This chapter discusses student learning outcomes, well-being and expectations for the future in Eastern Partnership (EaP) countries and economies. It first focuses on student performance in mathematics, reading and science, including OECD Programme for International Student Assessment (PISA) 2022 results and trends over time. It then examines student well-being by looking at their sense of belonging and feelings of safety at school. Finally, the chapter explores students’ expectations about their level of education attainment and aspirations for their future occupation.
Education in the Eastern Partnership: Findings from PISA
Abstract
This chapter analyses student learning outcomes, well-being and expectations for the future in Eastern Partnership (EaP) countries and economies as measured in the OECD Programme for International Student Assessment (PISA) data for 2022 (PISA 2022) (see Figure 2.1).1 Student learning is analysed for 2022 and over time, relative to the OECD average and Estonia and Singapore as selected aspirational benchmarks. The average for European Union (EU) countries is typically similar to the OECD average on a number of indicators. The analysis highlights that EaP students trail behind the average OECD student as well as students in Estonia and Singapore as selected aspirational benchmarks. The average for European Union (EU) countries is typically similar to the OECD average on a number of indicators. The analysis highlights that EaP students trail behind the average OECD student as well as students in Estonia and Singapore in all subjects. Among the EaP countries/economies, Ukrainian regions (18 of 27)2 perform best, while Baku (Azerbaijan)3 and Georgia show the lowest performance across all domains.
All EaP countries/economies have a large proportion of 15-year-olds who lack basic skills in one, two or all three subjects assessed by PISA, and only a small minority of students who perform at the highest proficiency levels. Notably, EaP countries/economies maintained their mean performance in mathematics – the main assessment domain in PISA 2022 – between 2018 and 2022, except for Baku, where students’ mathematics performance declined. However, mean performance in reading and science declined in all EaP countries/economies except Georgia.
Additionally, the chapter examines students’ perceptions of belonging and safety within schools, and students’ educational aspirations. Overall, the sense of belonging at school is lower among EaP countries/economies than on average across OECD countries. However, looking at specific indicators of student belonging reveals a more differentiated picture across EaP countries/economies. Surprisingly, in the current context, students in Ukrainian regions report exceptionally high levels of perceived safety in and out of schools, contrasting with Moldova and other EaP countries, where students feel less safe compared to the OECD average. With regards to student expectations, key findings include a significant increase in educational aspirations across all EaP countries, except Ukrainian regions, where aspirations decreased, remaining high nonetheless. This raises questions about how these aspirations match opportunities for tertiary education and the labour market.
Student belonging and safety are all measures of student well-being, defined in terms of the environment a student is exposed to at school and outside of school (OECD, 2019[1]). Other dimensions of student well‑being, including their social relationships with teachers and family and their engagement with school – indicated by behaviours like skipping school or the classroom disciplinary climate – are discussed in depth in Chapter 4 of the report. Box 2.1 provides a summary of these and selected additional aspects of students’ school experience that sometimes receive less attention in policy discussions.
The analysis of student well-being in this report focuses on some important dimensions, including a sense of belonging to school and feeling safe at school, discussed in depth in this chapter. However, the information on student well-being collected by PISA goes beyond the dimensions covered in depth in this report. Table 2.1 summarises results for some of these and other aspects of student well-being for EaP countries/economies as proposed by work by PISA on a comprehensive framework of student well-being (Cignetti and Piacentini, 2024[2]). According to these data, the levels of psychological and social well-being, as well as students’ resilience, are often comparable to or better than the OECD average in EaP countries/economies.
Psychological well-being: Table 2.1 highlights three selected dimensions of psychological well‑being. More students in Baku report having “emotional control” (the ability to manage emotions to achieve goals, complete tasks or control and direct behaviour) than the OECD average. Students in EaP countries/economies (excluding Baku) report higher “life satisfaction” than the OECD average. Regarding a “sense of purpose” (the capacity to find meaning or purpose in life), more students in EaP countries/economies agree with the statement that “My life has clear meaning or purpose” compared to the OECD average.
Resilience: Resilience, defined as the ability to encounter adversity or stress and achieve positive outcomes, is generally higher among students in EaP countries/economies, with less “fear of failure” reported compared to their OECD peers, except in Baku. Students in Baku report higher levels of “belief in self” (self-perceived ability to handle difficult situations or solve complex issues) than the OECD average, whereas the opposite is true in Moldova.
Social relationships: Students in EaP countries/economies report similar levels of belonging at school to the OECD average (a more detailed analysis on this dimension is provided later in this chapter). Students in Georgia report better relationships with their teachers compared to other EaP countries and the OECD average, and students in Baku and Georgia indicate higher levels of family support than the OECD average (a more detailed analysis on this dimension is provided in Chapter 4).
