You are here

Published in: 08/06/2019

Is Brazil moving towards a statistical blackout?

  • Facebook
André Antunes - EPSJV/Fiocruz

The questionnaire that will be applied in the 2020 Demographic Census was divulged last week, amidst a debate between the technical body and the direction of the Brazilian Institute of Geography and Statistics (Instituto Brasileiro de Geografia e Estatística, IBGE, in Portuguese), which has been going on since the beginning of the year. Appointed by Minister of the Economy Paulo Guedes, the new president of the Institute, Susana Guerra, who took office at the end of February, was assigned the task of reducing by 25% the Census budget, which was initially of R$ 3.2 billion. The mission was then assigned to economist Ricardo Paes de Barros, from the Institute of Education and Research (Instituto de Ensino e Pesquisa, Insper, in Portuguese), and it displeased the technical body of the IBGE. Ever since the cuts were announced, the Association of IBGE Employees (Associação dos Trabalhadores do IBGE, Assibge-SN, in Portuguese) had been warning that a decrease in the questionnaire could cause a rupture in the historical series of data about the Brazilian population.

In fact, the questionnaire released by IBGE has cuts in the number of questions when compared to the Census performed in 2010. Both the basic questionnaire, which is applied to 90% of the more than 70 million Brazilian households, and the sample questionnaire—which is more comprehensive, and is applied to 10% of the households in every municipality—suffered a reduction. The 34 questions of the basic questionnaire were reduced to 26. As for the sample questionnaire, which in 2010 had 112 questions, will only have 76 next year.

During the public meeting at the Commission to Fight for the Rights of the Elderly (Comissão de Defesa dos Direitos da Pessoa Idosa, in Portuguese) in the Brazilian Congress on July 4th, Susana Guerra argued that the cuts in the questionnaire are not related to the reduction in the Census budget. According to her, the changes were made to reduce the time spent by the Census researchers during their visits to households. The estimate is that the duration of the interview for the basic questionnaire will fall from 7 to 4 minutes, and the interview for the sample questionnaire will fall from 24 to 19 minutes. ‘With this, we will be able to decrease the number of Census interviewers, and to pay them better wages,’ stated Susana.

The presidency of IBGE has been defending that the items cut off from the Census comprise information that can be collected through sampling researches, such as the National Continuous Research by Household Sample (Pesquisa Nacional por Amostra de Domicílio, Pnad Contínua, in Portuguese), and administrative records, such as those of the Federal Police.

Among the questions cut are those regarding topics such as housing, education, work, income, and international emigration, among other questions.

The Assibge-SN criticized the changes, which, according to the association, were made without the participation of the technical body involved in the development of the questionnaire, which includes IBGE public servants, researchers from universities and research centers, members of social movements, and managers. In a statement, the association of employees of the institute warned of the risk of a ‘statistical blackout’ in the demographic characterization of the country, with the rupture of the historical series of data and information that are used to grasp the socioeconomic reality of Brazil.

‘The development and evaluation of public policies at a municipal level will suffer a hard blow, with consequences for the Brazilian population and the Public Administration for the following ten years. A lower ability to make socioeconomic diagnoses, lower quality in the population estimates, and an irreparable loss in the design of public programs,’ warned the statement.

What has been cut

One of the main topics criticized by the Assibge was the exclusion of questions about international emigration from the basic and the sample questionnaires. Moreover, according to Dione de Oliveira, from Assigbe’s National Executive Office, the 2020 Census basic questionnaire did not incorporate questions about internal migration, as was the proposal of the institute’s technicians involved in the development of the questionnaire, which would be an expansion of the basic questionnaire applied in 2010. Thus, only the sample questionnaire will contain questions regarding this topic, just as it happened on the Census performed a decade ago. ‘Throughout the decade, we made studies that indicate that, to make the proper mensuration of this movement that has an impact on the population estimates, it would be highly positive if migration would have been included in the basic questionnaire, which did not happen,’ regrets Dione.

According to the association, this exclusion can generate distortions in the projections of the population who resides in the country, which are carried out in the periods in-between Censuses, and in which one of the variables included is migration. Population projections are used as the basis for the calculation of federal fund transfers such as the Municipality Participation Fund (Fundo de Participação dos Municípios, FPM, in Portuguese) and the State Participation Fund (Fundo de Participação dos Estados, FPE, in Portuguese), and they are among the main focuses of judicialization regarding the Census, according to Dione. ‘It often happens that a municipality suffers a change in its population level and ends up losing revenue from the fund. When this happens, they go to court against the IBGE. And the tendency is that this will worsen,’ she evaluates. One fifth of the FPM and of the FPE are set aside for the Fund for the Maintenance and Development of Elementary Education and Esteem for the Education Professionals (Fundo de Manutenção e Desenvolvimento da Educação Básica e Valorização dos Profissionais da Educação, Fundeb, in Portuguese), which is the main source of funds for this level of education in Brazil.

