Skip to main content


Gross Injustice

The Slave Trade by the Numbers

By James Williford | HUMANITIES, September/October 2010 | Volume 31, Number 5

350 years.
43,600 voyages.
12.5 million Africans forced aboard European and American slave ships.
10.7 million survivors of the Middle Passage disembarked in the New World.

When it comes to taking the global dimensions of so vast and enduring a phenomenon as the transatlantic slave trade, absolute precision is more a worthy dream than a practicable goal. Lost shipping records, suspiciously round inventories of human cargo, undocumented illegal slaving expeditions, and a number of other unknowns hinder the brute force methods of simple arithmetic from arriving once and for all at the grand total of Africans brought against their will to the Americas. Yet there is a consensus. Wildly speculative sums once bandied about—3.5 million, 20 million, even 100 million—have been set aside as forty years of new research, increasingly rigorous modes of statistical analysis, and advances in computer and communications technologies have caught up to what can only be described as an embarrassment of archival riches. Although there are still, and likely always will be, gaps in the annals, historians now have more and better information on the trade than at any time in the past.

In a sense, of course, some of that information was always there, buried in the long and scattered paper trail that the business of trading slaves, because it was a business, left behind. But documents are not themselves measurable data. The basic facts and figures needed to calculate the trade’s broad contours had to be laboriously extracted from tax registers, legal proceedings, captain’s logs, surgeon’s journals, published memoirs, and private letters found in archives around the Atlantic world. That today historians can estimate—not guess—that before 1820 four Africans arrived in the Americas for every one European, that during the same period four of every five women to come to the New World were from Africa, and that the Portuguese traffic amounted to 46 percent of the entire trade, attests their efforts to find and digest the extant raw materials. The availability of hard data has dramatically improved historians’ perspective on the trade as a whole; but it has also, and this is more important, refined that perspective, pushing them to address more substantive, narrowly focused questions.

The slave trade was not, after all, some plodding, homogeneous affair. Rather, as Philip Curtin, one of the first historians to use machine-readable punch cards to analyze slave-trade data, wrote in 1969, it was “a process, constantly changing and closely integrated with other processes in the Atlantic economy . . . a pattern of changing sources of supply, changing destinations, and changing rates of flow.” Obviously, trying to grasp the historical realities of such a system by way of grand totals alone is all but impossible. Which is not to say that summary statistics don’t have their place—that nearly 1.5 million Africans died as a result of the horrendous conditions of the Middle Passage is, among other things, a staggering reminder of the evils of an economy that regarded people as commodities—but they can suggest little, if anything, of the labyrinthine dynamics of the trade’s growth, evolution, impact, and abolition. Historical insights—unlike moralizing rhetorical gestures, however valid and necessary—come from the close study of more modest quanta: fluctuations and cycles in traffic volumes, regional differences in captive gender ratios, or the frequency of African resistance aboard ship, for example.

The most recent crest in the four-decade wave of hard-data-oriented research that has helped reveal patterns like these is a project called the Transatlantic Slave Trade Database. Now in its second edition, the TSTD represents 34,941 individual voyages, perhaps as much as 80 percent of all slaving expeditions to cross the Atlantic between 1514 and 1866. Each voyage entry in the database comprises 293 content-specific data cells (also known as variables) indicating, if known, such information as the name of the ship involved, its tonnage and rig, the flag under which it sailed, the African port or ports at which slaves were purchased, the number of captives embarked, the duration of the Middle Passage, the average selling price of the slaves in Jamaican sterling, and so on. It is the single most comprehensive collection of data on the transatlantic slave trade yet assembled—a massive resource, and one understandably long in the making.

The idea that became the TSTD was sparked twenty years ago by a chance encounter between two historians in the Map Room of the Public Record Office in Kew, England. Stephen Behrendt, who now teaches at Victoria University of Wellington in New Zealand but was at the time a graduate student at the University of Wisconsin–Madison, recalls the scene with an enthusiasm that marks just about everything he says: “There were probably twenty scholars up there. I was looking at Liverpool muster rolls because I was studying slave ships from Liverpool, and around the corner was someone who was looking at documents that I knew were from the Royal African Company,” the British slaving monopoly from 1660 to 1731. “You don’t know who’s up there when you’re in the archives, and you wonder which famous professors might be around. We both had laptops, I think, which in 1990 were still a bit of a novelty. So we were looking at each other, like, ‘What are you doing?’ And finally I said, ‘Who are you?’ And he said, ‘Who are you?’”

The other laptop-toting historian, it turned out, was David Eltis, a professor of Atlantic history at Queen’s University in Canada. The two quickly discovered that not only did they share an interest in the slave trade, they were involved in more or less complementary projects. Eltis, who had already established a database of illegal nineteenth-century slaving voyages, was in the process of gathering data on the early period of the legal British trade. And Behrendt, as part of his PhD research, was doing the same for the later period.