Average performance in PISA 2022
Learning outcomes in EaP countries/economies trail behind international standards, including the OECD average and countries like Estonia and Singapore that lead in PISA 2022 results (Figure 2.2). Ukrainian regions come closest to matching OECD averages in all subjects. Baku and Georgia have the most room to improve.
The learning gap relative to OECD countries is smallest in mathematics and science and largest in reading for all EaP countries/economies. In mathematics, the main domain assessed in PISA 2022, Ukrainian regions scored 31 points below the OECD average. In Moldova, the score point difference was 58, in Baku 75 and in Georgia 82.
By contrast, in reading, Ukrainian regions score 48 points behind OECD countries, a learning gap equivalent to about 2.4 years of schooling by an average 15-year-old student in an OECD country.4 The difference in reading was more than twice as large in Baku and Georgia. Only 10 systems performed lower than Baku out of all 81 participants in PISA (OECD, 2023[3]).5
Considering their economic development, EaP countries/economies perform as expected based on their per capita gross domestic product (GDP). This is analysed in depth in Chapter 1. Indeed, they perform better than other countries with similar or even higher income levels. For example, Baku and Moldova surpass Colombia and Costa Rica in mathematics despite having a lower GDP. Meanwhile, Mongolia, with an even lower GDP, still does better than Baku, Georgia and Moldova in the same subject.
Performance in mathematics, reading and science in PISA 2022
Notes: The mean score in science is not statistically significantly different between Baku and Georgia. All other differences between countries/economies, and between countries/economies and the OECD average, are statistically significant.
Scores can be compared within each domain but not between different domains (e.g. math scores are not comparable to reading scores).
Source: OECD (2022[4]), PISA 2022 Database, https://www.oecd.org/en/data/datasets/pisa-2022-database.html.
Table 2.2. shows countries with PISA 2022 scores similar to those of the EaP countries/economies.6 The differences in the countries being compared are small enough that they could have occurred by chance. For instance, in mathematics, the scores of the Ukrainian regions matched those of Brunei and Serbia. Georgia’s scores were comparable to Colombia, Costa Rica, Mexico, North Macedonia, Peru, Saudi Arabia and Thailand.
Notes: Countries and economies are listed in descending order of their mean score within cells.
Countries and economies are ranked in descending order of mean performance in mathematics.
1. Countries and economies whose mean score is not statistically significantly different from the EaP comparison country’s/economy’s score.
* Caution is required when interpreting estimates because one or more PISA sampling standards were not met (see Box 1.1 in Chapter 1).
Source: OECD (2022[4]), PISA 2022 Database, https://www.oecd.org/en/data/datasets/pisa-2022-database.html.
Trends over time in student performance
EaP countries/economies have participated in various PISA cycles, as detailed in Chapter 1. These data allow for the analysis of student performance trends over different periods. Short-term trends are examined through the 2022 and 2018 results, while long-term trends are drawn from earlier assessments.
Analysing short-term trends in student performance against long-term patterns, where the availability of the data makes this possible, is particularly important in the context of the disruptions caused by the COVID‑19 pandemic. This helps to understand the potential impact of the pandemic on education while accounting for deeper, longer-term processes or factors that might influence student performance.
Georgia and Moldova have participated consistently in PISA since 2009, except in 2012, allowing for an analysis of long-term performance trends. Baku7 and Ukraine only participated in PISA 2018 and 2022, limiting the analysis to short-term trends.
The case of Ukraine is unique, as almost the entire country participated in PISA 2018,8 but only 18 of its 27 regions participated in PISA 2022 (see Box 1.1 in Chapter 1). To account for this change, comparisons between the two cycles in this section are based on data from the 18 regions that participated in both cycles.
Short-term trends in student performance
Trends in student performance between PISA 2018 and PISA 2022 varied widely in direction (whether improving, declining or stable) and magnitude (how large these changes are) across different domains in EaP countries/economies (Figure 2.3).
PISA performance in mathematics, reading and science across cycles
Notes: White markers (dots, squares or triangles) indicate mean performance estimates that are not statistically significantly above/below PISA 2022 estimates.
The OECD average in this figure is the arithmetic mean across all OECD member countries excluding Austria, Chile, Colombia, Costa Rica, Estonia, Israel, Lithuania, Luxembourg, the Netherlands, the Slovak Republic, Slovenia, Spain, Türkiye, the United Kingdom and the United States. It includes only 23 member countries with non-missing values across all of the assessments between 2009 and 2022. This average is used to report on a consistent set of OECD countries.
Sources: OECD (2022[4]), PISA 2022 Database, https://www.oecd.org/en/data/datasets/pisa-2022-database.html; OECD (2018[5]), PISA 2018 Database, https://www.oecd.org/en/data/datasets/pisa-2018-database.html; OECD, PISA 2009, 2012 and 2015 databases.
