3. Learning and its outcomes

This chapter looks at the outcomes of Brazil’s education system to assess the extent to which young Brazilians develop the skills and knowledge to progress successfully into higher levels of education and succeed in life and work. This analysis draws on evidence from the OECD Programme for International Student Assessment (PISA) on the knowledge and skills of 15-year-olds in Brazil, and explores how they compare with peers across the world and changes over time. This evidence is augmented by national sources, which provide fine-grained evidence on student and school performance across the country. The chapter also looks at the implications of learning outcomes and educational attainment for students’ later employment and earnings.

Ensuring that everyone takes part in education — the topic of the previous chapter — is only half the battle. Students also need to profit from the experience, acquiring knowledge and skills of value both in the labour market and in other aspects of their lives. However, according to a recent World Bank report, too frequently participation fails to translate into learning. This report reveals that many emerging countries – including Brazil – are facing a challenge whereby, despite increased participation in schooling, many young people do not acquire basic numeracy and literacy skills. Moreover, as part of this pattern, learning outcomes tend to be highly unequal, with the most disadvantaged students being most likely to leave school without gaining basic skills. Schooling in the context of large variations in access and quality magnifies rather than reduces initial learning inequalities and therefore also amplifies socio-economic divides (World Bank, 2018[1]). This hampers both the lives of individuals and the efforts of governments to reduce poverty and spur growth (Hanushek and Woessmann, 2015[2]).

While Brazil has made good progress in universalising access to education in recent decades (documented in Chapter 2), a weakening economy combined with the COVID-19 crisis means that the necessary next step, that of ensuring access to quality education for all students, is particularly challenging. A successful outcome will depend on many factors, including the flow of financial and material resources to schools and families which will be discussed in Chapter 4, high-quality schools and teachers examined in Chapter 5 and the effective involvement of parents as described in Chapter 6.

While the Brazilian government has announced it will be taking part in the next cycles of the Progress in International Reading Literacy Study (PIRLS) and the Trends in International Mathematics and Science Study (TIMSS)1, these assessments will not be carried out until 2022 and 2023. Other regional assessments Brazil takes part in have not taken place since 2013. It follows that international comparisons of learning outcomes are, for the present, heavily dependent on PISA results. This is a triennial survey of 15-year-old students, focusing on proficiency in reading, mathematics and science. Brazil has taken part in all PISA cycles, beginning in 2000. PISA has been influential in Brazil (INEP, 2019[3]), and has, for example, been used to benchmark national targets in the National Education Plan (Plano Nacional de Educação, PNE) (see below and Chapter 1).

In PISA 2018, 10 691 students in 638 schools completed the assessment in Brazil. The student sample represents only two-thirds (65%) of Brazil’s 15-year-olds2 (OECD, 2019[4]). The other one-third are not represented, mainly because they were not in school at the time when PISA 2018 was carried out. The coverage rate is much lower than in OECD countries, where only 12% of the 15-year-old cohort are not covered (see Figure 3.1). However, it has improved in recent PISA cycles – it was only around 55% in 2003 (OECD, 2019[4]) – as a result of growing rates of educational participation in Brazil (see Chapter 2). The coverage rate has important implications for the analysis of PISA results. First, international comparisons can often conceal varying levels of coverage across PISA-participating economies. More importantly, increases in coverage make it more difficult to interpret how mean scores in PISA have changed over time. Expanding participation in schooling often implies that a larger proportion of disadvantaged low-performing students come to be included in PISA samples (Avvisati, 2017[5]). This will be discussed in detail below.

Brazil was one of the first emerging economies to introduce standardised student assessments. Since the early 1990s, Brazil’s Basic Education System (Sistema de Avaliação da Educação Básica, SAEB) has been used to assess and monitor the quality of Brazil’s basic education (Bruns, Evans and Luque, 2012[7]), and to inform policymaking (Table 3.1). It is widely accepted as a reference point for researchers and policymakers across the country. However, some critics have pointed out that most assessment items fail to capture higher-order skills (Mol, 2019[8]). The SAEB has undergone regular adaptations over the years and, more recently, given the need to align the new assessments to the National Common Curricular Base (Base Nacional Comum Curricular, BNCC). Extensive reforms of the SAEB are being discussed seeking to adapt the instrument to better support teaching and learning3. This includes, for example, gradually increasing the coverage and scope of the assessments, and piloting a digital application.

In higher education, Brazil uses the National Examination of Student Performance (Exame Nacional de Desempenho dos Estudantes, ENADE) to assess students’ acquisition of knowledge and skills as well as graduates’ assumed learning gains in relation to their entry level. Education experts, as well as the OECD Review of Quality Assurance in Higher Education in Brazil, have highlighted some of its design and implementation weaknesses (see Box 3.1).

Since 2007, SAEB results have been used, alongside a measure of school performance based on the flow of students through schools (i.e. based on repetition and approval rates), to calculate the national education quality index (Índice de Desenvolvimento da Educação Básica, IDEB). This biannual index provides performance averages on a ten-point scale at the school, municipality, state and national level for each cycle of education: initial years (Years 1-5), final years (Years 6-9) and upper secondary education (Grades 1-3/4). Since 2005, Brazil has used IDEB to set targets for actors across the system, from the school to the national level. Target 7 of the PNE sets the following national targets for the 2021 cycle of IDEB: 6.0 for the initial years; 5.5 for the final years; 5.2 for Grades 3 or 4 of upper secondary education (Ministério da Educação, 2014[11]). These targets were calculated based on the quality of education systems in OECD countries benchmarked through a comparison between PISA and SAEB scores (INEP, 2019[3]). IDEB is widely acknowledged as having increased public awareness of school quality issues, encouraged improvement efforts, supported greater transparency and accountability, and informed the design and implementation of education policies, as well as the allocation of resources at the national and sub-national levels (OECD, 2014[12]). However, its design and use have raised some concerns, first and foremost, because the index is overly simplified. For example, IDEB does not take into account socio-economic differences across schools and school networks nor factors that influence learning outcomes outside the formal education system. Another concern is that despite these limitations, IDEB has increasingly high-stakes consequences for stakeholders, especially with regard to funding allocations. Efforts to develop a more comprehensive and balanced set of indicators to address such criticisms have stalled. Other concerns that have been voiced in Brazil relate to the use of IDEB to produce rankings, or for political and electoral purposes

In recent years, most states and several municipalities have developed their own standardised student assessments. These assessments vary according to their primary purpose, subjects, testing population (i.e. sample or census), grade levels, and other design features (Grupo de Trabalho de Avaliação do CONSED, 2018[13]). In many cases, however, these sub-national instruments replicate SAEB, often with the aim of providing information on students’ performace more quickly that can be used by policymakers and school leaders for planning purposes and by teachers in the classroom to monitor and support the learning progress of individual students. The duplication of effort is potentially very wasteful.

Target 7 of the PNE is to “promote the quality of Brazil’s basic education at all stages and modalities, with better progression rates and learning outcomes” and it sets targets for learning outcomes in the schooling system in terms of the IDEB, explained above (Ministério da Educação, 2014[11]). The PNE sets – mostly in broad terms – approaches and initiatives for how these targets are to be achieved (see Table 3.2).

Although their implementation has been largely decentralised to states and municipalities, several federal programmes aim to support learning, including:

  • The National Literacy Policy (Política Nacional de Alfabetização, PNA), established in 2019, aims to support early literacy by developing high-quality pedagogical resources, offering specific training on literacy development to teachers as part of their initial teacher education, developing diagnostic and assessment instruments and encouraging family-based reading habits (Presidência da República, Secretaria-Geral, 2019[14]). Two connected family literacy initiatives Conta pra mim, and Tempo de Aprender have been described in Chapter 2.

  • A federal Technical Assistance Network (Rede de Assistência Técnica) helps states and municipalities develop, monitor and assess their education plans, in line with the PNE (MEC, n.d.[15]).

  • The Support Programme for the New Upper Secondary Education (Programa de Apoio ao Novo Ensino Médio, ProNem) offers technical and financial support to states to implement the new upper secondary education model and curriculum standards (MEC, n.d.[16]) discussed in Chapter 2.

  • The Support Programme for the Implementation of the New Curricular Standards (Programa de Apoio à Implementação da Base Nacional Comum Curricular, ProBNCC) offers technical support to states and municipalities to develop, revise and implement their curricula in line with the new national curricular standards (MEC, n.d.[16]).

  • The Interactive PDDE (Direct Money to School Programme) Platform (Plataforma PDDE Interativo) is an online tool that supports school planning and management. Schools can develop an action plan for improvement based on an initial diagnosis of its strengths and weaknesses (MEC, n.d.[16]).

  • The National Programme of Educational Material and Books (Programa Nacional do Livro e do Material Didático, PNLD) involves distribution of pedagogical, literary and other materials to support the work of public school teachers (FNDE, 2017[17]).

  • The Full-Time Upper Secondary Education (Ensino Médio em Tempo Integral. EMTI) offers financial support to states that provide full-time school days (FNDE, 2017[18]). This programme has been associated with a decline in grade repetition, as discussed in Chapter 2.

As in many other large federal countries, in Brazil, states, municipalities and the Federal District can take the initiative to shape and define education policy and lead improvement efforts. First, following the national framework determined in the PNE, states, municipalities and other sub-national entities establish or adapt their own strategic plans for how they intend to achieve the set targets (see Table 3.2). State- and municipal-level governments often establish their own strategies (e.g. the Strategic Plan 2019-22 for the State of São Paulo, Plano Estratégico 2019-22), laying out medium-term objectives and policy plans (Secretária de Educação, 2019[19]).

In recent years, some states and municipalites have recorded impressive progress in terms of learning outcomes, partly as a result of decentralised initiatives, but also more equitable funding allocations (see Chapters 1 and 4). Studies suggest that many of the states making the greatest progress in improving education results are those that work most closely with their municipal school systems (Bruns, Evans and Luque, 2012[7]). Box 3.2 offers some noteworthy examples of state-level initiatives from Brazil and other federal countries.

In PISA 2018, Brazil’s outcomes remain quite a long way behind the OECD average and most benchmark countries across all subjects (see Figure 3.2). Students in Brazil scored on average 413 points in reading, 384 points in mathematics and 404 points in science (OECD, 2019[6]). The OECD averages in these three domains were 487, 489 and 489 score points, respectively. Brazil is well behind high-performing emerging economies, such as the Russian Federation and B-S-J-Z (China). Brazil’s performance is quite similar to that of its Latin American (LATAM) neighbours, although weaker than that of Chile, Costa Rica, Mexico and Uruguay.

In addition to presenting student results as score points, PISA 2018 also classifies student performance across proficiency levels. Each proficiency level is associated with what students at that level are expected to know and be able to do. According to PISA, in reading, students need to master the skills associated with Level 2 performance (equivalent to scoring at least 407 points) in order to be successful in further studies or the workplace. Students below Level 2 are classified as low performers and can only complete basic tasks (e.g. evaluating the literal meaning of simple sentences). At the other end of the spectrum, students who perform at Levels 5 and 6 (equivalent to scoring 626 and 698 points, respectively) are classified as top performers and are able to complete difficult tasks (e.g. inferring neutrality or bias in a text) (OECD, 2019[6]).

Results from the latest PISA cycle reveal that half of Brazilian students failed to reach the minimum threshold of performance in reading (50%) (see Figure 3.3). The share of students who do not demonstrate basic proficiency levels is even higher in mathematics and science (68% and 55%, respectively). In comparison, high-performing countries and economies — such as some parts of China — have well under 10% of their students classified as low performers in reading (B-S-J-Z (China): 5.2%). On average, OECD countries have just over one in five (22%) of 15-year-olds below Level 2.

National assessments indicate that underperformance at the age of 15 has its roots earlier in students’ academic lives. In 20164, only around 45% of pupils in Year 3 had an adequate level of reading in the National Assessment of Literacy (Avaliação Nacional da Alfabetização, ANA – now under the SAEB umbrella)5, and a similar proportion (45.5%) had an adequate level of mathematics (Todos Pela Educação, 2020[28]). This share gets progressively smaller after Year 5 as students advance in education, notably in mathematics Figure 3.4).

At the other end of the spectrum, very few 15-year-olds in Brazil demonstrate the highest levels of proficiency, as measured by PISA. In 2018, only 1.8% scored above Level 5 in reading, and less than 1% in mathematics and science. While these results are similar to other LATAM countries, in OECD countries, 9% of 15-year-olds scored above Level 5 in reading, 11% in mathematics and 7% in science.

The fact that Brazilian students struggle to tackle anything beyond the most basic tasks can be partly explained by the practices Brazilian teachers use in their classrooms. PISA results suggest that certain practices that research shows are more effective at supporting learning such as adaptive instruction are not as common in Brazil as in OECD countries (see Chapter 5 for a full discussion). These practices are particularly important for developing higher-order skills that students require to perform at the highest levels.

Strategies that students use to learn also matter. In PISA for example, students who often use memorisation techniques are more likely to correctly answer easy questions whereas those who use more elaborate strategies, such as trying to connect new concepts to what they already know, have a greater chance to correctly answer more difficult items. In Brazil however, students opt predominantly for memorisation and drill strategies over more elaborate learning strategies, which are unlikely to help them with the more advanced tasks (OECD, 2016[31]).

Modest improvements have been realised in Brazil’s PISA scores since 2003, mostly in mathematics6 (see Figure 3.5). However, Brazil’s average scores may mask a more positive trend. As discussed in Chapter 2, higher enrolment rates in Brazil have meant that the proportion of 15-year-olds covered by the PISA sample increased from about 55% in 2003 to 65% in 2018 (OECD, 2019[6]). A slight improvement in average performance in the context of a rapid expansion in access represents a considerable achievement. An estimation by the OECD, which focused on the 25% of better performers who are unlikely to be affected by changes in participation rates, suggested that Brazil’s performance in PISA has been improving by roughly ten score points every three years in the case of mathematics, and slightly less in the case of science. The results of this estimation for reading skills were not statistically significant (Figure 3.5) (OECD, 2019[6]).

Disaggregated results show that the country’s improvement in mathematics is mostly attributed to improvements in performance at the bottom of the distribution (below Level 2)7. The share of low-performers in mathematics declined by 7 percentage points between 2003 and 2018, and in science by 5.6 percentage points between 2006 and 2018 (see Figure 3.6). This not only means that a larger share of 15-year-olds are acquiring the basic skills and knowledge needed in life and work, but that performance gaps among students have reduced.

Moreover, while Brazil did not report state-level scores in 20188, results from previous cycles seem to suggest that much of Brazil’s improvement has been localised in a few states in the North and Northeast regions, notably Amazonas, Ceará, Maranhão and Pernambuco (see Annex 1). These states have managed to improve their results from a very low base and are increasingly closing the gap with the national average.

Data from surveys undertaken across Latin America by LLECE suggest that Brazil has made progress in the first years of education, most notably in mathematics scores in Grade 3. The scores increase from 505 in 2006 to 540 in 2013. Some other Latin American countries however reported even stronger increases (see Figure 3.7).

Similarly, national evidence from SAEB shows that between 2007 and 2019, the percentage of students in Year 5 reaching adequate levels of learning grew from 28% to 61% in Portuguese, and from 24% to 52% in mathematics (Todos Pela Educação, 2020[28]; MEC, 2020[29]). Progress has also been made in Year 9 according to SAEB. During the same time period, the share of students reaching adequate learning outcomes in Portuguese doubled from 20% to 41%, while for mathematics, the share increased from 14% to 24% (Todos Pela Educação, 2020[28]; MEC, 2020[29]). Upper secondary education has seen a significant progress in Portuguese, with the share of students in Grade 3 reaching adequate learning outcomes increasing from 24% to 37%; while for mathematics there was virtually no change (9.8% to 10.2%) (Todos pela Educação, 2020[33]). Given that secondary education expanded rapidly during this period, as with the PISA results, the apparent lack of progress in scores may reflect the expanded student population.

In addition to targets for 2024, the PNE has set intermediary annual targets, as measured by IDEB. At the national level, targets for the initial years of basic education have not only been consistently met, but evensurpassed. However, achievements in the final years of basic education and in upper secondary education have been below the national targets since 2013 (see Figure 3.8). This lack of progress not only reflects slow improvements in learning outcomes, but also persistently low approval rates (discussed in Chapter 2). Disaggregated results also show considerable variation between entities at the sub-national level, with states in the North and Northeast regions falling significantly behind others and national targets.

In Brazil, like many other countries, student performance is closely linked to socio-economic background. In PISA, socio-economic background is defined in relation to an index of economic, social and cultural status (ESCS — explained in the note to Figure 3.9). Across PISA-participating economies, socio-economically advantaged students tend to outperform students from disadvantaged backgrounds. In Brazil, the performance gap between students in the top and bottom national quarters of socio-economic status is equivalent to 97 score points in reading, above the OECD average (89 score points) (Figure 3.9).

In many respects the association between socio-economic background and learning outcomes is stronger in Brazil and in Latin American emerging economies than in most OECD and benchmark countries. Two-thirds (67.3%) of 15-year-olds in the bottom ESCS quarter in Brazil failed to reach Level 2 in reading literacy, compared to just over 27.6% of those in the highest ESCS quarter (see Figure 3.10). For OECD countries, the equivalent figures are 36% and 11% on average. Very similar patterns are found in the mathematics and science domains.

The impact of socio-economic background in Brazil is also visible from the earliest years. ANA 2016 data show that in primary school, and using a seven-group indicator of schools’ socio-economic status9, only 23% of the students from the lowest socio-economic group achieved adequate levels of literacy in reading, 40% in writing and 25% in mathematics – compared to 68%, 86% and 70% of the students in the highest socio-economic group, respectively (Todos Pela Educação, 2020[28]).

Figure 3.11 shows that students in the highest ESCS quarter have seen statistically significant performance improvements in reading in the latest cycles of PISA, whereas the most disadvantaged students (from the lowest ESCS quarter) have seen no improvements.

This is a concerning trend, as it means that the achievement gap between the most advantaged and disadvantaged students is increasing, with the least advantaged falling even further behind. In 2009, this difference was at 84 score points, and has increased to 97 in 2018. Across the OECD, the achievement gap between the most advantaged and disadvantaged has also increased recently, but at a lower level, from 87 in 2009 to 89 in 2018.

While reliable data on the impact of school closures and distance education on Brazilian students’ learning outcomes as a result of the COVID-19 pandemic are not available yet, studies predict that disadvantaged students will have fallen even further behind their wealthier peers. Prolonged episodes of school closures can increase inequalities if governments do not effectively implement measures to ensure every child has sufficient resources to learn in good conditions, particularly in countries where non-school factors play a strong role in learning outcomes (Gouëdard, Pont and Viennet, 2020[37]; Saavedra, 2020[38]), as is the case in Brazil.

