Slavery was necessary for the United States’ cotton boom because productivity levels were not high enough to attract free labour.

The dominant view among economic historians is that American slavery was an unnecessary evil: nothing good came of it for the development of the United States after independence. Even if some reluctantly accept that the boom in cotton production may have had some benefits for Antebellum America, they argue that the cotton could have instead been produced by free labour. Here, however, I will argue that productivity levels were too low in cotton to attract free labour, so slave labour was a necessary evil for the cotton boom.

Gavin Wright has summarised the argument that slavery was an unnecessary evil:

The best evidence that slavery was not essential for cotton supply is what happened after slavery’s demise. The wartime and postwar years of ‘cotton famine’ were times of great hardship for Lancashire, only partially mitigated by high-cost imports from India, Egypt, and Brazil. After the war, however, merchants and railroads flooded into the south-east, enticing previously isolated farm areas into the cotton economy. Production in plantation areas gradually recovered, but the biggest source of new cotton came from white farmers in the Piedmont. When the dust settled in the 1880s, India, Egypt, and slave-using Brazil had retreated from world markets, and the price of cotton in Lancashire was back to its antebellum level […]. Moreover, the great majority of southern cotton farms in the postwar era were specialized, purchasing grains and meats from other parts of the country.⁷

The problem with this argument is that cotton production was able to expand thanks to the black proletariat that formed after emancipation. Without access to low-wage black labour, the white farmers of Upcountry Georgia and Louisiana would have struggled to expand production. As shown in Figure 1, the wages they paid were not sufficient to compete with wages in the North, so they relied upon the labour of the former slaves and their descendants.

Cotton was unable to attract free labour because productivity levels were lower than in the North and West. The problem is shown in Table 1, which draws on William Parker’s estimates of the number of man-hours required for several crops during the broad period 1840-60.⁸ Parker found that an acre of cotton required 92 hours of labour preharvest and 41 hours to harvest, whereas an acre of wheat required 12 hours of labour for each.⁹ When combined with the yield and price data, Parker’s findings suggest that 626 hours of labour were required to make $100 from cotton, compared to 268 hours for wheat. Cotton was thus a crop that required considerable quantities of labour, especially for the preharvest period, so free labour would generally prefer to cultivate more remunerative crops that required fewer man-hours to earn a decent income. The agronomy of the South’s staple crop in this way made it more suitable for production by slaves.¹⁰

If American cotton had been forced to compete with wheat for free labour, it would have rapidly become uncompetitive on the world market. To attract free labour, the price of cotton would have had to double in order for it to only take 300 or so man-hours to earn $100 during 1840-60. Doubling the price to 22 cents per pound would have then raised the price of American cotton in Liverpool from around 6 pennies to about 11-12 pennies per pound.¹¹ As such, American cotton would have become uncompetitive in the British market, since cheaper cotton from other countries would have been preferred. Brazilian slave owners would probably have been the main beneficiaries because Brazil was the country most suited to the production of the varieties of cotton required by the new spinning machines of Britain’s industrial revolution. Indeed, Brazil had seemed poised to become Britain’s main supplier of cotton around the turn of the nineteenth century, but it could not ultimately compete with the United States on price.¹² What the consequences of becoming the world’s main cotton exporter would have been for Brazil (other than the enrichment of its slave owners) is a question for another day. Similarly, it will require another blog post to explain why King Cotton was so important to the United States’ development. What can be said for now is that the cotton boom would not have been possible without slavery, or at least something like the racially segregated labour that replaced it after emancipation.

1. G. Wright, ‘Slavery and Anglo-American Capitalism Revisited’, Economic History Review, 732:2, 2020, p. 372.
2. On the now largely abandoned study of man-hours as a measure of labour productivity, see W.N. Parker, Europe, America, and the Wider World: Essays on the Economic History of Western Capitalism, Vol. 2, America and the Wider World, Cambridge, 1991, Annex A. For a useful compilation of the available estimates for all animal products and crops, see L.A. Craig and T. Weiss, ‘Hours at Work and Total Factor Productivity Growth in Nineteenth-century U.S. agriculture’, in K.D. Kauffman, ed., Advances in Agricultural Economic History, Vol. 1, New Frontiers in Agricultural History, Bingley, 2000, pp. 8-9, Table 1.
3. This is probably an underestimate of the man-hours required by cotton, since it is unlikely to take into account the substantial increase in the productivity of pickers due to the breeding of new varieties of cotton. See A.L. Olmstead and P.W. Rhode, ‘Biological Innovation and Productivity Growth in the Antebellum Cotton Economy’, Journal of Economic History, 68:4, 2008, pp. 1123-71.
4. The same applied to other slave-produced crops, such as tobacco and rice. C.V. Earle, ‘A Staple Interpretation of Slavery and Free Labor’, Geographical Review, 68:1, 1978, pp. 51-65.
5. Cotton prices from Board of Trade, 'Report on Wholesale and Retail Prices in the United Kingdom in 1902, with Comparative Statistical Tables for a Series of Years', House of Commons Papers, 68:321, 1903, p. 44. A cent was equal to about half a penny.
6. T.A. Zamberlan Pereira, ‘The Cotton Trade and Brazilian Foreign Commerce during the Industrial Revolution’, Ph.D. diss., Universidade de São Paulo, 2017; and ‘The Rise of the Brazilian Cotton Trade in Britain during the Industrial Revolution’, Journal of Latin American Studies, 50:4, 2018, pp. 919-49.
7. G. Wright, ‘Slavery and Anglo-American Capitalism Revisited’, Economic History Review, 732:2, 2020, p. 372.
8. On the now largely abandoned study of man-hours as a measure of labour productivity, see W.N. Parker, Europe, America, and the Wider World: Essays on the Economic History of Western Capitalism, Vol. 2, America and the Wider World, Cambridge, 1991, Annex A. For a useful compilation of the available estimates for all animal products and crops, see L.A. Craig and T. Weiss, ‘Hours at Work and Total Factor Productivity Growth in Nineteenth-century U.S. agriculture’, in K.D. Kauffman, ed., Advances in Agricultural Economic History, Vol. 1, New Frontiers in Agricultural History, Bingley, 2000, pp. 8-9, Table 1.
9. This is probably an underestimate of the man-hours required by cotton, since it is unlikely to take into account the substantial increase in the productivity of pickers due to the breeding of new varieties of cotton. See A.L. Olmstead and P.W. Rhode, ‘Biological Innovation and Productivity Growth in the Antebellum Cotton Economy’, Journal of Economic History, 68:4, 2008, pp. 1123-71.
10. The same applied to other slave-produced crops, such as tobacco and rice. C.V. Earle, ‘A Staple Interpretation of Slavery and Free Labor’, Geographical Review, 68:1, 1978, pp. 51-65.
11. Cotton prices from Board of Trade, ‘Report on Wholesale and Retail Prices in the United Kingdom in 1902, with Comparative Statistical Tables for a Series of Years’, House of Commons Papers, 68:321, 1903, p. 44. A cent was equal to about half a penny.
12. T.A. Zamberlan Pereira, ‘The Cotton Trade and Brazilian Foreign Commerce during the Industrial Revolution’, Ph.D. diss., Universidade de São Paulo, 2017; and ‘The Rise of the Brazilian Cotton Trade in Britain during the Industrial Revolution’, Journal of Latin American Studies, 50:4, 2018, pp. 919-49.