Wealth holding in southeastern Brazil, 1815-60

HAHR - Hispanic American Historical Review

Volumn 85, Issue 2, 2005, Pages 223-257

  Frank, Zephyr ^a  

^a NONE

Author keywords

[No Author keywords available]

Indexed keywords

EID: 65849479469     PISSN: 00182168     EISSN: 15271900     Source Type: Journal    
DOI: 10.1215/00182168-85-2-223     Document Type: Article

References (103)

1. See, e.g.,João Luís Ribeiro Fragoso, Homens de grossa aventura: Acumulação e bierarquia napraça mercantil do Rio de Janeiro, 1790-1830 (Rio de Janeiro: Arquivo Nacional, 1992)
2. See, e.g., Werner Baer, The Brazilian Economy: Growth and Development (Westport, CT: Praeger, 1995), 17-18
3. Warren Dean, The Industrialization of São Paulo, 1880-1945 (Austin: Univ. of Texas Press, 1969), 3
4. Zélia Maria Cardoso de Mello, Metamorfoses da riqueza: São Paulo, 1845-1895 (São Paulo: HUCITEC, 1982)
5. Nathaniel Leff, Underdevelopment and Development in Brazil: Economic Structure and Change, 1822-1947, vol. 1 (London: Allen & Unwin, 1982), 33 and 42. Note that Leff also provides an upper-bound estimate for growth: 0.8 percent nationwide, 1 percent in the Southeast
6. For the domestic-use agriculture sector and retarded economic growth, see Leff, Underdevelopment, 20-22
7. for schooling rates and the human capital argument, see Stanley Engerman, Stephen Haber, and Kenneth Sokoloff, "Inequality, Institutions, and Economic Growth: A Comparative Study of New World Economies Since the Sixteenth Century," in Institutions, Contracts, and Organizations: Perspectives from New Institutional Economics, ed. Claude Menard (Cheltenham, UK: Edward Elgar, 2000), 108-34, esp. 122-23 and 126-27
8. for transport costs, see William Summerhill, Order against Progress: Government, Foreign Investment, and Railroads in Brazil, 1854-1913 (Stanford: Stanford Univ. Press, 2003), esp. chap. 2
9. for terms of trade, see Celso Furtado, The Economic Growth of Brazil: A Survey from Colonial to Modern Times (Berkeley: Univ. of California Press, 1965), 116-18
10. Important examples of this literature include Amilcar Martins Filho and Roberto B. Martins, "Slavery in a Nonexport Economy: Nineteenth-Century Minas Gerais Revisited," Hispanic American Historical Review 63, no. 3 (1983): 537-68
11. Douglas Libby, Transformação e trabalho em uma economia escravista: Minas Gerais no século XIX (São Paulo: Editora Brasiliense, 1988)
12. João Fragoso, Homens
13. and Afonso de Alencastro Graça Filho, A princesa do oeste e o mito da decadência de Minas Gerais: Sāo João del Rei 1831-1888 (São Paulo: Annablume, 2002)
14. Anuário Estatístico (Rio de Janeiro: IBGE, 1939/40), 1358. Exports rose from 4,030,000 pounds sterling in 1822 to 4,941,000 pounds in 1844-45 and 10,439,000 pounds in 1854-55
15. Stanley Stein's classic study of Vassouras begins in 1845-well after independence-so his data do little to answer the question posed in this essay. Stein indicates that his data are derived from 93 plantation inventories over the period 1820-88; Vassouras: A Brazilian Coffee County (Cambridge, MA: Harvard Univ. Press, 1957), 76
16. Katia M. de Queirós Alattoso's study of Salvador, Bahia, século XIX (Rio de Janeiro: Nova Fronteira, 1992), is methodologically impressive but uses ad hoc categories rather than standard percentages or deciles in the presentation of data. This renders her findings hard to use in a comparative manner. Cardoso de Alello's Metamorfoses is methodologically problematic and also begins at the end of the period proposed in this essay. We learn, for instance, that certain estate inventories were included and others omitted - but we do not know precisely why this was so or what difference it made to the findings. She writes on p. 30, for instance, that "[i]nventories that contained low levels of wealth, that is, of very poor decedents without diverse possessions (for example, someone who owned just a house) do not interest us." Finally, João Fragoso's analysis of wealth holding in Rio de Janeiro suffers from small sample size. For instance, he based his understanding of early nineteenth-century wealth on just 36 observations for 1820; Homens, 255
17. The single important exception is Ricardo Antonio Rocha Silveira, "The Distribution of Wealth in Brazil - The Case of Rio de Janeiro: 1870s to 1980s" (Ph.D. thesis, Univ. of California at Berkeley, 1985)
18. For a discussion of this problem, see Peter Lindert and Jeffrey Williamson, American Inequality: A Macroeconomic History (New York: Academic Press, 1980), 13