The topics that measure the impact of public policies on the reduction of poverty and inequality have also been altered. The income topic on the basic questionnaire will only be applied to the so-called “head” of the household. In 2010, every household resident aged over 10 years informed their income.

In the group of questions regarding other sources of income in the sample questionnaire, which was divided into 4 questions in the 2010 Census, will now be only 1 question in the 2020 Census. Thus, it will no longer be possible to identify the source of this income: whether it comes from a pension, from programs such as the Family Grant (Bolsa-Família, in Portuguese), from Social Security benefits, such as the unemployment benefit or the Continuous Welfare Benefit (Benefício de Prestação Continuada, BPC, in Portuguese), or from financial investments or from charging rent. ‘Intramunicipal inequality studies will become more fragile without a proper measurement of the income of the population,’ warned the statement issued by the Assibge.

Additionally, the sample questionnaire will no longer assess the existence of goods such as television sets, telephones, refrigerators, motorcycles and cars in the residences; only a few questions about the existence of a washing machine and internet access remain. “Collecting these data at the census sector level and in the rural countryside is fundamental for the identification of pockets of social inequity within the municipalities, especially at this moment, in which there are indications that a growing number of families are once again using firewood to prepare food,’ highlights the statement.

Another loss indicated by the Assibge is regarding the calculation of the housing deficit. This is because the basic questionnaire will no longer identify the households in which the residents pay rent. As for the sample questionnaire, it will no longer ask the price of the rent, as it did in 2010. The excessive burden with rent (that is, families with a maximum income of 3 minimum wages and expenses higher than 30% of the family income) currently corresponds to more than half of the housing deficit in the country, according to the latest data released by João Pinheiro Foundation (Fundação João Pinheiro, in Portuguese), which refer to 2018. 

It will also no longer be possible to identify, in the case of the residents that are enrolled in daycare centers, schools or universities, if those institutions are public or private. ‘This topic is paramount to characterize the students and the school system in which they are enrolled; it also identifies places (municipalities, neighborhoods) where there is a need to expand the public school system,’ points out the Assibge statement.

Losses regarding social security

Researchers interviewed by the Joaquim Venâncio Polytechnic School of Health (Escola Politécnica de Saúde Joaquim Venâncio, EPSJV/Fiocruz, in Portuguese) internet portal say that there is a risk that these changes represent a setback for the areas of social security and education. Martha Sharapin, a professor and researcher for the EPSJV/Fiocruz, highlights that, even though there are few specific questions regarding health in the Census, they are fundamental in the calculation of indicators such as vaccination coverage and the rates of incidence and prevalence of many injuries. ‘The Census data are also used in the payment of many procedures made through the Brazilian Unified Health System (Sistema Único de Saúde, SUS, in Portuguese), and they also enable us to know the living conditions of the population, and to identify the most vulnerable groups, who need more specific equality policies in order to have their rights guaranteed,’ she explains. In her opinion, the cuts in the questionnaire that have been announced will hinder the calculation of the population projections, as well as the development of a social and economic profile of the Brazilian population.

Demographer Dalia Romero, a researcher at the Institute of Scientific and Technological Communication and Information in Health (Instituto de Comunicação e Informação Científica e Tecnológica, Icict/Fiocruz, in Portuguese), agrees. According to her, the cuts in the topics regarding migration and emigration may generate distortions in the population projections, and cause a cascade effect in the many surveys in the field of health that are based on the research. ‘The IBGE Census is the matrix to calculate the samples in every health survey, be it in the context of food security, of the National Health Policy, etc. To do so, one must have a great deal of clarity regarding the population numbers in the municipalities. If you cut a piece of information that is fundamental for the population estimate, we hypothesize that this will generate a great deal more of mistakes in the calculation,’ she remarks. 

The researcher also criticizes the fact that only the income of the head of the household will be collected in the 2020 Census. ‘In the field of health, we need to analyze poverty indicators, and we need data on the income of every household resident. An elderly person may have the highest income and not be the head of the household, because this role is played by their son-in-law, for example. The elderly person may have moved in with you, you will take care of them, and they will help you financially, but you are the head of the household. This is an example of difficulty in defining who is the person responsible for the household,’ ponders Dalia.

According to her, the example of the elderly is important to analyze the impact of another change in the Census: the elimination from the sample questionnaire of questions that enabled the identification of whether the residents of a certain household receive benefits such as the BPC and the Family Grant.

In the 2020 Census questionnaire, which was released last week, many benefits regarding social programs were grouped into a single topic, which also contains items such as retirement, rent, and income from financial investments. ‘Ever since the second decade of the 21st century, the coverage of the BPC increased greatly. Some works show that the households in which there are elderly people who receive the BPC are less poor. And how do we know that? Through the Census. It is our only source of information. This is something that we will no longer be able to know,’ regrets Dalia.

