Production by Enslaved Workers and the US GNP

 

Sad to say, the gulf between economic history and mainstream history is as wide today as ever.  Undoubtedly many forces have contributed to this state of affairs, but one historical breakpoint was the controversy over slavery during the 1970s, prompted by publication of Time on the Cross, by Robert Fogel and Stanley Engerman, in 1974.[1]  Perhaps because of the subsequent divergence, when a new round of studies appeared some years later, written by historians specifically concerned with economic aspects of slavery, the authors drew very little on research by economic .

One claim in recent literature that is often repeated is that in the antebellum period, enslaved workers produced an outsized proportion of the total value produced in the U.S. economy – the Gross National Product (GNP).   A case in point is this statement on the website of the Smithsonian Institution’s National Museum of African American History and Culture:  “Men, women and children, pushed by the whip, produced cotton, rice, sugar and tobacco valued at well over half the gross national product.”[2]  The exhibit provides no source for this claim, but it seems likely to originate with Edward Baptist, who wrote: “All told, more than $600 million, or almost half of the economic activity in the United States in 1836, derived directly or  indirectly from cotton produced by the million-odd slaves.”[3]  Baptist’s calculation is egregiously exaggerated, conflating inputs with outputs and adding items that are not even part of GNP.  As such, it has been roundly criticized by economists.[4]  But what would be a more accurate answer to the question?

It is not an easy question to answer because there is no direct aggregate data on the value of the goods and services enslaved people produced. Because enslaved persons represented only about 12 percent of the US population in 1860, one might simply dismiss the “One-Half” claim out of hand as a physical impossibility.  To get closer to the answer, Paul Rhode recently constructed a bottom-up estimate of the aggregate value of goods and services produced by enslaved people, adding their share of each of the major staple crops, agricultural improvements, home production, and domestic service.[5] Applying the same methodology to each of the antebellum census years, Rhode’s results are summarized in Table 1.  The bottom-line conclusion is that the enslaved produced about the same share of GNP as their share of the population.  On the one hand, one might have expected the share to be larger, because the “labor-force participation rate” of the enslaved was higher. This is economics-speak for the fact that enslaved women were compelled to do field work, while enslaved children began work in their pre-teen years.  On the other hand, most of the enslaved worked in agriculture and domestic service, where the value of output per worker was lower than the economy-wide average.

The Rhode article may be compared to another recent article, co-authored by economist Mark Stelzner and historian Sven Beckert, author of Empire of Cotton: A Global History.[6]  The objective is the same: to estimate the value of goods and services produced by enslaved workers as a share of GNP.  But the approach is entirely different.  Lacking direct aggregate data on enslaved production, the authors reason that the expected value of that production should have been reflected in the market prices of enslaved workers.  On its face, this method epitomizes model-based theoretical analysis, complete with references to “rational economic agents in a perfectly competitive economy” (144) and “present value [asset] pricing theory” (145), exactly the features that historians so often find objectionable in economic .  The approach seems particularly questionable in that slave prices reflected expectations of production value across many years into the future, whereas the objective here is to estimate the value of production in one particular year (so that it can be compared to GNP).

Despite these issues, the Stelzner-Beckert results invite comparison with those of Rhode.  Table 1 presents both sets of figures.  (The range for Stelzner-Beckert reflects alternative assumptions for the discount rate: the interest rate at which future returns are “discounted” because of their remoteness in time.)  As may be seen, both studies find that the share of the GNP produced by enslaved people was about equal to, or slightly below, their share of the population.  To some degree at least, it seems reassuring that two such different approaches yield roughly convergent results.

The share of GNP produced by enslaved workers is of course only one item in the larger conversation about the place of slavery in US history.[7]  But when a topic engages a broad segment of the public, as slavery does, mistaken or misleading factoids can have great staying power.  We cannot expect to control or curtail this process, but awareness of basic magnitudes belongs in the knowledge sets of historians of all stripes.

 

 

[1] The book itself is Robert W. Fogel and Stanley L. Engerman, Time on the Cross: The Economics of American Negro Slavery (Boston: Little, Brown, 1974).  A critique by economic historians was Paul A. David, Herbert G. Gutman, Richard Sutch, Peter Temin, and Gavin Wright, Reckoning with Slavery: A Critical Study in the Quantitative History of American Negro Slavery (New York: Oxford University Press, 1976).  For a recent overview, see Eric Hilt, “Revisiting Time on the Cross after 45 Years: The Slavery Debates and the New Economic History,” Capitalism: A Journal of History and Economics 1 (2020): 456-483.

[2] Agriculture: Nature’s Harvest https://www.searchablemuseum.com/a-nation-bound-by-slavery#section-start downloaded May 16, 2024.

[3] The Half Has Never Been Told (New York: Basic Books, 2014), p. 322.

[4]  For example, Alan Olmstead and Paul W. Rhode, “Cotton, Slavery and the New History of Capitalism,” Explorations in Economic History 67 (2018), p. 13.

[5] Paul Rhode, “What Fraction of Antebellum US National Product did the Enslaved Produce?” Explorations in Economic History 91 (2024): 1-15.

[6] Stelzner and Beckert, “The Contribution of Enslaved Workers to output and Growth in the Antebellum United States,” Economic History Review 77 (2024): 137-159. Sven Beckert, Empire of Cotton: A Global History (New York: Knopf, 2014).

[7] For my take on the role of slavery in US economic growth, see “Slavery and the Rise of the Nineteenth-Century American Economy,” Journal of Economic Perspectives 36 (2022): 123-148.