EUROPEAN COMMISSION

Brussels, 10.12.2021

SWD(2021) 369 final

COMMISSION STAFF WORKING DOCUMENT

IMPACT ASSESSMENT REPORT

Accompanying the document

Proposal for a Council Recommendation

on individual learning accounts

{COM(2021) 773 final} - {SEC(2021) 417 final} - {SWD(2021) 368 final} - {SWD(2021) 370 final}

List of acronyms

1. INTRODUCTION: POLITICAL AND LEGAL CONTEXT

2. PROBLEM DEFINITION

2.1 What are the problems?

2.2 What are the problem drivers?

2.3 What are the consequences?

2.4 How will the problem evolve?

3. WHY SHOULD THE EU ACT?

3.1 Legal basis

3.2 Subsidiarity: Necessity and added value of EU action

4. OBJECTIVES: WHAT IS TO BE ACHIEVED?

4.1 General objectives

4.2 Specific objectives

5. WHAT ARE THE AVAILABLE POLICY MEASURES?

5.1 What is the baseline from which options are assessed?

5.2 Measures discarded at an early stage

5.3 Description of the policy measures

5.4 Description of the alternative policy packages

6. WHAT ARE THE IMPACTS OF THE POLICY PACKAGES?

6.1 Analysing the main impacts

6.2 The choice of legal instrument

7. HOW DO THE POLICY PACKAGES COMPARE?

8. PREFERRED POLICY PACKAGE

9. HOW WILL ACTUAL IMPACTS BE MONITORED AND EVALUATED?

ANNEX 1: PROCEDURAL INFORMATION

1. Lead DG, Decide Planning/CWP references

2. Organisation and timing

3. Consultation of the RSB

4. Evidence, sources and quality

ANNEX 2: STAKEHOLDER CONSULTATIONS

1. Introduction

2. Consultation strategy

3. Inception Impact Assessment

4. Targeted consultations

5. Public Consultation

ANNEX 3: WHO IS AFFECTED AND HOW?

1. Practical implications of the initiative

2. Summary of costs and benefits

3. Summary of the SME test

ANNEX 4: ANALYTICAL METHODS

1. Quantification of impacts

2. Qualitative methods

3. Constraints

ANNEX 5: OVERVIEW OF THE INTERVENTION LOGIC

ANNEX 6: ADULT LEARNING: MEASUREMENT, KEY CHARACTERISTICS AND PARTICIPATION LEVELS

1. Measurement

2. Key characteristics

3. Participation levels and differences between groups

4. Recommendations to increase participation in adult learning in the European Semester

ANNEX 7: ADDITIONAL EVIDENCE ON THE PROBLEM DRIVERS

1. Driver 1: Insufficient financial support

2. Driver 2: Limited incentives and motivation of individuals

ANNEX 8: BASELINE SCENARIO

1. The existing EU instruments

2. Baseline scenario by Member State

3. Estimating adult learning participation in 2030 under the baseline scenario

ANNEX 9: EVIDENCE ON THE DISCARDED POLICY MEASURES

1. Tax incentives for individuals

2. Subsidised loans or savings schemes for adult learning

ANNEX 10: EVIDENCE ON THE BENEFITS OF ADULT LEARNING

1. Returns to individuals

2. Returns to enterprises

3. Returns to society

4. Selection of coefficients for impact estimates

ANNEX 11: EVIDENCE ON INDIVIDUAL TRAINING ENTITLEMENTS

1. Effectiveness of training entitlements

2. Evidence on the costs of providing training entitlements

ANNEX 12: QUANTIFYING IMPACTS

A. IMPACT ON ADULT LEARNING PARTICIPATION

B. COST-BENEFIT ANALYSIS OF SHORT TO MEDIUM TERM IMPACTS

C. LONG TERM AND INDIRECT IMPACTS

ANNEX 13: CASE STUDIES

1. Austria: Education Savings

2. Estonia: Unemployment Insurance Fund

3. Greece: Targeted training voucher schemes

4. Italy: Regional individual training entitlements

5. Netherlands: The STAP - from income tax deduction towards individual training entitlements

6. Singapore: SkillsFuture Credit - using personal accounts for a universal yet targeted provision of training entitlements

7. UK: Individual Learning Accounts

ANNEX 14: INDIVIDUAL LEARNING ACCOUNTS – THE FRENCH EXPERIENCE

1. Summary

2. General Description, Key Features, and Achievements

3. Performance of the CPF

ANNEX 15: OPERATIONAL OBJECTIVES AND MONITORING

List of acronyms

1. INTRODUCTION: POLITICAL AND LEGAL CONTEXT

On 25 June 2021, the European Council welcomed the EU headline targets of the European Pillar of Social Rights Action Plan and the Porto Declaration. 1 Leaders thereby supported the ambition that at least 60% of all adults should participate in training every year by 2030. 2  This contributes to making a reality the right to quality and inclusive education, training and life-long learning, as set out in the Charter of Fundamental Rights of the European Union (Article 14: Right to education) and principle 1 of the European Pillar of Social Rights on education, training and lifelong learning. 3 It also acknowledges the importance of improving the opportunities for developing skills throughout working life to achieve the EU’s ambitions for the coming decade: a swift economic recovery from the COVID-19 pandemic and a successful digital and green transition.

Notably, the EU will only achieve its digital transition if individuals have the skills to take advantage of emerging technologies and jobs. Equally, the green transition “can only succeed if the EU has the skilled workforce it needs to stay competitive”, as set out in the “Fit for 55 - delivering the EU's 2030 Climate Target on the way to climate neutrality” package. 4  As labour markets are reshaped by these transitions, the existing support systems for developing skills need to evolve as well to allow all individuals to thrive in the labour markets of the future.

This requires additional investments as well as the engagement of individuals. The European Council invited Member States to “explore possible models for public and private financing of lifelong learning and the development of skills on an individual level”, and asked the Commission to support Member States in these efforts. 5 The revised Employment Guidelines 6 call on Member States to strengthen the provisions on individual training entitlements and ensure their transferability during professional transitions including, where appropriate, through individual learning accounts.

Following up on the Mission Letter to Commissioner Schmit and the request to “explore the idea of individual learning accounts” 7 , the Commission’s Communication on the European Skills Agenda 8  announced its intention to explore individual learning accounts as a tool to support the up- and reskilling of working age adults, complementing other actions targeted at employers and education and training providers. 9  

The European Parliament shares the views of the European Council, the Council of the EU and the Commision and called to ensure access to skills development and life-long training opportunities in its Resolution on a European Pillar of Social Rights 10 and its Resolution on a strong social Europe for Just Transitions. 11 The Resolution on the European Skills Agenda 12  points out the potential of individual learning accounts as a step towards universal lifelong learning entitlements and a funding mechanism key to the implementation of the Skills Agenda.

The present initiative aims to put forward recommendations to Member States to close gaps in existing support systems for training and to increase individuals’ motivations and incentives to take up this support and participate in training. Recognising the complex reasons why individuals do not take up training (Section 2.2), the initiative aims to outline measures to integrate financial/demand side with non-financial/supply side support. It thereby aims to support the implementation of existing EU Recommendations on education and training, notably the Council Recommendations on VET, Upskilling Pathways and the validation of non-formal and informal learning. 13  The initiative does not seek to replace any currently well-working training policies, but to complement them and help close the remaining gaps. 

This proposal is related to two other Commission initiatives in preparation. Firstly, the initiative on Micro-credentials 14 complements the present initiative by promoting transparency and recognition of short training programmes and therefore incentivising uptake of training. Jointly, both initiatives aim to empower individuals to seek training by addressing challenges on the supply side and demand side. Secondly, the present initiative is also complementary to the initiative on improving the working conditions in platform work. The lack of access to training opportunities is an important challenge for platform workers 15 , and the present initiative considers concrete policy measures to support the training of individuals that do not depend on the current employer or employment status. 

2. PROBLEM DEFINITION

2.1 What are the problems?

Not enough adults participate in learning. In 2016 16 , only 37.4% of adults in the EU participated in learning. This is 22.6 percentage points below the 2030 target of 60% and participation levels fell also short of the 2020 EU-level target 17 in 21 of 27 EU Member States. Progress in increasing participation in adult learning in the EU over the last decade has been limited and uneven across EU Member States. The COVID-19 crisis has had a negative impact on training participation but already in 2019 20 Member States were below the target. The public and targeted consultations on this initiative showed a broad consensus by stakeholders that participation in adult learning is insufficient, even more so in light of the expected acceleration of the green and digital transitions. 18  

In addition to being low, participation in adult learning is unequal and strongly depends on the labour market status, employment relationship, company size, educational attainment and the exposure of sectors or occupations to automation. To varying degrees, this problem is present in all Member States and all of them received country-specific recommendations on skills in the context of the European Semester in either 2019 or 2020. 19

Across the EU, permanent employees have higher adult learning participation rates (45%) compared to other adults (29%). This pattern can be observed across all Member States, although with different sizes of the gap ( Figure 1 ). 83% of the respondents in the public consultation agreed that a lack of support for workers with no or loose links to an employer is an obstacle to a higher training participation. 20  

Lower shares of employees in SMEs (42%) and micro-enterprises (36%) participate in adult learning compared to those in large companies (55%). This pattern is visible across all EU Member States. In the public consultation, 83% 21 of respondents see the lack of capacity by small, medium-sized and micro-enterprises to organise training for their employees as an obstacle.

Adults with lower educational attainment participate less in learning, and participation rates are lower in sectors, occupations and types of employment with higher shares of lower qualified people. However, also within groups of similar educational attainment, the employment status remains a significant determinant of participation.

There are no pronounced gender differences in the participation of adults in learning, but men are somewhat more likely than women to report learning for job-related reasons (see Annex 6.3).

Adults in occupations and sectors with a higher risk of automation participate less in learning. This means that workers in the sectors and occupations where employment shares are predicted to decrease the most until 2030 participate the least in learning. Individuals facing a higher risk of skills obsolescence and unemployment are hence less equipped for these vulnerabilities in the near future.

22 Figure 1: Participation in adult learning for permanent employees vs. other adults in 2016

Source: Adult Education Survey 2016, ages 25-64. Participation figures presented in this IA exclude guided-on-the-job training to match the definition of adult learning used for EU benchmarking. See Annex 6 for details on the definitions used and disaggregated figures.

2.2 What are the problem drivers?

The reasons behind low participation in adult learning in the Member States and Member States’ policy responses to the issue are heterogeneous. 23 Yet, there are two broad problem drivers which can be observed across the EU: insufficient financial support for training, and insufficient incentives and motivation of individuals to take up training that go beyond a lack of financial support. This Section outlines how they limit progress in increasing participation rates of adults in learning and contribute to inequalities in participation between different groups of adults. Additionally, external drivers, outside of the scope of this initiative, amplify the incidence or the adverse consequences of low and unequal participation of adults in learning (Figure 2). 

Figure 2: Problem tree

2.2.1 Internal Drivers 24

1. Insufficient financial support

The costs of skills development are often concentrated on a specific actor, while the benefits (e.g. in the form of job progression, higher wages or productivity or a successful transition to new employment) only become apparent later and are typically shared among the individual, employers and society at large. 25  This results in insufficient investment overall, a limited coverage of existing support schemes in terms of groups of adults and types of training, barriers to devoting time to training and the fragmentation of the existing support schemes.

Insufficient investment levels in adult learning: According to the best available evidence, the overall level of investments (public and private) in adult learning in the EU is around 1.7% of GDP 26 with the largest share contributed by employers (1% of GDP), followed by public authorities (0.5% of GDP) and the remaining share contributed by individuals themselves (0.2% of GDP). The shares of overall investments in Member States range from 0.5% of GDP in RO to almost 2.5% of GDP in SE and FI. The public share of investments (0.5% of GDP) is relatively low when compared to the 4.7% of GDP that public authorities in the Member States invest on average in education. 27 While there has been rapid progress in increasing educational attainments over the recent decade (notably by reducing early school leaving an increasing the share of adults with tertiary education), there has been no corresponding increase in the public commitment to support the continuing skills development of adults after they have left initial education.  28  Also 23 of 27 national adult learning experts who have been consulted for the purpose of this impact assessment 29 consider the current levels of financing in their Member State as insufficient.

There is a positive correlation between the participation in adult learning and the estimated available funding in relation to GDP. The six Member States that scored above the EU participation targets for 2020 (SE, FI, DK, NL, EE and LU) spend on average 2.0% of GDP on adult learning, compared to an average of 1.3% of GDP for the other 21 Member States. Member States with higher levels of investments in adult learning also show smaller differences in participation between permanent employees and other adults (Annex 7.1), indicating that sufficient investments also ensure a more equal access to learning for all adults. This suggests that raising the level of participation in adult learning will also require increasing the level of investment in adult learning.

Current investments in adult learning are low also when compared against available estimates of their social and economic benefits. Most evidence in the academic literature suggests that the benefits of additional investments (such as wage and productivity gains) would outweigh their costs at the current participation levels. 30   However, even where countries have earmarked funds for adult learning, they are not always underpinned by sustainable financing sources, limiting their impact and leading to insufficient funding during recessions, when it is most needed. 31

Limited coverage of existing support schemes in terms of the groups of adults and types of training that are supported: Survey data and the public consultation point to the costs of training as a key barrier for individuals 32 leading to low and unequal adult learning participation (Section 2.1). As currently employers sponsor 88% of job-related non-formal training 33 , adults who would like to learn more and are not in permanent employment report costs as a barrier more often than those in permanent employment (31% vs. 22%). Among employees, those employed in SMEs receive less support for training than those in large companies 34 , mirroring the pattern of inequality in participation levels documented in the previous Section.

Another dimension of coverage gaps concerns the types of training undertaken. Any current employer is unlikely to have an incentive to provide support for training that will help a worker to change company or sector. However, such training is important to prevent skills gaps in light of the labour market transformations ahead. This results in a risk of underinvestment of employers in certain types of trainings and in particular for non-company-specific skills 35 , including transversal skills which are increasing in importance on the labour market. 36  Only in three Member States (AT, EE, IE), the available support for professional transitions was assessed as adequate by national adult learning experts. In the public consultation, 88% of respondents agreed that additional policy efforts are needed to support training in transversal skills, and 93% agreed with respect to digital skills. 37

Existing public support schemes in the Member States are often insufficient to close these coverage gaps. Income tax incentives for training exist in a majority of Member States, but require upfront investment by individuals and a sufficiently high income. Schemes providing individuals with training entitlements upfront such as training voucher schemes could help and exist in 21 Member States, however often with small target groups 38 such as the unemployed and hence not supporting a broader range of “preventive” trainings for transversal skills or professional transitions.

