1. On Cuban railroads, see Gert J. Oostindie, “La burguesía cubana y sus caminos de hierro, 1830–1868,” Boletín de Estudios Latinoamericanos y del Caribe 37 (Dec. 1984):99-115; Levi Marrero, Cuba: economía y sociedad (Madrid: Editorial Playor, 1984), 11:139-213; Patria Cok Márquez, “La introducción de los ferrocarriles portatiles en la industria azucarera, 1870–1880,” Santiago 4 (Mar. 1981):137-47; and Oscar Zanetti and Alejandro García, Caminos para el azúcar (Havana: Editorial de Ciencias Sociales, 1987). On the labor problems encountered by Cuban railroad builders, see Oscar Zanetti Lecuono, “Esclavitud i treball lliure: el problema laboral dels ferrocarrils cubans, 1837–1867,” L'Avenc, no. 101 (Jan. 1987):17–23. I want to thank Rebecca Scott for calling my attention to this article.

2. For a listing of all Cuban sugar ingenios and the technology they employed in 1859, see Carlos Rebello, Estados relativos a la producción azucarera de la Isla de Cuba, formados competentemente y con autorización de la Intendencia de Ejército y Hacienda (Havana: n.p., 1860).

3. This figure was derived from the data on Matanzas sugar ingenios listed in the sugar plantation census conducted in 1877. See “Noticia de las fincas azucareras en producción que existían en toda la isla de Cuba al comenzar el presupuesto de 1877–78,” Revista Económica, 7 June 1878, pp. 7–24. My article focuses on the most important sugar-producing regions of Cuba located in Matanzas province. Almost all of Cuba's great nineteenth-century mills were located in the zones considered here. The economic dynamics of sugar production in Matanzas do not apply to the less capital-intensive areas of marginal sugar output in eastern Cuba. But this inapplicability does not detract from the arguments outlined here because Matanzas was the center of Cuba's nineteenth-century slave-sugar complex.

4. In the aftermath of the Ten Years' War and final abolition of slavery, Cuban sugar production was substantially reorganized. The central factories that emerged specialized in processing. Cane was largely provided by colonos, who specialized in growing cane. The Cuban colonato was highly stratified into small-scale minifundia, medium-sized properties, and large plantations. On the colonato, see the classic study by Ramiro Guerra y Sánchez, Azúcar y población en las Antillas (Havana: Editorial de Ciencias Sociales, 1976 edition).

5. This argument has been advanced by the distinguished Cuban historian Manuel Moreno Fraginals, who emphasizes it in El ingenio: complejo económico social cubano del azúcar (Havana: Editorial de Ciencias Sociales, 1978), 3 vols. He also makes the argument in an article, “El esclavo y la mecanización de los ingenios,” Bohemia, 13 June 1969; and in two chapters of his La historia como arma y otros estudios sobre esclavos ingenios y plantaciones (Barcelona: Editorial Crítica, 1983): “Abolición o disintegración?,” 50–55; and “Plantaciones en el Caribe: el caso Cuba-Puerto Rico-Santo Domingo (1860-1940),” 56–117. Fe Iglesias García varies her interpretation but stresses the incompatibility of slave labor with technological modernization of refining in several works: “The Development of Capitalism in Cuban Sugar Production, 1860–1900,” in Between Slavery and Free Labor: The Spanish-Speaking Caribbean in the Nineteenth Century, edited by Manuel Moreno Fraginals, Frank Moya Pons, and Stanley Engerman (Baltimore: Johns Hopkins University Press, 1985), 54–76; “Formación del capitalismo en la producción de azúcar en Cuba (1860-1900),” manuscript; and “Changes in Cane Cultivation in Cuba, 1860–1900,” paper presented to the Symposium on Caribbean Economic History, University of the West Indies, Jamaica, 7–8 Nov. 1986 (forthcoming in Social and Economic Studies). Rebecca Scott skillfully challenges this interpretation by examining the abolition process from a regional perspective within Cuba. See her Slave Emancipation in Cuba: The Transition to Free Labor, 1860-1899 (Princeton: Princeton University Press, 1985). The persistence of slavery in the areas of intense sugar production after abolition began in 1870 is emphasized.

6. The use of the term productivity in this article refers to the income produced by each plantation. The documentary material utilized does not differentiate between agricultural and industrial sectors; no distinctions are made between output produced by slave, Chinese, or free labor, nor is productive output by task indicated. Although it would have been ideal to have access to these kinds of data, their nonexistence means that productive units must be considered as totalities, rather than according to their components. These factors do not compromise the conclusions of this article for reasons outlined in the text and subsequent notes.

