Behrendt, Stephen, David Eltis and David Richardson. "The Costs of Coercion: African Agency in the Pre-Modern Atlantic World". The Economic History Review, Vol.54, No.3 (2001): 454-476.

Behrendt, Stephen, David Eltis and David Richardson. "The Costs of Coercion: African Agency in the Pre-Modern Atlantic World". The Economic History Review, Vol.54, No.3 (2001): 454-476.

• Log books from the port of Nantes in France indicate that around 8.6% of all ships returning to that port had experienced a revolt or attack during some stage of engagement in the slave trade, this is likely an accurate figure, with the author estimated about 8% to 10% of voyages experienced some form of insurrection (456).
• The frequency of slave revolts was geographically varied, with slaves from Upper Guinea being the most disproportionately rebellious, accounting for 40% of all revolts despite only making up 10% of all slaves. Slaves from the Congo basis and Angola stand out as the least rebellious, making up only 10% of revolts despite representing nearly half of all slaves (457).
• The likelihood of slave revolts also depended on other factors of the voyage. For example, slave revolts were more likely to occur in less crowded vessels and more likely when women and children were on board than just men (459).
• The higher incidence rate of revolt on board ships carrying fewer men is linked to the fact that European slavers considered women weak and docile, therefore requiring fewer restrictions and safety measures. Female slaves were often kept near the arms store and could give supplies for an insurrection (461).
• Successful revolts very rarely ended with a return to Africa, with almost all insurrections ending in either death on board the ship or recapture, either by European or African slavers. Those few ships which did return to Africa were often seized by African slavers and their cargo resold (462-464).
• Almost 2/3 of revolts occurred when the slave ship was still in African waters, either in embarkment or resupplying near São Tomé and Principe. Most of the remaining third of revolts occurred during the Middle Passage (464).
• There is evidence to suggest that European trades tried to avoid purchasing slaves from the regions of Upper Guinea and Gabon because they were more likely to revolt, depending on these slave markets only in the period of 1751 to 1775, when other slave markets were in disorder because of war (466).
• The development of plantation and plantation forms of production dependent on slave labor were not developed in the Americas, but by the Portuguese in the Madieras and São Tomé . These African forms of production were then transplanted to Brazil and the Caribbean (473).
• Europeans originally did try to establish colonial plantation in Africa along the Gold Coast, but they soon found themselves unable to maintain control. On the African mainland slaves would escape, populations would rebel, and fierce competition from African merchants made commodity exports unprofitable. Europeans started plantations in the Americas to avoid these problems (474). 
• "Before nineteenth-century partition there was no question of Europeans organizing labour levies on a conquered population such as developed in early modern  Spanish America, or in colonial Africa, or in German-occupied Europe after 1940. Europeans had no choice but to buy slaves; they could not steal them in significant number" (474).
• The resistance of Africans, while both on board ships and before capture, determined the price of slaves sold to Europeans. This agency in revolt certainly affected the regional variations in slaving practices and the regional concentrations of the slave trade" (475).