In mathematics, EaP countries/economies generally maintained their performance between 2018 and 2022, contrasting with the average decline across OECD countries. In Georgia, Moldova and Ukrainian regions, the mean score in mathematics did not change significantly.9 Conversely, OECD member countries experienced a notable decrease in mathematics, with an average drop of 15 points. This decline was the largest ever observed in PISA history. To put this drop into perspective, until 2018, the change observed in mathematics in the OECD average had never exceeded four score points.
However, Baku aligns with the broader downward trend between 2018 and 2022. With a drop of more than 20 points, the decline in performance in Baku exceeded the average decline in performance across OECD countries. On average, mathematics remained the strongest domain in Baku since performance also declined in the other two subjects. Alongside Baku, 18 other education systems saw similar declines of at least 20 points. Some countries nevertheless experienced even more significant decreases, exceeding 30 points: Albania, Jordan, Iceland, Norway and Malaysia in descending order.
While most EaP countries/economies bucked the international trend for mathematics, they generally mirrored the short-term global decline in reading. In fact, most EaP countries/economies exceeded the drop in the OECD average.
The decline was most acute in Ukrainian regions, with a 36-point drop. Baku saw a 24-point drop, while Moldova saw a smaller but still substantial drop of 13 score points. For comparison, on average, the fall amounted to 10 score points across OECD countries. Georgia was the exception, with stable reading performance between 2018 and 2022.
In science, all EaP countries, except Georgia, experienced a decline, although less severe than in reading. The drop was again larger in Baku (18 points) and Ukrainian regions (17) than in Moldova (11). Science performance did not change significantly across OECD countries, similar to the situation in Georgia.
Long-term trends in student performance
Among the four EaP countries/economies covered in this report, only Georgia and Moldova have sufficient PISA data to analyse long-term trends in student performance. In these two countries, performance had increased from low overall scores since their first participation in 2009. However, this trend began to shift in Georgia starting in 2015 and in Moldova in 2018 (Figure 2.3).
In Moldova, the overall trajectory has shown significant long-term improvement with a recent reversal of this pattern. From the country’s first participation in PISA in 2009 up to 2018, performance had been improving in all subjects. However, between 2018 and 2022, as discussed, performance declined in reading and science, although mathematics scores remained stable. Despite these recent declines, Moldova’s average scores in PISA 2022 in mathematics and reading were still higher than in 2009. In science, scores returned to levels close to those of 2009.
In Georgia, the overall trajectory has been somewhat more uneven, with significant improvements again in earlier cycles followed by a reversal and decline, which appears to be stabilising. From 2009 to 2015, the country experienced a large increase in mean performance across all subjects. This period of improvement was followed by a sharp decline in reading and science and a gradual decline in mathematics. As discussed above, performance in all subjects remained stable between 2018 and 2022. Overall, when considering all PISA assessments in the country, average PISA 2022 scores are close to those observed in 2009 but below those observed in 2015.
The positive long-term trend in Georgia and Moldova up to 2015 and 2018 respectively contrasts with a longer-term declining trend in OECD countries. As discussed above, OECD countries experienced an unprecedented decline in mathematics and reading performance between 2018 and 2022. However, average scores had already declined before 2018 in many OECD countries. In reading and science, performance peaked in 2012 and 2009 respectively, before starting a downward trend.
This perspective also underscores the complexity of understanding the relationship between the COVID‑19 pandemic and performance trends. While the pandemic may have influenced recent developments, other factors are likely at play. This is also suggested by the considerable variation across countries in how short-term trends compare with previous performance trends.
For some countries and economies, the changes in PISA performance observed between 2018 and 2022 are significantly different from the trend observed in previous assessments. This seems to be the case for Moldova (where a positive long-term trend may be reversing) and Georgia (where a negative trend may now be stabilising). For others, they confirm or reinforce a trend that started before 2018.
Proficiency levels
In the previous section, student performance was measured in terms of average score points on the PISA scale; however, average scores do not indicate what students are capable of doing in each subject or the variation in capabilities across the student population. In this section, PISA scores are translated into proficiency levels to provide a meaningful interpretation of the types of tasks that students with higher or lower PISA scores are able to complete successfully.
In mathematics and reading, PISA 2022 categorised student performance into eight different proficiency levels, ranging from the highest (Level 6) to the lowest proficiency (Level 1c) (see Table 2.3).10 In science, PISA 2022 categorised performance into seven proficiency levels. Level 2 is considered the “baseline” level of proficiency across domains, the minimum level of knowledge and skills students need to acquire to progress in their education and participate fully in society. At this level, students begin to demonstrate the ability and initiative to use mathematics in simple, real-life situations. A high share of students below the baseline proficiency level can hinder long-term economic growth, productivity and innovation, as an under-skilled workforce may struggle to adapt to changing technological and economic demands. Annex 2.A illustrates the types of tasks students were confronted with in PISA 2022 to demonstrate proficiency at Levels 1a and 2 in mathematics.
