4 Slavery and the Economy in the Nineteenth Century

The evolution of Brazilian slave society in the nineteenth century is defined by several important developments. The first was the emergence of a new export crop, coffee, which would form the basis of a new slave plantation economy in the southeastern states. The second was the continued growth of the traditional colonial plantation crops. Sugar again became a world-competitive product with the decline of the sugar industries in both Saint Domingue and the British West Indies, and Brazilian cotton production, which, after suffering severe competition from the southern United States in the first half of the century, revived in the period of the world's cotton famine in the 1860s during the North American Civil War. Finally, there was also the continued growth of the slave population of Minas Gerais – the largest in Brazil – in a region that was primarily dedicated to production for the internal market and yet that retained and expanded its slave population even in the era of the internal slave trade.

The growth of all these traditional and newer export crops created an increasing demand for slaves, and the Atlantic slave trade reached its peak in the third decade of the nineteenth century. But the closing of the trade in 1850 led to both a secular rise in slave prices and the consequent expansion of the internal slave trade by both land and sea in the next two decades, until the interprovincial slave trade was finally abolished in the last decade before emancipation. All of these economic and social developments led to a progressive shift of the slave population within Brazil to the more dynamic regions from the less thriving centers and from the urban to the rural areas. With the ever-increasing expansion of free colored workers, the need for urban slaves declined. In fact, the very rapid growth of the free colored population in the nineteenth century meant that it became the largest single racial-legal group in the empire by the first census of 1872. Despite this growth of the free colored, plantation export agriculture in most regions remained primarily based on slave labor until the end of slavery.

The final major theme of the nineteenth century would of course be the slow abolition of slavery itself. Realizing that the institution could not survive in the world of the nineteenth century – especially after the Civil War eliminated African slavery in the United States – Brazil slowly moved toward emancipation of all slaves. In 1872, a law of free womb was declared and funds set up to purchase the freedom of slaves¹; in 1885 all slaves older than 65 were freed² and final abolition came sixteen years later. In anticipation of eventual abolition, and because of a declining slave population after 1850, by the last quarter-century before abolition, Brazil was slowly shifting its labor base to a more complex mix of slave and free labor in most regions and industries and, except for Minas Gerais, the declining number of slaves was ever more concentrated in export agriculture. Although abolition would come as a shock to the sugar and coffee economy of the southeastern region in 1888, the transition to a new form of wage labor based on new immigrant groups would be relatively quick, with only a minimum of disruption in the export sector.

All these complex developments would see the shift of the slave population by region from their eighteenth-century distribution. The regions of the Northeast and of the South lost their relative importance as places of slave residence, and the three coffee provinces of Rio de Janeiro, Minas Gerais, and São Paulo took primacy as the slave centers of late nineteenth-century Brazil (see Table 4.1). Although the internal trade was not of the same size as the Atlantic slave trade, it was sufficient to shift the balance of slave population to the three states of Minas Gerais, Rio de Janeiro, and São Paulo. Traditional producers for the internal market as well as planters in regions with abundant sources of landless free workers were forced to slowly abandon their exclusive reliance on slave workers and adopt mixed slave and free labor or even based themselves exclusively on free wage workers. Although the internal trade would eventually be brought to a halt, by then the center of slavery had dramatically shifted

TABLE 4.1. Change in Relative Importance of the Slave Population by Province, 1819, 1872, 1886/87 (sorted by % importance in 1886/87)

Notes:

* Includes Imperial Court city of Rio de Janeiro.

** No data available for Amazonas in 1886–87.

Source: 1819: Norberto (1870), p. 152; Census of 1872; Ministério da Agricultura, Relatório 1887, p. 24.

toward the southeastern coffee and sugar plantations. Whereas Minas, Rio de Janeiro, and São Paulo accounted for only 35 percent of the total slave population in 1818, by 1872 they had increased their share to 58 percent and continued to increase to the end of slavery, arriving at 65 percent by 1886–87.

The Afro-Brazilian population would expand dramatically in the nineteenth century in terms of both slaves and free persons. The century began with the volume of African importations increasing from decade to decade until the 1840s (see Graph 4.1). At the same time, the growth of the native-born slaves, now all of African descent, kept increasing as well, so that the age and sex biases within the resident slave population became less distorted by the impact of the Atlantic migrants, although

GRAPH 4.1. Estimated average annual African slave arrivals in Brazil by decade, 1781–1850.

Source: Emory data set, accessed August 18, 2008.

TABLE 4.2. Population by Color and Status in 1872; Provinces Ranked by Total Colored Population Size

Notes: * Includes the Munícipio Neutro (city of Rio de Janeiro).

Source: Census of 1872.

the resident slave population still favored adults and males over women and children. Finally, the free colored population increased faster than the slave population, and by the time of the first national census in 1872, it had become the single largest group in Brazil (see Table 4.2).

The first major development in the Brazilian slave-based economy in the nineteenth century was the renewed expansion of the sugar plantation slave economy. The collapse of Haitian slave production and the mid-nineteenth century decline of British West Indian sugar production had an impact in reviving local production in the traditional northeastern region as well as in the newer centers of the Southeast. For Brazil, the Haitian collapse came at a time of a classic export crisis in Brazilian history. The last decades of the eighteenth century had seen a decline of both the gold and diamond mining industries in the central interior, while the sugar economy found itself in serious competition with the booming French and British producers of the West Indies.

Thus, the immediate impact of the decline of Saint Domingue – a powerhouse of multiple slave-produced products – was to give new life to old industries such as sugar and cotton, as well as open up the possibility of new slave-producing industries for the colony. Within a decade, sugar production surpassed its old 15,000 to 20,000 tons per annum limit as world prices and demand began a long upward secular trend and Brazilian production responded, reaching into the 100,000-ton range by the late 1840s.³ The Haitian impact thus intensified the plantation system of the old Northeast by increasing the number of plantations and slaves in sugar production and by encouraging the expansion of the sugar fields in Rio de Janeiro and São Paulo. International demand was so intense that Brazil found itself once again competing on a world market. Confined mainly to Portugal and the Mediterranean for most of the eighteenth century because of restrictive tariffs on the part of Britain and France, Brazilian sugars again began to penetrate central and northern European markets, and above all, to supply the previously closed English market. Production also grew so steadily from decade to decade that Brazil once again moved into a leading position as a major world producer, accounting for 15 percent of world output by 1805, and until the 1830s, sugar remained Brazil's leading export in terms of value.⁴ Although Cuban sugar had taken the lead early in the century, Brazil became America's second largest producer, especially after the crisis of emancipation disastrously affected British West Indian production in the 1830s. In the early 1840s, when its production was just half of Cuban output, Brazil accounted for more than one-fifth of American production and 13 percent of world cane exports (see Graph 4.2).⁵

The expansion of world sugar production throughout the nineteenth century, especially to new regions, meant that Brazil's ratio of total world cane sugar production declined in importance despite its increasing production. Several major American producers now entered the cane sugar market, including Cuba and Puerto Rico. Even Peru and Mexico began exporting sugar in the second half of the nineteenth century with the use

GRAPH 4.2. Relative share of Brazil in world sugar production.

Source: Fraginals, El Ingenio, I, pp. 40–2; II:173.

of free Indian and Chinese indentured labor. Cuban competition especially had a profound impact on both prices and shares of the European markets attained by the Brazilians. There was also the growth of sugar production in Asia, which began to be a serious competitor to American cane production. African slaves were used by the French and British to produce large quantities of sugar in their Indian Ocean island possessions, but free labor was also used in India, Java, and later in the Philippines for sugar production. Even more important was the growth of the European beet-sugar industry, which fully came into its own in the 1850s. The beet-sugar industry, which accounted for roughly 15 percent of the world sugar market in mid-century, took half of the world sugar market by the 1880s. Nor could Brazil find an alternative market in the United States, as Louisiana production also expanded to meet national needs, with the shortfalls in U.S. consumption supplied by Cuba and Puerto Rico.⁶ Nevertheless, production remained a relatively stable one-tenth of world production until the end of the century. Thus, despite the major increase in the value and volume of Brazilian sugar production, its relative share of the international sugar market declined, and by the end of the century, Brazil accounted for only 3 percent of total world sugar output.⁷ This relative decline continued until well into the following decades, and Brazil would not recover its primary role as the world's leading cane-sugar producer for the world market until the late twentieth century (see Graph 4.3).⁸

GRAPH 4.3. Brazilian sugar cane production as a percentage (with linear trends) of world cane production and world sugar (cane + beet) production, 1839–1909.

Source: Deerr, The History of Sugar, II, pp. 113, 490.

Despite its relative decline in world markets, Brazilian sugar production expanded for most of the century. By the 1820s, national output was up to 40,000 tons and climbed to 70,000 tons by the next decade. A decade later, it was up to the 100,000-ton range, at which point it would remain for the next two decades as world prices were buffeted by the entrance of beet sugar into the European market (see Graph 4.4). But expansion got under way again with favorable world prices so that by the 1870s, Brazilian production averaged 168,000 tons and by the last decade of slavery output had climbed to more than 200,000 tons.⁹

Although the relative importance of Brazilian sugar on the world market and in the mix of Brazilian exports declined, the total value of sugar exports increased along with output until the early 1880s (see Graph 4.5). This meant that the sugar plantation slave economy actually increased throughout most of the nineteenth century, although there was a slow shift in production zones as southeastern producers became even more important in the total production picture.