PISA 2018 results show that 34% of Brazil’s variation in reading performance was observed between schools (OECD average: 29%) and 62% of Brazil’s variation was accounted for by differences within schools (OECD average: 71%) (see Figure 3.12). Internationally, in countries where there is substantial variation between schools, like Brazil, students tend to be grouped in schools in which other students perform at levels similar to their own. One implication is that, at the system level, there is less consistency with regards to performance, meaning that where parents enrol their children can have important implications on their future performance.

Internationally, students in private schools tend to perform better in PISA than those in public schools. The performance gap found in Brazil is however particularly large. Students from public schools in Brazil score below Level 2 on reading literacy over four times more often than students from private schools (57% as opposed to 13%). In Latin America the equivalent ratio is between 2 and 3; in the OECD it is only 1.5 (see Figure 3.13). National data from Brazil also show large performance differences between students in public and private schools at all levels of education (see Figure 3.14).

As discussed in Chapter 1, this gap is explained in part by differences in the intake of students between the public and private sector. Students in private schools tend to come from more advantaged socio-economic backgrounds (average ESCS: 0.2) according to the ESCS index than their peers in the public sector (average ESCS: -1.3). After taking into account the differences between students’ socio-economic background in the two sectors, the performance advantage of private schools in reading scores declines from 113 points to 32 points (OECD average: from 27 to -12.5).

However, even accounting for differences in school intake, a substantial and statistically significant difference remains between private and public schools in Brazil (see Figure 3.15). This contrasts with other LATAM countries, such as Chile, Colombia, Costa Rica and Peru, where there are no statistically significant differences in reading scores between private and public schools once ESCS is taken into account. This suggests that, once selection effects are taken into account, private schools in most LATAM countries do not deliver better quality education than public schools (OECD, 2011[41]), whereas in Brazil the evidence suggests real differences in quality between the two sectors, consistent with the findings of some researchers (Medeiros, 2016[42]). This means that parents who can afford to enrol their children in private schools can give them a learning advantage, which contributes to ensuring that socio-economic advantages are passed on to their children.

National and international assessments reveal large regional variations in performance. In PISA 2018, Brazil did not report state-level scores10, but results from previous cycles revealed significant disparities. In PISA 2015, the state of Espírito Santo scored 435 points in science, similar to Uruguay, whereas the state of Alagoas scored 360 points, below Albania and Kosovo. Such disparities are also reflected in national data. In the North and Northeast regions, average performance in SAEB was lower than the national average and lower than in the South and Southeast regions (see Figure 3.16). The gap between urban and rural locations are also significant, and particularly wide in the North, Northeast and Centre-West regions. This is most probably, at least in part, associated with the socio-economic and ethnic differences between schools and students across Brazil, whereby in schools in the North and Northeast states there are more disadvantaged students, and more students from minority backgrounds (see Chapter 1).

Seemingly, there is also a performance gap in PISA results between rural and urban schools in Brazil. However, once the socio-economic background of students is accounted for, this difference disappears or in fact suggests that rural schools, taking account of their student make-up do better. The score-point difference drops from around 80 points to the advantage of urban students to 33 points to the advantage of rural students, after accounting for ESCS (OECD, 2019[39]).

In Brazil, at age 15, girls are ahead of boys in reading by 26 score points, slightly behind boys in mathematics (9 score-point difference), and about the same level in science (1 score-point difference11) (see Figure 3.17). It is noteworthy that the gaps in favour of boys in mathematics and science observed in Brazil are much lower, whereas the advantage of girls in reading is much higher than in many LATAM countries, although similar to the average across OECD countries.

As in many other countries, there are large differences in career aspirations between boys and girls in Brazil. In OECD countries, more than three times as many boys as girls expect a career in engineering or computing (OECD, 2012[44]); conversely girls are three times more likely than boys to expect to work as doctors, veterinarians, nurses or other health professionals (OECD, 2017[45]). A similar pattern was observed in Brazil: among high-performing students in mathematics or science, about one in three boys in Brazil expects to work as an engineer or science professional at the age of 30, while only one in five high-performing girls expect to do so, but two in five of these girls expect to work in health-related professions (OECD, 2019[4]).

Unlike most PISA-participating countries, in Brazil, as in many other Latin American countries, students in pre-vocational or vocational programmes scored higher in reading than students in general or modular programmes. The difference is equivalent to 48 score points before accounting for students’ socio-economic background and 39 score points after accounting for it, which is significantly higher than in neighbouring countries and among the highest across PISA-participating countries (see Figure 3.18).

This disparity may be partly explained by the fact that Brazil’s vocational education and training (VET) systems have been, as discussed in Chapters 1 and 2, academic in focus and have attracted higher-performing students.

While the current reform to Brazil’s upper secondary education is expected to increase enrolment in VET programmes, this could have implications for student performance. International experience shows that the stigmatisation of lower status pathways (and this is often how vocational programmes are seen), can lead teachers to assign slower-paced and more fragmented instruction to students in these programmes. As a result, students adjust their expectations and efforts, leading to even lower performance. To counter such stigmatisation, it is crucial to maintain high expectations for all students by ensuring that all tracks offer an appropriately challenging curriculum and high-quality instruction.

National data reveal how race is linked to learning outcomes. White Brazilians generally perform better than average, mixed-race Brazilians around the average, and black Brazilians well below the average. For example in the 201712 Portuguese assessment for Year 9 (lower secondary education), around 52% of white Brazilians reached adequate learning levels, compared with 36% of mixed Brazilians, and 29% of black Brazilians (Todos pela Educação, 2020[33]). In mathematics, 32% of white Brazilians reached the expected standard, compared with 18% of mixed Brazilians and only 13% of black Brazilians (Todos pela Educação, 2020[33]). While these disparities are widely recognised, there seems to be limited research and few targeted policies that explicitly address this issue.

Internationally and in Brazil, labour market returns from education include both a better chance of getting a job and higher earnings once in work. These returns reflect, in part, greater productivity of individuals with more knowledge and skills, gained through education. The raw returns are also partly attributable to the “signalling” effect of education credentials, rather than to the additional knowledge and skills delivered by the associated education programme. Employers often use education qualifications as screening measures to identify the knowledge, skills and personality traits that granted individuals entry to an education programme and helped them successfully complete the programme, rather than the qualities developed through the education programme. Such signalled qualities include intelligence, self-control and persistence. Brazil does not yet participate in the OECD Survey of Adult Skills (PIAAC), which would offer a perspective on the actual levels of literacy, numeracy and digital literacy skills associated with higher levels of educational attainment.

