Transcript 0:00 [upbeat music] Next year, the Brazilian Institute of Geography and Statistics will carry out, as it does every 10 years, Brazil's national census. While that may sound straightforward, it is quite the feat. 0:21 [upbeat music] The project will demand 230,000 temporary workers who will have to visit 70 million households spread across Brazil's eight million square kilometers. 0:36 Just for the sake of comparison, that's bigger than Poland, Egypt, Ukraine, France, Spain, Germany, South Africa, the UK, and Italy all combined. 0:49 [upbeat music] However, as is the case with most things in Brazil right now, even the census is becoming a hot button issue. 1:00 Questions regarding the census budget have led government officials to propose a slimmer version of the survey, while opponents say that the move could compromise the data, which is vital for policy making. 1:13 The truth, of course, is much more complex. [upbeat music] My name is Gustavo Ribeiro, editor-in-chief of the Brazilian Report. This is Explaining Brazil. 1:26 [upbeat music] A major controversy is brewing over the 2020 census in the United States as the Donald Trump administration wants to include questions related to citizenship. 1:43 Challengers say the question could lead to millions of people not responding to the census. In Brazil, the problem is the exact opposite. 1:52 Economy Minister Paulo Guedes suggested reducing the number of questions in the 2020 census as a way of cutting public spending. 2:02 Some experts have been highly critical of the initiative, including the former boss of IBGE, Brazil's official statistics agency, a critique which reportedly led to him being sacked. 2:14 [upbeat music] But before we get into who is in the right, we have to understand how important the census is and how difficult it is to prepare one in Brazil. It's absolutely crucial. It's absolutely crucial. 2:30 This is Sergei Soares, a researcher who served as president of the Brazilian Institute of Applied Economic Research. We need to know how many people reside in each municipality. Without that, 2:42 you can't do policy of any kind almost. So it's absolutely crucial. It's, it's just, it's just a huge operation because, not only because of the size, but also because of the diversity, the heterogeneity. 2:54 It's not, um, I don't... It's not like everywhere in Brazil you have a little formal address. Often, people don't even have formal addresses. 3:01 If you live often in rural areas or in, uh, urban peripheries or in shanty towns, you don't have a formal address. I mean, you don't, you don't have a s- a street number with a, with a s- a street name that works and, 3:18 and you can list as a formal address. So the, the, the diversity of Brazil is perhaps as big a, a, a problem as the size of Brazil, you know? 3:27 So it's important to remember that we have a census which happens every 10 years, okay? And we have a household survey, which is a continuous household survey, which happens all the time. 3:40 The census itself is composed of two parts. One is indeed a, a census, which is called the, the, the contagem, or the count, where you, you have very little information about each person. 3:54 You basically, in each household, you have how many people live there, their sex, their age, and little else. And then you have something called the sample, which is usually something like 8% of the total population, 4:07 which answers a longer questionnaire, okay? Which has questions on schooling, income, various other things. Now, why is this important? 4:18 Well, uh, if it were not for the fact that we need geographically disaggregated data, nobody would care. You would just have the, the, the, the count and you wouldn't even need the, uh, the 8% sample questionnaire. 4:30 But the fact is that people need information at a municipal level or even at a district level, and for that you need to have this, um, the 8% sample of the census. 4:42 So basically, the household service will give you a snapshot of the country on a national level, and the question is, which kind of information do you need at a smaller level, like from a municipality or a district? 4:56 So for example, there's a plethora of income questions in the household survey, and this has to be that way because inclum- income is a complicated concept, and it's difficult to know, uh, um, exactly what your income is, particularly when you have multiple jobs and multiple sources of income. 5:13 Now, everybody will agree that all these questions should not be in the census. The question is, how many of them should be there? Uh, only a single question, what is your in- total income? 5:23 Or should we ask, what is your labor income? And just say, from all jobs, and whether it's monetary and non-monetary. Should we divide it into monetary and non-monetary income? And what about non-labor income? 5:35 Should we just as- aggregate everything together and say, "What is your non-labor income from pensions, Bolsa Familia, BPC, and everything else?" Or should we ask them separately? 5:43 That's, these are the kind of issues that people are grappling with. So a minimalist, a person who wants a, a, a short questionnaire would say, "Look, let's, let's do two questions. One, what's your labor income? 5:54 What's... And two, what's your non-labor income?" And that's that. I'm happy with that. The, uh, person who is not so minimalist would say, "No, maybe we would like a, um, a..." More than one labor income question. 