Previous Section Next Section

2. The Education and Training 2020 targets

2.1 Early leavers from education and training (ELET)

2.1.1 Progress towards the EU target

The ‘early leavers from education and training’ (ELET) indicator (also named ‘early school leavers’) measures the proportion of 18-24 year-olds with, at most, lower secondary educational attainment (i.e. ISCED 0-2 levels) and who are no longer in formal or non-formal education and training. According to the Europe 2020 target⁴⁹, the ELET rate should be lower than 10% by 2020.

In 2019, the ELET rate stood at 10.2% in the EU-27, down from 14.0% in 2009 and very close to the 2020 target. The three countries with the highest rates are Spain (17.3%), Malta (16.7%) and Romania (15.3%). High ELET rates – more than 3 pps higher than the EU average – can also be observed in Bulgaria (13.9%) and Italy (13.5%). In 19 countries, the ELET rate is below 10% and is below 5% in Croatia (3.0%), Lithuania (4.0%), Greece (4.1%) and Slovenia (4.6%).

Countries that had low proportions of early school leavers in 2009 mostly continued to do so in 2019, with the exception of Slovakia, where the ELET rate grew from 4.9% to 8.3%. Spain improved the ELET rate by 13.6 pps while Greece improved the rate by 10.1 pps. Greece has now reached and even surpassed the target (standing at 4.1% in 2019). Despite their substantial progress, Portugal and Spain have still not attained the 10% target (recording rates of 10.6% and 17.3% in 2019, respectively), as shown in Figure 15.

A closer look at the percentages of early leavers by sex, country of birth and degree of urbanisation provides further insights. In the EU-27, fewer young women than young men leave education early (8.4% v 11.9% respectively), and this gap has remained broadly constant over the last decade. Also, on average in Europe, native-born people have lower ELET rates than foreign-born people (8.9% v 22.2% respectively). As regards the urban/rural divide, the pattern is more nuanced. At EU level, the ELET rate is lower in cities (9.1%) than in rural areas (10.7%) or towns (11.2%). The rural disadvantage is very strong in Romania and Bulgaria, where the difference between the ELET rate in rural areas and in cities is 18.1 pps and 16.0 pps, respectively. By contrast, in Austria (+7.3 pps), Belgium (+4.8 pps), Cyprus (+2.9 pps) and Germany (+1.3 pps), more young people leave education prematurely in cities than in rural areas (see Figure 16), although this could also be an effect of demographic ageing.

2.1.2 How many young people complete upper secondary education?

Having an upper secondary qualification is the minimum requirement for a satisfactory employment prospects in today’s economy, and a passport to full participation in society.

Eurostat publishes data on the share of people aged 20-24 with at least upper secondary education (ISCED 3 level), which corresponds to completion of upper secondary education (the ‘completion rate’). People aged 20-24 (instead of 18-24, as in the ET2020 ELET indicator) is the most appropriate age group, as the statutory age for completing most ISCED 3 education programmes is between 18 and 19 years old⁵⁰.

The main difference compared to the ET2020 ELET indicator is that the focus here is on completion of formal education. Therefore, a person with an ISCED 0-2 qualification and still in (either formal or non-formal) education/training would be treated as an early leaver according to a completion indicator, while they would not be considered as an early leaver in the ET 2020 ELET indicator. Besides, the completion rate measures how many (young) people in a cohort get an education at a certain level (relevant for a country’s economy and economic growth) whereas the focus of ELET is on the ability of the education system, or education institutions, to keep people that are already in education from dropping out. Even if both measures are a reverse of each other, due to the limitations of surveys, they do not lead to the same results yet the youth enrolled even in informal training course would not be counted in the ELET indicator which may distort the picture of ‘educational poverty’. In the new, post-2020 strategy, there will be a change of focus away from ELET, over to the ‘completion rate’.

In 2019, 83.5% of people aged 20-24 in the EU-27 had at least upper secondary education, with an increase of 4.8 pps since 2009. In most countries, when the completion rate is higher (lower) than the EU average, the ELET rate is lower (higher) than the EU average. There are, however, a few exceptions to this pattern. In Luxembourg and the Netherlands, both the ELET rate and the completion rate are lower than the EU average, while the opposite is the case in Bulgaria and Hungary.

2.1.3 A policy framework to tackle early school leaving

In 2011, the Commission and Member States developed a comprehensive policy framework to reduce early school leaving⁵¹. Its three pillars are prevention, intervention and compensation.

Prevention measures aim to reduce the risk of early school leaving before problems start. They may include – in addition to high-quality early childhood education and care – an early screening of language competence, development problems and special education needs that allows action to be taken at an early stage.

Intervention measures aim to avoid early school leaving by improving the quality of education and training at the level of the educational institutions, by reacting to early warning signs and by providing targeted support to pupils or groups of pupils at risk of early school leaving. They may include staff support, involvement of parents and local communities, and extra-curricular activities in the youth field. In addition, career guidance can play an important role in easing transitions between different education levels and between education and the labour market.

Compensation measures aim to help those who have left school prematurely to re-engage in education, offering routes to re-enter education and training and gain the qualifications they missed. They may include second chance education programmes, and various routes back into mainstream education and training, as well as recognising and validating prior learning.

A recent study⁵² shows that compensation measures are comparatively well-established across Europe. Most EU countries offer ‘second chance’ education schemes of some description, often combined with career guidance and financial, childcare and/or psychological support.

Intervention policies are also relatively widespread within countries, although with more variation. Coverage is highest for intervention measures focused on in-school support, including targeted support for learners experiencing personal, social or academic difficulties, as well as continuing professional development for teachers and school leaders to manage diversity. Implementation of infrastructural measures shows the weakest overall coverage, including measures relating to school networks, early warning systems, and extra-curricular provision.

Although coverage of prevention policies within countries is also fair overall, most countries have some gaps, and around a quarter have more marked gaps. As for intervention measures, there seems to be less emphasis on implementing systemic policies (e.g. anti-segregation policies) appears less prevalent than measures implemented within schools or other institutions (e.g. improving accessibility of early childhood education and care to all, developing curriculum flexibility and choice).

Educational attainment is a major factor in determining employment prospects for young people. Early leavers from education and training and those lacking basic skills have particular barriers to employability. Therefore, early leavers from education and training should be brought to the scope of the Youth Guarantee where they can be helped to return to education or training, or referred to other relevant services.

The Youth Guarantee can have a role in prevention, intervention and compensation of early leaving. This aspect is strengthened in the Commission’s recent proposal for a Reinforced Youth Guarantee⁵³, which recommends that Member States strengthen their early warning systems and tracking capabilities to identify those at risk of early leaving from education and training. This requires close cooperation with e.g. the education sector, parents and local communities, and the involvement of youth policy as well as social and employment services.

Success factors in tackling early school leaving include the existence of:

• a comprehensive strategy;
• a national coordinating mechanism or structure;
• a corresponding set of policy measures – prevention, intervention and compensation;
• an implementation plan, with clear targets and milestones;
• proportionate resources for implementation;
• synergies with other EU and national funding opportunities;
• clear lines of accountability;
• systematic monitoring, evaluation and feedback.

Some important challenges have been insufficiently addressed so far. For example: (i) integrating measures to tackle early school leaving within broader educational policies; (ii) specific targeting of measures at disadvantaged groups (e.g. migrants, ethnic minorities, or people living in remote areas); and (iii) monitoring and evaluation.

2.1.4 What socioeconomic factors influence early school leaving?

Besides specific policy measures, a number of contextual socio-economic factors can influence early school leaving. A study of EU-28 data from 2006 to 2017 provides some evidence on how strong those factors are in the EU.

The following variables were selected based on the literature on early school leaving⁵⁴ and data availability. The proportion of low-educated⁵⁵ women aged 45-54 is a proxy for low parental education. Research has consistently found low parental education to be as a good predictor of poor educational attainment and suggests that mothers’ education has a stronger impact than that of fathers⁵⁶. The unemployment rate for low-educated 15-24 year-olds (the ‘youth unemployment rate’) captures the cyclical labour market conditions and indicates how difficult it is to find a job for potential early school leavers. The higher the rate, the higher the incentive to stay in education or training, which should translate into a lower ELET rate⁵⁷. Expenditure per student in secondary education as a percentage of GDP per capita⁵⁸ measures the amount of financial resources that a country spends on each student compared to its level of economic development. In principle, one could expect that more spending helps to prevent early school leaving, but in practice there is no guarantee that additional resources are used to support measures against early school leaving. Therefore, it is not possible to draw any firm conclusion about the impact of this variable.

Being born abroad increases the risk of becoming an early school leaver, as shown in Section 2.1.1 above. The model accounts for this by including the proportion of foreign-born people aged 15-24. However, this variable is available for the sub-period 2009-2017 only. The ‘at risk of poverty or social exclusion’ (AROPE) rate for 15-24 year-olds may capture other family disadvantages that go beyond low parental education or a migrant background. It can also be a useful additional indicator of the impact of socio-economic background on early school leaving.

The strongest impact comes from low parental education (Figure 19), where a 1 percentage point increase is associated with a 0.4 percentage point increase in the ELET rate. In the sub-period 2009-2017, an increase of 1 percentage point in the proportion of foreign-born young people is associated with a 0.3 percentage point increase in the ELET rate. The other variables are not statistically significant.

These results provide suggestive evidence of the key role of parental background in shaping educational outcomes. As a policy message, they suggest that finding ways to compensate for adverse background may help education systems prevent early school leaving. One possible measure is to involve low-educated parents in the activities of their children’s school. This is consistent with the recent literature focusing on parental engagement and student outcomes⁵⁹. For instance, results from a large-scale randomised experiment in France show that a series of meetings targeted at parents of low‐achieving students towards the end of lower secondary education helped families choose a better-suited upper secondary educational programme. One year after this exercise, dropout rates fell from 9% to 5% and the probability of repeating a grade decreased from 13% to 9%⁶⁰. Such programmes must of course ensure to avoid unduly reducing students’ aspirations (e.g. as a result of unconscious bias) as this would risk reinforcing existing inequalities⁶¹.

Previous Section Next Section

2.2 Tertiary educational attainment (TEA)

Graduating from tertiary education has become increasingly important as an ever-changing European labour market needs more people with academic degrees who can easily upskill and re-skill, and thereby better contribute to economic and societal innovation. In addition, higher educational attainment is associated with higher earnings, lower unemployment risk, better health and more active participation in society.

Therefore, back in 2009 – in the context of the Europe 2020 strategy on promoting economic growth and employment – EU countries agreed to raise the share of people with an academic degree and set a headline target of bringing the number of 30-34 year-olds with tertiary educational attainment to at least 40% by 2020. Today, 40.3% of people aged 30-34 hold a tertiary degree, so the EU has reached the 40% target.