In Brazil, the returns from educational attainment are high, and while similar to other LATAM countries, they are much higher than in OECD countries on average (Figure 3.19). In Brazil, those with bachelor’s degrees earn, on average, more than twice as much as those with upper secondary qualifications. In OECD countries on average, those with bachelor’s degrees earn only half as much as those with upper secondary education (OECD, 2020[46]). The returns are even higher in Brazil for those with a master’s or doctoral degree who on average have earnings more than four times higher than those with upper secondary education, much larger than the OECD average (twice as high) (OECD, 2020[46]).

One explanation for these comparatively high returns by international standards is the strong demand for, and weak supply of highly-qualified individuals (see Figure 3.20). Chapter 2 explained how the Brazilian labour market is increasingly demanding higher-level skills. At the same time, tertiary attainment levels, especially at the master or doctorate level, remain low by international standards.

Research also shows that the probability of informality decreases significantly with educational attainment (Mello and Santos, 2009[48]). The same authors also find that those whose first jobs are informal are more likely to remain in that condition for an extended period. This may result in long-term challenges, since informal jobs offer no protection against unemployment, sickness and old age. For the government, informality means that much economic activity is unregistered, unregulated and untaxed (OECD, 2014[12]).

As described in Chapter 2, Brazilian women have higher participation and completion rates than men. As a result, young women are now more likely to have completed upper secondary education or a tertiary degree. Similarly to what is observed internationally, women’s employment prospects improve considerably with better qualifications. However, it is noteworthy that, irrespective of their level of qualification, upon leaving education, women still face on average worse labour market prospects than their male peers:

  • The prospects of employment rise more sharply for women with better qualifications than for men. In Brazil, in 2018, under half (45%) of young women with less than upper secondary qualifications were employed, compared with over three-quarters (77%) of their male counterparts. At tertiary level the comparable figures were 81% and 90% (see Figure 3.21) (OECD, n.d.[47]). While the male-female gap in employment chances narrows at higher levels of educational attainment, data suggest that a relatively high share of well-educated women remain inactive or unemployed. This represents an important untapped resource for the country.

  • Women are more likely to be unemployed. According to data from 2018, in Brazil, the unemployment rate is 19% among young women with less than upper secondary education, compared with only 12% for males. Among the tertiary-educated, the comparable figures are considerably lower at 9% for women and 7% for men (OECD, n.d.[47]).

  • Despite wage returns from education, women still earn less than their male peers at all qualification levels. In 2015, women’s earnings were 69% of men’s earnings, among those with below upper secondary education. Among those with upper secondary and tertiary qualifications, women’s earnings were equivalent to 65% of men’s (OECD, 2017[49]). Discriminatory barriers in the labour market may explain at least part of these gaps (OECD, 2014[12]).

Young people who are neither in employment nor in education or training – commonly referred to as NEETs (jovens nem-nem in Portuguese) – are at greater risk of becoming socially excluded, with income below the poverty line and poor career prospects. In OECD countries, around 11% of young adults (aged 15-24) were in this category in 2019, but in Brazil, as in many other developing countries, rates are much higher. In 201813, in Brazil, nearly one-quarter of individuals in this age group were NEET (24%). The pattern is similar across Latin America: in Colombia 24% of 15-24 year-olds were NEET, as were 20% in Argentina and in Costa Rica (OECD, n.d.[47]).

NEET rates are strongly associated with individual’s education, socio-economic status, gender and ethnicity.

  • In 2018, almost half (47%) of those aged 18-24 and NEET had not completed basic education (ISCED 1 and 2). This suggests that many of those who drop out early on from school end up in the NEET category.

  • Unlike the majority of OECD countries, in Brazil, NEET is primarily a female phenomenon often attributed to early pregnancy, housework, childcare or other family responsibilities (OECD, 2014[12]). In 2018, nearly a third of young Brazilian women (aged 15-29) (31%) were classified as NEET, 12 percentage points higher than for men (19%) (see Figure 3.22). While the gap between men and women has decreased since the early 2010s, this is primarily due to a large increase in NEET rates among men rather than in any change of female NEET rates (OECD, 2014[12]). In 2012, an estimated 20% of Brazilian women aged 15-29 were NEET, almost twice the rate for young men (12%). In OECD countries, female NEET rates in 2018 (on average 16%) are only a little higher than those for men at 11% (see Figure 3.22).

  • Of those who are NEET, 42% were in the lowest household income quintile, while only 7% come from the richest quintile (IBGE, 2019[50]).

  • Black or mixed women are the most likely to be NEET (32%), compared to 22% of white women, 20% of black or mixed men and 15% of white men (IBGE, 2019[50]).

In 2018, white people earned on average 74% more than mixed and black people (IBGE, 2019[50]). In addition,mixed and black individuals are concentrated in economic sectors with low wages (e.g. domestic services and construction) (IBGE, 2019[50]). Even after taking into account educational attainment, the earnings of mixed and black Brazilians lag behind those of whites, with larger gaps at higher attainment levels. Earnings of the tertiary-educated white population were 45% higher than those of their mixed and black counterparts with the same level of qualification (IBGE, 2019[50]).

Ensuring that everyone takes part in education is only the beginning. Students also need access to quality education, where they are able to learn and develop knowledge and skills of value for the labour market and beyond. However, this is often not the case in Brazil. Low levels of skills can not only undermine individuals’ career opportunities, but also slow Brazil’s development and growth.

This chapter highlighted four main issues that Brazil faces:

  • First, relatively low outcomes in PISA at the age of 15 reflect a school trajectory where students are progressively falling behind. Results from national and international assessments show that students’ underperformance has its roots in the early years of schooling. Without mastering foundational skills, students are unable to progress and accumulate more advanced skills, falling increasingly behind expectations. Children from disadvantaged backgrounds are more vulnerable to this vicious cycle, having often entered school ill-prepared for the demands of formal education (World Bank, 2018[1]).

  • Second, despite improvements, many students leave school without the skills and knowledge they need for their future careers and lives. Brazil’s expansion in access to education and school enrolments has been accompanied by improvements in terms of learning outcomes. While this is a significant achievement, student performance, as measured by PISA, remains well behind OECD countries. Prolonged school closures and distance learning models put in place during the COVID-19 pandemic may also lead to learning losses. Sustaining and accelerating progress will become increasingly challenging as participation in education continues to rise, and larger numbers of disadvantaged students, who are more likely to have fallen behind in school, stay on in education. But national and international evidence has shown that this is not an unsurmountable challenge.

  • Third, performance gaps between the private and public sectors risk further accentuating socio-economic disparities. At present, Brazil’s education system does not offer a level playing field for children and young adults. Families that can afford to enrol their children in private schools are able to offer them better academic and professional prospects. Those from less well-off backgrounds, on the other hand, do not have the same opportunities and tend to attend lower-quality public schools. This damages their life chances, as measured by their employment status and earnings, both reflecting and compounding inequalities in the country. The COVID-19 pandemic and its socio-economic consequences add greatly to this challenge for the current youth cohort.

  • Fourth, Brazil still faces important gender gaps. In education, girls and women are more likely to be enrolled in school and progress to more advanced levels than boys. In the labour market, however, the trend reverses. Women show lower levels of employment and face a significant wage penalty, relative to men. More research should focus on the potential barriers women face to join and progress in the labour market (e.g. lack of high-quality and affordable childcare; scarcity of full-day schooling; discrimination in the labour market; unawareness of labour market opportunities for women; etc.). Overcoming these barriers would help Brazil make fuller economic use of its female working population and ensure women are able to successfully enter and remain in the labour market, realising their full potential.


[20] Aragon, S. and E. Workman (2018), Emerging state turnaround strategies, Education Commission of the States, https://www.ecs.org/wp-content/uploads/12139.pdf (accessed on 10 September 2020).

[5] Avvisati, F. (2017), “Does the quality of learning outcomes fall when education expands to include more disadvantaged students?”, PISA in Focus, No. 75, OECD Publishing, Paris, https://dx.doi.org/10.1787/06c8a756-en.

[7] Bruns, B., D. Evans and J. Luque (2012), Achieving World-Class Education in Brazil: The Next Agenda, World Bank, Washington DC, https://doi.org/10.1596/978-0-8213-8854-9.

[26] Candido, M. (2020), Qual o segredo de Goiás para liderar o melhor ensino público do país? [What is Goiás’ secret to lead the best public education in the country?], https://www.uol.com.br/ecoa/ultimas-noticias/2020/09/18/qual-o-segredo-de-goias-para-liderar-melhor-ensino-publico-do-pais.htm (accessed on 19 November 2020).

[21] Centre for Public Impact (2020), A política de Educação em Tempo Integral no Estado Brasileiro de Pernambuco [The policy of Full-time Education in the Brazilian State of Pernambuco], Centre for Public Impact, Instituto Natura, https://www.institutonatura.org/wp-content/uploads/2020/08/A-poli%CC%81tica-de-Educac%CC%A7a%CC%83o-em-Tempo-Integral-no-Estado-brasileiro-de-Pernambuco..pdf (accessed on 19 November 2020).

[18] FNDE (2017), Programa de Fomento às Escolas de Ensino Médio em Tempo Integral (EMTI) [Full-Time Schools Support Programme (EMTI)], https://www.fnde.gov.br/index.php/programas/programas-suplementares/ps-ensino-medio/ps-emti (accessed on 26 November 2020).

[17] FNDE (2017), Programas do Livro [Book Programmes], https://www.fnde.gov.br/index.php/programas/programas-do-livro/legislacao/item/9787-sobre-os-programas-do-livro (accessed on  2020 November).

[37] Gouëdard, P., B. Pont and R. Viennet (2020), Education responses to COVID-19: Implementing a way forward, OECD Publishing, Paris, https://doi.org/10.1787/8e95f977-en.

[13] Grupo de Trabalho de Avaliação do CONSED (2018), GT CONSED Avaliação da Educação Básica Relatório[CONSED WP Evaluation of Basic education Report, CONSED, Instituto Ayrton Senna, Instituto Unibanco, Fundação Itaú Social, http://www.consed.org.br/media/download/59ef86cca3349.pdf (accessed on 25 November 2020).

[2] Hanushek, E. and L. Woessmann (2015), The Knowledge Capital of Nations: Education and the Economics of Growth, The MIT Press, Cambridge, Massachusetts, London, England, https://www.jstor.org/stable/j.ctt17kk9kq (accessed on 19 November 2020).

[50] IBGE (2019), Síntese de Indicadores Sociais: Uma análise das condições de vida da população brasileira 2019, Instituto Brasileiro de Geografia e Estatística, Rio de Janeiro, https://biblioteca.ibge.gov.br/visualizacao/livros/liv101678.pdf (accessed on 23 April 2020).

[34] INEP (2020), Ideb - Resultados e Metas [Ideb - Results and Targets], http://ideb.inep.gov.br/resultado/ (accessed on 23 November 2020).

[9] INEP (2020), INEP se prepara para implantação do Novo Saeb em 2021 [INEP prepares for the implementation of the New Saeb in 2021], http://inep.gov.br/artigo2/-/asset_publisher/GngVoM7TApe5/content/inep-se-prepara-para-implantacao-do-novo-saeb-em-2021/21206?inheritRedirect=false (accessed on 6 July 2020).

[35] INEP (2020), Resumo Técnico Tesultados do Índice de Desenvolvimento da Educação Básica, Versão Preliminar [Technical Summary Results of the Basic Education Development Index, Preliminary Version], Instituto Nacional de Estudos e Pesquisas Educacionais Anísio Teixeira, Ministério da Educação, Brasília, http://download.inep.gov.br/educacao_basica/portal_ideb/planilhas_para_download/2019/resumo_tecnico_ideb_2019_versao_preliminar.pdf (accessed on 22 November 2020).

[3] INEP (2019), O que são as metas de qualidade educacional [What are educational quality goals], Instituto Nacional de Estudos e Pesquisas Educacionais Anísio Teixeira, http://portal.inep.gov.br/educacao-basica/ideb/metas (accessed on 20 April 2020).

[51] INEP (n.d.), Ações Internacionais, PISA: Resultados [International Actions, PISA: Results], http://portal.inep.gov.br/acoes-internacionais/pisa/resultados (accessed on 23 July 2020).

[40] INEP (n.d.), Resultados [Results], http://inep.gov.br/educacao-basica/ideb/resultados (accessed on 23 November 2020).

[27] Instituto Unibanco (2019), Como Goiás Superou Meta do Ideb no Ensino Médio [How Goiás Overcame Ideb’s Goal in High School], Instituto Unibanco, https://www.institutounibanco.org.br/aprendizagem-em-foco/45/ (accessed on 19 November 2020).

[24] Loureiro, A., E. Di Gropello and O. Arias (2020), There is no magic: The formula for Brazil’s Ceará and Sobral success to reduce learning poverty, https://blogs.worldbank.org/education/there-no-magic-formula-brazils-ceara-and-sobral-success-reduce-learning-poverty?token=53176c4095d917916aa31ea735b5ceaa (accessed on 19 November 2020).