6:05 We would want l- your labor income from your main job and your labor income from your other jobs. 6:11 And we would like to divide up the, uh, non-labor income question into Bolsa Familia, into pensions, into capitals income, and so on. 6:20 So it's how many questions you should have on each particular issue that is the, that is the, the, the, the d- the debate centers around that. But even that is not a black-and-white issue. 6:33 There are other ways for the country to produce reliable data in a small geographic area besides taking a census. The issue here is how the government will take a new direction. 6:45 Administrative records in Brazil have been getting better and better and better. And the direction that IBGE is going in is to try to merge census data with administrative records. In my opinion, this is, th- they... 7:00 This is the entirely the correct direction. The correct direction is to have a smaller and smaller census, uh, with better administrative records and more questions in the household survey. 7:12 The, the big issue is the speed at which we're going to do it, and I think relatively few people disagree that the dir- that, in, a- as to the direction that we're going in. 7:25 The big disagreement is the speed at which we're doing this. Some people think that we're going too fast, and some people think that, you know, we should really scrap the 8% questionnaire or make it very, very small. 7:38 That's, that's... The issue is, is basically there. It's not whether we should, um, whether it should be a huge questionnaire. Everybody agrees that it shouldn't. 7:47 But, uh, how fast we should be reducing this questionnaire. What a lot of people argue is that, "Look, this is the first time that we're trying to do the link between the census, uh, data and the administrative records. 8:02 So we're not sure this is gonna work out. 8:04 So let's keep the big ques- questionnaire for this census, and if this link works well, then for the next census in 10 years' time, we can s- think about reducing the questionnaire." 8:15 That, that's what a lot of people are saying. Um, so, you know, it's, it's, it's still a debate, and, uh, I don't think, um, there is necessarily a wrong or a right side. 8:26 Sergey Soares is not losing any sleep over the 2020 census. He does believe things will get done, and he rejects the idea that Brazil is risking a statistical blackout, as some experts have suggested. 8:40 It's a complete and total overstatement. I mean, people are totally exaggerating. 8:46 You may discuss whether you want a larger or a smaller census questionnaire, but if you're going to have the census, and nobody's suggesting not having the census, you're not gonna have a statistical blackout. 8:57 You may, you may be missing detailed information on one or another issue at the municipal level that you had before. But please, statistical blackout, that's, you know, this is, uh, hysterical. That's a hysterical term. 9:09 We have excellent statistical production from, um, other institutes that are not IBGE. So we have an excellent, uh, statistical, um, institute r- related to education, INAPI. It's really good, top-notch. 9:24 Absolutely top-notch. Maybe only Chile has anything as good as o- ours in Latin America. We have excellent labor market data produced by the Ministry of Labor. We have... In terms of statistical, in terms of data, you... 9:37 It's difficult to find a developing country which has as good and as much data and as high quality data as Brazil. 9:45 So while it's regrettable that, you know, we're losing funding, and this is m- mostly due to our terrible, terrible, um, macroeconomic management over the last few years. But, uh, it's not like we're having... 10:00 It's not like we have a, a bad statistical system at all. We have an excellent one. And, uh, we don't wanna lose it, so it's important that we don't lose more funding, but, um, it's not... 10:12 There's no, no, no chance of a statistical blackout or anything like this. But even if Brazil won't reach the point of a statistical blackout, things aren't particularly rosy at the IBGE, the official statistics agency. 10:26 [gentle music] That's after the break. 10:29 [drumbeats] Ponto Futuro Rio is the B2B conference about technology for decision-makers. 10:45 On June the 5th in Rio de Janeiro, 25 speakers from Brazil and abroad will get together to discuss how technology transforms businesses and society. 10:56 The theme for this year's edition, the third already, is augmented humanity or amplifying human abilities through technology and how this transformation shapes new relations and calls for adaptation. 11:10 For more information, go to futuro, that's future with an O at the end, dash rio.com, futuro-rio.com. 11:22 [drumbeats] [gentle music] This is a recurring theme in Brazil, but like pretty much every single other state agency, IBGE is battling a lack of funds. 11:38 Since 2014, the agency has closed 16 regional bureaus as civil servants have retired without anyone to replace them. 11:48 Of IBGE's 583 regional offices across the country, 232 operate with only two people. Another 61 have only one staff member and could be shut down within the next few years. 12:05 The cuts in the, the regions take away their eyes, uh, their local eyes Marcelo Soares is a data journalist and regular contributor to The Brazilian Report. 12:16 One thing is to, to do a, a demographic research if you, if you have a base there, if you know the population, if you are in touch with them. 12:23 And the other thing is to do that kind of research visiting the places every once in, in a while. 