The first recorded landing of enslaved Africans in what is today the US did not occur in Jamestown in 1619, but rather Puerto Rico a century earlier.Footnote ¹ While the slave vessels involved began their voyages from the Iberian peninsula and followed a triangular route, the phrase “triangular trade” does not come close to capturing the complex reality of the routes that distributed enslaved Africans across the Atlantic world. A decade after these first voyages, the United States mainland received its first enslaved Africans, but not from Africa. A Spanish expedition from Hispaniola landed near Sapelo Sound, Georgia, and established the settlement of San Miguel del Gualdape, if “established” is the correct word given that it survived less than two months. This mix of transatlantic and intra-American origins of the US’ Black population was therefore in place within three decades of Columbian contact. This bifurcated pattern of arrivals held true for every part of the Atlantic littoral of North and South America but was not reflected on the slavevoyages site until 2018 when the intra-American slave trade database joined its transatlantic counterpart.Footnote ²

Across the US and indeed across the Americas, it is likely that every port large enough to organize and dispatch even a coastal commercial voyage was involved in buying and selling slaves before slave trading became illegal. Trading slaves in the Americas was as ubiquitous on water as it was on land. Some smaller communities such as Bristol, Rhode Island, even entered the business after it became illegal, because the abolition laws were easier to avoid in the minor maritime centers. Between 1519 and 1866 transatlantic slave voyages set sail from places as far north as Ventspils in modern Latvia (ID 26441) and as far south as Rio de la Plata, a latitude range of 95 degrees. The east-west span of departures was almost as large, ranging from Odessa in the Crimea to Cartagena. Every continent bordering the Atlantic, including Africa, dispatched slaving expeditions.Footnote ³ These vessels serviced slave markets that were no less scattered – from Boston, Massachusetts, to Carmen de Patagones in southern Buenos Aires province, and east to the Iberian peninsula. And most of these destinations served as both entrepôts and final transatlantic markets. Thus, as suggested above, some 15–20 percent of the enslaved underwent further movement by land and/or sea that might see them crossing the Caribbean on another vessel or entering the Pacific via Cape Horn, trekking over the Andes, or the Panama Isthmus, with final disembarkations in Chile, Callao in Peru, and, for a very few, across the Pacific to the Philippines. Vivid accounts of two such journeys into the Pacific – separated by two centuries and at either end of South America – are now in print.Footnote ⁴

Written and media presentations on this slave trade, both fictional and scholarly, remain fixated on the supposed triangular pattern of the typical slave voyage with the apex anchored in Europe.Footnote ⁵ The first half-century of the transatlantic traffic was indeed very much based in Europe, mostly Lisbon and Seville and their satellite ports. Well over 90 percent of the slaves in this early period came from Senegambia, in the northern part of sub-Saharan Africa and the region closest to the optimal sailing route from Europe to America. The early triangular trade initially comprised a very flat triangle entirely confined to the North Atlantic. Ships sailed from Iberia to the Canary Islands followed by the circum-Caribbean and then a return to Iberia. Although the labor force in the silver export sector was mainly Indigenous and heavily concentrated on the American mainland, vessels sailing to the Caribbean from Spain did not have to go far out of their way to obtain African slaves. The sixteenth-century Caribbean and circum-Caribbean became populated by people of Senegambian descent based on trade patterns that were a relatively simple extension of what had been established in the Mediterranean Atlantic since the mid fifteenth century – all based firmly in Europe.Footnote ⁶

Nevertheless, a quick overview of the traffic shows that while Europe colonized the Americas, it was not in fact the center of the transatlantic traffic in people. We can now say that probably more slaves were carried off from Africa in vessels with home ports on the western side of the Atlantic Ocean than in Europe. Table 2.1 distributes the overall TSTD estimates of captives carried off from Africa by the ports from which the slaving expedition was launched. As the final row in the table indicates, while the organization of the transatlantic traffic was split between the continents of Europe and America, most of these ventures were based in the Americas.

Figure 2.1 shows the trend over time in terms of the percentage shares that European and American-based slave voyages held (the shares sum to 100). The surge in the Brazilian traffic in the early seventeenth century temporarily put the control of the trade in the Americas, but the explosive growth of the Caribbean plantations quickly reversed this pattern mid-century. Abolition of the slave trade initially affected Europe more than the Americas and this largely explains the re-emergence of the Americas’ dominance after 1800. Except for the early trade and the late eighteenth century, the organizational bases of the trade were distributed more or less equally across the Atlantic traffic, but overall, the Voyages database shows that a traffic that had begun in the Iberian peninsula in first half of the sixteenth century eventually shifted to the Americas and for the last forty years of its existence was located there almost entirely.

Figure 2.1 Share of transatlantic captives taken from Africa by location of home ports of vessels carrying them

Source: Alex Borucki, David Eltis, and David Wheat, “Atlantic History and the Slave Trade to Spanish America,” American Historical Review, 120 (2015): 433–61.

Column 4 of Table 2.1 shows that more than half of all Africans who crossed the Atlantic disembarked in Cuba and Brazil, but column 2 also shows that Europe was not heavily involved in how they got there. Slave ships sent out from other American slave-holding regions supplied a smaller but still significant share of their incoming slaves. Brazilian and Rio Platense merchants owned the vessels that introduced 30 percent of the Rio de la Plata inflow. On the North American mainland, almost one-fifth of arrivals direct from Africa disembarked from colonial (and later US) ships. Even in the British Caribbean traffic, dominated as it was by Liverpool, London, and Bristol owners, nearly 6 percent of all Africans arrived on colonial vessels. But this figure understates West Indian involvement given that the Elizabeth and Sarah (id 75411), leaving from London in 1724, had five owners, four of whom were residents of St. Kitts.Footnote ⁷ Moreover, thirty of the wills of merchants dispatching slave ships from London are described as “of Barbados,” “of Jamaica,” or “of Antigua.” Two thirds of British colonial slave vessels left from West Indian ports – mainly Bridgetown, Barbados, and Kingston, Jamaica – rather than from New England. The transatlantic slave traffic based in the colonial Caribbean has largely escaped scholarly attention. While Rhode Island’s role in the business prior to US independence is well known, the so-called New England rum ships brought far fewer slaves to the Americas than did their Caribbean counterparts.

As well as occupying the organizational center of the slave trade, the Americas were central to the business in a much more profound sense. Three continents shared access to the Atlantic Ocean: Africa, Eurasia, and AmericaFootnote ⁸ though people in only two of the three were aware of each other before 1492. For more than 350 years after that year the largest maritime coerced migration in history bridged the Atlantic and bound the three together. The keystone in that bridge was America rather than Europe. It may seem a commonplace to say that without that continent there would have been no slave trade, but there are deeper meanings. The fact that there was a continent called America that had remained isolated from significant interaction with the rest of the world for tens of thousands of years virtually guaranteed the epidemiological apocalypse that enveloped the Indigenous population after 1492. It also made highly likely that in the absence of Western Europeans the resources of America would have become available to one or more of the many other post-Iron Age cultures of the Old World (both east and west) as soon as the invaders had figured out a way of making return transoceanic voyages. A long list of peoples has claimed to have reached the Americas in the medieval millennia, some with stronger evidence than others.Footnote ⁹ The counterargument is not the absence of documentation, but rather that there is no archeological evidence of a demographic disaster before the Spanish invasion. Any sustained contact with Old World biota – Asian, African, or European – would surely have triggered such an impact, just as it did between 1492 and 1680. And when this demographic catastrophe did occur, the scale would surely have been lessened after 1492 if contact with the Old World had happened in preceding centuries. In addition, because the New World had much lower population densities than any other continent in the world (holding climate zones constant), any labor used to exploit America’s abundance of resources would very likely have been coerced in the aftermath of renewed contact between the continents. For four centuries after 1492 America continued to have by far the largest land-to-labor ratios in the temperate world, and historically these are not conditions under which individuals usually work voluntarily for others for any extended period.Footnote ¹⁰

The implication of the argument is that a Eurasian culture other than Western Europe could have navigated its way to the resources of America, perhaps an Asian, Russian, or African society. Likewise, there were possible alternatives to sub-Saharan Africa as a source of labor. More systematic use of convicts and indentured servants from within Western Europe, or perhaps Slavic or South Asian labor was possible. Moving such labor to where its value to the elite was maximized would of course have required a long sea journey that few would have undertaken voluntarily. Without Western Europe, and without Africa, the development of the Americas would have taken a different path. But several millennia of isolation and the “portmanteau biota,” of micro-organisms and animals that accompanied migrants from the Old World would have ensured depopulation of the continent in any event.Footnote ¹¹ Old World technology and the ubiquity of slavery globally, already discussed, made some form of coerced labor in the aftermath of demographic collapse in America highly probable. Without Europe, exploitation of American resources would have occurred later and proceeded more slowly than it did, but exploitation and development there would certainly have been. To believe that America’s resources would have entered the global economy without empire, slavery, racism, and unequal income distribution is to ignore global history. Only if Columbus really had sailed to the East Indies in 1492 – if indeed America had not existed – could there have been less forced migration in the world after that date. A transoceanic slave trade would have happened without Western Europe and without Africa, but not without America. This, together with the fact that the Americas dispatched most of the transatlantic slave voyages and was the recipient of almost all the resulting captives, is the reason why the New World takes precedent over the Old in this reinterpretation of Atlantic slave trades.

No existing interpretations of the slave trade allow for this counterfactual emphasis on the Americas. The last century of scholarship on Atlantic history has been structured around European expansion and – in the last several decades – the pushback against that expansion on the part of the peoples it subjugated, as well as the new nations that eventually emerged there. Europe is still seen as the main instigator of the plantation complex it sustained. Goods traded for slaves may have come from around the world, but in the traditional view they passed through European ports first. There, they were sorted into appropriate cargoes, incorporated into slaving ventures, and the slaves for which they were exchanged were distributed across America under European flags. A small North American-based trade has a bit role in this interpretation and claims far more scholarly attention in the English-language literature than does its many times larger Brazilian-based counterpart. Scholars therefore see the resources (capital, expertise, financial services, and most of the merchandise) as coming predominantly from the North Atlantic, and the benefits (plantation produce, bullion, and, above all, profits) accruing mainly to Europe.

Dispatching Transatlantic Slave Voyages

As already noted, there is some truth to the traditional narrative in the first half-century of the traffic. The distribution of natural resources across the Americas was just as important as the oceanic gyres described in the previous chapter. The Spanish conquistadors looked for connections with the Indies for the simple reason that transoceanic journeys were expensive and only exotic products could bear the costs of transportation to Europe. While extending imperial control over first, the Caribbean, then New Spain and Peru and Bolivia, it was logical that Spain’s primary interest would be in precious metals. Agricultural produce, whether animal or vegetable, could not by itself bear the cost of shipping from the Caribbean to Europe until the seventeenth century. Moreover, after the plunder phase of the conquest the earliest export complexes of America were isolated. After the Spanish found silver at Zacatecas, Mexico, and Potosí, Bolivia, in the mid-1540s, the most lucrative of these resources were remote – located hundreds of miles from the Atlantic Ocean and in awkward topography.