[23] Loureiro, A. et al. (2020), The State of Ceara in Brazil is a Role Model for Reducing Learning Poverty (English), World Bank, Washington, D.C, http://documents.worldbank.org/curated/en/281071593675958517/The-State-of-Ceara-in-Brazil-is-a-Role-Model-for-Reducing-Learning-Poverty (accessed on 19 November 2020).

[29] MEC (2020), Saeb Resultados [Saeb Results], https://www.gov.br/inep/pt-br/areas-de-atuacao/avaliacao-e-exames-educacionais/saeb/resultados (accessed on 22 November 2020).

[15] MEC (n.d.), Rede de Assistência Técnica [Technical Assistance Network], http://pne.mec.gov.br/8-videos/27-rede-de-assistencia-tecnica (accessed on 6 July 2020).

[16] MEC (n.d.), SEB, http://portal.mec.gov.br/secretaria-de-educacao-basica/programas-e-acoes (accessed on 6 July 2020).

[42] Medeiros, M. (2016), Income inequality in Brazil: new evidence from combined tax and survey data, UNESCO, Paris, https://unesdoc.unesco.org/ark:/48223/pf0000245825.

[48] Mello, R. and D. Santos (2009), Aceleração Educacional e a Queda Recente da Informalidade [Educational Acceleration and the Recent Fall in Informality], Instituto de Pesquisa Econômica Aplicada (Ipea), http://repositorio.ipea.gov.br/handle/11058/4066 (accessed on 26 August 2020).

[11] Ministério da Educação (2014), Plano Nacional de Educação - Lei N° 13.005/2014 [National Education Plan - Law No. 13.005/2014], http://pne.mec.gov.br/18-planos-subnacionais-de-educacao/543-plano-nacional-de-educacao-lei-n-13-005-2014 (accessed on 24 August 2020).

[8] Mol, S. (2019), PROVA BRASIL: Uma análise da complexidade cognitiva de itens de Matemática por meio da Taxonomia SOLO [PROVA BRASIL: An analysis of the cognitive complexity of Mathematics items through the SOLO Taxonomy], Universidade Federal de Ouro Preto, Mariana, Minas Gerais, https://www.repositorio.ufop.br/bitstream/123456789/11597/1/DISSERTA%c3%87%c3%83O_ProvaBrasilAn%c3%a1lise.pdf (accessed on 17 December 2020).

[46] OECD (2020), Education at a Glance 2020: OECD Indicators, OECD Publishing, Paris, https://doi.org/10.1787/69096873-en.

[43] OECD (2020), PISA 2018 Results (Volume V): Effective Policies, Successful Schools, PISA, OECD Publishing, Paris, https://dx.doi.org/10.1787/ca768d40-en.

[39] OECD (2019), PISA 2018 database, https://www.oecd.org/pisa/data/ (accessed on 26 August 2020).

[6] OECD (2019), PISA 2018 Results (Volume I): What Students Know and Can Do, PISA, OECD Publishing, Paris, https://dx.doi.org/10.1787/5f07c754-en.

[36] OECD (2019), PISA 2018 Results (Volume II): Where All Students Can Succeed, PISA, OECD Publishing, Paris, https://dx.doi.org/10.1787/b5fd1b8f-en.

[4] OECD (2019), Programme for International Student Assessment: Results from PISA 2018: Brazil, https://www.oecd.org/pisa/publications/PISA2018_CN_BRA.pdf (accessed on 24 August 2020).

[10] OECD (2018), Rethinking Quality Assurance for Higher Education in Brazil, Reviews of National Policies for Education, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264309050-en.

[49] OECD (2017), Education at a Glance 2017: OECD Indicators, OECD Publishing, Paris, https://dx.doi.org/10.1787/eag-2017-en.

[45] OECD (2017), “What kind of careers in science do 15-year-old boys and girls expect for themselves?”, PISA in Focus, No. 69, OECD Publishing, Paris, https://dx.doi.org/10.1787/76e7442c-en.

[31] OECD (2016), Ten Questions for Mathematics Teachers… and How PISA Can Help Answer Them, PISA, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264265387-en.

[12] OECD (2014), Investing in Youth: Brazil, Investing in Youth, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264208988-en.

[44] OECD (2012), “What Kinds of Careers do Boys and Girls Expect for Themselves?”, PISA in Focus, No. 14, OECD Publishing, Paris, http://dx.doi.org/10.1787/5k9d417g2933-en.

[41] OECD (2011), Private schools: Who Benefits?“, PISA in Focus, No. 7, http://dx.doi.org/10.1787/5k9h362mhtkd-en.

[47] OECD (n.d.), Education and Training / Education at a Glance, https://stats.oecd.org/ (accessed on 22 November 2020).

[14] Presidência da República, Secretaria-Geral (2019), decreto nº 9.765, de 11 de abril de 2019 [Decree nº 9.765, of April 11 2019], http://www.planalto.gov.br/ccivil_03/_ato2019-2022/2019/Decreto/D9765.htm (accessed on 11 September 2020).

[22] Rosa, L. et al. (2020), The effects of public high school subsidies on student test scores, https://www.sonhogrande.org/storage/the-effects-of-public-high-school-subsidies-on-student-test-scores-the-case-of-a-full-day-high-school-program-in-pernambuco-brazil.pdf (accessed on 19 November 2020).

[38] Saavedra, J. (2020), Educational challenges and opportunities of the Coronavirus (COVID-19) pandemic, https://blogs.worldbank.org/education/educational-challenges-and-opportunities-covid-19-pandemic (accessed on 4 September 2020).

[19] Secretária de Educação (2019), Plano Estratégio 2019-2022, Governo do Estado de São Paulo, https://www.educacao.sp.gov.br/wp-content/uploads/2019/07/plano-estrategico2019-2022-seduc_compressed.pdf (accessed on 6 July 2020).

[25] Secretária de Estado da Educação de Goiás (2018), Educação: Investimentos fazem de Goiás o 1º lugar na Educação pública do País [Education: Investments make Goiás the 1st place in public education in the country], https://site.educacao.go.gov.br/educacao-investimentos-fazem-de-goias-o-1o-lugar-na-educacao-publica-do-pais/ (accessed on 19 November 2020).