12:30 A- and that's a problem if you, if you lose the, if you lose your, your feet on the ground IBGE has lost 40% of its civil servants in eight years, and more losses are to come. 12:41 There's a, a huge share of the workers in IBGE who are about to retire. They can retire, uh, if they want. And, and many more, uh, will soon be able to retire if they want, like, a one-third or so. 12:58 Not having money will... is a problem for any activity, any government activity. I, I don't know of any, any government activity which if you, if you cut its funding in 40%, it will have zero impact. Okay? 13:13 But IBGE is a spectacular statistical agency. It's very good. It still is very good. You are hard-pressed in Latin America to find another country which has as, 13:28 as a good a statistical production as Brazil, maybe Mexico, in terms of, of, of what is produced by the, uh, by the statistics institute. 13:37 Even if Brazil's IBGE remains a highly respectable agency, some cracks are starting to show. 13:44 For the 2010 census, for instance, recruiting began two years in advance, and that's because it takes time to train a few hundreds of thousands of census workers. 13:55 For the, the 2010 census, in 2007, they were already training the, the, the people who, who, uh, take the census. Now, they are only now, uh, uh, trying to, to hire people. 14:11 So it's, uh, it's a, it will be a, a difficult situation. Brazil has a checkered story with censuses. Uh, uh, we, we missed some. We, we got some late. Uh, in t- in 1910, we didn't have a census. 14:27 Then in, in the, the, in 1930 also we didn't. Uh, in 1990, uh, we delayed for one year the census. It's, uh... In 2010, 14:41 uh, I was not, uh, I was not interviewed by the census, and I was mad about it. Then, uh, I called IBGE and they said, "Well, but you were. You were. We have your records here." 14:54 Then I found out, uh, they, they sent, uh, the, an interviewer there, and they found out the, the interviewer had, had talked to the doorman in the building, uh, to cut corners. Uh, some information was wrong. 15:09 Statistically, it would not make, make my home too, too different from what it was, but it was not what, uh, it was not me. 15:17 With, uh, with 200,000 people on the streets, uh, uh, making interviews, it's difficult to, to keep track of what, uh, each one of them is doing, what, uh, how much, how much they are cutting corners and so on But political interference in the statistics agency can be as damaging as a lack of funds. 15:38 IBGE has had to endure a lot of criticism from President Jair Bolsonaro, who seems not to agree with how the agency calculates unemployment data, which by the way, follows methods recognized worldwide. 15:52 Since last year, he has suggested changing how the data is calculated. 15:56 But according to Sergei Soares, we must be prudent with how we look at this, as he doesn't see government interference in data production, at least not in the way we saw in Argentina in previous years. 16:07 In 2006, then Argentine President Néstor Kirchner intervened in the country's statistics firm, leading to a lack of trust by the international community. 16:19 After Kirchner's move, respected publications such as The Economist stopped using Argentina's official data altogether. Not at, not, not at the level of Argentina. 16:30 I mean, in Argentina it was a blatant, you know, blatant political, um, sacking of a, of a, of a sta- gutting of a statistical institute. It wasn't due to saving money or anything like that. It's just, you know... It was 16:46 a government trying to put forth, you know, you know, uh, fake facts in order to, to, to, um, further its political, its political agenda. It was, you know, a, a despicable act on, on, on that government's part. 17:01 Um, of course, in, a- anywhere you go, there is always a tension between the government that wants to show good news and the statistical institute that wants to show the right news. So this is the, this exists in Brazil. 17:15 This exists everywhere. Um, but one thing that has to be said is that, uh, anything close to what happened in Argentina has n- not happened in Brazil, at least since we've returned to democracy. 17:31 [instrumental music] In a country as huge as Brazil, data is an indispensable tool in forming public policies. 17:47 Without vast and detailed studies, we wouldn't be able to pinpoint the pressing needs facing the nation and come up with the solutions which can be justified, tested, and measured. 18:00 As it stands, the census appears to be on track, even though it is likely to undergo some changes to cut costs. However, with the turbulence that comes with the current government, we'll just have to wait and see. 18:20 This podcast was written and prepared by me, Gustavo Ribeiro. Maria Martha Bruno produced this show, and Ewan Marshall edits the final script. 18:33 If you like this podcast, rate us on any platform you may be listening to Explaining Brazil. It takes only a second, but it is really important for us. 18:43 The best way to support this podcast is by subscribing to The Brazilian Report, the journalistic engine of this show. Every day we have new content about Brazilian politics, finance, and society. 18:54 We've also got exclusive newsletter services if you want to be briefed about what's going on in Brazil before starting your day. 19:01 Subscribe now for a free trial and enjoy all of our content for seven days, and it is really free. You don't have to submit any credit card information whatsoever. You can also follow us on Facebook and Twitter. 19:16 Our handle is @brazilianreport. That's all for now. See you next week. [outro music]