19. Note, however, that the opposite may just as well have been the case. Douglas Libby and Afonso de Alencastro Graça Filho have uncovered some evidence for censoring at the top end of the wealth distribution in Minas Gerais; Libby, "Minas na mira dos brasilianistas," in História quantitativa e serial no Brasil, ed. Tarcísio Rodrigues Botelho et al. (Belo Horizonte: ANPUH-AIG, 2001), 296
20. For an excellent overview of the problems of working with estate inventories (probate records) and some good advice on how to proceed, see Peter Lindert, "An Algorithm for Probate Sampling," Journal of Interdisciplinary History 11, no. 4 (1981): 649-68
21. Rio de Janeiro's municipal population was roughly 114,000 in 1822, rising to approximately 198,000 in 1860; Anuário estatístico do Brasil (Rio de Janeiro: IBGE, 1939/40), 1294
22. The city of São João del Rei contained about 8,000 people in 1831-32, while the population of neighboring São José was around 3,000; by the 1872 census, the município of São João had grown to 28,008 (town center, 10,901), of whom fewer than 200 were foreign nationals; São José (today known as Tiradentes) accounted for another 7,670 inhabitants. Population data for 1831/32 in Clotilde Andrade de Paiva, "População e economia nas Minas Gerais do século XIX" (Ph.D. thesis, University of São Paulo, 1996), 166-80
23. for 1872, see Recenseamento do Brasil (Rio de Janeiro: Leuzinger, 1872), Provincia de Minas Gerais, 1069-73
24. Laird Bergad's thorough study of slave demography in Minas Gerais, Slavery and the Demographic and Economic History of Minas Gerais, Brazil, 1720-1888 (Cambridge: Cambridge Univ. Press: 1999), contains age pyramids for 1808 and 1872. Although Bergad's figures show an anomalous bulge in the 21-30-year-old bracket for 1872, he argues that this change is a statistical mirage based on an overcount of the 21-30 group and an undercount of children, 107. Beyond this, Bergad does find strong evidence for shifts in the racial composition of Minas Gerais' various regions and in the concentration of slaves, 112-22
25. The most important institutional change occurred in 1847 with the adoption of a new law regarding inheritance rights for illegitimate children. The law gave fathers greater power to recognize illegitimate children but, at the same time, required more formal action to recognize such children and limited the circumstances in which such recognition was possible. Although this law was enormously important socially and politically, it probably did not lead to major changes in the size and distribution of wealth reported in the estate inventories after 1847. For a thorough description of the law and its ramifications, see Linda Lewin, Surprise Heirs II: Illegitimacy, Inheritance Rights, and Public Power in the Formation of Imperial Brazil, 1822-1889 (Stanford: Stanford Univ. Press, 2003), esp. 290-97
26. For example, the minimum value of the category of personal items in São João was four mil-réis in the first period and eight in the second; in Rio de Janeiro the values were four and one, respectively. In both cases, the threshold at which wealth was recorded was very low
27. Calculated from Mary Karasch, Slave Life in Rio de Janeiro, 1808-1850 (Princeton: Princeton Univ. Press, 1987), 61. A similar calculation for 1821 yields 10.8 slaves per slaveholding household. The sire of slaveholdings remained fairly stable, as did the percentage of the enslaved population in the city
28. Estate Inventories, Arquivo Nacional, Rio de Janeiro. The sample includes 120 estates, 95 slaveholders, and 1,174 slaves
29. According to the samples, the mean slaveholding was 8.44 in 1820, 10.25 in 1850-54, and 11.46 in 1855-60. In all of these cases, the rural slaves of urban decedents are included in the averages
30. Clotilde de Paiva, machine-readable census for São José (1831/32) and São João (1838). Kote that slaves are not enumerated in the city of São João
31. In order to test the relationship between slave wealth and other wealth, I ran simple OLS regressions with nonslave wealth as the dependent variable for the 1815-25 sample. The resulting R-squared values were small: 0.19 for Rio de Janeiro and 0.47 for São João. Slave wealth thus "explains" 19-47 percent of the variation in nonslave wealth
32. By strong links, I mean exact matches of names, with exceptions in two types of cases. First, unusual names were matched, even when the entire name was absent in one of the two sources. Second, in three cases I matched the most likely candidate from a short list of repeated names (2 in two cases, 3 in the other), with the added restriction that the resulting match not be significantly different from the other possible matches in any important dimension (age, wealth, etc.)
33. This calculation is based on three rather than four age categories in order to avoid very small cell sizes. Mean wealth for the younger group (15-39, N = 16) was 5:065$, for the middle group (40-59, N = 18) it was 17:235$, and for the older group (60+, N = 15) it was 13:598$. The average age of decedents in this small sample was 48
34. Lançamento de décima urbana de vila de São João del Rei, 1826, Museu Regional de São João del Rei (MRSJDR). I am grateful to Afonso de Alencastro Graça Filho for providing me with this data
35. Décima Urbana, Rio de Janeiro, 1849, Arquivo da Cidade, Rio de Janeiro. A total of 16,175 properties with individual addresses were listed in the tax rolls. Against this number, I was able to identify approximately 6,000 individual property owners. This latter figure is approximate, because a significant number of names in the database are sufficiently similar to be ambiguous (possible but not definite matches) and others may be homonyms. In addition, religious institutions, small in number, owned about 10 percent of urban property. Notwithstanding these limitations, the figure of 6,000 individual property owners is a conservative estimate, based on matching as many very similar names as possible, thereby biasing the estimate against finding too many individual property owners. It should be emphasized here that the estimate offered is rough. It assumes that each property owner occupies a single address and rents out the remainder of their properties to others without urban real property. It does not account for multiple occupancy of owned or rented property. In 1849, multiple-occupancy was still limited, but as tenements expanded in the city over the second half of the nineteenth century, this assumption becomes increasingly problematic
36. The assumption that 50 percent of households held zero wealth is considered an extreme lower-bound assumption by Silveira, "The Distribution of Wealth in Brazil," 106 and no
37. Alice Hanson Jones, Wealth of a Nation to Be: The American Colonies of the Eve of the Revolution (New York: Columbia Univ. Press, 1980), 365
38. The presence of negative net estates in the samples is one indication. Clearly, these debts were not cleared. In Rio de Janeiro during the 1850s, unpaid debts comprised 5 percent of the value of all debts in the sample. Second, there are a few instances in the records where the executor of the estate complained that debts owed to the estate were not paid. Third, casual inspection of lists of outstanding credits reveals a small but not insignificant number of cases in which a fraction of the total debt was repaid - some debts were only partly recouped
39. The exchange rate is a reasonable proxy for price levels in the tradable sector in Brazil during the nineteenth century and is in keeping with the common practice in the literature of rendering mil-réis values in pound sterling equivalents. Leff's purchasing power index (exchange rate and British wholesale prices) is statistically very similar to a comparable mil-réis/pound sterling index. For the PPP index, see Leff, Underdevelopment, vol. I, appendix table 4
40. Others have noted problems with Lobo's price index. For explicit criticism, see Mattoso, Bahia, 557-60
41. and Carlos Manuel Peláez and Wilson Suzigan, História vionetária do Brasil (Brasília: Editora Universidade de Brasília, 1976), 112-13
42. Lobo's 1919-weighted index shows a more moderate rise in prices over the nineteenth century. This index has been used recently, in lieu of the 1856-weights, by Summerhill, Order against Progress, 86-90. Summerhill also rejects Lobo's 1856-weighted index as unreliable
43. Price series generously provided in machine-readable form by Afonso de Alencastro Graça Filho. Graphical depictions of the series are found in Graça Filho, A princesa do oeste
44. At this point, it is worth asking how accurately the probate sample accounts for the probable level of wealth per head in Rio de Janeiro based on alternative measures. A simple exercise based on the prices for slaves and houses suffices to show that our estimates are near the mark. Circa 1821, there were about 10,943 houses and 86,323 residents in greater Rio de Janeiro (Karasch, Slave Life, 61), including 40,376 slaves. The mean values of slaves and houses reported in the inventory records were 114 (N = 153) and 1,603 (N = 175) mil-réis, respectively. Consequently, the total value of urban real estate and slaves, per head, was ((10,943 × 1,603) + (40,376 × 114)) / 86,323 = 257 mil-réis
45. For a discussion of inheritance law and its effect on wealth holding, see Frank, Dutra's World: Wealth and Family in Nineteenth-Century Rio de Janeiro (Albuquerque: Univ. of New Mexico Press, 2004), chap. 6. The evolution of inheritance practices in Brazil is best described in Lewin's two-volume study, Surprise Heirs
46. Jones, Wealth of a Nation to Be, 349
47. Renato Leite Marcondes, A Arte de acumular na economia cafeeira: Vale do Paraíba, Século XIX (Lorena: Stiliano, 1998), 311-18. The number of observations per period used in this calculation were 44 and 27 respectively
48. Mattoso, Bahia, 613-16 (tables 108-11). The nominal value was 10,025 mil-réis, which I converted to pounds at the 1822 exchange rate of 4.898
49. Converted at the average exchange rate of 8.93 mil-réis to the pound
50. Kevin O'Rourke and Jeffrey Williamson, Globalization and History: The Evolution of a Nineteenth-Century Atlantic Economy (Cambridge, MA: MIT Press, 1999)
51. Douglass North, Institutions, Institutional Change, and Economic Performance (Cambridge: Cambridge Univ. Press, 1990)
52. Emília Viotti da Costa, The Brazilian Empire: Myths and Histories (1985; Chapel Hill: Univ. of North Carolina Press, 2000), 183, hints at the importance of this institutional change but limits its presumed effects to the coastal provincial capitals
53. For a clear discussion of the importance of the Conservative Party (Party of Order) in the consolidation of the regime, see Needell, "Party Formation and State-Making: The Conservative Party and the Reconstruction of the Brazilian State, 1831-40," Hispanic American Historical Review 81, no. 2 (2001): 259-308
54. The best depiction of the reign of D. Pedro II, and his emergence as a major political force, is found in Barman, Citizen Emperor: Pedro II and the Making of Brazil, 1825-91 (Stanford: Stanford Univ. Press, 1999), esp. chap. 4
55. The literature on the reasons behind the new land law and its effects is large. For a Marxist modes-of-production account, see José de Souza Martins, Fronteira: A degradação do outro nos confins do humano (São Paulo: HUCITEC, 1997), 11-24, 35,and 151-52
56. Emilia Viotti da Costa, Da senzala a colônia (1966; São Paulo: UNESP, 1997), 35
57. Levy, História, 55
58. For internal sources of growth in São Paulo, see Elizabeth Kuznesof, Household Economy and Urban Development: São Paulo, 1765-1836 (Boulder: Westview, 1986)
59. for internal domestic growth in Minas Gerais, see Libby, Transformação; for Rio de Janeiro, see Fragoso, Homens
60. For a definition of Fragoso's terms, see ibid., 27-28; for the symbiosis between plantation agriculture and domestic markets, see 147
61. e.g., Winifred Barr Rothenberg, From Market-Places to a Market Economy: The Transformation of Rural Massachusetts, 1750-1850 (Chicago: Univ. of Chicago Press, 1992)
62. Rothenberg's argument is based on the idea that markets provide incentives (price mechanism) for households to allocate their resources more efficiently and to raise productivity, even in the absence of technical change or infusions of capital. For another useful study of internal domestic growth of the "industrious" type, see David F. Weiman, "Farmers and the Market in Antebellum America: A View from the Georgia Upcountry," Journal of Economic History 47, no. 3 (Sept 1987): 627-47
63. Jan de Vries, "The Industrial Revolution and the Industrious Revolution," The Journal of Economic History 54, no. 2 (June 1994): 249-70
64. Such growth may have been common throughout much of the New World during the first half of the nineteenth century. Lyman Johnson, for instance, records a fivefold increase in mean wealth holding in Buenos Aires province between 1829/30 and 1855/56; "The Pampa Transformed," ms. (generously provided by the author), 12. According to Johnson, these figures are subject to further review. What is clear, however, is that mean wealth holding grew at a healthy clip over a roughly analogous period in Argentina
