IBGE: Census lives absolute tragedy, says former president – 01/24/2023 – Market

The publication of preliminary data from the 2022 Population Census exposes the “absolute tragedy” that befell the survey, assesses former president of IBGE (Brazilian Institute of Geography and Statistics) Roberto Olinto.

In an interview with Sheethe advocates carrying out a comprehensive audit of the data to check whether they are valid, whether additional work needs to be done or, in extreme cases, whether it is the case to draw up a new Census.

“Why do you have to audit? Because these data are not reliable. It had all these problems, and a six-month collection is all that the demography demands. It cannot be like that”, he says.

Today an associate researcher at FGV Ibre (Brazilian Institute of Economics of the Getulio Vargas Foundation), Olinto chaired the IBGE between June 2017 and December 2018. Before that, he was director of research at the body and coordinator of the area of ​​national accounts, responsible for calculating the GDP (Gross Domestic Product) of the country.

A specialist with international experience in the area of ​​statistics, he criticizes decisions taken by the management that succeeded him at the IBGE and defends an investigation to hold possible culprits responsible for the failures in carrying out the Census. “When you do a Census and decide not to advertise [para a população colaborar], this goes against a rule. That was malicious, wasn’t it malicious?”, he asks.

The IBGE says in a note that the objections do not proceed. According to the agency, the methodology for the estimate presented was approved by the Census advisory board, made up of economists, demographers and statisticians as representatives of civil society.

According to the IBGE, the result released so far represents an effort “to deliver population data duly updated within the best statistical technique available with greater precision and reliability”.

What are the problems with the Census? I’ll try to make a timeline. When I left the presidency, Susana Cordeiro Guerra took over, coming from the United States, after 20 years outside Brazil and without ever having worked in the area of ​​statistics or management of anything. She began, in an extremely arrogant way, to tinker with the entire project. She reduced the questionnaire, interfering with a project discussed with civil society. She reduced the budget, from R$3.4 billion to R$2.3 billion, without even trying to keep what was planned. She dismissed the director of research and the director of information technology. A project that had already been worked on for five years involved not only technical intervention, but also exonerated very involved and experienced people.

Covid came, then they cut [de novo] the budget, Susana resigns. She had a whole period there that was kind of complicated. At least the IBGE would have two years to organize itself.

The IBGE knew from the start that it was going to have a payroll of 250,000 people. I had to prepare the system. The Census comes out in 2022, comments from census takers begin to come in. Five days of training only. Delay in payment, errors in the amount. You start to see census takers quitting, angry, quitting their jobs.

I feel ashamed of it. The Census is to be taken in two months [a coleta começou em 1º de agosto de 2022], and we are in the middle of January and it has not ended. There are only half of the census takers. The IBGE asking the city of Rio de Janeiro to put a municipal health agent to collect the Census. The guy wasn’t trained, he doesn’t know what he’s doing. Absolute tragedy.

Preliminary results are out. Absolutely inexplicable things. In Rio de Janeiro, the municipalities in the metropolitan area all dropped in population compared to 2010. In Rio Grande do Sul, a huge number of municipalities went to court, because they don’t know if the Census is right or wrong. I think it’s wrong.

Why? They imputed data for almost 20% of the population. Imputing data is saying: I don’t have data for 20% of the population and I’m going to use a process that is normally used for 2%. No Census covers 100% of the population, but you have high coverage. And it has problems with some variables. Income, population, age. It’s not like ‘I don’t have information on half the municipality’. And what are you doing? They are imputing entire questionnaires.

Usually about 2% of Census information is imputed? More or less that. That’s what was done in 2010. Why do you have to audit? because this data [do Censo 2022] are not reliable. There were all these problems, and a six-month collection is all the demography calls for. it can’t be like this [com tanta demora].

To have a reliable Census, you have to audit. It’s so much of a past problem that you have to stop, breathe and say the following: let’s evaluate. An independent commission, obviously with people from the IBGE and outside, to say the following: these data on the Brazilian population and their variables are valid and respectable. Or to say: they are not valid, there are problems and we have to discuss a solution for that. Question: do we throw R$ 2.3 billion in the trash or not?