Source: OECD (2022[4]), PISA 2022 Database, https://www.oecd.org/en/data/datasets/pisa-2022-database.html.
As shown in Figure 2.4, the proportion of students who scored below Proficiency Level 2 (i.e. “low-performing students”) in PISA 2022 is greater in every subject in all EaP countries/economies than on average across OECD countries. They are also far from the EU-level targets set out in the European Education Area strategic framework (adopted by the Council of the European Union in February 2021) (European Union, 2021[6]). These aim for fewer than 15% of 15-year-olds to be low achievers in reading, mathematics and science by 2030. In mathematics, the percentage of low-performing students ranges from 42% in Ukrainian regions to 66% in Georgia. In contrast, the average across OECD countries is 31%. In reading and science, approximately two-thirds of students score below Level 2 in Baku and Georgia.
In EaP countries/economies, most students who perform above baseline proficiency levels in mathematics, reading and science perform at Proficiency Levels 2, 3 and 4. Only a small minority of students in EaP countries/economies perform at the top levels of proficiency (Levels 5 and 6). Students who attain Proficiency Level 5 or 6 are commonly referred to in PISA reports as “top performers”. In Ukrainian regions (the top-performing EaP country), the share of top performers was 3% in mathematics and 2% in reading and science. In Baku, Georgia and Moldova, the proportion of top performers was 1% in mathematics and less than 1% in reading and science. By contrast, 9% of 15-year-old students across OECD countries were top performers in mathematics, 7% in reading, and 8% in science.
Percentage of students scoring at different levels of proficiency in mathematics, reading and science
Notes: Countries and economies are ranked in descending order of the percentage of students who performed at or above Level 2.
"Basic proficiency" is defined as students achieving Proficiency Level 2 or above in a subject.
Source: OECD (2022[4]), PISA 2022 Database, https://www.oecd.org/en/data/datasets/pisa-2022-database.html.
Figure 2.5 shows that the share of students not mastering basic skills generally increased in EaP countries/economies between 2018 and 2022. In mathematics, the share of these students increased by 5 and 6 percentage points in Georgia and Moldova. This increase is also observed on average across OECD countries and Estonia, for example. In Baku, the increase in the share of low performers in mathematics was about twice as large (11 percentage points). Analysis for Ukraine is impossible due to reduced participation in PISA 2022 compared to PISA 2018.
In Georgia, the share of students scoring below Level 2 in reading and science stayed consistent between 2018 and 2022, indicating that it is possible to prevent further students from falling behind in challenging periods such as the COVID-19 pandemic. This is also evident in top-performing systems like Singapore, which maintained a stable proportion of low performers.
Percentage of students who score below Proficiency Level 2 in each subject
Notes: Statistically significant differences between 2018 and 2022 are shown in a darker tone. The change between 2018 and 2022 is not statistically significant in reading and science for Georgia: is not significant in mathematics, reading and science for Singapore; it is not significant in science for Estonia.
The OECD average in this figure is the arithmetic mean across all OECD member countries, excluding Costa Rica, Luxembourg and Spain. It includes 35 OECD member countries with non-missing values in PISA 2018 and PISA 2022.
Sources: OECD (2022[4]), PISA 2022 Database, https://www.oecd.org/en/data/datasets/pisa-2022-database.html; OECD (2018[5]), PISA 2018 Database, https://www.oecd.org/en/data/datasets/pisa-2018-database.html.
Overlap of low performance and coverage of the education system
Understanding the true extent of low performance requires looking at the overlap of low performance across subjects because students who perform poorly in one subject often perform poorly in other subjects as well. Also, the implications of being a low performer in all three subjects are significant because it signals a more systemic issue within the education system, requiring comprehensive reforms. In all four EaP countries/economies, more students lack basic skills in all three subjects assessed by PISA 2022 than in two subjects or one subject (Figure 2.6).
Overlap of low performers in mathematics, reading and science
Note: Countries and economies are ranked in ascending order of the total percentage of 15-year-olds who are low performers in at least one subject.
Source: OECD (2022[4]), PISA 2022 Database, https://www.oecd.org/en/data/datasets/pisa-2022-database.html.