65. Fragoso, Homens, 146-47
66. For the original formulation of this idea, see Martins Filho and Martins, "Slavery in a Nonexport Economy," 537-68
67. Libby, Transformação, 153-55 and 196-214
68. For similar evidence from São Paulo, see Kuznesof, Household Economy, 150
69. In a recent article, Nicola Verdon has argued that failure to account for women's and children's contribution to household income has tended to skew our impression of standards of living and economic activity in preindustrial societies. She finds, for instance, that women and children accounted for fully 37 percent of family income in England in 1834
70. "The Rural Labour Market in the Early Nineteenth Century: Women's and Children's Employment, Family Income, and the 1834 Poor Law Report," Economic History Review 55, no. 2 (2002): 299-323, esp. 316-18
71. Bergad, Slavery, 40
72. Graça Filho, cited in Libby, "Minas na mira dos brasilianistas," 173
73. Paiva, "População," 52, reports a population in the southeastern zone of 69,098. The machine-readable databases created by Paiva indicate a total of 7,744 households in the municípios of São João and São José
74. Sample of credits and debts drawn from estate inventories of major credit entrepreneurs of São João del Rei, MRSJDR: Francisco José Santiago (N = 506), Francisco José Dias (N = 354), Francisco de Paula Almeida (N = 665), Manoel José Baptists (N = 155), Manoel José Machado (N = 41), and Carlos Baptista (N = 111)
75. Marcondes analysis of dívidas de rol and créditos in 1857 shows that the former accounted for 197 loans and 38 percent of total credit, whereas the latter (in larger denominations and at interest) made up 12 loans and 62 percent of credit. Marcondes, A arte de acumular, 217. Marcondes also reports that mean credit (not including dívidas de rol) extended in Lorena was about 1,900$ during the 1850s, p. 230
76. Sample drawn from Inventories, São José, MRSJDR
77. Philip T. Hoffman, Gilles Postel-Vinay, and Jean-Laurent Rosenthal, "Information and Economic History: How the Credit Market in Old Regime Paris Forces Us to Rethink the Transition to Capitalism," American Historical Review 104, no. 1 (1999): 69-94
78. Kuznesof, again, finds much the same pattern for São Paulo, where she notes that rural producers shifted dramatically from subsistence to market production, 150
79. Inventory of Antonio Joaquim de Oliveira, 1847, MRSJDR, caixa 179
80. I obtained these results by crossing the list of Antonio Joaquim's creditors with the census for 1831/32. Machine-readable dataset generously provided Clotilde de Paiva. For additional material on the status of the free colored population in Southeastern Brazil, see Herbert Klein and Clotilde de Paiva's articles on Minas Gerais, "Freedmen in a Slave Economy: Minas Gérais in 1831," Journal of Social History 29, no. 4: 935-62
81. and "Slave and Free in Nineteenth-Century Minas: Campanha in 1831," Slavery and Abolition 15, no 1: 1-21
82. For São Paulo and Minas, see Herbert Klein and Francisco Vidal Luna, "Free Colored in a Slave Society: São Paulo and Minas Gerais in the Early Nineteenth Century," Hispanic American Historical Review 80, no. 4 (2000): 913-41
83. For Bahia, see Bert Barickman, "As cores do escravismo: Escravistas "pretos," "pardos" e "cabras" no Recôncavo baiano, 1835," População e Família 2 (1999): 7-59
84. Bergad's data for exports from Minas Gerais indicate that, even after the rise of coffee exports in the 1850s (to the international market), roughly 25 percent of provincial exports continued to be pastoral products: cheese, bacon, pigs, and cattle to the domestic market, mainly Rio de Janeiro. Inferred from Slavery, 54, fig. 2.7
85. Marcondes's research on Lorena, in São Paulo's part of the Paraíba Valley, suggests that two-thirds of the debts owed to merchants and moneylenders in Lorena were mediated through Rio de Janeiro, Arte de acumular, 225-26
86. To be sure, growth in real wealth is not the same as growth in per capita income. It is possible that the rate of growth in the latter was lower and that the spread in the ratio of wealth to income widened over the period under review. Nevertheless, it is hard to believe that real incomes did not also increase, especially in São João, where the urban disadvantages of high food and housing prices were less pronounced than in Rio