2.2.1 Gaps between and within countries

The overall achievement, however, hides substantial differences between countries and between regions within countries, for historical, structural, accessibility or other reasons. Among the countries with a low proportion of tertiary graduates, only Romania and Italy have not reached 30%. In 2019, 12 EU Member States showed TEA rates of 40% to 50%. In the Netherlands, Sweden, Ireland, Luxembourg, Lithuania and Cyprus, more than 50% of the population aged 30-34 holds a tertiary degree.

Sub-national TEA levels according to the degree of urbanisation⁶² show a clear qualification gap between cities, towns and suburbs, and rural areas in all Member States. While urban labour markets evidently attract more people with a tertiary degree in any country, the urban-rural TEA gap is remarkably country-specific.

In the EU, the average tertiary education gap between rural areas and cities today is more than 20 pps. Moreover, this urban-rural divide is larger than 30 pps in eight Member States (Luxembourg, Romania, Slovakia, Bulgaria, Hungary, Denmark, Lithuania and Poland), while in only two countries (Belgium and Slovenia) is it smaller than 15 pps. Unfortunately, the gap is growing as, in most countries, TEA levels are increasing faster in cities than in rural areas. For example, in 2009, there was no country with an urban-rural TEA gap over 30 pps and the largest difference was only 19 pps.

Sex and migrant status also seem to be important factors in the EU when it comes to higher educational attainment. Today, women’s tertiary educational attainment among 30-34 year-olds (45.6%) is on average more than 10 pps higher than men’s (35.1%). This gender difference has built up in the EU over the last two decades by continuously faster increasing female TEA levels. Interestingly, in 2019, for the first time in 20 years, the annual TEA level increase for males (1 percentage point) outperformed the annual increase for females (0.8 pps). It remains to be seen if the widening of the gender gap has indeed been halted, and whether it could even be reversed.

Regarding migrant status, EU citizens whether from the reporting country or not, have a higher percentage of tertiary level education than migrants (non-EU citizens). This currently stands at 41.1% for EU citizens from the reporting country, 37.3% for EU citizens who are not from the reporting country, and 34.4% for non-EU citizens. The tertiary education gaps between national citizens, foreign EU citizens and non-EU citizens have persisted at EU level over the last decade, given that the increase in TEA levels was of a similar order of magnitude for all three groups (around 8-10 pps).

2.2.1 Progress towards the EU target

Targets that measure and compare progress in the field of education are perceived in Europe as one of the most powerful tools for motivating national governments to drive their reform agendas and improve education systems. Looking back, the ET2020 target of 40% for tertiary educational attainment was a realistic objective 10 years ago and the target has been successfully met. This achievement gives reason to review the course of progress since the setting of the target in 2009.

Tertiary educational attainment has grown in each Member State since 2009, on average by 9.2 pps. However, the increase varied significantly, from only 1.4 pps in Finland to an impressive 22.5 pps in Slovakia.

Although countries also set national targets back in 2009, some more ambitious than others, those countries that started with a TEA rate below the EU target in 2009 generally seem to have substantially boosted their performance since then. The opposite seems to be true for the group of countries with a TEA rate above the EU target back in 2009, i.e. Belgium, Denmark, Ireland, Spain, France, Cyprus, Lithuania, Luxembourg, Finland and Sweden.

The comparison of the progress of TEA rates between 1999 and 2009 and between 2009 and 2019 shows that the annual increase slowed down across the EU in the last decade to 0.9 pps from 1 percentage point in the decade before. The main factor is that, in the group of countries that already had a TEA rate in 2009 above the EU target, the average annual increase measured by the indicator slowed down significantly from 1.3 pps in 1999-2009 to 0.5 pps in 2009-2019. In contrast, in the 18 countries with a TEA rate below 40% in 2009, the annual increase was, on average, 0.3 pps higher in the 10 years after 2009 compared to the 10 years before (1999-2009).

Moreover, the group of countries that were challenged by the EU-wide TEA target in 2009 have since made much better progress in fighting the gender gap than the nine countries that already had a TEA rate over 40% at that time. The comparison shows that, in countries with a TEA rate below the EU target in 2009, the share of male graduates has grown annually by 0.9 pps since then. This was more than twice the annual increase of the male TEA rate in the group of countries that were already performing above the TEA target when it was set.

The opposite was the case in the preceding 10 years (1999-2009), when male TEA rates grew half as fast in the group of 18 countries which were performing below the EU target of 0.5 pps a year that was set in 2009. The group of nine countries that had a TEA rate above 40% in 2009 reported an annual growth rate of 1 percentage point for male graduates in the same period.

While the evolutions of rising male TEA rates were almost completely reversed in the two groups of countries – those with overall TEA rates above 40% and those with overall TEA rates below 40% in 2009 – the same comparison for female TEA rates does not show such a clear pattern. The slowdown in the increase of female graduates from the decade before 2009 to the decade after was much more pronounced in the group with overall TEA rates above 40% in 2009 (1 percentage point) than the speed of increase of female graduates in the group with TEA rates below 40% in 2009 (0.2 pps) for the same period.

Despite these promising trends, the widening of the gender gap from 0.4 pps in 1999 to up to 10.7 pps in 2018 has only stopped recently. The reason is that the increase of female TEA rates has continuously outperformed the increase of male TEA rates in the course of the last 20 years, no matter what the overall national TEA rate was. In 2019, the EU finally reported a slight narrowing of the gender gap, with an overall distance between male and female TEA rates of 10.5 pps.

The outcomes of the analysis could give reason to speculate that the share of people with an academic degree will reach around 50%, a saturation point in today’s developed societies. In particular, the evolution of female TEA rates would suggest so. The consequence would be that the closer education systems get to a population where every second person has a university degree, the lower would be the increase of TEA rates, which would eventually stagnate around 50%.

The ET2020 target of 40% for TEA may have played a role in supporting national efforts to succeed and make visible and measurable progress in increasing the number of graduates from tertiary education.

Previous Section Next Section

2.3 Early childhood education and care (ECEC)

2.3.1 Evolution of the early childhood education and care target

Promoting social integration of children and levelling educational opportunities through ECEC remain key objectives of the European policy agenda, and European policymakers have set several objectives to increase participation. In 2002, the Barcelona European Council⁶⁴ set two targets on the availability of high quality and affordable childcare facilities for pre-school children, i.e. 90% of children from the age of 3 until compulsory (primary) school age and 33% of children under 3 years-old. In 2009, the strategic framework for European cooperation in education and training set the scene for the ECEC target, aiming at participation of ‘at least 95% of children between 4 years-old and the age for starting compulsory primary education’⁶⁵. To be classified as early childhood education for the ECEC target, early childhood education and care services must include a certain minimum level of instruction activity – they cannot simply be formal childcare facilities as specified by the Barcelona target.

The ECEC target comprises participation in any childcare provision for children from the age of 4 through to compulsory primary education that falls within a national regulatory framework, classified as ISCED level 0⁶⁶ and included in reporting⁶⁷. Many differences apply to the age when education becomes compulsory⁶⁸ in different Member States; the compulsory starting age for primary education is generally around the age of 6 in Europe⁶⁹.

2016 marks the year in which the ECEC target was officially reached: 95.3% (94.9% in the EU without the United Kingdom) of children between the age of 4 and the age of starting compulsory primary education participated in ECE⁷⁰. From 95.4% in 2017, the EU-28 percentage dropped very slightly to 95.3% in 2018. Post-Brexit, the EU-27 average is similar, even though the numbers remain just below 95%: from 94.9% in 2016 and 2017 to 94.8% in 2018.

In 2018, 15 Member States had reached the target. In descending order, these were Denmark, Ireland, France, the United Kingdom (still a Member of the EU at that time), Belgium, Spain, the Netherlands, Luxembourg, Germany, Latvia, Austria, Sweden, Hungary, Cyprus and Malta. Of the countries with a participation rate below 95%, Finland, Poland, Slovakia and Slovenia saw an increase compared to 2017 of between 1 and 4 pps. While the decrease of most of the others remained below 2 pps, two countries took a bigger step backwards: Greece (6.3 pps; this is due to a break in the time series) and Romania (3.3 pps). Seven countries were close to the target, with rates between 91% and 95%: in ascending order, these were Lithuania, Czechia, Estonia, Poland, Slovenia, Portugal and Italy (which, having had a participation rate above 95% in the past few years, slipped just below the target in 2018). Greece (because of breaks in the data)⁷¹, Croatia, Slovakia and Bulgaria have the lowest participation rates among EU countries.

Of the countries with an attendance rate above 95%, Denmark, Ireland, France and the United Kingdom had a full participation rate, while participation among children from 4 years of age upwards is also close to 100% in Belgium and Spain. Most other Member States above 95% showed slight changes in the percentage from 2016 to 2018 (around 1 percentage point or less). Cyprus stands out, with a participation rate that went from 89.7% in 2016 to over 92% in 2017 and 95.3% in 2018⁷², as does Croatia, where the percentage of children participating in early childhood education increased from 75.1 (2016) to 81.0% (2018).

When considering a wider group, of children from 3 years of age upwards, Greece, Croatia, Slovakia and Bulgaria have participation rates in ECE below 80%. The highest participation rates, all over 95%, are seen in Ireland, France, and the United Kingdom (all three with 100%), Denmark, Belgium, Spain and Sweden. With a rate of 92.2%, the EU-27 average still has room for improvement.

Expanding the age range even further, to all children from 0 up to the start of compulsory primary education, allows us to look at the ECE system as a whole. Participation rates in the overall group are much lower than in the older age categories of 3 and above and 4 and above⁷³. Malta and Slovakia have the lowest participation rates, while Belgium, Sweden and Denmark are at the other end of the range, with participation rates above 70%. The EU-27 average for this age group is 57.10%.

Research suggests that integrated ECEC systems – i.e. where care and education are considered as a whole – offer more continuity, quality and coherence across ECEC policy (e.g. regulation and funding, curriculum, workforce education/training and working conditions, monitoring and evaluation systems) and allocate more resources to younger children and their families. In addition, the literature suggests that unitary systems, which cover the whole age range, from birth up until the start of primary education, lead to better quality, more equitable service provision and result in greater financial efficiency⁷⁴. Countries that provide integrated ECEC services for all children under primary school age were likely to guarantee a place in publicly funded provision for each child from an early age (6 to 18 months) and contribute to high standards across all ECEC services⁷⁵.