The most intensive growth of the sugar industry in the nineteenth century initially occurred in the older regions of the Brazilian Northeast, with Bahia and Pernambuco leading all other regions. Although the otahiti variety of sugar cane was introduced into Brazil at this time, just as it had been into Cuba, there were no other major technological inventions in the industry. Mills were not changed in structure, nor were steam engines or railroads introduced until late into the nineteenth century.¹⁰ Average output per mill in the northeastern sugar heartlands remained the same as it had been in the colonial period until the late nineteenth-century conversion to steam-driven mills. Thus, increased output in the key northeastern states was not based on major new technology for most of the nineteenth century, but rather on the expansion of the sugar zone into new lands beyond the traditional Recôncavo district and the increase in the number of mills. The existence of a growing national market successfully cushioned the local industry from severe world price shocks.

There was also a major shift in northeastern zones of production in this period. Between 1790 and 1820, Bahia doubled its mills to more than five hundred and increased its slave population to nearly 150,000 persons, with an important part of this population dedicated to nonsugar

GRAPH 4.4. Estimate of Brazilian sugar production output and linear trend, 1760 to 1909.

Source: Deerr, The History of Sugar, II, p. 113.

GRAPH 4.5. Average annual volume (in metric tons) and value (in pounds sterling) of Brazilian sugar exports, 1821–1906.

production as the economy diversified and expanded.¹¹ It also contained the largest of the sugar engenhos in Brazil. A unique census of the district of Santiago do Iguape in Bahia in 1835 showed that this famous sugar production center contained 22 plantations with an average of 123 slaves per plantation, probably the largest such sugar estates in Brazil.¹² By this date, Bahia alone was exporting close to half of the Brazilian output. Thereafter production slowed, and by the late 1840s, Pernambuco production surpassed Bahian output. The Bahian sugar industry revived in the last quarter of the century, with major capital inputs finally bringing steam mills to more than three-quarters of the province's engenhos. It is estimated that by 1877 there were 802 engenhos in Bahia and about 800 in Pernambuco.¹³

The impressive growth of Pernambuco in the nineteenth century had its origins in the revitalization of the local economy carried out in the late eighteenth century. The work of the Pombaline monopoly company in Pernambuco had been effective and placed the province in an advantageous position to respond to the post-1791 boom in sugar prices. There was an expansion of mills in both the traditional and frontier areas; at the same time, the slave trade became quite intense, and the local slave population increased to almost 100,000 by the second decade of the nineteenth century. Sugar production expanded with each passing decade, and at mid-century, Pernambuco had passed Bahian levels. By the mid-1880s, it was producing more than 100,000 tons of sugar and accounted for almost half of Brazilian exports. This growth was achieved with a declining slave population. In the 1850s, at the closing of the slave trade, Pernambuco had 145,000 slaves; by the census of 1872, this dropped to 106,000, and it declined further to 85,000 in the next decade. ¹⁴ The growth of the free colored population more than made up for this decline, some of which was due to the post-1850 expansion of the internal slave trade that resulted in an estimated 23,000 to 38,000 Pernambucan slaves being shipped south into the southeastern coffee plantations between 1850 and the beginning of the 1880s.¹⁵ Already by the 1850s, the plantations in the richest of Pernambuco's sugar zone were averaging seventy slaves and forty-nine free-wage workers in their labor force.¹⁶ This ratio of free workers increased just as the introduction of steam increased sugar output per worker, and northeastern planters had little difficulty replacing slaves with poorly paid and often seasonal wage laborers. Moreover, in Pernambuco, as in Bahia and throughout the northeastern sugar zone, the increasing complexity of milling led to the greater importance of independent planters (lavradores de cana) with their own slaves who supplied cane to the mills as well as the increasing use of free-wage workers, seasonal labor, and even squatters.¹⁷ Although few statistics exist on the lavradores, many of them were clearly poorer farmers. In a census of the Recôncavo parishes in 1816–1817, there were listed 165 senhores de engenhos, or mill owners, and 478 lavradores de cana, or cane farmers without mills. The lavradores averaged just ten slaves each, and a census of 1835 of one Bahian district suggests that almost a quarter of these cane producers held no slaves and that some 44 percent were free colored.¹⁸

A most impressive growth in the sugar industry in the post-1791 period was also registered in the province of Rio de Janeiro. Around Guanabara Bay and in the interior lowlands of Campos, a major industry developed. By the early 1820s, there were more than 170,000 slaves in the province, and in Campos alone, there were some 400 or so sugar estates with an average slave labor force of 36 slaves, or just under less than the contemporary Bahian rate of 66 slaves per engenho.¹⁹ Sugar engenhos in this south central region in general used the same technology as in the Northeast but were on average smaller than those of the Northeast. In the middle decades of the century, however, there was considerable growth in the sugar industry. Finally, in the last quarter of the nineteenth century, there occurred an important change in the sugar-producing areas with the creation of the first steam-driven mills (engenhos centrais), a revolution in technology that had already occurred in other countries. With the support of the government, which guaranteed loans to construct these steam-driven mills, this new technology permitted Brazil to again become a player on the international sugar market. It also led to profound changes in the organization of labor. The new and very expensive industrial central mills, which were eventually established in the more advanced areas, were operated by free wage labor, and this led to the complete concentration of slaves in cane production. Numerous central mills were created not only in the Northeast and in Rio de Janeiro, but even in São Paulo. However, the initial transformation was only partial, as the steam technology was often applied to old mills without a complete transformation in the technology into what were called central mills, with all their technical innovation in processing. There also seems to have been problems with a sufficient supply of cane to these new steam mills, because in contrast to the traditional mills, they depended heavily on others to furnish cane. But this supply of cane from the older engenhos did not occur as expected, in part because of transport problems, but also because of the boycott of the traditional and powerful senhores de engenhos who saw themselves being transformed into simple suppliers of cane and owners of modest mills that could no longer produce sugar. Thus, the central mills, despite the support of the government, did not have the success expected, but they did represent the beginnings of the process of destruction of the old social order that was based on the traditional mills. Eventually there would be the rise of the usinas, or large-scale industries based on modern technology, that produced almost all of the cane it processed, thus being immune to the boycott of the old planters. With the usinas and their larger scale of production came new social groups in command of the sugar industry. This slow transformation would also represent a rupture in the structure of existing power relations even before the end of slavery in the sugar-producing regions.²⁰

The final region to become a significant producer in this period was the captaincy of São Paulo. Along the coast near the port of Santos and in the previously mixed farming region around the city of São Paulo, sugar now began to be produced for world export. Although São Paulo always ranked a poor fourth in national output and accounted for no more than 5 percent of national production, sugar proved vital to the paulista economy. Sugar immediately became São Paulo's most valuable export, and even though output barely climbed into the 1,000-ton figure in this early period, it already accounted for well over half the value of all provincial exports. By the 1820s, sugar was the province's primary export and was then in the 5,000- to 10,000-ton range. Sugar exports continued to expand into the late 1840s, along with a major production of aguardente for which both São Paulo and Rio de Janeiro became known, especially in the trade to Africa for slaves.²¹ Wherever sugar was established it eventually developed as plantation crop, and it also had the highest number of slave workers per agricultural unit (see Table 4.4). Usually because of high start-up costs and the quick maturation of the plants, planters concentrated primarily on growing sugar cane in most regions of the world. But in southeastern Brazil, such sugar estates were fairly unusual by American standards in that they often continued to produce food crops as well. During the course of the nineteenth century, the introduction of new technology from the more advanced Caribbean producers greatly increased the investments and output of the sugar mills, all of which increased the need for even more sugar cane. As can be seen in early nineteenth-century data (see Table 4.3), the northeast still had the largest sugar estates in terms of average number slaves per unit – almost double that of even the premier sugar-producing center of Campinas in São Paulo, although as can be seen in the changes from 1804 to 1829, the southeastern producers were beginning to approach these traditional northeastern big estate levels. But none of these zones approached the average of more than two hundred slaves per unit, which was the norm in Jamaica.²²

TABLE 4.3. Slaves on the Sugar Engenhos in Various Regions

* Just the engenhos.

Source: Arquivo Público do Estado de São Paulo, Mapas de população; Schwartz (1985), p. 364; and Costa (1982).

