Chapter 3. The organisation of the school network

This chapter examines how the Portuguese school system is managed, and the school opportunities available to students from a variety of backgrounds. In particular it attempts to address the question of whether the system is currently structured to promote the success of all students, including those for whom traditional schooling structures have been ineffective. Portugal has made considerable advances in designing a well-organised school network over the past 15 years, including consolidating many under-resourced schools, investing in some of its school buildings and expanding access for students with special education needs. However, important challenges remain related to shifting demographics, many deteriorating buildings and substantial regional- and school-level inequalities. The chapter makes a number of recommendations to address these challenges, particularly as it relates to the governance of schools.


The statistical data for Israel are supplied by and under the responsibility of the relevant Israeli authorities. The use of such data by the OECD is without prejudice to the status of the Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of international law.

Context and features

System-level governance

As described in Chapter 1, the governance of the Portuguese education system is relatively centralised; however, there has been a long-term, gradual shift towards decentralisation. Decentralisation in the education sector is not isolated but constitutes part of a gradual delegation of responsibilities to the local level in public sectors such as health and transportation. The education sector was one of the first to begin decentralising after the Revolution. Initially, municipalities were responsible for student transportation, meals and facilities management. In 1984, municipalities also became responsible for the construction and maintenance of buildings and equipment for early childhood education and care (ECEC), primary education and adult education, and for managing socio-educational activities (Ministry of Education, 2018[1]). During the past decades, more financial authorities and executive powers have been assigned to municipalities concerning school buildings, equipment, non-teaching staff and Curriculum Enrichment Activities (Atividades de Enriquecimento Curricular – AEC) in pre-school and the 1st cycle of basic education. Since 2015, the decentralisation process has been accelerated by the introduction of a four-year pilot with inter-administrative contracts (contratos interadministrativos) between 13 municipalities and the central government. These contracts expand the financial authorities and executive powers of municipalities in such areas as hiring non-teaching staff, social support at school, construction, maintenance and equipping school buildings in the 2nd and 3rd cycles, transport, family support and AEC in pre-school and the 1st cycle of basic education.

The decentralisation process involves multiple subnational entities: inter-municipal associations, municipalities and parishes. Parishes operate under municipal structures. In the Lisbon Municipality, the responsibilities for pre-school education and the 1st cycle of basic education have been further decentralised from the municipal level to the civil parishes, by means of inter-administrative contracts. In this way, Lisbon parishes play an important role in education by undertaking maintenance of buildings, hiring non-teaching staff, organising study supervision and support, social support, extracurricular activities and school holiday activities, providing meals and launching specific educational projects. Outside Lisbon, parishes have fewer responsibilities, only organising study supervision and support, undertaking small repairs or providing school bus transport.

The decentralisation goal articulated by the current government in education is to provide autonomy to municipalities to distribute funding for current spending on education – except teachers’ salaries – and capital expenditures in schools under their jurisdiction. This would allow municipalities more discretionary power, thereby promoting responsive governance close to the needs of its citizens and efficient in its operation. Current government leaders stressed during the review visit, however, that they intended to keep centralised responsibilities related to teacher hiring, placement and pay, as well as curriculum and the planning of the school network.

Organisation of the school offer

Public schooling

The Portuguese public school offer is organised in three sequential levels described in Chapter 1. Both public and private providers guarantee the school offer in Portugal. The Continental school network consists of 5 729 public schools and 2 569 private schools (CNE, 2017, pp. 45, 49[2]). The public school network is structured in school clusters that integrate schools from different education levels in one organisation. In 2016, there were 713 school clusters and 98 non-clustered schools. Almost all non-clustered schools provide secondary education exclusively (Ministry of Education, 2018[1]).

The average individual school size varies significantly across regions, often as a function of the regions’ population density. Panel A of Figure 3.1 shows the average school size by region. In the Lisbon Metropolitan Area and in the upper northwest around Porto, individual schools enrol on average more than 600 students, whereas in the rural northeast and in the western part of Alentejo, schools enrol, on average, 350 to 425 students.

The average size of clusters also varies significantly, including substantial regional variation. The modal school cluster size is five to nine schools, but clusters range from as small as two schools to as many as 30 schools (Figure 3.1, Panel C). In the Lisbon Metropolitan Area, the average school cluster size consists of five to six schools, whereas in rural areas of central Portugal, school clusters are composed of, on average, more than ten schools (see Figure 3.1, Panel B).

Central authorities in Portugal define the legal standards for class size. In 2013, the maximal number of students per class was legally increased by 2 students per class in basic and general secondary education (typically from 24 to 26 students), and by 6 students per class in vocational courses (typically from 18 to 24 students) but the new government has prioritised class size reduction and plans to reduce the sizes to pre-2013 levels in 2018/19 (Ministry of Education, 2018[3]). However, as Chapter 4 discusses in detail, even after the changes to class size maximums, the actual class sizes in Portugal are around the OECD average of 21 and 23 in primary and secondary schools respectively.

A regular school day comprises of five to eight hours of classroom instruction, divided by some short breaks. School day extension is becoming more common in Portugal: an increasing number of students have lunch at school and spend additional time in extracurricular activities (Ministry of Education, 2018[3]). The majority of students (86% in 2015/16) enrol in the full-time school programme for the 1st cycle of basic education. The current government aims to extend this full-time school programme to other educational levels. In addition to standard academic programming, 797 of 809 school clusters and non-clustered schools offer school sports clubs (Desporto Escolar). Students remain at school after the standard school day free of charge and may participate in 36 different sports with over 7 000 teams across all municipalities. To date, however, no comprehensive evaluation of the impact of Desporto Escolar on students’ health or well-being has been conducted.

Simultaneous to the extension of the school day for many students, 10% of schools still work in double shifts where some groups of students have classes only in the morning, while others attend only in the afternoon, especially in the densely populated suburbs of Lisbon (Ministry of Education, 2018[3]).

Figure 3.1. Variation in average school size and number of schools per cluster, 2015/16
By NUTS III region

Source of administrative boundaries: Direção-Geral do Território (2016), Official Administrative Maps of Portugal - Version 2016 [Carta Administrativa Oficial de Portugal - Versão 2016],

Sources: CNE (2017), Estado da Educação 2016 [State of Education 2016], Conselho Nacional de Educação, Lisbon,, Tables 2.1.1 and 2.2.2; DGEEC administrative data, 2015/16.

Private schooling

According to the Constitution of the Portuguese Republic and the Comprehensive Law on the Education System (see Chapter 1), the state has the duty to provide public education to all, as well as the duty to allow and to certify private initiatives in education. So, families have the right to attend private schooling, but at their own expense, unless there is no public offer where they live or their child has a specialised need not met in the local public schools.

Almost a third (30%) of Portuguese schools are private, either government-dependent and privately run, or private independent. The share of private schools has increased substantially in recent years for various reasons. The number of public schools was nearly halved from 10 443 in 2006/07 to 6 078 in 2015/16, whereas the number of private schools grew from 2 587 in 2006/07 to 2 708 in 2015/16.

Private independent schools are self-financed through attendance fees, charged to students’ families. On the other hand, government-dependent private schools utilise a variety of contracted funding models with the government (see Chapter 2). Figure 3.2 presents total student enrolment numbers in public, government-dependent private and private independent schools by education level. Government-dependent private schools are most prevalent in the ECEC sector in institutions jointly financed and managed by the Ministry of Education and the Ministry of Labour, Social Security and Solidarity (MTSSS). Private independent schools serve students in all levels but especially in VET programmes in upper secondary, as well as at the ECEC and 1st cycle primary educational levels. Annex A describes this distribution by the number of schools and highlights the dominance of private providers in the ECEC sector.

Figure 3.2. Student enrolment by type of provision and programme orientation, 2015/16

ME: Ministry of Education; IEFP: Institute for Employment and Vocational Training.

Source: Ministry of Education (2018), OECD Review of Policies to Improve the Effectiveness of Resource Use in Schools: Country Background Report for Portugal, Ministry of Education, Lisbon,, Table 2.2.

Government-dependent private schools are intended to fill gaps in the public supply of schooling in oversubscribed locales, remote locations, specialised artistic areas or special education. Figure 3.3 displays geographic patterns of enrolment in all private schools (shaded areas) and in government-dependent private schools (numbers). While some regions of Portugal have high rates of private school enrolment (up to one-quarter of all students), for the most part, the provision of private school places subsidised by the government exists only in areas where public school capacity is strained. Across the 278 Continental municipalities, there are only 21 municipalities in which the rate of government-dependent private school enrolment is over 10% and where fewer than 20% of schools at any educational level are over-capacity (defined here as having a total number of classes that exceed the number of available classrooms by 5% or more). In 8 of these 21 municipalities, no educational level has more than 20% of schools operating over capacity. Thus, with the exception of these few municipalities with no capacity issues but high government-dependent private enrolment, government-dependent private schools seem to be authorised, in line with stated priorities, to respond to the demand the public schools are unable to meet.

Figure 3.3. Private school enrolment
By NUTS III region

Note: The shaded regions on the figure display the proportion of students enrolled in all private schools in that region. The percentages inside the regions display the proportion of all students enrolled in government-dependent private schools. Government-dependent private schools are private schools with an association contract (contrato de associação), sponsorship contract (contrato de patrocínio) or any other type of agreement with local or central education authorities that provides public funding directly to schools. Schools enrolling students who receive direct financial assistance from the government that covers tuition fees in private schools are not considered government-dependent.

Source of administrative boundaries: Direção-Geral do Território (2016), Official Administrative Maps of Portugal - Version 2016 [Carta Administrativa Oficial de Portugal - Versão 2016],

Source: DGEEC administrative data, 2015/16.

Government-dependent private schools must obey the same set of criteria for the selection of students as those for public schools. Private independent schools have no specific regulatory constraints for the selection of students. They are free to set their own criteria as long as these do not violate non-discriminatory principles. As a consequence, private schools may give preference to children from specific backgrounds. Access to, or exclusion from, private independent schools is not subject to government inspection (Glenn and De Groof, 2002[4]).

The pedagogical and academic level of the curricula in government-dependent private schools must meet the national standards of general education policy. Independent private schools can follow the national curriculum or offer an alternative to be approved by the Inspectorate-General for Education and Science (Inspeção-Geral da Educação e Ciência – IGEC). In order to gain this approval from IGEC, schools must disclose the qualifications of their non-teaching staff, account for their educational philosophy by spelling out its educational orientation, explaining the principles by means of its “educational project” (projecto educativo) and show adequate physical facilities (Glenn and De Groof, 2002[4]). All private schools have the authority to determine what will be taught in at least 20% of the instructional time and choose textbooks and other materials without prior government approval. Private schools regulate their teachers' salaries in function of a pay scale which is based off the one used at public schools, though some divergence is possible (Ministry of Education, 2018[1]). With respect to accountability, government-dependent private schools are subject to compulsory inspection of their teaching, learning and administrative processes by the Ministry of Education’s inspection services, though these take place outside of the regular external evaluation process. Inspection of private schools can include an evaluation of teaching methods, as long as the technical judgment is not of the ideological, philosophical or religious basis of the teaching (Glenn and De Groof, 2002, p. 422[4]).

School governance, performance evaluation and accountability

School governance

Due to the strong position of the central government in education and the relatively centralised decision-making, school-level governance in Portugal has traditionally been weak. It has long been characterised by limited school leadership skills and responsibilities. Teachers’ pedagogical approaches and curricular decisions tend to be dictated by central fiat (Ministry of Education, 2018[1]).

The nature of school-level governance changed in 2008 when the structure of schools and school clusters was recalibrated and professionalised; this provided more responsibilities to school cluster management and increased accountability. Today, a management team consisting of the principal and deputies manages each school cluster or non-clustered school. Most cluster principals now have some form of formal leadership training and ongoing professional leadership development courses are offered. Nonetheless, principals and other administrators continue to be conceived of within the profession as teachers on assignment, rather than professional managers (see Chapter 4). The management team shares its policy- and decision-making powers with and is assisted by a pedagogic council (Conselho Pedagógico), consisting of heads of subject departments and educational staff, and an administrative council (Conselho Administrativo), consisting of the principal, a deputy and financial staff. Chapter 4 explores leadership responsibilities and school management structures in more detail.

The formal governing body of each school cluster (and non-clustered school) is the General Council (Conselho Geral). The General Council is composed of teachers, non-teaching staff, parents, secondary students, representatives from the municipality and representatives co-opted by the elected membership. The General Council selects the school cluster principal, approves the educational improvement plan for the school, and conducts the school’s internal evaluation Since the beginning of the process of democratisation, school autonomy has been framed in terms of participation and democracy at subcentral levels; a primary justification for the inclusion of teachers, students, parents and local stakeholders in school governance is this goal of democratic participation. The inclusive representation on the General Council is broader in Portugal than in many other democracies (Glenn and De Groof, 2002[4]).

Despite the original intent of a broad-based, representative local governance structure, due to lack of community participation and shifts in General Council membership, concerns have arisen about whether the General Councils continue to promote a governance structure that is fully accountable to all stakeholders and student interests. In the 1990s, the predominance of teachers in decision-making and the powers they were granted became a topic of contention. Debates ensued about the extent to which parents and members from local communities should participate in school governance (Eurydice, 2007, p. 11[5]). At the turn of the century, the vision of school autonomy gradually shifted from a political-administrative reform to a mechanism to improve the quality of education, emphasising the link between pedagogical and curricular autonomy and school success.

As it stands today, provisions for school autonomy in the governance of educationally important decisions are limited. The current pilot autonomy contracts (Pilot Project of Autonomy and Curriculum Flexibility) in Portugal allow for the carving out of time in school schedules during which modest amounts of the national curriculum (up to 25% of teaching time) may be tailored by the school to the specific needs and interests of its students. All schools will be eligible to participate starting in the 2018/19 school year. Additionally, starting in 2016/17, seven school clusters launched a separate Pedagogical Innovation Pilot Project (Projeto-Piloto de Inovação Pedagógico – PPIP) providing more complete curricular and scheduling autonomies with the goal of improving learning outcomes and reducing year repetition rates to zero. However, other types of school autonomy such as the autonomy to manage financial resources within the school cluster by means of a lump-sum budget instead of earmarked funding, or autonomy to manage human resources such as the freedom to select teachers, are not typically part of the policy discussions around school autonomy.

School accountability

According to international classifications of accountability in education, the current school accountability mechanisms can best be characterised as strong regulatory school accountability complemented with modest elements of school performance accountability (Hooge, 2016[6]). Strong regulatory school accountability enforces compliance with laws and regulations and focuses on inputs and processes within the school by means of reporting to the central government (Hooge, 2016[6]). It is a vertical accountability mechanism: top-down and hierarchical. The traditionally strong emphasis on the use of regulatory school accountability in Portugal corresponds with its highly centralised government and legalistic approach of educational policy- and decision-making.

Recently, Portugal has introduced some elements of school performance accountability on top of regulatory school accountability. School performance accountability mechanisms focus on outputs of schools such as efficiency and effectiveness, holding schools accountable for the use of resources in relation to the quality of education they provide (Hooge, 2016[6]). The new forms of school performance accountability introduced in Portuguese education are periodic school performance evaluations. By design, they were intended to involve a five-year evaluation cycle which would include a visit from an external inspection team examining: i) student performance data; ii) the quality of education service provision; and iii) leadership of the school. Coupled with the internal evaluation, schools were required to draft improvement plans based on the results of these external evaluations and a follow-up was planned to assess the extent to which the school was making progress towards these goals (Santiago et al., 2012[7]).

However, the 2nd cycle of external evaluations concluded in 2017 and at the time of the site review, no external evaluations were underway. Stakeholders reported that, during the 2017/18 school year, they were conducting internal evaluations but no external evaluations were taking place. Furthermore, at the time of the drafting of the report, no overall cycle report had been completed for the 2012-17 evaluation cycle. Finally, though a working group was meeting to discuss details of the 3rd evaluation cycle, no framework or schedule of evaluations had been created (IGEC, 2018[8]). Thus, for the 2017/18 school year, no external evaluations were to be conducted. In practice, therefore, Portugal’s form of performance accountability recently has been accomplished by means of common student assessments and by public reporting of school performance. While the public reporting of performance is a form of performance accountability, even on paper Portugal’s inspection and testing regime are relatively low-stakes compared to contexts in which poor performance can result in sanctions and high performance may result in additional autonomies or incentives.

Distribution of students to schools

In the 2015/16 school year, more than 1.5 million children and youngsters went to school in Portugal. The enrolment in pre-school education is above international averages (see Chapter 1). Particularly the share of 3-year-olds in ECEC has increased by 17 percentage points during the last decade, from a 63% enrolment rate in 2006/07 to 80% in 2015/16 (CNE, 2017, p. 81[2]). The enrolment in Portuguese basic education overall has been stable during the past decade. Almost all 15-year-olds (97%) are enrolled in the 3rd cycle of basic education, which is at the OECD average (OECD, 2017, p. 257[9]). As Chapter 1 highlights, total student enrolment has remained relatively stable over the past decade, reflecting countervailing trends of increased ECEC and upper secondary enrolment and declining basic education enrolment. Given that Portugal is approaching the upper bounds of enrolment rates and it is experiencing a long-term decline in its school-age population, it is expected to see a decline in overall enrolment in the coming years.

As discussed in Chapter 1, Portugal relies on geographic assignment to schools. While siblings’ enrolment is considered first, this, of course, is dictated by the initial enrolment of the oldest child. After special educational needs are considered, students’ legal residence is the next and most influential factor in school assignment. Thus, for parents enrolling their children in the public school system, the primary mechanism by which they can influence the school their child attends is through their choice of residence and to a lesser degree their choice of employment location. As a result of concerns about some families manipulating their residential or occupational address to access a more desirable school, for the 2018-19 academic year, the criteria have been changed to require students to use their legal address, defined via the tax declaration process, for their public school application. Additionally, as noted in Chapter 1, new for the 2018-19 academic year, students receiving social support will have preferential status after the above factors are applied in an attempt to increase opportunities for low-income students and increase socio-economic integration in schools.

Parents’ choices of residence, and by extension schools, are heavily influenced by public reporting on school quality. The yearly coverage by the media of school rankings based on average scores in national tests during the past decade and a half has generated competition between schools and led to socio-economic- and achievement-based segregation between schools (Ministry of Education, 2018[1]). Until recently, the school rankings published in Portugal did not take students’ demographic characteristics into account. Currently, the Directorate-General for Education and Science Statistics (Direção-Geral de Estatísticas da Educação e Ciência – DGEEC) publishes performance ratings of schools that consider the demographic characteristics of their student bodies following recommendations from the OECD Review of Evaluation and Assessment in Education (2012[7]). The DGEEC also publishes indicators of school performance online in the Schools Portal (InfoEscolas) and the IGEC external evaluation reports of all public schools are available on the internet. However, the most intensive media attention remains on the raw rankings based solely on test scores published in leading newspapers.

The results of a recent study show that the reporting of school rankings in Portugal has had significant effects on families’ choices. Specifically, an improvement of 10 ranking places for the average school is associated with an increase of about 0.4 percentage points in the number of enrolled students. Additionally, lower performing schools, particularly private ones, are more likely to close since the public release of rankings (Nunes, Reis and Seabra, 2016[10]).

Residential segregation, geographic assignment and public rankings all contribute to the fact that the level of between-school socio-economic segregation in the Portuguese education system is substantial. Figure 3.4 and Figure 3.5 display the unevenness in the distribution of students across schools across regions in the country. Figure 3.4 reveals the high degree of between-school segregation, in which children receiving School Social Assistance (Ação Social Escolar – ASE) and who have low levels of maternal education are concentrated in particular schools. In half of continental Portugal’s NUTS III regions, including the most populated areas of Lisbon and Porto, the Gini index measuring between-school socio-economic variation is above 0.25. This is higher than the findings in a recent study that found between-school socio-economic variation in the 100 largest United States metropolitan areas registered a Gini index of 0.23 (Owens, Reardon and Jencks, 2016[11]). These findings accord with a recent DGEEC report finding high rates of socio-economic segregation within 2nd cycle schools. In particular, the report found a startling difference in the school enrolling the highest percentage of students receiving School Social Assistance (78%) in the Lisbon municipality and the school enrolling the lowest percentage (8%). Even more startling, in the 2nd cycle Lisbon school with the lowest levels of maternal education, 91% of students had mothers who had not completed secondary education. Conversely, in the school with the highest level of maternal education, only 2% of students had mothers who had not completed secondary education (Oliveira Baptista, Pereira and DGEEC, 2018[12]). Similar variation held in the Porto municipality as well.

Figure 3.4. Level of student segregation across schools by socio-economic status (Gini coefficient)
By NUTS III region

Note: Index of socio-economic need measured by rank percentile-ordering the proportion of students within a school receiving School Social Assistance A and average years of maternal education at the school level. The average of the 2 ranks was then demeaned (mean = 0) and assigned a standard deviation of 1. High values of the index indicate high levels of socio-economic challenge. A within-NUTS III region Gini index was calculated based on the standardised value of school-level disadvantage. The Gini coefficient ranges from 0 to 1, where 1 represents perfect inequality and 0 represents perfect inequality between subunits.

Source of administrative boundaries: Direção-Geral do Território (2016), Official Administrative Maps of Portugal - Version 2016 [Carta Administrativa Oficial de Portugal - Versão 2016],

Source: DGEEC administrative data, 2015/16.

Similarly, while the overall proportion of immigrants in Portugal tends to be low, immigrant students are more common in some regions and are concentrated in a small number of schools. Immigrant student enrolment is highest in Lisbon and southern Portugal, whereas there are relatively few immigrant students in eastern Alentejo, the Centre and the North. More striking is the range of immigrant enrolment across schools in those areas that do have more immigrant students. Schools in the Lisbon Metropolitan Area (AML) that enrol more immigrant students than 90% of all other AML schools are 15.1% immigrant. On the other hand, schools that enrol fewer immigrant students than 90% of all other AML schools are only 1.5% immigrant (Figure 3.5). Thus, it is evident that for a small but not marginal proportion of schools, meeting the needs of immigrant students is a central concern.

Figure 3.5. Immigrant student distribution
By NUTS III region

Note: The proportion of immigrant students is the weighted average of the proportion of the immigrant enrolment rate for all schools within a region. The interdecile range is the difference between the immigrant enrolment rate for schools in the 90th percentile for immigrant enrolment within the region and the immigrant enrolment rate for schools in the 10th percentile. For example, in the Metropolitan Area of Lisbon (AML), schools in the 90th percentile for immigrant enrolment have immigrant enrolment rates of 15.1%. Schools in the 10th percentile have immigrant enrolment rates of 1.5%. Thus, the interdecile range for immigrant enrolment in AML is 13.6 percentage points.

Source of administrative boundaries: Direção-Geral do Território (2016), Official Administrative Maps of Portugal - Version 2016 [Carta Administrativa Oficial de Portugal - Versão 2016],

Source: DGEEC administrative data, 2015/16.

The levels of between-school segregation evidenced by the school-system wide census data are borne out by PISA samples of 15-year-old students. Within the OECD, Portugal has the 5th highest rate of between-school socio-economic variation, trailing only Turkey, Spain, Chile and Mexico (Table I.6.10 in OECD (2016[13])). An important caveat is that Portugal also has some of the largest overall variations in students’ socio-economic status within the OECD which drives a large proportion of the level of between-school segregation (Figure II.5.12 in OECD (2016[14])).

Between-school segregation is also related to between-school performance variation. As noted in Chapter 1, student performance in Portugal is strongly correlated with the environment in which children grow up and the composition of the student population of each school is also closely related to average learning outcomes on national assessments. PISA data reveal in more detail how between-school segregation relates to between-school performance variation. There is a moderately strong, positively signed correlation between school systems that have higher rates of social inclusion across schools and school systems that have higher rates of performance inclusion across schools (Figure II.5.12 in OECD (2016[14]).

In addition to evidence suggestive of high degrees of public school segregation, there are also patterns of socio-economic segregation between public and private schools. Among schools enrolling 15-year-old students, schools in the top quartile of socio-economic status are 13 percentage points less likely to be public schools than those in the bottom quartile of socio-economic status (Table II.4.10 in OECD (2016[14])). Further, when examining census data of all Portuguese Year 6 students, Brás de Oliveira (2018[15]) found that 1.3% of students at private independent schools receive social support, while 45.4% in government-dependent private ones receive social support and 53.6% in public schools. Thus, there are clear socio-economic differences between students in private and public schools, and this is particularly true for private independent schools (Brás de Oliveira, 2018[15]).

With respect to the level of within-school segregation, Table 3.1 shows that in Portugal, tracking at the age of 15 is in line with most OECD countries. Year repetition in Portugal is almost 3 times higher than the OECD average (31.2% compared to 11.3%). Official ability grouping is much lower than the OECD average (4.3% compared to 7.8%) (Table II.5.22 in OECD (2016[14]). However, no system-wide regulations exist governing how class groups are formed that would prohibit class-level sorting on the basis of prior academic achievement. As a result, important concerns exist about these informal mechanisms of within-school sorting (see below).

Table 3.1. Grouping and student selection



OECD average


(Age of selection into different education types or programmes)



Year repetition

(Percentage of students who have repeated a year at least once in primary, lower secondary or upper secondary school)



Ability grouping

(Percentage of students in schools where students are grouped by ability into different classes for all subjects)



Source: OECD (2016), PISA 2015 Results (Volume II): Policies and Practices for Successful Schools,, p. 178.

Contributing to the relatively high levels of segregation, the Portuguese education system offers multiple curricular pathways for students with special profiles, starting in the first cycle of basic education. Besides the general path, attended by most students in basic education, about 8% of the total number of enrolled students complete the Year 9 under alternative offerings (Ministry of Education, 2018[1]). In particular, the provision of Alternative Curricular Pathways (percursos curriculares alternativos – PCA) is targeted to students in the 2nd and 3rd cycles of education who are overage, have learning difficulties and are at-risk of year repetition or dropping out. This type of offer is intended to be temporary in nature and aims to motivate students to re-gain interest in school. There are also programmes that target highly mobile students. A programme for Distance Learning for Itinerant Students (Ensino a Distância para a Itinerância) seeks to ensure basic school education to students that, due to their parents’ occupation or lifestyle, are forced to change schools frequently, particularly targeting students from Roma communities. Regular education primary and secondary programmes also offer Portuguese as a Second Language (Português Lingua Não Materna – PLNM) for students who need Portuguese language instruction. Students are initially orally screened and families complete a home language survey. Students then sit for a placement test if necessary to determine Portuguese proficiency levels. Students in vocational programmes do not have access to PLNM courses in all schools. Finally, home-schooling (ensino doméstico) is permitted, as long as it is under the responsibility of a qualified adult and the supervision of a school (Ministry of Education, 2018[1]).

For adults, several second-chance education programmes are available, depending on their needs and previous educational experiences, whose provision is under the umbrella of the recently created Qualifica programme (see Chapter 1, Box 1.2).

Adapted provision for students with SEN

School education in Portugal recognises the special educational needs (SEN) of some students. The law for inclusion (Decree-Law 3/2008), at the time of writing of this report, does not explicitly define special educational needs nor formally set criteria for identification of SEN, rather leaving to the discretion of specialised personnel operating according to international standards. Differentiated provision for SEN students translates into the adaptation or modification of curricula, and potentially the assignment of a student to a different learning environment. Education of students identified as SEN is almost exclusively provided in mainstream schools – 98.8% of SEN students were integrated into such settings as of 2016 – with special education schools being now almost entirely restricted to fulfil a role of resource centres for inclusion (see Chapter 2 and below). Students may only attend special education institutions, a total of 50 of which exist nationwide, under approval from the Ministry of Education and when learning limitations are sufficiently severe to require education in a separate school (EASNIE, forthcoming, p. 23[16]). The identification and enrolment of students with special education in regular schools have significantly increased in recent years, alongside a continuous decrease of enrolment in special education schools (Figure 3.6). Mainstream-school educated students with SEN grew by 76% between 2010 and 2017. In total, 5.8% of all students are identified as SEN, most of whom attend basic education (7.7% of all students in Years 1-9) (Ministry of Education, 2018, p. 88[1]).

Specialised support to students with SEN within schools is complemented by a network of 93 specialised resources centres for inclusion (centros de recursos para a inclusão – CRI) and 25 information and communication technology (ICT) resource centres for special education spread across the country. About 581 school clusters (72% of the public school network) receive support from CRIs. CRIs deploy a total of 2 251 technicians, such as occupational therapists, speech therapists, physiotherapists or psychologists, and 1 141 technicians work directly in schools (DGEEC, 2017[17]). Additional resources have recently been targeted to the support of students with SEN. In 2016/17, an additional 221 full-time psychologists were allocated to schools, with a commitment to contract an additional 200 psychologists in 2018 in order to improve the student-psychologist ratio to 1 140-to-1. The resource centres are designed to support the inclusion of children with disabilities, build partnerships with local actors and facilitate the access of students with SEN to different activities. The network of support is further complemented by the identification of schools specialised in teaching students with given disabilities. Finally, students with severe needs rely on two types of structured teaching units: one for the education of students with autism spectrum disorders and another for the education of children with multiple disabilities and congenital deaf-blindness (Ministry of Education, 2018[3]) where students are educated in largely separate classrooms, even if in a mainstream school building. Together, these types of specialised teaching units supported 5% of SEN students, as of 2016/17.

Figure 3.6. Change in the number and proportion of students with SEN by type of school, 2010-17

Source: DGEEC (2017), Necessidades Especiais de Educação [Special Educational Needs]

Since 2017, and while this report was being prepared, there has been public discussion of a new law on the inclusion of students with special educational needs. The reform aims to clarify the roles of the different actors involved in the identification of SEN students. In particular, the new law aims to make a medical evaluation an optional, rather than mandatory, procedure to identify a student as having SEN and trigger the development of an Individualised Education Plan (Planos Educativos Individuais – PEI). The intent of this reform is to recognise special educational needs as learning obstacles, rather than medical conditions.

Vocational education and training

Vocational Education and Training (VET) plays an important role in upper secondary education in Portugal. When entering upper secondary education, typically at 15 years of age, students choose between different tracks and fields of study (see Chapter 1). Just as several strands are available in the general track, initial VET is separated into five strands. For the purpose of this report, initial VET courses refer to a vocational offer that is targeted at young students – typically until 24 years of age – and at the upper secondary level of education. Table 3.2 provides the main features for each of these programmes.

Table 3.2. Characteristics of initial VET programmes

Type of programme

Theoretical duration of studies (in years)

Number of learning hours

Work-based learning (%)

Fields of study

Target age (years-old)

Professional programmes


3 200




Apprenticeship programmes


2 800-3 700




Specialised artistic



[in the 3rd year]



CEF courses


1 125-2 276




Vocational courses






.. : Information not available.