Generally, how a commodity is produced determines both settlement patterns and the social institutions that subsequently evolve.Footnote ¹² Of all sixteenth-century economic activities that required capital – ranging from hard-rock mining at one extreme to hunting and fishing at the other – the Spanish project was by far the most capital-intensive. Locating, then mining, precious metals, organizing the necessary Native American labor, and shipping the refined output safely back to Europe constituted activities well beyond the capacities of early sixteenth-century entrepreneurs, whether single or in companies – even those in northern Italian city-states. No city-state could have created the infrastructure necessary to exploit Potosi. With the single exception of the Iberian Americas, European overseas ventures in the 1500s comprised trade, plunder, and some fishing, rather than the establishment of new centers of production. Only the most powerful of European states could contemplate such an undertaking. By comparison, the Portuguese Empire, memorably described as “occupying a multitude of territories without possessing them,”Footnote ¹³ remained huddled in three pockets on the shores of the Atlantic in 1600, with sallies beyond these enclaves establishing varying and impermanent spheres of influence.Footnote ¹⁴ Not surprisingly, most American gold and silver output before 1600 was the product of a state monopoly, just as it had been under the Inca state.Footnote ¹⁵

As the conquest proceeded, Spanish America evolved into two distinct but mutually dependent economies. The Lowlands included the Caribbean and the low-lying mainland; the Highlands generated most of the transatlantic exports. The former produced some precious metal – gold in early Española and El Chocó, Columbia – but they mostly catered to the needs of the Highlands by supplying provisions, including sugar and tobacco, from large rancheros and other latifundia. The Caribbean, especially Havana, also provided a first line of defense against European competitors for the Spanish. A quarter of a million Spanish immigrants arrived in the New World before 1640, but twice this number of African captives disembarked. Not many of either group worked directly in the export sector. The state monopoly over production extended to licensing of the slave trade, with some licenses specifying monopoly rights, and some not. La Casa y Audiencia de Indias in Seville acted as the organizational center. But this did not prevent the merchant or landowner in the Spanish Americas dispatching the occasional slave voyage to Africa in this period. The first known slave voyage organized in the Americas left Puerto Rico in 1551 financed by vecinos in San Juan. The Nuestra Señhora de los Remedios (id 42987) brought between seventy and ninety captives from the Cape Verde Islands in December of the following year.Footnote ¹⁶ Thereafter almost every plantation colony established after that Spanish century began sending its own slave vessels to Africa even as European-based merchants entered the business, including Barbados, South Carolina, New York, Virginia, and Brazil.

After the earliest years, there were few barriers to entry to the slave-trading business. Compared to most transoceanic branches of maritime trade the ships themselves were small and the areas open to the traffic on the African coast were vast enough that no single power (European or African), much less an individual slave trader, could restrict access. Between 1562 and 1866 ninety-five separate American ports ranging from Maine in the North Atlantic to Montevideo in the South fitted out and dispatched slave ventures to Africa. These ports were concentrated overwhelmingly in the regions of the continent that imported the most slaves. Despite the widespread misapprehension that New England slave traders played a major role in the business, fewer than 7 percent of all departures from ports in the Americas set out from New York and points north.

Slave traders based in Europe began to lose ground to their American counterparts well before 1600. The key factor was the movement of the sugar plantation complex from São Tomé in the Gulf of Guinea to Pernambuco (Recife), Brazil, around 1560, and a little later further south to Bahia de Todos os Santos (Salvador).Footnote ¹⁷ At first this new export center also relied heavily on Indian labor. Planters in northeastern Brazil enslaved thousands of Native Americans, some from far to the south. Bandeirantes were the first slave traders to meet the relatively modest labor needs of the late sixteenth-century Brazilian sugar sector. For a few decades a south-to-north slave trade was operating on both sides of the Atlantic before the east-to-west transatlantic movement to Brazil got underway. On the eastern side of the ocean, it brought sub-Saharan Africans and Canary Islanders to Europe and, further south, from Angola to São Tomé in the Gulf of Guinea; while on the western side it brought Native Americans from temperate to tropical Brazil, and from the northern states to the English Caribbean. On the eastern shores of the Atlantic this south-to-north trade was based in Lisbon and Seville, and on the western side merchants in Brazilian and Eastern Caribbean ports organized the traffic.

The exact genesis of the Americas-based slave trade to Africa will likely never be known. Probably some enslaved people with sugar-making skills accompanied milling equipment carried from São Tomé to Pernambuco and Vera Cruz around the mid sixteenth century, but the evidence is circumstantial.Footnote ¹⁸ Northeast Brazil was the point in the New World closest to Africa, but the very first engenhos (sugar-mills) in Brazil and in Vera Cruz have left evidence only of Amerindian, not African slaves. However, the first arrival of enslaved people from Africa was probably in Bahia in 1550, even though it was the captaincy of Pernambuco that saw the most rapid development of Brazilian sugar. The Portuguese queen regent issued a royal decree in 1559 permitting the governor of São Tomé to sell up to 120 slaves to each senhor de engenho in Pernambuco.Footnote ¹⁹ By 1580, one third of engenhos labor was of African origin, and 5,800 lived in the captaincy. But not until 1620 was Africanization of the workforce complete.Footnote ²⁰ After the São Tomé slave rebellion in the late 1590s almost all sugar in European markets was produced in Brazil.Footnote ²¹ Information on specific slave ventures for this early period is scattered, with the first recorded transatlantic slave voyage dating to 1574, but clearly African slaves had been arriving for some years prior to this. TSTD has records of ten voyages bringing or intending to bring slaves into Brazil between 1574 and 1620, and a further ten for the following decade, ending early in 1630 with Dutch occupation of Pernambuco. These twenty voyages carried off an estimated 5,200 slaves, though the projected total for arrivals in Brazil in these years, at 232,000, is many times greater than this small sample. The derivation of this estimate is explained elsewhere,Footnote ²² but the important point here is that every one of the twenty ventures in the sample was launched from Brazil – Pernambuco, Bahia, or Rio de Janeiro.Footnote ²³ The chronology of the first recorded slaving ventures from each follows what we know to have been the development sequence of the sugar complex in the Americas: Pernambuco, Bahia, and then Rio de Janeiro by the end of the century.

Cultivating and processing sugar for export to overseas markets was certainly capital-intensive, but it was not hard-rock mining as in New Spain or Peru. Mining gold and silver in the highlands of New Spain and Peru and getting it to Europe remained beyond the reach of individual merchants, in contrast to a plantation located on the coast of the South Atlantic or a Caribbean island, both of which were clearly well within grasp. If merchants could finance a plantation they could certainly invest in a slave-trading venture direct to the African coast. In Brazil the wealth that sugar and later tobacco generated, initially from Native American labor, also provided the capital that enabled a direct trade with Africa. As well as the complexities of growing and processing, sugar in effect sustained the cost of two transatlantic journeys – one to bring labor from Africa, and the other to carry plantation produce to Europe. Of all sixteenth-century Europeans, the Portuguese possessed the most extensive pool of expertise in both making sugar and trading slaves. Portuguese São Tomé was Europe’s main source of sugar in 1550, to be replaced by Brazil before the end of the sixteenth century.Footnote ²⁴ In addition, Portuguese officers, ships, and trading posts on the African coast had undergirded the Spanish slave trade from its inception. While the Portuguese state granted land and the Dutch much of the shipping, private merchants established both the plantations and the slave trade. By 1575, the Portuguese had founded what was to become the largest transshipment point for African captives ever at Luanda. Except for the brief Dutch occupation of Pernambuco and Angola, the Portuguese, more precisely the Brazilian grip on the South Atlantic gyre remained secure for the better part of the next three centuries.

The Dutch attacked and attempted to conquer what in 1630 was the largest of the three Brazilian sugar complexes – Pernambuco. Portuguese resistance and several years of Dutch consolidation meant a temporary suspension of the transatlantic traffic, at least to Pernambuco. When a Dutch slave trade to Brazil eventually got underway in 1637, it, too, was based entirely in Brazilian ports. Between 1631 and 1654, not one of the 142 slaving ventures to Brazil in www.slavevoyages.org set sail from Europe. The Dutch were the most powerful maritime nation in the world at the time, but they had quickly adopted the Portuguese pattern and located their Atlantic slave trade in the Americas. The Portuguese retook Pernambuco while the Dutch conquered Surinam from the English, a cession recognized in the Treaty of Breda in 1667. Whereas Pernambuco was located just south of the equator, the conquered territory was located just north of the line. Here, too, the Dutch organized a plantation complex complete with a transatlantic slave traffic that this time endured to the late eighteenth century. Strikingly, almost all slave voyages to the ceded territory left from Europe. Thus, two branches of the slave trade conducted largely under the same flag – that of the Netherlands – using the same marine technology and hauling thousands of people from and to the same continents nevertheless used organizational bases separated by almost 5,000 miles of ocean. Moreover, the two systems were largely mutually exclusive throughout most of the slave trade era – slave ships based in Europe rarely supplied Pernambuco, and, until the 1790s, slave ships based in the Americas never supplied the Dutch Guianas. And what held for Pernambuco also held for the whole of Brazil south of Amazonia. Max Menz and Diego Martins have found details of 75 previously unknown vessels leaving Lisbon for Luanda and Benguela between 1719 and 1807. Assuming all these were slave ships we need to note that over the same period 2,442 voyages left Angola for Brazil.Footnote ²⁵ Virtually all nineteenth-century voyages were based in Brazil. Likely fewer than 10 percent of the Africans taken to Brazil over three centuries were on vessels based in Europe.Footnote ²⁶

The organizational heart of the transatlantic slave trade in the Americas was already located in Brazil by the mid seventeenth century, a dominance that continued unchallenged through to the ending of the Brazilian traffic in 1851. And in the wider Atlantic world over the whole slave-trading era, vessels from the top three South Atlantic ports of Rio de Janeiro, Bahia de Todos os Santos (now Salvador), and Pernambuco (now Recife) carried off far more enslaved Africans than the combined total of ships leaving the three leading British ports of Liverpool, London, and Bristol together with the major French port of Nantes. More than one in four of the enslaved Africans brought to the Americas over 360 years disembarked from slaving ventures that had set out from these three Brazilian ports, almost all of them on bilateral voyages.Footnote ²⁷ Lisbon traders did have a big presence in the trade, but it was limited to the African embarkation ports of Luanda, Benguela and Bissau. These merchants even provided credit to bring captives from the African interior. But the transatlantic phase of the traffic remained at the risk of merchants in Brazil.Footnote ²⁸ Nothing better illustrates the misplaced emphasis of the literature, both scholarly and popular, on the triangular trade and Europe as the dominant centers of the slave-trading business.