Barriers to devoting time to training, related to a lack of income to cover the costs of living during training periods: Lack of time for training also limits participation. 39  Again, the employer plays an important role in providing support by giving the opportunity to participate in training during working hours, or allowing individuals to make use of existing paid training leave provisions. Paid training leave allows workers to keep their salary or a replacement income during periods of training that may range from a few days to several months. Related provisions of some sort already exist in 24 Member States, of which 12 are bound by their ratification of the ILO Paid Educational Leave Convention. 40  However, access to paid training leave is often restricted based on features of the employment relationship. Take-up of existing provisions is usually below 1% of the employed population, driven by restrictive eligibility criteria (e.g. excluding employees without permanent contracts), a lack of information about available training opportunities and a lack of trust in the quality of training among employees and employers. Employees in SMEs are also at a disadvantage since smaller enterprises often face greater difficulties in handling temporary absence of staff. 41  Moreover, the self-employed (including the growing and diverse groups of the “dependent self-employed”) are usually excluded entirely from schemes that provide income support during periods of training. 42

Fragmentation of existing support systems: The fragmentation of existing support systems 43  further contributes to gaps in financial support. One example of fragmentation are weak links between sectoral support systems. A worker may contribute to a sectoral training fund for most of his/her working life, but cannot use it to train for a professional transition and loses access to such entitlement the moment the worker changes to a different sector. 44 Fragmentation can also create or perpetuate support gaps if responsibilities for support are strictly divided by target group and the coordination between different support schemes does not keep up with labour market changes. 45 An example are platform workers who often neither benefit from the training support offers for employees nor from those for the unemployed. Fragmentation can also prevent individuals from making use of available support, as individuals need to be well informed about the eligibility conditions to be able to benefit from it when support is linked to the current employment or income status. 87% of respondents in the public consultation highlight that a lack of awareness of available financial support for training hampers participation. 46

2. Limited incentives and motivation of individuals to participate in training

80% of adults who did not participate in learning during the previous year say that they did not want to participate, 47 showing that financial support is necessary but not sufficient.  This is not due to a lack of appreciation of adult learning, as 96% of adults agree that learning throughout life is important. 48  The reasons why this does not always translate into action are most notably a limited awareness of one’s own skills needs, limited transparency about available training offers, uncertainty about the quality of training provision and recognition of training outcomes in the labour market and an insufficient tailoring of training offers to individual needs.

Limited awareness of own skills needs by individuals: There is a large divide between employers’ and employees’ views on the adequacy of skills: Only 16% of employers report that their workers have all the required skills, while 76% of adults do not think that they lack general skills and 71% do not think they lack technical skills. 49  The literature suggests that the truth may lie somewhere in the middle as employees may be reluctant to acknowledge missing skills while employer reports must be assessed in light of the wages and working conditions offered. 50  

Limited transparency about available training offers: The advent of digital and blended learning opportunities creates an increasing number of potentially relevant training opportunities. 51  Since employers play an important role also for signalling skills needs and suggesting concrete training programmes, navigating the training offer is particularly difficult for individuals or trainings without a strong link to a single employer. 89% of respondents 52  to the public consultation pointed to limited transparency of information on training opportunities as a reason preventing individuals to train, consistent with the 90% of respondents in the Cedefop adult learning perceptions survey agreeing that more information and guidance would encourage participation in learning.

According to the adult learning expert network, 20 Member States already have legally mandated career guidance offers in place, which can help individuals with increasing the awareness of skills needs and training opportunities. However, they often have low effective outreach, as measured by the percentage of adults who have received career guidance in the past year in the AES. 53 Most national public authorities are already conducting awareness-raising campaign on existing adult learning opportunities. 54  Survey respondents identify the internet as most frequently used source of information on training opportunities. This points to the potential of user-friendly digital registries of training opportunities that are recognised on the basis of their quality and labour market relevance in the provision of career guidance services more broadly understood. However, such registries only exist in 12 Member States, and are less likely to exist in Member States with lower adult learning participation rates. 55

Uncertainties about quality and recognition of a training programme: To engage in training, individuals need to believe that it will pay off for them personally, but they need to make choices among a large number of training offers based on imperfect information and with substantial uncertainty about individual returns. The lack of quality assurance has been pinpointed as a reason for low participation in recent national initiatives. 56  Almost a quarter of respondents across the EU (24%) do not have sufficient confidence in the overall quality of adult learning in their country, and 87% of respondents are of the opinion that increased quality standards would encourage participation in work-related training. 57   80% of respondents 58 to the public consultation agreed that uncertainties about the quality of training prevent individuals from accessing training. Although a number of international quality labels exist, the quality assurance reports underlying the certification are usually non-public, leading to little visibility as to what the labels actually represent. 59  Moreover, there is a risk that the coexistence of multiple quality labels hampers their ability to reduce uncertainties about quality and recognition of training programmes. 60  

The currently high uncertainty about the recognition of learning outcomes further reduces incentives for training. If learners cannot evidence the value of newly acquired skills and competences, they may not be able to reap the full benefits of training (i.e. negotiate wage increases or improve job content and prospects). The 2020 evaluation of the Council Recommendation on validation of informal and non-formal learning concluded that while there has been some progress towards providing more validation opportunities in the Member States, strengthening support to individuals is necessary to promote the take-up of these opportunities. 61 Historical evidence suggests that in particular in times of rapidly changing skill requirements, a lack of recognition of skills reduces the bargaining power of workers and reduces wages. 62  The existence of digital registries of recognised training opportunities could support the standardisation and recognition of skills, but as mentioned above, such registries do not exist in most Member States.

Insufficient tailoring of training offers to individual needs: It reduces the motivation of individuals to take up training in view of the heterogeneous needs of adult learners in terms of the timing and the delivery mode of training. Related barriers to training participation are among the most frequently mentioned in both the AES and the Cedefop adult learning perceptions survey 63 , and managers in more than 40% of enterprises report that employees have no or a small influence on management decisions concerning training and skills development. 64  77% of respondents 65 to the public consultation agreed that insufficient tailoring of available training to individual needs prevents individuals from accessing training, and 87% of respondents to the Cedefop adult learning perceptions survey reported that better adaptability to individual learning needs would encourage participation in work-related training. Adult learning experts indicate that supporting individuals directly has been recognised as a policy priority by two thirds of Member States in the top tier of adult learning participation, as compared to one third of Member States in the bottom tier (Annex 8.2).

2.2.2 External Drivers

In addition, a number of external drivers are relevant insofar as they increase the importance of training or risk reducing its provision in the absence of a policy response.

Technological change: It makes some jobs increasingly susceptible to substitution by digitalisation and automation while at the same time creating new jobs. Estimates suggest that around 9% to 14% of jobs in the EU face a high risk of automation, and that a further 32% to 40% of jobs are likely to be transformed by automation, entailing corresponding changes in their skill requirements. 66   Changes in the technology used at work are already felt by 43% of adult employees in Europe according to the European Skills and Jobs Survey. 67   This indicates that many workers may require training to perform job tasks which are potentially very different from those they have previously undertaken. On the upside, the digital transformation also creates many new online learning opportunities for adults that can reduce cost and distance barriers and allow for better targeting of learning to individual needs, further accelerated by COVID-19 pandemic. 68  However, the present lack of certainty about quality and recognition prevents adults from fully seizing these opportunities (Section 2.2).

The green transition: It is driving a profound transformation of labour markets, not only displacing workers but also by changing the task requirements within occupations and creating new jobs that require new skills. 69  Substantial investments are needed to provide opportunities for acquiring new skills to those who are at risk of losing their jobs in high-emissions industries. 70  Skills shortages already constrain the transition to a greener economy 71 , highlighting the importance of skills investment also to reach the ambitious EU emission reduction targets.

Demographic ageing: Europe has an ageing and shrinking labour force. With people living longer, there is an increasing need for people to update their skills. Ageing is also expected to further increase old-age dependency ratios. 72  This suggests that there will be an increased demand to raise employment rates of older workers as well as productivity of employed adults across the board, whereby training will play a key role. At the same time, older people participate significantly less in adult learning than younger workers (see Annex 6.3), which highlights the urgency of creating learning opportunities that are also tailored to the diverse needs of older learners.

High atypical employment and more frequent labour market transitions: The share of workers in atypical forms of work is high and the frequency of professional transitions has increased. Around 4 in 10 adults are in atypical forms of work (defined as not having full-time permanent employment contracts). However, most groups of atypical workers participate less in learning (cf. Section 2.1). While the share of atypical workers has remained rather stable in recent years (cf. Section 2.4) the number of professional transitions from one labour market status to another throughout one’s working life has increased considerably. 73  This increases the need for training to adjust to new job tasks; it is, however, less likely to be supported by the current employer who has fewer incentives to support such transitions into new jobs (cf. Section 2.1). A high share of atypical forms of work and an increasing frequency of labour market transitions both risk reducing the coverage of employer support for training on the labour market: individuals may not have a single employer to turn to for support, or this employer may be more reluctant to grant it for fear of not being able to capture the returns of their investment once their employee moves on.

2.3 What are the consequences?

The low and unequal participation in adult learning across the EU reduces prosperity and produces inequalities between those who benefit from learning and those who do not. Beyond that, socio-economic consequences are also felt by enterprises and the society as a whole (see Annex 10 for a literature review, and Section 6 for quantitative estimates).

2.3.1 For individuals

Many immediate consequences of low and uneven participation in training are observed at the individual level. Low-skilled individuals are more likely to face unemployment and exclusion from the labour market. 74 These consequences are particularly pertinent given that workers facing the highest risks of displacement in the future are also the ones that currently participate less in learning (Section 2.1). For those in employment, less training means foregoing wage increases. 75 In terms of risks to well-being more broadly, the literature points to negative associations with personal and job satisfaction, health and social inclusion for individuals that participate less in training. 76  

2.3.2 For enterprises

73% of EU employers considered difficulties in finding workers with the right skills as an obstacle to investment in 2020. 77  Skills shortages reduce productivity and enterprises’ capacity to adopt new technologies 78 , which is particularly costly in the face of the digital and green transitions. There is also evidence that companies who fail to invest sufficiently in the skills of their workforce are more likely to go out of business. 79  

These consequences are particularly relevant for SMEs and micro-enterprises, whose staff participate in training comparatively less than that of larger enterprises. This reduces their opportunities to develop personally and professionally and impacts on the attractiveness of SMEs as a workplace. It also reduces the productivity levels of the SME, increasing the competitiveness gap with larger companies. 80  

2.3.3 For society as a whole

At a macroeconomic level, income and productivity losses by individuals and enterprises and a less efficient reallocation of labour add up to lower GDP levels (ceteris paribus) and higher unemployment with associated transfer spending. 81 Skills are also associated with a range of non-monetary externalities concerning civic participation and social cohesion, and there is growing evidence to suggest that the relationship is causal. 82  

Taken together, this suggests that low and unequal levels of participation in adult learning make the EU less resilient in the face of economic downturns and structural change. This is of particular concern in light of the rapid digital and green transition as both transitions will require both an openness to adapt to change and an ability to shape the change, for instance in the form of innovations that advance the EU on the path to carbon neutrality.

2.4 How will the problem evolve?  83

Extrapolating recent trends, adult learning participation in the EU could increase from 37.4% in 2016 to 48.6% in 2030, but would stay below the EU-level target of 60%. Large differences in participation rates between Member States are likely to persist, reflecting differences in investments in and policy attention for adult learning. Thus, existing support gaps will not disappear without further policy action. External drivers like the green transition, digital transformation and demographic trends will persist and even partly accelerate in light of the ambition to reach climate neutrality by 2050.

At the same time, labour market transitions from one job or employment status to another are expected to become more frequent. This is expected to result both from increasing life expectancy and extrapolating the significant decrease in job tenure observed across EU Member States over recent years (Section 2.2.2). While there is no compelling evidence that atypical forms of work will increase further overall 84 , some forms of it such as platform work are expected to grow. 85 These are typically associated with fewer benefits in terms of support to training.

There is broad recognition at EU and Member State levels of the importance to increase adult participation in learning in order to increase the resilience of labour markets and ease necessary professional transitions (Section 1). The measures adopted to tackle the impact of the COVID-19 pandemic, notably Next Generation EU, could provide a strong impetus for reform and some Member States have included reforms of adult learning policies in their recovery plans (see Annex 8.1). However, pressures on public budgets in the medium-term increase the importance to devise sustainable funding arrangements through a comprehensive policy framework.

3. WHY SHOULD THE EU ACT?

3.1 Legal basis

The envisaged legal basis is Article 292 combined with Article 149 TFEU. 86  Article 149 allows for non-binding measures in the field of employment aimed at providing comparative analysis and advice as well as promoting innovative approaches. It contributes to the implementation of Article 145 TFEU, according to which, “Member States and the Union shall (…) work towards developing a coordinated strategy for employment and particularly for promoting a skilled, trained and adaptable workforce and labour markets responsive to economic change”.

Article 166 TFEU specifies that Union action shall aim to “facilitate adaptation to industrial changes, in particular through vocational training and retraining”. However, the ultimate and predominant purpose of the initative is to support Member States in reaching the objectives of employment policy (cf. also Section 4), leading to Article 149 TFEU as the legal basis (next to Article 292). Article 153 TFEU, which would allow for measures to “support and complement the Member States [on] working conditions”, was also discarded. Its relevant paragraph 1) b. on working conditions would only cover workers whilst leaving out self-employed, unemployed, and inactive people. In light of the problem analysis, it was essential to choose a legal basis which could cover the entire population on the labour market.

3.2 Subsidiarity: Necessity and added value of EU action

The problems described above are widespread across EU Member States and have similar underlying causes, as shown in Section 2. The structural transformation of labour markets in light of the digital and green transitions and an ageing European population greatly increase the stakes for all Member States. An EU initiative in this area is warranted by the EU’s interest in having a skilled labour force as reflected in its objectives (Section 4.1). This is particularly important for upward socio-economic convergence across the economic and monetary union 87  and the EU in general in a context of rapid shifts in skills demand within and across sectors. Increasing access to training opportunities can also improve the resilience of labour markets to shocks such as the one induced by COVID-19 pandemic.

While Member States can take measures to improve the situation at national level (see baseline scenario in Section 5.1), an EU initiative can support and accelerate national efforts by sharing experiences and promoting innovative approaches (see Section 6.2 for a discussion of the available instruments). It can also help set expectations and create trust among Member States and stakeholders for an increase in public and private investment in skills, sending the message that all of them will contribute to, and benefit from, the shared asset of a skilled EU labour force. 84% of respondents 88 to the public consultation agreed that the initiative could also lead to a more efficient use of EU funds for skills development.

While the main EU added value of the initiative is to support Member State reforms to reach EU level objectives in the field of employment policy and the European Pillar of Social Rights, there are a number of additional dimensions of EU added value: The initiative can facilitate access to training for the growing number of EU citizens who reside in another EU Member State, contributing to labour mobility within the EU and a further integration of the single market.  89  The initiative can also contribute to creating a level playing field for companies operating on the single market by enhancing the quality and transparency of training markets across the EU. 75% of the public consultation respondents 90 agreed that the initiative can improve the portability and recognition of training outcomes across Member States, and 71% 91 agreed it can increase the transparency of national training markets for companies operating on the single market.

Article 149 TFEU precludes measures that would lead to a harmonisation of the laws and regulations of the Member States in this area, respecting subsidiarity by allowing Member States to adapt the measures taken to national circumstances.

4. OBJECTIVES: WHAT IS TO BE ACHIEVED?

4.1 General objectives

The general objective of the initiative is to support Member States with reforms to empower adults to participate in training in order to increase participation rates and reduce skills gaps. The initiative thereby contributes to the EU's objective to promote a highly competitive social market economy, aiming at full employment and social progress (Article 3 TEU).

The initiative also aims to contribute to the implementation of several individual social rights that have been derived from this general objective in the context of the European Pillar of Social Rights: the right to “quality and inclusive education, training and life-long learning” (first principle), the right to “active support to employment” (fourth principle), and the right to training “regardless of the type and duration of the employment relationship” (fifth principle). Alongside the other inititives from the European Skills Agenda, the initiative aims to support Member States in making progress towards the EU-level target of 60% adult participation in training every year by 2030. This target has been set by the European Pillar Social Rights Action Plan and endorsed by EU leaders in the Porto Declaration and in the European Council conclusions of June 2021 (see Section 1).