Although free persons were employed as specialized workers or as supervisors of slave labor on the typical sugar estates, the majority of the workers, either in the field or in the mills, were slaves. There was of course a difference in skills and in autonomy between slaves employed in agricultural tasks and those dedicated to the industrial process. In the fields, the work was heavy, little specialized, usually executed in gangs, and with a significant participation of women. The only specialized tasks in this area were supervision of the field workers. In the mills, the work was organized into specialized tasks categorized by skills and physical difficulties. Although some of these more specialized tasks involved free workers, the majority of work, even those jobs that were most complex and required greater training, were carried out by slaves who were of great value to their owners, which was reflected in their market prices. One of the best studies of these labor divisions on a northeastern sugar estate comes from the eighteenth century but was typical as well of those of the nineteenth century. There was a large group of specialized workers, and many of them were slaves.²³ In a sample of engenhos from the Bahian Recôncavo in the eighteenth century, among a group of 1,331 slaves, the majority worked in the fields (984), in the mills (126), and in transport (87); another 42 were artisans and 10 had supervisory functions, while 82 were employed as household domestic slaves.²⁴ A study of sugar engenhos in Minas Gerais in the 1830s found that among a sample of 589 slaves employed in sugar, 326 were field hands, 183 were artisans, 39 in domestic service, 28 in transport, and 13 in other occupations. In this case, an even higher percentage of slaves were to be found in skilled and semi-skilled labor beyond the unskilled field workers.²⁵

Whereas the northeast remained concentrated in sugar and to a lesser extent in cotton during most the nineteenth century, the regions of Rio de Janeiro and São Paulo dramatically and rapidly opened up a new plantation crop that quickly competed with traditional sugar production. In terms of structural change and growth, it was not sugar production but rather coffee that was most affected by the Haitian experience. Although coffee had been produced in Brazil since the early eighteenth century and was already a minor but growing export at the end of the century, the halving of Haitian production in the new century and the growing demand for coffee in the North American and European markets created a major demand for new American production. It was coffee, greatly pushed by the collapse of Saint Domingue, that would be the slave crop par excellence in nineteenth-century Brazil.

Although coffee estates in the second half of the nineteenth century would finally rival the big sugar producers in terms of slave workers, they often started as mixed farming enterprises and for several decades were based on relatively small slaveholdings or were even produced by free wage laborers and family farmers. Given the long period of gestation for the coffee trees to mature, it was the norm for alternative food crops and other commercial products to be produced on the coffee estates. Even after these estates grew quite large and their trees were in full production, food crops were still planted in significant quantities even after the abolition of slavery.

The production of coffee in Brazil had initially been widely dispersed over the colony. But it was in the captaincy of Rio de Janeiro that the beans became a major product. What is impressive about this growth of coffee production in Brazil was how late it was in terms of American coffee development, how quickly Brazil came to dominate world production, and how concentrated the plantations were within Brazil itself. Thus, in a complete reversal from the experience of sugar, it was from the West Indies that Brazil learned to cultivate coffee. First from Saint Domingue and later from Cuba, Rio de Janeiro planters learned the techniques of producing coffee on a commercial scale. It was the combination of the crisis created by the elimination of Saint Domingue and the post-1815 rise in European and North American demand that sent prices rising, which finally got the industry into its mature phase.²⁶

Before the end of the Napoleonic Wars, production was negligible, and even as late as 1821 the planters of Rio de Janeiro were exporting no more than 7,000 tons. This was a third of Cuban and Puerto Rican output and was nowhere near the 42,000 tons that Saint Domingue had been producing in 1791. Even within the province itself, coffee did not replace sugar as the most valuable export until the 1820s. But in this decade the industry began its dramatic growth. In 1831 coffee exports finally surpassed sugar exports for the first time in Brazil, and they finally surpassed the tonnage record of Saint Domingue set in 1791. By the middle of the decade, Brazil was producing double the combined output of Cuba and Puerto Rico and was the world's largest producer. In the 1840s, output climbed to more than 100,000 tons per annum and doubled again to more than 200,000 by the 1850s (see Graph 4.6), by which time Brazil had become the world's largest producer and accounted for more than half of world production.²⁷ By the early 1830s, it had become the single most important export (see Graph 4.7) and had replaced sugar and cotton as the most valuable Brazilian commodity being exported, a position it would dominate for the rest of the century.

Although the costs of entrance into coffee production permitted participation of non-slave-owning farmers, coffee was produced by slave labor from the beginning. Moreover, as farmers increased their share of coffee production in their mix of outputs, slaves became an increasingly important part of the coffee labor force. Thus, the growth of coffee in the central provinces of Rio de Janeiro, Minas Gerais, and São Paulo, the top three producers, was closely associated with the growth and expansion of the Atlantic slave trade to Brazil, which reached its largest volume in the nineteenth century. The expansion of the coffee frontiers up from the coastal valleys and into the interior highlands was also typical of a slave plantation economy. Virgin lands were the crucial variable in determining productivity and, with no serious fertilization carried out, soil exhaustion made for a continuously expanding frontier. From the 1820s to the late 1860s, the central valleys of Rio de Janeiro were the core zones of exploitation, with the Paraíba Valley being the heartland of the new industry. From there it spread westward into the southeastern region of Minas Gerais, whose declining mining economy was revived first by sugar production and then by coffee. By the 1860s, local production expanded so rapidly that Minas Gerais temporarily replaced São Paulo as the nation's second-largest producer, with more than one-fifth

GRAPH 4.6. Average annual volume (in metric tons) and value (in pounds sterling) of Brazilian coffee exports, 1821–1910.

GRAPH 4.7. Relative share of major crops in total value of exports, 1821–1860.

of total coffee exports. For most of the 1860s and early 1870s it gained ground on Rio de Janeiro and maintained its lead over São Paulo.²⁸ It was only in the late 1870s and the 1880s, at the very end of the slave era, that the coffee frontier finally moved into the West Paulista plains area, and former sugar plantation regions like Campinas became the centers of coffee production. Even abolition and a shift to free labor did not stop the moving coffee frontier, which by the end of the nineteenth century had finally reached south of São Paulo into the province of Paraná.²⁹

In its earliest days, coffee was produced on relatively small farms in the coastal region of Rio de Janeiro and in the interior Paraíba Valley. Initially coffee was produced on mixed crop farms and only slowly began to dominate local production. Given the long period for the trees to mature – coffee trees only began producing beans in the third or fourth year of growth – farmers reduced their risk by producing corn and other crops besides coffee. Even after major fazendas were created with hundreds of thousands of trees, intercropping of corn and other foodstuffs were common to all coffee plantations. There was wide variation in soil quality and in the ages of trees, which made annual production vary quite dramatically among plantations. In a study of representative São Paulo coffee municipalities, average production rose from 400 kilos per worker on the earliest farms of the 1820s to several thousand kilos per worker a decade or two later. In the mature phase of the industry in the last quarter of the century, the average field hand produced just over 2,000 kilos of coffee per annum.³⁰ The number of slaves per coffee estate throughout Brazil slowly increased over the rest of the century, so that by the early 1880s, a survey of some 707 coffee fazendas in the leading producing province of Rio de Janeiro indicated an average of 43 slaves per unit.³¹ This was a size of holding that was still lower than the average of 128 slaves per coffee estate achieved in Jamaica in 1832,³² a figure approached only by a minority of Rio de Janeiro and São Paulo plantations even in the last decade of slavery.

The search for better and virgin soils was constantly drawing the coffee frontier inland. The interior valleys of the province were heavily forested and contained excellent soil. For this reason, initial output from the coffee trees was extremely high in the first fifteen or so years. However, denuded of forest and improperly planted, these steep valley lands were subject to soil erosion and rapid decline of productivity of their coffee groves. Thus, a boom–bust cycle accompanied coffee in these early centers. Typical of this first stage was the interior fluminense valley of Paraíba and its central district of Vassouras. Initially settled in the 1790s, Vassouras did not develop coffee fazendas until the 1820s. Yet the richness of the local soils, the high prices for coffee on European markets, and the availability of large amounts of capital and labor previously generated by the sugar plantation system allowed for the development of a new type of coffee plantation regime. It was in Vassouras that these first large coffee plantations made their appearance.³³

During the course of the nineteenth century, the productivity of slaves increased as more stable virgin lands were opened in the highland plateaus and as more experience in planting was developed. It was estimated that an average output of a slave at the leading fazendas was close to thirty-four sacks of coffee per annum.³⁴ In the pre-railroad era – that is, in the period before the 1850s – transport costs were a large part of the final price of coffee, so a large percentage of the slave labor force was engaged in the moving of the coffee sacks to market on the backs of mules. A good one-third of the slave labor force on a coffee estate prior to the introduction of rail connections was off the plantation at any given time transporting goods to and from the distant port markets. The railroads eliminated these mule trains in the second half of the nineteenth century, which were replaced by ox-driven carts and feeder roads that led to the nearest railroad sidings.³⁵ This revolution in transport reduced costs considerably but did little to change the actual structure of the labor force on the coffee fazendas. In surveys from coffee plantations, the number of field hands never reached more than 58 percent of the total number of slaves on a given estate.³⁶ These field hands were divided into gangs and driven under supervision of local slave or white overseers. Just like the sugar estates of the Caribbean and the rest of Brazil, it turned out that the majority of workers on these field gangs were women. In coffee as in sugar, it was men who were exclusively given all the skilled occupations, and it was men who were underrepresented in the unskilled field tasks of planting, weeding, and harvesting. Given the large pool of free black and white labor available even in the most densely settled coffee zones, it was left to hired free laborers to do all the dangerous tasks of clearing virgin forests, a task usually reserved for male slaves in the British and French Caribbean.

The growth of the coffee fazendas and of the slave population devoted to them was not confined to the traditional areas but was constantly on the move throughout the Paraíba Valley complex. Older zones with aged trees and a low percentage of virgin forest lands, such as Vassouras, saw declining numbers of slaves in the working-age categories as these younger workers were exported to the newer production zones. Thus, although an intense African slave trade kept the fazendas supplied with slaves until the 1850s, reaching a high of fifty thousand African arrivals in all of Brazil per annum in the 1820s (see above Graph 4.1), most of the post-1850 plantation growth was accomplished through an intense internal slave trade migration of slaves at the intercounty, interprovincial, and interregional levels.³⁷ As older zones like Vassouras matured, their ratio of Africans among the slaves declined, the average age of the slaves declined and ratio of men to women became more balanced as the population became more native born. Also, many of these older and less efficient producers began losing slaves to the expanding coffee producers in the interior.