“I guess we got into a conversation,” says Eltis, now the Robert W. Woodruff Professor of History at Emory University, “and decided that the logical thing to do would be to put everything together on a CD-ROM”—which in those days, he points out, wasn’t exactly a cheap proposition.

Two years later, Eltis and Behrendt approached Henry Louis Gates Jr. with their idea. He helped the database team apply for funding from the National Endowment for the Humanities and the Mellon Foundation, and secure an office at Harvard University’s W. E. B. Du Bois Institute for African and African American Research. They soon found some assistants, set up shop, and went to work.

“Eltis and I just sat down for a year and tried to put together all the data we had,” says Behrendt. Using a sophisticated software program called Statistical Package for the Social Sciences (known as SPSS) that allows users to organize, sort, tabulate, graph, and analyze large quantities of data, they consolidated not only their own databases, but those other historians had already compiled in digital and print formats. Among the first to contribute data was David Richardson, whose materials on slaving voyages originating in Bristol and Liverpool filled out the middle period of the British trade not covered in Eltis’s or Behrendt’s original sets. Richardson, a professor of economic history at the University of Hull in England, has remained an integral part of the TSTD team, helping to plan and execute the database, as well as raise funds.

When asked about the difficulties they faced getting the database off the ground, Eltis mentions eliminating double counting—that is, determining whether various primary sources refer to the same or different voyages—as one of the principal challenges then and now. But what Behrendt remembers was having to cope with the slow processing speeds of early nineties computers. “We would work all morning on the dataset, a big file of, say, twenty thousand voyages, we would hit save, and then go have a long lunch, because it would take over an hour for the program to save.”

All the hard work and waiting paid off: The first edition of the TSTD, a CD-ROM containing data on 27,233 voyages, was released in 1999. Widely praised in the academic press, the project was the centerpiece of a well-attended conference in 1998 (at which it was unveiled) and, in 2001, a special issue of The William and Mary Quarterly, one of the oldest and most esteemed scholarly journals of Atlantic history. As one reviewer of the first edition wrote, “It seems a foregone conclusion that the database will be the new benchmark for quantitative studies of the slave trade.”

And indeed it was. But that didn’t mean that there wasn’t room for improvement. One of the advantages of piecing together such an extensive picture of what was then known was being able to see clearly just how much remained unknown. Specifically, the 1999 database showed marked weaknesses in the data related to the Portuguese and Spanish branches of the slave trade, as well as a number of other less pronounced deficiencies. Over the next several years, more than a dozen historians, some of whom had no formal connection to the project, began searching archives in South America, Europe, and Africa for materials that might help flesh out those portions of the database. And by 2008, when, with the help of another grant from NEH and supplementary funds from the Du Bois Institute, the second edition of the TSTD came out, they had turned up information that added 8,232 new slaving expeditions (more than 5,000 of which sailed under the Portuguese flag) and revised 19,729 existing voyage entries.

With the new edition, the TSTD moved from the Du Bois Institute to Emory, and from CD-ROM to a website called Voyages ( The database is now available free of charge either as a fully fledged downloadable SPSS file, or via a simplified online interface that gives users access to ninety-nine of the most useful variables.

“This project, unlike a lot of scholarship, really relies on collegiality,” says Behrendt. “Lots of people study the slave trade. And we’ve had a lot of people very willing to basically give up their research for the greater good of centralizing the data.”

It’s a good thing, too. As a historical approach that seeks to trace economic, political, and cultural connections across four continents and as many centuries, slave trade studies demand mastery of more disciplines and archival materials than any one historian could hope to achieve in a lifetime. The number of languages involved—English, Spanish, Portuguese, Dutch, German, French—is, by itself, intimidating. Add in the intricacies of precolonial African cultures, transoceanic shipping practices, the New World plantation complex, European and American legal history, naval architecture, the latest methods in demographic analysis—the list could go on—and any lone-scholar scenario becomes highly improbable. “From a time perspective, there’s simply no way,” says Philip Misevich, an assistant professor of history at Denison University in Ohio, and one of the fifty-seven contributors of data acknowledged on the TSTD website. Even as a collective endeavor, he says, “it’s an extreme challenge,” adding, “but those of us who do it are sort of turned on by that challenge, and by recognizing that we’re putting history into a really big picture.”

Eltis, who has been working the slave trade from a quantitative angle since the 1970s, says he was impressed early on by the need to “bring in as many people as possible and get something going which everyone could contribute to.” This summer, the collaborative spirit that has sustained the TSTD all along expanded beyond the academy. The Voyages website now boasts a “Contribute” page that allows anyone to submit data related to existing records or even present evidence of slaving expeditions not yet identified in the database. “Let’s say you find a correction to the dataset,” says Behrendt, “maybe we’ve got a name wrong, or we’re missing that a vessel went to Jamaica, you can add that information to the set.” Changes won’t appear online immediately, of course; the TSTD editorial team plans to vet the submissions and, every two years or so, upload en masse whatever legitimate data have surfaced. “We’re moving into a sphere of scholarship that simply didn’t exist when I began my career,” says Eltis, “tapping into the knowledge of a much broader range of people than was ever possible before.”