Significant progress on these dimensions will be crucial in order to support the just transition to a carbon neutral Europe, to equip citizens with the skills they need to grasp the opportunities resulting from the digital transition, and to adapt to change and uncertainties such as those ensuing from the COVID-19 pandemic.

4.2 Specific objectives

To reach the general objective, the specific objectives of the initiative are to support Member State reforms to:

-Close gaps in existing support systems so that all adults have access to support for training, including for professional transitions and irrespective of their labour force or professional status. 92

-Increase the incentives and motivation of individuals to seek training.

5. WHAT ARE THE AVAILABLE POLICY MEASURES?

5.1 What is the baseline from which options are assessed?

Under the baseline scenario, adult learning will continue to be an important strand of the EU’s policy work and funding. The most important recent Skills Agenda initiatives include the 2020 Council Recommendation on Vocational Education and Training and the Pact for Skills 93 launched in November 2020. Several of the Recovery and Resilience Plans to address the socio-economic impact of the COVID-19 pandemic agreed between the Member States and the Commission include relevant initiatives 94 , and the European Social Fund+, the European Regional Development Fund and Erasmus+ remain important funding sources for skills development. 

However, existing EU policy initiatives focus mostly on the design of the provision of education and training for adults (the “supply side”), rather than on making this provision accessible to adults and encouraging them to use it (the “demand side”). The 2019 stocktaking report on the implementation of the Upskilling Pathways Recommendation and the 2020 evaluation of the Council Recommendation on the validation of non-formal and informal learning show that a key challenge is to ensure a sufficient uptake of new opportunities that are created on the supply side.  95  Moreover, existing EU policy initiatives are often limited in scope. For instance, the 2016 Council Recommendation on Upskilling Pathways focuses primarily on one level of skills (basic skills) and one target group (low-skilled adults). Hence, under the baseline scenario, the EU’s policy support will not fully cover Member State policy reforms and programming of EU funds to reach the objectives of the initiative.

At the national level, 24 Member States have identified increasing participation of adults in learning as a priority in legal acts, policies, or strategies, which can be expected to translate into further concrete measures relevant to the general and specific objectives of the initiative. 96    However, on the basis of past experience, overall progress is expected to be uneven across Member States and groups of adults. While most Member States have for instance training voucher schemes in place (giving individuals an entitlement to a specific type of training), target groups are usually small and the overall impact on participations levels limited. This is also reflected in the Recovery and Resilience Plans by Member States: among the 18 plans that had been endorsed by the European Commission by August 2021, 7 include schemes providing individuals with training entitlements, however often with a limited number of expected beneficiaries (ranging from 3 500-30 000, with the exception of 500 000 expected beneficiaries in EL- cf. Annex 8.1). Some of the supporting services that could address the limited incentives and motivation of individuals to take up training (such as registries of recognised training opportunities, validation and guidance services) exist in many Member States, but they often have low effective outreach and are not systematically linked to financial support instruments. 97

The baseline scenario for this impact assessment hence assumes that adult learning participation until 2030 stays on its modest trend growth observed between 2007 and 2016, reflecting the ongoing policy efforts at Member State and EU level and the combined impact of the external drivers (Section 2.2.2). While adult learning participation dropped in 2020 following COVID-19 restrictions, direct impacts of the pandemic on adult learning participation are are likely to be transitory (Annex 6.1). However, expected growth in adult learning participation under the baseline scenario is not fast enough to reach the EU-level targets 98 and to address the challenges facing the EU. Moreover, the inequalities in access to learning opportunities observed between different groups of adults (Section 2) are not expected to narrow, given the inconclusive trends in participation gaps (Annex 8.3) and the systematic patterns of under-representation also observed in Member States with high participation rates.

5.2 Measures discarded at an early stage

A number of policy measures have been discarded for inclusion in this initiative, in light of the existing policy initiatives at EU and Member State level in the baseline scenario and the specific objectives of this initiative. 99

Tax incentives for individuals: Income tax incentives for training exist in 16 Member States. 100 Their potential to increase incentives and motivation of non-participating adults is limited by a significant delay between expenditure and reimbursement (via the income tax return), requiring significant forward planning by individuals. While income tax incentives have a large potential target group, progressive income tax regimes imply that they often favour high-income earners, and by design they cannot support individuals who do not pay any or little income tax.  This limits their potential to close existing support gaps. This is further exacerbated by a structural lack of integration with measures that could provide information on training opportunities and their quality and recognition. 101

Subsidised loan or savings schemes for adult learning: Their high administrative burden makes them more appropriate for one-off formal education rather than shorter and more frequent non-formal learning to update skills throughout working life. They are better suited for already motivated and informed individuals, and less effective in incentivising and motivating current non-learners. They are also less well suited for low-income learners, who may face restricted access to credit and saving. Existing schemes point to generally limited outreach. 102

Financial incentives for training for employers and training funds: Both instruments are an important part of training support systems in many Member States. 103  They can increase training provision by employers by internalising the broader benefits that their training investment will confer to other employers, individuals and society at large, and related initiatives are supported by the Commission. 104 They are outside of the scope of the present initiative in light of its objective to close access gaps to training so that all adults receive support to training, including those not well-served by training from an employer, as well as its focus on individual’s incentives and motivation to take up existing training opportunities.

Supply side policy guidance and funding: Public funding to education and training providers is used to support adult learning in many Member States, and plays an important role in particular for training for the unemployed and vulnerable groups. 105  Other policy initiatives from the Skills Agenda and EU funding support Member States to further develop the supply side of their adult learning systems. However, the evaluation of the Council Recommendation on the validation of informal and non-formal learning highlights that better training offers on the supply side need to be complemented with incentives on the demand side to significantly increase outreach to individuals. 106  The present initiative covers policy measures to ensure that the substance of existing initiatives on the supply side reaches individuals and empowers them to participate in training.

5.3 Description of the policy measures

Identified policy measures need to contribute to closing the existing gaps in financial support. This requires particular attention to groups that are often not well served by the current support systems as described in Section 2. Measures need to also increase individuals’ motivation and incentives to seek training as evidence shows that financial support is important but not sufficient for participation. They are grouped in three areas and combined into packages (Section 5.4). Respecting the principles of subsidiarity and proportionality, choices on key design parameters of all measures are left to the Member States as well as the Advisory Committee on Vocational Training (ACVT) 107  to respect national collective bargaining for instance concerning the funding of training.

Area 1: Individual training entitlements

For the purpose of this impact assessment, individual training entitlements are defined as a recurrent personal budget that is at the individual’s disposal to cover the direct costs of his/her training/course fees within a set time period. Also guidance, skills assessment and validation of offers may be eligible for funding from this budget. Training entitlements can close gaps in financial support by channelling funding for skills directly to the individuals in need of training. They can also increase individuals’ incentives and motivation to seek training by providing salient financial support 108 and by allowing for content, timing and delivery mode of training that is more tailored to individual needs and professional ambitions. In the public consultation, more then 80% of respondents agreed that individual training entitlements are effective at tackling financial constraints and at increasing incentives and motivation influencing participation in training. 109  

There are two main delivery modes to provide individual entitlements and they can be targeted at different groups, as discussed in the sub-section below. The precise amount of entitlements, the funding source and other key design features would be left to Member States. As stated in the ACVT Opinion, Member States are best placed to make these decisions in light of differences in national costs for training and existing arrangements for the funding of training. Member States would be asked to ensure that the implementation of the Recommendation does not reduce existing financial support for training by employers or training providers, but brings about a significant net increase in skills investments in order to satisfy the training needs not well covered by existing support systems.

Area 1.1: Delivery mode of training entitlements

This impact assessment analyses training vouchers (measure 1.1.1) as some kind of them existed in 21 EU Member States in 2020. Another delivery mode are personal accounts (measure 1.1.2). Whilst not widespread in the EU, personal accounts are an innovative approach that has been receiving increasing attention in the policy debate. 110

Measure 1.1.1: Training vouchers

For the purpose of this impact assessment, we define training vouchers as a delivery mode for training entitlements through which a specific funder (e.g. PES) funds a specific type of individuals’ training (e.g. in digital skills). Thus, the individual’s flexibility in using the entitlement is limited. There may be also other modalities and conditions depending on the scheme (e.g. the period during which the entitlement has to be used). This policy measure addresses concerns voiced by stakeholders, in particular employers 111 , to consider the correspondence of training to labour market needs.  

Measure 1.1.2: Personal accounts

For the purpose of this impact assessment, we define personal accounts as a delivery mode for training entitlements that allows individuals to accumulate training entitlements over time in order to use them on whichever training, guidance or validation opportunity they deem the most useful and whenever they see fit within the boundaries of the scheme. 112  They de-couple training entitlements from their original funder and give individuals full ownership over the entitlements. They allow for a portability of entitlements during professional transitions and can facilitate cost-sharing between different funders, such as public authorities and employers, by allowing different funders to contribute to the account. This delivery mode addresses concerns voiced by some stakeholders (including the European Trade Union Committee for Education and civil society organisations such as the AGE Platform and FEPS/Jaques Delors Institute 113 ) about the need for broad autonomy in how individuals can spend their training entitlements in order to promote a culture of developing skills throughout working life. The possibility to accumulate entitlements over several years was also specifically supported by some stakeholders. 114

Area 1.2: Target group for training entitlements

Training entitlements could be provided to a specific target group (measure 1.2.1) or to all working age adults (measure 1.2.2).

Measure 1.2.1: Defined priority target groups

This approach can ensure that those who are in most pressing need of training can access it. These target groups can be defined on the basis of the employment or contract status, level of skills or the sector of employment, as well as other circumstances (i.e. disability). This measure addresses stakeholders’ concerns about potential displacement effects of private skills investments (“deadweight loss”) 115  when opting for a broader target group. Such effects can be reduced by successfully targeting only those individuals who would not have been able to participate in learning otherwise. It corresponds to the approach followed by most Member States to date 116 and is in line with the view expressed by some stakeholders, such as AONTAS (Ireland’s national adult learning organisation) or SMEunited (an employers organisation), that attention should be focused on specific vulnerable groups. 117

Measure 1.2.2: All working age adults 118

This approach includes e.g. all those currently in employment who would like to train for new tasks to be able to keep their job or prepare for professional transitions. It addresses issues related to the fragmentation of support schemes for adult learning, and can facilitate the promotion of a culture of developing skills throughout working lives. This is the approach taken by FR, NL and Singapore. 119 It is in line with the view that the initiative should promote a universal right and be available to all learners, expressed in the public consultation by stakeholders such as the Lifelong Learning Platform (LLP), the European Trade Union Confederation (ETUC) and Eurocadres (trade unions) or WKÖ (an Austrian employers’ organisation). In its dedicated opinion, the ACVT agrees that while the initiative should focus on the upskilling and reskilling of the workforce (workers/employees/self-employed), it may also tackle training for other target groups such as the unemployed and inactive individuals, retired people or young persons not in employment, education or training. 120  

Area 2: Career guidance

Career guidance services support individuals in their training and career choices, direct individuals to skills assessment and validation procedures, and can provide assurance concerning the quality of training opportunities and the subsequent recognition of training outcomes on the labour market. This can increase individuals’ incentives and motivation to seek training by raising awareness of training needs and available offers, improving the “match” between individuals and training opportunities and training outcomes.

The importance of guidance, and its integration into broader strategies to increase adult learning, has been emphasised by the Council. 121  It was also underlined by resondents from all stakeholder groups in the targeted consultations and positions papers submitted in the public consultation. 122  92% of public consultation respondents 123 agreed that guidance is effective in increasing incentives and motivation to participate in training, reflecting near universal support from all groups of stakeholders. Also the ACVT’s opinion on the initiative emphasises the importance of quality, effective and inclusive guidance and counselling services are essential.

Measure 2.1: Making career guidance available to all

Under this measure, Member States would be recommended to ensure that career guidance is available to all. Career guidance services already exist in almost all Member States, though with a large variation in effective outreach (Section 2.2). Such services often accompany initiatives that provide individual training entitlements. 124  This policy measure would underline the importance of outreach and cooperation with all relevant stakeholders as well as up-to-date labour market and skills intelligence to inform guidance. The detailed modalities of career guidance provision would be left to the Member States (and may be made available to individuals free of charge or considered as eligible for funding from training entitlements while ensuring it is available to all).

Measure 2.2: Career guidance as a pre-condition for the mobilisation of the training entitlement

Under this policy measure, the use of the training entitlement would be subject to compulsory guidance, with a view to improve the match between the selected learning opportunity, the labour market and the individual skills needs. This reflects current provisions or policy reflections in some Member States 125 , and addresses stakeholder concerns (in particular by employers such as SMEunited) about a lower efficiency of spending on training entitlements without specific guidance.

Area 3: Enabling framework

Beyond career guidance, additional policy measures can create an enabling framework complementing individual training entitlements and thereby contributing to reaching the objectives of this initiative.

Measure 3.1: Public registry of recognised training, validation and career guidance opportunities

A public registry of recognised training opportunities can increase individual’s incentives and motivation to seek training. It does so by centralising information on available training opportunities, thus facilitating the “match” between individual and training opportunities. It has also the potential to reduce uncertainty about the quality, labour market relevance and recognition of the registered training on the labour market by delegating the relevant checks to a third party, tackling the problem of imperfect information by individuals. 126  A comparison of the experiences of the English and Scottish training entitlement schemes also point to the importance of having such a registry for the success of related initiatives (see Annex 13). A registry of recognised offers provides an opportunity to ensure that the substance of existing recommendations, notably the Council Recommendation on VET, becomes tangible for individuals. 127

Comparable registries already exist in some Member States (Section 2.2), and their establishment has been specifically supported in the targeted consultations by some stakeholders such as the European Mentoring & Coaching Council. 128 In the public consultation, 86% of respondents 129  agreed that such registries are effective in increasing incentives and motivation to participate in training. Respondents also largely agreed that skills intelligence should play a strong role in the selection of training opportunities included in the registry, and that public authorities, employer organisations and trade unions should play a strong role. While there was broad consensus among stakeholders that strong quality assurance mechanisms are essential to ensure trust in and acceptance of a scheme of individual training entitlements, employer organisations (such as SMEunited) pointed out that eligibility rules need to be sufficiently flexible to ensure a correspondence of the training offer to (changing) skills needs on the labour market.

Under this policy measure, Member States would be recommended to establish a public registry of recognised training, validation and career guidance opportunities. Member States would be recommended to ensure adherence to quality assurance and the labour market relevance of the included training offers by establishing a durable governance mechanism that considers evidence from skills intelligence and social partners and foresees periodic updates to the registry, with the view to support flexibility. Member States would be recommended to also make the registry easily accessible, notably by digital means, and user friendly for individuals.

Measure 3.2 Paid training leave

Paid training leave schemes allow individuals to take time off work for training while still receiving their salary in its entirety or in part, or alternatively an allowance to cover the costs of living from public or social partner funds. They can cover gaps in financial support concerning the indirect/opportunity costs of training that is not covered by regular employee training during working hours, and for individuals with weak or no links to an employer. They can therefore increase incentives and motivation for individuals to train by addressing the time obstacle to training.