Source: DGERT (2016), Vocational Education and Training in Europe - Portugal

Each type of VET programme combines specific learning components. The development of general skills is built into socio-cultural or scientific components, which correspond to learning that occurs in the classroom context at school. On the other hand, technical, technological or practical components are typically learned on the job or in a simulated working environment. As Table 3.2 displays, each programme has different dosages of work-based learning, often dependent on the field of study. The fields of study vary considerably. Professional courses, the most popular among VET students (Figure 3.7, Panel C), offer as divergent study options as computer sciences, electronics, engineering, tourism, business administration, construction or applied arts, among forty different possibilities. In turn, apprenticeship programmes focus on priority areas of training such as computer sciences, construction and repair of motor vehicles or manufacture of textiles, among others. As in most other OECD countries, completion of any of the VET offerings leads to a double certificate allowing students to pursue further studies in post-secondary education. However, access to tertiary academic education is subject to the same entrance requirements as for upper secondary students from the general track (DGERT, 2016, pp. 17-19[18]).

Upper secondary VET is currently offered through different networks of provision. Building on the strategic orientation to improve the status of VET and the extension of compulsory education, Ministry of Education schools have recently expanded their offering of VET courses. As of the 2015/16 school year, 453 schools offered VET. On the other hand, a network of public, private independent and publicly-funded professional schools (escolas profissionais) also provides upper secondary and post-secondary education VET. There were 224 professional schools offering upper secondary VET, only 16 of which were strictly public. Finally, 33 professional training public centres managed by the Institute for Employment and Vocational Training (IEFP), as well as other private providers certified by the institute, offer apprenticeship programmes, in which over 26 000 students are enrolled annually.

Figure 3.7. Trends and distribution of the provision of VET programmes

Note: Panel A depicts the trend in the number of young students in each of the tracks as a proportion of the total enrolment of young students in upper-secondary education. In Panels A and B, the proportion of students in the regular track includes those enrolled in technological courses, a type of vocationally-oriented offer that is included in the regular track and has been progressively discontinued (see Chapter 1). In Panel B, IVET refers to initial vocational education and training, i.e. VET offer targeted at young students. IVET offer includes professional, apprenticeship, specialised artistic, vocational and education and training courses (cursos de educação e formação – CEF). In Panel C, enrolment in adult learning courses include individuals in recurrent upper secondary classes (ensino recorrente), education and training courses for adults (Educação e Formação de Adultos – EFA), certified modular training (Formaçãos Modulares Certificadas – FMC) and in the national system of prior learning assessment and recognition (Sistema Nacional de Reconhecimento, Validação e Certificação de Competências – RVCC).

Source: DGEEC (2016), Estatísticas da Educação 2015/2016 [Education Statistics 2015/16]$clientServletPath%7D/?newsId=145&fileName=DGEEC_DSEE_2017_EE201520164.pdf.

Current government leaders are eager to improve the status of VET as an alternative track to general education. The current government has stated the goal of reaching a 50-50 distribution of upper secondary students across general and vocational programmes by 2020. But the historical reputation of vocational education in the country is weak. After a backlash against the development of a “technical education” sector following the democratisation process started in 1974, it was not until 1989 when publicly regulated professional schools offering VET were created. The opening of professional schools occurred alongside the emergence of professional courses offered in IEFP training centres, mainly created to address local employment needs. The integration of VET courses in the mainstream education system only occurred in 2006, with the development of an alternative vocational track in upper secondary public schools (Ministry of Education, 2018[1]). Since then, the share of young upper secondary students enrolled in initial VET programmes has been steadily increasing, reaching about 41% in 2015/16 (Figure 3.7, Panel B). The surge has been followed by a decrease in the provision of technological courses, offered alongside sciences-humanities programmes in the general track. Technological courses, despite being vocationally-oriented, do not offer any work-based learning and have been essentially discontinued, enrolling only about 1% of upper secondary youth (Figure 3.7, Panel A).

As in most OECD countries, enrolment in VET programming is tied to students’ socio-economic status. Upper secondary schools with higher proportions of socio-economically disadvantaged students concomitantly have higher proportions of students enrolled in VET programmes. In fact, as Figure 3.8 shows, in schools that enrol the most students receiving social support and with the lowest levels of maternal education, an average of 47% of students enrol in VET programmes. On the other hand, in the most advantaged schools, only 22% of students study in VET courses. A key policy consideration is whether the greater concentration of VET students in schools with higher levels of socio-economic need is an attempt to design curriculum that better engages this group of students and helps them progress towards secondary school completion, or whether the differences reflect a lowering of expectations and a circumscription of educational and professional opportunities for students from low socio-economic status backgrounds (OECD, 2018[19]).

The strategic priority to develop the vocational education and training system in Portugal has been extended to lower secondary education. CEF courses are offered to students who have not yet reached upper secondary education, targeting those at risk of early school leaving at the 2nd or 3rd cycles of basic education. In 2013, pre-vocational programmes (cursos vocacionais no ensino básico) were introduced for students 13 or older, who have repeated at least twice in the same cycle or three times overall. Classes were organised into modules and teaching was intended to have a strong involvement of private companies from communities surrounding the school. The expansion of this type of offer was limited to schools with technical and pedagogical capacity and did not have full national coverage. However, this type of offer did not gather political support due to concerns with the effectiveness and equity of tracking students earlier and has been discontinued since 2016. Offerings in CEF courses for students 15 or older are still an option available to schools to tackle early school leaving (DGERT, 2016[20]; Ministry of Education, 2018[3]).

Figure 3.8. Proportion of upper-secondary students enrolled in VET programme by level of school disadvantage

Note: The figure shows the distribution of the proportion of VET students by quartile of an index of socio-economic disadvantage. Index of socio-economic disadvantage is measured by rank percentile-ordering the proportion of students within a school receiving School Social Assistance A (ASE A) and average years of maternal education at the school level. The average of the 2 ranks was then demeaned (mean = 0) and assigned a standard deviation of 1. High values of the index indicate high levels of socio-economic challenge.

Source: DGEEC administrative data, 2015/16.

Planning of the VET network is centrally steered and monitored by ANQEP (Agência Nacional para a Qualificação e o Ensino Profissional), an inter-ministerial public agency (see Chapter 1, Annex B). The recently created System for Anticipation of Qualification Needs (Sistema de Antecipação de Necessidades de Qualificações – SANQ) is the cornerstone of the process of strategic decision-making with regards to VET. Together with the broader architecture of the National System of Qualifications (SNQ), which extends to adult learning (see Box 3.1), the SANQ plays an important role in closing the gap between VET courses on offer and labour market needs. SANQ aims at increasing both vertical co-ordination of VET and promoting horizontal co-operation among relevant actors at subnational levels (in this case, the 23 inter-municipal communities, CIMs - Comunidades Intermunicipais). SANQ has three main strategic priorities: i) a diagnosis of past and forecasted labour market dynamics; ii) a planning module aimed at building a ranking of qualification priorities; and iii) a regional co-ordination module. The planning module includes the characterisation of the VET offer profile, relating the current number of enrolled students by type of offer and qualification priorities. The ranking of priorities is defined by ANQEP based on the assessment of labour market needs, in the first stage. The regional co-ordination module combines the two first priorities at the regional level: inter-municipal communities are asked to adjust centrally-defined priorities according to SANQ’s methodology, with priority given to local stakeholders’ perspectives. The CIMs that decide to not participate in this phase of the process abide the prioritisation established at the central level. After updating the priorities according to CIM feedback, ANQEP has the legal power to define and regulate the minimum and maximum number of classes to open in each CIM. According to ministry officials, currently the criteria are only being applied to providers under the Ministry of Education supervision.

Box 3.1. National System of Qualifications (SNQ)

Overview of main instruments

  • The National Qualification Framework (NQF) defines a hierarchical structure of qualification levels aligned with the European Qualifications Framework (EQF).

  • The National Qualifications Catalogue (Cátalogo Nacional de Qualificações – CNQ) is a catalogue of non-tertiary qualifications under the purview of the National Agency for Qualifications and Professional Training (Agência Nacional para a Qualificação e o Ensino Profissional – ANQEP). For each qualification, the CNQ identifies relevant skills, which are broken down as “skill units” (unidades de competência) as well as relevant training units of short duration (unidades de formação de curta duração – UFCD). The catalogue is regularly updated and managed through the collaboration of 16 different Sector Councils for Qualification co-ordinated by ANQEP. Each sector council integrates several stakeholders, such as employers, trade unions, schools and VET providers, technology and innovation centres, sector regulators, professional associations and invited experts. The participants in each sectorial council, as well as other entities registered in SNQ, can at any time propose the integration of new qualifications or the deletion of existing ones, through an online tool – the Open Model of Consultation.

  • The National Credit System for VET seeks to exploit the modular structure of the CNQ to implement flexible training paths. Every skill unit or UFCD corresponds to certain credits, enabling users to easily capitalise on prior skills acquisition when pursuing new qualifications. It is used only in double-certification training courses.

  • The Integrated System of Information and Management of Education and Training Supply (‘SIGO’) is a platform created for registering training activities of individual students with providers in the SNQ, whether their qualifications are part of the CNQ or not. Upon conclusion of any training activity, a certificate is issued by SIGO.

Source: OECD (2018), Skills Strategy Implementation Guidance for Portugal: Strengthening the Adult-Learning System,


Clear governmental priorities and targets with respect to expanding access to schooling

The current and previous Portuguese governments have set very clear education policy priorities and targets which have pinpointed precisely how and for whom school places are needed across the education system. These include:

  • Universal expansion of upper secondary access in 2008-09, followed by increases in graduation rates to 90% by 2020.

  • Universal access to early childhood education and care (ECEC) for 3-5 year-olds by 2019.

  • 90% participation in extended school day (full-time schooling).

  • Increase in enrolment in VET programmes to 50% by 2020.

  • Involvement of 600 000 adults in education by 2020.

These priorities will provide education access at nearly universal levels. The realisation of these priorities and targets require sufficient solid physical infrastructure, appropriate network planning and recruitment of a new wave of teachers (see below and Chapter 4). Importantly, when these priorities are realised, it will create additional pressure to ensure the quality provision of the expanded educational offer.

Articulated priorities for decentralisation in education

In spite of its highly centralised governance of the educational system, there has been broad political and societal support for decentralisation in Portugal since the democratic revolution of 1974, favouring the development of civil society and participation at the local level. The primary priorities the ministry and government currently articulate for decentralisation in education relate to the construction and maintenance of school buildings, the hiring and employment of non-teaching staff, and peri-educational activities such as full-day enrichment activities, sports, etc.

The inter-administrative contracts between central and local governments (see above) seem flexible enough to explore different approaches to decentralisation and to conduct experiments. The review team observed promising policies and practices at the municipality level, such as the development of local targeted educational projects, the promotion and fostering of education networks across teachers, school leaders, parents and families, educational stakeholders and private partners and the involvement of private entities in training school administrators, ad hoc management consulting, school visits to companies and timely traineeship opportunities for students and engagement in the design of VET and general schooling offerings.

Despite clearly articulated priorities around decentralisation in education, the political leadership of the ministry is clear that the following core three areas are not under consideration for local control: hiring and placement of instructional staff; curriculum; and the organisation of the school network.

There is an ambitious and overarching strategy to respond to declining student populations

The Portuguese education system has witnessed a major process of consolidation in the past decade, leading to a considerable reduction of educational institutions in the public school network. Between 2004 and 2014, Portuguese educational authorities shuttered more than 47% of public education institutions – a total of 5 600 schools (Figure 3.9, Panel A), compared to a decline of about 15% of students enrolled in primary education during the same period (Figure 1.7, Panel A). In 2006, a new policy let municipalities recalibrate the 1st cycle school network and thousands of small schools were closed. Prior to the reform, the school network was dominated by small primary schools with poor facilities and low performance – particularly in rural areas (Ministry of Education, 2018[1]). As Panel B of Figure 3.9 shows, while the school network in the densely populated Lisbon Metropolitan Area shrunk by only 6%, the consolidation process meant the closure of almost half the schools in the rural North (45%) and Centre (46%) regions.

Central authorities have recently been employing additional efforts to also reduce the number of government-dependent private schools. Contrary to the tendency in the public school network, the number of private schools increased in all regions but Alentejo. In the Lisbon Metropolitan Area, for every 10 public schools that were closed, a net average of 7 new private providers entered the school network (Figure 3.9, Panel C). While there are no strict regulations to discourage the growth in independent private provision, such increase in the face of declining student populations can lead to greater misallocations of school resources, double offering and greater pressure to keep consolidating the public network (see Challenges section). In order to curtail double provision of school places, the current government has further tightened the criteria for the public funding of private providers, de-funding classes in areas with available public offer, thus signalling the intention to keep rationalising the overall offer. Availability of centrally collected data allowed the decision to be based on a study of available infrastructure capacity and travelling distances to the closest public schools, effectively reducing duplicate offer (DGEEC, 2016[21]). These approaches align with the OECD recommendations in the Responsive School Systems: Connecting Facilities, Sectors and Programmes for Student Success report (OECD, 2018[19]) to tie the licensing of government-dependent private providers to the local need for additional school places as an effective means to enhance the efficiency of school networks, particularly in areas characterised by lack of available places in public schools.

Consolidation can be a disruptive experience for students and families having to relocate to a new school, often further removed from their home. The experience of OECD review countries has demonstrated the importance of public authorities and schools working together to make this transition as smooth as possible. Careful planning can ameliorate negative impacts on students’ well-being and learning outcomes, for example by ensuring the provision of school transport services where needed (OECD, 2018[19]). In Portugal, educational authorities implemented complementary policies alongside the reduction of the number of schools to address the challenges of consolidation. These policies improved the quality and capacity in the school network, to potentially cushion the shock caused by school closures.

First, the close co-ordination between central and local authorities, responsible for identifying the underperforming schools in need of consolidation and ensuring the transport of students to their new schools, was instrumental to the success of the process. Municipalities, in partnership with civil parishes, played a prominent role in reducing the disruptive nature of school closures.

Second, at the time of the initial round of school closures, new school centres that offered youth and community programming were built to serve as small town hubs. Further, a new school transport system was developed to ensure students who suddenly had to commute to other localities could do so with no additional costs for families.

Figure 3.9. Change over time in school facilities, 2004-14

Source: DGEEC (2017), Educação em Números 2016 [Education in Numbers 2016], DGEEC.

Third, the Secondary Schools Modernisation Programme, led by the Parque Escolar agency, invested in refurbishing and modernising secondary school buildings. The additional capital investment stream permitted the expansion and requalification of school buildings that would otherwise lack the capacity to take new incoming students. All stages, from the planning process to the maintenance of completed buildings, included a thorough community involvement process (Veloso, Marques and Duarte, 2014, p. 410[22]). At the local level, this engagement took the form of information sessions and consultation meetings before and during the construction process, which brought together parents, teachers, students, school boards and non-teaching staff on the one hand, and engineers and architects on the other hand (Blyth et al., 2012, p. 47[23]) (see below for more on Parque Escolar). Independently from the Parque Escolar interventions, but as part of the overall strategy of consolidation, many small schools have also been replaced by new buildings with a minimum capacity for 150 students (Ares Abalde, 2014[24]).

Fourth, the organisation of school administration into clusters has also facilitated the consolidation process. Clustering allows students to be integrated into larger school communities, increasing access to a greater variety of services, such as extracurriculars or student counselling and career guidance. Relatedly, school clustering eased transitions of students across education levels. In the same cluster, students could more easily progress through the years within the same extended school community, allaying concerns associated with moving to different school environments and facilitating a sense of belonging at school. Furthermore, as resource planning is made at the cluster level, variations in demand for a particular school can be more easily dealt with by shifting human and material resources across commonly managed school buildings.

Finally, the new organisation of the school offer provided greater pedagogical coherence across education levels. Teachers are now more aware of student needs as they progress through different educational levels, provided that additional opportunities for professional collaboration in an extended school community are actually fulfilled (see Chapter 4).