In Georgia, for example, 20% of all 15-year-olds lack basic skills in one or two subjects, whereas more than twice that share lack basic skills in all three subjects. By contrast, in OECD countries, the overlap of low performance is not as pronounced as in EaP countries/economies. On average, across OECD countries, the share of 15-year-olds who lack basic skills in all three subjects assessed by PISA is about the same as those who lack basic skills in one or two subjects.
In addition, to understand the low performance of a system more broadly, it is important to consider the coverage of the education system because PISA does not test 15-year-olds who are not enrolled in school or who are severely delayed in their school-grade progression. Research shows that these individuals are more likely to lack basic skills in all three subjects assessed by PISA than their peers who are enrolled in and progressing through school (OECD, 2023[3]).
As discussed in Chapter 1, the coverage of the education system is comparatively low in Baku, high in Moldova and similar to the average across OECD countries in Georgia (see Table 1.3 in Chapter 1). In Ukrainian regions, education system coverage was very low in 2022, as measured by the coverage of the PISA sample. However, this should be interpreted with caution since schools in several regions of Ukraine were not accessible in 2022 due to the war. In PISA 2018, the coverage of the education system in Ukraine was higher and similar to the average across OECD countries (Coverage Index 311 was 86.7% in Ukraine in 2018). In Moldova, education system coverage was high in PISA 2015 and has increased by a small margin since then (Coverage Index 3 was 92.9% in 2015 and 97.4% in 2022).
In Georgia, education system coverage increased by 7.6 percentage points between 2015 and 2022 (Coverage Index 3 was 78.7% in 2015). This increase brought education system coverage in Georgia closer to the average across OECD in PISA 2022.
Sense of belonging at school
PISA also provides insights into students’ sense of belonging and social experiences at school. To measure students’ sense of belonging at school, students were asked whether they agree (“strongly disagree”, “disagree”, “agree”, “strongly agree”) with the six school-related statements included in Figure 2.7. Some of these statements are “negatively” worded (e.g. I feel lonely at school), meaning that disagreement with the statement indicates a higher sense of belonging at school. These statements were combined into an overall index of sense of belonging at school, the average of which is zero across OECD countries. According to this index, students’ sense of belonging at school is lower among EaP countries/economies than on average across OECD countries. However, except for Baku, it is stronger than in Singapore than in Estonia.12
This picture has not changed much since the last PISA survey in 2018. Sense of belonging at school remained largely unchanged in both EaP countries/economies and, on average, across OECD countries.13 More substantial changes were observed between 2015 and 2018. During this period, the sense of belonging worsened in Georgia and Moldova.14
Figure 2.7 shows specific aspects of students’ sense of belonging at school. Across EaP countries/economies, Moldova and Ukrainian regions stand out in that four out of five students agree or strongly agree with the statement: “I feel like I belong at school”. In Baku, three out of four students feel they belong at school, which is similar to the OECD average and Singapore.
Percentage of students who disagreed/strongly disagreed or agreed/strongly agreed with statements about a sense of belonging at school
Source: OECD (2022[4]), PISA 2022 Database, https://www.oecd.org/en/data/datasets/pisa-2022-database.html.
Georgia has a significantly lower share of students who feel that they belong at school (51%). Moreover, a relatively modest share of students in the country feel liked by their peers (70%). Yet, on average, Georgian students are less likely to report feeling lonely or like an outsider at school than their counterparts in other EaP countries/economies and OECD countries.
In Baku, students experience relatively high levels of sense of isolation and discomfort at school. About one in four students reported feeling lonely, like an outsider (or left out of things), or awkward and out of place at school. This exceeds the OECD average and is higher than in all of the other EaP countries/economies. It is also higher than in Singapore, where these feelings of isolation are already quite common.
Across all EaP countries/economies, making friends is a common experience, according to PISA data. Over three-quarters of students feel at ease making new friends in all EaP countries/economies. This is in line with the OECD average (76%).
Feeling safe at school
PISA 2022 also asked students about their perceptions of safety in different environments, which can influence students’ stress levels, willingness to attend school and ability to concentrate while at school (Lacoe, 2016[7]). While no causal claims are possible, higher levels of perceived safety at school are associated with higher mathematics performance, life satisfaction and sense of belonging at school on average across OECD countries and in Baku, Georgia and Ukrainian regions.15
The data show a clear hierarchy of perceived safety in and out of school. Students from the regions of Ukraine feel the safest and most secure compared to the OECD average, followed by students from Georgia, Baku and Moldova.
Ukrainian regions show exceptionally high levels of perceived safety, with over 92% of students feeling safe in the classroom, elsewhere at school and when travelling to and from school. It is plausible that schools and school communities became a source of social-emotional protection for students in the context of war, which could explain the perception of schools as a safe space. All other EaP countries/economies fall below the OECD average in terms of students’ perceptions of safety.
Percentage of students who agree or strongly agree with the statements below
Note: Items are ranked in ascending order at the OECD average.