7. The data for this article are contained in the following sources: for Colón, Archivo Nacional de Cuba (hereafter ANC), Miscelánea de Expedientes (hereafter ME), legajo 4120, no. M, “Repartos municipales de la jurisdicción de Colón, 1859”; ANC, Gobierno General, legajo 405, no. 19209, “Padrón de fincas rústicas de la jurisdicción de Colón, 1865”; ANC, Gobierno General, legajo 270, no. 13563, “Padrón general de fincas rústicas de este distrito, año de 1875 a 1876.” The data for Cárdenas are in ANC, Gobierno General, legajo 945, no. 16724, “Padrón general de la riqueza rústica para regir en los años económicos de 1866 a 1867”; ANC, Gobierno General, legajo 269, no. 13554, “Jurisdicción de Cárdenas. Padrón general de la riqueza rústica de esta ciudad y su jurisdicción formado para los años económicos de 1875 a 1876.” The 1877–78 sugar census is found in “Noticias de las fincas azucareras en producción que existían en toda la isla de Cuba al comenzar el presupuesto de 1877–78,” Revista Económica, 7 June 1878, pp. 7–24. Unless otherwise indicated, all of the data and statistical tables presented in this article were derived from these sources.

8. Refacción contracts were the main means used by planters to secure credit and supplies for their plantations at harvest time. They were also the legal mechanism that tied planters to particular merchants for marketing sugar and molasses. These contracts stipulated several important points. First, merchants would supply the capital and supplies needed to maintain the ingenio. This amount would include salary disbursements to free workers, capital to purchase slaves if needed, food and clothing for the slave population, capital for the medical maintenance of the slave population, all kinds of supplies (from containers to hold processed sugar to the most elementary items such as nails and lumber), and transportation services to move sugar from plantation to port or capital to pay for transportation. Second, planters usually mortgaged some or all of their harvests to the merchant granting the refacción contract. Planters were forbidden to market their products to other merchant houses but were usually paid the current price in the port at the time of delivery. Warehousing fees, commissions on marketing, and transportation costs were deducted at this time. The terms of these contracts are important to take into consideration, especially when the 1877–78 census is examined. This document listed refacción contract expenses for each plantation. Usually, 65 percent of gross income was deducted for refacción costs to determine net income, although in some cases 50 or 60 percent was deducted. The importance of this procedure is that slave maintenance costs were thus deducted from gross income, allowing this variable to be factored into the formula to determine net income per slave for 1877–78.

9. On inflation in the 1870s, see Fe Iglesias García, “Azúcar y crédito durante la segunda mitad del siglo xix en Cuba,” Santiago 52 (1983): 119–44.

10. For sugar prices in the middle of the 1859 zafra, see Biblioteca Gener y del Monte, Matanzas, Aurora del Yumurí, 23 Mar. 1859; and for 1878, see Biblioteca Nacional José Martí, Gaceta de la Habana, 1 May 1878. Evidence that local sugar prices were quoted in gold, rather than in paper money, is contained in the invoices sent to U.S. importers of Cuban sugar. See, for example, New York Public Library, Moses Taylor Collection, Individual Letters for 1870 (formerly boxes 86 and 123), invoice of Zaldo and Company to Moses Taylor, 1 Jan. 1870.

11. A fundamental problem in determining slave profitability is ascertaining the ratio of productive or prime-age slaves to unproductive or young and old slaves who contributed marginally to production. Unfortunately, data on the age structure of the slave population in these jurisdictions are lacking for the years under consideration.

12. The data for Matanzas, which shows 83.5 percent of all workers as slaves, is suspect. With the exception of Ceiba Mocha, Matanzas partidos (jurisdictional subdivisions) did not report on rented or freed slaves. The only other category noted was Chinese contract laborers. Thus the number of slaves as a percentage of all workers was in all likelihood artificially inflated.

13. The economic data on which this study is based, covering ingenios in Colón and Cárdenas between 1859 and 1878, may be obtained by writing to the author directly.

14. A large number of mills in Colón did not report information on their work forces for the 1877–78 census, while all mills reported income data. In order to calculate income per worker without statistical distortion, only those estates reporting data on their labor forces were utilized. They accounted for 70 of 116 sugar ingenios.

15. See Public Records Office, FO 84/965, for this document. I would like to thank David Eltis, who provided this indirectly via Stanley Engerman.

16. See Stanley Engerman, Manuel Moreno Fraginals, and Herbert Klein, “The Level and Structure of Slave Prices on Cuban Plantations in the Mid-Nineteenth Century: Some Comparative Perspectives,” American Historical Review 88, no. 5 (Dec. 1983): 1201–18.

17. See David Eltis, Economic Growth and the Ending of the Transatlantic Slave Trade (New York: Oxford University Press, 1987), Appendix C, “Prices of Slaves in the Transatlantic Slave Trade after 1810,” 261–64.