For ports in the Atlantic world most dependent on slavery and the slave trade for their existence we must turn to Brazil, rather than northwest Europe or Iberia. Even after ignoring the virtual absence of hard data for the first century of slave-trading data in Bahia and Rio de Janeiro, each port dispatched more voyages to Africa than did Liverpool, or London, or Amsterdam. And whereas these and other European ports were developing a multitude of trading connections around the globe during the slave trade era, Brazilian coastal cities had only the coastal trade and a tenuous connection to the Far East as alternatives to ferrying enslaved people and the produce they created. Like their European counterparts, both Brazilian ports filled their departing slave ships with trading merchandise generated in their hinterlands. Recent work by Kristin Mann, Lisa Castillo, and Luis Nicolau Parés establishes that given the range and diversity of investors, few people in Salvador were far removed from the commerce. Salvador de Bahia was a slave-trading society. This was a city entirely dedicated to bringing in enslaved people from Africa, and to facilitating the export of plantation produce as well as the importation of whatever merchandise could be obtained in return. Kristin Mann concludes, “that few free people in Salvador were far removed from the slave trade, including after it became illegal, and many freed men and slaves had connections to the commerce, as well.”Footnote ²⁹ To anticipate a later discussion, with such an exclusive and centuries-long focus on slavery and the slave trade and the economic growth that both delivered, it is indeed odd – from the perspective of the new historians of capitalism – that industrialization did not begin in Brazil rather than 5,000 miles to the northeast.Footnote ³⁰

Sugar production spread from Brazil to the most easterly of the Caribbean islands, Barbados, in the early 1640s. A plantation economy based on tobacco and cotton, drawing on European indentured labor, was already in existence there, but by 1643 the island was exporting some sugar.Footnote ³¹ If the Brazilians moved from Indigenous to African labor over a sixty-year period, the British and French made a transition from White indentured to Black slave labor in half this time. The first ship on record carrying slaves direct from Africa arrived in Barbados in 1641. Outfitted in London, it predated the first recorded sugar exports from the island. However, a continuous slave traffic based on the island began only in 1644, with according to George Downing of Downing Street fame, 1,000 slaves arriving the following year.Footnote ³² In August of 1644, and therefore in the vanguard of the British slave trade, the first known slave venture dispatched from British America left Bridgetown for the Rio del Rey in what is now southeastern Nigeria. In the following year a small vessel left from St. Kitts for the same African region. In the first decade of the Barbados traffic a quarter of all captives brought to the island arrived on Barbadian vessels and a further 12 percent on ships based in New England.Footnote ³³ Just 25 years after the Plymouth Colony settlers disembarked, Boston was dispatching slave vessels to Africa, including one named Gift of God that no doubt had Puritan owners. By the end of the century Barbados had dispatched more than a hundred vessels to Africa,Footnote ³⁴ and merchants in Antigua and Nevis had dispatched others direct to the African coast. Overall, a quarter of the 356 slave voyages recorded as disembarking slaves in Barbados down to 1700 began their voyages in the Americas, mainly Barbados.Footnote ³⁵ Such a ratio is certainly a lower-bound estimate given that most pre-1698 voyages violated the Royal African Company (RAC)’s English monopoly over the slave trade. Illicit Caribbean-based ventures are more likely to have left no trace in the historical record.

We have much less detail of slave ships leaving the French Lesser Antilles. But we do know that the Black population of St. Kitt’s, Martinique, and Guadeloupe came to match that of the White in the early 1660s, shortly after that same pattern had emerged in Barbados. But here the slave traders were more likely to have been Dutch than French. All but one of the twenty-one recorded disembarkations of captives in the French Antilles before 1666 were from vessels setting out from Dutch ports.Footnote ³⁶

As the frontier of the sugar complex moved west through the Caribbean so, too, did the home ports of Caribbean slave traders. Kingston, Jamaica, exported sugar to London for the first time in 1669, but except for a single voyage in 1687 the slave trade based on the island did not get underway until the Company’s monopoly was relaxed in 1698. Beginning in 1699 there followed eighty documented slave voyages from Kingston, mostly in the first half of the eighteenth century. The smaller ports of Montego Bay and Martha Brae quickly joined in. The RAC was nevertheless a latecomer to the traffic based in the Americas, dispatching its first voyage – from Bridgetown – in 1679, but managing an average of only one voyage a year from that location for the rest of the seventeenth century.

The major Caribbean organizational center of the slave trade developed late in the era. Havana merchants had established a major presence in the intra-American slave traffic in the later eighteenth century. Aided by British and US withdrawal from the transatlantic business and a rapidly expanding sugar economy after the collapse of St. Domingue, they were able to evolve into major transatlantic slave traders in the decade after 1807. Cuban-based merchants dispatched more than 1,500 voyages to Africa in the nineteenth century and carried off almost half a million enslaved people. Ninety percent of these voyages either left from Havana or were Cuban-owned. In just sixty years Havana became more important in the slave trade than any European port except for the three major English centers and Nantes, all of which had traded on the African coast for a century prior to Cuban involvement. Merchants in Santiago de Cuba, Matanzas, Trinidad de Cuba, and other Cuba out-ports organized their own ventures. A wide investors base evolved in the nineteenth century. The British Commissioner to the Mixed Court reported 108 slave-ship departures in 1828 and 1829, the failure of some of which “made many shopkeepers in Havana bankrupt because of their slave trade speculations.” A quarter-century later the successful disembarkation of 479 captives in Havana from the Pierre Soulé meant that $85,000 in profits were shared among forty-eight investors, some of them in Africa.Footnote ³⁷

Although the northern continental colonies got into the business quite early, they remained minor players for the first century relative to their competitors to the south.Footnote ³⁸ The early vessels brought slaves to the New England and New York markets – the very first perhaps carrying Africans, not from Africa but from the Caribbean in 1638. One was the Desire, which first brought the few male survivors of the Pequot War in 1637 to Barbados and then returned to Boston with a miscellaneous cargo that included Africans.Footnote ³⁹ Down to 1700 only one transatlantic slave voyage left the mainland British colonies for every three from the Caribbean. All mainland ports together sent out at least thirty-six slaving ventures to Africa in the seventeenth century, two-thirds from Boston and New York, the latter beginning its two-century-long span as a transatlantic slave-trading base in 1663 while it was still Dutch. Initially New Englanders sold their slaves in the Caribbean; New Yorkers, by contrast, brought slaves back to their own city. Yet as with their sugar-planting Caribbean compatriots, tobacco farmers and rice planters, they were able to sponsor transatlantic ventures to fetch the labor they needed. Chesapeake planters dispatched a small slave vessel to Africa in 1652 from the Wye River and followed up with a further four recorded ventures down to 1698, all of them in violation of the RAC’s English monopoly. Other Chesapeake merchants partnered their London counterparts on ventures that left from the English capital.Footnote ⁴⁰

Between 1680 and 1721, New York merchants established a slave trade with Madagascar, and here, too, the last such venture, was owned jointly by New York and London merchants, in this case Jewish.Footnote ⁴¹ Further south within a couple of years of the first rice exports from Charleston, Charleston vessels were reported buying slaves in the Gambia without paying the 10 percent duty to the RAC required by the 1698 legislation.Footnote ⁴² None of this activity came close to supplying the labor needs of the rapidly expanding British North American plantation sector, but even in its earliest phase of growth, some plantation owners had the capital and expertise to blur the line dividing planters and slave merchants, and effectively met some of their own labor requirements.

The North American mainland slave-trading business began to expand in the 1720s and overtook the West Indian competition in the 1740s. The first Newport voyage dates from 1701, but it took several decades for New England rum to compete with the Barbadian product. Not until the 1720s did Newport become a significant base for slaving ventures. Rhode Islanders nevertheless held two-thirds of the mainland business by mid-century. As Sean Kelley points out, while eleven merchants owned one-fifth of the voyages dispatched, ownership in Rhode Island was broadly based and few of the merchants specialized in the business. Moreover, New England became a major shipbuilding center during the eighteenth century. While historians have focused on the role of US-built ships in the illegal slave trade of the nineteenth century, it is not generally recognized that over one-third of the British slaving fleet in the third quarter of the eighteenth century was built by North American mainland shipbuilders.Footnote ⁴³

By the end of the eighteenth century, Charleston-based slave ships had become more numerous than those sailing from New England, aided by a developing northern abolitionist movement that had begun to restrict the northern traffic. For a decade before 1808, Charleston became the entry point for the largest inflow of slaves into the mainland of any decade in American history as the new South, created by the 1803 Louisiana Purchase, became available to US slaveholders. For a few years Charleston and Havana were competing for slaves directly: some slave-ship captains checked out both ports before deciding where to sell. Between 1808 and 1820, the year that the US made slave trading a capital offence, US citizens and ports remained active in the slave trade though at a seriously reduced level and without introducing many slaves into the US itself. They were particularly involved in the Cuban slave trade where some ship owners like the d’Wolfs had plantations and, as with the well-documented voyage of the Abaellino in 1818,Footnote ⁴⁴ sent out their vessels with Spanish registration papers. Cuban and Spanish owners became dominant in the Cuban trade between 1810 and 1820, even before slave trading became a capital offence. US participation declined sharply, before reviving again in the late 1830s. Renewed participation took the form of supplying fast sailing ships and, in a reversal of the pattern in the 1810s, facilitating the supply of US registration papers to Brazilian and Cuban slave ships. US papers would allow vessels to escape the attentions of the British navy on at least the outbound voyage. Direct US ownership of transatlantic slave ventures was rare by this time.Footnote ⁴⁵

The advantages of using the Americas as an organizational base are obvious. In the eighteenth century, round-trip voyages from Caribbean home ports lasted nine months, compared to 15 months for voyages setting out from Europe. Brazilian and North American slave ships returned to home port within a year, and by the mid nineteenth century in just five months.Footnote ⁴⁶ The most important part of the voyage from the standpoint of its profitability – the return trip – lasted on average forty-two days compared to the seventy-seven days for the middle passages of European-based vessels following the triangular route. Mortality and morbidity data – the key elements governing profitability in the business – are not up to the task of allowing us direct comparisons but given the much faster Atlantic crossing they likely favored the Americas-based trafficker. America also had the advantage of abundant supplies of timber. While the British carried almost eight times more slaves across the Atlantic than vessels setting out from their colonial possessions, as already noted, the latter built over one-third of the British slaving fleet during the slave trade era. It is probable that a similar ratio of colonial to metropolitan built vessels held for slavers sailing under the Portuguese/Brazilian flags.

