Reestimating Brazil’s Gdp Growth From 1820 To 1980: A Note On Ongoing Research

Introduction 

Our research on reestimating Brazil’s GDP and GDP per capita growth from 1820  to 1980 currently consists of three papers: “Reestimating Brazil’s GDP Growth  from 1900 to 1980,” published in Revista Brasileira de Economia in 2023; “A Note  on Brazil’s Historical GDP per Capita Growth Rates,” published in the same  journal in 2024; and “Secular Stagnation? A New View on Brazil’s Growth in the  19th Century,” published in Revista de Historia Económica/Journal of Iberian and  Latin American Economic History in September 2025. 

These papers have generated considerable interest in the research community, as  exemplified by Villela (2025), for 1900-1980, and by Bogart et al. (2024), for 1820- 1900. This note elaborates on their content and outlines new related research that  our papers have stimulated or that is being conducted independently. 

Completed research 

The first paper presents new estimates for Brazil’s real GDP and GDP per capita  growth rates between 1900 and 1980, focusing on four subperiods with  significantly distinct growth rates—1900–1919, 1919–1947, 1947–1966, and 1966– 1980. Our figures are obtained by adding slow-growing service sectors to Haddad’s  (1980) estimates for 1900–1947, and to IBGE’s (1990) estimates for 1947–1980.  For 1947–1980, we applied downward revisions to IBGE’s estimates that reduced  yearly GDP growth from 7.4% to 6.2%. For the 1900–1947 period, our revisions  reduced yearly GDP growth from Haddad’s 4.4% per year to 4%. 

The paper was presented in an online seminar at the Brazilian Institute of  Economics (IBRE/FGV-Rio). Partly as a consequence, IBRE has recently  established a research team led by Claudio Considera and Roberto Olinto to  undertake a revision of Brazil’s national accounts from 1947 to 1995. Results— expected for early 2027—should help assess the validity of our suggested revisions  for 1947–1980.

Our second paper is a research note comparing the GDP per capita growth rates  from 1820 to 1980, derived in our two other papers, with those in the Maddison  Project Database (2020) for 1820–1980 and in Ipeadata for 1900–1980—the latter  simply merging the Haddad (1980) and IBGE (1990) series. 

The third paper proposes new estimates for Brazil’s GDP per capita growth rates  from 1820 to 1900. It extends (for 1820–1850) and refines (for 1850–1900)  Goldsmith’s (1986) estimates for 1850–1900. Briefly, the methodology consists of  aggregating the growth rates of government revenue and expenditure, exports and  imports, and money supply (all in current prices), then dividing the resulting  aggregate series by an overall price index (and population estimates) to obtain a  proxy for real GDP per capita growth. Goldsmith’s estimates include an urban  wage series that we left out because we could not get data for the first half of the  century. A regression analysis for our GDP proxy indicates a trend growth rate of  0.9% per year from 1820 to 1900, with essentially the same rate for both the 1820– 1850 and 1850–1900 subperiods. 

Our estimates draw on new data by Carrara (2022) for government accounts and  by Absell-Junguito (2018) for trade accounts. Money supply growth is based on the  Peláez–Suzigan series in IBGE (1990), and the price indexes come from Lobo et  al. (1971), Buescu (1973), and Catão (1992). Data for the Brazilian population in  1820-1915 are from Mortara (1941), and in 1916–1947, from Ipeadata (2022). 

Ongoing improvements in measurement of money, prices, and population  

The limitations of the Peláez–Suzigan money supply series are well recognized but  it has long been used in the absence of a better alternative. The main shortcomings  reflect the omission of coins (a major component of the money stock in the early  19th century) and of notes and deposits of banking houses and private banks  (important at least from the mid-century through 1864). We are now collaborating  with a research team led by Angelo Carrara (UFOP) and Fernando Cerqueira Lima  (UFRJ) to produce a new money supply series incorporating these missing  components. Results are expected by mid-2026. 