Although consolidation plans across OECD countries are often met with strong opposition from students, parents, teachers and staff (Ares Abalde, 2014[24]), most stakeholder groups the review team spoke with did not express particular concerns with the general direction of the process, recognising it as a necessary step to rationalise provision in face of falling birth rates. However, particularly at the school level, stakeholders with whom the review team spoke voiced concerns related to uneven administrative oversight over the schools in the cluster, greater weight given to administrative tasks distributed to teachers and an unequal utilisation of collaborative potential across clusters (see Chapter 4). In fact, the clustering process has also led to unequal distributions of enrolled students in each cluster and number of schools per cluster (see Challenges section).

The physical infrastructure of schools requalified by Parque Escolar is of high quality

In 2007, the Portuguese Government launched an unprecedented plan to modernise and improve public secondary school infrastructures, implementing a new management and maintenance model and optimising the allocation of resources and the sustainability of school infrastructures (Ministry of Education, 2018[1]). The state-owned company Parque Escolar was established to carry out the programme. A total of 173 secondary schools have been involved, which means that the school infrastructure of more than two-thirds of all secondary school students in 2016 has been improved (Ministry of Education, 2018, pp. 67-68[1]).

In only one decade, the central government succeeded in building, renovating, upgrading and pedagogically modernising 173 upper secondary schools throughout the country. A strength of the Parque Escolar model is linking the design of the school building to the development of innovative and modern instructional spaces, such as advanced laboratories and flexible classroom layouts. Blyth and colleagues provide an extensive description of the Secondary Schools Modernisation Programme (Blyth et al., 2012[23]).

During school visits to sites intervened by Parque Escolar, the review team heard from students, parents, teachers and school administrators a high-level satisfaction with their new secondary school infrastructure. They stated that they felt the new buildings met the requirements of teaching and learning, well-being and social safety in school. Parque Escolar involved strategic and thorough infrastructural planning and maintenance. Nevertheless, the successful process and results of Parque Escolar highlight the poor physical condition of most other schools (see Challenges section).

There is increased access to, and attainment in, upper secondary education

Graduation rates from upper secondary education have been climbing and are approaching OECD averages. Between 2005 and 2015, the proportion of upper secondary students under 25 years of age that graduate from secondary schooling jumped from 51% to 83% – by far the largest increase among countries for which there is available data (Figure 3.10). Interestingly, the increase in the secondary enrolment and graduation rates preceded the increase in the compulsory age of education to 18 years of age in 2009. In fact, between 2005 and 2015, the proportion of 18-year-olds enrolled in education increased from 66% to 81% (Figure C1.2 in OECD (2017, p. 251[9])). Declining rates of early school leaving and year repetition in basic education are important drivers of this change.

Figure 3.10. Change in graduation rates from upper secondary education, 2005 and 2015
Students below 25 years old

Note: Countries with missing data for both periods or with missing data for 2015 are not presented. The change for OECD average is not presented due to missing data.

Source: OECD (2017), Education at a Glance 2017: OECD Indicators,, Table A2.3.

It should be noted that the surge in graduation rates and expansion of the upper secondary schooling occurred in spite of efficiency-driven measures aimed at consolidating the school network and declining number of teachers (see also Chapter 4). Costs associated with the greater need of provision were also contained through a general increase in average class size, associated with the increased legal maximum number of students per class in 2013. Therefore, it seems that, while holding resources constant in the face of a growing secondary population, Portuguese secondary schools still increased the proportion of students graduating successfully. Additionally, clearly articulated standards – e.g. through a stable policy of national exams at the end of the 3rd cycle and during upper secondary education – have steered the quality and rigour of the curriculum delivered in the classroom. Greater emphasis should now be given to continue decreasing the still high rates of year repetition and early school leaving (see Challenges section).

Efforts have been made to improve the profile of VET pathways

While Portugal has traditionally favoured enrolment in general education programmes in upper secondary education, education officials have gradually improved the profile of VET programmes. In particular, the number of students opting to enrol in professional programmes has grown simultaneously with the expansion of VET courses in mainstream educational institutions. According to administrative data for 2016, the great majority of public upper secondary schools in the country (89%) offers some type of VET programme – a consistent pattern across regions. In fact, the proportion of schools offering VET is above 70% in every NUTS III region (Figure 3.11, Panel A). Mirroring this pattern, there is no region in the country in which the proportion of students enrolled in school-based VET programmes is lower than 20%, with regions such as the Algarve having enrolment rates over 35% (Figure 3.11, Panel B).

Figure 3.11. Geographic variation in provision of vocational education and training (VET)
By NUTS III region

Note: Figures present only school-aged, initial VET programmes and enrolment by school-aged secondary students (under 24 years of age).

Source of administrative boundaries: Direção-Geral do Território (2016), Official Administrative Maps of Portugal - Version 2016 [Carta Administrativa Oficial de Portugal - Versão 2016],

Source: DGEEC administrative data, 2015/16.

According to recent evidence from Portugal, the choice to enrol in VET depends on such factors as the pathway options available in schools, students’ previous achievement, as well as the guidance and supervision of students by psychology and orientation services (Serviços de Psicologia e Orientação - SPOs) at school. Importantly, peers, family and teachers also influence the choice of track (Vieira, Pappámikail and Resende, 2013[25]). Similarly to other OECD countries, students that opt to enrol in VET tracks generally come from relatively more disadvantaged socio-economic backgrounds and have lower chances of accessing higher education (Cruz and Mamede, 2015[26]; Henriques et al., forthcoming[27]). However, despite attracting students from lower socio-economic backgrounds, Cruz and Mamede (2015[26]) find that the expansion of VET pathways in Portugal has had significant positive impacts on the progression and outcomes of students in the labour market. In line with evidence from other countries, enrolment in the VET track has also been shown to increase the chances that low-performing students will remain enrolled in school compared to low-performing performing students who opted to enrol in the regular track (Henriques et al., forthcoming[27]). Stakeholders with whom the review team met repeatedly emphasised the expansion of vocational offer in schools as a good opportunity to attend the needs of students in risk of dropping out or willing to experience more school success in an alternative offer to the academic track.

Alongside the expansion of VET offer, graduation and completion rates have increased. The proportion of students graduating from VET programmes in Portugal increased from only 13% in 2005 to 56% in 2014 – the largest growth among OECD countries (Figure 3.12). Likewise, the percentage of students completing professional courses in a given year climbed from 55% in 2002 to 74% in 2016 (Figure 1.4.5 in DGEEC (2017, p. 50[28])). Across all VET programmes, 51% of students conclude their studies during their theoretical duration, above the 49% among those in general programmes in 2015. But despite the progress, the proportion of VET students completing their studies on time is still considerably lower than other education systems with available data (Figure A9.3 in OECD (2017, p. 158[9])).

Portuguese educational authorities have taken specific steps to increase the status of VET programming through a range of initiatives to match the VET offer to labour market needs. First, work-based learning has been gradually built into professional programmes, while VET programmes have preserved academic coursework requirements in a school setting. In 2013, a reformulation of VET upper secondary syllabi introduced changes aimed at providing more training hours in a work context, continued expansion of vocational pathways in lower and upper secondary aimed at students in risk of dropping out and further harmonisation of VET programmes across upper secondary schools and IEFP training centres (Ordinance 276/2013).

Second, Portuguese authorities set, in 2014, the legal framework for Professional Business Reference Schools (Escolas Profissionais de Referência Empresarial – EPRE). EPREs are intended to be created by companies or employers’ associations willing to develop courses directly related to their activity (Decree-Law 92/2014). According to the government’s intentions, 50% of the teaching hours in the EPREs – excluding work-based learning – would be provided by companies’ employees with adequate pedagogical qualifications. Nonetheless, since the publication of the legal framework, no Reference Schools have yet started to operate. While the design of these schools appears promising, and despite review team inquiries into this policy development, it is unclear why none have opened.

Figure 3.12. Change in graduation rates of VET students in upper secondary education, 2005 and 2014

1. Year of reference is 2013.

2. OECD average is calculated only for countries with available data for all reference years.

Source: OECD (2016), Education at a Glance 2016: OECD Indicators,, Table A2.4.

Third, Portugal has gradually developed a coherent national qualifications framework, which is data-driven and informed by consultation with labour market stakeholders. The procedure for ensuring the VET network has labour-market relevance is facilitated by a tradition of dialogue and stakeholder involvement in the discussions. In particular, vertical and horizontal co-ordination mechanisms have been put in place to steer the VET offer. The system for anticipation of skills, coupled with regulations on the limit of classes per field of study, help centrally steer the VET network of provision according to labour market needs. While vertical steering mechanisms can be effective tools to streamline vocational offer, the involvement of subcentral actors in the process – through the regional module of SANQ – incentivises local commitment and the sharing of common goals (OECD, 2018[19]). In turn, horizontal co-ordination can be leveraged through a network of almost 300 Qualifica Centres that help implement VET programmes at the school-level and guide young and adult students through the National Qualifications Catalogue (see also Chapter 1).

Fourth, Portuguese authorities have set higher quality standards for vocational programming. Alignment with the European Quality Assurance in Vocational Education and Training (EQAVET) system has provided an internationally agreed-upon framework for quality assurance of VET providers. An increasing number of schools are already included in this system, improving their esteem among students, families and companies. In particular, EQAVET certification is used as a criterion in decisions pertaining to opening new programmes or to extend the existing ones (Ministry of Education, 2018[1]).

Finally, all VET tracks lead to a double certification qualification in recognition of students’ mastery of both academic and vocational skill sets. Double certification holds the potential to foster transitions to academic pathways in post-secondary levels while remaining relevant for entering the labour market. Professional qualifications are organised into stackable modules that help facilitate the progression of students throughout the duration of their studies and to higher education levels. But challenges still remain regarding the fulfilment of such potential. A recent follow-up survey on the status of students one year after graduation has found that for those completing upper secondary education through VET tracks, only 34.1% were not employed but studying, 6.7% were both working and studying, while 18.9% were neither studying nor working (DGEEC, 2018[29]). Nevertheless, the number of VET students employed one year after graduation has been improving considerably in comparison to other waves of the survey (Ministry of Education, 2018[1]).

Extremely high rates of inclusion of students with Special Educational Needs (SEN)

Since the Revolution of 1974, inclusion and integration have been leading principles in the public sector and have a broad-based commitment from society, especially in education. This is an important explanation for the long history of prioritising inclusion of students with special educational needs (SEN) resulting in an extremely high level of school-level inclusion by international standards: 98.8% of SEN students were assigned to regular schools in 2016/17.

Despite efforts to identify all students who have special educational needs, specific support is primarily targeted to students with severe SEN – i.e. students with “multiple disabilities” (multideficiências), congenital deaf-blindness, deafness, blindness or autism spectrum disorders. Specific support strategies generally imply the deployment of additional resources – either human or specific ICT equipment – but also changes in pedagogical strategies (Decree-Law 3/2008 was in force at the time of writing). For students with severe SEN, a reduced class size maximum of 20 students applies.

Various strategies and supports exist to support teachers in successfully identifying and teaching students with SEN. Initial teacher preparation includes a required course for all teachers on how to support students with SEN. Box 3.2 describes the process for identifying students with SEN.

Box 3.2. Identification and intervention processes for students with SEN


Identification of SEN students is currently based on the World Health Organization’s International Classification of Functioning, Disability and Health (ICF). The ICF works as the framework for the labelling of the specific student need and the basis for delineating specific pedagogical interventions in the case of students with severe SEN. Students can be initially signalled as in need of an evaluation by any member of the school community, such as parents, teachers and non-teaching staff, or a doctor.

The signalled student is evaluated according to the ICF along four main dimensions: i) body functions (e.g. mental and sensorial); ii) body structure (e.g. physical impairments); iii) activities and participation (e.g. learning or socialisation difficulties); and iv) environmental factors (e.g. unsupportive home environment).

The process for evaluating the extent of the need is under the responsibility of the special education team within the school cluster, mainly composed of specialised SEN teachers and, often, other specialised personnel. The SEN team within the school cluster can also access health centres or specialised resource centres for inclusion (CRI).

A new law that would de-emphasise the medical diagnosis component of the identification process in favour of identifying students with SEN based on persistent struggles in school despite tiered interventions was being finalised at the time of the drafting of the report.


SEN teachers in the school and CRI staff determine the pedagogical intervention. It contains two main instruments: i) an Individualised Education Plan (IEP); and if relevant ii) an Individual Transition Plan. The IEP applies to all identified SEN students and sets the need of additional services to be offered to the student, the changes in educational goals or evaluation (if any) and the need to learn in a setting different than the regular class (if any). According to the law at the time of writing, the provision of individualised additional support is dependent on a full medical evaluation. The individual transition plan applies only to students with permanent disabilities and is a complementary support plan to facilitate the transition of the student to an independent life outside school.

Source: Decree-Law 3/2008.

Depending on the assessment of SEN teachers in schools, most students with special educational needs are fully integrated into regular classes, for which reductions in class size are allowed according to national regulations. On the other hand, students requiring greater care are educated in separate learning environments within the school, adapted to their functional disabilities and according to different curricular standards. Nonetheless, in mainstream public schools, students with severe SEN are frequently integrated into regular classes with the individual support of a SEN teacher. So too are students with more moderate disabilities, but without either a reduction in class size or much additional support.


Shifting demand for school places remains a challenge that needs a system-wide and differentiated strategy

A sophisticated strategy is required in order to meet the contradictory demand for school places in the near future. On the one hand, there is a high demand for school places. This is due to the current ambitious education policy goals of 100% access to ECEC for 3-5 year-olds and 90% participation in the full-time school day. The increasing demand for school places is also caused by intensification and expansion of upper secondary and adult education. Along with recent growing demand for more school places, longstanding shortages of school places in the densely populated suburbs of Lisbon is not resolved yet as 10% of the 1st cycle schools in that area still work in double shifts (CNE, 2017[2]).

On the other hand, there is a much lower demand for school places, stemming from the massive decline of school-age population (see Figure 1.6, Chapter 1). Further complicating the challenge is that this demand will be more scattered throughout the country as rural locations become more isolated and scarcely populated. A system-wide and differentiated strategy of providing school is needed to address this complex and varied demand for school places.

These contradictory developments in the demand for school places pose challenges to central and local authorities. Increasing challenges on planning capacity stem from complex, inter-related pressures including:

  • Pressures on schools whose operation and funding are gradually affected by decreasing enrolment levels due to poor performance in the rankings.

  • Greater challenges to allocate students across schools that are over- and under-capacity, in a context where enrolment patterns are heavily determined by where families live.

  • Physical capacity challenges are particularly concentrated in schools serving few socio-economically disadvantaged students (see Figure 3.13), which interact with family concerns around poor performing schools and produces situations in which some families become increasingly dissatisfied with their inability to access high-performing, advantaged schools and other families become increasingly dissatisfied with the low status of their poor-performing, disadvantaged schools. Figure 3.13 further highlights that the new student assignment criteria developed for the 2018/19 school year that prioritises school choice requests for students receiving social support are unlikely to substantially reduce school segregation. The most socio-economically advantaged schools are already over-subscribed. Since the newly introduced criterion comes last behind other criteria, there are unlikely to be places remaining in high-wealth community schools when non-resident students apply to attend them.

  • Pressures to better align incentives pertaining to enrolment in private and public schools. For instance, the regional variation in the proportion of students enrolled in private schools (see Figure 3.3), in parallel with the regional variation in over-capacity schools (especially in the Lisbon Metropolitan Area), will force concerned policy makers to face trade-offs between increasing enrolment capacity of public schools, or to find more mechanisms to direct demand to (potentially government-dependent) private providers.

  • The challenge to best balance between the democratic principle of responding to parents’ choice preferences for enrolment in particular schools, with a highly centralised governance system in education.

