The Atlantic slave trade is a central topic in African history and has been debated as to how transformative its impact was on Africa.
Lovejoy frames the debate around three core questions:
What was the volume of the Atlantic slave trade? (regional origins, ethnicity, gender, age)
What were the demographic trends of the trade?
What was the impact of the slave trade on Africa?
Core positions in the debate:
Lovejoy’s stance: the Atlantic slave trade marked a radical break in African history and was a major influence in transforming African society. Slavery interacted with domestic slave use within Africa, creating a system integral to the political economy across many parts of the continent, expanding until the late 19th century.
The transformation thesis: slavery and the slave trade were central to Africa’s transformation over the past millennium, with the Atlantic trade as a major, but not sole, influence. Also emphasizes the Muslim slave trade and other internal African developments.
Eltis/Jennings revisionism: argues the Atlantic slave trade was not large enough in scale or impact to markedly shape African history; suggests climate and human genius were more significant than export-led growth. Emphasizes per-capita analyses and argues that Africa’s exports were a small share of world trade.
Lovejoy’s position in the debate: critiques Eltis/Jennings, arguing that their conclusions are flawed and that the Atlantic trade and its suppression had a significant, largely negative impact on Africa.
Key methodological points:
Importance of integrating volume, regional origins, ethnicity, gender, and age to assess impact.
Need to reassess missing data and regional revisions as new archival material becomes available (French, British, Portuguese data; Mettas; Curto; Richardson).
The article proceeds to synthesize recent literature, seeking to test revisionist claims and to place new findings in a broad context of African demographic and economic change.
The Volume of the Atlantic Slave Trade
Early quantitative baseline: Philip Curtin’s census (The Atlantic Slave Trade: A Census, 1969) estimated about 9.566 million Africans imported to the Americas, Europe, and Atlantic islands across all periods, with a flexible margin of error due to non-archival sources.
Subsequent debates and revisions:
Some scholars challenged Curtin’s figure; revisionists argued for much smaller or larger totals, depending on data sources and methods.
Revisionists like Eltis (with Jennings) argued the absolute and relative scale of the trade was not large enough to drive major African economic changes.
Lovejoy’s synthesis of new data (as of 1989):
Total slave exports from Africa: about 11{,}863{,}000 slaves (to the Americas, Europe, and Atlantic islands) across the entire Atlantic era.
Estimated Middle Passage death rate: around 10 ext{-}20 ext{ extperthousand}? (text indicates a 10–20 percent loss; expressed as 10–20% mortality) — i.e., about 10–20% of those embarked died during the Middle Passage, reducing the number who actually arrived in the Americas.
Net impact in the nineteenth century revised upwards by about 81{,}000 slaves (Eltis’ revision) and additional modest upward adjustments from Curto in Luanda/Benguela data.
The updated total remains within Curtin’s original plausible range; the best current estimate: 11{,}863{,}000 exports from Africa, with variations depending on missing data and unrecorded shipments.
French sector: based on Mettas’ shipping data; Becker’s substitution of Mettas data suggests the real level could be around 1.15 million for the French trade (including Mascarene imports) and Becker’s higher estimates for the Americas; Richardson’s synthesis yields roughly 1.0–1.1 million for 1700–1810; Lovejoy’s preferred total around 1.15 million, with caveats about gaps in data.
British sector: Richardson’s new data show roughly 3.12 million slaves exported from Britain between 1700–1810, about 342,700 more than Lovejoy’s earlier estimate; adjusting for non-deliveries (roughly 5% non-slave shipping and another ~5% losses) could lower this to around 2.96 million, a revision of about -86{,}000 ext{ to }+Y ext{%} depending on how allowances are applied.
North American trade: Richardson reinterprets prior estimates; estimates around 208{,}000 slaves total for the eighteenth century, with debates about how many were shipped before 1760 and how many to Cuba; Lovejoy’s synthesis aligns more closely with Anstey and Drescher on the upper-lower bounds depending on source.
Portuguese Angolan trade (Curto): modest upward revision to about 7 ext{–}7.5 ext{ million?}? (text indicates about 7% higher total than earlier estimates for the Angolan trade during 1710–1830; include approx 12,736 more slaves to Luanda and 10,393 from Benguela as part of the revision per Curto).
