughout the survey, sugar attracted much more attention than coffee. Still, all testimonies related to coffee had agreed that Brazilian slavery was at the heart of the post-abolition coffee crisis in Jamaica and Demerara, threatening to do the same to the recently built coffee plantation economy in Ceylon.³⁷ The Committee’s final proposal, which was in favour of returning to previous protectionist tariffs, was not taken up by the Russell Cabinet. Nonetheless, the material presented to Parliament between February and May 1848 had a direct impact on British foreign policy. The campaign against the transatlantic slave trade that the Foreign Secretary Lord Palmerston promoted in the following years was an explicit official response to the British sugar and coffee colonial lobby. In 1849–1850, the British Navy, after decades of fighting the slave trade in African waters or on the high seas, launched a frontal attack against Brazil by capturing or sinking at least ten slave ships in Brazilian territorial waters, and even threatening to directly bombard its ports. In 1850, faced with the risk of going to war with the greatest naval power in the world, the Empire of Brazil finally ended the secular – and infamous – transatlantic slave trade.³⁸

The constant fall in world prices between 1823 and 1848, a direct result of brazilian supply, helped popularise coffee consumption in the North Atlantic. After 1850, however, Brazilian coffee planters no longer counted on the unlimited supply of enslaved Africans. The Brazilian Empire’s commitment to ending the transatlantic slave trade, which Tytler proposed to examine in 1852, had become real and effective as early as 1851, largely because of a campaign in the British metropole that was supported by Ceylon’s coffee interests. Also, in 1851, British customs tariffs for colonial and foreign coffee were finally equalised (3d. per pound weight),³⁹ helping to increase the consumption of the beverage in a market that had been dominated by tea since the 1730s. Between the 1850s and 1870s, as part of a cyclical upswing of the capitalist world economy, international coffee prices would rise sharply, which boosted the definitive take-off of coffee production in Ceylon. In 1850, the British colony had exported about 17,000 tons of coffee. Two decades later, that number rose to 50,000 tons (45,000 tons of plantation coffee), closely approaching Java with its 70,000 tons per year. By then, the pace of coffee growth in Brazil had slowed significantly compared to what it had been between 1821 and 1850, but, even so, Brazil exported 208,800 tons to the world market in 1870.

2 Slavery and the Kangany System, 1850–1870

The patterns of labour that were created and the landscape management practices that were developed in the Paraíba Valley in the 1830s, and that quickly led Brazilian coffee to dominate the world supply, remained unchanged in the second half of the nineteenth century, despite the end of the transatlantic slave trade. In this section, I will briefly describe and contrast these patterns with the patterns of the Ceylon coffee economy.⁴⁰

Just like in all areas of tropical agriculture, the techniques of swidden agriculture for clearing the ground for coffee planting in the Paraíba Valley and in Central Ceylon were almost the same. In the undulating hilly topography of the Paraíba Valley (Mar de Morros de Meia-Laranja; literally, “sea of half-orange hills”), the planting of the seedlings was done in vertically aligned, well-spaced rows, from the base to the top of the hills. The unit of area measurement used in the Paraíba Valley was the alqueire (11.95 acres); usually, roughly four thousand coffee trees were planted per alqueire. Following the rules of the former slave coffee economy in the Caribbean, the trunks were cut to facilitate harvesting (known as “topping”) by slaves, but in Brazil the side branches of the coffee trees were rarely pruned. In the first three years after the trees had been planted, maize, beans and manioc were grown between the coffee rows, both to shade them and to facilitate weeding. Manure was never used in the Paraíba Valley. The initial productivity of coffee trees planted in frontier areas was very high (from ninety to one hundred arrobas⁴¹ per one thousand trees), but a few years later it was common for it to drop by one-third or even half. The productivity in the oldest coffee zones was in general around thirty arrobas per one thousand trees.