Expecting much genuinely new material from the public may sound a bit farfetched, but the TSTD team is optimistic. “I can’t tell you,” says Behrendt, “how many people contact me who have the odd document related to the slave trade.” And every odd document that comes to light has the potential to fill in some of the database’s blank data cells.

Of course, the idea isn’t just to collect data, but to use them. And somewhat surprisingly, one of the most productive ways of using the data is to allow them to use you first, so to speak. Misevich recalls a colleague telling him in grad school, “Let the quantitative material raise the questions.” The phrase is practically a mantra among those who work with the TSTD. The data only offer answers to the most dry questions about the trade: How many? From where? When? But they set helpful limits, give intriguing clues, and suggest other, more compelling queries of the Why? How? and So what? variety. Behrendt describes one way to approach the TSTD as “playing with the dataset: running programs in SPSS, plotting data, making little figures and charts, looking for interesting patterns.” Eventually, things start to come together, “and you can say, ‘Aha! Why did so many slaving ships go to Bonny in the Bight of Biafra”—a port town in what is now Nigeria—“and then to Kingston in Jamaica?’ or ‘Why did so many go to Luanda and then Rio de Janeiro?’” At that point, he says, a historian might return to the very documents from which the information was originally culled. “The data often lead you to reread primary sources differently.”

A quick survey of recent slave trade scholarship gives a good sense of the kinds of intricate historical questions—and answers—the TSTD has made possible. Richardson, for example, has explored the conditions under which slave rebellions were likely to occur on board ship. To pluck just one hypothesis from his multifaceted essay: Voyages trading south of Upper Guinea that embarked a greater than average number of female slaves were more likely to experience resistance, perhaps because the women captives, who were allowed more freedom than the men, had better opportunities to steal weapons and keys from their captors. Oscar Grandió Moráguez, taking more of a port-to-port cultural approach, has inquired into the African origins of slaves brought to Cuba between the years 1789 and 1865. Although the specific African ethnic identities of these slaves remain difficult to assess, Moráguez’s research indicates that most were taken from the Congo River and Bonny ports, and not from those in the Bight of Benin, as was once thought. A number of historians have investigated the role Africans played in introducing rice agriculture to the Americas.

Behrendt describes this last as “one of the big questions in American history now. It’s a debate over whether a disproportionate number of Africans from rice-growing areas—Senegal, Sierra Leone—came to places like Georgia and the Carolinas. The question is, Is that a real pattern or not?” A recent article on the subject coauthored by Eltis, Richardson, and Philip Morgan uses evidence from the TSTD and contemporary documents to challenge the notion that African slaves were the sole source of the risiculture methods found in the Americas. Instead, they argue for a more nuanced historical perspective that would take into account the agricultural contributions of migrants, enslaved and free, from around the Atlantic world. The jury is still out, it seems; “but,” says Behrendt, “you couldn’t even have had that debate before the database.”

Professional scholars aren’t the only ones the TSTD is designed to benefit. Although the volume and complexity of the data—not to mention the level of expertise needed to produce original research from them—can make the database seem a bit daunting for those of us without PhDs in history, the Voyages team has made sure that their resource is adaptable to a wide variety of educational settings. The website offers helpful introductory essays, an extensive user guide, suggestions for further reading (including some extremely well-written overviews of the trade), and a series of lesson plans for middle and high school teachers. Misevich is quick to point out the value of the database to historians-in-training as well. While he is careful, he says, not to force his own projects on his students, last year more than half of the undergraduates in his upper-level history courses ended up using the TSTD for their final papers. “If someone is interested in North American history, they can use the database to understand the North American slave trade,” he says, “or if someone is interested in Jamaica, they can see what it tells us about Jamaica. The possibilities are really endless.”

Even with such a comprehensive and ever-growing resource as the TSTD at their disposal, the precise grand totals of the transatlantic slave trade may always hover just beyond the ken of historians. But, then, calculating the big numbers was never among the foremost goals of the project. From its inception, says Eltis, the fundamental questions driving the database have been the same as those that lie at the heart of the entire field of slave trade studies, “the impact of the trade and the nature of the interaction between Africans and Europeans.” To the exploration of these questions, the TSTD has already made impressive contributions, from inspiring scholars to dig deeper in the archives and in their own intellects, to suggesting new and sometimes unexpected lines of historical research. And given the extent and diversity of the dataset, it is hard to imagine that its potential to support and direct new slave trade scholarship will be exhausted anytime soon.

James Williford is an editorial assistant at HUMANITIES magazine.

The Transatlantic Slave Trade Database has received three grants since 1993, totaling $810,773, to support the development of the CD-ROM and the current database website project.