The weaknesses of the Lobo and Buescu price series are also well known: Lobo’s  figures overestimate the inflation rate, while Buescu’s come from distinct non comparable goods samples. Thales Pereira has recently proposed a new price index  extending Catão’s well-established 1870–1900 series back to 1824. Once published,  this new index—combined with the forthcoming money supply series—should  enable a revised reassessment of real GDP growth in the 19th century.

Furthermore, Botelho (2025) has recently proposed revised population estimates  for the first half of the 19th century that differ substantially from Mortara’s (1941)  widely-used figures. Botelho’s estimates imply a much higher population growth  rate in the 19th century than that proposed by Mortara. Given the recency of this  work, further academic evaluation will be needed to determine whether these  revised estimates should be incorporated into the official historical series  maintained by IBGE and Ipeadata. 

Ongoing improvements in GDP measurement 

To contextualize our methodological strategy, it is useful to note that national  accounts methodology recognizes three distinct approaches to measuring GDP:  the expenditure approach, the production approach, and the income approach. In  principle, all three should yield identical results. However, in practice—particularly  for historical periods with limited data—each approach relies on different source  materials and thus serves as an independent check on the others. 

Following Goldsmith’s (1986) pioneering work on Brazilian financial development,  our 19th-century estimates adopt the expenditure approach. This method infers  GDP growth from the evolution of monetary aggregates, government fiscal  accounts, and foreign trade, deflated by price indices. The logic is that increases in  monetary circulation, government revenues and expenditures, and trade volumes  should collectively reflect underlying economic expansion. Our approach could be  improved with the addition of a wage series once new data is made available on  this critical private consumption related variable. 

The production approach was used by Haddad to extend the national accounts to  1900–1947. Applying a similar supply-side method, Guilherme Tombolo is  currently preparing new GDP growth estimates for 1820–1947. His preliminary  results for 1820–1900 point out to a 0.8% trend growth rate, very close to that  reported in our 19th-century paper. An IEPE/Casa das Garças discussion paper  presenting Tombolo’s estimates should be available by mid-2026. 

Apparently, an income-based approach is being developed by Guilherme Lambais  and Nuno Palma in their forthcoming study “How a Nation Was Born: Four  Centuries of Brazilian Economic Growth in Comparative Perspective.” This  inference is based on a recent working paper by the same authors, “African Slavery  and the Reckoning of Brazil,” which has a figure (Figure 2, p. 9), displaying the  behavior of the real wages of unskilled male workers in Brazil from 1574 to 1920.  In the period of our interest, from 1820 to 1900, this variable exhibits substantial volatility but with a clear upward trend. A 27-year moving average, for example,  suggests a compound annual growth rate of 1.3%. This variable is not directly  comparable to the GDP per capita proxies, but is directionally consistent.  Confirmation must await the publication of the Lambais–Palma GDP work, but  this preliminary wage evidence suggests a convergence across methodologies  regarding Brazil’s 19th-century growth performance. 

Conclusion and outlook 

At present, the expenditure, production, and income approaches apparently all  converge in supporting the conclusion advanced in our 2025 paper. In opposition  to what current economic historiography has until now asserted, Brazilian GDP  per capita did not stagnate during the 19th century but instead would have grown at a rate comparable to those of Europe and other Latin American countries. If  confirmed by ongoing research, this finding will have important implications for  understanding Brazil’s long-run development trajectory and its comparative  position within Latin America during the nineteenth century. Together with our  revised estimates for the 1900-1980 period, it suggests more continuity between  the imperial and republican periods than previously recognized. 

This note makes it clear that substantial research efforts are ongoing to refine  Brazil’s long-run GDP and GDP per capita series since 1800. We are in contact  with Jan Luiten van Zanden of the Maddison Project, who has expressed interest  in our work and acknowledged that the Project’s current depiction of Brazil’s  historical development is unconvincing—suggesting that revisions such as ours are  not only welcome but necessary.