2.3.2 Inclusion/equity in access to early childhood education and care

Scientific literature has long shown that children participating in qualitative ECEC enjoy long-term emotional, social and cognitive benefits⁷⁶. More study is needed, however, on the impact of policy changes to scale up universal ECEC for very young children and especially on the links between scaling up and the quality of this process.⁷⁷ To benefit children’s early development and subsequent school performance, labour market participation, social mobility and social integration, ECEC needs to be of a high quality⁷⁸.

The beneficial effects of participation in ECEC seem to be especially pronounced for children from disadvantaged backgrounds⁷⁹. ECEC broadens the educational experiences of children, has been shown to facilitate access to employment and may have a positive impact on parental aspirations and behaviour.

Administrative enrolment data currently does not track any information on children’s socio-economic background, so it is not possible to track the extent to which vulnerable children are participating in ECEC services with a minimum educational component. It is clear from participation rates in the wider formal ECEC sphere (which includes services both with and without a minimum educational component). It is clear from participation rates in formal ECEC, however, that there is a clear tendency towards lower participation rates among children from a lower socio-economic background than for those from a higher one, i.e. a social gap in ECEC attendance, which is evident in several Member States⁸⁰. In many European countries, children from socially disadvantaged groups do not fully enjoy the benefits of ECEC. Recent OECD analysis⁸¹ reveals that several countries continue to struggle with equity issues regarding 0 to 2 year-olds’ participation rates in ECEC. Approximately half of the countries in the most recent available OECD Family Database (2017 or latest)⁸² show considerable differences in access depending on the income range of the families.

Recognising the importance of all children having access to basic services, the European Commission 2020 work programme announced the development of a European Child Guarantee, to be included in the next EU budget (2021-2027). Also, the Council Recommendation on High-Quality Early Childhood Education and Care Systems advocates the improvement of inclusiveness of ECEC⁸³. In order to reach as many children as possible, many countries have worked over the past 5 years on extending a legal entitlement to ECEC or introducing compulsory ECEC of at least 1 year before primary education⁸⁴. Most EU countries guarantee a place from a certain age, but only seven EU Member States (Denmark, Germany, Estonia, Latvia, Slovenia, Finland and Sweden) guarantee a place in ECEC for each child from an early age (6-18 months), often immediately after the end of childcare leave. From the age of 3, a place in publicly subsidised ECEC is ensured in Belgium, Czechia, Spain, France, Luxembourg, Hungary and Poland. ⁸⁵

Like availability, affordability contributes to ensuring access for as many children as possible. In EU, most families have to pay ECEC fees for the youngest children⁸⁶. The older the children, the more countries provide ECEC free of charge for everyone. For children aged 3 or older, almost half of European countries offer a form of free ECEC. During the last year before compulsory primary education, free places are almost universal in Europe. Many countries provide targeted measures regarding availability or affordability, to facilitate ECEC access for children living in poverty. Price reductions (through lower fees and/or free meals) and priority admission are the most common measures for young children. Measures tackling inequalities by increasing affordability, through fee reductions, are more common than those that increase availability. Targeted groups are children living in poverty, children of single parents, children whose parents’ work situation puts them at a disadvantage, number of siblings, children with disabilities/difficulties (SEN), children from migrant backgrounds and those from regional or ethnic minorities.

Previous Section Next Section

2.4 Underachievement in basic skills in the digital age

2.4.1 The 2018 PISA study

The strategic framework for European cooperation in education and training (ET 2020) set a 15% target for 15 years-olds’ underachievement⁸⁷ in reading, mathematics and science. The results from PISA 2018 show that the majority of EU Member States perform below the ET2020 target. As highlighted in the sections below, since 2009, the EU share of underachievement has increased in both science and reading, while remaining stable in mathematics. More than one in five 15 year-olds in the EU cannot complete even simple tasks in the three subject areas tested under PISA. Specifically, PISA 2018 shows that 22.5% of pupils in the EU-27 underachieve in reading⁸⁸, 22.9% in mathematics and 22.3% in science. Underachieving in basic skills implies not being equipped to thrive in the labour market and the broader society. Therefore the cost of underachievement is significant both for the individual and for society at large.

2.4.2 Underachievement in reading

Reading literacy was a PISA 2018 core subject⁸⁹ and reading performance shows a large variation between EU Member States (Figure 27).

Four EU countries met the 15% ET2020 target for underachievement: Estonia (11.1%), Ireland (11.8%), Finland (13.5%) and Poland (14.7%) and were in fact among the 10 best OECD performers globally⁹⁰. Within the EU, Denmark was just above the target (16.0%). By contrast, the underachievement rate exceeded 30% in Malta (35.9%), Slovakia (31.4%) and Greece (30.5%), and even 40% in Bulgaria (47.1%), Cyprus (43.7%) and Romania (40.8%).

Since 2009, performance has not significantly changed in most countries, signalling that no improvement has been registered since the ET2020 framework was adopted (Figure 28).

In eight countries (the Netherlands, Slovakia, Greece, Hungary, Finland, Latvia, Belgium and Luxembourg), the underachievement rate increased in a statistically significant way. Only Ireland and Slovenia experienced a statistically significant decline. Overall, EU reading performance deteriorated, with an average underachievement rate of 19.2% in 2009.

2.4.3 Underachievement in mathematics and science in the EU

The 2018 pattern of underachievement in mathematics is similar to that of reading.

Four countries met the 15% ET2020 target: Estonia (10.2%), Denmark (14.6%), Poland (14.7%) and Finland (15.0%). Ireland (15.7%), the Netherlands (15.8%) and Slovenia (16.4%) were just above the target. The underachievement rate exceeded 30% in Romania (46.6%), Bulgaria (44.4%), Cyprus (36.9%), Greece (35.8%), Croatia (31.2%) and Malta (30.2%). Performance remained rather stable in many Member States between 2015 and 2018.

A slight majority of countries experienced a decline in the underachievement rate, but it was statistically significant⁹¹ only in Cyprus (-5.7 pps) and Latvia (-4.1 pps). The only statistically significant increases took place in Romania (+6.6 pps) and Germany (+3.9 pps). Consequently, the EU average, at 22.9%, remained stable compared to 2015, when it stood at 22.2%.

Similarly, underachievement in science also shows a mixed picture across EU countries.

Four countries met the 15% ET2020 target in 2018: Estonia (8.8%), Finland (12.9%), Poland (13.8%) and Slovenia (14.6%). In contrast, the underachievement rate was higher than 30% in Bulgaria (46.5%), Romania (43.9%), Cyprus (39.0%), Malta (33.5%) and Greece (31.7%). In a few Member States the underachievement rate had increased in a statistically significant way since PISA 2015 (+8.6 pps in Bulgaria, +3.0 pps in Spain, +2.8 pps in Denmark), while Cyprus and Poland experienced a statistically significant decline (-3.2 pps and —2.4 pps, respectively).

2.4.4 Top performers

PISA also provides an important insight into top performers’ share. This indicator captures the extent to which a school system can produce excellent results in basic skills. Top performers are pupils who reach PISA Level 5 or above. For instance, top performers in reading are able to distinguish between facts and opinions, while the ability to discern the source of information is emerging as a crucial skill in the digital age. PISA 2018 has shown that top performers in reading ranged from 14.2% in Finland to 1.4% in Romania.

In only six countries did the proportion of top performers exceed 10%: Finland (14.2%), Estonia (13.9%), Sweden (13.3%), Poland (12.2%), Ireland (12.1%), and Germany (11.3%).

For mathematics, the proportion of top performers is somewhat higher than for reading in most countries.

In the Netherlands (18.4%), Poland (15.8%), Belgium (15.7%) and Estonia (15.5%), more than 15% of pupils are top performers. Compared to 2015, this percentage increased significantly in Poland (+3.6 pps), Latvia (+3.3 pps), the Netherlands (+2.9 pps), Slovakia (+2.9 pps), Czechia (+2.3 pps), and Cyprus (+1.2 pps), while it decreased significantly in Malta (-3.4 pps).

In science (Figure 34 below), the proportions of top performers are the lowest among the three subject areas. The countries with the highest proportions are Finland (12.3%), Estonia (12.2%), the Netherlands (10.6%) and Germany (10%). In many countries, the percentage decreased between 2015 and 2018. This decline was statistically significant in Slovenia (-3.3 pps), Malta (-3.2 pps), Finland (-2.1 pps), Portugal (-1.8 percentage point), Luxembourg (-1.5 percentage point), Bulgaria (-1.4 pps), Italy (-1.3 pps) and Greece (-0.8 pps). No country experienced a statistically significant increase.

2.4.5 Underachievement by sex

As in previous PISA cycles, and in line with the OECD global trend, girls significantly outperform boys in reading in all EU countries⁹². The gap in underachievement between boys and girls ranges from 6.6 pps in the Ireland to 21.5 pps in Cyprus. The EU average is 27.3% for boys and 17.4% for girls (Figure 35). Specifically, the gender gap increased by 1.7 pps at EU level between 2015 and 2018⁹³.

No innate gender-related ability explains gender differences in reading literacy. These differences rather depend on the social and cultural context, pupils’ non-cognitive abilities (motivation and self-esteem), and gender stereotypes that translate into parents’, teachers’ and pupils’ gender-oriented expectations⁹⁴. These factors play their role as early as during the first grades of primary education⁹⁵. Disengaged adolescent boys suffer from a lack of male role models, both in school and outside. In European schools, women account for the large majority of teachers. Outside schools, boys may perceive reading as a female activity, not fitting a young man’s self-image⁹⁶. Attracting more men into the educational professions, and promoting reading styles that are appealing to boys and that involve male reading partners, are all effective measures to close the gender gap in reading⁹⁷. Finally, it is important to highlight the potential of integrating of digital technologies in European curricula in closing the gender gap. In fact, evidence suggests that the combination of digital tools, social interaction and formative feedback effectively reduces both learning gender gaps and underachievement trends in literacy (and mathematics)⁹⁸.

The picture in mathematics is more mixed than in reading. The differences between boys and girls are much smaller than in reading and vary from country to country. Only a few countries stand out as having statistically significant differences: girls perform better than boys in Malta (8.2 pps), Cyprus (6.0 pps), Finland (3.8 pps) and Lithuania (3.6 pps), while the opposite is the case in Belgium (2.7 pps). Underachievement at EU level is similar among girls (22.9%) and boys (22.8%), while in 2015 boys still outperformed girls⁹⁹

The picture for science is quite similar to mathematics (Figure 37). Gender differences are rather small, with the proportion of underachievement generally higher among boys than girls. This gender gap (in favour of girls) is statistically significant in Cyprus (10.7 pps), Malta (10.2 pps), Bulgaria (7.8 pps), Finland (7.7 pps), Greece (6.3 pps), Latvia (5.1 pps), Lithuania (5.0 pps), Slovenia (4.4 pps), Slovakia (3.5 pps), Sweden (3.5 pps), the Netherlands (3.2 pps), Denmark (3.1 pps), Germany (2.6 pps) and Poland (2.2 pps). At EU level, the advantage of girls over boys stood at 2.0 pps in 2018, with an increase of 1.6 pps compared to 2015.

