Pre-transitional mortality indicators for a Brazilian region in the - - PDF document

1

Pre-transitional mortality indicators for a Brazilian region in the 19th century Dayane Julia Carvalho Dias – Unicamp – dayanejuliacd@gmail.com Luciana Conceição de Lima – UFRN – limamarx@gmail.com Luana Junqueira Dias Myrrha – UFRN – luanamyrrha@gmail.com 1) Introduction This article is a demographic study of mortality in Rio Grande do Norte (currently a state placed at Northeast of the Brazil) during the nineteenth century. It contributes to a still scarce set of studies in Brazil that are dedicated to the pre-demographic transition which was marked by the occurrence of epidemic diseases. In Brazil, the 19th century was marked by the occurrence of epidemic diseases due to the poor environmental and medical-sanitary conditions of the time (CHALHOUB, 1996; DINIZ, 1997; MARTINS, 2012). The most effective control of these diseases by the national government would

• ccur only a century later, more specifically in the 1940s, when explicit policies of

public health and basic sanitation contributed decisively to the beginning of the transition of mortality in the country (PRATA, 1992; WOOD & CARVALHO, 1998; ALVES, 2002). Thus, until this period that starts the process of demographic transition in the country, infectious diseases played an important role in the structure of causes of death of the population. At the beginning of the 20th century, infectious diseases accounted for 60% of the causes of death in the North Region, 49% in the Northeast Region, 43% in the Southeast Region, 40% in the Central-West Region, and 39% in the Southern Region. The external causes were not significant, representing only 2% of deaths in areas such as the North, Northeast and South regions (PRATA, 1992). Infectious diseases reach younger and younger people, especially those with unfavorable socioeconomic conditions (FRENK et al, 1991). This article is divided into five sections. The first one presents a brief overview of the mortality in the Rio Grande in the late eighteenth and early nineteenth centuries. The second section describes the methodology used (materials and methods), the discussion

• f the main results is found in the third section, the fourth section presents the

conclusions of this study, and finally, the fifth present the discussion about the results

• btained.

2

2) The context of mortality in Rio Grande do Norte in the 19th century: socioeconomic and sanitary conditions Important transformations involved Rio Grande do Norte between the end of the eighteenth century and the first decades of the nineteenth century: replacement of the Portuguese commercial monopoly by the commercial freedom highlighted by the presence of the English economy; the transition from colony to politically independent nation; and the need for political reorganization, with a view to maintaining the monopoly of large landowners and, at the same time, ensuring territorial unity (MONTEIRO, 2002). However, amidst these important changes, the succession of droughts plagued the captaincies of the north. Between 1791 and 1793 occurred the greatest drought of the eighteenth century, reaching not only the captaincy of Rio Grande do Norte but also Ceará, Pernambuco, Bahia, Sergipe, Alagoas, Paraíba and Piauí. In addition to the pests

• f locusts, snakes and rats, droughts were accompanied by epidemics such as smallpox

(popularly known as bladder), producing even more dead (Guerra, 1981). During the provincial period, the health conditions of Rio Grande do Norte were not different from the other Brazilian provinces. The Empire of Brazil suffered constant epidemics and, in this context, a series of measures were taken to combat them, initially in the Court, Rio de Janeiro, and later in the other provinces of the Empire. The action

• f the provincial government was no more than a count of the victims, since the medical

conditions of the population were precarious. Medical performance was limited, since it was difficult to reach underserved areas in a short period of time, and the scientific knowledge offered at the time was insufficient (MATTOS, 1985). In the case of the province of Rio Grande do Norte, the state of health was precarious, with irregular sanitary conditions that facilitated the spread of infectious diseases (ARAÚJO and MACEDO, 2011). In the second half of the 19th century, more specifically, the province of Rio Grande do Norte was affected by epidemic outbreaks

• f cholera-morbidity, yellow fever and smallpox (SANTOS, 2013). The Charity

Hospital, built in the city of Natal in 1856, "was a house of suckers for 40 sick men and as many women. Media 176 by 56 palms wide and occupied all the masons and carpenters of Natal "(CASCUDO, 1984). Even considered a great achievement, because before the capital had only a military ward and a pharmacy (BRAZIL, REPORT OF