Source: OECD (2022[4]), PISA 2022 Database, https://www.oecd.org/en/data/datasets/pisa-2022-database.html.
In Moldova, around 40% of students feel unsafe not only on their way to and from school but even in their classrooms. In Baku, still, around 20% of students feel unsafe in their classroom. This contrasts with other school systems, including high-performing ones such as Estonia and Singapore, where less than 5% of students feel unsafe in their classroom or other places at school.
PISA 2022 surveyed students on their expected highest education level and anticipated occupation at age 30.
Educational expectations
In the EaP countries/economies, the share of students expecting to complete tertiary education (International Standard Classification of Education [ISCED] Level 5 or above) has increased significantly between PISA 2018 and PISA 2022, except in Ukrainian regions. By contrast, across OECD member countries, the share of students expecting to complete tertiary education has remained stable at almost 70% between PISA 2018 and PISA 2022.
Baku experienced a very large increase in educational aspirations (17 percentage points). While the share of students expecting to complete tertiary education in Baku was lower than the OECD average in PISA 2018, it now exceeds the OECD average. Four out of five students expect to complete tertiary education in Baku in 2022.
Moldova also experienced a significant increase (14 percentage points) but from a much lower base. In Moldova, the share of students expecting to complete tertiary education remains below the OECD average in PISA 2022.
Baku and Georgia were the EaP country/economy with the highest educational expectations in PISA 2022. In Georgia, the share of students expecting to complete tertiary education increased moderately between 2018 and 2022. As in Baku, this share is above the OECD average.
Ukrainian regions show a decline, probably related to the upheaval caused by the war, affecting young people’s views on the feasibility and value of pursuing tertiary education. The proportion of students expecting to complete tertiary education in Ukrainian regions fell by 25 percentage points. Almost all 15-year-old students (95%) are expected to complete tertiary education in 2018, as opposed to 70% in 2022. Educational expectations nevertheless remain relatively high and are now similar to the OECD average.
In EaP countries/economies where aspirations have been increasing, this has a positive dimension as these might fuel students’ motivation to invest in school and beyond. It theoretically also provides the pool of young people that EaP countries/economies require for social and economic transformation. At the same time, they raise important questions. First, there is still a gap between students’ high educational aspirations and their actual academic performance. This might make it difficult for students to achieve their goals. Where students succeed in entering tertiary education, they might drop out, which would result in costs for themselves and society. Second, it might require a substantial increase in the number of places available in tertiary education institutions in EaP countries/economies. In Azerbaijan, for example, higher education opportunities have been expanded but remain low (World Bank, 2020[8]). Finally, the labour markets in EaP countries/economies might not yet be prepared to offer skilled jobs to a growing number of highly educated young people.
Percentage of students who expect to complete tertiary education (ISCED Level 5 or above) in 2018 and 2022
Notes: Statistically significant differences between 2018 and 2022 are shown in a darker tone. The change between 2018 and 2022 is not statistically significant for the OECD average.
The OECD average in this figure is the arithmetic mean across all OECD member countries, excluding Costa Rica, Luxembourg and Spain. It includes 35 OECD member countries with non-missing values in PISA 2018 and PISA 2022.
Sources: OECD (2022[4]), PISA 2022 Database, https://www.oecd.org/en/data/datasets/pisa-2022-database.html; OECD (2018[5]), PISA 2018 Database, https://www.oecd.org/en/data/datasets/pisa-2018-database.html.
Occupational expectations
The following analysis contrasts the future occupational aspirations of 15-year-olds, focusing on two key sectors: health (including doctors, nurses and veterinarians) and information and communication technology (ICT)-related fields (such as software and web developers and data miners).
The health sector, spotlighted during the COVID-19 pandemic for its professionals’ critical and challenging work, has been subject to mixed perceptions. The recognition of the essential nature of this work came alongside awareness of the intense demands and stress it involves, as well as the relatively low remuneration for nursing and support roles. These factors have influenced student interest differently across countries.
Among EaP countries/economies, Baku stands out with a marked increase in students’ interest in health professions (+9 percentage points). Along with Panama, this is the largest increase of all PISA-participating countries/economies. Nevertheless, Baku started from a small pool of interested students. Only 11% of the population is expected to work as a health professional in 2018.
Conversely, Georgia, Moldova and Ukrainian regions all saw a decline in students’ interest in health careers, but only in Ukrainian regions was this decrease statistically significant (-3 percentage points). This reflects broader global patterns. On average, across OECD countries, interest in health careers declines slightly by 1 percentage point, with the percentage of interested students dropping to 16% in 2022. The share of students interested in working as health professionals decreased in 17 out of 73 countries/economies with available data between 2018 and 2022 (OECD, 2023[9]).