Paid training leave provisions of some sort already exist in 24 Member States, but with variable coverage and often low take-up (Section 2). In the public consultation, 82% of respondents agreed that paid training leave for employees is effective to address time constraints to participation in training, and 85% agreed to make allowances for the cost of living during training available to non-employees. 130 Trade unions emphasised the need to strengthen the existing paid training leave provisions in the Member States in position papers and the targeted consultations. 131  

Member States would be invited to introduce paid training leave provisions or reassess the existing provisions to ensure they cover all types of employed individuals, including on financial support for employers whose employees make use of paid training leave (in particular SMEs who face greater challenges, cf. Section 2). They would also be recommended to strengthen the links between paid training leave provisions and other available support measures. The detailed modalities of provisions would be left to the Member States.

Measure 3.3 Effective governance for continuous improvements

Effectively closing support gaps and increasing individual’s incentives and motivation requires an effective governance to ensure sustainable funding and efficient spending. It involves a constant monitoring and evaluation of the success of the existing support schemes and possible adjustments to them. For instance, priority target groups may change as the labour market evolves. Also systematically integrating relevant EU Recommendations 132 and the experience from evaluations and observed take-up rates among different groups of adults can help a support scheme reach its objectives more effectively and efficiently, including concerning effective outreach to vulnerable groups. 133  With this policy measure, Member States would be recommended to set up funding and governance arrangements that pay particular attention to the needs of micro-entreprises and SMEs and allow for an effective integration of financial and non-financial support for training. They would also be recommended to ensure outreach and awareness-raising tailored to the needs of the potential beneficiaries of the scheme for individual learning accounts, jointly with social partners, civil society organisations, regional and local organisations and other relevant actors and networks.

In line with the principle of subsidiarity, these governance arrangements would be set at national level, with the view to evaluate and adjust as necessary, for instance, concerning the amount of training entitlements, priority target groups, the registry of recognised training opportunities or dedicated outreach and awareness campaigns. The importance of including external stakeholders, such as social partners, was underlined by employer representative organisations and trade unions. 134   This was also underlined in the ACVT’s dedicated opinion. 

Annex 5 provides an overview of the intiative’s intervention logic, linking the policy measures to the identified problems and the objectives of the initiative.

5.4 Description of the alternative policy packages

The policy measures are grouped in two policy packages (packages A and B). Packages are designed to be coherent and their policy measures are mutually reinforcing in achieving the specific objectives. For instance, making career guidance a pre-condition for the mobilisation of the training entitlement is more feasible when the provision of training entitlements is limited to defined priority target groups. At the same time, they highlight the main trade-offs that emerge on the basis of the available evidence and the stakeholder views (cf. the discussion above under Area 1).

Package A combines measures 1.A.1 (vouchers as a delivery mode) and 1.B.1 (defined target groups). It hence aims to empower adults to participate in learning by closing gaps in existing support systems in a targeted way, focusing financial support on those most in need of training and on providing the skills which are the most needed on the labour market at a certain point in time. Vouchers are the delivery mode that fits best this approach because they enable the funders to decide who they give the voucher to and what type of training the voucher should cover. Measure 2.2 (Career guidance as a precondition for the mobilisation of the training entitlement) is also included in package A, as it allows to ensure that the funder’s assessment of labour market relevance is reflected in the choice of training.

Package B combines measures 1.A.2 (personal accounts) and 1.B.2 (covering the whole adult population of working age), with the view to operationalise the right to training as a universal right. It responds to the argument that in a context and changing skills requirements and more frequent labour market transitions, everyone should receive incentives to participate in training, no matter their current employment status or their level of skills. Personal accounts are the delivery mode that fits best this approach as they allow for entitlements to follow individuals throughout their careers and they facilitate cost-sharing, which is essential when the entitlement covers the whole adult population of working age (employed, unemployed, atypical workers, platform workers, etc.).

Policy package B has 2 two sub-options, B.1 and B.2: under package B.1, all beneficiaries receive the same amount of entitlements (measure 1.2.2), whereas under package B.2 training entitlements are modulated according to the target group, with those having particular training needs as determined by the Member States receiving a higher amount of entitlements (new measure 1.2.3). This is due to the fact that during the analysis, the complementarity between some elements of package A and some elements of package B became clear. This also follows the OECD’s and the ILO’s recommendation for universal support that is however modulated based on recipient characteristics. 135  

Both packages A and B include measure 2.1 and measures from area 3 because they are considered as essential to the success of the initiative irrespective of the package chosen. Only training, validation and career guidance opportunities included in the public registry of recognised opportunities would be eligible for funding from training entitlements under all packages, to promote quality, labour market relevance 136 and recognition of the training undertaken. Under package A, this condition complements the requirement of guidance as a condition for mobilising the training entitlements. However, the public registry would be accessible also for those who do not receive training entitlements.

Table 1: Overview of the policy packages

6. WHAT ARE THE IMPACTS OF THE POLICY PACKAGES?

6.1 Analysing the main impacts

This section presents the main social and economic impacts of the policy packages as well as impacts on fundamental rights. No direct environmental impact was identified, and potential indirect impacts are discussed in Section 7 under “coherence with other EU objectives”.

Significant efforts have been made to collect the necessary data and provide a quantitative assessment. However, this was not always possible. A first reason are limitations concerning the availability of evidence, notably on possible interactions between policy measures. A second reason is that respecting the principle of subsidiarity underlined in Article 149 TFEU, choices on key design parameters of the policy options are left to the Member States, including on the amount of training entitlements and priority target groups. This precludes a full measurement of their social and economic (including fiscal) impacts. Finally, impacts will depend on the degree of implementation of the recommendations by Member States and on the extent to which it can build on already existing provisions.

Therefore, the approach taken is to outline a quantification of expected impacts for specific implementation scenarios on the basis of assumptions on the key design parameters derived from the available evidence, complemented with a qualitative assessment. Impacts are assessed relative to the baseline scenario (Section 5.1), based on a view of how the problem may evolve (Section 2.4) and taking into account differences in Member States’ estimated adult learning participation levels and training costs. This is complemented by a qualitative mapping of relevant already existing provisions in the Member States in Annex 8.2. The methodology used is described further in Annex 4, and details on these scenario analyses for the quantification of impacts referred to throughout this Section, including sensitivity checks on the key parameters, can be found in Annex 12.

Across impact categories and target groups, the horizontal recommendation of effective governance mechanisms with monitoring and evaluation is expected to have positive impacts by allowing for an adjustment of the design parameters on the basis of “lessons learned”. 137  

6.1.1 Social impacts

On individuals

Participation in learning: All packages are expected to stimulate reforms in Member States that increase the participation of adults in learning compared to the baseline scenario. The impact on participation rates will depend on the size of the target group that receives training entitlements, on the number of potential beneficiaries who spend their training entitlement (“take up”), and the extent to which individuals purchase training they would otherwise not have undertaken (no “deadweight loss”, cf. Annex 11 for details). The scenario analyses below are based on the existing evidence from related schemes and from training voucher experiments. A limitation is that they do not fully consider interactions between training entitlement and the other policy measures that are included in packages A and B, and whose expected impact is discussed qualitatively below.

Package A could result in an additional 5.8 million adult learners every year across the EU if training entitlements were provided to all low qualified. This would increase their annual participation rate in 2030 considerably (from a predicted 23.4% under the baseline scenario to 35.0%). The overall participation rate would increase from 48.6% to 51.0% due to the relatively small size of the target group. Similar changes are expected for additional priority target groups considered in the scenario analyses on packages A and B.2 (the inactive, unemployed, workers not in permanent employment and workers in SMEs), with the support provided to workers in SMEs (the largest of these priority target groups) leading to the highest overall participation rate of 56.2%. 138  Package A would, therefore, not reach the Porto target of 60% participation by 2030 for plausible choices of priority target groups.

Packages B would be expected to significantly increase adult learning participation across all groups of working age adults. Package B.1 could result in an additional 33.6 million adult learners across the EU and increase the participation rate to 62.7% by 2030. Under package B.2, the corresponding figures could be an additional 36.6 million learners and a participation rate of 63.9% when providing all workers in SMEs with additional entitlements. 139  Both scenarios under package B would, therefore, achieve the Porto target of at least 60% participation by 2030.

By targeting specific groups, package A could also reduce inequalities in learning participation across groups of adults: for example, the gap in the participation rate of low qualified adults compared to the average in 2030 could decrease by 9.2 percentage points (from 25.2 to 16.0 percentage points). The impacts of packages B on inequalities in participation across groups are ambiguous: they may increase if take-up rates among under-represented groups are significantly lower, or decrease them due to the lower deadweight loss for priority target groups. Modulating support according to the target group in package B.2 tends to reduce inequalities. The larger target group of packages B allow them to better address support gaps that cannot be easily anticipated by policy makers on the basis of individual characteristics (for instance, the ambition to prepare for a professional transition).

Both packages could also reduce inequalities in adult learning participation across Member States, with stronger impacts under packages B due to its larger target group. The reason is that in Member States where adult learning participation rates are lower under the baseline scenario, deadweight loss can be expected to be less prevalent, resulting in larger increases in participation rates. Under package B.1, the increase could range from 11.6 percentage points in the Member State with the highest participation rate under the baseline scenario (SE) to 17.7 percentage points in the Member State with the lowest (RO). 140  

The limited available evidence suggests that making prior career guidance a condition for the mobilisation of training entitlements in package A may have a small negative effect on the uptake of training. 141 The evidence from the French CPF suggests that when individuals can accumulate their training entitlements in personal accounts over multiple years (as foreseen in packages B), they use this flexibility to undertake less frequent yet longer or more expensive trainings, which would reduce the impacts on annual participation rates in packages B. However, aligning the intensity of training to the current skills needs and circumstances is expected to increase its effectiveness overall. 142

Registries of recognised training opportunities can be expected to increase the uptake of training by making it less time-consuming for individuals to identify a suitable training offer. The review and strengthening of paid training leave is expected to increase the share of beneficiaries who make use of their training entitlements as it relaxes time barriers, in particular among groups often not yet covered by existing schemes such as atypical workers (Section 2.2).

Wages and employment: The impacts on wages and employment follow from the changes in participation rates and are lower for package A and highest for package B.2. Studies from a large number of countries suggest that there are substantial positive wage returns to training. Return estimates also remain significant in studies that are able to account for the selection of high-ability adults into training, and a review of these estimates suggests returns around 1% for a training course of 30 hours (Annex 10).

Regarding the impact on employment, a review of more than 200 recent studies of active labour market programmes by Card et al. (2018) identified asignificant increases in employment probability of the previously unemployed after training participation. 143  Results from macroeconomic modelling that also takes into account general equilibrium effects via higher productivity and a more efficient reallocation of workers across jobs suggest that giving low-qualified individuals training entitlements under package A could yield an additional 30 000 jobs across the EU by 2030, compared to 120 000 under package B.1 and 140 000 under package B.2 when providing additional support to all low qualified, with job gains increasing to 100 000 - 400 000 by 2040 (see Annex 12C for details).

There will be substantial heterogeneity in wage and employment returns both across individuals and training courses. The conditions on quality assurance and labour market relevance for inclusion in the registry of recognised training opportunities and effective governance provisions are expected to ensure that average returns are in line with those found in the literature. While giving individuals greater freedom in their choice of training opportunities under packages B compared to package A can be expected to lead to a higher share of training conferring more transversal skills, there is no evidence that it is less labour market relevant (Annex 11).

All packages are expected to have positive indirect effects on wages and employment outcomes by improving the participation in guidance and take-up of skills validation opportunities, which can facilitate wage negotiations or finding employment. The evaluation of the Council Recommendation on validation of informal and non-formal learning concluded that strengthening the financial support to individuals can promote the uptake of the new validation offers that have been created by the Member States. 144 81% of public consultation respondents agreed that the European initiative on individual learning accounts can add value on the validation of non-formal and informal learning outcomes, and 77% agreed on the provision of guidance services. 145

Working conditions and social dialogue: All packages are expected to improve the working conditions of the beneficiaries of training entitlements by improving access to training. Providing individuals with training entitlements is already used by some employers, mainly large companies, to improve the working conditions and motivation of their employees to take up training. Under package A, positive impacts would concern only some target groups selected by the Member State. Under packages B, this would cover all adults on the labour market, whereby the groups most in need could receive additional support under B.2.

The recommendation to revisit paid training leave provisions is expected to improve the working conditions of employed adults, benefitting particularly those who are not well covered by the current paid training leave provisions (notably those in atypical work and employees of SMEs). It would also benefit employees at large by making it easier to identify quality training opportunities and cover the direct cost of training, strengthening the take-up of existing training leave offers beyond the currently low levels (Section 2.1).

In the consultations, trade unions, such as ETUC 146 , have expressed their concern that the initiative could reduce the training provision by employers, and employers, such as SMEunited 147 , have argued that resulting national schemes should not regulate such training provision. Both packages would recommend to Member States to ensure that the implementation of the Recommendation does not reduce existing financial support for training by employers, mitigating this concern. Moreover, the decision to spend training entitlements would lie with the individual, ensuring that the resulting training would complement the existing training provision by employers.

Packages can strengthen social dialogue, by recommending to give social partners an important role in the governance of the registry of recognised training opportunities.  Packages B could further strengthen the role of skills in social dialogue by creating the infrastructure for top-ups to the personal accounts of employees in specific companies or sectors. 148

Other social impacts on individuals: Packages A and B are expected to have a range of positive impacts on individuals’ health and well-being (Section 2.3). The universal packages B is expected to create less stigma for vulnerable groups than the targeted package A in light of their approach that training is for everyone, no matter their current employment situation or skills level. 149

On education and training systems

All packages are expected to increase the size of the market for continuing training by mobilising additional skills investments, with a larger potential impact of packages B stemming from their larger scale. This creates opportunities for training providers, who are also expected to benefit from enhanced visibility for their training offer after inclusion in the registry and from a rapid invoicing process with the authority managing the training entitlements as the single contact point.

The registry of recognised training opportunities is expected to increase the quality and transparency of training offers, in particular for non-formal training offers where the scope for improvements is largest (Section 2.2). While small training providers may face greater difficulties to meet the requirements for inclusion in the registry, the effective and inclusive governance of the registry recommended to the Member States is expected to remove unnecessary access barriers. This form of governance is also expected to improve the dialogue between training providers, public authorities and social partners, which can help improve the tailoring of the training offers to labour market needs, including the skills needs of employees of SMEs. 150  

The increased transparency from the registries of recognised training opportunities can be expected to increase competition among training providers and reduce the cost of training, thereby having a positive impact on the funder of the training. This is confirmed by first indicative evidence from the French CPF, where the average price of more standardised training offers such as English language certificates or driving licences has decreased following the most recent reform (Annex 14).

On society as a whole

Packages A and B are expected to have a range of positive impacts on civic participation and social cohesion (Section 2.3). Since these benefits are expected to be proportional to the increase in adult learning participation, they can be expected to be stronger for packages B. The evidence also suggests that increasing inequality of opportunity has played an important role in the recent rise of anti-establishment sentiment.  151 This hints at broader positive impacts of closing gaps in the access to training and thereby equalising opportunities.