Challenges with the overall quality of school infrastructure

Portugal spends at or near the bottom of all OECD countries on the construction and maintenance of school facilities. As Chapter 2 notes, municipalities have responsibility for building and equipment maintenance for ECEC and the first cycle of basic education, whereas the maintenance and renovation of infrastructure for the remaining cycles of basic education and upper secondary (with the exception of Parque Escolar schools) are the responsibility of the Directorate General for Schools (Direção-Geral dos Estabelecimentos Escolares – DGEstE). Due in part to tight budgets over the past years, the investment in renovating and improving school buildings and equipment at all levels of education has been low. In 2014, Portugal spent only 3.1% of its total public education expenditure on capital expenditure, ahead of only the United Kingdom in the OECD (OECD average: 8.4%) (Figure B6.1 in OECD (2017[9])). Additionally, Portugal spends the smallest proportion of all current expenditures on items other than staff compensation, which would include facility maintenance. Portugal devotes only 5% of its current expenditures at the primary and 8% at the secondary level to non-compensation areas, compared to OECD averages of 22% and 21% respectively (Figure B6.2 in OECD (2017[9])).

Figure 3.13. Proportion of over-subscribed schools by quartile of disadvantage, 2015/16

Note: The figure shows the distribution of the proportion of over-subscribed schools by quartile of an index of socio-economic disadvantage. Over-subscribed schools are those in which the ratio of defined classes to classrooms is greater than 1.05, implying too many classes for the available physical space. The index of socio-economic disadvantage is measured by rank percentile-ordering the proportion of students within a school receiving School Social Assistance A (ASE A) and average years of maternal education at the school level. The average of the 2 ranks was then demeaned (mean = 0) and assigned a standard deviation of 1. High values of the index indicate high levels of socio-economic challenge.

Source: DGEEC administrative data, 2015/16.

PISA 2015 results reflect that Portuguese school buildings and physical infrastructure are in poor condition. Portuguese principals representing 15% of 15-year-old students report that the schools’ capacity to provide a sound learning environment is hindered by a lack of physical infrastructure – e.g. building, grounds, heating/cooling, lighting and acoustic systems (OECD, 2016, p. 388[13]). At the time of the review visit in January, the team observed several schools struggling to maintain facilities warm enough for students and teachers. Students and teachers alike attended classes wearing winter coats and sometimes hats and gloves. School administrators shared with the review team that the school either did not have a heating system or that they could only afford minimal utility costs, and so kept the heat low out of necessity. Efforts exist to record the state of facilities and prioritise intervention needs, including the removal of asbestos, yet these have not yet resulted in consistently high-quality school buildings.

ICT resources and services for the general operation of public schools – computers, internet, networking equipment, cloud services, access control system and video surveillance system – are acquired and managed centrally by the DGEEC. In 2007 a major Technological Plan for Education was launched by the central government, aiming at a ratio of 2 students per computer, the availability of high-speed broadband and 90% of teachers being qualified in ICT in all public schools across the country. However, in 2015 PISA surveys of principals recorded an average of only 0.43 computers per student, 5th worst in the OECD (Table II.6.4 in OECD (2016[14])).

The poor state of educational infrastructure conflicts with the ambitious policy goals to expand and intensify schooling, and to modernise and innovate teaching and learning processes. Offering more students full-time schooling from ECEC to upper secondary education and VET requires sufficient school buildings equipped with appropriate classrooms and spaces for teaching and learning, and facilities such as school cafeterias, toilets, outdoor space and playgrounds of at least basic quality. Students and staff in remote locales are hindered in their ability to participate in distance learning and staff collaboration due to the limited ICT resources. Opportunities for innovative learning strategies, such as blended learning, are limited by the lack of modern digital learning environments and appropriate hardware and software in all schools.

The articulated goals of decentralisation in Portugal do not align with promoting school autonomy and equity in student outcomes

Portugal is in the midst of a clearly outlined comprehensive effort to decentralise. As a result, municipalities will have more discretionary powers to shift non-teaching staff over different school clusters, to influence the state of school buildings and equipment, and to play a greater role in developing extracurricular activities. These operational responsibilities tend to be the most typical kinds of decentralised responsibilities for which municipalities assume responsibility.

Alongside these decentralisation efforts that allocate powers and responsibilities across governmental levels, Portugal is also undertaking efforts to promote school autonomy. However, the kind of school autonomy being favoured in Portugal in practice is fairly circumscribed with respect to educational responsibilities. School autonomy is largely limited to tailoring a part of the national curriculum to the specific needs and interests of its students but does not include broader curricular and pedagogical autonomy or other types of school autonomy such as local responsibility for financial or human resources. At the same time, the major process of school consolidation has created more operational responsibilities for school cluster leaders: sometimes managing up to 30 schools, responding to facility problems, engaging with families and the community, supervising greater numbers of students and so on.

Together, these patterns create a risk that both municipal and school-level actors will understand their key autonomies to be related to the operational and management side of educational endeavours. As local levels of government (municipalities and parishes) receive additional authorities over schools, they may hold these responsibilities closely, inhibiting school-level decision-making. Power plays and conflicts in responsibility in education may arise at the local level, with different local authorities crowding out each other’s autonomy.

The decentralisation processes in education may also lead to undesired effects with respect to equity in education. During the site visit, different stakeholders expressed to the review team that, despite the broad political and societal support for decentralisation in Portugal, there is a lack of confidence in the governance and management capacities of municipalities, particularly with respect to safeguarding values such as neutrality, sense of due processes, equity and efficiency. Evidence exists that decentralisation damages coherence and exacerbates inequities (Dafflon, 2015[30]; Bullock and Thomas, 1997[31]). Locally fragmented and scattered education policies and decisions may cause losses of coherence and knowledge, which are needed to improve and innovate education. Equity in education can be undermined due to different capacity across the country (Dafflon, 2015[30]). Particularly in Portugal, where strong regional and school-level variation in student outcomes give rise to substantial equity concerns, and where educational disadvantage also is of geographical character, the risk that decentralisation perpetuates, and even reinforces, existing inequity in education is real. Moreover, the traditional weak investments in education from the private sector, civil society and local communities in Portugal (Ministry of Education, 2018[1]) present an extra challenge to the risks of decentralisation for equity in education.

Challenges remain with the school clustering process

Despite the impressive accomplishment of reducing the number of schools and clustering schools with little public outcry (see below), the process of school clustering has highlighted the importance and challenges of greater school autonomy in practice. The large and complex school clusters require an appropriate level of educational governance and leadership capacities, including stronger and more professional school administration than before. This is challenging for small groups of minimally trained school administrators.

A particular challenge the review team noted during their school visits is the large geographic stretch of some clusters. This results in difficult working conditions for teachers and staff as they have to travel across the cluster during the school day. If the school cluster has remote, isolated and poorly accessible schools, and public transport is also limited, travel is difficult and this becomes a threat to the offer of high-quality education. Mega-clusters with a large number of schools at different education levels impose a challenging governance and leadership task for the teams of school leaders and principals, requiring strong and appropriate governance and educational leadership capacities at the cluster level, which is a condition that is not often met.

Further unification and integration is still a challenge in many newly-established school clusters. There are limited system-wide interventions or measures to promote cross-school collaboration within clusters. The lack of cross-school collaboration in many clusters results in the uneven quality of collaboration, planning and joint practice around student trajectories across schools. To be able to take advantage of school autonomies and leverage capacity across many schools, the large and complex school clusters require an appropriate level of educational governance and leadership capacities. There were some cases during school visits in which the review team observed strong practices of local school autonomy. For instance, one school cluster had developed a combination of psychosocial and curricular interventions intended to fight dropout and school failure. It had also developed a relationship with a local higher education institution to provide strategy and professional development support. However, these examples of innovative practices were not observed uniformly across schools.

The potential for further unification and integration of school clusters is dependent on the nature of decentralisation processes and the promotion of school autonomy in Portugal. The current decentralisation measures in education are primarily of an operational nature, focusing on the administrative management of schools, rather than on core educational improvement efforts. Therefore, the influence of municipalities on the local core processes of education remains indirect and limited, and it seems unlikely that municipalities’ policies and decisions will directly touch on the primary teaching and learning processes of schooling. Under the current decentralisation conditions, it is thus not very likely that municipalities can engage in the core educational enterprise of schooling at the local level. Moreover, the role of municipalities in education is strongly counterbalanced by the central government that retains responsibility for core matters such as hiring, placement and pay of teachers, responsibility for the curriculum, examinations, selection and tracking policies, and planning of the school network under its jurisdiction.

The challenge with respect to further unification and empowerment of school clusters is to remove the impediments at the school cluster level to design and implement an integrated educational strategy, directed at the quality of teaching and learning processes and educational improvement and innovation. This will require meaningful school-level autonomies, leadership capacity development and responsive, agile municipality actors with clearly distinct responsibilities from school decision-makers.

High levels of variation by background and region in student outcomes

The strong individual, school-level and regional variation in student outcomes in Portugal is suggestive of substantial equity concerns in school quality and effectiveness. Variations in student performance, rates of year repetition and student dropout rates are persistent challenges. As noted in Chapter 1, Portuguese students’ performance on international and national exams is tightly linked to their socio-economic background and the extent to which their repetition rates are associated with socio-economic background and gender are well above the OECD average.

Portuguese 15-year-olds with a higher socio-economic status are only 0.65 times as likely to have repeated a year than students from lower socio-economic backgrounds (OECD, 2016[14]). Importantly, the strong influence of students’ backgrounds in their likelihood of repetition is true even when students’ academic performance on PISA is taken into account. This implies that even when students have the same level of academic skills, they are much more likely to have repeated a year if they are from a socio-economically disadvantaged background. The influence of background on year repetition is significantly higher in only two other OECD countries.

In addition to the common reasons why poverty and the concentration of poverty in a classroom affect student performance, the extensive use of private tutoring in Portugal exacerbates gaps between haves and have-nots. Significant evidence exists that students from low socio-economic backgrounds are exposed to more limited vocabulary as young children (Hart and Risley, 1995[32]), receive fewer educational resources (OECD, 2017[33]) and experience more frequent trauma (Souers and Hall, 2016[34]) than youngsters from more advantaged backgrounds. Additionally, concentrations of poverty in a community or school deprive students of networks with high social capital (Bayer, Ross and Topa, 2008[35]) and expose students to weak or caustic learning environments (Epple, Newlon and Romano, 2002[36]; Sacerdote, 2011[37]; Duncan and Magnuson, 2011[38]; Carrell and Hoekstra, 2010[39]). In Portugal, around 40% of upper secondary students in the general pathway participate in private tutoring, paying up to EUR 200 per month for this support (Costa, Ventura and Neto-Mendes (2003) and Costa et al. (2007) as cited in Ministry of Education (2018[1])), which further exacerbates socio-economic performance gaps.

In addition to socio-economic gaps, it is also more likely for male students to be retained or drop out from schools than female students at all education levels. Just over a fifth of male students in upper secondary education were retained or dropped out from school during as of 2015/16, while only 15% of females have done so (CNE, 2017, pp. 166-175[2]). PISA 2015 estimates suggest gender influences repetition more heavily than in all but one other OECD country (Table II.5.13 in OECD (2016[14])).

Additionally, some immigrant student groups in Portugal experience consistently worse outcomes than native students, though this pattern varies considerably both within and across immigrant groups. Portugal is a particular case concerning the socio-economic status of students with an immigrant background and their school success, running counter to the general patterns observed in other OECD countries. The OECD report on the Resilience of Students with an Immigrant Background (2018[40]) identifies several unique characteristics of the Portuguese immigrant school population:

“First, both second-generation immigrants (i.e. those born in Portugal with two foreign-born parents) and native-born students of mixed heritage (i.e. those born in Portugal with one foreign-born parent and one parent who was born in Portugal) are more advantaged, on average, than native students. Second, differences in socio-economic status between native students and first-generation immigrant students (i.e. foreign-born students whose parents are also foreign-born) are not statistically significant. Third, foreign-born students with at least one parent who was born in Portugal are more disadvantaged than native students” (OECD, 2018, p. 160[40]).

A number of high-skill, high-maternal-education Eastern European immigrants since the 1990s have contributed to another unique characteristic of Portuguese immigrants: there are no significant differences in the socio-economic status of native students and first-generation immigrant students. Nonetheless, Portuguese immigrant 15-year-olds underperform their native-born peers and the extent to which this is explained by their socio-economic status is below the OECD and EU average (Table 6.2 in OECD (2018[40])). As Figure 4.1 of the Resilience of Students with an Immigrant Background report highlights, students’ country of origin plays an important role in their likelihood of attaining baseline academic proficiency. Unfortunately, too few immigrant students were sampled to report results for immigrants to Portugal. However, Cape Verdean first-generation immigrants to Luxembourg are 34 percentage points less likely to achieve baseline academic proficiency than French first-generation immigrants. Thus, it would be helpful for Portugal to explore the extent to which students from countries representing some of their largest immigrant groups – Brazil (over 12 000 students), Cape Verde (over 6 000 students) and the remainder of the African Portuguese-Speaking Countries (Países Africanos de Lingua Oficial Portuguesa – PALOPs) (around 14 000 students) – are experiencing success in school (CNE, 2017[2]).

In addition to the impacts of immigrant status, Portugal struggles to successfully promote the success of its Roma students. Over 48% of Roma students had repeated a year in 2017, and 28.4% had repeated a year more than once. Further, Roma students have annual dropout rates of 5.9%, leaving school before the end of compulsory education in rates substantially above system-wide averages (DGEEC, 2018[41]). An important caveat to the data presented here is that the questionnaire that produced the preceding data relied on a non-representative set of schools’ self-reports (69.6% of all schools), but nevertheless captures some of the challenges facing the Roma community that the Portuguese school system has not yet sufficiently addressed.

In Portugal, educational performance is not only related to socio-economic status, gender, and ethnicity or immigrant background but it is also of geographical character (Abrantes, 2013[42]; Marôco, 2017[43]). Figure 1.13 in Chapter 1 highlights regional and school-level differences in 3rd cycle school repetition rates, a pattern that is similar across levels of education. Poverty and deprivation related to geography appear persistent and hard to fight. Low community expectations, negative attitudes toward school and low trust in public service delivery impact students’ motivation and expectations towards school and future professional status, which in turn leads to low academic performance, absenteeism, year repetition and school dropout (Abrantes, 2013[42]). Substantial variations in student achievement across regions and schools also exist. Chapter 1 highlights some of these regional performance differences on international and national assessments (see Figures 1.14 and 1.15 and Table 1.E.1).

Beyond geographic performance constraints, some students find their opportunities limited because of the classroom and academic track in which they study. As noted above Portugal has no system-wide criteria for class formation. Anecdotally during the review visit, stakeholders expressed concerns that classes in basic education were unofficially levelled with all of the strongest students grouped together in a single class. Some attributed this to the powerful voice that high-social-capital families had in directing choices made at the school. The review team is unaware of any study currently underway examining class-based sorting and the impact of peer effects on students, but the Ministry of Education Information System (MISI) data should permit just this sort of analysis. Additionally, stakeholders in some schools visited during the review reported that students in the vocational track in secondary schools did not have access to Portuguese as a Second Language (PNML) courses, despite the fact that a large proportion of non-native Portuguese speakers enrolled in the VET pathway. Finally, the review team learned during their school visits that the lack of fluidity across tracks in upper secondary education incentivises year repetition. The development of transferable skills is especially hindered by the fact that students who begin one specialisation in the sciences-humanities strand and seek to transfer into a different strand receive credit for only some of the courses they have already taken and must repeat different ones to progress towards their diploma, though the curriculum flexibility project may allow secondary students to substitute subjects from one strand into another.