18. Data on slave sales were extracted from the notarial protocols of Colón stored in the Archivo Histórico Provincial de Matanzas (hereafter cited as AHPM). See AHPM, Protocolos Notariales, Manuel Vega Lavarría, 1863, 1864, 1867, 1870, 1872, 1875, 1876, 1878, 1880. More extensive information on slave prices is contained in Laird W. Bergad, “Slave Prices in Cuba, 1840–1875,” Hispanic American Historical Review 67, no. 4 (1987):631–55. The inflation of the 1870s has been accounted for in the slave prices noted here. Slave transactions were listed in gold or paper money, and the ratio was 2 paper pesos to 1 gold peso. All values were adjusted to reflect the gold price.

19. In an economic report on the sugar industry sent by the British consul in Havana, Joseph Crawford stated that the average productive life of an individual slave was twenty years. He also noted a 5 percent yearly death rate among slave populations. See Public Records Office (London), FO 72/748, Joseph T. Crawford to Viscount Palmerston of the Foreign Office, 28 Jan. 1848.

20. Although no data on income per worker is available for 1870, between 1865 and 1876, income per worker increased in Colón 4.7 pesos per year from 309.6 to 361.6 pesos. Thus in the five years between 1865 and 1870, an increase of 23.5 pesos can be estimated, yielding an estimated income per worker of 333.1 pesos in 1870.

21. No existing formulas can readily convert the gross income generated per slave into profitability. Sufficient data are not available to estimate accurately rates of return on slave labor over time. A different series of data would be needed to calculate such figures, including prices, depreciation, and replacement costs of machinery, tools, animals, and slaves; land values and depreciation; maintenance costs of slaves; costs of supplies, warehousing, commission fees, taxes; and the reproductive rate of the slave population. Moreover, no data on net income is available for any year except 1878, nor is there any way to calculate net income based on the formula used in the 1878 documentation. Estimates on profitability and costs were made by Juan Poey, owner of Ingenio Las Cañas, one of Cuba's largest mills in the early 1860s. He estimated a 2.5 percent depreciation of slave values due to death and presented a detailed array of economic calculations deriving a final yearly loss of 4.13 percent on investment capital on an “average” ingenio cultivating seventeen caballerías of cane with 132 workers. The 7.7 workers per caballería are strikingly close to the estimates noted in this article. See Juan Poey, Informes presentados al excmo. capitán general Gobierno Superior Civil de la isla de Cuba sobre el proyecto de Colónizacion africana y al Illmo. Sr. Intendente de Hacienda de la propia isla sobre derechos de los azucares (Madrid: Imp. de la Compañía de Empresas y Libretos a cargo de D. A. Aurial, 1862), 141–45. These data, however, must be viewed with caution because Poey's purpose was to persuade Spain to reduce taxes on sugar exports and imports and garner support for his ill-fated project to import “free” African laborers. Planters were always complaining about losing money while increasing production, constructing lavish homes, and leading lives of luxury. As a result, their printed testimonies on the internal economic dynamics of their estates cannot be used reliably. Until detailed internal account books are located and scrutinized, rates of return on investments in slaves cannot be accurately ascertained.

22. In order not to distort these data, only mills with information on the number of slaves were utilized for this comparison. If the 1878 data on Colón are compared with the 1876 data, there seems to be a drop in the total number of ingenios. This decline is illusory because there were a large number of mills for which data on the labor force was missing for 1878, and thus they were not included in these calculations.

23. See the accounts of Cosme de la Torriente's Ingenio La Reforma from 1866 through 1871 in AHPM, Gobierno Provincial, Ingenios, legajo 3, no. 38; and Mauricio Alfonso's Ingenio la Vega in 1874, legajo 3, no. 37. In Slave Emancipation in Cuba, Scott found wage rates in the late 1870s to range between 22 and 27 pesos per month (p. 119). With respect to the critical question of the comparative labor productivity of free versus slave labor, the identical monthly salary rate for free labor (25 pesos) and the hire rate plus maintenance cost for slaves (17 plus 8 pesos) strongly suggest that planters perceived little difference in labor productivity between the two. I want to thank an anonymous reader for suggesting this important point.

24. This period was one of active ingenio construction in the jurisdictions considered in this study. Between 1865 and 1876, the number of mills in the Colón partidos of Macuriges, Hanabana, Macagua, and Palmillas increased by over 45 percent, from 73 to 106 percent. See ANC, Gobierno General, legajo 405, no. 19209, “Padrón de fincas rústicas de la jurisdicción de Colón, 1865”; ANC, Gobierno General, legajo 270, no. 13563, “Padrón general de fincas rústicas de este distrito, año de 1875 a 1876.” In a communiqué written in 1880, the British vice-consul in Matanzas noted that “the country round Matanzas has suffered little from the late insurrection, and that the sugar industry there is in a comparatively thriving condition.” See Public Records Office, FO 72/1576.

25. For sugar prices, see issues for the corresponding years of the Aurora de Yumurí in the Biblioteca Gener y del Monte, Matanzas, and the Gaceta de la Habana in the Biblioteca Nacional José Martí in Havana.