3

PRESIDENT OF PROVINCE, 1856), the place was described by Pedro Velho (health inspector at the time) in 1886, as: "In the conditions in which the Hospital is nowadays, it admires even how surgical operations are successfully performed there, such is the risk that an operative is in being in that infected environment" (CASCUDO, 1984). This description demonstrates the conditions of the only existing hospital in Natal, which, according to Cascudo (1984): "the poverty of the funds explained the misery of the environment, but did not justify the filth indignantly recorded in some administrative documents." Cascudo (1984) refers to the statement made by Vice-President Vicente Inácio Pereira, the first doctor in 1879: "The most filthy chapel, in complete abandonment of all the laws of the most coarse hygiene, can not take to the benefit of these unfortunate, by the state of penury in which the coffers of the province are found". 3) Metodology The data sources used were statistical population maps (population and death data) and data from the Census of 1872 and 1890 (population data). The statistical maps refer to the years 1801 and 1805 were deposited in the Historical Archive Ultramarino de Lisboa (AHU) and made available for purposes of this article by the project Counting Colonial Populations: Demography and the use of statistics in the Portuguese Empire, 1776-1890 New from Lisbon. States have always needed to enumerate, measure, quantify their populations, wealth, and resources in order to know the nation through statistics. Initially, this was due to the recruitment of soldiers into wars and then to legislating and administering many aspects

• f the public and private spheres. With the development of the modern world and

capitalism, the difficult issues of birth, mortality, longevity, public health, housing, migration, among others emerged (SENRA, 2006). The first census survey of Brazil occurred in the year 1776 and constituted a landmark

• f change for the proto-statistical period (BOTELHO, 1998). This survey was carried
• ut by means of population maps in the management of the Marquis of Pombal, which

brought a modernizing character inspired by the European ideas of population survey (PAIVA et al, 2012). It aimed to 'know the Brazilian reality' to promote the colony's agricultural renaissance and the production of raw materials for the industrialization of Portugal (BOTELHO, 1998).

4

According to Paiva (2015), the maps that present the greatest detail of their information are those derived from the royal diplomas of 1796 and 1797, such as the maps of 1801 and the maps of 1805. Based on the characteristics of the maps used for the Captaincy

• f the Rio Grande in 1801 and 1805, it was observed that these documents included

information on the population by age group (quinquennial and decennial) and sex (male and female), divided into color categories (white, black and mulatto), marital status (married, single and widowed) and condition (free, slave or indigenous), the natives being those 'domesticated' and not those who actually inhabited the captaincy. It also contains birth, marriage, and death information. Marriage information is divided by 10- year-old age groups (10 to 70 years old), sex, color (white, black and brown) and condition (free and captive). For births, information is given by sex, color (white, black and brown), condition (free and captive), and live, dead, and twin births. By death, information is recorded by age groups (five-year-olds up to 10 years and up to 100 years

• ld), sex, color (white, black and brown) and condition (free and captive). In addition,

there are information of occupations of the inhabitants. Despite the making of statistical maps, simple work was not synonymous with quality. Little followed the standard model imposed by the provincial administration for data collection and those responsible for it were not always sufficiently qualified for such a

• function. There was no statistical rigor, and data did not always encompass the entire

province, resulting in poor quality of information. To evaluate the quality of the information, demographic techniques of population correction and death were used. Correction of population data The population of 1805 was estimated based on the population counted in 1872 considering the proportional distribution by age group of the population counted in

• 1890. In preliminary analysis, it was verified that the age structure of the population of

Rio Grande do Norte registered by the Census of 1872 was older in comparison with the population of 1890, which in principle represents a contradiction. Thus, based on the proportional distribution of the population by age group of the 1890 Census, the proportional distribution of the population by age group of the Census of 1872 was

• recalculated. Thus, assuming constant population growth by age group in the period of

pre-transition, from 1801 to 1890 the population was considered to have grown at a rate

• f 2.2% per annum (value adopted as a parameter in previous work for the period from