The key new growth state for coffee was the province of São Paulo. A late exporter of any major products, São Paulo initially entered the sugar race as a minor producer but ranked fourth in total output by the early decades of the nineteenth century. By the late 1830s, the province held 79,000 slaves, most of whom were in rural occupations, and it was sugar that was the primary occupation. But in the 1840s, coffee finally passed sugar in importance. In that decade the number of slaves in sugar probably numbered twenty thousand, but those attached to coffee fazendas reached twenty-five thousand, while the province as a whole accounted for almost one-fourth of national production with its 53,000 tons of coffee. Shipping half of its coffee sacks through the port of Rio de Janeiro and the other half from its own coastal ports, São Paulo's provincial production finally passed mineiro levels by the late 1840s and began to approach Rio de Janeiro output by the 1880s. By 1854, São Paulo had more than 2,600 coffee fazendas worked by some 55,000 slaves, with the average estate holding just over 20 slaves (see Table 4.4).

Coffee not only moved southwest into São Paulo from Rio de Janeiro but also northwest into the southeastern region area of Minas Gerais known as the Zona de Mata. In the late 1820s, Minas was no longer a major mining center but was involved in a complex mix of farming and cattle ranching and was exporting everything from cotton and hides to sugar when coffee started to emerge as a major crop. By the 1850s, coffee finally became the major provincial export in terms of total value and boomed (except for the crisis with emancipation in the late 1880s) until the early twentieth century (see Graph 4.8). Although the quality of mineiro coffee was considered quite good, the average size and the total number of workers involved in coffee were somewhat smaller than in Rio de Janeiro and São Paulo.³⁸ Using the same techniques of planting, harvesting, and slave gang labor as Rio and paulista planters, the mineiro

TABLE 4.4. Fazendas, Workers, and Production in Coffee, Sugar, and Animals in São Paulo in 1854

Source: José J. Machado de Oliveira, “Quadro Estatístico de alguns Estabelecimentos Ruraes da Província de São Paulo,” Documento com que o Ilustríssimo e Excelentíssimo Senhor Dr. José Antonio Saraiva Presidente da Província de S.Paulo abrio a Assembléia Legislativa Provincial, no dia 15 de fevereiro de 1855 (São Paulo: Typ. 2 de Dezembro de Antonio Louzada Antunes: 1855).

GRAPH 4.8. Minas coffee exports, 1830–1906.

Source: Blasenheim, “A regional history,” p. 38, table 1.

fazendas tended to be far more mixed crop endeavors, even with some cattle ranching, compared to the standard coffee fazenda in the other major coffee provinces. Like these other coffee regions, and in contrast to most other slave-based economies, their servile population expanded up to the day of total abolition itself. The slave population in the coffee municípios of Minas reached a maximum of 90,000 by the early 1870s and 96,000 by 1886–1887.³⁹

As time went on, coffee increasingly absorbed more workers and finally became the largest single employer of Brazilian slaves in the last two decades of the slave era. After several failed experiments with imported European immigrant labor, the coffee planters, unlike their sugar compatriots in the northeast, abandoned all attempts to mix free and slave labor and concentrated exclusively on a slave labor force.⁴⁰ Thus, the end of the Atlantic slave trade in 1850, the consequent secular growth of coffee production, and the rise in slave prices all lead to a massive shift of slave laborers out of all parts of the empire into the coffee regions. This was most evident in the expansion of the internal slave trade after 1850. Although an internal slave trade had always existed, with the close of the African trade at mid-century there developed a significantly more important internal interprovincial and intraregional slave trade for the next two decades, which resulted in important shifts in the location of slaves within Brazil. It is estimated that the internal seaborne slave trade alone probably involved the migration of between 100,000 to 200,000 slaves from northeastern and far southern ports to Rio de Janeiro and Santos between 1850 and 1888.⁴¹ There were also important shifts between poorer zones within a given province and their richer neighboring districts.⁴² It has been estimated that by the 1870s, the coffee counties of São Paulo contained some 81,000 slaves, and the coffee counties (municípios) of Rio de Janeiro another 148,000.⁴³ Together with the previously estimated population of coffee-based slaves for Minas Gerais, it would mean that coffee production absorbed something like 320,000 slaves. Of the 1.2 million economically active slaves listed in the first national census of 1872, some 808,000 were employed in agriculture, and they represented only a fifth of all slaves in the empire.

By the early 1880s, the coffee estates had reached their maximum size, but with significant regional variations. Clearly Rio de Janeiro still dominated with the largest estates and a significantly larger average size plantation labor force than was the norm in the other two major provinces of coffee production (see Table 4.5). In fact, the average Rio estate had twenty more slaves than the Minas and most of the paulista plantations, which meant that by the last decade of slavery, the carioca planters were approaching the size of the biggest sugar estates in the Northeast. As the well-known analysis of Van Delden Laerne showed, slaves were not only a fundamental part of the labor force, but their value also represented more than 40 percent of the total value of the estates and their coffee trees. Moreover, this number was close to the general value of almost all postmortem inventories for plantations of sugar or coffee for all regions

TABLE 4.5. Basic Statistics on Coffee Plantations in Rio de Janeiro, Minas Gerais, and São Paulo, 1881–1883

Notes:

* We have excluded all mixed coffee and other crop fazendas.

** The survey divided fazendas into those related to the port of Rio de Janeiro, no matter what the province; Santos includes only those municípios in central and southern São Paulo directly linked to the port of Santos.

Source: Van Delden Laerne, Brazil and Java, pp. 222–3.

of Brazil and was considerably more than the value of slaves in cattle ranching and mixed farming estates.⁴⁴

But despite this ever-increasing concentration of slaves in coffee, the majority of Brazil's slaves still did not work in coffee or sugar fazendas, even in the central-south zone. For example, despite its dominant position in the total value of provincial exports, coffee absorbed only a small fraction of even the rural slaves within the province. Minas Gerais was unique in that its mixed farming sector, which exclusively produced for a local market, absorbed the majority of the slaves in this largest slave province in the empire. In the 1870s, there were 382,000 slaves in the province, of which 279,000 were listed as rural workers, and in the province as a whole, slaves on coffee fazendas represented fewer than a quarter of all the local slaves. Nor were the counties with the most slaves those most associated with coffee.⁴⁵ Cattle ranching, food processing, and the production of grains and root crops were all slave- as well as free-labor activities. There were even a significant number of slaves in skilled and semi-skilled occupations. In a detailed analysis of the unpublished census of 1831–32, some 14 percent of the 88,000 or so adult slaves were listed as skilled and semi-skilled nonagricultural workers, with the single largest area of activity being spinning and weaving of cotton. Moreover, slave artisans made up 20 percent of all workers in spinning and weaving and 28 percent of those in construction. It was estimated that almost one-quarter of all female adult slaves were in these nondomestic service occupations.⁴⁶ Thus, Minas Gerais represents the rare case within America of a large-scale employment of rural slaves for primarily local, regional, or national production. Minas Gerais was also unusual for a Center-South state in its distribution of slave ownership. As in the case of slaveholdings throughout mainland America, the average number of slaves held per slave owner was quite small, but, unlike most American slave zones, the number of slave owners was quite large, and they represented a much more sizable percentage of the free population. Big fazendeiros owning large numbers of slaves were few, and they controlled a relatively small share of the total provincial slave labor force.

However, Minas was not alone in this more complex distribution of slaves. It was found that even in Pernambuco, with its high concentration of slaves in sugar, between 30 to 40 percent of the rural slaves were found outside the plantation zones throughout the nineteenth century. These were engaged in cotton production, ranching, and food production. In

TABLE 4.6. Distribution of Slave Holdings by Region, Pernambuco, 1800–1887

Notes: Distribution based on inventories that contained slaves.

Source: Versiani and Vergolino, “Slave holdings in nineteenth-century Brazilian northeast” (2002), table 3.

fact, cotton plantations in the interior agreste region actually produced a significant quantity of foodstuffs as well as cotton.⁴⁷ Moreover, as might be expected, these interior small farming, cattle, and cotton estates had much smaller holdings of slaves, and most slave owners had on average fewer slaves and there were far fewer larger units than found in the sugar-producing region (see Table 4.6).

As already noted, aside from producing sugar, which was still Brazil's second-most-valuable export, slaves were to be found producing rice,⁴⁸ cotton, poultry, pigs, and pork products as well as various grains and corn and general food production for the internal market. There was also a significant use of slaves along with free workers in the cattle industries of the traditional southern grazing lands, but also in the expanding cattle ranches of the northeastern provinces. Brazilian cotton, which had been a vital colonial product and still supplied the European market with a major share of its raw cotton until the first decade of the nineteenth century, revived in the 1860–1880 period. The U.S. Civil War created a cotton famine for European mills, and the result was a revival both of the Maranhão cotton plantation sector as well as the growth of new cotton regions such as those of Minas Gerais, with even some production coming out of São Paulo. Although impressive in financial terms, this temporary growth in the value of cotton exports had no long-lasting or significant impact on local labor distribution. Production only doubled in the 1860–1880 period, whereas the value of these exports more than quintupled. The result was a temporary local shift of slaves into cotton, which had little long-term impact on slaves engaged in the other plantation crops.