The size of plantations in the Paraíba Valley varied widely in terms of their spatial dimensions as well as in terms of their human scale, but available data indicates that the bulk of coffee production came from properties with over one hundred slaves and with two to three hundred alqueires. Comparatively, these properties were much larger than the slave-based coffee plantations in Saint-Domingue, Jamaica and Cuba. At the beginning of the exploitation of the Paraíba Valley plantations (on both small and large units), there was always much more land available for planting than was effectively used for cultivating coffee. This land structure was directly related to the landscape management strategy used there. Vertically aligned, well-spaced planting always implied rapid erosion: in about twenty years, not only would the coffee trees become unproductive, but the original soil cover would be completely lost. In response, planters periodically proceeded with new plantings on their properties; the internal landscape of the plantations thus came to resemble a patchwork quilt, with a somewhat chaotic combination of scattered coffee plantations of different ages and varying productivity, depleted land converted into pasture, and belts of diminishing forest reserves for the maintenance of the future productivity of the plantation. All of this resembled mining more than agriculture.⁴²

The harvest took place in the driest and coldest months of the year, between May and October. Given the vastness of the plantations and the large number of coffee trees allocated to each worker, the planters’ strategy was to extend the harvest for as long as possible, allowing the slaves to practice the so-called derriça – the quick harvesting of all the green, the ripe and the dry cherries at once (with leaves included), from the stem to the tip of the branch. Here, we can understand how landscape management was related to labour management through a specific spatial economy. As is always the case with slavery, the plantation labour force in the Paraíba Valley was quantitatively rigid for the whole year. Given the needs for ancillary services and the considerable number of children and elderly people on the plantations, not all slaves worked in the coffee fields. In general, field slaves comprised 50 to 60 per cent of the overall enslaved population in a given plantation. Nor did the slave owner have any means to dismiss his workers during the dead season of the agricultural cycle. The solution was to operate according to an equation which combined the production of foodstuffs during the coffee off-season (corn, beans and manioc interspersed in new coffee rows or in lower quality land, with the cultivation of rice in the lowlands in the small valleys), two to three annual weedings of the coffee fields, and overwork at the time of gathering coffee. Hence, the logic of extending the coffee harvest period with the use of the derriça resulted in an increasing allocation of coffee trees per slave. Compared to the Caribbean, these figures were impressive: while between one thousand to one thousand and five hundred coffee trees were usually attributed to each field slave in Saint-Domingue, Jamaica, and Cuba, this proportion more than doubled between 1830 and 1870 in the Paraíba Valley. In the second half of the century, it was not uncommon to find plantations that attributed four to five thousand coffee trees to a field slave, which also meant that these slaves would have to weed much more extensive areas, considering the large spacing between the trees.⁴³

The large spacing of coffee rows, which were vertically aligned on the hills, was informed by the wish to increase the capacity to watch and control slaves. Weeding was done in gangs under the command of a foreman; the fastest slaves were placed at the last rows in order to dictate the pace of the group; if a worker was slower than the others, he or she could easily be seen by the foreman at the bottom of the hill and, consequently, punished. During the harvest, there was a minimum individual daily harvest quota for each slave based on their individual capacities and the estimated total volume of the crop. If the slave did not meet the quota, he would be whipped; if he exceeded it, he would receive small monetary incentives. As a result of these schemes, field slaves had to individually harvest from three alqueires of coffee (here, an alqueire counted as a unit of volume, equal to sixty litres) per day in bad years to up to nine alqueires in good years. (Coffee trees planted without shade have strong annual variations in crops). It should be stressed again that the green, the ripe and the dry cherries would all be harvested at the same time. When taken to the plantation headquarters, the cherries were placed in the vast Brazilian drying platforms to dry the pulp and the parchment for up to two months, that is, until they were dry enough to be processed at the rough, but highly effective, pestle mills. The end result was low-quality, but very cheap coffee that was produced through a spatial economy focused on maximising the productivity of slave labour.

All contemporary reports – including those by A. Scott Blacklaw and G.A. Crüwell – noted the harsh daily disciplinary codes that made this brutal workload feasible. Given the absence of reliable agricultural statistics for nineteenth-century Brazil and the inner variations of the coffee agronomy, it is difficult to accurately establish the average coffee productivity per slave in the Paraíba Valley. Estimates that take the total slave population of the coffee districts – and not only the field workers – as a basis indicate an annual productivity of 1,000 to 1,200 kilograms of processed coffee per slave.⁴⁴ A detailed cliometric study focused on the specific productivity of field slaves in 1873, however, gives 2,280 kilograms of coffee per worker.⁴⁵