Overall synthesis (as of 1989):
The combined adjustments lead to an estimated total of 11{,}863{,}000 slaves exported from Africa, with a plausible error margin around a few hundred thousand; this total remains within Curtin’s original range and supports a view that the Atlantic slave trade was large enough to matter, though not uniformly or equally across all regions.
Key points and implications:
The Atlantic slave trade’s volume is now more precisely quantified, but there remains significant regional variation and data gaps, especially pre-1700 and in certain ports.
The revisions tend to support a more nuanced view than the early, sweeping claims of universal devastation for all African societies; however, they also underscore substantial regional impact and demographic changes.
Regional and Ethnic Origins of Slaves
Richardson’s revisions (building on Mettas data) shift our understanding of origins, destinations, and the timing of regional involvement:
The Bight of Biafra (Delta) played a much earlier and more substantial role than previously thought, with implications for Igbo and related groups and for backward projections of Igbo political history.
West-central Africa’s (Loango coast, Angola/Benguela region) prominence: British and French buyers were heavily involved from at least the 1740s; Loango becomes a major source, with as much as 40% of all slaves possibly coming from the interior of Angola and the Zaire river basin in some periods.
The Bight of Benin (Slave Coast) and Sierra Leone: French shipments largely sourced from Benin; North American ships sourced more from Sierra Leone with secondary concentrations in Senegambia and the Gold Coast.
The Loango coast becomes a key site for the shift toward west-central Africa as a major source, influencing the global pattern of slave origins.
Ethnic origins and regional patterns (Table references summarize findings):
Ethnic groupings commonly identified in slave shipments include six major categories: Congo, Gbe (Ewe-Fon), Yoruba, Igbo, Bambara, Akan, with other groups forming smaller shares. This six-category set accounts for roughly three-quarters to nearly four-fifths of reported origins in sampled data.
In the French plantation inventories (Geggus), the distribution shows Congo, Gbe, Yoruba, Igbo, Bambara, and Akan as the dominant groups, with a notable over-representation of Senegambia and Sierra Leone origins relative to shipboard data, indicating some sampling bias in plantation inventories.
In the Slave Coast (Senegambia–Windward Coast) samples, a large share of Africans came from interior regions, reflecting long-distance captures.
Ethnicity across the Atlantic world:
In Brazil and the Caribbean, ethnicity data show clustering into: Mina (Bight of Benin subset, often tied to interior origins), Congo, Angola (Luanda), Cabinda, and other Western African groups; some mixing and amalgamation occur in the Americas as slaves adapt to new social environments.
In Bahia (Brazil), distribution patterns show significant representation of Bight of Benin origins, with Yoruba, Igbo, and Hausa represented in notable shares, but with substantial internal African origins also present.
In the Americas, ethnicity shows both sampling bias (plantation records overrepresent coastal origins) and genuine diversity (a mosaic of West and Central African groups).
Implications of origins for understanding slavery in the Americas:
The geographic origin of slaves helps explain the ethnic composition of enslaved communities in the Americas, including language, religion, and cultural retention.
The interior origins (west-central Africa, Loango, Angola) suggest strong pre-existing political and social structures that fed into the transatlantic routes.
Variation by region indicates how different colonial networks and trade routes shaped the slave supply and the demographic makeup of enslaved populations.
From Bight of Benin: large shares of Gbe and Yoruba, with Gurma (Chamba) among others; Sierra Leone and Windward Coast show overlap with Senegambia in some patterns.
From Biafra: Igbo predominance among reported ethnicity, with significant Congo and other groups; interior origins common in the Benin/West Africa cluster.
From West-Central Africa (Loango/N’Gola regions): higher shares of Congo, Yoruba, and Igbo categories as the region evolves into a major source in the 18th century, and increasingly so into the 19th century.
Methodological notes on ethnicity data:
Different data sources yield varying levels of ethnic diversity depending on data collection methods (plantation inventories vs. free/slave population data, Koelle inventories, census data, and port/ship records).
Koelle’s data (free Africans in Sierra Leone) shows greater ethnic diversity than plantation inventories, which tend to reflect coastal export regions more strongly.
