Brazil’s growth in the 20th century was overestimated, argues new research

The growth of the Brazilian economy from 1900 to 1980 may have been overestimated, according to a study that reviews data currently used as a reference for the period.

The new calculations are presented by economists Edmar Bacha, Guilherme Tombolo and Flávio Versiani in the article “Reestimating Brazil’s GDP growth from 1900 to 1980”, published on the Iepe/Casa das Garças website.

For the period from 1900 to 1947, a study from the late 1970s by economist Claudio Haddad estimated an average growth of 4.4% per year, now revised by the three researchers to 4%.

For the years 1947 to 1980, historical estimates for GDP were carried out by Fundação Getúlio Vargas, which pointed to an annual growth of 7.4%. The calculation by Bacha and his colleagues shows an expansion of 6.2%.

For the entire period 1900-1980, the study lowers the estimate for the annual growth rate of GDP from 5.7% to 4.9%—still above the estimate of 3.2% for the global economy average. in the same period.

“This article argues that currently accepted figures overestimate Brazil’s GDP growth from 1900 to 1980,” economists say.

The main reason for the stronger result in the past was the use of a methodology that privileged in the calculation activities producing goods with higher growth, to the detriment of lower performing service activities. These include government expenditures, financial intermediation and rents.

In some cases, previous calculations assume that these activities grew at the same rate as others that performed better.

Since 1986, GDP has been calculated by the IBGE (Brazilian Institute of Geography and Statistics), with a historical series retroactive to 1980.

Among the consequences of the review is the assessment that the country grew less during the period of industrialization through import substitution in the middle of the last century.

The growth during the so-called “Economic Miracle” (1968-1973) of the Military Dictatorship would also be overestimated, mainly because of a methodological change adopted in 1969 that changed the accounting of services with lower growth.

The authors also state that the new calculations should lead to a revision of the numbers for the 19th century, which is pointed out as a period of decline in Brazilian per capita income compared to global income, a process that would have been reversed from 1900 to 1980, according to data. from the Maddison project — a database created by the British economist Angus Maddison (1926-2010) and continued by the University of Groningen, in the Netherlands.

“We believe that some of these extraordinary structural breaks are statistical illusion,” the researchers say.

This year they will publish an article with new statistical evidence on the Brazilian economy in the 19th century showing that the country’s performance was better than that calculated by the Maddison project.

The article will be presented by Bacha this Thursday (29) at an event at the Academia Brasileira de Letras, within the cycle of conferences on the bicentennial of the Independence of Brazil, in the lecture “What GDP is this? For a reinterpretation of the Brazilian economy in the 19th and 20”.