2.4.6 Pupils’ performance and socio-economic context

Education systems can be one of the main drivers in breaking negative social heritage and equipping pupils with the skills necessary to achieve their full potential in life. However, this does not happen in most EU Member States, where socio-economic background is a strong predictor of educational attainment. In PISA, pupils' socio-economic background is estimated by the PISA index of economic, social and cultural status (ESCS)¹⁰⁰, which is based on information about the pupils' home and background. As Figure 38 shows, the proportion of underachievers in reading in most countries is much larger in the bottom quarter of the ESCS index compared with pupils in the top quarter, rising to more than 40 pps in Romania and Bulgaria.

On the other hand, some countries seem better able to counter the impact of socio-economic background on the educational success of pupils: for example, Estonia, Ireland, Finland, Poland, Croatia and Latvia. Overall, countries with a low share of underachievers in reading also tend also to have a smaller difference in the proportions of underachievers at the top and bottom of the ESCS scale. Cyprus is an exception to this pattern. It has a very high share of underachievers, but socio-economic background seems to have a smaller impact on educational attainment compared to other similar Member States.

Addressing underachievement among socio-economically disadvantaged pupils is key to improving the overall performance of EU education systems. This requires a concerted effort involving many actors and resources. Any successful strategy should start from early childhood education and care. In fact, social inequalities affect pupils’ academic outcomes from the early stages of their schooling. Lack of intervention in the early years will likely widen the performance gap throughout school, eventually resulting in underachievement and lack of social mobility across generations¹⁰¹.

2.4.7 Pupils’ performance by migrant background

The proportion of underachievers in reading¹⁰² among pupils with a migrant background is much higher than for pupils without a migrant background in many EU Member States¹⁰³. Not speaking the language of instruction at home can play a negative role in the reading performance of pupils with a migrant background, to a greater extent than for the other two tested subjects. The situation is usually worse for pupils born abroad (their underachievement rate exceeds 50% in Greece, Germany, the Netherlands and Sweden) than for native-born pupils with parents born abroad¹⁰⁴. Greece has the highest underachievement rate in the EU among foreign-born pupils (58%), while Germany has the widest gap in underachievement rates in reading between pupils born abroad and pupils without a migrant background (40 pps).

Being born and growing up in the country of assessment is an advantage compared to moving there as a child or as a young person. It may help with learning the language of instruction and getting familiar with the country and its education institutions, but it is not usually sufficient to reach the same levels as pupils with a non-migrant background. However, patterns are quite different among EU Member States. A few countries (Germany, Sweden, Slovenia, France and Estonia) face a large gap between pupils born abroad and non-migrant pupils, but native-born pupils with parents born abroad largely catch up. In Finland, Austria, the Netherlands and Greece there is some catching up, but the gap remains wide also between native-born pupils with parents born abroad and non-migrant pupils. In countries like Italy, Denmark and Luxembourg there is little variation between the two groups of pupils with a migrant background. Finally, only in Ireland, Croatia, Latvia, Malta and Cyprus are the differences small between both groups with a migrant background and pupils with a non-migrant background. A possible explanation is the specific composition of migrant populations in those countries (related to e.g. knowledge of the language of instruction or cultural similarities).

Member States can use a variety of education policies to promote inclusion of migrant pupils, ranging from language support for pupils whose mother tongue differs from the language of instruction, to education and career guidance, to increasing the flexibility and permeability of educational pathways. Participation in high-quality ECEC is crucial for achieving better educational outcomes. It is also important to promote a culture of inclusion in schools where diversity is increasing, and the availability of high quality resources and extracurricular activities has proved beneficial in this respect. Finally, equipping teachers with the skills they need to teach multicultural and multilingual classrooms requires appropriate initial teacher education and continuing professional development¹⁰⁵.

2.4.8 The urban-rural divide

PISA 2018 shows that the difference in reading performance between pupils attending schools in cities and those enrolled in schools in rural areas¹⁰⁶ is statistically significant and rather large in many Member States. In Hungary, Bulgaria, Romania, Slovakia and Portugal it even exceeds 100 PISA score points, corresponding to approximately 3-4 years of schooling.

Schools in rural areas often struggle to provide quality education due to their geographical isolation and small size, which increase the risks of suffering from insufficient infrastructure, a limited educational offer and a lack of experienced teachers. Policies to counter these risks may include adjusting the school network, making effective use of technology and better preparing teachers and school leaders to work in rural locations¹⁰⁷.

Previous Section Next Section

2.5 Employability of recent graduates

2.5.1 The target on the employment rate of recent graduates

While the employment rate of recent graduates (80.9% in 2019) has still not reached the ET2020 target of 82%, it was still at its highest level since 2008 (it stood at 81.8%, just before the financial crisis). Compared to 2010, the situation has improved in most countries, particularly Poland, Hungary, Ireland, Slovakia, Estonia and Latvia – all of which have surpassed the EU target value of 82%. That said, the situation is likely to worsen in the wake of the COVID-19 crisis. However, the situation in Spain, Greece and Italy has improved only to a limited extent, while it has deteriorated in France and Portugal. Separating employment rates of higher education graduates from other graduates gives allows us to see more change over time. There was a positive evolution of the employment rate among recent graduates from higher education (ISCED 5-8), from 77.8% in 2016 to 85.3% in 2019. The employment rate of recent higher education graduates is 85.2%, being above the EU-28 average (85%) and currently representing 9.5% more than in 2016.

Employment rates tend to be linked to education level as well as labour market relevance of particular programmes. Among recent graduates with different levels of education (as compared to the average employment rate of young adults aged 20-34 who are not in further education or training), it is evident that those with only a medium-level generalist qualification, compared to all recent graduates, suffered a significant employment penalty in all countries except Slovakia and Finland. In contrast, those with a high-level of qualification had a higher employment rate compared to the average employment rate of all young adults in almost all countries except for Greece. A more diverse picture emerges for employment rates of recent graduates with a medium-level vocational (VET) qualification. On average in the EU-27, recent VET graduates had a slight employment rate premium of 1.7 pps compared to the average employment rate of young adults not in education, with the highest such premium in Luxembourg (11.9 pps), Germany (10.6 pps) and Slovakia (8.1 pps). In some other Member States, though, the recent VET had a lower employment rate than young adults in general. This is the case in Greece (—15.6 percentage points), Latvia (—15.5) and Lithuania (—15.4). This pattern is also found in Cyprus, Romania Portugal, Spain, Slovenia, Italy, France, Croatia, Belgium, Bulgaria, Ireland and Poland. These employment penalties signal challenges in the VET systems

On the back of a broad improvement in economic condition, the EU-27 employment rate of all three main categories of recent graduates increased between 2014¹⁰⁸ and 2019 by 5 pps, though the employment rate of recent graduates with a medium-level vocational qualification increased the most. In all countries, except Malta, the employment rates of recent tertiary graduates have increased. However, in Czechia, Austria, France, Luxembourg and Estonia, the employment rates of recent graduates with a medium-level general qualification have declined, while in Latvia, Lithuania and Malta this was the case for recent graduates with a medium-level vocational qualification. Together with the data from the previous analyses, this suggests that medium-level VET systems in Latvia and Lithuania are facing particular challenges and that the situation has further deteriorated in recent years.

One should remember that the size and composition of learners at the different levels of education, particularly vocational education and training, as well as tertiary education, differ significantly between countries. This probably also has a strong impact on, or relationship to, graduates’ employment prospects. In a situation where in one country medium-level VET (ISCED-3) is restricted only to a very narrow set of occupations and targeted primarily at disadvantaged learners compared to another country where medium-level VET prepares learners for a broad range of occupations, attracting well performing learners, this would be reflected in the employment rates of graduates, but may not fully reflect the quality of such programmes when taking into account their purpose, the profile of targeted learners or targeted jobs. It is often the case that in some countries a certain occupation – i.e. nursing – requires a medium level of education, while in others it requires a high level of education. Even though data on which programmes prepare learners for which occupations is very limited, certain trends can be observed when analysing the data on graduates by detailed education level and programme orientation. Two of the countries that stood out in the above analysis in terms of less positive performance by their medium-VET systems (Lithuania and Latvia) indeed have, since 2013, progressively reduced the provision of VET at upper secondary level and increased the provision at higher levels, which suggests that changes in learner composition may partly explain the drop in employment rates.

2.5.2 Young adults’ labour income and its distribution

Figure 43 below shows the wage¹⁰⁹ distribution by educational level¹¹⁰ in several EU Member States in 2018 and provides for three main observations. Firstly, workers with a higher level of education earn more than those with a lower level of education. Secondly, several countries (e.g. Germany, Luxembourg, Austria and the Netherlands) display a larger absolute earnings premium between the highest and lowest education groups. Thirdly, in many countries (e.g. Cyprus, Spain, France, Luxembourg and the Netherlands) there is considerably more wage dispersion among the higher educated than among the low and medium educated.

This means the labour market returns for tertiary education may differ between fields of study and that graduating from a prestigious institution may be associated with a larger wage premium. Furthermore, productivity and wages are also affected by individual traits such as non-cognitive skills, resilience, work ethic, unobservable cognitive skills, etc. In a number of countries (e.g. Denmark, Sweden, Austria and Italy), the top-half (in terms of wages) of medium-qualified workers earn more than the bottom-quarter (in terms of wages) of the highly qualified workers.

In most countries, upper secondary education is split into general and vocational education tracks. While vocational education gives students specific job-related skills, general education provides students with broad knowledge and basic competences that are required for more advanced educational programmes. Austria there is no clear hierarchy between general and vocational education; most VET programmes open pathways to advanced educational programmes. Figure 44 shows the full-time equivalent gross monthly wage distribution by type of upper secondary education¹¹¹ in several Member States in 2018¹¹². In some countries (including Bulgaria, Hungary and Czechia), the wages associated with vocational and general education are practically the same in the 25th, 50th and 75th percentiles of the distribution of earnings. In Sweden, Finland, Austria¹¹³ and Denmark, vocational education appears to pay off more than general education, though the size of the wage difference is largest in Denmark. In these countries, vocational learning takes place at school and in the workplace as part of the programme. An advantage of this system is that it helps students acquire especially marketable skills, since firms may exert a direct influence on the content and organisation of vocational training. On the other hand, in the Netherlands and Luxembourg workers with general education at the top of the wage distribution are found to earn more than their counterparts with vocational education. Finally, while the above figures focus on young adults, some studies demonstrate that the employment premium, enjoyed by VET graduates compared to general education graduates in a number of countries, becomes less pronounced at a higher age.