The administration and provision of VET programmes is still fragmented

The public system of VET is offered by different networks of providers, raising some efficiency concerns. Two almost parallel systems – those of IEFP training centres and upper secondary VET – often fulfil overlapping functions. Such organisation leads to similar qualification profiles, with varying intensity of work-based learning and quality of training. The separate governance of these networks, often operating under different regulatory frameworks and overseen by different ministries, hinders the ability to plan the courses that are offered, to decide which courses are to be offered by each provider and to plan the consolidation of provision that leads to low prospects of employment. The Ministry of Education is responsible for regulating the offer of professional programmes in schools. On the other hand, the Ministry of Labour, Solidarity and Social Security co-ordinates, through IEFP, the delivery of apprenticeship courses in certified training centres. Additionally, a network of professional schools under the direction of the Ministry of the Economy offers upper secondary and post-secondary VET. While the continued integration of the vocational offer in comprehensive schools allows for sharing resources across general and vocational tracks, further opportunities remain for VET students in schools to take advantage of the relatively well-resourced network of IEFP training centres (Araújo, 2017[44]). The type of apprenticeship programmes offered through these centres – by requiring a formal agreement between the student and an employer – are more dual in nature and provide opportunities to further engage the private sector. Stronger apprenticeship schemes enable employers to have access to a pool of skilled and potential future employees, particularly in labour markets with strong barriers to entry, such as Portugal (OECD, 2017[33]).

The rationalisation of VET programmes for greater relevance requires effective monitoring and evaluation mechanisms. Portugal has taken considerable steps to develop an evidence-based methodology to steer the provision of VET programmes, particularly through the introduction of a system for the anticipation of qualification needs (SANQ and as described above). However, the effectiveness of SANQ is still hindered by limited commitment at the political level and lack of human resources with relevant knowledge and experience (OECD, 2018[45]). Data analysis can be further deepened and nuanced. In particular, the full potential of data is insufficiently explored, given the availability individual-level information in upper-secondary education and labour market outcomes. Although current employment data helps regulate the VET offer through the description of employability prospects by field of study, there is no estimation of levels of student demand or labour market outcome for various VET programmes. Insufficient integration of information on students’ profiles – through currently existing student-level datasets – in the qualifications assessment exercise limits the ability to anticipate the demand for VET courses and to tailor outreach to specific student groups.

Furthermore, local level horizontal co-ordination is hampered by unequal and low levels of participation by inter-municipal communities (CIM). In fact, at the time of the review, few CIM (4 out of 23) were actively involved in the regional co-ordination module of SANQ. Such uneven commitment of the subcentral level in the definition of priorities raises both equity concerns with respect to capacity and efficiency concerns about the ability to steer the network and share resources at a regional level. Some stakeholders with whom the review spoke with also reported concerns about the low frequency of meetings aimed at defining strategic guidelines for the sector. Nevertheless, some improvements are already apparent, with an additional nine inter-municipal communities expressing commitment to participate in the regional co-ordination module. Finally, despite the online presence of the national VET agency (ANQEP) and the publication of SANQ’s results, there is no concerted effort for their widespread dissemination. Skills’ anticipation exercises in Portugal often directly feed into specific training policies or public employment programmes, without transparent communication of its outcomes (OECD, 2018[46]).

Too few resources are invested in support of students with moderate Special Educational Needs (SEN)

Despite high rates of inclusion, the quality of the education students with SEN receive is hampered by insufficient and inadequate resource investment and teaching skill gaps. Particularly SEN students with moderate disabilities risk being overlooked as there are insufficient resources available to effectively include them. During their initial training, mainstream content teachers are provided with some training and coursework on the academic needs of students with SEN (one course). However, there are no ongoing professional development requirements for teachers to continue to build their skills in teaching students with SEN. Furthermore, during the review team visit, multiple stakeholders reported that minimal supports existed for students with moderate special educational needs, primarily due to the fact that SEN teachers were overburdened with full caseloads and responsibilities for teaching classes of students with severe disabilities. Stakeholders reported frequent failures to implement and monitor the Individual Educational Plans (IEPs) of students with moderate disabilities. Furthermore, class size maximums are lower if a single student with severe SEN is present in a class, but they are not any lower if there are multiple students with severe SEN or only students with mild- or moderate-SEN needs.

Figure 3.6 highlights the increasing rates of SEN identification in Portugal; however, identification rates continue to exhibit cross-regional and school variation. Figure 3.14 presents regional and school-level variation in the rates of identified students with SEN. While between-region variation in average rates of SEN students only differ by about 4 percentage points, within a region some schools have substantially more students with SEN than others. In 17 of 23 continental NUTS III regions, schools that enrol more students with SEN than 90% of other schools in their region have at least 8 percentage points more students with SEN than schools that enrol fewer students with SEN than 90% of other schools in their region. These widely varying rates of SEN identification bear further study to understand whether they result from programmatic placement, regional variation in rates of incidence or inconsistent application of evaluation standards. If some students are identified as having SEN because schools have simply not successfully taught them, this robs resources from other students who truly need them (OECD, 2018[19]).

Figure 3.14. Between- and within-region variation in rates of special education identification
By NUTS III region

Note: The interdecile range is the difference between the Special Educational Needs (SEN) enrolment rate for schools in the 90th percentile for SEN enrolment within the region and the SEN enrolment rate for schools in the 10th percentile. For example, in the Metropolitan Area of Lisbon (AML), schools in the 90th percentile for SEN enrolment, have SEN enrolment rates of 10.7%. Schools in the 10th percentile have SEN enrolment rates of 1.9%. Thus, the interdecile range for SEN enrolment in AML is 8.8 percentage points.

Source of administrative boundaries: Direção-Geral do Território (2016), Official Administrative Maps of Portugal - Version 2016 [Carta Administrativa Oficial de Portugal - Versão 2016],

Source: DGEEC administrative data, 2015/16.

Furthermore, while the rapid increase in identification of students with special educational needs has been supplemented by an upsurge in the number of certified SEN teachers working in schools, the numbers of special education teachers have been growing at lower rates than the number of identified students. In the 2010/11 school year, there were roughly nine students with special educational needs for each SEN teacher. In the 2016/17 school year, there were roughly 11.4 students with special educational needs for every SEN teacher or an increase of over 2 students per teacher. Crucially, most of the newly hired special education teachers were temporary and part-time, rather than permanent, SEN teachers (DGEEC, 2017[47]).

Policy recommendations

Integrate decentralisation in education and promotion of school autonomy into a comprehensive strategy for effective governance of the education system

The currently proposed decentralisation and school autonomy strategies in Portugal are relatively narrow in scope. The decentralisation measures are primarily of an operational and administrative nature, focusing on more efficient subcentral administration of buildings, equipment, and hiring and allocating non-teaching staff, rather than on core educational improvement efforts. Portugal must review its priorities around decentralisation and school autonomy and determine whether these established priorities are likely to accelerate school improvement outcomes.

The central argument in favour of decentralisation is that it can make government more accountable and responsive to the governed (Faguet, 2014[48]; Bullock and Thomas, 1997[31]). Following this argument, decentralisation in the Portuguese educational system has the potential to improve government service delivery and to reduce inefficiency and bureaucracy as municipalities have better insight in and are more closely linked to schools and local communities. Municipalities and parishes may reduce waiting time, allocate concerned resources more efficiently and effectively, improve selection and hiring processes of non-teaching staff, and maintain and renovate school infrastructure more appropriately. Decentralisation in education may also address specific geographical educational issues and needs, implemented collaboratively by municipalities, schools and community stakeholders.

However, this general assumption that by placing government closer to people, decentralisation fosters greater responsiveness of policy makers to the tastes of individual jurisdictions must be nuanced (Dafflon, 2015[30]). The smaller the local authority, the greater the risk that unrepresentative groups dominate and control local policy- and decision-making to their own benefit. Moreover, issues of coherence, equity and governmental capacities problematise the decentralisation processes.

Decentralisation in education involves much more than delegating responsibilities to municipalities and parishes by regulating legal powers properly and providing appropriate contractual frameworks. Portugal should rather perceive decentralisation in education as shared governance including a balancing act of promoting autonomy in education at the local level and fostering responsiveness to local diversity, and at the same time guaranteeing system-wide education quality, accessibility and equity. Like in many OECD systems, this is a delicate equilibrium that is difficult to achieve given the complexity of the Portuguese education system (Burns, 2016[49]). Portugal should consider integrating its current decentralisation processes in education into a comprehensive strategy for effective governance and embrace the challenge to shift its current legalistic approach with a focus on governance structures to a more systemic approach focusing on processes and governance culture.

Portugal could explore various alternative governance structures. For instance, the education system could take advantage of its strong central decision-making powers, retain control of educational issues at the central level and work to promote system-wide steering strategies with a particular emphasis on promoting equity. The substantial equity concerns in Portugal represent a strong argument to keep the responsibility for the curriculum, examinations, selection and tracking policies, and planning of the school network under the jurisdiction of the central government. In order to mitigate the risks of inequality of opportunity in education and fight school failure, Portugal could improve its central steering policies by integrating and aligning them better and improve policy making by means of a robust knowledge infrastructure and sound policy evaluations (see recommendations in Chapter 2). The possibility to allocate higher skilled and/or more experienced teachers, specialised to serve struggling students, to under-resourced communities (see recommendations in Chapter 4) could justify keeping the responsibility of selection, hiring and pay of teachers under the jurisdiction of central government. In addition, Portugal might consider re-centralising and bringing the selection, hiring and pay of non-teaching staff with instructional and socio-emotional responsibilities (teaching assistants, psychologists, laboratory assistants, etc.) back under the jurisdiction of central government, because instructional and student support staff matter for good quality education and improvement and innovation, especially if they collaborate and jointly focus on core educational goals in practice. In parallel, Portugal could prioritise school cluster autonomy over budget, pedagogical innovation and strategic planning. The nature and degree of school cluster autonomy could be aligned with, but not overlapping, with the powers of central government in education.

On the other hand, Portugal could devolve most operational and instructional responsibilities to the municipal and school levels. It would need to articulate a clear division of responsibilities between actors such that municipalities and/or parishes were not tempted to infringe on the educational autonomies of schools. One sensible division would be to assign municipalities responsibility for all operational matters, including non-teaching staff responsible for operational management (security, cleaning, food services, etc.). Schools would then be granted further control over all resources (financial and human) which contribute directly towards student learning and development. The central government role would be to support municipalities and schools with capacity building efforts, with a particular eye towards assisting schools and communities in which weak governance and leadership skills had been identified. Other blends of such approaches could potentially be successful.

Key to the success of decentralisation will be that all governance structures are aligned. So far, this kind of alignment is at odds with the current decentralisation and autonomy approaches which appear muddled and conflictual.

A strategy in line with such a systemic approach should include four elements (Burns, 2016, p. 229[49]), described below and applied to the case of Portugal:

  1. 1. Stakeholder involvement and ownership of agreed goals and principles.

As Portugal is in the midst of a double process of delegation of responsibility to subcentral authorities and the enhancement of school-level autonomy, this first element of involving different stakeholders seems particularly relevant (see Box 3.3). The traditional weak investment in education from the private sector, civil society and the local communities (described above) requires special efforts to involve external local stakeholders. At the same time, involving internal stakeholders such as school principals and deputies is critical.

  1. 2. A whole-of-system vision that keeps the focus on processes and is flexible enough to adapt to change and unexpected events.

This second element is of importance for Portugal due to the challenges associated with the shifting demand for school places and its contradictory trends (see above).

  1. 3. Alignment of roles and responsibilities across the system as well as a way to address any potential conflict or overlap.

This is critical given the discussion in the preceding paragraphs.

  1. 4. The ability to identify needs and develop capacity in a realistic and timely manner, based on a system vision and informed by research evidence.

This last element is also highly relevant to Portugal, as adequate and sufficient levels of educational governance and leadership capacities are still lacking at both the municipal level and the school cluster level. Moreover, a firm knowledge infrastructure including data and policy-evaluation, accessible to support administrators and professionals at the local levels and school clusters is lacking. Knowledge plays an important role in capacity building: directly, as input, instruments or resources as well as indirectly, the “knowledge capital” of actors in and around (Burns, 2016, p. 34[49]) (see also Chapter 2).

Leverage school autonomy to improve the educational capacity of school clusters

Whether decentralisation and school autonomy are meaningfully realised and how they become linked to educational strategies for school improvement and lifting student achievement largely depend on the educational governance and leadership capacities in school clusters.

The review team observed great variation in leadership capacities within school clusters. At some school clusters, the principal, school leaders, co-ordinators and others involved in school governance develop a clear vision on teaching and learning and an overarching strategy for achieving quality education and school success in collaboration and consultation with teaching and non-teaching staff, students and parents. Importantly, they appear to be able to implement this vision. At these school clusters, the review team noticed maximal deployment of granted autonomy and decision-making powers and an attitude of social entrepreneurship. In contrast, other school clusters demonstrated a rather bureaucratic and legalistic attitude towards school governance and leadership. Here, principals, school leaders, co-ordinators and others involved in school governance operated merely according to the legal framework but did not use the available autonomies effectively. At these school clusters, legislation is often interpreted in terms of limitations instead of opportunities, capacity to make use of curricular and pedagogical autonomy is lacking and governance and leadership do not constitute a lever for quality education. Chapter 4 expands on these observations with system-wide data and a focus on the development of individual professional capacity. In addition to the need to invest in generalised leadership competencies outlined in Chapter 4, this section highlights the structures necessary to promote school autonomy, which requires particular leadership traits and supports.

Box 3.3. Stakeholder involvement in education

Like in many other OECD countries, government actors in Portugal are not necessarily at the centre in educational policy- and decision-making anymore. Processes of decentralisation and promoting school autonomy make different internal and external stakeholders of schools more important, turning education from a “state-government enterprise” into multi-actor governance processes. Many actors such as the local communities, media, parents, researchers, school boards, NGOs, training providers, etc. all are potentially relevant stakeholders to school clusters.

Figure 3.15. Potential stakeholders in education

Source: Burns, T. (2016), Education Governance in Action: Lessons from Case Studies, OECD Publishing, Paris.

Box 3.4 displays characteristics of high-performing school districts in Ontario. Comparing school districts with school clusters, the example of Canadian school districts provides valuable insights for building educational governance and leadership capacity in Portuguese school clusters.

Box 3.4. Taking advantage of autonomy in Ontario, Canada

Syntheses of evidence about school system conditions contributing to improved student learning in the Canadian state of Ontario have led to the identification of actions that maximise the benefits of local control to generate system-level improvements.

The 9 practices of successful local autonomy

  1. 1. Establish broadly shared mission, vision and goals.

  2. 2. Provide coherent instructional guidance.

  3. 3. Build capacities and skill to seek out and use multiple sources of evidence to inform decisions.

  4. 4. Create learning-oriented organisational improvement processes.

  5. 5. Provide job-embedded professional development.

  6. 6. Align budgets, personnel policies/procedures and uses of time with mission, vision and goals.

  7. 7. Use a comprehensive performance management system for leadership development.

  8. 8. Advocate for and support a policy-governance approach to local governing boards (equivalent to Portuguese General Councils).

  9. 9. Nurture productive working relationships with staff and stakeholders.

Source: Leithwood, K. and C. McCullough (2016), Leading High-Performing School Districts, Education Canada.

Continue efforts to rationalise the school network, while preserving support for unique schools meeting the needs of under-served populations

Portugal must address the heterogeneous needs for school places in the near future and formulate a sophisticated strategy, meeting the strong demand for new school places in some locations and anticipating a much lower and more scattered demand for school places in others. A system-wide and differentiated strategy to provide schooling places is needed to address this complex and varied demand. While rationalising the school network, Portugal should consider additional investment in schools providing high-quality schooling for unique populations, particularly because many under-served populations in Portugal live in poor, remote areas. For instance, the process for de-funding government-dependent private schools takes into account the location and capacity of public schools in the region, the school infrastructure and transportation services but, crucially, does not consider specialised need or school performance. Therefore, Portugal must shift away from a unified, one-size-fits-all approach and take a diversified, tailor-made and situational approach in the provision of schooling options. In particular, the consolidation process, which disproportionately affects remote, poor and under-served areas, should consider multiple criteria for the decision of closing down or de-funding schools that meet particular needs.