The overall pattern shows that ethnicity in the slave trade was shaped by the route of capture, the slave market, and the evolving supply networks, not simply by the Africa-to-Americas diaspora.
Tables referenced in the article (summary descriptions):
Table 1: Region origins for the French slave trade (eighteenth century) showing distribution across Senegambia, Sierra Leone, Windward Coast, Gold Coast, Bight of Benin, Bight of Biafra, West-Central Africa, and Southeast Africa.
Table 2: Ethnic identifications on French plantations (the main six categories plus ‘Other’), illustrating the concentration of Congo, Gbe, Yoruba, Igbo, Bambara, Akan, and the remainder.
Table 3: Ethnic identifications of slaves from the Bight of Benin (1721–1797), with Gbe, Yoruba, Gurma, Hausa, Kotokoli, Nupe, Bariba, and totals.
Summary of ethnic findings:
By region, the slave trade involved a mix of coastal and interior African groups, with patterns that shifted over time toward west-central Africa as a major source.
The distribution of ethnic origins correlates with the destinations and routes of European traders, the presence of slave markets, and the reach of interior trading networks.
Age and Gender Profile of the Exported Slave Population
Key data sources and contributions:
Geggus compiled data on age, gender, and ethnicity for a large sample of slaves exported to the Americas from 1636–1867, based on Mettas’ shipping data and plantation records.
Eltis provided similar analyses for the nineteenth-century trade, and Joseph Miller analyzed age/sex patterns for west-central Africa (18th–19th centuries).
Overall sex ratio (males to females) across the period:
From the seventeenth century to the end of the trade in the nineteenth century, the ratio of males to females was roughly 181:100 (i.e., 64.4 ext{ extpercent} male, 35.6 ext{ extpercent} female).
The eighteenth-century French trade had a similar overall sex ratio: about 179:100 (64.2% male).
The nineteenth century saw higher male proportions, approaching or exceeding two males for every female in many regions (
West-central Africa and south-eastern Africa showed very high male shares, e.g., around 67 ext{ extpercent}–69 ext{ extpercent} male in various samples; some regional totals approach 70 ext{ extpercent} or higher).
Regional variation in sex ratios:
West Africa (overall): I58:100 (61.2% male) for eighteenth-century French shipments; Sierra Leone: around 57.3 ext{ extpercent} male; Bight of Benin: around 61.7 ext{ extpercent} male (French and other data combined).
Bight of Biafra: among combined samples (French, British, Dutch), about 129:100 (56.4% male) overall; lower than many other regions.
West-central Africa (Loango, Angola interior): much higher male shares; by the eighteenth century, roughly 212:100 (≈ 67.9% male) for west-central and south-eastern Africa combined; by the nineteenth century, male shares rose toward ~70 ext{ extpercent}.
Age distribution and child slave proportions:
Children (defined by age thresholds) were under-represented early but became more common over time, especially by the nineteenth century.
In eighteenth-century samples, children accounted for around 16 ext{–}26.5 ext{ extpercent} depending on the region and dataset, with French trade showing higher child shares pre-1800 (~26.5 ext{ extpercent}) than the average non-French trade (~16.2 ext{ extpercent}).
By the nineteenth century, the share of children rose substantially, particularly from west-central Africa (e.g., as high as 52 ext{ extpercent} in some regions) and from the Bight of Benin (around 33.2 ext{ extpercent} in the nineteenth century).
In areas far from the coast (inland sources), the share of children tended to be higher as distance increased, reflecting supply and transport dynamics.
The age–sex profile and its implications:
The nineteenth-century shift toward higher male and higher child shares is associated with the emergence of more aggressive, large-scale export markets and the capacity to secure more robust male workers and younger cohorts.
The Dutch were distinctive in achieving very high male shares earlier (e.g., 228:100 male-to-female ratio during 1675–1740, about 69.5 ext{ extpercent} male), suggesting strategic differences in sourcing and competition for male slaves.
The shift toward higher male shares appears to reflect market competition and “tight packing” strategies, especially in west-central Africa in the 1740s onward, enabling west-central Africa to become a major source for Dutch and other European buyers.