Previous Section Next Section

2.6 Adult learning

2.6.1 The target on adult learning

Recognising the value of learning not only during early years, but also throughout a person’s working and post-working life¹¹⁴, the EU has sought to increase the frequency and extent to which adults engage in learning. To monitor progress and set a target, the Council in 2009 reinforced the target on adult learning, adopted earlier as part of the Lisbon agenda in 2003. This raised the level of ambition to reach, by 2020, an adult learning participation rate of 15% on average in the EU (up from the earlier 12.5% target).

However, progress towards that target has been slow and uneven. The participation rate of adults in learning, increased from 7.8% in 2010 to 10.8% in 2019 on average in the EU-27, which is a sizeable relative increase yet still well below the target.

However, there are limitations to the indicators. For example, the way participation in adult learning is currently measured is quite restrictive, as it measures ‘the share of population who report having participated in any formal or non-formal learning activity during the last 4 weeks prior to being interviewed’. This is problematic in the context of adult learning which is still a rather sporadic activity, often taken-up only once or at most twice a year for a short duration.. Looking at available data on learning activities during the last 12 months, the average duration of a non-formal learning activity (the most prevalent type of learning during adulthood) in 2016 was 75 hours, or around 10 working days¹¹⁵.

Consequently, due to the low frequency of learning activities, the measurement of adult learning by using a 4-week reference period may not be conducive to monitoring with sufficient precision participation in learning. Also, a recent review by the Joint Research Centre¹¹⁶ indicates, that while such a measure may indeed be appropriate to indicate the average number of persons who have had a recent exposure to adult learning, it may not be sufficient to represent the average participation rate over a longer time period (such as a calendar year), which is nevertheless how this indicator is most often being defined and interpreted.

Therefore, a more comprehensive measure of adult learning has been included in the EU Adult Education Survey (AES), registering participation in learning over a longer – a 12-month reference period and includes a broader definition of adult learning¹¹⁷, i.e. also covering guided on the job training (for more details on different measures of adult learning, please see earlier editions of the Education and Training Monitor, particularly the 2018 edition¹¹⁸). When analysing this more comprehensive measure, a significantly larger proportion of the population is reported to participate in learning. This was also the case when analysed over a shorter period (over 5 years between 2011 and 2016). At the same time, relative differences between Member States are smaller than in the LFS survey; in most cases we are dealing here with differences of at most two and a half times¹¹⁹. Still, the low frequency of the EU Adult Education Survey (every 5 years before 2016 and every 6 years since 2016) does not allow for a more regular monitoring of adult learning through this more inclusive indicator.

A broader definition of non-formal education is also included in the EU Continuing Vocational Training Survey (CVTS). As in the AES results, CVTS does not show in general such large differences between countries as are seen in the LFS results (with the exception of some forms of learning at the workplace). This could be another argument for the incomplete identification of non-formal education in the LFS survey. This is especially important because adult participation in education and training depends mainly on participation in non-formal education and training. Participation in formal education is low (a little above 3% in the 4 weeks before the survey) for many years in LFS, and almost 6% (in the 12 months before the survey) in AES 2016, with a downward trend since 2007. Participation in non-formal education and training is significantly higher, and in LFS it is just above 8% in the last few years, and in the AES 2016 it reached 42.6%, showing an upward trend since the AES 2007 (31.6%). Given that the AES and CVTS have a wider coverage of the adult learning forms, these surveys seem to provide a more realistic picture on the learning of adults than LFS survey. It is likely that the LFS results do not identify the entire spectrum of non-formal education in EU, as in many countries the process of this identification is not completed, and in some of them this process has not yet started.¹²⁰ Still, the low frequency of the AES (every 5 years before 2016 and every 6 years since 2016) and EU Continuing Vocational Training Survey (every 5 years since 2005) did not allow for regular monitoring of adult learning through this more encompassing indicator.

2.6.2 Improving the measurement of adult learning

The Commission has been working towards better data on adult learning. In particular, in the context of developing an integrated framework for European Social Statistics (IESS)¹²¹, and the revised legal base of the EU Labour Force Survey (LFS)¹²², adopted at the end of 2019. The latter provides for a more regular (every two years) collection of data on participation in adult learning with a 12-month reference period, starting in 2022. As of 2023 (the first year of data release), this will allow Member States to report every second year on participation in adult learning.

At the same time, this new data will not be identical to the data collected via the AES. Apart from general differences between the two surveys, the newly collected data via the LFS will exclude one specific form of non-formal adult learning: guided on the job training (GOJT). GOJT is characterised by planned periods of training, instruction or practical experience, using normal tools of work, either in the immediate place of work or in a work situation with the presence of a tutor¹²³. It is possible, from the data available in the AES, to estimate the prevalence of GOJT as part of adults’ total participation in learning and thus calculate participation rates, which should be more comparable with the forthcoming data from the LFS in 2022.

In 2016, in the EU-27, the participation rate of adults in learning, excluding GOJT, was 37.9%, with GOJT representing an additional 6.5% participation rate (i.e. the share of adults who say they have been participating in learning through GOJT but not any other types of learning). The largest share of GOJT was reported in Czechia (23.3%, more than all other types of learning combined), Spain (13%) and Bulgaria (12.8%). In most other countries the participation in GOJT represented a more limited part of total adult learning.

When analysing changes in the participation rate over time (i.e. during the two latest waves of the AES in 2011 and 2016) and separating GOJT, a rather mixed picture emerges among Member States. On average in the EU-27, the share of individuals participating only in GOJT remained stable, while participation in other forms of non-formal and formal learning increased. Major changes in Ireland and Luxembourg might have been (at least in part) driven by methodological changes. Otherwise, there was a major increase in GOJT in Czechia and a decrease in a number of Member States, including Slovakia, Cyprus, Denmark and Bulgaria. A decline in other forms of learning was also registered in Sweden (also possibly due to methodological changes), Czechia, Estonia and Denmark. In relation to the remarks included in the above-mentioned report of the ET 2020 Working Group on Adult Learning 2018-2020 ‘Achievements under the Renewed European Agenda for Adult Learning’, it is also important to analyse in more detail the differences between EU countries in identifying adult participation in non-formal education in three European surveys: LFS, AES and CVTS (it is also worth considering the Eurofound surveys). These differences are critical to national differences across adult learning and to the achievement of the EU's ambitious adult learning targets. Particularly in-depth analyses should concern practical forms of learning, such as learning at the workplace and learning during the implementation of social projects and activities, including those necessary, such as extending the functions of schools children and youth with non-formal education offers for adults in local surroundings and deinstitutionalization of forms of care for older seniors.

2.6.3 The profile and quantification of adult learners and non-learners

The two key socio-demographic factors influencing adult learning are employment status (and sector of employment) and qualification level. From this, it is clear that participation of employees in the private sector, particularly medium-qualified employees, is likely to have the most significant impact on adult learning participation rates, due to their large number and moderate participation rates. The second significant group are inactive adults, which is the second-largest population group as well as the group with the lowest participation rate in the EU. Indeed, these insights are confirmed when the number of non-learners are quantified in each of these socio-demographic groups.

The largest groups of non-participating adults in 2016 were medium-qualified private sector employees (27.1 million), medium-qualified inactive adults (16.6 million), low-qualified inactive adults (14.5 million) and low-qualified private sector employees (10.7 million). These four groups (covering nearly 70 million individuals) together represent more than 50% out of all (nearly 130 million) adults aged 25-64 who did not participate in any formal or non-formal education or training activity that year. Beyond the total EU-27 figures, it is also relevant to look at the distribution of non-learners across EU Member States.

In line with the previous analysis, the data indicates that the two largest groups of non-participants on average across the EU are private-sector employees (34.8%) and inactive adults (26.4%). Overall, employed individuals represent more than half (56.2%) of all non-participants. It is also possible to further disaggregate this data, by looking, for example, at the largest group of non-participants (private sector employees) and analysing their composition in terms of educational attainment. Such analysis indicates that, out of all private sector employees in the EU-27 who did not participate in any formal or non-formal learning activity in 2016, the largest proportion were medium-qualified private sector employees (representing 20.2% of all non-learners), followed by low-qualified non-learners (representing 8.0% of all non-learners) and high-qualified non-learners (representing 6.7% of all non-learners). However, some countries stand out as having the majority of their non-learners low-qualified, in particular Portugal and Malta, because of the high share of low-qualified people in the population. Conversely, the largest share of high-qualified private sector employees not participating in learning were in Lithuania (representing 14.7% of all non-learners).

2.6.4 The determinants of participation in adult learning

Beyond a descriptive analysis of who is and who is not participating in adult learning, it is also possible, using regression analysis, to extract which socio-demographic characteristics most influence the likelihood of participation in learning the most. Such determinants can be divided into three categories: (i) personal characteristics (sex, age¹²⁵, migrant status, married/cohabiting status and degree of urbanisation of place of residence¹²⁶; (ii) educational attainment (low, medium, and high¹²⁷); and (iii) job-related characteristics (occupation, firm size, work situation¹²⁸, professional status¹²⁹ and sector). Note that here there is no distinction between formal and non-formal education and training.

These are put together and referred to as ‘adult learning activities’ (ALA)¹³⁰ in the remainder of the analysis. Figure 52 illustrates the relative contribution (in percentages) of the above three categories in accounting for participation in ALA in individual Member States¹³¹. Although no consistent pattern can be observed across different countries, job-related characteristics seem to be, overall, the most important determinant of participation in ALA. They play an especially crucial role in the Netherlands, Romania, Hungary, Finland, Lithuania and Slovakia. On the other hand, personal characteristics appear to matter more than job-related characteristics in Cyprus, Estonia, Sweden, Croatia and Germany. In the latter, educational attainment clearly turns out to be the most relevant predictor of participation in ALA. In Germany, as in the majority of EU countries (the exceptions being Bulgaria, Denmark, Estonia, Finland and Hungary), the probability of being involved in ALA increases with the level of education. This could signal an important challenge for those Member States in which upskilling and re-skilling of low-skilled/low-educated workers are especially necessary.