The other 370,000 or so slaves in rural occupations could be found scattered throughout the empire in activities that went to feed the growing market of 9.9 million Brazilians. Slaves still remained important in the jerked beef industry in the southern provinces of Rio Grande do Sul, which successfully competed throughout the nineteenth century against the free-labor-based Uruguayan jerked beef industry. Although free workers were always present in this industry, the use of slaves remained important and profitable until the 1880s.⁴⁹ The same occurred in the expanding cattle ranches of the northeast where slave and free workers were the norm. They were also important in the manufacture of dairy and pork products for local and regional markets in Minas Gerais. Along with cotton, for example, northeastern farmers grew a host of crops, many of them for export. Bahia, for example, had a thriving tobacco and manioc flour production, a portion of which was produced by slave labor. Although the majority of tobacco and cassava producers were mostly small free farmers with a few agregados, or hired hands, a significant minority of these producers owned slaves and produced for export. In addition, the more successful of these farmers often began to purchase slaves as their profits increased. In the tobacco district of Cachoeira in the 1846–1860 period, two-thirds of the farmers owned nine or fewer slaves. And none held more than forty slaves. Even fewer slaves were used on the cassava farms, although these produced major quantities for sale in regional

TABLE 4.7. Ratio of Slaves Owned by Size of Slaveholding (Plantel) in the 1870s by Region

Source: Marcondes, “Desigualdades regionais brasileiras,” (2005), tabelas 4–5, 5–5, 6–5, 7–5, 8–5, 9–5.

markets. However, even here a few individuals had more than ten slaves producing the flour.⁵⁰

A recent attempt at a national evaluation of slave ownership throughout Brazil in the 1870s shows how widely but also how unevenly distributed were the slaves among the agricultural communities.⁵¹ Sugar and coffee municípios had the highest average number of slaves per unit and the most unequal distribution of slaves, but everywhere else slaves were fairly widely distributed in small holdings in a host of agricultural, grazing, and other rural activities (see Table 4.7).

Finally, every major urban center was surrounded by truck gardens, many of which were run with small numbers of slaves often as independent units along with poor free farmers.

The 345,000 of the economically active slaves not directly engaged in agriculture in 1872 were often closely allied with plantation life. The most obvious example was the 95,000 slaves listed as day laborers, some of whom were probably employed in the fazendas alongside the resident slave forces. Some of the seven thousand artisans listed as working in wood and metal crafts, especially carpenters and blacksmiths, may also have been employed on plantations. But as the example of Minas Gerais reveals, there was also within the slave labor force a significant proportion of slaves who were not directly related to export agriculture yet still played a significant economic role in the economy. Thus, slaves made up 10 percent of the 126,000 workers found in the textile factories, which were then coming into prominence as Brazil's first major industrial activity. The 175,000 slaves who were in domestic service accounted for 17 percent of all persons employed in that activity and made up 15 percent of the economically active slaves. Slaves also exceeded their 15 percent share of the laboring population in such activities as construction (4,000 of whom made up 19 percent of all such workers), masonry, stonework, and allied crafts (18 percent), just as they held more than their share of day laboring (23 percent). Finally, there were some occupations in which, even though slaves represented a small share of all workers, the absolute number of slaves was impressive. This was the case with seamstresses, in which the 41,000 slave women represented only 8 percent of the total workers in this occupation.

A great many slaves also lived in cities, in which, like the country at large, they formed a minority of the total colored population. A much higher percentage of the 4.2 million free colored than of the 1.5 million slaves lived in urban centers. Nevertheless, slaves were important in the labor force of every city. Of the 785,000 persons who lived in cities of 20,000 or more in 1872, a minimum of 118,000, or 15 percent, were slaves. This was probably not the highest number of urban slaves ever reached, as such urban-based slaves were in decline at this time, just as was the total slave population. Ever since the end of the slave trade at mid-century, the total number of slaves had declined from their peak of 1.7 million. As in most of the American slave states, the abolition of an intense Atlantic slave trade initially led to a negative growth rate of the resident slave population. The decline in total numbers also led to a shift in their distribution. The steep rise of slave prices as a result of the ending of the Atlantic slave trade, the increasing impact of manumission, and the continued expansion of coffee meant that ever more slaves would be sold from the city into the countryside. In 1849, for example, the city of Rio de Janeiro had 78,000 slaves, whereas in 1872 there were only 39,000. But slaves in Rio still represented more than one-fifth of the city's 183,000 people. Salvador, the second-largest center in 1872 with 108,000 residents, had 13,000 slaves, and Recife ranked third with 57,000 persons, of whom 10,000 were slaves. Even the still quite small city of São Paulo had 3,000 slaves out of a population of 28,000.⁵²

Urban slavery in Brazil had both the standard forms of rural master–slave relationship with direct ownership and resident employment along with direct ownership and rental to a third party. But there were also a fairly large number of slaves who were self-employed, or, as the Brazilians called them, escravos de ganho.⁵³ These slaves spanned the occupational spectrum from the least skilled and most dangerous of jobs to the most highly remunerative occupations. Such slaves even had multiple occupations. In the city of Salvador in the 1850s and 1870s, there were several cases noted of multiple skills or occupations occupied by one slave. Thus, the young slave boy André of the Nagô nation was both a sedan porter and an apprentice carpenter, while Ricardo, another young Nagô slave, worked both as a farmhand (serviço de roça) and as an escravo de ganho. Amália, a 20-year-old slave who was self-employed, worked as a cigar maker and as an ironer.⁵⁴ This same pattern of multiple occupations was found for the rental slaves of the city of Salvador in the mid-nineteenth century.⁵⁵ All information points out that these escravos de ganho porters, vendors, and semi-skilled and skilled artisans were not only self-employed and paid a fixed income to their owners, but they also took care of their own housing or lived as apprentices in the homes of master craftsmen who were not their owners. Studies of the major cities point out that local municipal records constantly show the existence of independent-slave households.⁵⁶ Such independent slaves even created complete families in these separate quarters, as occurred in the case of seven of the eight slaves owned by widow Dona Catarina de Oliveira in the paulista district of Curitiba in 1776. Three of her slaves not only lived apart but also had their own families and independently farmed some lands, whereas the others simply lived apart and took care of their own household needs.⁵⁷ Such slaves were also let out in apprentice contracts to local free artisans. This was the case, for example, of four out of the five slaves owned by Antonio de Souza Ferreira, who died in Rio de Janeiro in 1824; two of these slaves were apprentices to a carpenter, one to a shoemaker, and one was already a journeyman carpenter. Artists and musicians were often self-employed slaves, although some of these resided with their masters. This was the case of the African-born Antonio José Dutra, a barber in Rio who on his death in 1849 owned thirteen slaves, mostly Africans, the majority of whom formed a band and brought him far greater income than either of his two rental properties or his barbering. These slave musicians also served as barbers in his shop.⁵⁸

Many slaves were also rented to factory owners. In 1840, José Vieira Sarmento, for example, had a small tortoiseshell comb factory in the city that employed eleven of his own slaves (four journeymen and seven apprentices) and nine rented slaves.⁵⁹ In the Rio de Janeiro city of Niterói in 1855, for example, there were 130 slaves employed in the local factories, of which 85 were owned by the individual factory that employed them and 45 were rented. In the neighboring city of Rio de Janeiro, half (527) of the 1,039 workers employed in the local factories were slaves, although there is no corresponding breakdown into rented and owned. The largest textile firm in Brazil, the Companhia da Ponta D’Arêa, in late 1855 had a workforce of 622 workers, of whom 181 were slaves.⁶⁰ Moreover, despite the steady increase of foreign-born free workers in industry as late as 1872, the census listed 2,135 slave industrial workers compared to 9,458 foreign born in a total industrial workforce of 18,091.⁶¹ In the detailed, unpublished census of Minas Gerais in 1831–1832, there were listed some 23,000 slaves who worked in such activities. These represented 17 percent of the provinces’ slaves (compared to 41 percent of free persons who were working in manufacturing activities). Half of the slaves, like half of the free workers, were listed as working in manual and mechanical activities, but slaves were higher than their provincial average among metal workers, construction workers, and in textile activities, with 95 percent of the textile slave workers (or 9,998 of them) being women.⁶²

All this flexibility made for a quite complex pattern of slave activity and for a much more pronounced intervention of the slave in the market economy as a consumer and earner of income. Although municipalities often complained about the relative freedom and lack of financial support for self-employed slaves, they proved so lucrative an investment for their masters that the practice was never abolished. Estimates from both urban and plantation slave rentals suggests that rented-out slaves – after paying for their own upkeep of housing and food (estimated at 20 percent of their gross income) – provided an annual profit of some 10 to 20 percent to their masters on their initial investment.⁶³ They were, however, only moderately used in the rural area and so were one of the major features that distinguished urban from rural slavery in Brazil. There was even a strike organized in Bahia in 1857 of the porters known as ganhadores. These haulers of objects, water, persons in sedan chairs, and innumerable other cargoes were organized into work groups usually along African nationalities. They also included both slave and free Africans, as well as Creoles both slave and free, although the dominant element was Nagôs or Yorubas. In that year, an attempt by the municipal government to tax their labor and create other forms of limitations on their freedom of contract led to a weeklong strike of all active workers, which effectively caused the urban government to abandon most of its tax and control efforts.⁶⁴