Relatedly, Portugal should consider setting the class size minima for opening a class higher than the class size minima for closing a class. The current regulations set these at the same level, which creates inherent instability if students transfer during the year. Once a class has been opened, it should remain so unless the enrolment drops so low as to make it educationally unsustainable.

Tackle between-school segregation through complementary policies

Between-school segregation in Portugal is persistent and substantial. Worldwide, structural factors such as neighbourhood segregation and long-standing and/or informal school admission policies appear to be main drivers of segregation of students between schools. Between-school segregation mirrors persistent and ingrained social phenomena, which require innovative policies and measures to tackle them. As discussed, new from 2018/19, the official residence (rather than the declared one), and receipt of income support, will be taken into account to assign students to schools in Portugal.

Multiple strategies are employed internationally to promote integration and inclusion in education systems, with varying degrees of success and sustainability. Examples include: re-drawing of school catchment areas to include more diversity in the neighbourhoods they encompass, revising school assignment policies to deprioritise proximity of residence to school as the dominant factor in placement decisions, offering increased choice to families between public schools with complementary educational and informational campaigns to empower all families to make best choices, establishing controlled choice systems that offer expanded choice only when selecting a school other than students “home school” will contribute to increasing between-school diversity, or the siting of schools with unique profiles (e.g. “magnet schools” offering unique curriculum, arts focus, or other specialised offer) in low-income neighbourhoods (OECD, 2018[19]). Box 3.5 contrasts the first two approaches, one of which attempts to ensure integrated settings, while the other attempts to preserve choice while promoting integration. While considering whether Portugal can learn from these policies and measures and potentially adopt them, it should always be kept in mind that “good policies travel badly” (Harris, 2012[50]). Careful and accurate translation to the Portuguese context is essential.

Box 3.5. Student assignment policies to reduce segregation

Many school systems struggle with problems of segregation, both between- and within-schools, along various dimensions: socio-economic, language, immigrant status, race, ethnicity and more. Most OECD school systems use place of residence as the primary mechanism by which to assign the majority of students to schools. Given high rates of residential segregation, systems interested in breaking patterns of school segregation must explore strategies that deprioritise place of residence. Two primary models in existence are: i) to mandatorily assign students to a location remote from their residence; or ii) to allow families to choose schools other than the closest one, but to constrain that choice in a manner that promotes integration. Each approach has strengths and weaknesses.

Mandated, non-residential assignment

Mandatory assignment to schools other than ones closest to students’ home first came to prominence in the United States as a response to government-sanctioned segregation of black students (Swann v. Charlotte Mecklenburg Bd. of Ed. (1971) 401 U.S. 1.). Court orders required students to be bussed to schools outside their neighbourhoods. In some cases, predominantly black and white schools were paired and students attended the first years of an educational level at one school and then transitioned together to the paired school, ensuring both communities would bear portions of the commuting burden. These desegregation efforts resulted in improved academic and life outcomes for black students (Guryan, 2004[51]; Johnson, 2015[52]), but were met with intense political resistance and faded out in the 1990s and 2000s, resulting in increased segregation (Reardon et al., 2012[53]) and declining academic outcomes (Liebowitz, 2018[54]).

Controlled choice

In other contexts, particularly where a rich tradition of private schooling and family choice exist, policy makers who recognise that segregation can be exacerbated in the context of school choice, have instituted strategies to preserve elements of choice while promoting integration. They have constrained school choice such that families whose school choice decisions improve integration are prioritised over others’ choices. In the Flemish Community of Belgium, officials cognizant of the segregating effects of choice instituted new student assignment rules in 2013 that gave priority to certain places in oversubscribed schools to both disadvantaged and non-disadvantaged schools, in proportion to the socio-economic composition of each school’s neighbourhood (Nusche et al., 2015[55]). Preliminary evidence indicates this approach reduced segregation (OECD, 2015[56]), but not as much as mandatory assignment policies.

Sources: Swann v. Charlotte Mecklenburg Board of Education. (1971). 401 U.S. 1.; Guryan, J. (2004), “Desegregation and black dropout rates”, American Economic Review, Vol. 94/4,, pp. 919-943; Johnson, R. (2015), “Long-run impacts of school desegregation and school quality on adult attainments”, NBER Working Paper, No. 16664,; Reardon, S. et al. (2012), “Brown Fades: The End of Court-Ordered School Desegregation and the Resegregation of American Public Schools”, Journal of Policy Analysis and Management, Vol. 31/4, pp. 876-904,; Liebowitz, D. (2018), “Ending to what end? The impact of the termination of court-desegregation orders on residential segregation and school dropout rates”, Educational Evaluation and Policy Analysis, Vol. 40/1, pp. 103-128,; Nusche, D. et al. (2015), OECD Reviews of School Resources: Flemish Community of Belgium 2015, OECD Publishing, Paris,

Focus on improving the quality of learning for struggling students, rather than simply adding more teaching time

The longstanding tradition, and relatively high degree, of year repetition in Portugal reflects culturally-ingrained and strong societal beliefs about the benefits of simply adding more teaching time. The ways current strategies and programmes for addressing students’ struggles in school such as Priority Educational Intervention Areas (Territórios Educativos de Intervenção Prioritária – TEIP) and National Programme to Promote School Success (Programa Nacional de Promoção do Sucesso Escolar – PNPSE) are translated into practice frequently mirror this tradition of, and belief in, the benefits of augmenting teaching time. As part of school success strategies and programmes, schools frequently provide students with extra instruction, tutorials and support in small groups inside or outside the classroom. But these approaches do not address the relevance and effectiveness of the extra instructional time. They beg the question: “If it did not work the first time, would more of the same result in different outcomes?”. This is particularly true if the additional instruction or personalised tutoring does not present the information in different ways (Chapter 4 presents evidence on the kind of personalised learning/tutoring shown to generate significant learning gains). Furthermore, the overwhelming evidence on the negative effects of year repetition, not only on academic achievement but also on students’ attitudes and behaviour towards school (OECD, 2018[19]), challenges the Portuguese education system to move away from simply adding more teacher instruction time.

Portugal should consider promoting the implementation of schoolwide projects and interventions that are evidence-informed and tailor-made to the specific local situation of the school. The OECD report on Responsive School Systems: Connecting Facilities, Sectors and Programmes for Student Success (2018[19]) reviews the evidence on strategies to support students struggling to progress. Portugal should reflect on whether strategies such as smaller classes, more instructional assistants, dividing classes into levelled groups, extra Portuguese and mathematics classes and others that involve more time or fewer students per teacher are the most effective. The following approaches have the best evidence on their positive impacts with struggling learners:

  • Investigate and evaluate the reasons why, and with what, students struggle and design deliberate educational strategies aligned with the needs, motivation and talents of students (McAdams, 2006[57]; Marzano, 2004[58]; Bruckere, Kirschner and Hulshof, 2015[59]; Wojcicki, 2015[60]).

  • Provide appropriate teacher training and support to improve the quality and not just the quantity of instruction (see Chapter 4).

  • Invest in student counselling, including behavioural insights, to orient students to second-chance education, accelerated education, post-secondary education and the labour market (OECD, 2018[19]).

  • Prepare clear, external criteria, involving cognitive and non-cognitive factors to determine the appropriateness of a student for year repetition; limiting repetition to a module, subject area or failed course in secondary schooling (OECD, 2018[19]).

  • Continue improving early-warning data systems to identify students at risk of failure and assign trained staff to intervene with support early (OECD, 2018[19]).

Some Portuguese schools have effectively translated policy priorities around innovative instructional practices, teacher collaboration and ongoing professional practice learning to effectively utilise the additional TEIP resources, European-funded professional development or curricular autonomies to innovate their interventions for struggling students. National authorities should identify examples of schools of best practice and use networks of learning to spread these success stories.

Reduce fragmentation of governance and operation of the VET network

Portugal should consider rationalising the double offer across vocational programmes provided in comprehensive or professional schools and apprenticeship programmes provided in IEFP training centres. Reducing fragmentation of VET provision requires a thorough evaluation of available resources and the potential for sharing agreements across providers. For instance, economies may be obtained by systematically organising specialised training in IEFP training centres close to schools. Increased efficiencies could be also leveraged through closing less successful VET programmes. Freed resources could be allocated to reinforce the quality of human and physical resources of the remaining providers.

While SANQ has gradually permitted data-informed steering of the existing offer, greater efforts should be made to make provision more efficient. Given the fragmented governance of VET, Portugal should reflect on giving the National Agency for Qualification and VET (ANQEP) greater legitimacy to develop inter-ministerial policy. In particular, it can capitalise on the agency’s capacity to co-ordinate across education and labour market stakeholders to develop a strategic vision for the sector. Developing the technical capacity to estimate the costs and benefits of offering each strand of VET courses across different networks of providers is a key first step. Whereas responsibility is granted to regulate the offer, no oversight is granted for resource allocation. In order to attain efficiency gains, Portugal should further align the authorities of network planning with those of resource distribution. The misalignment is particularly evident for the case of the offer dependent on European funds. Current legislation does not facilitate an effective co-ordination between agencies such as PO CH – responsible for distributing European monies – and ANQEP. For example, the co-ordinating council of PO CH is required to have only 1 representative from ANQEP, among more than 30 different stakeholders (Dispatch 2906-A/2015). Increasing the participation of the agency in decision-making processes would allow a distribution of school resources further grounded on evidence of current and future network needs.

Another major challenge in the organisation of the VET network relates to the insufficient participation of regional actors. While current legislation envisages the involvement of inter-municipal communities (CIMs) in ranking priorities and planning the network at the regional-level, only 13 of 23 CIMs are currently engaged or planning to engage in the process. In order to increase participation and improve vertical co-ordination, Portugal should build more binding and effective contracts to signal the costs of not participating in the regional co-ordination module of SANQ. Moreover, it can leverage agreements across different upper-secondary VET providers by relying on the potential of Qualifica centres to foster horizontal co-ordination. While the focus of the Qualifica Programme is on adult education, the territorial outreach of its centres can be further used to engage schools with nationally defined strategies for VET and sharing learning resources. Qualifica centres could thus be used to optimise the use of specialised equipment and facilities across comprehensive schools, professional schools and IEFP training centres.

Finally, sector leaders can continue improving the capacity of SANQ to guide policy and better inform students. Alongside the legitimacy of ANQEP to steer the network, the relevance of SANQ can be further developed. Consider endowing ANQEP with additional capacity to produce more sophisticated anticipation exercises. Incentives to involve universities and research centres in the continuous improvement of the methods and the incorporation of international best practices would be helpful for this purpose. The improvement of SANQ’s methodology will generate more robust evidence to assess the value of each VET programme. As part of the strategy to promote system learning, links across datasets administered by different ministries can be developed and made available to researchers, according to best practices of data protection. More efforts can be made to rigorously estimate the economic impact of shifts in the demand for and supply of skills, better informing policy making and public debate. Such evidence could effectively feed a communication strategy on the value of VET, tailored to the profiles of different students. SANQ already has sufficient capacity to combine local labour market data and skills needs to better inform school-level services of psychology and counselling (SPOs) on available fields of study and local employment trends. Finally, given its national outreach, SANQ could be expanded to help regulate the offer of adult learning courses in schools, defining a long-term strategy for the sector (OECD, 2018[19]).

Develop a differentiated profile of VET, while pursuing the strategy of integration within secondary schools

The expansion of provision of VET in upper secondary schools has helped to increase enrolment in the sector. However, the rapid growth in non-specialist schools merits greater attention with regards to its effectiveness. While convergence with European monitoring mechanisms (EQAVET) has gradually allayed concerns about the quality of such provision, stakeholders in the VET sector have insisted on a more thorough evaluation of this type of offer. The effectiveness of VET depends on the consideration of what are the benefits of offering VET in comprehensive schools vis-à-vis offering it in dedicated professional schools. Transparent assessments of graduation rates, employability prospects and participation in the further education of VET students across comprehensive and professional schools should be regularly performed and made available so that strategic priorities regarding the organisation of the vocational offer are based on rigorous cost-benefit analyses (OECD, 2018[19]). Importantly, stakeholders in Portugal continue to stress the low status of VET as a pathway to success. Improving the status of VET may crucially depend on a combination of different strategies.

First, Portugal should consider facilitating transitions across programmes. Portugal has been taking steps in this direction. Indeed, integrating VET in comprehensive schools can increase the fluidity in transitions across and within general and vocational tracks. Increased opportunities for transition would serve to reduce perceptions of the VET pathway as a dead-end track. However, the current organisation of the upper secondary offer hinders such transitions. Students who wish to change tracks during their upper secondary studies often have to halt progression and re-sit previous school years. Allowing students to move easily across tracks can help reduce repetition. Information and guidance for students not only in the transition from 3rd cycle to upper secondary education but also throughout the duration of upper secondary studies, can help students more easily achieve their potential (OECD, 2018[19]). As detailed above, through the curriculum flexibility project, students in upper secondary will, starting in 2018/19, be allowed to receive credit for some courses even when they transition across general education pathways. These efforts should be sustained and expanded.

Second, the organisation of the school offer should also facilitate transitions of VET students to post-secondary education, in particular, tertiary education. While the simultaneous recognition of academic and vocational skills known as double certification is a step towards an improved status, more opportunities should be provided for VET students to progress into higher levels of education. The current tertiary education access regime is traditionally oriented towards students from the sciences and humanities track. A recent study shows that only 6% of graduates from professional courses in 2013/14 were enrolled in higher education 1 year after graduation, including tertiary education courses in polytechnic universities, typically geared towards vocational studies. Likewise, only 10% of graduates from professional programmes access short-cycle tertiary education (ISCED level 5) courses (Cursos Técnicos Superiors Profissionais CTeSP), even though these courses have different admission standards than university or polytechnic offerings (DGEEC, 2016[61]).

A possible way to increase participation in higher education could be by creating specific access pathways – based on alternative criteria – for students coming from vocational programmes in upper secondary education (OECD, 2018[19]). Such pathways could, at least in the short term, help VET build a differentiated profile and increase enrolment of highly motivated students. Alternatively, the OECD Review of the Higher Education, Research and Innovation System in Portugal (Forthcoming[62]) – in line with assessments from national entities – notes that the existing regime of entrance examinations could be modified, by re-structuring the contents of access examinations in two parts: a general component and a component of modules specific to each type of upper secondary education offer. In particular, entrance examinations should be designed to include modules aligned with VET curricula. As the Review of Higher Education notes, simultaneously improving secondary VET students’ skills and the quality of higher education support will be critical to increase access and success (OECD, Forthcoming[62]). Higher education institutions and VET sector leaders can collaborate to identify VET graduates’ skill and knowledge gaps and take steps to improve the transition by improving VET programmatic curriculum and providing supports and transitional courses in higher education.