Children and region-specific patterns:
The share of children was especially high inwest-central Africa in the nineteenth century, with the southern and eastern Africa regions also showing substantial child export shares.
The Bight of Benin had relatively lower child shares in the nineteenth century (~33.2 ext{ extpercent}) compared to west-central Africa regions, which approached or exceeded 38 ext{ extpercent} to 52 ext{ extpercent} in various datasets.
Interpreting the data:
The age and gender composition reflects labor demand, supply constraints, and social/physical costs of moving young and male labor across the Atlantic.
While some authors claimed a consistent target of two males for every female, the data show substantial deviations by region and period; the nineteenth century especially shows high male shares across many regions.
Regional distance from the Americas and the structure of market competition (Dutch, French, British, Portuguese) influenced the gender composition of exported slaves.
Impact of the Atlantic Slave Trade on Africa
Core question: Did the slave trade have dramatic economic, demographic, and social impacts on Africa?
Economic costs and per-capita impact:
Lovejoy critiques Eltis/Jennings’ per-capita income framework, arguing that it inadequately captures internal African price structures and the true economic costs of slavery.
Eltis argued that Africa’s share of world trade and per-capita income were low, implying limited external impact; Lovejoy contends that the ratio of export value to per-capita income understates the social and political costs, and the real costs included loss of life, dislocation, war damages, and long-term economic disruption.
Manning’s simulation model and demographic impact:
Manning’s demographic model uses a structured simulation to estimate population decline in source areas under continuous enslavement and export.
Assumptions: starting population around 22–25 million in early 18th century; growth rates limited to around 0.5% to counterbalance the loss of slaves.
Findings: with a 0–0.5% growth rate, populations could have grown to 46–53 million by 1850, which would have represented more than double the actual 1850 populations if war, famine, and disease were omitted; Manning estimates that roughly 3 million slaves were in the Americas by the early 19th century and that roughly 21 million people were captured in Africa, with about 7 million in domestic slavery and 5 million dying within a year of capture.
Conclusion: the trans-Atlantic slave trade contributed to significant population decline in the exporting areas and increased the incidence of slavery within Africa, consistent with the broader “transformation thesis.”
Miller’s Angola analysis (west-central Africa):
Miller estimates that roughly 40% of all slaves in the Americas came from west-central Africa, with death in Africa related to capture and enslavement potentially equaling the number exported (≈ 50,000–60,000 per year in late 18th century).
He suggests total population displacement on the order of 100,000–120,000 per year, implying substantial demographic disruption.
Miller emphasizes the gendered and age structure of the population and argues that matrilineal social structures reinforced the effects of the slave trade and export.
Suppression of the Atlantic trade (19th century) and its effects:
Eltis argues the suppression of the Atlantic trade did not necessarily reduce slavery in Africa, suggesting internal production and Islamic markets continued to sustain slavery.
Lovejoy remains skeptical of Eltis’s timing and concentration of the suppression impact, pointing to jihads, Islamic expansion, and intra-African slave trading as drivers of continued enslavement in the 19th century.
Prices and demand dynamics: Eltis notes that prices for slaves declined after the suppression of Atlantic demand, suggesting a reduced external demand; Lovejoy argues that persistent internal demand and export markets kept slavery viable in many regions.
Internal African slavery and broader demographic consequences:
Slavery persisted and grew in many areas of Africa during the 19th century, driven by internal and regional slave economies, jihads, and external demand from other parts of the world.
Comparative population estimates indicate large African slave populations in regions like Haut-Senegal-Niger and the Sokoto Caliphate; for example, the Sokoto Caliphate might have contained a large slave population, potentially representing a significant share of the total population there (~¼ of 10 million in some estimates).
The transformation thesis revisited:
Lovejoy argues that the overall impact of slavery in Africa is best understood as a transformation of African societies through the dual processes of enslavement and slave trade, where slavery and the slave economy both shaped and were shaped by Africa, Europe, and the Islamic world.
He suggests that internal African social structures, political power, and economies were altered by the presence of slavery and by the demand for enslaved labor, even if some external economic indicators (like export shares of world trade) appear modest.