Finally, slaves could be found in all parts of Brazil's complex transportation network, from muleteers in the mule trains that formed the basis of interior transport until well into the second half of the nineteenth century⁶⁵ to sailors in the shipping in both long-distance, coastal trade and among local fishermen. As we previously noted, slaves were to be found aboard slave ships bound for Africa as members of the regular crew and were vitally important in the coastal trade in the late eighteenth and early nineteenth centuries.⁶⁶ Even in the nineteenth century, slave sailors remained a fundamental part of the shipping industry. For example, in the port of Rio de Janeiro in the period of 1815 to 1826, officials issued passports for 2,463 slave sailors engaged in the coastal trade with southern ports.⁶⁷ In the southern Brazilian port town of Porto Alegre in 1857, there were 3,193 sailors listed as working from the port involved in either coastal trading, fishing, or local rowboat transport, 1,157 of whom were slaves. These slave sailors represented 49 percent of the sailors engaged in coastal trade, their most significant participation, and they accounted for 30 percent of the sailors engaged in long-distance oceanic trade, 27 percent of the river and port sailors, and only 10 percent of the fishermen.⁶⁸ More than a decade later in 1869, there was only a modest decline in the ratio if not the number of slave sailors. In that year the government listed 3,638 sailors as working from the port involved in oceanic and coastal trading, fishing, and local rowboat transport, of whom 1,168 – or almost a third – were slaves.⁶⁹ They were also a significant part of the boatyard workers in the same year of 1869, accounting for 72 of the 247 workers making and repairing ships in the port's naval yards.⁷⁰

The relative decline in urban slavery, if not of the urban colored population, was part of a larger process of geographic redistribution of slaves that occurred in the post-Atlantic slave trade period. Not only were a higher proportion of slaves found in the most productive industries such as coffee but also in those regions in which those industries were concentrated. At mid-century, fewer than half of the slaves were to be found in three major coffee provinces, but by 1872 more than half were located there. An active post-1850 internal slave trade helped to concentrate slaves in the center-south district, with both the Northeast and the far southern provinces shipping their slaves to Rio de Janeiro, Minas Gerais, and, above all, to São Paulo. On the eve of abolition in 1887, almost three-quarters of the remaining 751,000 slaves could be found in these three provinces. Thus slavery, as in Cuba, was most heavily concentrated in the most dynamic regions of their respective societies on the eve of emancipation. With the cost of slaves rising and the growth rates of the slave populations still negative, ever more slaves were shifted into the export sectors of the two largest slave states in Latin America.

¹ On the debate about the law, see Martha Abreu, “Slave Mothers and Freed Children: Emancipation and Female Space in Debates on the ‘Free Womb’ Law, Rio de Janeiro, 1871,” Journal of Latin American Studies, 28, no. 3 (October 1996), pp. 567–80.

² On the so-called Lei dos Sexagenários, see Joseli Maria Nunes Mendonça, Entre a mão e os anéis: A Lei dos sexagenários e os caminhos da abolição no Brasil (Campinas: Editora da UNICAMP, 1999).

³ For estimates of Brazilian sugar production in the eighteenth century, see Noel Deerr, The History of Sugar (2 vols.; London: Chapman and Hall, 1949–1950), II, p. 113.

⁴ Peter Eisenberg, The Sugar Industry in Pernambuco: Modernization without Change, 1840–1910 (Berkeley: University of California Press, 1974), p. 5, table 1.

⁵ The data on American cane-sugar production come from Deerr, The History of Sugar, II, pp. 113, 131, 193–204. For an estimate of world output, see Manuel Moreno Fraginals, El Ingenio: Complejo económico social cubano del azúcar (3 vols.; Havana: Editorial de Ciencias Sociales, 1978), II, p. 173.

⁶ Galloway, The Sugar Cane Industry, chapter 7.

⁷ Eisenberg, The Sugar Industry in Pernambuco, table 7, p. 20.

⁸ Today, of course, Brazil is both the world's largest producer and exporter of cane sugar.

⁹ Deerr, The History of Sugar, II, p. 113.

¹⁰ Eisenberg, The Sugar Industry in Pernambuco, chapter 3.

¹¹ Schwartz, Sugar Plantations in the Formation of Brazilian Society, pp. 423ff.

¹² B. J. Barickman, “Revisiting the Casa-grande: Plantation and Cane-Farming Households in Early Nineteenth-Century Bahia,” Hispanic American Historical Review, 84, no. 4 (2004), p. 626.

¹³ Deerr, The History of Sugar, I, p. 111.

¹⁴ Eisenberg, The Sugar Industry in Pernambuco, table 22, p. 147.

¹⁵ Eisenberg, The Sugar Industry in Pernambuco, p. 157.

¹⁶ Eisenberg, The Sugar Industry in Pernambuco, p. 146.

¹⁷ See Schwartz, Sugar Plantations in the Formation of Brazilian Society, chapter 15; and Eisenberg, The Sugar Industry in Pernambuco, chapter 8.

¹⁸ Schwartz, Sugar Plantations in the Formation of Brazilian Society, p. 454, table 16–6; and Barickman, “Revisiting the Casa-grande,” p. 627.

¹⁹ Schwartz, Sugar Plantations in the Formation of Brazilian Society, pp. 428, 446, table 16–3. On the early evolution of sugar engenhos in Campos, see Arthur Cézar Ferreira Reis, “A provincial do Rio de Janeiro e o Município Neutro,” in Sérgio Buarque de Holanda, ed., História Geral da Civilização Brasileira (11 vols.; São Paulo: Difusão Européia do Livro, 1960), tomo II, vol. 2, p. 317.

²⁰ Deerr, The History of Sugar, I, p. 111; and Alice P. Canabrava. “A grande Lavoura,” in Sérgio Buarque de Holanda (org.), História Geral da Civilização Brasileira, vol. 6 (Rio de Janeiro: Bertrand Brasil, 1997). Recent analysis of the government promotion of these central mills in the period from 1875–1889 is found in the works of Roberta Barros Meira, “O processo de modernização da agroindústria canavieira e os engenhos centrais na Província de São Paulo,” História e Economia Revista Interdisciplinar (São Paulo), 3, no. 1 (2° semestre 2007), pp. 39–54; and Roberta Barros Meira, Bangüês, engenhos centrais e usinas: O desenvolvimento da economia açucareira em São Paulo e a sua correlação com as políticas estatais (1875–1941) (Dissertação de mestrado, Faculdade de Filosofia, Letras e Ciências Humanas, da Universidade de São Paulo, 2007).

²¹ The best overall survey on the paulista sugar industry is Maria Thereza Schorer Petrone, A lavoura canavieira em São Paulo. Expansão e declínio (1765–1851) (São Paulo: Difusão Européia do Livro, 1968). Other studies of this period include Suely Robles Reis de Queiroz, “Algumas notas sobre a lavoura de açúcar em São Paulo no período colonial,” Anais do Museu Paulista, vol. 21 (1967), pp. 109–277; and the several regional studies that treat sugar in this period. Among these are the outstanding work of Lucila Herrmann, Evolução da estrutura social de Guaratinguetá num período de trezentos anos (São Paulo: Ed. Facsimilada. Instituto de Pesquisas Econômicas [IPE/USP], 1986); as well as Carlos Almeida Prado Bacellar, Os senhores da terra: Família e sistema sucessório entre os senhores de engenho do oeste paulista (1765–1855) (São Paulo: UNICAMP, 1987) and Ramón Vicente Garcia Fernandéz, “Transformações econômicas no litoral norte paulista (1778–1836)” (Tese de doutorado, FEA: Universidade de São Paulo, 1992). On this theme, also see Meira, Bangüês, engenhos centrais e usinas.

²² In 1832, Jamaica had 531 sugar estates, with an average of 223 slaves per estate, of which only 4 produced another product in addition to sugar. B. W. Higman, Slave Populations of the British Caribbean, 1807–1834 (Baltimore: Johns Hopkins University Press, 1984), p. 14.

²³ For a description of these skilled positions, see Schwartz, Segredos Internos, chapter 12; and also André João Antonil, Cultura e Opulência do Brasil. Introdução e vocabulário por A. P. Canabrava (São Paulo: Ed. Nacional, s/d).

²⁴ Schwartz, Segredos Internos, p. 137, tabela 13.

²⁵ Marcelo Magalhães Godoy, “Fazendas diversificadas, escravos polivalentes – Caracterização sócio-demográfica e ocupacional dos trabalhadores cativos em unidades productivas com atividades agroaçucareiras de Minas Gerais no século XIX,” XIV Encontro Nacional de Estudos Populacionais, ABEP (2004), p. 20, tabela 13.

²⁶ On the evolution of the coffee industry in Brazil, see Affonso de E. Taunay, História do Café no Brasil (20 vols.; Rio de Janeiro: Departamento Nacional do Café, 1939), especially vols. 1 and 2; and Stanley J. Stein, Vassouras: A Brazilian Coffee County, 1850–1900 (Cambridge: Harvard University Press, 1957). The best study on coffee from the second half of the nineteenth century is the work by Antônio Delfim Netto, O problema do café no Brasil (São Paulo: IPE-USP, 1981).