As for trade goods, some European manufactures reached Angola direct from Lisbon and in the late eighteenth century these averaged about 40 percent of Luanda’s imports. The South Atlantic gyre ensured that East Indian textiles were cheaper in Rio de Janeiro than in Lisbon.Footnote ⁴⁷ There were ready products at hand in the Americas in high demand on the African coast. Cachaça (alcohol), tobacco, bullion, and provisions were plentiful, with cachaça predating Barbadian rum exports by at least half a century.Footnote ⁴⁸ The first recorded New England slave vessel disembarked slaves in Barbados in 1645, but it was provisions, rather than rum, that underpinned the early North American mainland trade. A New England ship carrying rum was captured off Succondee on the Gold Coast in 1687 by the RAC, but it may have got its rum in Barbados.Footnote ⁴⁹ The typical trading cargo of these vessels, including those sailing from New York to Madagascar in the late seventeenth century was “bread, flour, butter, pork, Barbados rum, and Madeira wine.”Footnote ⁵⁰ Barbadians had drunk “mobby,” fermented from sweet potatoes, before the sugar regime, but by the late 1640s rum had taken over and indeed became the island’s currency.Footnote ⁵¹ Where ships from Europe had always included brandy or English spirits among their trade goods, Africans preferred Barbadian, and eventually, Jamaican rum when it became available – the equivalent in Africa of cachaça from Brazil.Footnote ⁵² Finally, for several decades from the beginning of the eighteenth century the commodity that trumped all others was Brazilian gold. Some London-owned ventures would sell their slaves to Bahian vessels on the Gold Coast and in the Bight of Benin and then return directly to London with gold.Footnote ⁵³

The Eurocentric thrust of slave trade historiography has ensured the sidelining of the Americas, especially on the issue of the funding and crewing of voyages. The wider literature recognizes high land-labor ratios in the Americas relative to Europe and the consequences of this for higher wages and capital costs in the New World. As Jacob Price noted, the Americas “retained the frontier characteristics of cheap land, scarce capital, and dear labor.”Footnote ⁵⁴ But the literature does not consider how slave traders adjusted to such resource patterns. One crude proxy for wage costs is the tons per crew ratio that TSTD makes it possible to calculate. At the height of the slave trade – 1751–1800 – slave vessels leaving the North American mainland had similar ratios to those leaving Brazil. These were 23 percent higher than for crews leaving British and French ports. But the big difference between the Americas and Europe in labor productivity was in the number of enslaved people per crew member. On Portuguese vessels there was one crew for every 14, a ratio which was between 40 and 70 percent greater than their northwest European competitors.

On the capital front, mercantile credit ultimately based in London, Amsterdam, and Paris underpinned the British, Dutch, and French slave trades at relatively low interest rates. In addition, creditors in most of the British Caribbean could seize plantations in the event of non-payment of debt. This was not an option in Latin America, where creditors’ claims were limited to the produce of the plantation, and debt financing of voyages was therefore more restricted.Footnote ⁵⁵

Distinct regional solutions to these financing issues evolved. Before 1775, New England merchants operated the smallest slave ships in the business, averaging below 70 tonsFootnote ⁵⁶ and often sailing between Upper Guinea and North America. They were among the smallest of all transoceanic sailing vessels. New England ships also had the highest tons per crew ratios in the slave trade. Although the average British slaver was more than twice the size of its colonial counterpart, the two groups (New England and British) had a comparable slaves per crew ratio. In addition, colonial ships could buy slaves at major British forts on the Gold Coast, where fewer men on board were required (see Chapter 4) and sell at major British Caribbean markets. Given the more than one hundred British American shipbuilding sites specified in the slavevoyages database, it seems clear that cheaper and smaller ships, and a degree of British naval protection off Africa enabled the New Englanders to compete with the more heavily capitalized British slave trade centers.Footnote ⁵⁷

A second response to relative labor and capital scarcity in the Americas emerged in the Caribbean. The TSTD shows that in the seventeenth century 128 slaving ventures set sail from British Caribbean ports compared to just thirty from the North American mainland. Here, too, vessels dispatched from British islands to Africa averaged just half the size of their London-based counterparts and were thus cheaper to fit out. But in this instance European capital was available. In the early British and French Caribbean there were strong links between establishing plantations and sending out slave voyages, with both activities funded in part by metropolitan merchants. London and Bristol merchants bought up land and introduced slaves at the same time.Footnote ⁵⁸ Of thirty-seven documented slave-ship owners (other than the London-based RAC) setting out from seventeenth-century Barbados, half were proprietors of sugar plantations and only five can be identified as local merchants without English connections. Later in the eighteenth century, Caribbean involvement was limited to the funding or part-funding of ventures that left from Europe. For example, the Freemasons of Guadeloupe dispatched the Franc-maçon (ID 32744) from Le Havre, which was destroyed off Gabon by a slave uprising.Footnote ⁵⁹ This was the branch of the slave trade of the Americas with the strongest connections to a major European slave-trading center, but that branch was also a relatively small component of the larger picture. The problem with small was the risk, as the Freemasons discovered. On mainland North America the owner of the snow Tulip (ID 25371) wrote in 1762 after a third unsuccessful venture from Annapolis, “there are more disasters in those Voyages than any others whatever.”Footnote ⁶⁰

The major Caribbean organizational center for the slave trade turned out to be not British but rather Spanish, and here ties to European capital were much weaker. Havana was the port in the Americas with longest-lasting connection to Africa, Cuba being one of the first markets for enslaved Africans to open in the Americas and certainly being the last to close. Before the 1790s the island depended on foreign slave traders for almost all its supply of enslaved people. Yet 87 percent of Cuba’s total influx from Africa arrived between 1790 and 1866, putting Havana second only to Rio de Janeiro as the world’s premier slave trading port of the nineteenth century. The development of Havana in the thirty years after the gradual implementation of the Bourbon reforms in 1789 and the parallel opening up of trade in Brazil after the Portuguese court moved to Rio de Janeiro in 1807 was just extraordinary. The income generated by Cuba’s sugar and Southeast Brazil’s coffee economies reached unprecedented levels by the second quarter of the nineteenth century.Footnote ⁶¹

In the mid 1820s only three ports in the Americas had populations larger than 100,000 – Havana, Rio de Janeiro, and New York – figures that put them in the top five most populous cities in the Americas. All three were larger than Liverpool at the time of British abolition.Footnote ⁶² Such port cities backed by the plantation output of their hinterlands – sugar, coffee, and cotton respectively – could raise the capital and provide the myriad services necessary to launch slave-trading expeditions to Africa.Footnote ⁶³ As we have seen in the case of Bahia, that the hinterlands comprised intensive agriculture rather than industrial manufacturing mattered little. Both could support the capital requirements of a vibrant slave trade. Moreover, by the nineteenth century the laws restraining creditors from taking possession of plantations no longer existed. Eventually, therefore, New York and Havana joined the Brazilian port cities as major centers serving prosperous hinterlands that could finance their own ventures. Between them New York, Havana, and Rio de Janeiro sent out 2,460 slaving ventures in the last half-century of the traffic, even though the New York component was severely restricted by increasing US sanctions against the traffic.Footnote ⁶⁴ John Harris and Leonardo Marques have laid out the innovative risk-averting strategies that New York and Havana traders adopted.Footnote ⁶⁵ In the final quarter-century, these voyages – two dozen of them on steam ships carrying more than 1,000 captives at a time – were double the size of their eighteenth-century counterparts, averaging 552 enslaved persons at embarkation as opposed to 283.Footnote ⁶⁶

The three environments for launching slaving ventures described above nevertheless applied to fewer than half of all voyages leaving the Americas for Africa. Seven thousand ventures set out from Brazil (excluding nineteenth-century Rio de Janeiro) and these drew on a yet different pattern used only in the Portuguese South Atlantic.Footnote ⁶⁷ As argued at greater length in Chapter 4, the South Atlantic slave-trading system had lower costs that were unrelated to wage rates for crew and expensive credit. Distances were shorter, voyages faster, slave rebellions fewer, losses from wars and piracy rarer, and slaves carried, both per crew member and per ton, greater than for ventures originating in the North Atlantic. In addition, as first European comers in the South Atlantic, the Portuguese claimed a tenuous, but enduring and exclusive sovereignty over two of the three biggest African embarkation points in the trade, in the form of Luanda and Benguela. Between them, these ports with a combined population of 4,000 in the late eighteenth century, accounted for almost 30 percent of all transatlantic departures from Africa. Just as the Spanish were able to occupy the richest parts of the Americas for themselves, so the Portuguese on the other side of the Atlantic were able to lay first claim to the ports that supplied the most enslaved people. In doing so they developed relations with Africans in the hinterland that no other European nation could match.Footnote ⁶⁸ Lançado traders of mixed Portuguese and African heritage helped link African sellers in the interior with Brazilian buyers on the coast who maintained barracoons where Brazilian vessels embarked captives. Dutch, English, and French slave-ship captains typically dealt directly with Africans resident on the coast, although both Dutch and English contacts did involve limited numbers of traders of mixed descent.

But if the funding requirements of a voyage in the South Atlantic were more modest, a second advantage was that Brazilians were able to tap into a wider range of small investors than any of their competitors. The contrast between the major French and English ports on the one hand and the Brazilians on the other was indeed dramatic. The www.slavevoyages.org site contains records of 15,000 owners of slaving ventures located in ports around the Atlantic. Slavers from Bahia carried off approximately the same number of Africans as vessels from Liverpool – between 1.3 and 1.4 million – though of course Bahia was active in the traffic for a longer period. However, the ownership structure and funding in the two ports was completely different. For Bahia, TSTD identifies 1,091 investors in 2,799 ventures. The Liverpool records are somewhat more complete, with 1,876 individuals investing in 4,669 voyages. The largest Bahian owner was João Ferreira Souza, active between 1709 and 1736, followed by Jose de Cerqueira Lima trading between 1801 and 1839. They had ownership stakes in vessels that embarked 17,053 and 12,618 slaves respectively. However, such numbers as these would not have placed the pair in the top sixty ranking of Liverpool slave traders. William Boat’s ships transported 55,352 slaves in the years 1744–1795, and Thomas Hodgson’s vessels carried off 52,339 captives between 1752 and 1805.Footnote ⁶⁹

The more interesting question is the structure of investment in the two ports. This can be explored with the Gini coefficient, commonly used to measure income or wealth inequality, with zero indicating complete equality and one, complete inequality. Instead of income, here we assess investment in the slave trade. Using the raw data, the Gini index for Liverpool is 0.35, and for Bahia, 0.31, suggesting that investment in the second-largest Brazilian slave-trading center was only slightly more equally distributed than in Liverpool. But the raw data in TSTD includes only the principal investors and ignores the large number of minor participants – including slave-ship crew, shopkeepers, and indeed even the enslaved crew on board the typical Brazilian slaving vessel.

The earliest examples of the participation of small investors comes from the customs records of the Cape Verde Islands in the early sixteenth century. The Madanela Cansyna left Santiago in December 1513 with 125 slaves bound for Castile. Twenty-five owners sailed with their chattel, all but two were Portuguese and among them were both merchants and crew.Footnote ⁷⁰ This pattern has recently been brought to center stage for the later period by Mary Hicks. It emerges from trials of captured vessels rather than shipping records where only the principals are identified. Hicks’ impressive data supports her claim that “[s]laving merchants drew extensively on a variety of familial and patronage ties to kin, friends, and dependents in order to complete cargoes,” and gives instances of trading cargoes owned by dozens of investors some of whom were themselves enslaved persons.Footnote ⁷¹ Very few of these individuals are included in the PAST databases in TSTD, which is biased toward the large investor. It will require a major research effort to ever identify more than a handful of them. But the practice was extensive enough that the typical Brazilian-based slaving venture had a great many more investors than French, Dutch, and English voyages. In the major northwestern European ports partnerships of ten or more are rare. The TSTD Enslavers database shows that only 93 out of 21,700 voyages sailing from northwest European ports have ownership information that falls into this category.