Ceylon’s coffee economy exhibited a clear continuity with the eighteenth-century Caribbean agronomy, but the consolidation of local agronomic techniques did not take place until the 1850s. The topography of the highlands of the ancient Kingdom of Kandy – the most suitable for coffee – was even more irregular than the Paraíba Valley and its climate was subject to the monsoon regime, which is much more humid. Here, the coffee trees were also vertically planted, but with trenches and drains in contour lines to mitigate the effects of erosion. The most common planting density used in Ceylon was c. 1,200 coffee trees per acre; in the Paraíba Valley it was c. 335. In addition to topping, the trees in Ceylon were constantly pruned because of the high planting density. In the early years of the shrubs’ growth, not only was nothing planted between the rows, but weeds were also taken out by hand. Subsequently, weeding was supposed to be carried out monthly. Another major difference in relation to the Paraíba Valley was manuring, which, in Ceylon, was often practised right from the early days of a new coffee plantation. Incidentally, the topic of manure may have been the most discussed subject of the agronomic publications in the British colony; by contrast, almost nothing was written about it in Brazilian coffee manuals. In spite of all these intensive agricultural treatments, coffee plantations in Ceylon had a remarkably low initial productivity compared to the Paraíba Valley. In Ceylon, productivity was calculated per acre, not per coffee tree. Converting the numbers, productivity in Ceylon was always between fifteen to twenty arrobas per one thousand trees, regardless of whether the plantations were on virgin soil or on terrain that had already been exploited.⁴⁶

The relatively small spatial scale of its plantations was one of the reasons that made intensive cultivation in Ceylon possible. As is almost always the case in coffee economies, there was a significant number of small coffee estates (i.e., under fifty acres) on the highlands of the Central Province, but the standard units were between one and two hundred acres (or 8.5 to 17 Brazilian alqueires as a unit of area). The allocation of the estate surface to coffee was overwhelming; once an estate got into full production, there were very few areas for grazing or forest reserves, and almost none for the production of foodstuffs. The basis for calculating the demand for labour was just under two coolies per acre, leaving six to eight hundred coffee trees per worker.

Intensive cultivation was also made possible by a vast supply of labour. After the 1847/1848 crisis was over, Southern Indian coolies returned to the Ceylon plantations in large numbers, this time through an arrangement directly induced by the planters or the superintendents called the “Coast Advance System,” but better known as the kangany system (a label that appears in the British official documentation).⁴⁷ This system differed from what had been practised before (1837 to 1847). From the 1850s on, the kangany would receive a cash advance from the coffee entrepreneur in Ceylon, which he would use to recruit the Tamil coolies in South India. With prior knowledge of the communities and the local power hierarchies in India (being a member of those same communities or not), the kangany would search for lower caste coolies already enmeshed in indebtedness relations with native landlords. The higher the local demographic pressure and the more acute the periodic cycle of hunger was, the easier it would be to recruit the coolies. Attracted by money advances offered by the kanganies, the coolies would go to Ceylon as a gang (consisting of twenty to fifty coolies, overwhelmingly men of the same castes), temporarily leaving their families behind. The long journey from the peasant community in India to the plantation where they would work in Ceylon would take 15 to 30 days of walking (interspersed with a short boat crossing of no more than a day). Throughout the journey, the coolie would increase his debt to the kangany in exchange for food (exclusively rice). For every worker, the kangany would earn a commission upon arrival at the plantation. Recruitment in India and migration to Ceylon followed a clear seasonal pattern. The coffee harvest took place in the upper lands of the Central Province (where the bulk of Ceylon’s coffee production was concentrated) from October to mid-January, and in its lower lands from August to December. The gangs arrived at the coffee plantations between June and September, returning to India between January and April. There was, therefore, a certain complementarity between the agricultural cycles of rice in India and coffee in Ceylon, which allowed the surplus of Tamil workers to move seasonally to harvest coffee without compromising the Indian economic and power structures to which they were subjected and to which most of them would return.