6.1.2 Economic impacts

Direct costs of training entitlements

The direct costs of training entitlements depend on the size of the target group, the take up and the amount of entitlements. The scenario analyses introduced above suggest that implementing package A in all Member States could result in a direct annual cost of €4.2 billion (or 0.03% of 2020 GDP) for EU-27 when targeting the low-qualified. Packages B costs would be higher, ranging from €16.8 billion /0.13% of GDP (package B.1) to around €24.5 billion / 0.18% of GDP (package B.2 with additional entitlements for SME workers). The amounts will depend on decisions by Member States regarding target groups and levels of entitlements. 152 Depending on the financing arrangement chosen by the Member States, these costs would be partly or fully borne by public authorities, employers or individuals. 153  

Additional costs for enterprises including SMEs

Additional costs may result from staff absence during training, which can lead to a decrease in productivity. Scenario analyses suggest that possible additional costs of undertaking training during working hours (either with informal permission from the employer, or making use of paid training leave provisions) would not outweigh training benefits (see Annex 12B). In the consultations, SME representatives have highlighted that the absence of staff during periods of training is a particular challenge for SMEs. Therefore, both policy packages include a recommendation to also include the financial support provided to employers whose employees make use of paid training leave in the review of the adequacy of existing paid training leave provisions, with particular attention to the needs of SMEs. 154 They also include a recommendation to consider the needs of micro-entreprises and SMEs in the design of funding arrangements.

Additional costs for public authorities

Under all packages, the set-up, maintenance and governance of the registry of recognised training opportunities would result in costs to public authorities. Additional costs may be limited for the 12 Member States where similar registries already exist. Under packages B, the set-up and maintenance of personal accounts would create additional costs. Limited information is available on the magnitude of these costs, as Member States do not systematically report them in a sufficiently disaggregated way. Estimates for the French CPF point to costs of running the system of personal accounts of about €100 million over the three-year period 2020-2022. A review of the limited available evidence suggests that administrative costs as a share of direct costs of training entitlements decrease with the number of beneficiaries from a scheme. 155  

Providing career guidance for free could result in additional costs for the guidance providers, often public authorities. However, public career guidance already exists in 25 Member States at least to some extent (Section 2.2). The simplified cost options used for reimbursement of guidance services under the ESF 156 place the cost for an hour of in-person guidance at on average €18 across the Member States, ranging from €2 to €39. Thus, no or minimal set-up costs would arise and most additional costs would result from a higher uptake. Similarly, modest additional costs for public authorities can be expected due to the increase in the use of validation opportunities.

Both packages allow for a provision of guidance services by providing orientation through the registry of recognised opportunities. Under package A, in-person guidance is a precondition for the use of training entitlements, and additional costs of this can be substantial. Therefore, expected additional guidance costs per beneficiary are significantly lower under packages B compared to A. 

The recommendation to revisit paid training leave provisions included in both packages could also lead to additional costs for public authorities. In the Member States where public authorities are already involved in the funding of paid training leave, additional costs would be limited to those resulting from expanding coverage or higher take-up resulting from the other policy measures included under packages A and B. 157  

Costs for education and training providers

Compliance with the criteria required for inclusion in the registry of training offers may create costs for training providers, which can be a relatively higher burden for smaller providers. To minimise compliance costs for training providers, both packages recommend to establish these criteria in line with existing national quality assurance frameworks and a governance of these registries that is responsive to these concerns.

Other economic impacts

On enterprises and SMEs: All packages are expected to lead to an increase in the productivity of workers and hence in the competitiveness of their employers. The literature review suggests that productivity impacts are about double the size of wage impacts, so that a 30 hour training increases productivity by about 2%. 158 In the public consultation, 77% of respondents agreed 159 that the initiative could increase the productivity and competitiveness of companies. As employees of SMEs currently participate less in training, the potential impacts of both packages on the productivity and competitiveness of SMEs are larger. Yet, while it is clear that packages B would cover SMEs, the impact of package A on SMEs would depend on each Member State’s choice of target groups, which could or could not cover SME employees.

On individuals: Individuals participating in training may face indirect costs related to their inability to earn income during periods of training or costs of transport. The initiative reduces such indirect costs by including the paid training leave provisions in all packages, and in packages B by the flexibility given to individuals concerning the timing of training, allowing them to select the most convenient time (such as a period of low economic activity to minimise income losses).

On public authorities: The increase in employment foreseen for both packages can be expected to entail in the long run higher tax returns and cost savings resulting from lower spending on unemployment benefits or healthcare (Section 2.3), with the smallest impacts for package A and the strongest impacts for package B.2.

On society as a whole: By facilitating training in transversal skills and for professional transitions, both packages are expected to improve the employability of workers and thus support successful labour market transitions in particular from declining to growing sectors of the economy. 81% of public consultation respondents 160 agreed that the initiative could support the digital and green transitions by providing relevant skills. Impacts of packages B are expected to be stronger as they can support a broader range of training to support professional transitions, with the strongest impacts for package B.2 due to its additional support for target groups with specific needs.

Both packages are expected to increase investment and business growth, in light of the evidence of skills shortages as a barrier for companies (Section 2.3). They can be expected to have indirect positive impacts on research, innovation and technology diffusion 161 throughout the economy. Impacts of packages B could be expected to be larger due to the fact that it is expected to trigger more participation in learning than package A. This is in part also due to the fact that personal accounts can stimulate additional skills investments by companies by facilitating a flexible sharing of training costs among multiple stakeholders. 162  

The estimated macroeconomic benefits include a higher level of GDP per capita in 2030 estimated at 0.23% (package A targeting the low qualified), 0.87% (package B.1) and 0.99% (package B.2 with additional entitlements for the low qualified) in comparison to the baseline, with increases between 0.33-1.4% of GDP by 2040 (cf. Annex 12C). The set-up of personal accounts in packages B is expected to contribute to macroeconomic stabilisation, as it allows individuals to accumulate training entitlements over time and spend them on training during economic downturns and also creates the infrastructure to quickly provide individuals with additional public support for skills investments during downturns. 163

6.1.3 Fundamental rights impacts

Article 14(1) of the Charter of Fundamental Rights of the EU states that “everyone has the right to education and to have access to vocational and continuing training”, which is also reflected in principles 1, 4 and 5 of the European Pillar of Social Rights. Both packages are expected to have a positive impact on fundamental rights by bridging gaps in the access to continuous training. Fundamental rights impacts are stronger under the universal but targeted support provided under package B.2, as it foresees coverage of the entire adult population of working age while also providing specific support to those most in need.

6.2 The choice of legal instrument

In light of the objective of the initiative to support reform in the Member States and the envisaged legal basis, only non-binding instruments are under consideration. A first option would be a revision of the Employment Guidelines to support Member State reforms through the European Semester process and country-specific recommendations. The most recent revision from October 2020 164 already asks Member States to “strengthen the provisions on individual training entitlements and ensure their transferability during professional transitions, including, where appropriate, through individual learning accounts”. However, as they are general by nature, they do not provide further guidance on how to do this. A Commission Communication or Recommendation could provide further guidance to the Member States and inform country-specific recommendations. However, it would not have any political ownership by the Member States, and would therefore be insufficient to incentivise the required reforms in the Member States.

The preferred instrument is a proposal for a Council Recommendation. A Council Recommendation would provide a common framework for action that could serve as basis for subsequent analysis of reforms and progress. It would provide Member States with commonly agreed recommendations on a concrete tool that can help them to make the necessary progress towards the objectives of this initiative, with a view to implement the right to education, training and life-long learning from the European Pillar of Social Rights and make progress towards the ambitious 2030 EU headline target on training participation (Section 1). A Council Recommendation would be complementary to country-specific recommendations in the European Semester, as it would provide a commonly agreed framework for the implementation of recommendations that are well suited to a Member States’ starting point. In the public consultation, 61% 165 of respondents agreed that the introduction of EU legislation to be adhered by Member States on a voluntary basis such as a Council Recommendation would be suitable to reach the objectives of this initiative.

7. HOW DO THE POLICY PACKAGES COMPARE?

Table 2 summarises the results of comparing the baseline scenario to the alternative packages A, B.1 and B.2, on the basis of the analysis from the previous Section.

Effectiveness

All packages are expected to support reforms in Member States to increase overall participation in training and reduce skills gaps compared to the baseline scenario. The increase in training participation is expected to be smallest under package A and greatest under package B.2.

Regarding the specific objective of closing support gaps, effectiveness is expected to be of medium strength for packages A and B.1. In package A, this is because of the expected small target group receiving training entitlements. In package B.1, the positive impacts of universal coverage are reduced by a risk of insufficient support for individuals who are in need of more fundamental reskilling. Package B.2 is expected to have significant positive impacts by supporting all adults with training entitlements while also envisaging additional support for individuals with additional training needs. Through personal accounts, packages B also reduce the fragmentation of support and the portability of training entitlements during professional transitions, and facilitate the cost-sharing among different funders of training entitlements.

With regards to increasing the incentives and motivation of individuals to take up training, the recommended individual entitlements, strengthened guidance, increased transparency about quality assured training offers, improved access to paid training leave and effective governance are expected to contribute to effectiveness across all packages. Impacts are expected to be moderate under package A but significant under packages B. This is because packages B allow for a greater tailoring of training to individual needs by allowing individuals to freely select among eligible opportunities and at the time that suits them best. Hence, packages B provide incentives to train also in situations that cannot be easily anticipated by policymakers for targeted support at the level of groups of adults 166 (e.g., a desire to learn new skills to prepare for a professional transition, or a sudden period of low economic activity and work intensity), and are expected to trigger a positive change in the overall approach of adults to learning.

Table 2: Comparison of the packages

Note: For the purpose of comparing the impacts of the packages with the baseline scenario, a seven-stage qualitative grading scale is used: significant positive impact (+++), moderate (++), small (+), no impact (0), small negative impact/cost (-), moderate (--), significant (---).

Efficiency

Scenario analyses on the basis of the available evidence on the quantifiable impacts discussed above suggest that the benefits of packages A and B would outweigh their costs. The estimated ratio between benefits and costs is around 3 for most packages 5 years after implementation. 167 This is in line with the finding of significant increases in GDP (Annex 12C) and the evidence from other cost effectiveness calculations in the literature. 168  It highlights that the additional skills investments foreseen under all packages can support sustainable public finances, as additional costs can be paid for with higher tax revenues from and lower transfer spending on the beneficiaries of the training. Benefit-cost ratios are estimated to be somewhat higher in Member States with lower adult learning participation under the baseline scenario (due to lower expected deadweight loss) or lower costs of training, but above one for all Member States. 169  

No clear differences between packages A and B concerning efficiency is identified in comparison to the baseline scenario: aiming at priority target groups can be expected to increase the efficiency of package A, as it is less likely that these groups would have been able to fund their training from private sources (lower deadweight). However, the smaller target group over which administrative fixed costs are spread and making prior guidance a condition for the mobilisation of entitlements can be expected to reduce the ratio of benefits to costs. There is no conclusive evidence on heterogeneity of the returns to training across groups of adults or returns to scale (Annex 10), and no evidence that giving individuals greater autonomy in their choice among quality assured and labour market relevant training opportunities under packages B reduces positive impacts or the ratio of benefits to costs (Annex 11).

Main impacts

All main impacts (social impacts, direct costs, other economic impacts and fundamental rights impacts) of the policy packages are expected to stem from resulting increases in adult learning participation, and are hence expected to be smallest under package A and greatest under package B.2. These main impacts are identified in Section 6.1.

Coherence

European Pillar of Social Rights (EPSR): All packages help to make the first, fourth and fifth Pillar principles tangible for individuals, whereby the strength of impacts is expected to be roughly proportional to the expected increase in training participation. By covering the entire adult population of working age, packages B.1 and B.2 are most fully in line with the principle 1, which states that “everyone has the right to quality and inclusive education, training and life-long learning”, while packages A and B.2 support best the implementation of principle 4 on active support to employment and principle 5 on equal treatment regarding working conditions. Hence, the impact of package A is the smallest and the impact of package B.2 is expected to be highest.

Other EU objectives: All packages are expected to stimulate reforms in the Member States that contribute to the objectives of European VET policy 170  to promote a lifelong learning culture and make continuing training more learner-centred, whereby impacts are stronger for packages B.1 and B.2 as these would recommend more significant departure from the baseline scenario in most Member States. The packages build on past DG EMPL initiatives and evaluation results. 171 They complement Commission objectives such as those of the Digital Education Action Plan 172 by proposing concrete policy measures that support an effective adaptation of education and training systems of EU Member States to the digital age. Packages B outline a tool for implementing the European Parliament’s calls for a “Skills Guarantee as a new right for everyone, at every stage of life, to acquire fundamental skills for the 21st century”. 173

All packages are expected to contribute to the green and digital transition. The fundamental reorganisation of the EU economy required to reach the related EU objectives in this area such as climate neutrality by 2050 depends on significant increases in training participation: to equip the existing workforce with the necessary skills, and to ensure social fairness and continued public support. Since the impacts of the digital and green transitions are expected to be comprehensive yet heterogeneous across sectors, package A is likely to have small impacts due to the expected small target group, whereas the universal but modulated approach for support of package B.2 is expected to have the most significant impacts.

Stakeholders’ views: Respondents in the public consultation expressed clear support for the measures included in packages B over those of package A: 77% 174 agreed that all working-age adults should receive training entitlements as compared to 43% 175 agreement to a model corresponding to package A where only priority target groups receive training entitlements. There was also broad support for allowing individuals to select their training offer independently from the registry of eligible training opportunities, again implying support for packages B over package A. 176  A majority of respondents agreed that this initiative addresses relevant constraints to training and expected positive impacts, with particular strong agreement among Member States with lower participation rates. 177  A majority of respondents also expected positive impacts of the initiative, with particularly high agreement among citizens/individuals. 178  

Subsidiarity and proportionality

Both packages respect subsidiarity by leaving decisions on key design parameters to the Member States, notably decisions on the funding source, the amount of entitlements, priority target groups or eligible training opportunities. Packages B make the more specific recommendations to provide training entitlements to all working age adults and to set up personal accounts to ensure their transferability.

These recommendations can be considered proportionate in light of the urgency of the challenge, reflected in Member States’ ambition to increase adult learning participation levels significantly by 2030 beyond levels that can reasonably be expected to be reached under the baseline scenario (see Section 5.1) or an initiative in a different form (such as a Commission Recommendation without buy-in by Member States). Specifically, the implementation scenarios presented in Section 6.1 show that the recommendation to provide support to all working age adults is proportionate to the Porto Declaration target 179 that at least 60% of all adults should participate in training every year by 2030: EU training participation would not be expected to reach 60% in 2030 with more narrow target groups for support.

They are also proportionate in view of Member States’ stated ambition to strengthen the provisions on individual training entitlements and ensure their transferability during professional transitions (see Section 1). The recommendation to set up personal accounts for training entitlements is proportionate because such accounts allow to de-couple training entitlements from their original funder and give individuals full ownership over the entitlements, which is essential to ensure the transferability of entitlements. Personal accounts also facilitate flexible funding models with cost-sharing across different funders of training entitlements and modulated support by target groups, accommodating a wide range of possible national funding and implementation models.

8. PREFERRED POLICY PACKAGE

The preferred policy package is B.2, i.e. recommending Member States reforms to have in place individual training entitlements for all working age adults modulated according to selected target groups and in the form of personal accounts. Training entitlements and personal accounts are complemented by career guidance available for all, a public registry of recognised training opportunities, paid training leave provisions and effective governance for continuous improvements. It is the package that is expected to be most effective in reaching the objectives of the initiative with a high degree of efficiency and coherence with other EU policy objectives. The practical implications and expected impacts of package B.2 are summarised in Annex 3.

The elements of package B.2 are compatible with the different starting points in terms of adult learning participation and support systems in the Member States. Most Member States already have experiences with schemes providing individuals with training entitlements and relevant infrastructure, career guidance and paid training leave offers (Annex 8.2). The implementation of the recommendations is expected to increase their impacts on training participation by enhancing their efficiency and by overcoming problems deriving from fragmentation, insufficient integration of the different elements of financial and non-financial support, or ineffective governance.