²⁷ Pedro Carvalho de Mello, “The Economics of Labor in Brazilian Coffee Plantations, 1850–1888” (Ph.D. dissertation, Department of Economics, University of Chicago, 1977), table 14b, p. 42.

²⁸ There were some 5,200 slaves on the coffee fazendas in Juiz de Fora, the largest coffee county in Minas Gerais, at the end of the slave era, with thirteen coffee estates holding more than 100 slaves. Luiz Fernando Saraiva, “Estrutura de terras e transição do trabalho em um grande centro cafeeiro, Juiz de Fora 1870–1900,” X Seminário sobre a Economia Mineira (2002).

²⁹ For the evolution of coffee in the São Paulo province, see Warren Dean, Rio Claro: A Brazilian Plantation System, 1820–1920 (Stanford: Stanford University Press, 1976); the dissertation of Carvalho de Mello, “The Economics of Labor in Brazilian Coffee Plantations, 1850–1888,” and the Portuguese translation, “A economia da escravidão nas fazendas de café: 1850–1888,” PNPE, 1984; mimeo; José Flávio Motta, “A família escrava e a penetração do café em Bananal (1801–1829),” Revista Brasileira de Estudos Populacionais, 5, no. 1 (1988), pp. 71–101; José Flávio Motta, Corpos Escravos e Vontades Livres – Estrutura da Posse de Cativos e Família Escrava em um Núcleo Cafeeiro (Bananal, 1801–1829) (São Paulo: Annablume/Fapesp, 1999). José Flávio Motta and Nelson Hideiki Nozoe, “Cafeicultura e acumulação,” Estudos Econômicos, 24:2 (Maio/Ago. 1994), pp. 253–320; Armenio de Souza Rangel, “Escravismo e riqueza – Formação da economia cafeeira no município de Taubaté – 1765/1835” (Tese de doutorado, São Paulo: FEA-USP, 1990); and Renato Leite Marcondes, A arte de acumular na economia Cafeeira (Lorena, São Paulo: Editora Stiliano, 1998); and Francisco Vidal Luna and Herbert S. Klein, Slavery and the Economy of São Paulo, 1750–1850 (Stanford: Stanford University Press, 2003), chapter 3.

³⁰ This was expressed in an average of 33.6 bags of coffee at 60 kilos per bag. See Carvalho de Mello, “The Economics of Labor in Brazilian Coffee Plantations, 1850–1888,” p. 18.

³¹ C. F. Van Delden Laerne, Brazil and Java: Report on Coffee-Culture in America, Asia, and Africa (London: W. H. Allen, 1885), pp. 222–3.

³² B. W. Higman, Slave Population and Economy in Jamaica 1807–1834 (Cambridge: Cambridge University Press, 1976), p. 13, table 1.

³³ The classic study of the Vassouras coffee economy and society is Stein, Vassouras.

³⁴ See Carvalho de Mello, “The Economics of Labor in Brazilian Coffee Plantations, 1850–1888,” p. 18.

³⁵ For an analysis of the relative roles of mule, wagon, and rail in the transport of coffee, see Herbert S. Klein, “The Supply of Mules to Central Brazil: The Sorocaba Market, 1825–1880,” Agricultural History, 64, no. 4 (Fall 1990), pp. 1–25.

³⁶ For Rio de Janeiro coffee plantations in the 1870s and 1880s, Carvalho de Melo estimates that field hands were 57 percent of the total labor force. See Carvalho de Mello, “The Economics of Labor in Brazilian Coffee Plantations, 1850–1888,” p. 14.

³⁷ A detailed estimate of the internal slave trade is found in Chapter 6.

³⁸ For a good overall survey of one of the principal coffee zone of Minas Gerais, see Rafael Rangel Giovanini, “Regiões em movimento: Um olhar sobre a Geografia Histórica do Sul de Minas e da Zona da Mata Mineira (1808–1897)” (Dissertação de mestrado, Geografia, Belo Horizonte: Universidade Federal de Minas Gerais, 2006), pp. 103ff; and Peter Louis Blasenheim, “A regional history of the Zona da Mata in Minas Gerais, Brazil, 1870–1906” (Ph.D. dissertation, Stanford University, 1982).

³⁹ Giovanini, “Regiões em movimento,” table 16, p. 149.

⁴⁰ On the pre-abolition attempts to incorporate free immigrant labor, see Dean, Rio Claro, chapter 4.

⁴¹ One contemporary source reported that 26,622 slaves were imported into Rio de Janeiro from other states between 1852 and 1859, which would give an annual average of 3,327 slaves. Sebastião Ferreira Soares, Notas estatísticas sôbre a producção agrícola e carestia dos generos alimenticios no Império do Brazil (Rio de Janeiro: J. Villeneuve, 1860), pp. 135–6. The British Minister to Rio de Janeiro, W. D. Christie, reported that 34,688 slaves were imported by sea into Rio de Janeiro between January 1852 and July 1862, which would give a rate of 275 per month or 3,300 per annum. W. D. Christie, Notes on Brazilian Questions (London: Macmillan, 1865), p. 93. J. H. Galloway estimates the total loss of the Northeast at 90,000 slave migrants between 1850 and 1880, or 3,000 slaves per annum shipped by sea from the region to the port of Rio de Janeiro. See Galloway's “The Last Years of Slavery on the Sugar Plantations of Northeastern Brazil,” Hispanic American Historical Review, LI, no. 4 (November 1971). His estimate may be considered the lower limit, because it excludes southern exports and any post-1880 figures. On the internal trade in one year to the port of Rio de Janeiro, see Herbert S. Klein, “The Internal Slave Trade in Nineteenth-Century Brazil: A Study of Slave Importations into Rio de Janeiro in 1852,” Hispanic American Historical Review, LI, no. 4 (November 1971), pp. 567–8.

⁴² See José Flávio Motta, “Escravos daqui, dali e de mais além: o tráfico interno de cativos em Constituição (Piracicaba), 1861–1880,” Anais do XXXIII Encontro Nacional de Economia ANPEC (2005); “O tráfico de escravos na Província de São Paulo: Areias, Silveiras, Guaratinguetá e Casa Branca, 1861–1887” (Texto para Discussão. São Paulo: IPE/USP, 21, 2001); “Tráfico interno de cativos: o preço das mães escravas e sua prole,” XI Encontro Nacional de Estudos Populacionais, ABEP (1998); and José Flávio Motta and Renato Leite Marcondes, “O Comércio de Escravos no Vale do Paraíba Paulista: Guaratinguetá e Silveiras na Década de 1870,” Estudos Econômicos, 30, no. 2 (Abril–Junho 2000), pp. 267–99. Also see the several studies on Minas: Cláudio Heleno Machado, “O tráfico interno de escravos na região de Juiz de Fora na segunda metade do século XIX,” X Seminário de Economia Mineira (2002); Rômulo, Andrade, “Havia um mercado de famílias escravas? (A propósito de uma hipótese recente na historiografia da escravidão),” Revista de História (Juiz de Fora), 4, no. 1 (1998), pp. 93–104; Camila Carolina Flausino, “Negócios da escravidão: Tráfico interno de escravos em Mariana: 1861–1886” (Dissertação de mestrado, História, Universidade Federal de Juiz de Fora, 2006).

⁴³ Carvalho de Mello, “The Economics of Labor in Brazilian Coffee Plantations, 1850–1888,” tables 21 and 22, pp. 76–7.