5

1818 to 1872 as stated in the provincial estimates for the period 1818 to 1872 for the Rio Grande do Norte (Rodarte, 2008)). Therefore, the population of Rio Grande do Norte in 1872, redistributed proportionally by age group according to the age distribution of 1890, the population of each age group was decremented by year, until 1805, using an annual growth rate of 2, 2%. The next item presents the methodology presented for the correction of deaths. Correction of death data For the analysis of mortality, information on deaths contained in the statistical map of 1805 was used, since the information in this chart was more coherent than the information in the statistical map of 1801, which presented few cases, especially for the infant population. Initially, for the correction of adult mortality (age groups over 20 years), Brass's Growth Balance method was adopted (1975). This method is based on the assumption that it is possible to adjust the difference in coverage between deaths and population, making it possible to calculate the specific mortality by age-adjusted rates, considering that the population is stable and closed. Thus, the coverage of deaths does not vary with age. The main idea is that in a closed population or with small migratory flow, the forms of entry of the population are the births, and the exits, the deaths. To estimate infant and juvenile mortality (below 20 years), the methodological strategy was to use an already existing mortality function with the criterion of choice that was closest to the estimated mortality function for adult mortality. After analyzing the different levels of Coale and Demeny (1996) tables of the west (6) and east (18) models, and 19 different mortality functions from different countries available on The Human Mortality Database, we adopted function of Iceland in 1860. Finally, in order to reconcile the two parts of the mortality function (infant and juvenile and adult), indirect standardization was used, assuming that the structure of the mortality function chosen to represent the mortality of the ages under 20 years was also adequate to represent the mortality of the ages over 20 years. Indirect standardization is based on the assumption that in order to estimate a given function, it is necessary to adopt a known function of specific rates from another population and assume that both functions have the same structure (CARVALHO,

6

SAWYER and RODRIGUES 1998). In order to adjust the level of the function as a whole, only mortality information over 20 years was used, since for such age groups we have "corrected observed deaths" and estimated deaths based on the function adopted as standard. Thus, a new mortality function was obtained for the state of Rio Grande do Norte in 1805, which was used to estimate life expectancy by age group. If the population of Rio Grande do Norte is considered to be quasi-stable during the period from 1801 to 1890, it means that mortality was also constant and, therefore, the estimated mortality function for 1805 should represent the behavior of mortality throughout the period. The next section presents the mortality indicators obtained: the estimated mortality function after the adjustment of the adjusted functions of adult and child mortality from 1801 to 1890. Also, the life table for Rio Grande do Norte including life expectancy at birth. 4) Results Graph 1 presents the estimated mortality function that represents mortality levels and standards in Rio Grande do Norte between 1801 and 1890, assuming that the population

• f Rio Grande do Norte was stable during this period. According to the estimated

specific mortality rates, mortality in Rio Grande do Norte was more intense in the infant population and in the population from 50 years of age, more susceptible to adverse conditions such as droughts and infectious diseases. made present during much of the colonial and provincial period. There are differences between observed and estimated specific mortality rates in all age groups, especially in the age groups 0 to 4 years and

• ver 50 years. These results suggest that mortality intensity in Rio Grande do Norte was

higher than that recorded in the population map and that underreporting was concentrated at younger and advanced ages.

7

Graph 1 - Functions of mortality of Rio Grande do Norte: observed in 1805 and estimated (1801-1890)

0,00100 0,01000 0,10000 1,00000 0-4 5-9 10-19 20-29 30-39 40-49 50-59 60-69 70-79 80-89 log (TEM) Age groups Estimated age-specific mortality rate TEM observada Source: 1801: AHU. ACL. CU. B. Rio Grande do Norte, Caixa 9, Documento 565; 1805: AHU. ACL.

• CU. B. Rio Grande do Norte, Caixa 10, Documento 629; 1839, 1846 e 1870: Relatórios de Presidentes da

Província do Rio Grande do Norte; 1872: Censo Imperial de 1872; 1890: Recenseamento de 1890.

With the acquisition of the new mortality function for Rio Grande do Norte, it was possible to estimate the life expectancy of the population during the whole study period, which is part of the demographic pre-transition phase. According to Table 1, estimated life expectancy at birth for Rio Grande do Norte was 32.6 years, a value very close to Mortara's (1941) estimates for Brazil between 1870 and 1890 using two different tables

• f mortality: 32.7 and 33.9 years.