Figures 54 and 55 show the relative importance of the different factors included in the categories of personal and job-related characteristics. For the former, Figure 53 shows that there is a lot of heterogeneity across Member States. Age appears to be the most relevant personal characteristic affecting participation in ALA in Croatia, Denmark, Italy and Lithuania. Migrant status seems to be important in Sweden, Germany, Cyprus, Greece, Ireland, the Netherlands, France and Malta. In all these Member States, except for the Netherlands and Malta¹³², migrants are significantly less likely to participate in ALA than natives. Married/cohabiting status turns out to be a significant determinant in Bulgaria, Spain and Hungary. In these countries, apart from Hungary¹³³, married or cohabiting workers are found to have a lower probability of participating in ALA than single workers. Finally, the degree of urbanisation of one’s place of residence is found to be influential in Latvia, Austria, and Portugal.

Moving on to the different job-related characteristics, it is possible to observe large variations between Member States in terms of their relative importance. Firm size is found to be the most important predictor of participation in ALA among job-related characteristics in Cyprus, Ireland, Portugal and Bulgaria, with workers in larger firms are consistently more likely to participate in ALA than those in smaller firms. Sector is important in Lithuania, Greece, Austria, Italy, Belgium and Hungary. Professional status is very relevant in Slovakia, France, Finland, the Netherlands and Romania. Occupation appears to be a meaningful determinant in Czechia, Luxembourg, Estonia and Croatia. In all these countries, in line with expectations, workers with high-level occupations broadly display a higher participation rate in ALA than those with low-level occupations. Overall, work situation is the least important job-related characteristic influencing participation in ALA. However, one exception is Denmark, where it is actually the most important factor.

Analysing different determinants of participation in adult learning could allow for a better targeting of policies, in particular by identifying target groups in specific countries based on the most important characteristics. For example, in countries where job-related factors are the most important determinants of adult learning participation, policies could relate to the labour market status/situation of the individuals and aim at increasing the learning participation of the underrepresented groups. In other countries, where personal characteristics (such as age) seem to be the driving force, interventions could be designed that specifically target age criteria. This could allow for policies to be adapted for specific contexts, driven by historic, economic, cultural and other country-specific factors.

Previous Section Next Section

2.7 Learning mobility

2.7.1 Progress towards the EU target on learning mobility

In 2011, EU countries agreed on a target that ‘by 2020, an EU average of at least 20% of higher education graduates should have had a period of higher education-related study or training (including work placements) abroad, representing a minimum of 15 European Credit Transfer and Accumulation System (ECTS) credits or lasting a minimum of three months’¹³⁴. This target refers to worldwide outward mobility, i.e. mobility from EU countries to both EU and non-EU destinations. It covers two types of mobility: credit mobility and degree mobility¹³⁵. Regrettably, due to the lack or incompleteness of inward degree mobility data for some destination countries, the computation of this target remains underestimated¹³⁶.

In 2018, 13.5% of higher education graduates in the EU were mobile, meaning that they studied abroad, partly or entirely (Figure 55). Overall, 9.1% of the graduates had a temporary experience abroad, known as ‘credit mobility’, and 4.3% graduated in a country which was not the one where they received their upper secondary school diploma¹³⁷, known as ‘degree mobility’. Luxembourg, Cyprus, the Netherlands, Germany and Finland (in descending order) have the highest shares of outwardly mobile graduates.

Figure 55 highlights the different indices of the two mobility components across Member States. Among the best performers, the prevalence for degree mobility and credit mobility differs. In the Netherlands, Finland and Germany, the percentage of credit-mobile graduates (22.5%, 15.1% and 14.5%, respectively) is higher than the percentage of degree-mobile graduates (2.8%, 4.1% and 5.3%, respectively). In Luxembourg and Cyprus, in contrast, the percentage of degree-mobile graduates (74.1% and 35.2%, respectively) is higher than the percentage of credit-mobile graduates (12.7% and 2.2%, respectively).

About half of the credit-mobile graduates in the EU had their stay abroad funded by the Erasmus or other EU programmes (Figure 56). In six countries (Cyprus, Greece, Latvia, Romania, Malta and Bulgaria), this share is 95% or higher. Credit mobility through EU programmes accounts for more than half the credit mobility in 15 additional countries where data is available. Only four countries have a higher frequency of credit mobility through non-EU programmes (the Netherlands, France, Sweden and Denmark), with shares of credit mobility through non-EU programmes ranging from 64% to 77%.

2.7.2 Inward mobility

In 2018, the highest rates of inward degree-mobile graduates were recorded in Luxembourg (24.2%), the Netherlands (18.8%), Austria (16.0%) and Denmark (15.1%) (Figure 57). Values between 10% and 15% were present in Ireland (14.6%), Belgium (12.7%), Czechia (12.4%), Estonia (11.7%), France (10.6%) and Sweden (10.3%). The rate of inward degree-mobile graduates is below 10% in the remaining Member States, with rates of 5% or lower recorded in nine countries (Figure 57).

At the EU level, the rate of inward degree-mobile graduates increases with the level of educational attainment: short-cycle (ISCED 5) has a rate of 2.4%¹³⁸; bachelor’s level (ISCED 6) has a rate of 5.0%; master’s level (ISCED 7) has a rate of 12.9%; and PhD level (ISCED 8) has a rate of 20.8%. In 17 countries, the highest inward degree mobility rate is at ISCED level 8¹³⁹, while seven countries have the highest rate at ISCED level 7¹⁴⁰. Only one country has the highest inward degree mobility rate at ISCED level 6, Greece, while two countries have the highest rate at ISCED level 5, Cyprus and Malta. At ISCED level 8, Luxembourg is in an atypical situation where the inward degree mobility is higher than 100%. This is due to the peculiarity of the country, which attracts more inward degree-mobile PhD students than the number of PhD graduates with origin in the country.

Figure 58 presents inward degree-mobile graduates across the EU by area of origin, distinguishing between the EU and the main macro-areas outside the EU. Overall, 30.3% of inward degree-mobile graduates to the EU originate in the EU, followed by graduates originating in Asia (22.2%), Africa (15.9%) and non-EU European countries (13.5%). Intra-EU degree mobility accounts for less than 50% of the inward degree mobility in 18 countries¹⁴¹. In France (14.0%), Poland (14.1%) and Croatia (14.1%), less than 15% of the inward degree-mobile graduates originate in another Member State. This is in contrast to Austria (74.2%), Luxembourg (71.9%) and Denmark (66.9%), where two out of three inward degree-mobile graduates originate in the EU.

Historical ties are important for explaining mobility patterns between countries. This is clear in the case of Spain and Portugal, where, respectively, 51.7% and 38.5% of the inward degree-mobile graduates come from the Caribbean, Central America and South America (Figure 58). Geographical proximity and common language are two other important factors. In Belgium, 19.8% of the degree-mobile graduates come from France, and, similarly, 43.6% of the graduates in Austria come from Germany. In Czechia, 57.6% of the inward degree-mobile graduates originate in Slovakia, while 47.2% of the inward degree-mobile graduates in Slovakia originate in Czechia. High mobility from non-EU countries may also be driven by geographical proximity, as in Poland, where 53.0% of the inward degree-mobile graduates come from Ukraine, and in Croatia, where 70.6% of the inward degree-mobile graduates come from Bosnia and Herzegovina.

2.7.3 Recent policy responses

Although mobility in higher education is being increasingly recognised, obstacles hindering the free movement of students are still present in the form of economic, administrative and linguistic barriers. In a 2011 Council Recommendation promoting learning mobility, Member States were encouraged to implement structural reforms to create a positive learning environment for all students¹⁴². The Recommendation also provides a framework for Member States to monitor progress in this area and for the European Commission and the Eurydice network to gather national data and compile it into a ‘mobility scoreboard’.¹⁴³.

Limited access to funding is one of the main obstacles to mobility, and consequently the portability of domestic support (grants and/or loans) can be a major factor in a student’s decision to study abroad. Portability can be differentiated between portability for credit mobility and portability for degree mobility. Degree portability is the extent to which students can take domestic support to pursue a full degree abroad, whereas credit portability is the extent to which students can take domestic support to pursue study visits abroad that lead to credits in the framework of a home country programme¹⁴⁴.

In 2018/2019, portability of public support measures for the first and second cycle was more widespread for credit mobility than degree mobility. Within the EU, 14 education systems (the Flemish and German communities of Belgium, Denmark, Germany, Ireland, France, Cyprus, Luxembourg, Malta, the Netherlands, Austria, Slovenia, Finland and Sweden) guarantee portability for both credit and degree mobility within the European Higher Education Area (EHEA)¹⁴⁵. An additional 12 education systems (Czechia, Estonia, Spain, Croatia, Italy, Latvia, Lithuania, Hungary, Poland, Portugal, Romania and Slovakia) provide portability of public support for credit mobility only¹⁴⁶. Most Member States apply some restrictions, whether geographical (e.g. Germany limits degree portability to the EU and Switzerland) or scheme-based (e.g. Greece, Spain, Latvia, Lithuania and Portugal only allow portability of grants to programme exchanges with recognised schemes such as Erasmus+ for example. In Lithuania, this is also the case for loans).

Availability of public support for learning mobility has a particular impact on students from disadvantaged socio-economic backgrounds and students with disabilities, who have been identified as less likely to participate in such activities¹⁴⁷. All Member States but one (Bulgaria) provide targeted mobility grants, needs-based portable grants or universal portable grants to disadvantaged learners¹⁴⁸. Since the release of the previous mobility scoreboard background report in 2016¹⁴⁹, several countries have introduced new measures. Romania introduced new legislation in 2017 making the portability of need-based grants possible for credit mobility. Additional need-based grants have been introduced on top of Erasmus+ grants in Latvia and Slovenia.

To provide the right support for disadvantaged students, information on the extent to which different groups participate in learning mobility is key. Although all countries participating in the Erasmus+ programme are required to monitor participation for this specific programme, there are some countries that go beyond this obligation. Six education systems already had comprehensive monitoring practices across all major mobility programmes in 2015/2016 (French and Flemish communities of Belgium, France, Germany, Austria and Italy).

Additional policy measures supporting learning mobility include foreign language education and automatic recognition of qualifications and the outcomes of learning periods abroad. Foreign language preparation is considered in Chapter 3.1 of this report. Automatic mutual recognition is the right for: (i) the holder of a qualification of a certain level issued by one country to be considered for entry to a higher education programme at the next level in another country, without having to go through a separate recognition procedure; or (ii) a person who has completed a mobility period abroad to have their learning outcomes recognised without having to go through a separate recognition procedure. The automatic recognition of qualifications and the outcomes of learning periods abroad is particularly important, as it is a necessary precondition for large-scale mobility. At present, eight Member States (Denmark, Germany, France, Italy, Malta, Poland, Finland and Sweden) report that they operate on the basis of automatic recognition of degrees issued in all other EHEA countries. A further 15 education systems in Member States report that they have automatic recognition for some of these countries¹⁵⁰.