The broader synthesis: the costs of slavery in Africa were large and multifaceted, encompassing demographic losses, social dislocation, wars, and the destruction of property. The extent of the impact varied regionally, but the evidence points to substantial negative consequences in many areas.
The Final Synthesis and Conclusions
Overall conclusion: the Atlantic slave trade, and its suppression, had a significant, negative impact on Africa, particularly in demographic terms and in social/political transformation, even if some macroeconomic indicators (e.g., export shares of world trade) were modest.
Economic costs:
Slavery and slave trade imposed severe social and political costs and disrupted economies; the low per-capita income and the small share of external trade did not negate the heavy costs associated with the enslavement and displacement of large parts of the population.
Demographic impact:
The slave trade contributed to substantial population decline and population movements in source regions, including large-scale displacement and deaths related to capture and enslavement.
The age/sex structure indicates a substantial gendered and generational impact, with long-term demographic consequences for many West and West-central African societies.
The role of suppression:
Suppression of the Atlantic slave trade did not immediately reverse slavery or eliminate slave economies in Africa; internal slavery and Islamic slave markets continued to be important, and the broader transformation of African slavery persisted into the 20th century.
Comparative global perspective:
The scale of African slavery in the 18th–19th centuries was large enough to compare with populations in the Americas; estimates suggest that by the late 19th century Africa still housed substantial enslaved populations (e.g., in the Sokoto Caliphate and other regions) alongside large enslaved populations in the Americas.
Key empirical takeaways:
The total number of slaves exported from Africa was revised upward, to about 11{,}863{,}000, with regional revisions showing stronger roles for the Bight of Biafra and Loango coast; the distribution of ethnic origins is more nuanced than earlier simple models suggested.
The demographic profile indicates a gradual but significant shift toward more males and more children among the exported populations, with pronounced regional variation and a notable rise in male shares from the eighteenth century onward, especially on the Loango coast.
The interrelationship between trans-Atlantic, intra-African, and Islamic slave trades created a complex system of enslavement and political economy that helped transform multiple African societies.
Final takeaway: The debate about the Atlantic slave trade’s impact on Africa is nuanced. While some macroeconomic indicators may point to a limited external impact, the demographic, social, and political changes driven by slavery and the slave economy were substantial and lasting, supporting Lovejoy’s transformation-based view of African history during the Atlantic era.
Summary of Key Figures and Data Points (quick reference)
Total slave exports from Africa (to the Americas, Europe, and Atlantic islands): 11{,}863{,}000
Regional shift toward west-central Africa as major sourcing region (Loango coast) beginning in the 1740s; west-central Africa potentially contributed up to ~40 ext{ extpercent} of all slaves.
French slave trade total (adjusted ranges): approximately 1{,}15{,}000–1{,}50{,}000 slaves exported from Africa (the count affected by Mascarene trade inclusion and data gaps).
British slave trade (1700–1810): around 3{,}12{,}000 exported; when adjusted for non-deliveries, around 2{,}96{,}4{,}0{0} (roughly 2.96{ ext{ million}}).
North American trade (18th century): estimates around 208{,}000 slaves (with debate about sub-periods and data sources).
Sex ratio (males to females) overall (17th–19th centuries): about 181:100 (64.4% male); eighteenth-century French shipments ~179:100 (64.2% male); nineteenth-century data show higher male shares, approaching 70% in some regions.
Children share ( eighteenth century → nineteenth century trend): ~16 ext{-}26.5 ext{ extpercent} in the 18th century; rising to as high as ~33 ext{–}52 ext{ extpercent} in various regions by the 19th century, depending on region.
Africa’s share of world trade and per-capita income (Eltis/Jennings): per-capita income in western Africa was low; export share as a percentage of income was relatively modest, though this is a debated metric for impact.
Population-scale estimates (Manning): potential West/West-central African population around 22–25 million in the early 18th century; simulation suggests that even modest growth rates could not offset population losses due to slavery; estimates of total people captured in Africa around ~21 million, with ~7 million enslaved domestically and ~5 million dying within a year of capture.
Slavery in Africa (19th century): large slave populations in the Sokoto Caliphate and other regions; the scale of Africa’s internal slavery remained substantial even after suppression of the trans-Atlantic slave trade.