8

Table 2 – Life table, Rio Grande do Norte, 1801 a 1890

x n

nmx nax nqx

lx

ndx nLx

Tx ex 5 0.10272 2.5 0.4087 100000 40867 397832 3257507 32.58 5 5 0.01412 2.5 0.0682 59133 4031 285587 2859675 48.36 10 10 0.00691 5.0 0.0667 55102 3678 532628 2574088 46.72 20 10 0.00972 5.0 0.0927 51424 4767 490405 2041460 39.70 30 10 0.01091 5.0 0.1034 46657 4826 442443 1551055 33.24 40 10 0.01644 5.0 0.1519 41832 6355 386543 1108612 26.50 50 10 0.02259 5.0 0.2029 35477 7200 318773 722069 20.35 60 10 0.04479 5.0 0.3659 28277 10347 231038 403297 14.26 70 10 0.08685 5.0 0.6055 17930 10857 125014 172259 9.61 80 10 0.14147 5.0 0.8286 7073 5860 41426 47245 6.68 90 ∞ 0.20837 4.8 1.0000 1212 1212 5819 5819 4.80

Source: 1801: AHU. ACL. CU. B. Rio Grande do Norte, Caixa 9, Documento 565; 1805: AHU. ACL.

• CU. B. Rio Grande do Norte, Caixa 10, Documento 629; 1872: Censo Imperial de 1872 e Relatórios de

Presidentes da Província do Rio Grande do Norte.

This result is compatible with that expected for a period of demographic pre-transition, and the similarity between the values of life expectancies estimated by this study and that estimated by Mortara (1941) also for the nineteenth century, corroborates the argument that in the period before the consistent decline in the volume of deaths, mortality varied little over time, and at least based on these findings, also between

• regions. In the international comparison, the life expectancy at birth found about 33

years from 1801 to 1870 for Rio Grande do Norte is also a consistent result. Also, according to Mortara (1941), in more advanced countries during the nineteenth century such as England and France, life expectancy was much higher (44.16 and 42.13 years respectively), while in Eastern European countries, less developed, the values were lower: 27.65 years in Russia and 27.65 years in Austria. 5) Discussion In general, the present article reached the objective of estimating mortality indicators for Rio Grande do Norte in the context of demographic pre-transition. Although the task of dealing with information whose sources were not designed specifically for demographic analysis (in the case of population maps) or with censuses delineated before the modern era of census surveys has been difficult, the results found were consistent. Although it is important to recognize the possible limitations of this work, such as the absence of data information from other components of population dynamics for the period that could have been incorporated in analyzes such as fecundity and migration,

9

the results obtained contribute to the quantitative reflections on the Brazilian past. It is therefore suggested that future work should focus on the collection of information sources in which data on migratory flows and fertility experiences can be obtained for the proto and pre-statistical periods, such as parish registers and new maps population. Another important point to highlight is the methodology used. For example, to estimate mortality there are several methods other than that adopted in the present study that can be used such as the Exton Generations of Preston et al (1980), Bennett and Houriuchi (1981), Hill (1987) and Hill and Choi ( 2004). And the testing of other methods may be a demand for future work to check whether the results found are still consistent or not. Even in this respect, it should be mentioned that despite the study by Mortara (1941), which chronologically precedes the methods currently used for estimating mortality as the method of Growth Balance by Brass (1975), the results found by the author for Brazil were very similar to those of Rio Grande do Norte in relation to life expectancy at birth, which initially reinforces the consistency of the indicator found. Finally, new studies on the demographic pre-transition period are recommended for

• ther localities in the country and beyond captaincies (and provinces) in Central-

southern Brazil. Formal demography has many options of techniques for dealing with data with problems as a record, and they can be very useful for quantitatively understanding demographic dynamics in pre-transition times. References ALVES, J. E. D. A polêmica Malthus versus Condorcet reavaliada à luz da transição demográfica. Rio de Janeiro: Escola Nacional de Ciências Estatísticas, 56 p. 2002. ARAÚJO, A. I. C.; MACEDO, M. K. O sertão febril: impacto microbiano e escravidão nos espaços (in)salubres da Província do Rio Grande do Norte, Ribeira do Seridó (1856-1888). Mneme (Caicó. Online), v. 12, p. 30, 2011. BENNETT, N. G.; HORIUCHI, S. (1981) Estimating the completeness of death registration in a closed population. Population Index, vol.47, n.2, p. 207-221.