How would this data analysis work? It’s a long job. First, statistical coherence, comparing with the past and cross-referencing with other information, including returning to the collection in some municipalities to understand what happened. If the Census had been carried out with all the quality standards, the IBGE could say ‘it’s over, that’s the data, period’. But it was made full of mistakes and generating distrust.

I don’t think he’s right. All indications are that it needs to be revised. This audit would have to say what it will do in the future. Will collect in 2025 [ano previsto para a nova contagem populacional]? Are you going to have to expand it? Even at the extreme, you have to think: will you have to do a new Census?

Is it time to do a new Census? This audit would answer that. It has three paths. The audit will say it’s good — I doubt it. The audit will say ‘there are parts that are usable, but we have to correct it like this, like this, like this’. And there’s the third alternative: the Census is complete crap, let’s redo it. You can’t treat the Census as if it were a joke. It is fundamental to the history of Brazil.

We are already three years late. For what mr. are talking about, there are a series of inconsistencies that lead to the real risk of needing a new Census. Wouldn’t it be time to make a decision quickly about this? Then there is a question of government, Budget, and you still don’t have the dimension of what you would need to do the new Census. You could reason as follows: we are in a statistical catastrophe and we are going to put money in the Budget to carry out a new Census. If not, it saves money. If necessary, we do a new Census. But, for that, observe one fact: Minister Simone [Tebet] has not yet chosen an IBGE president.

How should a Census be prepared? A demographic census is based on excellent training of census takers. They have to be experts in what’s on the questionnaire to what people aren’t talking about. How will they deal with a denial or harassment from an informant? They have to know how to deal with it, and it hasn’t been done. And, key, this is fundamental: a lot of propaganda, a lot of clarification of the population to collaborate. Nothing has been done. Nothing! They used Instagram and Facebook.

In the 2010 Census, you had a census taker on the 9 o’clock soap opera. You put a banner on the soccer field. This is the basics of a Census. If you don’t clarify the population, you won’t get an answer. Why was this not done? Why was there no publicity? Why did you operationally delay the collection by six months? It was all disorganized. There are indications that there were wrong management decisions and that perhaps deserve an investigation.

For possible liability? Exactly. An accountability. I think the Court of Auditors will have to evaluate this whole thing. They didn’t pay the census taker, where is that money?

There is a need to evaluate the decisions that were taken by the direction of the IBGE. [É preciso verificar] if this four-year process had correct decisions or decisions that deserve a warning, a punishment. I’m not even talking about a crime, but an administrative process. When you do a Census and decide not to advertise, that goes against a rule. That was malicious, wasn’t it malicious?

The IBGE postponed the publication of the Pnad Contínua, which brings data on employment and income in the country, under the justification of concentrating efforts on the Census. What is the consequence of this? Never had a conjunctural survey been delayed. It went against everything. It breaks up the Demographic Census and, not satisfied, it breaks up the PNAD as well. Why did it have to affect the PNAD?

The big damage is that it starts to affect the credibility of the research. The statistics patterns are very clear. Conjunctural surveys must adhere exactly to their timetable. The problem leaking to a fundamental search, which is the job search, I think is unacceptable. Assume your problem in the Census and protect the other surveys.

Not that there will be a problem with the results of the PNAD, it’s not that, but it’s the issue of contamination and credibility. There is already a feeling of uneasiness in relation to the Census, will it create a uneasiness in relation to the PNAD? Soon it will create a malaise in relation to the [índices de] prices? Boundaries have to be very well defined.

---

X-RAY

Roberto Olinto, 70

An engineer with a master’s and doctorate in the area, he joined the IBGE in 1980. He served as manager of National Accounts from 1998 to 2005, during which time he helped to develop the new system for calculating the quarterly GDP in Brazil. In 2005, he became coordinator of the area. In 2014, he became Director of Research. In June 2017, he assumed the presidency of the IBGE, remaining until the end of 2018. He has already been a member of the group of advisers on national accounts of the UN (United Nations) and a consultant for the Department of Statistics of the IMF (International Monetary Fund). Today, he is an associate researcher at FGV Ibre.