For countries where individual training entitlements are negotiated as part of collective bargaining, the recommendations under package B.2 do not suggest a departure from well-established national funding models. The recommended policy measures can instead be expected to strengthen the available non-financial support to facilitate an effective use of these entitlements by individuals, in addition to closing support gaps for those not benefiting from collective bargaining agreements.

9. HOW WILL ACTUAL IMPACTS BE MONITORED AND EVALUATED?

The Recommendation will invite Member States to work jointly with the Commission on fully exploiting and where necessary improving the scope and relevance of the collection of data at Union level concerning the investments in adult learning in national accounts and in public budgets, as well as further evidence on the factors influencing individuals’ incentives and motivation to take up training.

The Commission will be invited to establish a monitoring framework with agreed common quantitative and qualitative indicators jointly with the EMCO, to assess the implementation of this Recommendation and enable its review. An initial proposal of operational objectives and corresponding indicators can be found in Annex 15 of this impact assessment.

Member States will be recommended to implement the principles set out in the Recommendation as soon as possible and submit a plan setting out the corresponding measures to be taken at national level. The progress made in the implementation of those plans should be discussed in the context of the multilateral surveillance of the European Semester in the EMCO.

The Commission will be invited to assess and evaluate the progress made in the implementation of this Recommendation, in cooperation with the Member States and after consulting the stakeholders concerned, and report to the Council within five years from the date of its adoption.

ANNEX 1: PROCEDURAL INFORMATION

1. Lead DG, Decide Planning/CWP references

Lead DG: Employment, Social Affairs and Inclusion.

Decide planning number: PLAN/2020/7916.

Commission Work Programme reference: Promoting our European Way of Life; initiative No. 36 b): Follow-up to the European Education Area and the updated Skills Agenda/ Individual learning accounts. 180

2. Organisation and timing

The preparation of the impact assessment was supported by the Inter-services Steering Group (ISG) to which the following DGs were invited: CNECT, EAC, ECFIN, ESTAT, GROW, JRC, JUST, LS, REFORM and SG. The ISG group met three times: on 15/06/2021, on 16/07/2021 and on 18/08/2021. The ISG also steered an external study supporting the Impact Assessment.

3. Consultation of the RSB

An “upstream” meeting with the RSB took place on 23 June 2021. The impact assessment report was submitted to the RSB on 1 September 2021, and discussed with DG EMPL in a meeting on 29 September 2021. On 1 October, the RSB issued a positive opinion with reservations.

The RSB reservations and the revisions introduced in response to them are summarised in the Table below:

4. Evidence, sources and quality

The following evidence has fed into the Impact Assessment:

·External study “Study to support the Commission explorations for a possible EU initiative on Individual Learning Accounts” by Fondazione Giacomo Brodolini. The Terms of Reference for the external study and the reports were shared with the ISG.

·A validation workshop on the preliminary findings of the supporting study with national adult learning experts and representatives from Member States public authorities and social partners.

·Country reports by the network of independent national adult learning experts located in the EU Member States on the “environment for providing direct financial incentives for adult learning to individuals”. The IA also draws on substantive prior work of this network that has supported DG EMPL since 2016 as referenced in the footnotes.

·Results from the forthcoming update of the Cedefop “Financing adult learning database”, used for the analysis of existing provisions in the Member States in Annex 8.2.

·Literature review on “The state of play of evidence about the conditions under which individual-oriented instruments for incentivising adult participation in learning are effective” by Sarah Baiocco on behalf of the European Expert Network on Economics of Education.

·Policy brief and analytical report on “Adult learning during Recessions in Europe” by Giorgio Brunello and Marco Bertoni on behalf of the European Expert Network on Economics of Education.

·Analytical input by the European Commissions Joint Research Centre on forecasting adult learning participation for the baseline scenario by Federico Biagi, Giorgio Di Pietro and Zbigniew Karpiński.

·Study to determine the necessary basic features of ILA models for an efficient use of off-the-shelf tools and simplified cost options under the ESF+. 

·Ad Hoc Report on the feasibility of ‘off-the-shelf’ tools for Individual Learning Accounts in the context of the external study “‘Off-the-shelf’ solutions for post-2020: A study complementing the ESF+ impact assessment” by PPMI.

·Targeted stakeholder consultations and a public consultation, as referred to in footnotes and summarized in Annex 2.

·Relevant literature and data as referred to in footnotes.

ANNEX 2: STAKEHOLDER CONSULTATIONS

1. Introduction

This synopsis outlines the consultations that were organised as part of the work on the initiative on individual learning accounts (ILAs) and presents their main findings in support of the impact assessment.

2. Consultation strategy

In line with the Better Regulation Guidelines wide stakeholder consultations were carried out. The mapping exercise identified the following groups of individuals or organisations:

·those having an interest in the matter (e.g. national public authorities, social partners, industry/businesses, training providers, NGOs)

·potential beneficiaries of an ILA (adults on the labour market)

·experts, e.g. researchers, consultancies and advisors, international organisations

The stakeholder consultation included targeted consultation events and the public consultation. Stakeholders could send comments on the Commission’s inception impact assessment as well as provide written statements, in response to the targeted consultations and the public consultation.

Table A2.1: Overview of the stakeholders reached through each consultation tool/method

Eleven targeted consultation events were held, with different groups of stakeholders. The list includes the high-level forum on ILAs (nearly 800 participants from 48 countries) and the validation workshop focused on the problem and impact analysis (around 25 participants, mostly NGOs and public authorities).

Table A2.2: Targeted Consultations

The 12-week public consultation was launched by the European Commission on 23rd April through an online questionnaire (including both open-ended and closed questions) and completed on 16 July.

216 respondents replied to the consultation comprising 78 citizens, 26 public authorities, 46 business associations and enterprises, 26 trade unions and 40 NGOs/other respondents. The majority of answers came from Belgium (35), of which 23 belonged to EU-level organisations, followed by Italy (24). In addition, 38 unique written responses to the public consultation were received from selected stakeholders.

3. Inception Impact Assessment

The Inception Impact Assessment was available online for public feedback between 23 March 2021 and 20 April 2021. 23 contributions were received, mostly from trade unions (9), NGOs (5) and EU citizens (3). Almost half of the responses stressed the importance of designing ILAs in conjunction with social partners, building on their knowledge. The quality assurance of training was also highlighted as were guidance and information. Almost half of the contributions stressed the importance of targeting vulnerable groups.

4. Targeted consultations

Problem definition

There was a general over-arching agreement on the main challenges (increasing transitions; automation, digitisation and decarbonisation; skill shortages etc.). Indeed, some stakeholders (including employers’ organisations and trade unions) underlined the urgency of these challenges, from their ‘on the ground’ perspectives. There was a general view, and especially from practitioners (including employer organisations and trade unions) that the Covid-19 pandemic was contributing to the acceleration of change (especially digitisation).

Most trade unions and some academics and NGOs stressed that the underlying problem was the insufficient application of fundamental rights to participate in training across Member States and urged the Commission to focus on ways to ensure compliance with principles 1 and 4 of the European Pillar of Social Rights. 182 A response from the European Trade Union Federation set out a 12-point plan to guarantee these rights across all Member States.

The general opinion of stakeholders on the barriers to training was that time was as important as finance. Trade unions in particular stressed the importance of employees being allowed to pursue training during work-time (reporting mixed experiences, especially the variances in the availability of paid training leave). There was also a general opinion across all stakeholder groups that a lack of information on training opportunities, and how to access them, was a barrier to participation. Those stakeholders with a particular involvement in training systems highlighted as a barrier the uncertainties (by potential trainees) as to the quality of training and its value in the labour market (dependent on recognised qualifications).

Policy options

The interest of the EU in promoting adult learning and placing the individual at the centre through training entitlements, was welcomed. In this respect, there was a general agreement amongst adult learning experts 183 that the ILA approach offered more opportunities than a voucher scheme to encourage lifelong learning and to improve employability. The majority of stakeholders (including social partners), and especially those from Member States with more advanced training systems e.g. Nordic countries, stressed the importance of integrating new initiatives with the existing national (or regional) training infrastructure, respecting the collective agreements for training already in place, normally involving employers, trade unions, and governments. Positive examples were given of collective bargaining and agreements.

There was a general agreement that the training focus of a new training entitlements initiative should be labour market oriented. This was stressed in particular by employer organisations, while some experts and NGOs saw ILAs as an opportunity to promote wider lifelong learning including civic responsibilities and other citizenship skills.

The target selection for a new training entitlement initiative drew varied views. There was support for a universal approach, but also a concern that such an approach might favour the higher skilled and those better resourced to navigate the labour market, unless there were safeguards to avoid disadvantaging priority groups (e.g. low-qualified, unemployed etc.).

There was some support from trade unions for a focus on older workers and high skilled (e.g. in IT sector where the technologies rapidly develop) – the argument being that these groups were strongly affected by the changes at the labour market and by the digital transition.

Some trade unions felt whilst putting the individual at the centre was positive, the onus on the individual to anticipate and plan for employment transitions in the context of changing labour should not absolve employers and public authorities from their responsibilities.

There was a general consensus amongst stakeholders on the importance of:

·quality information, especially for low skilled groups; whilst websites and digital applications 184  were useful, they need to be accompanied by other outreach and other activities

·free and easily accessible guidance

·quality assurance and accredited training; they promote participation raising trust in training and its perceived value

·the validation of non-formal and informal learning

·social dialogue and the involvement of social partners in ILA governance

Impact

There was a general expectation that training entitlements could be one tool as part of a wider package of measures to motivate more adults to participate in learning and close gaps in training systems by incentivising individuals from priority groups. Additional skills training would lead to economic benefits. There was general agreement that impacts would be enhanced where entitlements are coordinated with free guidance, quality assured systems and accredited training.

5. Public Consultation

Problem Analysis

The direct (tuition fees) and indirect costs of training (including loss of income due to time taken up by training) were highlighted as barriers to accessing training respectively by 192 89% and 176 82%. 185 There was also a high level of agreement that there was insufficient awareness of available financial support for training (187, 87%). Other barriers mentioned included poor accessibility to training opportunities in rural areas and a lack of access to IT.

Concerning low motivation to train, 175 respondents (82%, including 60 citizens) fully or somewhat 186 agreed that insufficient awareness of the benefits of training was a factor whilst 188 88% mentioned uncertainty about which skills were needed to improve employment and income prospects. 192 (89% – the highest response to this question) highlighted fragmented or insufficiently transparent information on available training opportunities, with 171 80% highlighting uncertainty of the quality of the training available. 177 82% agreed that uncertainty about whether training outcomes will be recognised by employers was a factor and 166 77% highlighted insufficient tailoring of training to individual needs. 189 88% agreed that a lack of time was a constraint with 178 83% agreeing that rigid time patterns of training provision was a factor.

When asked about obstacles to higher level of training provision in the labour market, 150 69% agreed that a lack of instruments to share training costs between companies, individuals and public authorities was a factor, 180 83% (including 34 enterprises/business associations) highlighted a lack of capacity within small, medium and micro businesses to organise training for their employees. 179 83% (including 65 citizens) thought that a lack of support to atypical workers was an obstacle.

Objectives and EU added value

Respondents were asked whether a European initiative on ILAs could add value. Respondents agreed that there could be added value related to:

·increased transparency about national training markets – 154 71%

·portability and recognition of training outcomes across Member States – 163 75%, 22 fully disagreed

·more efficient use of EU funds for skills development – 181 84%, 10 fully disagreed

·development of registries of quality assured training opportunities at the national level – 168 78%, 15 fully disagreed

·implementation of quality assurance for non-formal training opportunities – 168 78%, while 20 fully disagreed

·validation of non-formal and informal learning outcomes – 176 81%, 15 fully disagreed

·career guidance – 166 77%, 12 fully disagreed

·provision of paid educational leave and its take up by individuals – 167 77%, 15 fully disagreed 

There was a general plea to avoid complexity and bureaucracy (also highlighted in position papers e.g. by employer organisations).

Policy efforts to support learning amongst adults were widely supported by respondents, with very few disagreements:

·short job-related training - 185 86%

·more fundamental job-related training - 195 91%

·basic, soft and inter-personal skills - 189 88%

·digital skills - 200 93%

·skills for green transitions - 198 92%

Support was less strong for non-job-related training: 158 73%

Entitlements and individual learning accounts (ILAs)

On approaches for tackling financial constraints influencing participation in training, the majority agreed with:

·establishment of ILAs – 182 84% (including 21 public authorities and 22 trade unions)

·establishment of training entitlements in other forms, such as vouchers -177 82% 

·tax incentives for companies 188 87%

·subsidies to training and education providers - 148 68%

·increased overall public funding to support training - 189 88%

·facilitate the sharing of training costs between companies, public authorities and individuals - 167 77%

Motivation to participate in learning could be increased through:

·ILAs - 178 82% 

·training entitlements in other forms, such as vouchers – 177 82% 

·public registry of quality assured training opportunities - 185 86% (including 72 citizens and 38 NGOs)

·one-stop shop digital platform (e.g. smartphone app) linked to a registry of training opportunities - 189 88% (including 71 citizens, 34 enterprises and business associations)

·subsidies to education and training providers - 158 73%

·awareness raising campaigns - 190 88%

·in-person advice and guidance - 198 92%

·opportunities for the validation of informal and non-formal learning - 192 89%, (notably 37 NGOs)

·increased modularisation of training offers (e.g. creating opportunities for the certification of short courses) - 181 84%

Respondents agreed that time constraints to participation in training could be addressed through:

·ILAs – 135 63%, 24 fully disagreed

·training entitlements in other forms – 136 63%, 22 fully disagreed

·modularisation of training – 181 84%, 4 fully disagreed

·paid educational leave – 177 82%, 9 fully disagreed

·allowance to cover living costs -182 85%, 5 fully disagreed

On universal, targeted or mixed approaches to individual training entitlements, 166 77% agreed in a universal approach open to all working age adults (15 fully disagreed), with 168 78% for a universal approach combining a greater level of entitlement to those with particular training needs. 92 43% agreed to targeted support only for those individuals with particular training needs 187 , but more respondents (111 52%) somewhat or fully disagreed. 188 The position papers also reflected the debate between universal and targeted entitlements, with concerns that universal schemes could widen gaps in the labour market, benefiting those with higher skills, and excluding vulnerable groups without the resources and knowledge to access training systems, whereas others arguing that access to training was a fundamental right for the benefit of all citizens.

On the rules for spending training entitlements, 191 88% agreed that there should be a free selection by individuals from the registry of eligible training offers and that training should be allowed to take place during working hours with the agreement of the employer. Only 83 38% agreed that there should be a requirement for training to take place outside of working hours, and only 77 35% agreed that individuals should have to follow compulsory guidance on the selection of training (with 118 55% somewhat or fully disagreeing).

In respect of sources to increase funding for training, there was broad agreement on the use of public funding by Member States (197, 91%). 147 (68% including 24 trade unions) agreed with an employer’s levy (with 31 fully disagreed) and 115 53% agreed with individual contributions, but 85 (including 22 trade unions and 24 citizens) somewhat or fully disagreed.

111 80% agreed that strengthening monitoring and exchange of experience would be effective in increasing the uptake in training. 84 61% agreed that the initiative should be promoted via an instrument on a voluntary basis (such as a Council Recommendation but with countries deciding what and how to implement the recommendation), 43 31% somewhat or fully disagreed. 61 44% agreed that there should be no additional instrument/ that the existing EU-level framework is sufficient, but slightly more respondents (64 46%) somewhat or fully disagreed. Position papers from employer representative organisations and trade unions argued for a role in the development and governance of any new initiatives, given their experience in developing collective agreements.

