The toll of the pandemic on early childhood education in São Paulo

Entry type: Single project

Country/area: Brazil

Publishing organisation: TV Globo and G1 (TV Globo’s news website) publish this work as a three-part series. First, the story aired on television news program SP1. Then, the online version was published and embedded the original video alongside the texts, photos and graphics.

Organisation size: Big

Publication date: 2022-11-16

Language: Portuguese

Authors: Ana Carolina Moreno (producer and online reporter)
Rafael Ihara, Fernanda Elnour, Henrique Silva (television reporters)
Marcelle Sansão, Amanda Ferreira (text editors)
Hélio de Olivera, David Faria, Edgar Rocha (cinematographers)
Kayan Albertin, Elcio Horiuchi, Juan Silva e Wil Nogueira (arts and graphics)
Fábio Rodrigues, Glauber Sousa (image editing)


Ana Carolina Moreno is a journalist with over 15 years experience covering public policy regarding access to basic human rights (mostly Education, Health/Covid-19 and Refuge). She graduated in Journalism and Portuguese and French from University of São Paulo (Brazil) and holds a Master in Journalism Editing from Universidade da Coruña (Spain). After specializing in Education since 2011, when she joined Grupo Globo, in 2017 she began using data science in her coverage, mainly R and SQL. Since 2019 she is a member of R-Ladies São Paulo and teaches basic programming to gender minority people interested in learning data journalism.

Project description:

This multiplatform series combined data science and on the road investigation to measure the impact that closing schools during quarentine had on children who were still not old enough for remote schooling and needed constant mediation from their parents to join a virtual classroom. Time and financial difficulties led many families to unenroll their children, causing a record number of private preschools to close and depriing children of social activities that are crucial for their development. We focused on three main actors: schools, families and, finally, teachers who found ways to bypass this problem and guarantee children’s right to education.

Impact reached:

This multimedia (TV and online) project was selected for the Dart Center 2022 Early Childhood Latin America Fellowship, which focused on inequality and Covid-19.

Over the course of 2021 and 2022, we had already covered the impact of the pandemic on other levels of education, mostly on school evasion of teenagers, the obstacles for young and adult education (a relevant phenomenon in Brazil), and on 6 and 7 year-old children who had migrated from the private sector to public schools and provoked an expansion on the public netwoork to absorb this new demand.

Common sense was clear: education was suffering losses both in access and in quality. Commom sense also said that, during an emergency, the younger children didn’t need to be prioritized because they would have “more time” to catch up.

But we wanted to show just how big a loss we were facing, both on the structure of the system (the schools that closed), on the individual level (how a child’s development would be halted or delayed, especially the ones from low income families) and on society in general, because science has already showed that early childhood education has the potential to mitigate the social inequality that only increases with age.

We also wanted to hold the government accountable for their constitutional role guaranteeing these children rights to education. Our data showed that, in the State of São Paulo, especially in the Metropolitan Region of the Capital, private schools, more so daycare centers, are much more predominant, which can be explained by the fact that these regions concentrate more women who are economically active. Many of them have to pay for what is a public right because they don’t find public options, especially options that will take their children for the entire day.

Techniques/technologies used:

Data analysis and mapping was done using R. The graphics fot the television version, used the Adobe packages. The online version had dataviz from Flourish.

The first part of the series envolved applying data science to cross the different editions of the official School Census, published every year by the Brazilian federal government. The consolidated version of the Census simply counts the number of schools by city. state and country, but our analysis envolved finding the daycare and preschools in the São Paulo State that were open and running and had at least one students in 2020, but in 2021 listed their status as inactive/extinct, without enrollments. We then applied geolocation techniques to map these schools and used OSINT to track previous school owners of to begin shapping the story.

Other analyses were made to better describe the private to public ratio of daycare and preschools in São Paulo and the rest of Brazil, to go beyond common sense and better explain the phenomenon.

For the second part, the focused was on the parents. We used quarterly data from Pnad, a questionnaire that asked a sample of household if they had children from the age of 5 that attended preschool. With help from specialists we applied techniques to weigh the responses and achieve an equivalente percentage of the population. The data showed that, from July 2020 until March 2022 the percentage of cchildren not attending preschool was twice to three times as high as before the pandemic, which pointed to a strong backset in a trend of access to education and risked the goals established by the National Education Plan.

The third part used concepts of Solutions Journalism to tell successful stories from teachers by explaining step by step their projects and results in ways that could be replicated.

Context about the project:

Access to data was the main obstacle for this project. In 2021, the Brazilian government, under Bolsonaro, decided that the Education disaggregated databases were not in accordance to the new personal data protection legislation. However, instead of adapting the databases, they simply stopped publishing them, and removed over ten years worth of historical data from their website.

Nowadays, the only database available is the one where each line is a school. They have added consolidated number, such as the total of enrollments and some details like how many enrollments are from each race. But there are no more databases where each line was a student, but anonymized, or each line was a teacher, also anonymized. It is impossible now, for example, to discover the average age of students by group of race, or the percentage of black teaches according to their education.

We tried to request information using the Information Access Law, Brazil’s equivalente to the Freedom of Information Act, to ask the government to send us the number of students who were enrolled in each category of school (private or public) in 2020, and where they were in 2021 (in the same category, in a new category, or not at all enrolled). But our request was denied on all instances, in a process that took a few months.

The alternative they gave the public was to apply to use the “Safe Room”, a service where a researcher can be allowed in the physical data storage to write SQL queries and search and cross the non-anonymized databases. However, this is a service created with academic researchers in mind, which require a scientific project to submit a request, and several months waiting for the authorization. The query itself would involve travelling to Brasília and most likely spending weeks in the “Safe Room” trying to collect the data, then waiting for a review of the results before actually receiving them to begin the analysis. Paper and pen are forbidden, and there is no internet connection, so the queries need to be memorized. In conclusion, this alternative is impractical for the journalism’s timing and needs.

What can other journalists learn from this project?

Here are five lessons:

1- Data journalism can be useful in all areas of coverage, not just Politics, Elections or Police. These areas might get more audience or a more visible impact, but using data to tell other stories can be just as helpful to sensitize the population on subjects that get almost no space in the news.

2- Education journalism in Brazil, for example, has been an entry door to data journalism because, over the last 15 years, the federal government had done a remarkable job in making very rich educational databases public and accessible.

3- The transparency backlash that affected many areas, including Education, need to always be on the journalist’s mind when deciding what to write about. The more people requesting and pressuring for more transparency, the better our odds to reverse it.

4- Different news formats impose different data journalism challenges. Television is a format with specific limitations. Time is the first one: whereas on a website or on print a reader can spend unlimited time reading a graphic, on TV the graphic is on the screen for mere seconds, so they need to be as simple and straight-forward as possible to be understood by the public. The audience is also different, especially on open channels, where the news is consumed by a majority of lower income households. Nevertheless, this audience also deserves to be informed, so data journalism needs to be especially aware of how to communicate effectively with them. In this project, we show the same graphics, but designed in different ways because they were conveyed in two different formats, TV and online.

5- If I may leave a suggestion: Education is not on the list of Tags on the Sigma Awards submission form. Maybe it should be, just like Agriculture, Health and Culture.

Project links: