Author
Abstract
The Gun-Slave Hypothesis is the long-standing idea that European gunpowder technology played a key role in growing the transatlantic slave trade. I combine annual data from the Transatlantic Slave Trade Database and the Anglo-African Trade Statistics to estimate a Vector Error Correction Model of the 18th century British slave trade that captures four versions of the Gun-Slave Hypothesis: guns-for-slaves-in-exchange, guns-for-slaves-in-production, slaves-for-guns-derived and the gun-slave cycle. Three econometric results emerge. (1) Gunpowder imports and slave exports were co-integrated in a long-run equilibrium relationship. (2) Positive deviations from equilibrium gunpowder “produced” additional slave exports. This guns-for-slaves-in-production result survives 17 placebo tests that replace gunpowder with non-lethal commodities imports. It is also confirmed by an instrumental variables estimation that uses excess capacity in the British gunpowder industry as an instrument for gunpowder. (3) Additional slave exports attracted additional gunpowder imports for 2-3 more years. Together these dynamics formed a gun-slave cycle. Impulse-response functions generate large increases in slave export in response to increases in gunpowder imports. I use these results to explain the growth of slave exports along the Guinea Coast of Africa in the 18th century.