Impact

The consultation asked whether impacts related to fundamental and social rights could result from an ILA initiative. The majority of respondents thought that ILA could:

·improve access to secure and adaptable employment – 162 75%

·improve health and well-being – 161 75%

·promote active citizenship – 141 65% 

·make it easier for individuals to manage labour market conditions – 179 83%

·improve employment prospects for the unemployed – 174 81%

·help tackle discrimination in respect of access to training, employment and career progression - 162 75%

A second question focused on labour market and economic impacts. Lower majorities fully or somewhat agreed that the initiative could:

·support labour mobility - 143 66% with 54 in somewhat or full disagreement

·reduce skills gaps and mismatches – 172 80%

·improve productivity and company competitiveness – 167 77%

·support efficient reallocation of labour – 164 76%

Finally, on societal and environmental benefits 174 81% agreed that the initiative could support green and digital transitions by offering the relevant skills whilst 147 68% agreed that cohesion in society could be improved. These generally positive views are reflected in many of the position papers (from NGOs, employer organisations and trade unions) but with the caveat that implementation is critical, as ILAs need to be developed in conjunction with national training systems supporting existing arrangements.

ANNEX 3: WHO IS AFFECTED AND HOW?

1. Practical implications of the initiative

If this initiative is adopted, its key obligations will apply for the Member States’ public authorities. Public authorities will have to consider policy and/or legislative changes to:

(a)set up governance arrangements to determine the amount and modulation of training entitlements (for all working age adults and with higher amounts of entitlements for target groups with particular training needs) in the form of personal accounts, as well as corresponding funding arrangements;

(b)set up the public registry of recognised training, validation and career guidance opportunities;

(c)ensure career guidance are available for all working age adults;

(d)introduce or reassess paid training leave provisions and their integration with the other support measures.

Changes should consider inclusive digital solutions taking into account needs of person with disabilities or less digitally skilled persons. Training providers who want to receive funding from training entitlements will need to ensure that their training offers meet the requirements to be included in the public registry of recognised training, validation and career guidance opportunities. For enterprises or individuals, additional obligations may result depending on the funding arrangements chosen by public authorities.

2. Summary of costs and benefits

Table A3.1: Overview of benefits - preferred option

Table A3.2: Overview of costs - Preferred option

3. Summary of the SME test

Table A3.3: The "SME test"- summary of results

ANNEX 4: ANALYTICAL METHODS

This impact assessment draws on a variety of quantitative and qualitative methods and data sources, (Annex 1). Stakeholder opinions have been captured via extensive targeted consultations and a Public Consultation, which closed on 16 July 2021 (Annex 2). Overall, care has been taken to validate information via triangulation from different sources and reflect minority opinions.

As set out in Section 6, the report assesses social impacts (including participation in adult learning, wages, employment, working conditions and social dialogue, health and well-being etc.), economic impacts (including the costs and benefits for individuals, education and training providers, enterprises, public authorities, as well as the wider impacts on the economy in terms of GDP for instance) and the impact on fundamental rights. This integrated approach covering both the micro- and macroeconomic perspectives requires a combination of quantitative and qualitative assessments.

In this annex, we outline: 1) the quantification of impacts; 2) qualitative approaches and 3) constraints to the assessment of impacts and how they have been addressed.

1. Quantification of impacts

The quantitative assessment of this impact assessment consists of three parts:

-an assessment of the impact on participation rates, which feeds into Section 6.1.1. on social impacts, and in turn provides the basis for the cost-benefit analysis and macroeconomic assessments described below;

-a cost-benefit analysis which feeds into Section 6.1.2 on economic impacts and contributes to the assessment of efficiency of the packages in Section 7;

-a macroeconomic assessment of the impacts on output and employment, which feeds into Section 6.1.2 on economic impacts and also contributes to the assessment of efficiency of the packages in Section 7.

First an assessment is made of the impact of providing training entitlements on levels of participation in training (Annex 12A). The aim here is to estimate the impact of training entitlements on training participation, bearing in mind that not all people who receive training entitlements will use them (take up) and that some people who use their training entitlements might have trained in any case without the entitlement (deadweight loss). The estimates of take-up and deadweight loss have been taken from experiences with past schemes giving training entitlements to adults. The selection gives preference to a small number of experimental studies (see “constraints” for an explanation), while placing them in the broader context of experiences from larger-scale schemes.

The approach to estimating impacts on training participation for packages A and B.2 (where it is up to Member States to select priority target groups for training entitlements) is to present estimates for a range of plausible priority target groups, selected on the basis of the problem analysis (Section 2). Impacts are estimated for the reference year 2030 by applying the take-up and deadweight rates identified. 2030 is chosen so as to leave sufficient time for impacts to materialise and to allow for a comparison of impacts to the 2030 EU targets on adult learning participation. Impacts are assessed relative to the baseline scenario, factoring in expected baseline changes in training participation rates (Annex 8.3) and population changes (derived from Eurostat projections of population change to 2030). The estimates presented in the main report are based on “middle ground” estimates on the key parameters (take-up and deadweight loss), and Annex 12A also presents sensitivity checks for reasonable “pessimistic” and “optimistic” values on both parameters.

The analysis of likely impacts on participation rates then feeds into the cost-benefit analysis (CBA) of the policy packages (Annex 12B). The CBA estimates the short to medium term costs and benefits of the policy packages on individuals, employers and public authorities. Key costs are the direct costs of training entitlements, administrative costs and the costs of devoting time for training, while key benefits concern higher productivity and wages as well as higher tax revenues for public authorities and savings on social benefits resulting from positive employment impacts. Overall estimated ratios of benefits to costs are derived on the basis of the participation rate estimates from Annex 12A, relevant estimates from the literature and Eurostat data. A limitation of these estimates is that they are of a “partial equilibrium” nature, i.e. they do not take into account some more complex impacts that are likely to affect GDP and employment levels in the longer run.

Macroeconomic modelling using the BeTa model (see Annex 12C) is hence used to complement the cost-benefit analysis. It is able to control for the complex economic interactions which might result from, for example, impacts of additional taxes used to fund the training entitlements and automation in response to higher productivity. In looking at the benefits of training it is also able to capture the wider impacts and is not limited to quantifying the impact on those directly affected by the entitlements (i.e. it is able to capture various externalities). Accordingly, the BeTa model is able to estimate how increased levels of participation in training might affect output and employment by adjusting the stock of human capital available to an economy, and the impact this may have on the efficiency with which people are matched to available jobs. As a forecasting model, it is able to provide an assessment of medium to long-effects on output and employment, providing impact estimates for 2030 but also extending them to 2040.

2. Qualitative methods

While the quantification of impacts is primarily used to support Sections 6 and 7, qualitative methods were used throughout the impact assessment to support the problem definition, the definition of policy measures as well as to assess the impacts of the policy packages, in particular the social impacts and the impacts on fundamental rights, in addition to the economic impacts, where quantitative data is limited.

An in-depth systematic literature review was implemented in order to collect detailed and comprehensive evidence about the main problems and issues as well as the effects of different training schemes and interventions.

Seven case studies were conducted (see Annex 13) with the objective of (i) gaining a better understanding of the context (socio-economic and policy framework) in which financial instruments for adult learning are being implemented in selected Member States, (ii) identifying recent and relevant policy initiatives as well as “lessons learned” that could inspire other Member States, and (iii) explore the working mechanism for such policy instruments and their potential to respond to specific contextual challenges and policy problems. Country experts were engaged to collect the required information and institutional stakeholders’ opinion. In selecting the Member States, a set of criteria was adopted to ensure geographical representativity at the EU level and reflect other dimensions such as heterogeneity of adult learning systems and good coverage of the EU population but also relevant international experiences.

In addition, an in-depth analysis of the French Compte Personnel de Formation was carried out (see Annex 14). The development and dynamic adjustments of the French scheme in response to experiences made represents an interesting example in particular to assess the costs and benefits of personal accounts for training entitlements.

The outcome of the consultations of experts and stakeholders also contributed to the qualitative analysis in the report, as described in Annexes 1 and 2.

3. Constraints

Since this impact assessment supports a non-binding proposal for a Council Recommendation, a first challenge is uncertainty about Member State’s responses and degree of implementation. The approach taken in the context of this impact assessment is to estimate impacts on the basis of clearly stated assumptions derived from the available evidence. These estimates hence do not attempt to estimate the output of the relevant political processes in the Member States after the adoption of a Council Recommendation (e.g. by scaling down impacts according to some estimate of partial non-implementation), in order to provide useful and independent input into these processes. This approach avoids the risk of “circularity”, whereby a Member State would not adopt a recommended measure on the basis of evidence of limited impacts from an impact assessment that already assumes this lack of adoption.

The assessment of impacts faced additional challenges related to limited availability of quantitative data on which to base estimates:

·Experimental data is needed for a rigorous identification of the key parameters of interest, as observational data does not allow to observe deadweight loss (due to self-selection of motivated individuals into the scheme) and often also does not provide good evidence on take up rates (as existing schemes are often targeted at specific groups). Therefore, the quantification draws on evidence from a limited set of field experiments, in particular a voucher experiment that assigned training entitlements to a randomly subset of the entire Swiss adult population, which allows for a rigorous assessment of take-up and deadweight loss. 193  

·Estimates of deadweight loss are often not able to factor in impacts of training entitlements on the intensity of training in response to receiving training entitlements or future training intentions (i.e., individuals would have undertaken some training even in the absence of entitlements, but train longer or more frequently over a certain period thanks to entitlements), leading to an over-estimation of deadweight loss.

·There is some uncertainty in the literature as to whether returns to training are constant, increasing or decreasing (Annex 11), and this matters in particular for the estimates of long-run impacts in the BeTa model (Annex 12).

·Relatively little data is available on interactions between different policy measures, e.g. the impacts of training entitlements in the presence of strengthened guidance of paid training leave provisions and vice versa.

·Availability of data on costs (set-up and operational) is limited, and where these data are available they tend to reveal wide variation across countries and there is always a degree of uncertainty as to whether like is being compared with like.

·The benefits of wider benefits of adult learning (on job and life satisfaction, health etc.) and on some policy measures (e.g. paid training leave) are difficult to quantify, and large parts of the evidence base are of a correlational nature.

The following steps have been taken in this impact assessment to address these constraints:

·Deadweight loss estimates have been adjusted to ensure that deadweight estimates reflect differing levels of training participation in the Member States as explained in Annex 12A.

·Annex 11 situates these assumptions in the broader literature. Sensitivity checks on the key parameters are conducted throughout Annex 12, to show the robustness of key results to alternative assumptions concerning take-up rates and deadweight loss, wage and employment impacts and administrative costs. In the BeTa model, decreasing returns to scale have been selected as a cautious benchmark assumption for long-run impacts of higher training participation.

·Wider expected benefits of increased participation of adults in learning are not quantified but discussed qualitatively.

More generally, qualitative data has been drawn upon in order to provide a more holistic analysis of the likely additionality resulting from the introduction of a training entitlement and capture some of the costs and benefits where quantitative data are limited. There are constraints from using these data too, including:

·Questions about the representativeness of such data (e.g. adult learning experts may have a tendency to favour greater expenditure on adult learning);

·The comparability of data across countries given the highly specific nature of some of the more qualitative inquiries.

This impact assessment recognizes these constraints and addresses them to the extent possible by triangulating information from different sources to increase the robustness of conclusions, e.g. by providing expert assessments next to survey evidence and stakeholder views. The constraints are further recognised in the design of the policy options, which include a recommendation of constant monitoring and evaluation of the success of the policy measures in reaching the objectives of the initiative in order to allow for appropriate adjustments to them where necessary (measure 3.3).

ANNEX 5: OVERVIEW OF THE INTERVENTION LOGIC

Figure A5.1 – Intervention logic

ANNEX 6: ADULT LEARNING: MEASUREMENT, KEY CHARACTERISTICS AND PARTICIPATION LEVELS

1. Measurement

For the purpose of EU-level benchmarking, adult learning is measured as participation in formal or non-formal learning among adults aged 25-64 during the past 12 months. 194  This is also the definition used for monitoring progress towards the EU’s 2030 target that at least 60% of adults should participate in training every year by 2030 (Section 1).

Formal learning is characterised by an official recognition of the learning outcome by public authorities (such as a University diploma). Non-formal learning aims to improve knowledge and skills in any areas in an institutionalised setting and may take the form of courses, workshops or private lessons, sometimes resulting in credentials and certificates that are recognised by employers. It excludes informal learning (such as learning from a friend, colleague, book, guided tour or library visit). 195  Non-formal learning is much more prevalent among adults than formal learning: in 2016 (the most recent year with available data), 34.8% of adults participated in non-formal learning, compared to only 5.0% for formal learning.

The three main EU statistical data sources on adult learning are the Labour Force Survey (LFS), the Adult Education Survey (AES) and the Continuing Vocational Training Survey (CVTS) on training provision by enterprises.

The LFS provides annual data on adult learning participation during the last 4 weeks, which has been used to monitor progress towards the 15% participation target by 2020 set out as part of the strategic framework for European cooperation in education and training (“ET 2020”). 196 However, the short reference period is considered sub-optimal for a comprehensive measurement of the short learning periods that are characteristic of adult learning. 197 Therefore, the LFS will also measure adult learning participation during the last 12 months every second year starting in 2022 198 , which will be available to monitor progress towards the EU’s 2030 target (Section 1).

The AES collects more detailed information on adult learning than the LFS, however at a lower frequency: the most recent available data are for 2016, and further data are available for 2011 and 2007 (with limited comparability between the 2007 “pilot survey” and the subsequent survey waves). It uses a 12 months reference period. The coverage of non-formal learning activities in the AES is slightly broader than in the LFS as it includes “guided on the job training”, which is closer to informal learning than the other types of non-formal learning activities and not included in the EU benchmarking framework. Most adult learning participation data presented throughout this impact assessment are hence taken from the 2016 AES, adjusted to correspond to the definition of adult learning used for EU benchmarking by excluding guided on the job training. 199

The CVTS complements the individual-level survey data from LFS and AES with information on enterprises’ training needs, planning, provision and financing from the perspective of enterprises. The results of the fifth and latest wave of CVTS are available for the reference year 2015.

2. Key characteristics

The AES allows for a characterisation of adult learning in terms of its purpose, duration, providers and funders. 200

Purpose: Concerning the purpose of adult learning, AES asks participants of non-formal learning activities whether these activities have been “mainly job-related” or “mainly personal/ non-job related”. 201  About 80% of non-formal learning in the EU was mainly job-related, a pattern that is consistent across EU Member States for both men and women, with somewhat higher shares for men (85%) as compared to women (76%). An analysis of the instruction hours spent by field of study reveals that about one fifth (18%) of adult learning in the EU concerned business, administration and law, followed by services (16%), and health and welfare (14%), arts and humanities (14%), engineering, manufacturing and construction (12%) and information and communication technologies (6%).

Duration: An average adult learning activity was 118 hours in 2016, with a considerably shorter average duration of non-formal (75 hours) as compared to formal (469 hours) learning activities (and an average closer to the duration of non-formal learning activities due to the small share of formal learning in overall adult learning). The average duration noticeable decreased compared to 2007 when it was 134 hours. 202  This decrease may reflect a move towards more frequent but shorter learning activities facilitated by digital learning offers, consistent also with the overall modest increase in participation rates over this period (Section 2.4).