Hicks also describes the common practice of allowing crew to make private investment in slaves, or sometimes allotting them cargo space, as an alternative to receiving wages. English and French principals typically allowed captains and, less often, surgeons small numbers of “privilege” slaves to be transported free, but rarely extended the option to ordinary seamen and never to seamen who were also slaves. Remuneration for non-officered crew in English, French, and Dutch slave ships took the form of wages.Footnote ⁷² One exception in the very early English slave trade is the Brazil Frigate (ID 21750), which set off from Luanda for Pernambuco in 1658 with 1,200 Africans – easily the largest number of people ever carried by any English slave ship. Ordinary seamen on this venture did indeed buy slaves on their own account. However, this English ship was hired by the Governor of Angola, João Vieira, so that the voyage can be classed as Portuguese rather than English, and it is thus probable that Portuguese practices prevailed. Such strategies meant that principal investors in Brazil could adapt to both the high-wage and the capital-scarce environments of the Americas relative to their European and North American competitors. If the crew were investors, albeit small, and half their number at least on Bahian slave ships, were of African descent – including many crew members – then the complications for a class-based (or even race-based) interpretation of the slave trade, or even the Atlantic world, seem clear enough.

As for the racial composition of crews, the emergence of Black sailors in the English-speaking plantation Americas has received considerable attention and was far more common than on voyages leaving Old World ports. Enslaved sailors on privateers based in North America could receive part of any prize money that captures generated.Footnote ⁷³ But in North America, Black sea-captains were rare, and instances have yet to emerge of Black crew members on slave ships having roles as investors or in reassuring the enslaved in their own language that they were not to be eaten at the end of the voyage. There is certainly no parallel in the English, Dutch, and French slave trades to the NS da Penha de França, which left Pernambuco in 1812 with a pardo (mixed-origins) captain and fifteen crew. The only White on board was a sailor from Porto in Portugal.Footnote ⁷⁴ The extensive appendices to Mary Hicks’ book suggest that just under half the crew on slave voyages leaving Brazil were Black and half of these were enslaved. We know that the average crew on these voyages was 21.6,Footnote ⁷⁵ and that TSTD shows at least 9,628 ventures left Brazilian ports for Africa. If just ten of the crew on average were pardo or preto (Black) then close to 100,000 Africans and Black Brazilians crewed slaving voyages to Brazil.Footnote ⁷⁶

We do not know how many of these were investors, in the sense of receiving all or part of their wages in the form of enslaved people or fractions thereof, but Hicks’ work makes it clear that the practice was not only extensive but also supplemented by a traffic in African products, especially textiles that crew members were allowed to trade for and to transport to Brazil in their sea chests.Footnote ⁷⁷ But small investors were not confined to Brazil. In Havana, accionistas (shareholders) with “shares as low as $100 … eagerly sought for by clerks in public and mercantile houses.”Footnote ⁷⁸ There were more small investors in the North Atlantic slave trades than the literature recognizes, but nothing on the scale of the Brazilian and Cuban business. These financing and hiring strategies were critical to the ability of the slave traders of the Iberian Americas to defend their dominant role in Atlantic slave trading in the face of competition of the ports of northwestern Europe. There can be little doubt that their competitors had far greater access to credit and capital markets.

Indebted planters feature prominently in the histories of the French, British, and Dutch tropical colonies. Distressed estate sales and spectacular bankruptcies are almost a trope.Footnote ⁷⁹ Given that most slave traders advanced credit to plantation owners and, especially in the French case, waited for years for repayment, how could planters afford to finance slave voyages? Successful businesses have always drawn heavily on credit and renewing the slave labor force always constituted the largest single expense for owners in the first century or so of a sugar colony’s existence. Borrowing was an essential part of normal business. The value of labor force purchases may be viewed against a backdrop of total output. Table 2.3 shows the value of captives arriving in the Caribbean in 1770 relative to the total product of the British, French, and Dutch Caribbean in that year. Overall, even assuming most of these purchases were bought on credit, the debt load does not appear crippling, and Lorena S. Walsh’s assessment of the Chesapeake plantocracy that serious planter indebtedness was limited to early careers and was rare in later life was probably generally true across the Americas.Footnote ⁸⁰ Nothing in the French, British, and Dutch slave trades matched the financing system in the Portuguese traffic. Hicks’ findings revise our understanding of that traffic.

The Intra-American Slave Traffic

Sugar plantations and their ancillary activities accounted for most of the transatlantic slave trade and almost all of them were located within thirty miles of salt water. The survivors of the transoceanic passage would therefore have reached the property of their new owners on foot in much less time than it took for them to reach the African coast from their point of enslavement. Some plantation records link the enslaved persons with name of the vessel that brought them from Africa.Footnote ⁸¹ In 1787 Robert Hibbard, agent to the Duke of Chandos purchased “twenty prime young men” in Kingston from the Brooks (ID 80666, and later to become the iconic subject of a misleading abolitionist pamphlet) for the Hope estate located just three miles to the north and now the Mona campus of the University of the West Indies. In St. Domingue, Jean-Joseph de Laborde, perhaps the wealthiest slave owner in the eighteenth-century Caribbean with his 1,400 slaves, had two plantations in the southern peninsula of St. Domingue, one just six miles from les Cayes where Laborde bought Africans from the Naïade in 1769 (ID 31564). Only four of them were still alive a quarter-century later.Footnote ⁸²

However, not all prospective slave buyers had access to markets large enough to accommodate transatlantic slaving ventures. And some that did, like Cuba, were subject to imperial regulations that for centuries inhibited direct trade with Africa. Some minor plantation products such as tobacco, rice, indigo, and coffee were also accessible by water. But the rest of the market for slave labor, amounting, perhaps, to 20 percent of the total demand in the Americas, lay far away from the ocean in the Spanish American highlands and the gold-bearing regions of Minas Gerais, Goiás, and Mato Grosso of Brazil. As Figure 2.1 shows, panning for riverine gold was not as capital intensive as hard rock mining for silver.Footnote ⁸³

Figure 2.2 Panning for gold in Mato Grosso.

“Prospecto da lavagem do ouro na fábrica de minerar da mesma lavra,” Source: Universidade de Lisboa/Museu Nacional de História Natural e da Ciência, PT-MUL-RMJBA-TC-02-0007.

A workforce of about fifty is displayed, not much different from the requirements of a moderate sized engenho or a seventeenth-century Barbados plantation. The remote location of the “mining factory” depicted was a challenge, given that western Mato Grosso is not just the center of Brazil, but of South America, too. But gold was the least bulky of the precious metals and compared to a sugar-mill, the required infrastructure for its extraction was not beyond the resources of a single investor. Leo Marques has written of the owner, José Paes Falcão das Neves, who a few years later owned two “mining factories and a total of 116 slaves some of whom were employed in a cachaça distillery.” However, for the enslaved Africans forced to travel to these remote regions, the passage across the ocean constituted less than half their total journey in terms of time.Footnote ⁸⁴

A further underlying pattern saw slave emporia in the tropics such as those in the Caribbean and in Rio de Janeiro supplying slaves to smaller markets in both the tropical and temperate Americas. The first transatlantic voyages did not reach the Spanish American mainland until the 1540s. Small groups of Africans arriving in New Spain before that point transshipped from Cuba and Hispaniola. The Rio de la Plata began receiving captives from Rio de Janeiro before the end of the sixteenth century. Between 1640 and 1790, Spanish America was heavily dependent on the intra-American trade. Many vessels carried just a single slave, others such as the San Fortunato el Nuevo carried hundreds – in one instance, 973 people from Curacao to Portobelo in 1669.Footnote ⁸⁵ Even before the mid seventeenth century a major intra-American trade in slaves developed. We now know that Spanish supplies of enslaved people were sustained by a trans-imperial intra-American slave trading operation first established by the Domenico Grillo and Ambrosio Lomellino partnership of Genoa, the basic structure of which endured into the late eighteenth century. First the Dutch and then the British and Portuguese were the dominant vendors.Footnote ⁸⁶ But Spanish-flagged ships were far more prominent in this business than in the transatlantic trade. In sharp contrast to the transatlantic business almost as many Africans moved to their ultimate destination under the Spanish flag as under the British.Footnote ⁸⁷

Yet the Dutch accounted for more than 60 percent of all the recorded intra-American traffic before 1700, most of it originating at Curacao with a smaller stream from St. Eustatius. Curacao, a near desert of an island in the southern Caribbean, along with the similarly infertile St. Eustatius, received 82,000 captives direct from Africa in the second half of the seventeenth century, almost as many as the whole of the French Caribbean plus Jamaica combined. Ninety-five percent of these Africans were destined for reshipment to the Spanish American mainland, which, for some, meant a further ocean voyage from Panama to Callao in Peru.Footnote ⁸⁸ Barbados was the British counterpart to Curacao until the second decade of the eighteenth century. Strikingly this small island, along with Jamaica, was in the top five jurisdictions in the Americas in terms of arrivals of Africans, following Brazil, Cuba, and St. Domingue.Footnote ⁸⁹ Unlike the Dutch islands, Barbados combined a thriving plantation sector with a major slave-trading entrepôt role. In 1715, John Ashley, the RAC agent for Barbados, wrote that typically 5,500 slaves per annum arrived at Bridgetown (remarkably close to the mean slavevoyages annual estimate for the island, 1711–1715), of whom 1,400 were sold on to other sugar islands, mainly French, and 900 to the North American mainland.Footnote ⁹⁰ However, Jamaica quickly took over as the major slave entrepôt in the Caribbean during the eighteenth century. Of the 1 million Africans brought to the island between 1701 and 1800, 200,000 were subsequently re-embarked on smaller vessels, three-quarters to the Spanish-speaking regions. With the Dutch largely abandoning the internal Caribbean traffic after 1730 to focus on Surinam, the British became the dominant slave trader within the Caribbean. In addition to Jamaica’s expanded role, they opened free ports in Grenada and Dominica acquired from France in 1763 where ships of any nation could buy slaves free of duty.Footnote ⁹¹ The Danes followed suit and the island of St. Thomas became yet a further, if minor, slave entrepôt, especially during wartime.