The justification given by the directors of the IBGE, that the information that has been cut from the Census can be obtained through other sources, does not convince the Fiocruz researcher. ‘What they argue is that what has been cut can be obtained through the surveys of the National Household Sample Survey (Pesquisa Nacional por Amostra de Domicílio, PNAD, in Portuguese). But none of those surveys provide information about the municipalities, which is where lies most of the country’s poverty. This information never appears in a household survey,’ she states.

Moreover, the researcher highlights that, from the point of view of the researches that deal with Census data, it is important to group into the same database as much information as possible concerning the population. ‘Once the information is divided among different surveys, it is impossible to delineate the profile of the population of a territory, because these are different sources of information that often do not communicate with each other. We are on the verge of having an uninformed country, and the lack of information does not contribute to Brazilian public health,’ she states.

Losses regarding education

From the point of view of educational policies, the cuts are also cause for concern. According to the coordinator of the Educational Data Laboratory of the Federal University of Paraná (Universidade Federal do Paraná, UFPR, in Portuguese), Gabriela Schneider, the Census is important, for instance, to calculate data such as the average rate of schooling in years of the population, and the amount of people out of school, which the Census still allows to compare with the income. ‘It is because of the Census that we can get an outline of the access and quality, so that we can contribute to develop and evaluate the educational policy, if it reaches or not, and to what extent, the population in the municipalities,’ explains the researcher.

The monitoring of the goals of the National Education Plan (Plano Nacional de Educação, PNE, in Portuguese), which will end in 2024, depends on the precision of Census data, according to Gabriela. As an example, she mentions pre-school education, which would represent a bottleneck. ‘The PNE has the goal of ensuring that 50% of the children aged between 0 and 3 years are enrolled in daycare centers until 2024. And how can one do that without even knowing how many children are there in a municipality? It is impossible to develop public policies without using these data. And only the IBGE Census enables us to reach the municipal level,’ states Gabriela.

In her opinion, the cut in topics regarding income is also a problem for education. ‘The researchers have been performing many comparisons regarding income and access to education, and what we perceive is that the poorest population is also the one with the lowest access to education. In the previous Census, the data regarding income and benefits from social programs were more detailed. Without these data, it will be impossible to make more complex analyses of this relationship,’ she warns.

The UFPR researcher considers that the rupture in the historical series of the Census is a severe loss for the scientists, as well as for the managers who work with Census data. ‘It is a problem for those who perform historical studies that all of a sudden a question will no longer be asked. This is one of the main foundations on which to develop public policies in Brazil. Everyone says that we must develop efficient policies, but to do that we need to have a notion of how is the population, what are its characteristics, what is its reality. Is the financial impact high enough to justify these cuts?,’ Gabriela questions. 


The idea is that the questionnaire released last week will be applied at the Experimental Census, which is performed in the city of Poços de Caldas (in the state of Minas Gerais) between September and October. The 2020 Census is scheduled to begin in September of next year, and it should take 3 months until its conclusion. But there still is some uncertainty regarding that.

According to Dione de Oliveira, the community of IGBE researchers will remain mobilized to try to ensure that the 2020 Census happens as it was planned by the technical body of the Institute. ‘One cannot ignore a project that dates back at least to 2016 to do something completely new. One cannot alter methodologies this close to the Census, which by definition has a long maturation period. This is irresponsible,’ she criticizes.

Still according to the Assibge-SN representative, one of the stipulations of the public hearing held on July 4th to debate the cuts in the questionnaire was that the Commission to Fight for the Elderly (Comissão em Defesa da Pessoa Idosa) would present an amendment during the vote for the Budgetary Guidelines Law (Lei de Diretrizes Orçamentárias, LDO, in Portuguese) in order to ensure the budget initially provided for the 2020 Census, which was around R$ 3.2 billion. 

Federal Congressman Marcelo Freixo (who is affiliated to the PSOL-RJ party) informed after the hearing that he would file a complaint with the Federal Public Prosecutor’s Office (Ministério Público Federal, MPF, in Portuguese) to cancel the budget cuts. In May, the MPF sent a notification to the presidency of the IBGE requesting information on the process of development of the questionnaire.

According to Dione, there is also uncertainty regarding the content of research that is performed in parallel with the Census, which investigates the living conditions surrounding the households, such as street pavement, the number of trees, accessibility, urban mobility, lighting, among other aspects. ‘The research regarding the surroundings is a set of features that are important for city planning, for urban planning. It had been canceled, but, due to pressure from our movement and the technicians, it was reinstated. The thing is that we are not sure about what data will be effectively collected,’ points out Dione.

The EPSJV/Fiocruz internet portal sent questions made by the researchers to the public relations office of the IBGE, but it did not get any response back by the conclusion of this news article.