Previous Section Next Section

2.8 How has BREXIT changed the EU performance on the ET2020 indicators?

United Kingdom’s departure from the EU has changed the aggregate EU performance on the Europe 2020 and ET2020 benchmarks. Below is a summary table of these changes. The column ‘EU-28’ shows a score with the United Kingdom. The numbers in bold show a better result. For example, in tertiary educational attainment, the EU-27 has only 40.3% of tertiary diploma holders which means a 1.3 percentage point decrease compared to the EU-28 with the United Kingdom.

Notes

⁴⁹ This is both a headline target of the Europe 2020 strategy – the EU agenda for growth and jobs from 2010 to 2020 – and a target of the ET2020 strategic cooperation framework.

⁵⁰ It would probably be worthwhile exploring slightly older age groups (25-34 or 30-34) as some young people leave formal education temporarily and finish secondary education in schools for adults (often up to the age of 30).

⁵¹ Council of the European Union (2011). Council Recommendation of 28 June 2011 on policies to reduce early school leaving.

⁵² European Commission (2019). Assessment of the Implementation of the 2011 Council Recommendation on Policies to Reduce Early School Leaving.

⁵³ Proposal for a COUNCIL RECOMMENDATION on A Bridge to Jobs - Reinforcing the Youth Guarantee and replacing Council Recommendation of 22 April 2013 on establishing a Youth Guarantee, COM/2020/277 final.

⁵⁴ See Flisi, S., Goglio, V. and Meroni, E. (2014). Monitoring the Evolution of Education and Training Systems: A Guide to the Joint Assessment Framework for a literature review.

⁵⁵ With at most lower secondary education (i.e. ISCED 0-2 levels).

⁵⁶ Black, S.E., Devereux, P. J. and Salvanes, K. G. (2005). Why the Apple Doesn’t Fall Far: Understanding Intergenerational Transmission of Human Capital, American Economic Review, Vol. 95, No 1, 2005, pp. 437-49.

⁵⁷ The ELET rate includes people who may have left education in previous years and their decision would have been influenced by the labour market conditions at that time, not by the current ones. To account for this, we use the unemployment rate of the previous year (technically speaking, the unemployment rate with a lag of one period) as a variable.

⁵⁸ Expressed in purchasing power standards.

⁵⁹ Behaghel, L., Gurgand, M., Kuzmova, V. and Marshalian, M. (2018). European Social Inclusion Initiative review paper, Abdul Latif Jameel Poverty Action Lab J-PAL.

⁶⁰ Goux, D., Gurgand, M. and Maurin, E. (2017). Adjusting Your Dreams? High School Plans and Dropout Behaviour. The Economic Journal, Vol. 127, No. 602, pp. 1025-1046.

⁶¹ Weinberg, D et al. (2019). The pathways from parental and neighbourhood socioeconomic status to adolescent educational attainment: An examination of the role of cognitive ability, teacher assessment, and educational expectations, PLOS ONE.

⁶² The degree of urbanisation classifies local administrative units (at LAU2 level) as cities, towns and suburbs, or rural areas, based on a combination of geographical contiguity and minimum population thresholds applied to 1 km² population grid cells. More details on the methodology can be found at Eurostat, Statistics Explained.

⁶³ This definition has its limitations as in many EU Member States compulsory education starts at 4. See Figure 67.

⁶⁴ European Council (2002). Presidency conclusions. Barcelona European Council, 15-16 March 2002.

⁶⁵ Council conclusions of 12 May 2009 on a strategic framework for European cooperation in education and training (ET2020).

⁶⁶ There are two categories of ISCED level 0 programmes: early childhood educational development (ISCED 01) and pre-primary education (ISCED 02). The former has educational content designed for younger children (in the age range of 0 to 2 years), while the latter is designed for children from the age of 3 to the start of primary education. For the specifics of the definitions, see ‘International Standard Classification of Education ISCED 2011’; for further details on the characteristics of the programme content of ISCED 0, see section 9.

⁶⁷ It is, however, worth noting that there is not always a perfect overlap between ISCED 0 and the definition of ECEC used in the ECEC target. ISCED 0 covers children up to the start of primary education, while the target takes into account children up to the start of compulsory primary education. These concepts overlap in all EU target except Ireland, where primary education starts before compulsory education, and therefore the calculation of the target includes levels ISCED 0 and 1.

⁶⁸ In France and Hungary for example, education is mandatory from the age of 3 since 2019. Another example is Belgium, where education has become mandatory from the age of 5 since September 2020. For more examples, cf. Eurydice, Key indicators 2019, p. 66. Cf. also European Commission/EACEA/Eurydice (2020). Structural indicators, 2020, p. 9 on the legal framework.

⁶⁹ European Commission/EACEA/Eurydice (2019). Key indicators 2019, p. 66.
Cf. also European Commission/EACEA/Eurydice, The structure of the European education systems 2019/20.
European Commission/EACEA/Eurydice (2019). Compulsory Education in Europe – 2019/20. Eurydice Facts and Figures.

⁷⁰ European Commission. Education and training monitor. 2018 edition.

⁷¹ There is under-coverage for Greece in the 2018 data, because some 3 and 4 year-olds are not included. Better coverage was reported in 2017, and Greece hopes to provide full coverage in the next data collection (and revise the 2018 data).

⁷² Mostly due to a rise in attendance in private facilities, as is clear from educ_uoe_enrp01.

⁷³ ESTAT calculation on from educ_uoe_enra02 2018 data. Data is missing for Greece, because data on 0-2 year-olds enrolled in ISCED 01 are incomplete. Participation rates are underestimated for Belgium (under-coverage, only the Flemish Community has reported data on ISCED 01) and Malta (under-coverage, enrolments in ISCED 01 are not included).

⁷⁴ Kaga, Y., Bennett, J., and Moss, P. (2010). Caring and Learning Together: A Cross-National Study of Integration of Early Childhood Care and Education within Education. Paris: UNESCO. Proposal for key principles of a Quality Framework for Early Childhood Education and Care, Report of the Working Group on Early Childhood Education and Care under the auspices of the European Commission, DG Education and Culture, 2014.

⁷⁵ Key Data on Early Childhood Education and Care in Europe, 2019 Edition (European Commission/EACEA/Eurydice, 2019).

⁷⁶ Literature review on the effects of early childhood education and care, in Vandenbroeck, M., Lenaerts, K., Beblavý, M. (2018). Benefits of early childhood education and care and the conditions for obtaining them. An EENEE Analytical Report No. 32, January 2018. Utrecht University and CARE consortium (2017). CARE: Curriculum Quality Analysis and Impact Review of European ECEC. Kottelenberg, M. J., and Lehrer, S. F. (2017). Targeted or universal coverage? Assessing heterogeneity in the effects of universal child care. Journal of Labour Economics, 35(3), 609-653.

⁷⁷ OECD, Literature review on early childhood education and care for children under the age of 3 (forthcoming).

⁷⁸ Schleicher, A. Helping our Youngest to Learn and Grow. Policies for Early Learning. OECD (2019). Council Recommendation on High-Quality Early Childhood Education and Care Systems of 22 May 2019.

⁷⁹ Schleicher, A. Helping our Youngest to Learn and Grow. Policies for Early Learning. OECD (2019). Vandenbroeck, M., Lenaerts, K., Beblavý, M. (2018). Benefits of early childhood education and care and the conditions for obtaining them. An EENEE Analytical Report No. 32, January 2018. Utrecht University and CARE consortium (2017). CARE: Curriculum Quality Analysis and Impact Review of European ECEC.

⁸⁰ Flisi, S. and Blasko, Zs. (2019). A note on early childhood education and care participation by socio-economic background, JRC Science for Policy Report. See also European Commission (2019). Education and training monitor. 2019 edition.

⁸¹ OECD, Literature review on early childhood education and care for children under the age of 3 (forthcoming).

⁸² OECD Family Data Base, chart PF3.2.B. Participation rates in early childhood education and care by income, 0 to 2 year-olds (OECD estimates, based on EU-SILC data).

⁸³ Council Recommendation of 22 May 2019 on High-Quality Early Childhood Education and Care Systems, 9014/19.

⁸⁴ Key data on early childhood education and care in Europe – 2019 edition. European Commission/EACEA/Eurydice. Structural indicators for monitoring education and training systems in Europe – 2019 and 2020.

⁸⁵ Participation is compulsory from the age of 3 in France and Hungary.

⁸⁶ See more details in Key data on early childhood education and care in Europe – 2019 edition. European Commission/EACEA/Eurydice, pp. 54-62.

⁸⁷ Underachievers in PISA are those pupils who fail to reach proficiency Level 2, i.e. the minimum level necessary to participate successfully in society.

⁸⁸ All EU averages in reading exclude Spain, because Spanish data are not available at the time of writing.

⁸⁹ OECD defines reading literacy, the core subject for PISA 2018, as ‘understanding, using, evaluating, reflecting on and engaging with texts in order to achieve one’s goals, to develop one’s knowledge and potential, and to participate in society’. OECD, PISA 2018 Results (Volume I) – What Students Know and Can Do, 2019, p. 15.

⁹⁰ OECD, PISA 2018 Results (Volume I) – What students know and can do, 2019, p. 57.

⁹¹ The results of the PISA assessments are estimates, because they are based on samples of pupils, rather than on the total pupil population, and on a limited set of assessment tasks rather than all possible ones. An observed difference between two estimates based on samples is called ‘statistically significant’ if it is likely that a real difference exists in the populations from which the samples are drawn.

⁹² OECD, PISA 2018 Results (Volume II) – Where all students can succeed, 2019, p. 142.

⁹³ European Commission, (2019). PISA 2018 and the EU: Striving for social fairness through education.

⁹⁴ Pansu, P. Regner, I. Max, S. Cole, P., Nezlek, J. B. and Huguet, P. (2016). A burden for the boys: Evidence of stereotype threat in boys’ reading performance. Journal of Experimental Social Psychology, 65, pp. 26-30. Marcenaro-Gutierrez, O. Lopez-Agudo, L. Ropero-Garcia, M. (2018). Gender Differences in Adolescents’ Academic Achievement. Young, 26 (3), pp. 250-270. Smith, M. and Wilhelm, J. (2012). ‘Reading don’t fix no Chevys’: Literacy in the lives of young men, Portsmouth: Boynton/Cook. Freedmon, B. (2003). Boys and literacy: Why Boys? Which boys? Why now? Paper presented at the Annual Meeting of the American Educational Research Association.

⁹⁵ OECD (2010). Learning to Learn: Student Engagement, Strategies and Practices.