For ICT careers, Baku is again an outlier but in the opposite direction, with a slight decrease in student interest (-1 percentage point). This makes it one of only two PISA-participating countries, alongside the Netherlands, to see a dip in this otherwise growing field of interest. ICT careers can be attractive to students seeking rewarding professional paths. The ongoing digital transformation is creating new and exciting career opportunities, often associated with higher salaries due to the increasing demand for skilled professionals (OECD, 2020[10]). The share of students interested in working as ICT professionals increased in 39 out of 73 countries/economies with available data between 2018 and 2022 (OECD, 2023[9]).
All other EaP countries/economies experienced an uptick in interest for ICT occupations in PISA 2022, surpassing the overall OECD trend. Moldova and Ukrainian regions saw an increase of 5 percentage points, while Georgia experienced a 4-percentage point rise.
In PISA 2022, Georgia, Moldova and Ukrainian regions are among the top ten participating countries/economies for the proportion of students reporting a desire to work in ICT. All three had more than 10% of students showing this interest, nearly double the OECD average of 6%. Ukrainian regions had, together with Estonia, the highest share of students interested in this field across all PISA 2022 countries/economies: 15% of 15-year-old students aspired to an ICT career.
Percentage-point change of students who expect to work as the following when they are approximately 30 years old between 2018 and 2022
Notes: Statistically significant differences between PISA 2018 and PISA 2022 (PISA 2022 - PISA 2018) are shown in a darker tone.
The OECD average in this figure is the arithmetic mean across all OECD member countries, excluding Costa Rica, Luxembourg and Spain. It includes 35 OECD member countries with non-missing values in PISA 2018 and PISA 2022.
Countries and economies are ranked in descending order of the change between 2018 and 2022 in the percentage of students who expect to work as health professionals.
Sources: OECD (2022[4]), PISA 2022 Database, https://www.oecd.org/en/data/datasets/pisa-2022-database.html; OECD (2018[5]), PISA 2018 Database, https://www.oecd.org/en/data/datasets/pisa-2018-database.html.
The following examples illustrate the types of tasks students were required to solve to demonstrate Levels 1a and 2 of proficiency in mathematics within the “Triangular Pattern” unit. This unit of assessment has three questions. In all questions, students are presented with a drawing that consists of rows of alternating red and blue triangles. The same image is used across all three items in the unit (OECD, 2023[3]).
Task: Students are asked to compute the percentage of blue triangles shown in the first four rows of the pattern.
Process: Students count the blue triangles, which number 6, out of a total of 16 triangles. Using these data, they calculate the percentage of blue triangles as 37.5% (6 ÷ 16 = 0.375).
Focus: This is an easy item, designed to engage students in basic pattern recognition and mathematical calculation using a straightforward algorithm with all information shown.
Task: The second item in the unit builds off the first item by again asking students to compute the percentage of blue triangles, but this time based on five rows of the pattern.
Process: Since the fifth row is not shown, students have to extend the pattern by one row to determine new values for the number of blue triangles and the total number of triangles. With 5 rows, the percentage of blue triangles is 40.0% (10 blue triangles ÷ 25 total triangles).
Focus: This item is intended to be easy and to get students thinking about extending the pattern beyond what is shown but not extending the pattern so that it requires generalisation. This is a Level 2 item, so it is slightly more difficult than the first item in the unit, possibly because it requires working with a part of the pattern that is not shown; it is still an overall easy item for students. This is the baseline level of proficiency as assessed by PISA 2022.
References
[11] Avvisati, F. and P. Givord (2023), “The learning gain over one school year among 15-year-olds: An international comparison based on PISA”, Labour Economics, Vol. 84, https://doi.org/10.1016/j.labeco.2023.102365.
[2] Cignetti, M. and M. Piacentini (2024), “Beyond grades: Raising the visibility and impact of PISA data on students’ well-being”, OECD Education Working Papers, No. 313, OECD Publishing, Paris, https://doi.org/10.1787/806233fe-en.
[6] European Union (2021), Council Resolution on a strategic framework for European cooperation in education and training towards the European Education Area and beyond (2021-2030) 2021/C 66/01, https://eur-lex.europa.eu/legal-content/EN/ALL/?uri=CELEX%3A32021G0226%2801%29.
[7] Lacoe, J. (2016), “Too scared to learn? The academic consequences of feeling unsafe in the classroom”, Urban Education, Vol. 55/10, pp. 1385-1418, https://doi.org/10.1177/0042085916674059.
[3] OECD (2023), PISA 2022 Results (Volume I): The State of Learning and Equity in Education, PISA, OECD Publishing, Paris, https://doi.org/10.1787/53f23881-en.