87. Arte de acumular, 171.I converted Marcondes's percentage from net to physical wealth by dividing his percentage of wealth held in slaves (27.8) by 0.779 to account for the weight of credits and debts in his original figure
88. Graça Filho, Princesa do oeste, 122-34. Graça Filho found, for instance, that the mean number of cattle on a fazenda in São João was 146 in the first half of the nineteenth century and 133 in the second half. Additionally, his analysis of production in the agricultural sector highlighted the tendency toward policulture, with plantings of rice, beans, and corn interspersed with sugar and coffee
89. In 1825, the average price of a male slave aged 21-40 was 161 mil-réis (N = 62) in Rio and 189 (N = 25) in São João. A generation later, in 1855, the values were 831 (N = 84) and 918 (N = 91) mil-réis, respectively. Interestingly, roughly the same premium obtained in São Paulo: 194 (N = 15) and 883 (N = 18) mil-réis in the same years; Estate inventories from AN, Rio de Janeiro; Museu Regional, São João del Rei; Arquivo do Tribunal do Judiciário, São Paulo
90. For a discussion of the effects of the commercial code on company formation after 1850, see Richard Graham, Britain and the Onset of Modernization in Brazil, 1850-1914 (Cambridge: Cambridge Univ. Press, 1968), 25
91. the idea that rising slave prices pushed investors toward other investments is broached by Emilia Viotti da Costa in Da senzala a colônia, 35
92. This adjustment is similar to Jones's concept of "non-probate wealth."
93. After adjustment, 42 percent of the sample were slaveholders and 65 percent owned real property. Among the living, circa 1831/32, 38 percent owned at least one slave; in neighboring São João, 68 percent of houses were owner-occupied
94. See, e.g., Schwartz, "The Colonial Past: Conceptualizing Post-Dependentista Brazil," in Colonial Legacies, ed. Jeremy Adelman (London: Routledge, 1999), 188, citing Alice P. Canabrava
95. Libby is one of the few scholars to have noted this problem, Transformação, 132n30
96. Schwartz, "Patterns of Slaveholding in the Americas: New Evidence from Brazil," American Historical Review 87, no. 1 (Feb. 1982): 55-86
97. and Francisco Vidal Luna and Herbert Klein, Slavery and the Economy of São Paulo, 1750-1850 (Stanford: Stanford Univ. Press, 2003), 109-113
98. I know of no systematic study of the relative importance of slaves to decedents at different levels of wealth. Most studies merely report the percentage of total wealth accounted for by slaves. Cardoso de Mello, Metamorfoses, does provide a rough indicator when she disaggregates her data into three classes of wealth (top 33.3% and so on), 93-98. Her figures, however, are suspect on several accounts, begin only in the 1840s, and are still too aggregated (she should have reported her findings by decile)
99. Percentages calculated with Gini coefficients to three decimal places
100. The Gini coefficients for total physical wealth among wealth holders in the Rio de Janeiro sample (minus debts and credits) was 0.704 in 1820 and 0.711 in 1855; the coefficients for total net wealth (with debts and credits) were 0.78 and 0.737 respectively
101. Although a strong caveat is in order regarding the sample size, especially for the earlier period, the data for São Paulo indicate a rise in the portion held by the richest decile from 42.6 to 62.4 percent of physical wealth
102. Stein, Vassouras, 64: "By the mid-thirties the finances of the Empire were based on the prosperity of the coffee planter, a dependence stronger than pressure of English diplomacy applied for twenty years. Increased slave imports in the forties profited planters and government coffers alike." Here and throughout, Stein refers to slaveholders as "planters" - fair enough given the subject of his book, but misleading if taken to indicate that planters were the only slaveholders or supporters of the institution with a vested interest
103. Richard Graham, Patronage and Politics in Nineteenth-Century Brazil (Stanford: Stanford Univ. Press, 1990), 32-33