⁴⁴ Among senhores de engenho in Salvador in the nineteenth century, the value of the slaves was 51 percent, and for those owning agricultural properties, it was 32 percent of total value. For most urban professionals, the figure was less than 10 percent. Katia M. de Queirós Mattoso, Bahia, século XIX, Uma província no império (Rio de Janeiro: Editora Nova Frontiera, 1992), tabela 116, p. 629. In the Agreste region of Pernambuco, slaves were worth 51 percent of all slave-owning estates (92) in the period of 1820–1849 – in this case much like the Bahian plantation owners – and in both cases, they represented the single most important wealth source. In a sample in the same period from the Sertão region of Pernambuco, which contained cattle ranches, slaves were second in importance to animals and represented 31 percent of the total wealth of these landowners. Flávio Rabelo Versiani and José Raimundo Oliveira Vergolino, “Posse de Escravos e Estrutura da Riqueza no Agreste e no Sertão de Pernambuco: 1777–1887.” Estudos Econômicos, 33, no. 2 (Abr./Jun. 2003), tabelas 10 and 14, pp. 373, 375. In the cattle-growing and farming zone of Barbacena in Minas Gerais in some 302 postmortem inventories of leading planters and cattle ranchers from 1791–1822, slaves represented 35 percent of the total value of their estates. Adriano Braga Teixeira, “População, sistema econômico e poder na transição do século XVIII para o XIX em Minas Colonial – Barbacena – 1791/1822” (Dissertação de mestrado, Universidade Federal do Rio de Janeiro – UFRJ, 2007), tabela 14, p. 78; and in the São Paulo município of Mariana, which was dedicated to mixed farming production whose output went into the internal market, on the 227 farms that used slave labor, the slaves represented 39 percent of the total wealth of these farm families and were the most significant part of the family wealth by far. Heloisa Maria Teixeira, “Reprodução e famílias escravas, em Mariana, 1850–1888” (Dissertação do mestrado, USP-FFLCH, 2001), tabela 3, p. 32. In other cattle-growing regions, slaves were less important. Thus, slaves were of the same relative importance as animals and lands in the cattle regions of the Rio Grande de Sul province, and in the first half of the nineteenth century, they averaged about 20 percent of the value of all estates that contained slaves. Luiz Paulo Ferreira Nogueról, Diego Rodrigues, Ezequiel Giacomolli, and Marcos Smith Dias, “Elementos Comuns e Diferenças entre os Patrimônios Registrados na Pecuária Gaúcha e na Pernambucana no Início do Século XIX,” VIII Encontro de Economia da Região Sul – ANPEC SUL (2005), tabela 9, which accounts for 326 postmortem inventories that listed slave ownership in the period from 1820–1849, for the districts of Pelotas, Rio Pardo, Rio Grande, and Porto Alegre. A study of Pelotas alone in the period of 1850–1871 found that slaves were still just second in importance to land and housing, but at 30 percent of the value of all slaveholding ranches, they were more important than cattle in this period. Luana Teixeira, “Trabalho escravo na produção pecuária: São Francisco de Paula de Cima da Serra (Rio Grande de São Pedro, 1850–1871),” III Encontro Escravidão e liberdade no Brasil Meridional (2007), tabela IV. Even in the far southern district of Lages in Santa Catarina, among the cattle ranchers, slaves were worth 18 percent of the total value of some 149 estates, with slaves being more valuable the wealthier the unit. Nilsen C. Oliveira Borges, “Terra, gado e trabalho: Sociedade e economia escravista em Lages, SC (1840–1865)” (Dissertação de mestrado, Universidade Federal de Santa Catarina, Florianópolis, 2005), tablea VIII, p. 86.

⁴⁵ Giovanini, “Regiões em movimento,” tables 12 and 14, pp. 139, 144.

⁴⁶ Marcelo Magalhães Godoy, “Uma província artesã: O universo social, econômico e demográfico dos artífices da Minas do oitocentos,” Anais do XII Encontro Nacional da ABEP (2004), tabelas 1, 6, and 7.

⁴⁷ Versiani and Vergolino, “Posse de Escravos e Estrutura da Riqueza no Agreste e Sertão de Pernambuco,” pp. 353–93; and by the same authors, “Slaveholdings in nineteenth-century Brazilian northeast: Sugar estates and the backlands” (paper presented at XIII Congress of the International Economic History Association, Buenos Aires, July 2002), which adds the sugar zones to their previous regional selections.

⁴⁸ A partial analysis of colonial slave-produced rice in Maranhão is found in Judith A. Carney, “‘With Grains in Her Hair’: Rice in Colonial Brazil,” Slavery & Abolition, 25, no. 1 (April 2004), pp. 1–27. A more complete survey of an example of slaves working in rice production – this time in São Paulo in the nineteenth century – is found in Agnaldo Valentin, “Arroz no Vale do Ribeira (1800–1888)” (Tese de doutorado, FFLCH, Universidade de São Paulo, 2006).

⁴⁹ Leonardo M. Monasterio, “FHC errou? A economia da escravidão no Brasil meridional,” História e Economia Revista Interdisciplinar (São Paulo, BBS), 1, no. 1 (2° semestre 2005), pp. 13–28.

⁵⁰ B. J. Barickman, A Bahian Counterpoint: Sugar, Tobacco, Cassava, and Slavery in the Recôncavo, 1780–1860 (Stanford: Stanford University Press, 1998), chapter 6.

⁵¹ Renato Leite Marcondes, “Desigualidades regionais brasileiras: Comércio marítimo e posse de cativos na década de 1870” (Tese livre-docência, FEA, USP, Riberão Preto, 2005).

⁵² See Table 6.3 and its sources.

⁵³ On the self-renting slaves, see among others the works of Luiz Carlos Soares, “Os escravos de ganho no Rio de Janeiro do século XIX,” Revista Brasileira de História V, no. 16 (1988), pp. 107–42; and Leila Mezan Algranti, “Os ofícios urbanos e os escravos de ganho no Rio de Janeiro Colonial (1808–1822),” in Tamás Szmrecsányi, ed., História Econômica do Período Colonial (São Paulo: Ed. HUCITEC/FAPESP, 1996), pp. 195–214.

⁵⁴ Maria Cristina Luz Pinheiro, “O trabalho de crianças escravas na cidade de Salvador, 1850–1888,” Afro-Ásia, 32 (2005), p. 162.

⁵⁵ As Katia Mattoso noted, “…é difícil distinguir entre escravos de ganho, os que vão para as ruas mercadear, e escravos domésticos, pois seus proprietários utilizam seus serviços ou os alugam de acôrdo com suas necessidades do momento: um mesmo escravo pode, muito bem, ser ganhador e doméstico ao mesmo tempo.” Katia M. de Queirós Mattoso, “Sociedade escravista e mercado de trabalho: Salvador–Bahia, 1850–1868,” Bahia Análise & Dados (Salvador), 10, no. 1 (2000), p. 17.

⁵⁶ See, e.g., Leila Mezan Algranti, O feitor ausente: Estudo sobre a escravidão urbana no Rio de Janeiro, 1808–1821 (Petrópolis: Vozes, 1988); and Ynaê Lopes dos Santos, “Além da senzala: Arranjos escravos de moradia no Rio de Janeiro (1808–1850)” (Dissertação de mestrado, Universidade de São Paulo, 2006).

⁵⁷ Adriano Bernardo Moraes Lima, “Trajetórias de Crioulos: Um estudo das relaçes comunitárias de escravos e forros no têrmo da Vila de Curitiba (c. 1760 – c. 1830),” (Dissertação de mestrado, Universidade Federal de Paraná, Curitiba, 2001), p. 41.

⁵⁸ Zephyr Frank, Dutra's World: Wealth and Family in Nineteenth-Century Rio de Janeiro (Albuquerque: University of New Mexico Press, 2004).

⁵⁹ Luiz Carlos Soares, “A escravidão industrial no Rio de Janeiro do século XIX,” Anais do V Congresso Brasileiro de História Econômica ABPHE (2003), p. 4.

⁶⁰ Artur José Renda Vitorino, “Operários livres e cativos nas manufaturas: Rio de Janeiro, segunda metade do século XIX,” I Jornada Nacional de História do Trabalho, (Florianópolis: Laboratório de História Social do Trabalho e da Cultura, 2002), p. 7.

⁶¹ Luiz Carlos Soares, “A manufatura na sociedade escravista: O surto manufatureiro no Rio de Janeiro e nas suas circunvizinhanças (1840–1870),” in F. Mauro, ed., La préindustrialization du Brésil (Paris: CNRS, 1984), p. 42.

⁶² Godoy, “Uma província artesã, p. 4.

⁶³ The best estimates of profitability of slave ownership in the nineteenth century comes from Carvalho de Mello, “The Economics of Labor in Brazilian Coffee Plantations,” chapter 4.

⁶⁴ João José Reis, “The Revolution of the Ganhadores: Urban Labour, Ethnicity, and the African Strike of 1857 in Bahia, Brazil,” Journal of Latin American Studies, 29, no. 2 (May 1997), 355–93.

⁶⁵ African-born slaves seem to have been overrepresented among the tropeiros, and slaves made up a third of the labor force raising pack animals. See Marcelo Magalhães Godoy, Mario Marcos Sampaio Rodarte, and Clotilde Andrade Paiva, “Negociantes e tropeiros em um território de contrastes; o setor comercial de Minas Gerais no século XIX,” Anais do V Congresso Brasileiro de História Econômica, ABHPE (2003).

⁶⁶ Also see Jaime Rodrigues, “Cultura marítima: Marinheiros e escravos no tráfico negreiro para o Brasil (sécs. XVIII E XIX),” Revista Brasileira de História, 19, no. 38 (1999), pp. 15–53. There are even cases of fugitive slaves claiming to be free and enlisting in the Brazilian Navy and serving aboard warships for many months. Álvaro Pereira do Nascimento, “Do cativeiro ao mar: Escravos na Marinha de Guerra,” Estudos Afro-Asiáticos, no. 38 (Dez. 2000), pp. 1–25.

⁶⁷ Fábio W. A. Pinheiro, “Tráfico atlântico de escravos na formação dos plantéis mineiros, Zona da Mata, c. 1809–c. 1830” (Dissertação de mestrado, UFRJ, Rio de Janeiro, 2007), tabela 6, p. 75.

⁶⁸ Vinicius Pereira de Oliveira, “Sobre o convés: Marinheiros, marítimos e pescadores negros no mundo atlântico do Porto de Rio Grande/RS (século XIX),” IX Encontro Estadual de História – ANPUH-RS (Porto Alegre, 2008), pp. 3–5

⁶⁹ Paulo Roberto Staudt Moreira, Os cativos e os homens de bem: Experiências negras no espaço urbano, Porto Alegre 1858–1888 (Porto Alegre: EST edições, 2003), p. 74.

⁷⁰ Moreira, Os cativos e os homens de bem, p. 75.