A plantation would mobilise different gangs, seldom from the same caste. The debt contracted by the gangs of coolies directly with their kangany – and not with the plantation – guaranteed immediate discipline in the labour process: the work was supervised by this kangany, with all the kanganies of the different gangs under the general supervision of a European superintendent. During the harvest, each coolie was responsible for one row of coffee plants, with a fixed task of picking three bushels of ripe fruit (= 36.36 litres, which equals c. 1.8 alqueires [unit of volume] of coffee in Brazil). The payment of 8d. per day only took place if the labourer picked the specified amount. If a coolie picked more, he would receive a little more money in proportion to the extra amount collected. As part of the coolies’ wages, planters or superintendents provided rice rations. Wages were paid at the end of each month of work or only at the end of the harvest (the season generally lasted three months), discounting the advance given to the kangany for recruitment in India and transportation to Ceylon and the amount of rice rations given at the plantation. The official correspondence sent to the Colonial Office in London and the agronomic publications of the period were, in general, silent about the physical coercion of the coolies, but indirect evidence in that same documentation indicates that the whip was used with some frequency.⁴⁸ The disciplinary measure that was considered the most effective, however, was the suspension of payments, which further tied the coolie to the cycle of indebtedness that kept the entire system moving. This same logic facilitated the retention of at least one-third of the total number of annual migrants for the off-season, when they were necessary for weeding, pruning, manuring, or even for setting up new plantations. In the case of weeding, it became common to hire so-called weeding gangs, that is, coolies who would remain in Ceylon after the end of the harvests under the control of their kanganies, who, in turn, would be contracted by the plantations for maintenance services during the off-season. These services were paid directly to the kangany according to the area to be weeded.

The high humidity in the Ceylon coffee zone and the seasonal variation in the labour supply called for innovations in processing. In the Central Province plantations, the fruits were just pulped and, after only two or three days of drying, immediately sent on parchment to the port city of Colombo. In the 1830s, Ceylon’s colonial administration had started to construct excellent roadways connecting the plantations to the coast. In the warehouses of Colombo’s exporting houses, parchment coffee was dried in the sun and, finally, peeled in large steam-powered circular mills. This final stage was independent of the coffee crop’s seasonality. In the final years of the coffee economy, about twenty thousand workers who lived on the island permanently (mostly women and children) were engaged in this job.⁴⁹

Faced with the uncertainties inherent to the seasonal migration of coolies mobilised by the kangany system, Ceylon planters and superintendents always opted to recruit an oversupply of workers for the harvest. Such a strategy, in order to function, depended on the push factors of Southern India, which, in turn, resulted directly from the transformations produced by British colonialism there. The dependence of the lower castes on, and their vulnerability in relation to, the local landlords was reinforced throughout the nineteenth century, which made them particularly susceptible to being captured by the advances offered by the kanganies.⁵⁰ In other words, the debt nexus was the mainspring of the entire system. Indian misery allowed the coffee wages – which were very low and stable throughout this period – to be attractive enough to keep moving the sequence of advance (in India) – debt (on the way to Ceylon and during work at the coffee plantations) – payment (in Ceylon, so they could return to India).

What was the final result of this system in terms of labour productivity? The statistics produced by the British colonial apparatus and on the private initiative of the coffee planters themselves, which are much more reliable than those available for Brazil, suggest a slight increase in coffee production between the first and the second half of the 1860s. In this period, the annual coffee production per worker rose from 144.45 kilograms in 1861–1865 to 160.45 kilograms in 1866–1870.⁵¹ For the sake of comparison: if we consider the total amount of slaves in the Brazilian coffee zones (and not just the actual number of field slaves in the coffee plantations), a Brazilian slave produced almost eight times more coffee than an Indian coolie. However, if this comparison takes into account only field slaves, the difference rises to almost fifteen times more.

The higher productivity of coffee trees in Brazil does not explain this sharp contrast. Its explanation lies in the distinctions between the two types of spatial economies and the fundamental differences between nineteenth-century slavery and coolie work under the kangany system. The logic of the spatial economy in the Paraíba Valley was based on the intensive exploitation of enslaved labourers through the extensive exploitation of natural resources; in Ceylon, the rationale was to combine intensive coffee cultivation with the extensive use of coolie labourers. This difference was not only due to the fact that their geographies and environments were distinct, but was also caused by the social relations of production that were specific to each space. Slavery was based on an interpersonal, individual relation of power between the master and the slave. The public and private determinants of this relationship, that is to say, the nature of the state and the society in question, and the nature of the social relations of force between masters and slaves established the limits for the exploitation of the slave in the labour process. at the height of the “second slavery” of the Empire of Brazil (c. 1830–1870), coffee planters in the Paraíba Valley controlled these public and private determinants with an iron fist, within a profoundly asymmetrical framework of social relations that made possible the individualisation of the slave labourer and, therefore, his greater exploitation.⁵²