⁹⁶ Freedmon, B. (2003). Boys and literacy: Why Boys? Which boys? Why now? Paper presented at the Annual Meeting of the American Educational Research Association.

⁹⁷ European Commission (2012). EU high level group of experts on literacy. Final report.

⁹⁸ Genlott, A. A. Gronlund, A. (2016). Closing the gaps – Improving literacy and mathematics by ICT-enhanced collaboration. Computers & Education, 99, pp. 68-80.

⁹⁹ European Commission, (2019). PISA 2018 and the EU: Striving for social fairness through education.

¹⁰⁰ The OECD measures the ESCS index taking into consideration multiple variables related to pupils’ family background, namely: parents’ education, parents’ occupation, home possessions, number of books and educational resources available at home.

¹⁰¹ Emma Garcia, Elaine Weiss, ‘Education inequalities at the school starting gate, Gaps, trends, and strategies to address them’, Economic Policy Institute, September 27, 2017. Duckworth, K. et al. (2009). Influences and leverage on low levels of attainment: a review of literature and policy initiatives. Centre for Research on the Wider Benefits of Learning Research Report 31, London, DCSF.

¹⁰² Results by migrant background are available only in the main subject area tested in each PISA round.

¹⁰³ The proportion of pupils with a migrant background varies widely between EU Member States. To avoid calculations based on very small sample sizes, this report shows results only for EU Member States where the percentage of pupils with a migrant background is at least 5%.

¹⁰⁴ The definition of pupils ‘born abroad’ and pupils ‘native-born with parents born abroad’ employed in this report corresponds to what the OECD defines respectively as ‘first-generation immigrant students’ and ‘second-generation immigrant students’.

¹⁰⁵ OECD (2018). The resilience of students with an immigrant background, OECD Reviews on Migration Education.

¹⁰⁶ According to PISA 2018, cities have over 100 000 inhabitants, while rural areas have fewer than 3 000 inhabitants.

¹⁰⁷ Echazarra, A. and Radinger, T. (2019). Learning in rural schools: insights from PISA, TALIS and the literature, OECD Education Working Paper No. 196.

¹⁰⁸ The data differentiating the employment rate of medium-level graduates by the orientation of their qualification is available only as of 2014, after the implementation of the latest ISCED-2011 classification of education in the EU LFS.

¹⁰⁹ The wages are the ‘full-time equivalent gross monthly wages’.

¹¹⁰ Three different level of educational levels are considered: low education (pre-primary education, primary education and lower secondary education), medium education (upper secondary education) and high education (tertiary education).

¹¹¹ Only individuals aged 16-34 years old are considered. The information on vocational v general secondary education is not available for DE.

¹¹² The values presented in Figure 44 do not take into account the type of secondary education. General and vocational education students may systematically differ in terms of both observable and unobservable characteristics affecting the choice of the type of secondary education and earnings.

¹¹³ VET in Austria has apprenticeship-programmes in the dual system as well as school-based VET. There is no marked prevalence of the dual system.

¹¹⁴ The importance of seniors' learning cannot be overstated It is important that younger seniors have sufficient skills to remain active economically. On the other hand, it is beneficial for older seniors not to be grouped in retirement homes, but to participate in less institutionalised forms of care, which requires new skills of organisers and users of such forms.

¹¹⁵ Eurostat online data code [trng_aes_151].

¹¹⁶ Vera-Toscano, E., Urzì Brancati, C. (2020). Towards an improved adult learning monitoring framework.

¹¹⁷ European Commission / DG EAC (2019). Achievements under the Renewed European Agenda for Adult Learning, a report of the ET 2020 Working Group on Adult Learning 2018-2020.

¹¹⁸ Education and Training Monitor 2018.

¹¹⁹ See the report of the ET 2020 Working Group on Adult Learning 2018-2020: Achievements under the Renewed European Agenda for Adult Learning.

¹²⁰ According to the Working Group on Adult Learning 2018-2020, the differences in LFS, AES and CVTS results are not only related to the period of participation in education and training (4 weeks vs 12 months). The narrow definition of non-formal education in the LFS survey is also important. In some countries, this restriction may not be limited to guided on the job training.

¹²¹ Regulation (EU) 2019/1700

¹²² C/2019/8809 final

¹²³ See Eurostat, Statistics Explained.

¹²⁴ Eurostat, Statistics Explained.

¹²⁵ The relationship between our dependent variable and age is allowed to vary over the life cycle. In fact, the estimated age profile turns out to be inverse U-shaped, signalling that the probability of being involved in formal or non-formal learning activities first rises and then declines with age.

¹²⁶ Three options are considered: cities, towns or suburbs, and rural areas.

¹²⁷ Low education corresponds to ISCED11 levels 0-2, medium education to ISCED11 levels 3-4, and high education to ISCED11 levels 5-8.

¹²⁸ A distinction is made between full-time and part-time work.

¹²⁹ There is a distinction between employees with a permanent contract (or a contract of undefined duration), employees with a temporary contract (or contract of limited duration), the self-employed and family workers.

¹³⁰ The dependent variable ‘ALA’ takes a value of 1 if the individual has been involved either in formal or non-formal learning or in both; it has a value of 0 otherwise.

¹³¹ The data shown in this figure is based on the results of a logistic regression (run for each Member State separately) where the dependent variable captures participation in ALA v non-participation and the independent variables are grouped into the three categories mentioned above: personal characteristics, educational attainment and job-related characteristics. The relative importance of each of these three categories in accounting for participation in ALA is identified by comparing the reduction in deviance attributable to all the independent variables belonging to each category. The relative contribution of each group of determinants could not be presented in terms of the ‘proportion of variance explained’, because this concept is not well-defined in the context of logistic regression. Logistic regression is suitable for modelling binary dependent variables, such as the participation in ALA (i.e. a variable with only two answer categories such as 0 or 1 / Yes and No), but it does not allow for ‘decomposing’ the variability of the dependent variable into an explained and unexplained part.

¹³² In NL and MT, the difference in participation in ALA between migrants and natives is not statistically significant.

¹³³ In HU, there is no statistically significant difference between married/cohabiting workers and single workers in terms of ALA participation.

¹³⁴ Council conclusions on a benchmark for learning mobility. OJ C 372, 20.12.2011, p. 31–35.

¹³⁵ Credit-mobile graduates are those who have had a temporary study period and/or work placement abroad and return to their ‘home institution’ to complete their degree. Degree-mobile graduates are those whose country of origin (i.e. the country where their upper secondary diploma was obtained) is different from the country in which they graduate. While data on credit mobility is collected in the countries to which students returned after their credit mobility stay, data on degree-mobile graduates is collected at the level of the destination country. Consequently, the calculation of outwardly mobile EU graduates relies on figures provided by all EU and non-EU destination countries. For an overview of the learning mobility target, see Flisi, S. and Sanchez-Barrioluengo, M. (2018). Learning Mobility II: An estimation of the benchmark. A JRC Science for Policy Report JRC113390.

¹³⁶ For the academic year 2017/2018, information on inward degree mobility is available for 44 destination countries. See note to Figure 55 for an overview of countries with available data. The main outstanding missing destination by far is the US. See Flisi, S. and Sanchez-Barrioluengo, M. (2018). Learning Mobility II: An estimation of the benchmark. A JRC Science for Policy Report JRC113390. This gives an estimation of the effect the missing US data has on the computation of the target.

¹³⁷ While country of origin would ideally be defined as country of prior education, i.e. country where upper secondary diploma was obtained, there are a number of countries using a different definition in the data collection. For the academic year 2017/2018, definitions of country of origin include country of prior education (ES (ISCED 5), LV and PL (ISCED 8)), country of upper secondary diploma (BE, BG, CZ, DK, DE, EL, HR, CY, LT, LU, MT, NL (ISCED 5-7), AT, PL (ISCED 6-7), PT, RO and FI), country of citizenship (IT, HU and SK), the country of usual residence (EE, IE, ES (ISCED 6-8) and SI) and country estimations (NL (ISCED 8)). In FR country of upper secondary diploma or the country of citizenship is used. In SE, international students are defined as students who have a student residence permit or are either non-residents or have moved to Sweden not more than six months before starting their studies. For students at ISCED 8 the time limit is 24 months. Students with student residence permit are reported by country of citizenship while other students are reported by country of birth. Homecoming nationals are reported as national students.

¹³⁸ This level is not applicable for BG, EE, EL, LT, PL, RO and FI.

¹³⁹ BE, BG, CZ, DK, DE, IE, FR, HR, IT, LU, NL, AT, PT, SI, SK, FI and SE. The inward degree mobility rate for ISCED level 8 is unavailable for ES.

¹⁴⁰ EE, ES, LV, LT, HU, PL and RO.

¹⁴¹ BE, BG, DE, EE, IE, ES, FR, HR, IT, LV, LT, HU, MT, PL, PT, RO, FI and SE.

¹⁴² Council Recommendation of 28 June 2011 on 'Youth on the move' – promoting the learning mobility of young people, OJ C199, 7.7.2011, C199/4.

¹⁴³ The data on policy measures in this section is mainly taken from European Commission/EACEA/Eurydice (2020). Mobility Scoreboard: Higher Education Background Report 2018/19.

¹⁴⁴ European Commission/EACEA/Eurydice (2020). Mobility Scoreboard: Higher Education Background Report 2018/19.

¹⁴⁵ The European Higher Education Area (EHEA) is an international collaboration on higher education, encompassing 48 countries and the European Commission.

¹⁴⁶ Estonia, Latvia, Hungary and Slovakia provide publicly subsidised loans that are portable for both credit and degree mobility. In addition, Estonia allows portability for both credit mobility and degree mobility for two grant schemes: needs-based study allowance and scholarships for students with special needs.

¹⁴⁷ Hauschildt, K., Vögtle, E.M. and Gwosć, C., (2018). Social and Economic Conditions of Student Life in Europe. Eurostudent VI 2016-2018: Synopsis of Indicators. Edited by DZHW (German Centre for Higher Education Research and Science Studies. Bielefeld: W. Bertelsmann Verlag; European Commission, 2019. Studying abroad - benefits and unequal uptake. Science for Policy Briefs, Joint Research Centre.

¹⁴⁸ The definition of what constitutes a disadvantaged learner has been expanded to include both students from low socio-economic backgrounds and students with disabilities for the 2020 edition of the mobility scoreboard. This is the same as the approach taken in the Bologna Process Implementation Report.

¹⁴⁹ European Commission/EACEA/Eurydice (2016). Mobility Scoreboard: Higher Education Background Report.

¹⁵⁰ BE (fl), BE (fr), BE (de), CZ, EE, LV, LT, LU, HU, NL, PT, RO and SK.