After the low-lying sugar regions, the gold- and silver-producing highlands of the Americas comprised the second largest source of demand for enslaved labor in the Americas before 1800. Between 1540 and 1640 more than four out every five Africans brought into the Spanish Americas disembarked at Veracruz, Cartagena, Portobelo or Rio de la Plata, venues that were starting points for a strenuous further journey often by sea or through rugged landscape.Footnote ⁹² The enslaved were employed in the alluvial extraction of gold in Chocó, Columbia, and from the late seventeenth century in Minas Gerais, Brazil. However, the labor force in the much larger silver-mining sector of Spanish America was not African, but rather overwhelmingly Indigenous and came from levies on local communities called mita labor – a system that, in South America at least, the Spanish took over from the Inca Empire.Footnote ⁹³ The processing of silver in Zacetecas employed some Africans, as indeed did sugar plantations in the lowland Vera Cruz region. However, Africans, mostly enslaved, flooded into these areas not to labor in mines or plantations, but rather to urban areas to work at a wide range of tasks – including manufacturing. When the Panama City council complained of the shortage of labor in the 1640s after Portuguese withdrawal from the Spanish Empire, mining was only one of the sectors primarily affected. Sawmilling, pearl fishing, ranching, agriculture, and transisthmian transportation were listed as more important.Footnote ⁹⁴ But all this activity was ultimately sustained by the prosperity stemming from the high levels of silver production throughout the colonial era.Footnote ⁹⁵ Like almost everywhere else in the Americas, free people who could afford to buy slaves generally did so.

Portuguese merchants probably managed the greatest share of the intra-American slave trade over the whole era of Atlantic slave trading. The Portuguese colony of Côlonia do Sacramento facilitated the transfer of at least 47,000 slaves to the Rio de la Plata region between 1680 and Spanish annexation of Côlonia in 1777, and thereafter Portuguese sold a further 26,000 to Platense merchants down to 1810. Many of these Rio de la Plata arrivals were then taken to the Potosí region and others by sea to Spain’s Pacific colonies.Footnote ⁹⁶ But the major Portuguese role in this traffic played out on land rather than the ocean. After the discovery of gold in Minas Gerais in the late 1690s, hundreds of thousands of African slaves arriving in Bahia and Rio de Janeiro were sold on to the newly opened gold producing regions. The traffic in captives to these two ports during the thirty-five years before 1696 saw a combined total of 6,100 people per year disembarking. But in the sixty-five years 1696 to 1760, when gold production is thought to have peaked, the annual average more than doubled to 14,415, suggesting that Brazilians perhaps moved half a million Africans inland, a distance of at least 215 miles.Footnote ⁹⁷ Over the whole colonial era this phase of coerced migration could easily have reached a million. Initially the route from Rio went through Paratí and took six weeks through rugged terrain, but in 1701 a new road cut the journey by half. But even then such a journey would have excluded the very young as well as older people. The Jesuit André João Antonil described these two routes, plus a third from Bahia, in his foundational work on Brazilian slavery. Bahia was further away from the gold production sites in the interior, but the route was easier to travel.Footnote ⁹⁸ The enslaved moved in batches not dissimilar to those grouped together on a voyage and, given that police approval of such movements was necessary, records for the nineteenth century at least have survived.

The intra-American traffic attained its greatest importance in the nineteenth century, and again the Portuguese had a major role. After 1800 the major plantation crops of the Americas became cotton, grown overwhelmingly in the US South, sugar centered on Cuba, and coffee in southeast Brazil, all heavily dependent on slave labor. However, in the wider Atlantic world an abolitionist movement developed almost in lockstep with these new slave markets, the impact of which was to gradually restrict the international slave trade and more specifically access to Africa. Two of the three new major plantation centers, the US and Brazil, not coincidentally the largest countries in the Americas, turned to domestic sources of slaves. A major movement of slave labor occurred from the US Old South to the New after the invention of the cotton gin and Jefferson’s Louisiana purchase. More than 600,000 slaves were moved to the New South in the forty years after 1820, of whom 150,000 were traded, perhaps one half by sea.Footnote ⁹⁹ This was a maritime slave trade that lay beyond the reach of the British navy as it attempted to suppress the slave trade after 1807.

A parallel situation developed in Brazil with the northeast of the country assuming the role of the Old South and the rapidly expanding coffee plantations of the southeast providing the new market for enslaved labor. Whereas US slave traders could no longer access Africa after 1807, Brazilian planters maintained their access to African captives until 1851. Thereafter, a major interstate traffic quickly developed that according to Robert W. Slenes saw the transfer of 200,000 slaves from the northeast to the southeast, though as Slenes is careful to point out, most of these came from employment outside the sugar-producing sector.Footnote ¹⁰⁰

Cuba, by contrast, had no reserves of slaves to match those in Brazil and the US. Cuban planters nevertheless tried to access slaves that were already in the Americas. Cuban plantation owners certainly drew on the labor force of the declining Cuban coffee sector as it faded in response to Brazilian competition, but this was scarcely a replacement for Africa. They also provided a welcome new home for some Jamaican and Bahamian planters and their slaves in the 1820s as British slave owners tried to escape their government’s tightening restrictions on slavery in the British Empire.Footnote ¹⁰¹ The British were by now attempting to shut down not just the slave trade but slavery itself. David Turnbull, English consul in Havana in the early 1840s, claimed that enough slaves had moved illegally from the British islands that plantations in the Holguin area of Cuba used English as the working language.Footnote ¹⁰² Numbers from this source cannot have exceeded a few thousand, however, and the Cuban authorities turned to other sources outside Cuba. A traffic with the Yucatán peninsula brought in a few thousand Indigenous slaves. Between 1809 and 1818 at least sixteen Portuguese ships brought three thousand African slaves from Bahia, but this trade, too, failed to take off.Footnote ¹⁰³ In the 1840s, Cuba opened a traffic in indentured labor from China, but onboard abuses were such that the Chinese government suspended the trade. And when these efforts failed the slavers attempted to defeat the abolitionists by using a range of subterfuges to revive slaving links with Africa,Footnote ¹⁰⁴ and did so with some success. Cuba was the last region in the Americas to give up the transatlantic slave trade.

The British government largely stymied a fourth major nineteenth-century branch of the intra-American trade. An abolitionist-inspired Order-in-Council in 1805 followed by legislation in 1806, 1807, and 1824, severely restricted the movement of slaves to fertile underdeveloped territories acquired in the Napoleonic Wars. The British had obtained possession of Trinidad, Demerara, Essequibo, and Berbice, all of which had small populations and unexploited soils. Planters in the older sugar colonies of the British Eastern Caribbean and Jamaica immediately began to migrate to these newly acquired colonies, a movement that was at first restricted and then banned outright except for enslaved personal servants. In the event about 25,000 British slaves were forced to make the move between 1803 and the abolition of slavery in 1833. Migrations such as this were always associated with higher mortality and lower fertility for the enslaved. But as an indication of what might have happened one has only to look at later developments. From the late 1830s the British allowed the sugar colonies to bring in indentured labor from India. Ninety percent of this inflow went to these territorial prizes of war. Trinidad and Demerara between them produced nearly half of British sugar by 1860, a figure that was one-third of Cuba’s output.Footnote ¹⁰⁵ Without restrictions this intercolonial slave traffic might easily have seen one-third or more of Britain’s 700,000 Caribbean slaves forcibly relocated in the nineteenth century.

Almost all those caught up the intra-American traffic before the nineteenth century had recently arrived from Africa. The main exceptions were the post-1807 branches from the Old to the New South, and from northeast to southeast Brazil where the African share of the population was low. We can therefore get some idea of the probability of an African being involved in a significant post-oceanic journey by comparing the rough estimates discussed above (omitting the late US and Brazilian data) with the 10.7 million enslaved Africans that survived the transatlantic slave trade. Approximately 1.5 million were subjected to a further long-distance journey in the aftermath of disembarking from a transatlantic vessel, about 14 percent of all arrivals. The only full-length study of any branch of the intra-American is Greg O’Malley’s foundational coverage of the British traffic between the Caribbean and mainland colonies. O’Malley estimates that about 15 percent of all Africans entering what became the US arrived via the Caribbean, a ratio that this discussion suggests might be appropriate for the whole of the Americas.Footnote ¹⁰⁶

In summary, the Americas was the center of a huge intra-American trade in slaves, conducted by both land and sea. All this business was organized in the Americas by merchants with little direct involvement in the transatlantic slave trade. It required substantial capitalization and credit, very little of which came from Europe, and it falls naturally into four branches. First, the temperate regions to both the north and south of the tropical Americas always received a substantial portion of their African captives from slave emporia located in the tropical Americas. A second branch was in Spanish America. Because it lacked footholds in Africa and because of its own misguided attempts to regulate the slave trade, Spanish America came to depend heavily on such entrepôts, almost always Dutch or British, prior to 1789. A third stream developed in response to the discovery of silver in what is now Bolivia in the mid sixteenth century, and gold in Minas Gerais 150 years later. A huge overland traffic in slaves flourished in South America, long before the much better-known coerced migration from the Old South to the New in the nineteenth-century US. The Buenos Aires/Montevideo axis acted as the hub for the first of these, and Rio de Janeiro and to a lesser extent Bahia anchored the second. A final intra-American stream was triggered by attempts to end the transatlantic slave trade, beginning in 1807. Abolition of the transatlantic business triggered a large internal trade in slaves both in the US and Brazil. and would have induced yet a third from the older to the newly acquired British Caribbean colonies, but for the actions of imperial officials and abolitionists in 1805–1807.Footnote ¹⁰⁷ Much of the nineteenth-century intra-American traffic is best understood as a surrogate for a transatlantic traffic that was no longer permitted.