Wage labour on the Ceylon coffee plantations was undeniably coercive and, therefore, not exactly free labour, given the structuring role of indebtedness relationships for its entire system.⁵³ The nature of this coercion, however, was profoundly different from chattel slavery in the Americas. That is why wage labour in Ceylon could be ideologically constructed and represented as free labour by multiple actors (colonial administrators, planters, abolitionists) of the nineteenth-century British “Anti-Slavery Empire.” However, for this reason, too, there were clear limits to increasing the exploitation of coolies. Each of the different gangs of coolies acted as a collectivity on the plantations, coming from the same Indian peasant communities and belonging to the same castes, and this collective character protected them to some degree. The relationships between British planters and superintendents with the gangs of coolie labourers were always established through a kangany, which freed the former from taking charge of the immediate control of the labour process and its corresponding supervision costs. Planters and superintendents were only interested in the quantitative and steady supply of workers, always at a low cost, and not the individual performance of each one of them. Even though the kangany kept each coolie in debt, he had neither the motive nor the means to individualise each labourer in the coffee labour process in order to make him or her work harder. The kangany received a fixed commission for each coolie that was mobilised; the coolie received a fixed daily payment for the minimum quota harvested. The monetary incentives for harvesting above the minimum quota were not so significant to stimulate coolies’ overwork. After receiving their wages at the end of the coffee harvest, ending their debt relation with the kangany, and coming back to India with all that they could save, the coolies would return to their previous lives, moving back into all the previous ties of dependence and subordination. They would then be no longer subjected to the work of the plantation – until they were eventually forced to go back to it by the reproduction of poverty and the debt nexus within the local power hierarchies in India.

As in other times and places of the capitalist world economy, the proletarianisation of the Tamil coolie in Ceylon was thus incomplete. The slave in Brazilian coffee plantations, in turn, was indeed exploited as a proletarian in the immediate labour process, to affirmatively answer the question posed by Sidney Mintz in a famous essay.⁵⁴ It was this apparent paradox – the slave labourer as a proletarian; the coolie wage labourer as a non-proletarian – that allowed Brazil to continue to command the world coffee market in the 1850s and 1860s, even without the previous and almost infinite supply of enslaved Africans.

3 Epilogue: The 1870s Crisis

The appearance of leaf rust disease (Hemileia vastatrix) in Ceylon coffee plantations in 1869 would completely change the prospects of its coffee economy in just a few years. After 1870, as the plague spread throughout the dense coffee plantations in the highlands of the Central Province, harvests began to show great oscillation, with a continuous fall in the trees’ productivity (which was already low compared to Brazil). Still, in the first half of the 1870s, major investments were made with the goal of opening of new plantations. Coffee prices were high in the world market; the launching of the Colombo – Kandy railway, which had started to operate in 1867, diminished transportation costs; and, above all, everyone expected that the pest would soon pass.⁵⁵

But it did not, and in 1875 panic set in. It was time to move with Tytler’s 1852 plan. At the end of 1875, The Ceylon government sent G.A. Crüwell, an expert on agronomic and economic issues and a journalist for The Ceylon Observer (co) to Brazil. At the same time, the newspaper had already mobilised A. Scott Blacklaw as a local correspondent, an experienced Ceylon coffee superintendent who had been hired by the London and Brazilian Bank in 1873 to run a large plantation seized by debt two years earlier. The Angélica plantation was located in the confines of the Brazilian new coffee frontier, the so-called Oeste Novo de São Paulo (São Paulo’s New West), whose terrain, which was less undulating than that of the Paraíba Valley, was even more fertile.⁵⁶ After visiting plantations in the Paraíba Valley of Rio de Janeiro in January 1876, Crüwell met with Blacklaw on the Angélica plantation. Crüwell’s stay in Brazil lasted only three months (25 December 1875 to 30 March 1876). The two would send several articles and letters to the co in the following years. Crüwell died in Ceylon in November 1877. Blacklaw sent his last letter to the co in January 1878. In that same year, the owner of the co gathered all this material in a single 150-page volume.⁵⁷