[28] Todos Pela Educação (2020), Anuário Brasileiro da Educação Básica 2020 [Brazilian Yearbook of Basic Education 2020], Todos Pela Educação, https://www.todospelaeducacao.org.br/_uploads/_posts/456.pdf?1969753478/=&utm_source=content&utm_medium=site-todos (accessed on 7 July 2020).

[30] Todos Pela Educação (2020), Em 10 anos, aprendizado adequado no ensino médio segue estagnado, apesar dos avanços no 5° ano do fundamental [In 10 years, adequate learning in high school remains stagnant, despite advances in the 5th year of elementary school], https://todospelaeducacao.org.br/noticias/meta-3-em-10-anos-aprendizado-adequado-ensino-medio-segue-estagnado-avancos-5-ano-fundamental/ (accessed on 6 August 2020).

[33] Todos pela Educação (2020), Relatório Anual de Acompanhamento do Educação Já! [Annual Monitoring Report of Education Now!], Todos pela Educação, https://www.todospelaeducacao.org.br/_uploads/_posts/417.pdf?229296618%2F=&utm_source=Download-Relatorio-anual (accessed on 18 May 2020).

[52] Todos pela Educação (2019), Anuário Brasileiro da Educação Básica 2019 [Brazilian Yearbook of Basic Education 2019], Todos pela Educação, https://www.todospelaeducacao.org.br/_uploads/_posts/302.pdf (accessed on 4 May 2020).

[32] UNESCO (2014), Comparación de resultados del Segundo y Tercer Estudio comparativo y explicativo: SERCE y TERCE, 2006-2013, UNESCO Office Santiago and Regional Bureau for Education in Latin America and the Caribbean, https://unesdoc.unesco.org/ark:/48223/pf0000244239 (accessed on 24 August 2020).

[1] World Bank (2018), World Development Report. Learning to Realize Education’s Promise, https://www.worldbank.org/en/publication/wdr2018.


← 1. Brazil, like many Latin American countries, takes part in regional assessments of younger children, coordinated by the UNESCO Regional Bureau of Education for Latin America and the Caribbean, and the Latin-American Laboratory for Assessment of the Quality of Education (Laboratorio Latinoamericano de Evaluación de la Calidad de la Educación, LLECE). The First Regional Comparative and Explanatory Study (Primer Estudio Regional Comparativo y Explicativo, PERCE) was implemented in 1997 in 13 countries. Nine years later, in 2006, the Second Study (Segundo Estudio Regional Comparativo y Explicativo, SERCE) tested students in 16 countries plus one Mexican state. The Third Study (Tercer Estudio Regional Comparativo y Explicativo, TERCE) was implemented in 2013 in 15 countries and the same Mexican state. PERCE, SERCE and TERCE have measured learning outcomes of reading and mathematics of children in Grades 3, 4 (only in 1997) and 6. Brazil has participated in all rounds of the studies. Brazil also takes part in the International Civic and Citizenship Study (ICCS).Their participation in these assessments was until very recently one of the reasons why Brazil had not undertaken PIRLS and TIMSS. However, the Brazilian government recently announced that Brazil will take part in PIRLS 2021 and TIMSS.

← 2. This refers to PISA Coverage Index 3, which represents the coverage of the national 15-year-old population. This estimates the proportion of the national population of 15-year-olds covered by the non-excluded portion of the student sample. The index is below 1.0 to the extent that 15-year-olds were excluded, or not enrolled in Grade 7 or higher. For further information on Brazil’s student sample in PISA, please access: https://www.oecd.org/pisa/data/pisa2018technicalreport/

← 3. A proposal to reform the SAEB had been put forward in early 2021, but will no longer be rolled out as planned. The government still plans to implement many of the proposed changes, albeit under a different timeline. A forthcoming OECD report provides insights to this discussion.

← 4. In the most recent, 2019 edition of SAEB, Year 2 students were assessed in literacy for the first time. Previously, SAEB assessed Year 3 students. This change reflects Brazil’s new common curricular base, which establishes that students should be literate by the end of Year 2. However, given that the assessment of the Year 2 was sample-based, rather than census-based, the authors have chosen not to include these results in the graph.

← 5. Up until 2016, Grade 3 students who reached Levels 3 and 4 in ANA were considered to have adequate proficiency levels in literacy. However, the federal government has not explicitly defined adequate proficiency levels for other grades or subjects for SAEB 2019. The OECD has opted to use the approach that other national stakeholders (such as Todos Pela Educação) take, whereby: in Portuguese, students in Years 5, 9 and in Grade 3 or 4 of upper secondary education who reach Level 4 on the SAEB proficiency scale are considered to have adequate levels of learning; in mathematics, students in Years 5 and 9 who reach Level 5, and students in Grade 3 or 4 of upper secondary education who reach Level 6 are considered to have adequate levels of learning.

← 6. The data show improvements in reading performance between the PISA 2000 cycle and the PISA 2018 cycle.

← 7. Only for reading was there a statistically significant increase in the share of high achievers between 2012 and 2018

← 8. PISA 2018 results allow however a comparison across regional averages. Results show, as discussed, that the South, Southeast and Centre-West regions have higher scores across all subjects than the North and Northeast regions.

← 9. This refers to the National Institute of Educational Studies and Research Anísio Teixeira (Instituto Nacional de Estudos e Pesquisas Educacionais Anísio Teixeira, INEP) indicator of socio-economic level, which is based on data obtained from students' responses to contextual questionnaires. Students are categorised into eight socio-economic levels, with Level 1 being the lowest and Level 8 the highest on the scale. Schools are then categorised by their students’ socio-economic profile, ranging from 1 to 6 (1 to 7 until 2016). Schools at Level 1 have the highest concentration of students from the lowest socio-economic backgrounds and schools at Level 6 have the highest concentration of students from the highest socio-economic background (Todos pela Educação, 2019[52]).

← 10. PISA 2018 results allow however a comparison across regional averages. Results show, as discussed, that the South, Southeast and Centre-West regions have higher scores across all subjects than the North and Northeast regions.

← 11. The difference is not statistically significant

← 12. SAEB 2017 instead of SAEB 2019 results were used in this paragraph because at the time of the draft of this publication, the microdata for SAEB 2019 with information disaggregated by race for example, was not yet available.

← 13. Most recent year with available data for Brazil in the OECD statistics database.

Metadata, Legal and Rights

This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. Extracts from publications may be subject to additional disclaimers, which are set out in the complete version of the publication, available at the link provided.

© OECD 2021

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at http://www.oecd.org/termsandconditions.