[9] OECD (2023), PISA 2022 Results (Volume II): Learning During – and From – Disruption, PISA, OECD Publishing, Paris, https://doi.org/10.1787/a97db61c-en.
[4] OECD (2022), PISA 2022 Database, OECD, Paris, https://www.oecd.org/en/data/datasets/pisa-2022-database.html.
[10] OECD (2020), OECD Digital Economy Outlook 2020, OECD Publishing, Paris, https://doi.org/10.1787/bb167041-en.
[1] OECD (2019), “PISA 2018 Well-being Framework”, in PISA 2018 Assessment and Analytical Framework, OECD Publishing, Paris, https://doi.org/10.1787/38a34353-en.
[5] OECD (2018), PISA 2018 Database, OECD, Paris, https://www.oecd.org/en/data/datasets/pisa-2018-database.html.
[8] World Bank (2020), Survive, Learn, Thrive: Strategic Human Capital Investments to Accelerate Azerbaijan’s Growth, World Bank, Washington, DC.
← 1. The EaP countries/economies forming part of this report are Baku in Azerbaijan, Georgia, Moldova and 18 of the 27 regions in Ukraine. Any reference to EaP countries/economies, as well as the EaP average, specifically pertains to Baku, Georgia, Moldova and Ukrainian regions. Armenia is also part of the EaP but has not yet participated in PISA, although participation is underway for PISA 2025.
← 2. Ukraine joined PISA for the first time in 2018. However, in the context of war, only 18 of the country’s 27 regions were able to participate in the 2022 assessment. Box 1.2, included in Chapter 1, provides detailed information on Ukraine’s participation.
← 3. Azerbaijan joined PISA for the first time in 2006 but has only participated with its capital city, Baku, since PISA 2018.
← 4. According to estimates based on PISA data, 20 score points is approximately equivalent to the typical annual learning gain by students around the age of 15 across OECD countries (Avvisati and Givord, 2023[11]).
← 5. These countries/economies were the following (in descending order of their mean score): North Macedonia, Albania, the Dominican Republic, the Palestinian Authority, the Philippines, Kosovo, Jordan, Morocco, Uzbekistan and Cambodia.
← 6. A difference in mean scores (or in other population-level estimates of performance in PISA) is called statistically significant if it is unlikely that such a difference could be observed when, in fact, no true difference exists in the populations from which student samples were drawn.
← 7. Azerbaijan as a whole country participated in PISA 2006 and PISA 2009.
← 8. PISA 2018 was not conducted in the break-away regions of eastern Ukraine, the Autonomous Republic of Crimea or the city of Sevastopol, all of which were outside the control of the government.
← 9. Again, a difference in mean scores is called statistically significant if it is unlikely that such a difference could be observed when, in fact, no true difference exists in the populations from which student samples were drawn. Statistical uncertainty in trend comparisons has three different sources: the sampling of students and schools; the design of PISA tests (measurement precision); and the use of a common scale to report the results of tests that were scaled independently. Link errors represent uncertainty around scale values (“is a score of 432 in PISA 2022 the same as 432 in PISA 2018?”). These three independent sources of uncertainty are combined in the estimates of standard errors for trend indicators. Standard errors are then used to construct “confidence intervals”, a range of values that excludes only 5% of the differences that would be observed in the absence of true change.
← 10. For a description of the levels of reading and science proficiency in PISA 2022, see Chapter 3 in OECD (2023[9]).
← 11. PISA’s Coverage Index 3 captures the proportion of the national population of 15-year-olds (enrolled and not enrolled in school) represented by the PISA sample. Low values of Coverage Index 3 may be attributed to 15-year-olds who are no longer enrolled in school or who were held back in primary school. Coverage Index 3 may also be lower due to student exclusions from the PISA test and dropouts during the school year.
← 12. In PISA 2022, values in the PISA index of sense of belonging are the following: -0.05 in Georgia; -0.06 in Moldova; -0.08 in Ukraine regions (18 of 27); and -0.17 in Baku. By contrast, the value of the index is and -0.14 in Estonia and -0.22 in Singapore.
← 13. Statistically significant changes of small magnitude in the BELONG index did occur between 2018 and 2022 (OECD, 2023[9]). Sense of belonging improved by 0.06 index points or less between 2018 and 2022 in Baku and Georgia. By contrast, sense of belonging decreased by 0.02 index points on average across OECD countries and by 0.06 index points in Singapore. The sense of belonging did not change significantly in Moldova between 2018 and 2022.
← 14. Values in the BELONG index declined by 0.30 in Georgia and by 0.10 in Moldova between 2015 and 2015. No change occurred on average across OECD countries for the same period. No data are available for Baku or Ukraine.
← 15. For Moldova, these data are reported missing.