This is an exceptional document on the integration of and the divergence between the coffee economies of Brazil and Ceylon, with several keen insights into what was happening with “The Mistress of the Coffee Markets of the World.” First, and contrary to what Crüwell and Blacklaw themselves had initially assumed and written in their 1876 letters, by the end of 1877 it had become clear that coffee in Brazil had not been affected by leaf rust disease. Second, the internal slave trade had been operating for a quarter of a century in the opposite direction from what Tytler had predicted in 1852, by bringing slaves from economic sectors in crisis (such as sugar, cotton and cattle ranching) to the thriving coffee plantation economy in South-Central Brazil. Third, the railroads that were being built to serve São Paulo’s new frontiers demonstrated that Brazil had an incalculable potential for rapidly increasing its coffee exports; its plantations were much larger and were undergoing major technical improvements when it came to the processing processes of the cherries. Fourth, observing the labour process directly, notably how many trees a coolie and a slave cultivated in Ceylon and Brazil and how much each one harvested, it was found that “the work demanded and performed by the slaves is enormous. It is beyond the utmost of what human beings are capable of performing without derangement to the physical resources of the individual.”⁵⁸

But precisely because of this last point the Brazilian coffee economy would eventually also experience a crisis. The end of the transatlantic trade in 1850 had been compensated by the internal slave trade, but after the passage of the Free Womb Law in 1871, slaves were no longer born in Brazil. With great accuracy, Blacklaw observed that, in view of the failure of the experiments with European immigrant workers on the frontier coffee plantations, “now comes the question of where to find labour to employ on the plantations of Brazil, a question which will have to be answered within the next ten years.”⁵⁹ Crüwell reaffirmed the point, giving it a racial connotation: “The dark object in the picture is the slave question, the labour question. The colonist system, you see, is a failure. Europeans are unable to perform the work needed on coffee estates. They have tried, but they cannot do it: climatic influences prevent them.”⁶⁰ In both Brazil and Ceylon, then, coffee cultivation was in crisis; however, Blacklaw and Crüwell noted that these were distinct crises: one was environmental, and the other was a crisis of the social relations of production.

In this period, there was also an ecological crisis in the Paraíba Valley due to its extensive agronomic methods. However, the prospects of Western São Paulo – still based on slavery – promised to soon surpass the old Brazilian coffee zone. Indeed, in the short run the ecological crisis of Ceylon was much more urgent. The last pages of the 1878 book explained how rising international prices during the 1870s, the recent railway expansion, and the increase in the internal slave trade had led to the establishment of new coffee plantings in the province of São Paulo. An article published in the co on 28 March 1878, summarising the latest information collected by Blacklaw, predicted that if Brazil could send all the coffee it would soon be able to harvest to the world markets, “those markets would certainly be, to a large extent, ‘swamped’ …. All experience favours this conclusion, but Mr. Blacklaw’s statements are before our readers, who can form their own judgment as to the immediate future of coffee in view of his astounding figures.”⁶¹ The prognosis proved to be right: between 1871 and 1875 and 1881 and 1885, Brazilian exports almost doubled, jumping from 174,000 to 322,000 tons per year. The consequences were immediate: international prices fell by half between 1878 and 1885 and Ceylon’s coffee economy’s final hour had struck.

Blacklaw knew the average labour productivity figures of the coolies in Ceylon and the slaves in Western São Paulo but, despite the insistence of the Brazilian planters, he chose not to present them to the Congress. A minute and a half after this dialogue (the time it takes today to read the rest of the document aloud), he ended his speech. However, information about the individual harvesting capacity of the coolies was enough. In the crisis of Brazilian slavery (1878–1888), the Indian coolie would no longer be considered an alternative. The solution would be found in an innovative state-subsidised immigration scheme, which brought in many members of the Italian peasantry, which had been devastated by the post-1870 world crisis of wheat, to work on the São Paulo coffee frontiers. A new turning point in the world coffee economy gave birth to the fourth global coffee complex, in which Latin American producers would take the place of Ceylon and Java while post-slavery, Western-oriented São Paulo would replace the Paraíba Valley, which had been ravaged by its scheme of extensive agronomy and the intensive exploitation of enslaved workers. The latter proved to be rational in the short term but profoundly irrational in the long run.