References 

Absell, C. D., and A. Tena-Junguito. 2018. “The Reconstruction of Brazil’s Foreign  Trade Series: 1821–1913.” Revista de Historia Económica / Journal of Latin  American Economic History 36 (1): 87–115.  

https://doi.org/10.1017/S0212610917000143. 

Bacha, E. L., G. A. Tombolo, and F. R. Versiani. 2023. “Reestimating Brazil’s  GDP Growth from 1900 to 1980.” Revista Brasileira de Economia 77 (3):  e132023. 

Bacha, E. L., G. A. Tombolo, and F. R. Versiani. 2024. “A Note on Brazil’s  Historical GDP per Capita Growth Rates.” Revista Brasileira de Economia 78 (3):  e182024. 

Bacha, E. L., G. A. Tombolo, and F. R. Versiani. 2025. “Secular Stagnation? A  New View on the Brazilian Economy in the 19th Century.” Revista de Historia  Económica – Journal of Iberian and Latin American Economic History 43 (2):  279–308. 

Bogart, D., L. Chaudhary, and A. Herranz-Loncan. 2024. “The Growth  Contribution of Colonial Indian Railways in Comparative Perspective.” Economic  History Review, 1–26. 

Botelho, M. 2025. “População e imigração no Brasil (1808–1920).” História  Econômica & História de Empresas 28 (1): 165–201. 

Buescu, M. 1973. 300 Anos de Inflação. Rio de Janeiro: APEC. 

Calógeras, J. P. 1960 [1910]. A Política Monetária do Brasil. Rio de Janeiro:  Companhia Editora Nacional. 

Carrara, A. A. 2022. As Finanças do Estado Brasileiro: 1808–1898. Belo Horizonte:  Fino Traço. 

Catão, L. 1992. “A New Wholesale Price Index for Brazil during the Period 1870– 1913.” Revista Brasileira de Economia 46 (4): 519–533. 

Goldsmith, R. W. 1986. Brasil, 1850–1984: Desenvolvimento Financeiro sob Um Século de  Inflação. São Paulo: Harper & Row. 

Haddad, C. L. 1980. “Crescimento Econômico do Brasil, 1900–1976.” In Economia  Brasileira: Uma Visão Histórica, edited by P. Neuhaus, 21–43. Rio de Janeiro: Editora  Campus.

IBGE (Instituto Brasileiro de Geografia e Estatística). 1990. Estatísticas Históricas do  Brasil: Séries Econômicas, Demográficas e Sociais de 1550 a 1988. 2nd ed., rev. and  updated of vol. 3 of Séries estatísticas retrospectivas. Rio de Janeiro: IBGE. 

IPEA/Ipeadata. 2022. Dados Macroeconômicos e Regionais.  

http://ipeadata.gov.br/Default.aspx. 

Lambais, G., and N. Palma. 2025a. African Slavery and the Reckoning of Brazil. Lewis  Lab Working Paper 2023-01. Arthur Lewis Lab for Comparative Development,  University of Manchester. 

Lambais, G., and N. Palma. 2025b. How a Nation Was Born: Four Centuries of  Brazilian Economic Growth in Comparative Perspective. Work in progress. 

Lobo, L., O. Canavarros, Z. Feres, S. Gonçalves, and L. B. Madureira. 1971.  “Evolução dos preços e do padrão de vida no Rio de Janeiro, 1820–1930 – resultados preliminares.” Revista Brasileira de Economia 25 (4): 235–265. 

Maddison Project Database (MPD). 2020.  

https://www.rug.nl/ggdc/historicaldevelopment/maddison/releases/maddison project-database-2020. 

Mortara, G. 1941. “Estudos sobre a utilização do censo demográfico para a  reconstituição das estatísticas do movimento da população do Brasil.” Revista  Brasileira de Estatística 3 (5): 39–89. 

Pereira, T. 2025. Prices in Imperial Brazil (1824–1889). Work in progress. 

Villela, A. A. 2025. História Econômica da Cidade do Rio de Janeiro: da Fundação ao Século  XXI. Rio de Janeiro: FGV Editora.