10

BOTELHO, T. R. População e nação no Brasil do século XIX. 1998. 241 f. Tese (Doutorado em História) - Departamento de Ciências Sociais da Faculdade de Filosofia, Letras e Ciências Humanas, Universidade de São Paulo, São Paulo, 1998. BRASS, W. Methods for estimating fertility and mortality from limited and defective

• data. The University of North Carolina at Chapell Hill, International Program of

Laboratories for Population Statistics, 1975 (an occacional publication). CARVALHO, J.A.M; SAWYER, D.O.; RODRIGUES, R.N. Introdução a alguns conceitos básicos e medidas em demografia. Belo Horizonte: Série Textos Didáticos, N.1 ABEP, 63 p., 1998. CASCUDO, L. C. Capítulo XI. In: História do Rio Grande do Norte. 2. ed. Fundação José Augusto; Achiame: Natal/Rio de Janeiro, 1984. p. 275-289. CHALHOUB, S. Cidade febril. Cortiços e epidemias na corte imperial. São Paulo:Cia das Letras. 1996. COALE, A. J; DEMENY, P. Regional model life tables and stable population. New Jersey, Princeton University Press, 1966. DINIZ, A. S. Cólera: representação de uma angústia coletiva (a doença e o imaginário social no século XIX no Brasil) / Ariosvaldo da Silva Diniz. Campinas, SP: [s. n.], 1997. FRENK, J.; FREJKA, T.; BOBADILLA, J.L.; STERN, C.;LOZANO, R.; SEPÚLVEDA, J.; JOSÉ, M. La transición epidemiológica en America Latina. Bol. Oficina Sanit. Panam., 111:485-96, 1991. GUERRA, F. Secas do Nordeste. In: ROSADO, V (org). Memorial da Seca. Brasília, Centro Gráfico do Senado Federal/ESAM. Coleção Mossoroense, 1981. HILL, K. e CHOI, Y. Death distribution methods for estimating adult mortality: sensitivity analysis with simulated data errors. Adult mortality in developing countries workshop. The Marconi Center, Marin County, California, July 2004. HILL, K. Estimating census and death registration completeness. Asian and Pacific Population Forum, 1(3): 8-12, 1987.

11

MARTINS, L. L. Varíola em Fortaleza: marcas profundas de uma experiência dolorosa (1877-1881). Dissertação, mestrado, 30, mai. 2012. MATTOS, M. R. M. F. Vila do Príncipe-1850/1890 Sertão do Seridó – Um estudo de caso da pobreza. Dissertação (Mestrado em História). Universidade Federal Fluminense, Niterói.1985. 247 p. MORTARA, G. Estudos sobre a utilização do censo demográfico para a reconstrução das estatísticas do movimento da população do Brasil. VII Tabelas de mortalidade e de sobrevivência para os períodos de 1870-1890 e 1890-1920. Cálculo, exame e comparações internacionais. Rev. bras. Estat., p. 494-538, 1941. PAIVA, C. A, et al. D. Publicação Crítica do Recenseamento Geral do Império do Brasil de 1872. Janeiro, 2012. PAIVA, Y. A demografia da Paraíba. Uma análise preliminar, 1774-1820. Anais de Historia de Alem-Mar, Issue 16, p205-226. 22p, 2015. PRATA, P.R. The epidemiologic transition in Brazil. Cad Saude Publica. 1992;8(2):168-75. SANTOS, A. R. Cemitérios no Seridó, século XIX: construindo de uma pesquisa. Revista inter-legere, 2013. SENRA, N. de C. História das estatísticas brasileiras. v. 1. Estatísticas desejadas (1822-1889). Rio de Janeiro: IBGE, 2006. 614 p. WOOD, C. H.; CARVALHO, J. A. M. de. A demografia da desigualdade no Brasil. Rio de Janeiro: IPEA, 1994.