Third, strong VET sectors crucially depend on the engagement of employers and the provision of work-based learning. A stronger work-based learning component should not preclude solid general skills to be built into the programmes (OECD, 2018[19]). Effective work-based learning depends on the effective balance between the skills attained in the context of work and those in the classroom (Kis, 2016[63]; Araújo, 2017[44]). While professional programmes in Portugal offer some type of pre-graduation work experience, the amount of hours spent in the work context is still below international practice. In order to foster connections with the labour market, VET sector leaders should reflect on the best way to incentivise companies to participate in apprenticeship schemes that go beyond the programmes offered in IEFP training centres. Apprenticeship programmes that incentivise companies to offer relevant training opportunities for students in upper secondary schools, by also allowing them to benefit from the investment in human capital, would be helpful. Potential strategies may include developing more formal agreements between employers and apprentices through the payment of an apprenticeship wage while maintaining non-binding conditions between the parties. However, financial incentives should be also tailored to different occupations and sectors, and according to firms’ size. Defining different apprentice minimum wages and subsidies across sectors could be an option. Alongside financial stimuli, non-financial incentives should not be underestimated. In particular, the duration of the training programme and the timing of the school-based and work-based components still affect the ability of firms to offer additional and better training opportunities, independently of public financial support (Mühlemann, 2016[64]). In order to do this, effective evaluation mechanisms to determine the appropriate duration of each of these components based on evidence rather than on administrative discretion are key.

Finally, system actors must act to ensure that students are not funnelled into VET programming simply because they are struggling in school or because they come from a disadvantaged background. The incentives are poorly aligned for secondary schools that offer both regular and vocational pathways. Students in VET pathways do not typically take the national examinations in Year 12. If a student is struggling in school and is likely to score poorly on the national assessment, there is a strong incentive to encourage him or her into the VET pathway so as to avoid that student’s mark contributing negatively to the schoolwide average. This concern was voiced by several stakeholders during the review visit and is a common phenomenon in other OECD countries, known as “waterfalling” due to students dropping down to an easier track when they struggle in a more rigorous one (OECD, 2018[19]). Clear screening processes, strong counselling services and most importantly supportive interventions to prevent school failure are key strategies to counteract this (OECD, 2018[19]). Additionally, Portugal should investigate whether the over-representation of socio-economically disadvantaged students in the vocational pathway results in constrained life opportunities for these students. In addition to labour market outcome analyses, Portugal could consider exploring Dutch models of guaranteeing access to higher education for students who have successfully completed secondary education, whether in the VET or general pathway (OECD, 2018[19]).

Build capacity and increase staff support for the mainstreaming of moderate-needs special education students

Portugal must support its teachers to better meet the needs of moderate-needs special education students through increased training and staffing. Currently, while special educational needs teachers receive extensive training in supporting the needs of SEN students, subject-area teachers have no requirements beyond their initial training to pursue expertise in this area. Subject-area teachers, as well as special needs teachers, can grow their skills in differentiating classroom content for all students. Mandating ongoing, job-embedded training with staff from resource centres for inclusion (CRIs) providing technical capacity building could be an effective strategy. See Chapter 4 for more strategies to support widespread, ongoing professional development efforts. In addition to training teachers on classroom instruction, Portugal should evaluate the reasons behind wide regional and school variation in special education identification rates. Should this prove to be due to different knowledge bases and assumptions made at the local level, instituting common standards, personnel and quality training for the evaluation process could prove effective at standardising the evaluation process (OECD, 2018[19]).

Finally, Portugal must invest additional resources in the form of special education teachers to support students in mainstream classrooms. A key premise in the inclusion of students with SEN in the mainstream classroom, in addition to its ethical justification, is that they will benefit from positive learning models. However, as TALIS data indicates, inclusion has proved to be a challenge in many contexts (OECD, 2018[19]). It requires thoughtful planning and intensive supports. Portugal comes to this challenge from a place of advantage as it spends almost no resources on segregated education settings. However, to realise the promise of inclusion, the Portuguese education system needs to ensure that all students who have Individualised Education Plans (IEPs) that call for additional support within reading or mathematics classrooms, in fact, receive these supports.


[42] Abrantes, P. (2013), “Born to fail? Some lessons from a national programme to improve education in poor districts”, International Studies in Sociology of Education, Vol. 23/1, pp. 17-38.

[44] Araújo, S. (2017), “Raising skills in Portugal”, OECD Economics Department Working Papers, No. 1405, OECD Publishing, Paris,

[24] Ares Abalde, M. (2014), “School Size Policies: A Literature Review”, OECD Education Working Papers, No. 106, OECD Publishing, Paris,

[35] Bayer, P., S. Ross and G. Topa (2008), “Place of work and place of residence: Informal hiring networks and labor market outcomes”, Journal of Political Economy, Vol. 116/6, pp. 1150-1196,

[23] Blyth, A. et al. (2012), Modernising Secondary School Buildings in Portugal, OECD Publishing, Paris, (accessed on 06 March 2018).

[15] Brás de Oliveira, J. (2018), Effectiveness of Private Schools Versus Public Schools: A Comparative Analysis in Portugal, NOVA - School of Business and Economics.

[59] Bruckere, P., P. Kirschner and C. Hulshof (2015), Urban Myths about Learning and Education, Elsevier Inc. - Academic Press, Amsterdam.

[31] Bullock, A. and H. Thomas (1997), Schools at the Centre? A Study of Decentralisation, Routledge, New York.

[68] Burns, T. (2016), Education Governance in Action: Lessons from Case Studies, OECD Publishing, Paris.

[49] Burns, T. (2016), Enhancing effective education governance, OECD Publishing, Paris.

[39] Carrell, S. and M. Hoekstra (2010), “Externalities in the classroom: How children exposed to domestic violence affect everyone's kids”, American Economic Journal: Applied Economics, Vol. 2/1, pp. 211-228,

[2] CNE (2017), Estado da Educação 2016 [State of Education 2016], Conselho Nacional de Educação, Lisbon,

[73] Cruz, D. and R. Mamede (2015), “Counterfactual Impact Evaluation of Vocational Education in Portugal”, [Master Thesis],

[26] Cruz, D. and R. Mamede (2015), “Counterfactual Impact Evaluation of Vocational Education in Portugal”, [Master Thesis],

[30] Dafflon, B. (2015), The assignment of functions to decentralized government: from theory to practice, Edward Elgar, Cheltenham.

[29] DGEEC (2018), Estudantes à entrada do Ensino Secundário em 2016/17 [Students upon entry to Secondary Education in 2016/17], DGEEC,$clientServletPath%7D/?newsId=256&fileName=Estudantes___entrada_do_secund_rio_2016_.pdf.

[41] DGEEC (2018), Questionário no âmbito da Estratégia Nacional para a Integração das Comunidades Ciganas [Questionnaire within the Framework of the National Strategy for the Integration of Roma Communities], DGEEC, Lisbon.

[67] DGEEC (2017), Educação em Números 2016 [Education in Numbers 2016], DGEEC.

[28] DGEEC (2017), Educação em Números 2016 [Education in Numbers 2016].

[17] DGEEC (2017), Necessidades Especiais de Educação [Special Education Needs],

[47] DGEEC (2017), Necessidades Especiais de Educação [Special Educational Needs],

[21] DGEEC (2016), Análise da Rede de Estabelecimentos do Ensino Particular e Cooperativo com Contrato de Associação [Analysis of the Network of Private Education Providers with Association Contract],$clientServletPath%7D/?newsId=649&fileName=Analise_da_Rede_de_Estabelecimentos_do_E.pdf.

[69] DGEEC (2016), Estatísticas da Educação 2015/2016 [Education Statistics 2015/16], Direção-Geral de Estatísticas da Educação e Ciência (DGEEC) , Lisbon,$clientServletPath%7D/?newsId=145&fileName=DGEEC_DSEE_2017_EE201520164.pdf.

[61] DGEEC (2016), Transição entre o Secundário e o Superior - Parte I [Transitions between Secondary and Tertiary Education - Part I],$clientServletPath%7D/?newsId=701&fileName=TransicaoSecundarioSuperior_DGEEC.pdf.

[20] DGERT (2016), Vocational education and training in Europe - Portugal, Cedefop ReferNet VET in Europe reports, (accessed on 27 February 2018).

[18] DGERT (2016), Vocational Education and Training in Europe - Portugal, Cedefop ReferNet VET in Europe reports,

[38] Duncan, G. and K. Magnuson (2011), “The nature and impact of early achievement skills, attention skills, and behavior problems”, in G. Duncan and R.J. Murnane (eds.), Whither Opportunity? Rising Inequality, Schools, and Children's Life Chances, Russell Sage Foundation, New York.

[16] EASNIE (forthcoming), Financing Policies for Inclusive Education Systems - Portugal Country Report, EASNIE.

[36] Epple, D., E. Newlon and R. Romano (2002), “Ability tracking, school competition, and the distribution of educational benefits”, Journal of Public Economics, Vol. 83/1, pp. 1-48,

[5] Eurydice (2007), School Autonomy in Europe. Policies and Measures, Eurydice, Brussels.

[48] Faguet, J. (2014), “Decentralization and governance”, World Development 53, pp. 2-13.

[4] Glenn, C. and J. De Groof (2002), Finding the Right Balance. Freedom, Autonomy and Accountability in Education, Lemma, Utrecht.

[51] Guryan, J. (2004), “Desegregation and black dropout rates”, American Economic Review, Vol. 94/4, pp. 919-943,

[50] Harris, A. (2012), “Leading system-wide improvement”, International Journal of Leadership in Education: Theory and Practice, Vol. 15/3, pp. 395-401.

[32] Hart, B. and H. Risley (1995), Meaningful Differences in the Everyday Experience of Young American Children, Paul H. Brookes Publishing, Baltimore.

[27] Henriques, R. et al. (forthcoming), “Vocational education: coursetaking choice and impact on dropout and college enrollment rates”, [Master Thesis] .

[70] Henriques, R. et al. (forthcoming), “Vocational education: Coursetaking choice and impact on dropout and college enrollment rates”, [Master Thesis].

[6] Hooge, E. (2016), Making multiple school accountability work, OECD Publishing, Paris.

[8] IGEC (2018), Segundo Ciclo de Avaliação Externa [Second Cycle of External Evaluation], General Inspection of Education and Science,

[52] Johnson, R. (2015), “Long-run impacts of school desegregation and school quality on adult attainments”, NBER Working Paper, No. 16664,

[63] Kis, V. (2016), “Work-based Learning for Youth at Risk: Getting Employers on Board”, OECD Education Working Papers, No. 150, OECD Publishing, Paris,

[74] Leithwood, K. and C. McCullough (2016), Leading High-Performing School Districts, Education Canada.

[54] Liebowitz, D. (2018), “Ending to what end? The impact of the termination of court-desegregation orders on residential segregation and school dropout rates”, Educational Evaluation and Policy Analysis, Vol. 40/1, pp. 103-128,

[43] Marôco, J. (2017), Assimetrias Educacionais em Portugal: Através das lentes do PISA [Educational assimetries in Portugal: Through PISA lenses], Concelho Nacional da Educação, Lisbon.

[58] Marzano, R. (2004), Building Background Knowledge for Academic Achievement: Research on What Works in Schools, ASCD.

[57] McAdams, D. (2006), What School Boards Can Do: Reform Governance for Urban Schools, Teachers College Press.

[1] Ministry of Education (2018), OECD Review of Policies to Improve the Effectiveness of Resource Use in Schools: Country Background Report for Portugal, Ministry of Education, Lisbon,

[3] Ministry of Education (2018), OECD Review of Policies to Improve the Effectivess of Resource Use in Schools: Country Background Report for Portugal.

[64] Mühlemann, S. (2016), “The Cost and Benefits of Work-based Learning”, OECD Education Working Papers, No. 143, OECD Publishing, Paris,

[10] Nunes, L., A. Reis and C. Seabra (2016), Será a Repetição de Ano Benéfica para os Alunos [Is Year Repetition Beneficial to Students?], Fundação Francisco Manuel dos Santos, Lisboa,

[55] Nusche, D. et al. (2015), OECD Reviews of School Resources: Flemish Community of Belgium 2015, OECD Publishing, Paris,

[46] OECD (2018), Building a National Skills Strategy for Portugal, Action Phase: Strengthening the Adult Learning System, OECD Publishing, Paris.

[19] OECD (2018), Responsive School Systems: Connecting Facilities, Sectors and Programmes for Student Success, OECD Publishing, Paris,

[45] OECD (2018), Skills Strategy Implementation Guidance for Portugal: Strengthening the Adult-Learning System, OECD Skills Studies, OECD Publishing, Paris,

[40] OECD (2018), The Resilience of Students with an Immigrant Background: Factors that Shape Well-being, OECD Reviews of Migrant Education, OECD Publishing, Paris,

[9] OECD (2017), Education at a Glance 2017: OECD Indicators, OECD Publishing, Paris,

[33] OECD (2017), Educational Opportunity for All: Overcoming Inequality throughout the Life Course, Educational Research and Innovation, OECD Publishing, Paris,

[65] OECD (2017), OECD Economic Surveys: Portugal 2017, OECD Publishing, Paris,

[66] OECD (2016), Education at a Glance 2016: OECD Indicators, OECD Publishing, Paris,

[71] OECD (2016), Historical Population Data and Projections (1950-2050),

[72] OECD (2016), PISA 2015 Results (Volume I): Excellence and Equity in Education, OECD Publishing, Paris,

[13] OECD (2016), PISA 2015 Results (Volume I): Excellence and Equity in Education, PISA, OECD Publishing, Paris,

[14] OECD (2016), PISA 2015 Results (Volume II): Policies and Practices for Successful Schools, OECD Publishing, Paris,

[56] OECD (2015), OECD Economic Surveys: Belgium 2015, OECD Publishing, Paris,

[62] OECD (Forthcoming), OECD Review of the Higher Education, Research and Innovation System in Portugal, OECD Publishing, Paris.

[12] Oliveira Baptista, J., P. Pereira and DGEEC (2018), Assimetrias de Contexto Socioeconómico entre Escolas Públicas – 2.º Ciclo, 2015/16 [Asymmetries of Socio-economic Context between Public Schools - 2nd cycle, 2015/16], Direção-Geral de Estatísticas da Educação e Ciência (DGEEC) , Lisbon,$clientServletPath%7D/?newsId=885&fileName=DGEEC_AssimetriasEntreEscolas_2_CEB.pdf.

[11] Owens, A., S. Reardon and C. Jencks (2016), “Income segregation between schools and school districts”, American Educational Research Journal, Vol. 53/4, pp. 1159-1197,

[53] Reardon, S. et al. (2012), “Brown Fades: The End of Court-Ordered School Desegregation and the Resegregation of American Public Schools”, Journal of Policy Analysis and Management, Vol. 31/4, pp. 876-904,

[37] Sacerdote, B. (2011), “Peer effects in education: How might they work, how big are they and how much do we know thus far?”, in E. Hanushek, S. Machin and L. Woessman (eds.), Handbook of the Economics of Education, Elsevier, San Diego, CA.

[7] Santiago, P. et al. (2012), OECD Reviews of Evaluation and Assessment in Education: Portugal 2012, OECD Publishing, Paris,

[34] Souers, K. and P. Hall (2016), Fostering Resilient Learners: Strategies for Creating a Trauma-Sensitive Classroom, ASCD, Alexandria, VA.

[22] Veloso, L., J. Marques and A. Duarte (2014), “Changing education through learning spaces: Impacts of the Portuguese school buildings’ renovation programme”, Cambridge Journal of Education, Vol. 44/3, pp. 401-423,

[25] Vieira, M., L. Pappámikail and J. Resende (2013), “Forced to Deal with the Future: Uncertainty and Risk in Vocational Choices among Portuguese Secondary School Students”, The Sociological Review, Vol. 61/4, pp. 745-768,

[60] Wojcicki, W. (2015), Moonshots in Education. Launching Blended Learning in the Classroom., Pacific Research Institute.

Annex 3.A. Private and public schools
Annex Figure 3.A.1. Number of pre-schools and schools by type of provision, 2015/16

Source: CNE (2017), Estado da Educação 2016 [State of Education 2016], pp. 45-49.

End of the section – Back to iLibrary publication page