Finally, we should note that the intra-American trade was critical for merchandise as well as slaves. Adrian Pearce has shown that the value of this traffic was three or four times larger than previously thought and that Spanish traders from the colonies sailing to the British Caribbean free ports after 1760 ran most of it. It was not just slaves sent to the British Caribbean that ended up in the Spanish Americas, but much of the merchandise showing up in British records as being sent to the British Caribbean as well. As we show in the following section, historians of the trade in both merchandise and people have made insufficient allowance for the scale of these intra-American movements.Footnote ¹⁰⁸

Receiving Slaves

Before 1540 transatlantic voyages disembarked overwhelmingly in Puerto Rico and Hispaniola, but the first gateway to the Pacific, Nombre de Dios, saw initial arrivals as early as 1531, while Cartagena’s dominance as entrepôt for the Pacific was a post-1570 phenomenon. The first recorded disembarkation in Veracruz occurred in 1546, but New Spain must have received captives before this date given that by mid-century the slave trade to Spanish America had largely switched from the Caribbean islands to the mainland. The 1550s also saw the first impact of a European war on the Atlantic slave trades when hostilities between France and Spain, 1552–1559, temporarily interrupted the steady half-century-long expansion of the business. Indeed, from this point down to 1815 dramatic swings in the volume of the traffic every three or four decades – readily observable in the timeline at http://slavevoyages.org/estimates/M85jaRPs – were invariably the result of conflict in Europe. The number of voyages carrying slaves doubled in the 1580s compared to the previous decade and almost doubled again in the 1590s. Vessels by now might still carry a range of cargo and passengers, but while the size of the vessels remained unchanged, the number of slaves on board as the ship left Africa now averaged 360.Footnote ¹⁰⁹ By this time mainland markets in the Spanish circum-Caribbean, the Pacific (accessed by the Isthmus of Panama), and, since the mid 1580s, Rio de la Plata, absorbed almost all the Africans that survived the crossing, all of whom now left from sub-Saharan Africa rather than the Iberian peninsula. With the slave trade to Brazil also well underway, the major features of the transatlantic traffic would remain unchanged for the next 250 years.Footnote ¹¹⁰

Origins and destinations would of course change over time. The intra-American traffic data allows us to examine the final distribution of Africans in the Americas. We begin by reiterating the now widely accepted point that far more Africans than Europeans arrived in the Americas prior to 1820. We do not have estimates of annual breakdowns for European migrants, but the disparity in the aggregate totals is so great that we are probably secure in saying that arrivals from Africa exceeded those from Europe in every one of the years between 1520 and 1820. On a cumulative basis, the European migrant total did not overtake the African until the mid nineteenth century. The two Atlantic slave trades distributed 10.7 million Africans across the Americas. Combined with the devastating decline in the Indigenous population that reached a nadir in the late seventeenth century, these slave trades ensured that demographically, at least, the Americas was an extension of Africa not of Europe for much of the early modern period.Footnote ¹¹¹

Overwhelmingly, slaves were sold on to the most prosperous parts of the Americas. Such regions were prosperous because they produced a range of products that consumers in the rest of the world wanted. Most of the plantation produce came to Europe, while southeast Asia received much of the gold and silver of the Americas, in exchange for high-end textiles, spices, and other luxury goods that Europeans desired. The evils of the slave trade therefore stemmed not from profit-seeking merchants, but rather from ordinary consumers coupled with their obliviousness toward how plantation labor was acquired and enforced. As argued in a later chapter, Europeans were well aware of slavery, but for most of the slave trade era were prepared to act against it only insofar as Europeans were among the enslaved, as in North Africa.

The relatively small net benefits that consumers derived from these slave-produced items compounded the evils of the traffic. Sugar, rum, coffee, and small quantities of rice and indigo added little to the feeding, clothing, and housing of Europeans. Cotton from the US was an exception but in no sense did it depend on the transatlantic slave trade for its labor. Production by the enslaved determined income in the Americas far more than in Europe. In a land-abundant environment slavery ensured that employers would have access to labor that free persons could choose to avoid. In the Spanish American highlands in the first 120 years of the slave trade landowners drew mainly on Amerindian Indigenous communities for labor, mostly corvée, for the silver mines. Silver production was nevertheless strongly correlated with arrivals from Africa in the Spanish American mainland (Rio de la Plata and Spanish circum-Caribbean combined). Some enslaved Africans did produce exports of cacao, pearls in Venezuela, gold in New Grenada, and hides in Cuba by mid sixteenth century and later the Rio de la Plata, but the greater share of African labor in Spanish America worked outside the export sector, producing many of the goods that were traded between the highland and lowland Spanish colonies. Compared to mainland North America, Spanish America was highly urbanized even before the conquest. Mexico City, Guanajuato, Querétaro, Lima, Buenos Aires, and Havana were larger than New York, Boston, and Philadelphia as late as 1820. Obrajes (workshops) in these cities and foodstuff production beyond city limits employed both free and enslaved Africans, but these activities could not have existed without the massive silver production in New Spain and Mexico. In the following century, an equally strong correlation emerged between slave arrivals in Bahia and Rio de Janeiro combined and Brazilian gold production between 1691 and 1800, but in this case African slaves did produce the gold.Footnote ¹¹²

The sugar-producing islands of the Caribbean absorbed 4.8 million people forcibly displaced from the African continent.Footnote ¹¹³ When the slave trade was no longer possible, planters introduced contract laborers from India and China instead, beginning in Trinidad in 1808 and ending only when first the Chinese government and then the government of India banned the traffic over a century later. Nevertheless, a further million migrants, mostly indentured, left for sugar-producing regions around the world – from Mauritius in the Indian Ocean to Hawaii in the Pacific, but largely to the Caribbean, Surinam, and British Guiana. There are no parallels in global history of a single crop triggering migration on such a scale, most of it coerced and all of it long-distance.Footnote ¹¹⁴

As noted earlier, Barbados was the Caribbean bridgehead for sugar cultivated by slaves. There is little evidence of either Dutch or Brazilian involvement on the island in the 1640s, but there is an indirect link. One side effect of the Dutch attempt to seize Brazil and Angola was a prolonged transatlantic war between Portugal and the Netherlands, beginning in 1630. Of the twelve quarter-centuries between 1500 and 1800, 1626–1650 is the only one in which the volume of the transatlantic traffic declined from the previous quarter-century – an undoubted side effect of hostilities between the two leading slave-trading countries of the time. We have no data but sugar output in Brazil certainly declined as well. With England at peace with both combatants until 1652, the war allowed London merchants to establish sugar plantations, with the first recorded slave ship disembarking in 1641 in Barbados. The “Compagnie des Îles d’ Amérique” commissioned a vessel that landed sixty-three Africans in Guadeloupe in 1643, presumably for tobacco cultivation.Footnote ¹¹⁵ There is no evidence of a significant sugar culture on the French islands until the following decade when the Dutch and Brazilian connection with Martinique and Guadeloupe was established. Dutch slave ships began to arrive in the former in 1651 and in Guadeloupe in 1659. In 1654, 250 Dutch Jews expelled from Pernambuco arrived in Martinique with sugar cane. The few enslaved that arrived before this date worked on tobacco and cotton and were outnumbered by engagés, the French equivalent of indentured servants.

The Caribbean received more African slaves than Brazil in the second half of seventeenth century – an estimated 463,000 – with the tiny island of Barbados, arguably the first to introduce gang labor to sugar cultivation, receiving the largest share (156,000).

Astonishingly, as Table 2.4 shows, over the three- and one-half centuries of the trade’s duration, only four other jurisdictions in the Americas received more African captives than Barbados. And if we divide the number of captives disembarking by the area of the hinterland of the receiving port then we can say that Bridgetown was easily the leading slave-trading center not only in the early modern Americas but probably in the three-thousand-year written record of sales of the enslaved around the globe. Even more surprising, the island achieved this dubious distinction despite entering the traffic late (1640 as opposed to Puerto Rico in 1519) and leaving it early (1807, as opposed to 1867 in the case of Cuba).Footnote ¹¹⁶

But Barbadian development was dwarfed when the plantation complex extended its range into the western Caribbean. The western half of Hispaniola became formally French in 1697 and proved to be extraordinarily fertile. Two western Caribbean colonies, Jamaica and St. Domingue, received a combined total of 1.6 million enslaved Africans in the eighteenth century alone. If we include Barbados and Martinique in this group then these fragments of Caribbean territories with a combined landmass less than half the size of Portugal, received over 2 million people, or one in five of all who were forcibly disembarked in the Americas. Not all the arrivals worked in the sugar sector. As already noted, many of those going to the English islands were soon dispatched to the Spanish possessions, and some of those who remained labored on other crops or worked in service industries ancillary to sugar. After 1750 St. Domingue became not only the premier sugar producer in the world but also the Caribbean leader in the production of all other plantation products – coffee, indigo, and cotton, a dominance that it took a violent revolution to end.

The rebellion that overthrew the St. Domingue plantocracy was a defining moment for the Atlantic world. However, historians have failed to recognize one unfortunate consequence. The conflagration – literally – that began in the northern province of St. Domingue in August 1791 had no perceptible impact on volume of the transatlantic slave trade. The major decrease in the traffic that occurred by 1794 was the result of war in the wider North Atlantic, not rebellion in what was to become Haiti. St. Domingue’s competitors – specifically, Jamaica, Cuba, and Brazil – immediately began to increase their slave purchases in Africa. The revolution kick-started the massive expansion of sugar production in Cuba and of coffee in southeastern Brazil. Jamaican planters also expanded their output of both crops. They all did so by drawing on the transatlantic slave trade. Jamaican annual arrivals from Africa, which had averaged 10,800 between 1782 and 1791, jumped to 15,900 in the next decade despite the transition from the peacetime 1780s to the war-torn 1790s. In the twelve months beginning in January 1793, just sixteen months after the revolution, the planters bought the largest annual number of Africans ever to arrive on the island – 28,300 or a staggering 8 percent of Jamaica’s total slave population.Footnote ¹¹⁷ Fourteen years later, British abolition of the slave trade in 1807 meant that they and their fellow planters in Trinidad and Demerara could no longer keep pace with their Cuban and Brazilian competitors. But even with the inability to bring in more African labor after 1807, Jamaican output did not fall far behind that of Cuba until the British abolished slavery itself in 1833.

The British occupation of Havana in 1762 is often mistakenly singled out as the trigger for the explosive growth of the Cuban sugar sector.Footnote ¹¹⁸ But the Haitian Revolution and the effective suppression of their own slave colonies by the British themselves first in 1807 and then in 1833 were the true catalysts, just as British orchestration of international opposition to the international slave trade helped cut off Cuba’s access to Africa in 1867. Together St. Domingue’s slaves and British abolitionists framed the rise and fall of the Cuban slave trade.Footnote ¹¹⁹ By 1860, Cuba was the leading global source of sugar, Brazil the global leader of coffee, and the US South similarly of cotton. Cuba had introduced nearly a million slaves, 90 percent of them after 1790, and Brazil another 2 million in just the first half of the nineteenth century. The Brazilian sugar economy held its own until late in the slavery era, and the cotton and rice sectors in Amazonia also used slave labor. As we have seen, the United States also depended on a slave trade, in this case the intra-American trade, but natural growth of the US slave population came close to matching that of White Americans after 1800 and ultimately the natural population growth of the US Black population proved to be the major source of new labor in most cotton-producing states.

Ever since slave trafficking in the Atlantic was shut down, research into the topic has overwhelmingly focused on the transatlantic business in the North Atlantic. The South Atlantic and the intra-American trades have received relatively little scholarly attention. Books by Mauricio Goulart and Frederic Mauro established the importance of the former in the middle of the last century, but even including all the late Joseph Miller’s formidable publications and the growing contributions of younger scholars Mariana Candido, Roquinaldo Ferreira, and Carlos da Silva, the imbalance between the north and south Atlantic historiographies has continued to grow.Footnote ¹²⁰ Post-1640 Spanish America is similarly neglected, with, for example, very little appearing on Africans in Mexico between Colin Palmer’s 1976 book and Pablo Miguel Sierra Silva’s monograph of 2018. The most recent update of www.slavevoyages.org does offer some corrective to this imbalance. Table 2.5 incorporates the latest data for the transatlantic slave trade into the European Americas as it existed in the eighteenth century. Row 6 shows estimates for that part of the intra-American traffic that disembarked slaves in Spanish territories and row 7 shows the distribution of total disembarkations after allowing for these intra-American flows.Footnote ¹²¹

Spanish America was the destination of the first as well as the last enslaved Africans to cross the Atlantic, but even so it is rather extraordinary that Africa added more to the Spanish American population than to the British colonies. Even more striking is the dominance of the Iberian Americas in this broad picture. Almost two-thirds of the survivors of the Atlantic slave trades spent the rest of their lives in Brazil and Spanish America.Footnote ¹²² The northwestern Europeans – the British, Dutch, and French – came into the business late, left it early but, while involved, sold on over half a million Africans to the Iberians from their American entrepôts. Measured by US conceptions of race, they helped Brazil to become the largest Black country in the world after Nigeria. By contrast, in Spanish America outside Cuba, Africans arrived early and lived among a large Indigenous population. For the most part an intermixing of Indigenous peoples, Africans, and Europeans (mestizaje) has made identification of their modern descendants difficult except for some locations in Mexico, Columbia, and Venezuela.

Finally, we should note the impact of differences in attitudes toward race between Iberian and other Europeans. These did help shape the Atlantic slave trades. The peoples of Europe closest to West Africa – the Portuguese and Spanish – established slave empires where it was possible for the enslaved, especially in urban environments, to buy their own freedom, even when their sugar and coffee plantation economies were at their expansionist peaks. Strikingly, this pattern continued even when abolitionist pressures restricted supplies of new labor.Footnote ¹²³ As noted in Chapter 1, slave codes in both empires incorporated the Castilian code, which provided for self-purchase. Coartación was always more likely in urban centers on which the scholarship necessarily focuses, and in Cuba some estimates put the incidence at just 1 percent of the slave population a year. Moreover, as in the Islamic world, obligations to the former owner continued after freedom. But there was nothing equivalent in the English, Dutch, and French Americas. In Brazil, the practice was certainly widespread. The growing literature on Africans returning to Africa after gaining their freedom is almost exclusively concerned with return voyages beginning in Brazil and Cuba. Except for rare cases like Catherine Zimmermann-Mulgrave (see Chapter 7), there were very few such returnees from the British, French, and Dutch Americas.Footnote ¹²⁴ In Brazil, especially, coartación boosted the slave trade to the Iberian Americas as owners sought replacement labor. By the 1870s, Brazil had 1.5 million free Blacks, by far the largest number of any jurisdiction in the Americas prior to abolition. Reliable estimates of the numbers that were freed in earlier periods are not yet possible but were certainly much higher than in Cuba.

Economic Significance

The literature on the slave trade focuses on the great benefits that Europe received from the transatlantic slave trade. But with all the mines and plantations and most of the slave trade based in the Americas, it is hardly surprising that the main beneficiaries of this massive forced system lived on the western side of the Atlantic rather than in Europe. Absentee ownership was relatively rare outside the British and French Caribbean and even within these two broad regions was less common than is usually imagined. Transatlantic trade and income were highly correlated with slavery in the Americas. As argued in a later chapter, Europe already had a diversified and well-developed economy by the early modern period. The addition of a few sectors that added so little to existing European diets and lifestyles – an alternative sweetener (sugar), additional stimulants (coffee, rum, tobacco), a new dye (indigo) – is unlikely to have generated a revolution in either consumer behavior or investment patterns. In the labor- and capital-scarce environment of the Americas, however, the impact on the Americas of producing these items was of much greater significance.

It is possible to assess this impact for benchmark years. Beginning with the North American mainland, Alice Hanson Jones’ detailed work on probate records showed that the southern colonies were better off than the north in the late colonial era. In the South, slave-grown tobacco, rice, and indigo production enabled significantly higher living standards for the free population. Wider intercolonial comparisons across the Americas reveal more dramatic differences. Barbados generated per capita product (including the enslaved) at least 50 percent greater than in England at the end of the seventeenth century. Seventy years later the net worth per free person in Jamaica was almost ten times greater than in the southern US (for slave and free together).Footnote ¹²⁵ All the export economies of the Caribbean probably had a higher per capita income than Britain in both 1700 and 1770, and Britain had the highest income of any of the eighteenth-century imperial powers.Footnote ¹²⁶ Juan Pérez de la Riva has suggested that the western region of Cuba, wherein was to be found most of the sugar complex, had a per capita income of $350 in 1862, about the same as Sweden, France, Switzerland, and England in 1955.Footnote ¹²⁷ This was maybe an overstatement, but in 1850 the per capita output of Cuba must have ranked among the top half-dozen of the world’s nations.Footnote ¹²⁸ As might be expected, the ending of slavery in both St. Domingue/Haiti (1802) and Jamaica (1834) resulted in dramatic declines in per capita output. Haiti is discussed later, and there is no national income data for the new nation in the nineteenth century. With per capita exports of no more than £0.7 pounds sterling and a large domestic subsistence sector, per capita product in Haiti was likely to have been the lowest in the Caribbean.Footnote ¹²⁹ Most of the present income gap between Caribbean and North Atlantic countries appeared in the late nineteenth and twentieth centuries. Eisner’s estimates for Jamaica in 1850 and Moohr’s for British Guiana for 1852 are much closer to the mid nineteenth-century British figure than is Eisner’s 1930 estimate for Jamaica to the British figure for that year. But both the Jamaican and the British trends were upwards.Footnote ¹³⁰ Why the elite established slavery and the slave trade is clear enough from these patterns.

Perhaps the most important finding is that Spanish American preeminence long predated the nineteenth-century Cuban sugar revolution. The English Committee for Foreign Plantations for the mid 1660s marvelled at

The comment, of course, ignored the contribution of the enslaved, but the idea of an island a little more than twenty miles long and fourteen miles across at its widest point producing more exports than the whole of the Spanish Americas is as striking today as it was then. The assessment receives some corroboration from the dramatic drop in the price of sugar in the second half of the seventeenth century as Barbados sugar reached Europe.Footnote ¹³² Nevertheless, the opinion, or rather boast, was simply wrong. Historians have failed to recognize that exports to Europe from the Spanish Americas had a far greater value than those from their British, French, Dutch, and Portuguese counterparts. In 1700, the total output of the non-Hispanic Caribbean amounted to £1.7 million or 7.6 million pesos.Footnote ¹³³ In the Spanish possessions, by contrast, bullion production alone averaged 8 million pesos annually from 1696 to 1700, an amount that made them also more valuable to Spain than Brazil was to Portugal, and more valuable than both mainland and Caribbean colonies together were to the British. It is no surprise that “nominal wages and prices were on average much higher than in Western Europe or in Asia,” and comparable to colonial North America.Footnote ¹³⁴

Seventy years later, the supremacy of the Spanish was only slightly eroded. The total annual value in pesos of French Caribbean output in 1770 was 23.1 million, and of British output, 16.2 million, whereas the Spanish Empire generated exports worth close to 31 million pesos – 29.2 million of which was bullion. Even if we include the thirteen mainland colonies in the British total, the Spanish Americas still come out well ahead – though they no longer outproduced all their competitors combined. The cession of Jamaica to Britain in 1655 and St. Domingue to France in 1697 did not apparently enable the British and French to catch up prior to the era of independence; Spanish America grew vigorously until at least 1800.Footnote ¹³⁵ Thereafter, as we have seen, the Cuban sugar replaced mainland bullion. Enslaved Africans were certainly used in the gold mines of New Granada, but “free” Indigenous labor predominated in silver mines of New Spain and Potosí.Footnote ¹³⁶ Even though African labor had never produced much silver, the income that bullion generated underpinned a demand for enslaved people throughout the Spanish Americas. It helped make Spanish America the second biggest destination for the slave trades of the Atlantic World.

We cannot easily locate Brazil in this league table of infamy except to reiterate that it easily comprised the major market for slaves in the Americas. The pre nineteenth-century data, when gold production was at its height, unfortunately does not permit us to say much more than this, particularly the lack of reliable population and pricing information. Obviously, it was home of the first and longest-lasting slave sugar plantation complex on the continent, and possibly the source of most of the world’s gold for a time in the eighteenth century, as well as coffee in nineteenth century. The transatlantic slave trade continued until 1852, and trafficking from the northeast to southeast of the country until the 1880s. Once gold exports began, then both as colony and independent nation Brazil must have rivalled Spanish America in the value of its exports and thus probably exceeded those of its better-documented rivals in the Caribbean. The Brazilian agricultural renaissance of the later eighteenth century and beyond is reflected in the time profile of arrivals from Africa, given that the decline in gold exports at this time occurred without any decline in the slave trade. Brazil has yet to be fully integrated into the history of slavery in the Americas. Comparative scholarship in Brazil on the African diaspora in the Americas has focused on race rather than slavery.Footnote ¹³⁷

Sugar plantation owners in the Americas generated considerably more income than did slave traders and we should return to the question of links between the two groups. The state-owned mines in Spanish America used Indigenous labor so we would not expect such links there. Likewise, gold producers in Brazil do not show up as owners of slave ships. As we have seen, Barbadian and Chesapeake planters did initiate slaving ventures, but once a plantation colony was established, there was usually little overlap between growing and processing produce and buying slaves in Africa. Few of the major English slave-trading families acquired significant holdings in plantations. Their names cannot be found among those compensated by the British government in the aftermath of the 1833 abolition act.Footnote ¹³⁸ When Portuguese slave traders sold up their assets and abandoned Rio de Janeiro to return to Portugal as a direct result of the ending of the Brazilian slave trade in 1852, they did so without any apparent impact on the plantation sector. The grand blancs of St. Domingue, many of whom lived in France, do not appear to have invested in the slave trade.Footnote ¹³⁹ Both the New England slave trade and its rejuvenated successor based in Charleston in 1804–1807 were no different. Even when South Carolina reopened its ports to the transatlantic traffic in 1803, the value added to the US national product (output values minus input values of a given activity) of the traffic could only have been small.Footnote ¹⁴⁰

Except in the very early days of a plantation colony, successful slave traders around the Atlantic world tended to diversify their activities into other forms of commerce, but even in the Americas the latter would rarely involve plantation ownership. Some exceptions appear in Cuba in the nineteenth century, but slave trading, both transatlantic and intra-American, was a service to a much larger sector producing much more highly valued exports. Julián Zuluetta invested in at least eighteen large slave-trading ventures in the last quarter-century of the Cuban slave trade, but his main business was the four large sugar estates with a combined labor force of over 2,000 captives, acquired after he emigrated from Spain in 1832. Likewise, the twenty voyages sent out by major plantation owner Pedro Martinez during the thirteen years 1828–1840 comprised a sideline. Most investors, including the now well-documented Portuguese Company in New York in the late 1850s, were not plantation owners. In contrast, the Zangronis, Hermano & Cía firm of slave traders, operating from Havana for a half a century from the 1810s, had some plantation investments but do not appear to have been plantation owners.Footnote ¹⁴¹ The firm dispatched twenty-seven slaving voyages to Africa, but these were unlikely to have been the core of their commercial activities. Trade, agriculture, and mining apparently required different skill sets and there is no evidence of vertical integration in the sugar business.Footnote ¹⁴²