Providers: Formal learning is provided by Universities, VET schools or similar publicly recognised institutions. The main provider of non-formal education are employers (35%), followed by non-formal education and training institutions (18%). All the other types of providers, from commercial institutions to non-profit associations, from trade unions to formal education institutions, etc. constitute a mosaic of providers of adult education, with none of them amounting to more than 10%.

Funders: It is common for adults to pay for the costs of formal learning themselves (55%). Non-formal learning activities are often for free (23%) or fully paid by someone else (53%). 88% of job related non-formal learning activities were at least partially fundered by the employer. 203  

3. Participation levels and differences between groups

3.1 Progress in adult learning participation towards EU-level targets

Progress in participation in adult learning over the last decade has been limited and uneven across EU Member States; the EU-level target of 15% (of adults participating in the 4 weeks reference period) was not met in 21 of 27 EU Member States in 2020 and reached 9.2% overall, compared to 7.8% in 2010 (see Figure A6.1). 204 In nine Member States, the participation rate even decreased over the last 10 years (DK, AT, ES, SI, CZ, CY, PL, SK, RO).  

Trends in adult learning participation over time need to be interpreted with caution as they are influenced by statistical breaks in some Member States (with a tendency to overstate progress over recent years). However, the overall finding of moderate progress is confirmed by analyses focussing only on those Member States for which comparable data series are available (cf. Annex 8.3). COVID-19 restrictions resulted in a drop in overall adult learning participation from 10.8% in 2019 to 9.2% in 2020 for EU-27. 205  However, adult learning participation is likely to rebound once COVID-19 restrictions are lifted.

Figure A6.1 – Progress to ET2020 objectives - participation of adults 2010 and 2020

Source: Labour Force Survey, trng_lfse_01 . Participation rate is calculated as the share of adults from 25 to 64 year in some form of learning in the last four weeks as percentage of total population. 

According to the OECD’s “Programme for the International Assessment of Adult Competencies” (PIAAC), adult learning participation in the EU is also not particularly high in international comparison, with EU average participation lagging behind the OECD average and countries such as the US, Canada or Singapore. 206  

The remainder of this Annex and the impact assessment in general will draw on the Adult Education Survey, as LFS data with a 12 month reference period is not yet available (see Section 1 of this Annex).

3.2 Participation differences between groups of adults

Access to adult learning opportunities is influenced by characteristics of the employment status, the characteristics of the job as well as individual-level characteristics (Figure A6.2). 207

Figure A6.2 – Overview of participation figures – key characteristics

Source: AES 2016.

Participation differences based on the employment status 

Statistics show consistently lower adult learning participation for groups of adulta that are likely to receive less support from an employer. For instance, permanent employees are more likely to participate in training than other adults across all EU Member States (45.4% vs. 28.7% during last 12 months, see Figure 1 in Section 2.1 of the report). Figure A6.2 above shows how this gap between groups has been relatively constant between 2011 and 2016. Details on the participation rates of the different sub-groups are provided below. To put the numbers into perspective, Figure A6.3 provides an overview of the composition of the EU adult population by labour market status.

Figure A6.3 – Population aged 25-64 years by employment status in 2020

Source: Eurostat Labour Force Survey 2020. The split of permanent workers between SMEs and larger enterprises was conducted on the basis of the AES 2016, as the LFS uses different company size categories.

Labour market status: The share of employed individuals that participated in adult learning (43.3%) is almost double that of unemployed (26%) and inactive persons (21.3%). Particularly large differences can be observed in BG, HU, LT, PL, PT, SI, and SK. The only exception to this pattern is found in RO, where the share of unemployed that participated in adult learning in the last year is higher than the share of employees. Generally, the share of inactive that participates in adult learning is lower than that of unemployed, except in DK, FI and SE.  The importance of labour market status-related variables is also apparent in regression analyses of the determinants of adult learning. 208  

Figure A6.4 - Participation in adult learning - by labour market status

Source: Adult Education Survey 2016. Participation rate in education and training – formal and non-formal, 25-64 years old. Excluding guided on the job training (GOJT).

Employed / self-employed: Employees (wage earners) are consistently more often engaged in adult learning (44.9%, group size estimated by LFS2020 as 152.1 million, or 64% of 25-64 population) than self-employed individuals (36.9%, group size estimated by LFS2020 as 25.1 million, or 11% of the 25-64 population). Particularly steep differences between these groups can be identified in RO, MT, and SK. Only in a handful of Member States, self-employed report more often to have participated in adult learning, such as in CZ, HR, EE, DE, and IT. The OECD PIAAC study found an even more substantial difference, with 35% of self-employed workers participating in adult learning yearly compared with 57% of full-time permanent employees. 209  

Figure A6.5 - Participation rate, employed and self-employed

Source: Adult Education Survey 2016. Participation rate in education and training of employees – formal and non-formal, 25-64 years old. Excluding guided on the job training (GOJT). Member States are sorted by overall participation in adult learning.

Contract status: Generally, employees with permanent contracts (estimated by LFS2020 as 134.7 million or 57% of 25-64 population) report more often to have participated in adult learning (45.4%) than employees with a temporary contract (41.4%, entire group estimated by LFS2020 as 15.8 million or 6% of 25-64 population). However, exceptions are found AT, HU, FI, IE, LU, BE, LV, EE, MT, ES, LT, and BG.

Figure A6.6 - Participation in adult learning – by contract status

Source: Adult Education Survey 2016 Participation rate in education and training – formal and non-formal, 25-64 years old. Excluding guided on the job training (GOJT). Member States are sorted by overall participation in adult learning.

Full-time / part-time: Participation patterns between employees with full-time and part-time contracts are relatively similar when aggregated at the EU level. In most Member States however, individuals with full-time contracts (estimated by LFS2020 as 146.9 million or 62% of 25-64 population) report more often to have participated in adult learning (43.8%) than employees with part-time contracts (43.5%, group size estimated by LFS2020 as 30.3 million or 13% of 25-64 population), except in SE, EE, CZ, LV, SI, EL, BG, and RO. Larger differences appear for part-time workers when comparing those with a permanent contract (48.8%), and those and without (38.8%). JRC regressions analysis on AES data shows that full time employees are significantly less likely to participate in formal education and training relative to part-time employees, whereas they are more likely to participate non-formal education and training. 210 This shows how the factors of employer support and time available affect each in their own way preferences for specific types of training activities.

Figure A6.7 - Participation in adult learning – by contract type

Source: Adult Education Survey 2016. Participation rate in education and training of employees – formal and non-formal, 25-64 years old. Excluding guided on the job training (GOJT). Member States are sorted by overall participation in adult learning.

Participation differences based on the job-characteristics

Participation in adult learning also differs considerably based on characteristics of the employer, the sector and type of occupation as presented below.

Figure A6.8 - Different patterns in participation for work-related characteristics – EU27

Source: Adult Education Survey 2016. Participation rate in education and training of employees – formal and non-formal, 25-64 years old. Excluding guided on the job training (GOJT). Member States are sorted by overall participation in adult learning.

Enterprise size: The size of the employing enterprise is consistently associated with participation in adult learning; Figure A6.8 shows how participation consistently increases for each increase in size. This patterns is consistent in every Member State. A particularly substantial difference is visible when comparing the participation (42.0%) of employees in SMEs (less than 250 employees, estimated group size 100 million, 41.7% of population aged 25-64) with the 55.1% of workers in larger enterprises (with at least 250 employees, group size estimated at 52.9 million or 21.9% of 25-64 population). 211  Differences between smaller enterprises (1-49 staff, group size estimated by LFS2020 as 78.8 million, 32.8% of population aged 25-64) and micro-enterprises (1-10 staff, group size estimated by LFS2020 as 37.9 million, 16.0% of population aged 25-64) are less profound (39.7% vs. 35.6%). The pattern of lower learning participation in smaller enterprises is confirmed by CVTS data and regression analysis controlling for other factors such as individual characteristics. Specifically, JRC regressions show that employees working in firms with 50 or more workers are found about 7.2 percentage points more likely to participate in non-formal learning compared with those employed in firms with 10 or less workers.

Sector: AES data shows steep differences between sectors, particularly between workers in the education / social sector (participation of 59.7%, group size estimated by LFS2020 as 34.9 million or 15% of 25-64 population) against agriculture (participation of 20.7%, group size estimated by LFS2020 as 7.7 million, or 3% of 25-64 population) and construction sectors (participation 29.7%, group size estimated by LFS2020 as 12.7 million, or 5% of 25-64 population). Industry (37.1%) and business services (43.5%) score on average in between these other sectors.

Occupation: Occupations that require low and medium level skills show the lowest participation figures in most Member States. Workers in occupations that require higher levels of skills (technicians, professionals and managers, estimated by LFS2020 as 80.0 million, or 34% of 25-64 population) show considerably higher participation figures (60.1%). Especially groups that run the highest risk of displacement by automation show the lowest participation figures in most Member States (Figure A6.9).

Figure A6.9 - Participation in adult learning and risk of automation

Source: Cedefop estimates from the Skills Panorama and Adult Education Survey 2016.

Estimated employment growth: Based on the evidence provided above, using Cedefop’s forecasts of future skill demand, it is possible to assess how participation is linked to occupations for which shortages are expected. Figure A6.10 shows projected growth in occupational employment (at the 2-digit ISCO level) projected to 2030 by occupation along with the percentage of people in that occupation who participated in adult learning. It reveals that there are a distinct set of occupations which are characterised by relatively low growth and low level of participation in adult learning (agricultural workers, machine operatives and assemblers, and skilled-trades workers). In contrast, there are occupations where relatively high levels of growth are projected and where the workers participate considerably more in adult learning. It may well be that the risks facing these different groups of occupations vis-à-vis their participation in adult learning may become even more differentiated in the future.

Figure A6.10 - Projected employment growth by occupation by participation in adult learning

Source:    Cedefop Skill Forecasts via Skills Panorama, Adult Education Survey 2016, own calculations

Figure A6.11 summarises the expected employment trends against participation figures at the sector level. It shows that there is a range of sectors where employment is expected to fall and where workers participate comparatively less often in adult learning (i.e. manufacturing, agriculture, mining, and water supply). The position of those currently working in such sectors may be a relatively precarious one: a relatively high risk of job loss with limited participation in training opportunities compared with people working in other occupations. This is particularly true against the background of different consequences of green jobs and occupational changes for different sectors per sector. Some sectors such as renewable energy and environmental goods and services (including water and waste management) have developed significantly and show potential for employment growth. For other sectors the impact varies, such as the construction sector, depending on the degree to which the existing built environment is greened through retro-fitting or, conversely, where the focus is on ensuring that new construction is greener. Some parts of manufacturing, notably the automotive sector, are gradually changing their output to produce more energy-efficient versions of the same product, with limited net employment gains. Other parts of manufacturing are producing green products and creating jobs in the supply chains of green sectors: an example is the production of wind turbines. 212

Figure A6.11 - Projected employment growth by sector and participation in adult learning

Source:    Cedefop Skill Forecasts via Skills Panorama, Adult Education Survey, own calculations

Participation differences based on individual characteristics

Besides the characteristic of the employment relationship, and characteristics of the occupation, individual characteristics play an important role explaining levels of participation as well, such as level of qualification, age, and gender. However, within groups of similar education attainment and age, differences observed above continue to be important determinants of participation as well.

Educational attainment: The level of education of individuals is widely regarded as one of the driving factors of participation in education and training. 213  Low-qualified adults (estimated by LFS2020 as 49.7 million, or 21% of 25-64 population) are considerably less likely to have engaged in adult learning (18.0%) than adults with secondary (33.4%) or tertiary qualifications (58.1%). Across all OECD countries, participation in adult learning by low qualified adults is 40 percentage points below that of high-qualified adults. 214 This broader pattern is largely confirmed for the EU by the AES as well, which are presented in Figure A6.12 below. Low qualified workers participate considerably less in EL, HR, PL, and RO; here the participation rates of lower qualified adults are less than a quarter of that of the general population.

Figure A6.12 - Participation in adult learning – by education level

Source: Adult Education Survey 2016. Participation rate in education and training – formal and non-formal, 25-64 years old. Excluding guided on the job training (GOJT).

Regression analysis performed by the Joint Research Centre confirms this. It also shows that the size of the effect of education is stronger for non-formal education and training and informal learning compared with formal education and training. The JRC regression analysis shows that individuals with tertiary education are about between 28 and 30 percentage more likely to participate in non-formal education and training and informal learning than those with lower secondary education or less. The corresponding figure for formal education and training is about 7 percentage points.

While education level is widely considered as one of the main determinants of an individual’s participation in adult learning, one needs a broader perspective if one seeks to address such differences in participation. The barriers experienced by individuals with lower qualifications vary, based on other conditions, such as the support available by their employer, or the need for specific training on the short term for their work. The figure below for instance highlights the differences in adult learning participation across different education levels, but also shows that the differences in support for individuals continue to contribute to participation as well; among each level of qualification, individuals with permanent contracts participate more in adult learning than workers who are not permanent employees. The figure also highlights how for each qualification level, workers in larger companies consistently participated more often in adult learning that individuals in smaller companies, while the unemployed and inactive participate consistently less. This underlines the need for better understanding the barriers that all individuals face, considering the extent to which these affect individuals with lower qualifications.

Figure A6.13 - Participation rate, by education level, employment status and size of enterprise

Source: Adult Education Survey 2016. Participation rate in education and training. Excluding guided on the job training (GOJT). Note that the two categories depicting size of firm overlap with the two categories depicting the type of contract of salaried workers\

Age: Participation tends to be consistently lower for higher age categories (Figure A6.14 below). This pattern is observed across all Member States. Particularly the higher age group (55 to 64 years – participation of 27.4%, group size estimated by LFS2020 as 60.1 million, or 25% of 25-64 population) reports considerably lower participation rates than other groups. Nevertheless, additional analysis again points to substantial differences when comparing participation in formal and non-formal learning. 215 For non-formal learning the estimates point to an inverse-U relationship: the probability increases up to age 31-32 and declines thereafter. Particularly for job-related non-formal learning, the mid-aged participate most, which is consistent with a need to update skills via non-formal learning for those with some distance to initial education. 216  On the other hand, for formal learning, a U-shape is observed, with the likelihood decreasing after age 21 and increasing again after age 61.

Figure A6.14 - Participation in adult learning – by age

Source: Adult Education Survey 2016. Participation rate in education and training – formal and non-formal, 25-64 years old. Excluding guided on the job training (GOJT).

Gender: Figure A6.15 shows that women participate slightly more (38.4%) in adult learning than men (36.4%) at the aggregated level, although the differences are small. In some Member States larger differences can be observed such as in SE (considerably more women than men participating in adult learning) or CY (considerably more men participating in adult learning). While these statistics do not suggest major differences in participation, larger differences exist in the types of adult learning that men and women participate. A total of 83.2% of non-formal learning activities of men was job-related, compared to 74.6 % for women. This pattern was found in almost all EU. Only in CY the trend is reversed, where the share of job-related non-formal learning activities was higher for women than for men (79.3 % against 66.3 %). 217

Figure A6.15 - Participation in adult learning – by sex

Source: Adult Education Survey 2016. Participation rate in education and training – formal and non-formal, 25-64 years old. Excluding guided on the job training (GOJT

Persons with disabilities: Data also suggest that persons with disabilities (estimated by EU-SILC as 14%, or roughly 33.2 million of the 25-64 EU adult population) participate only half of the average amount of individuals without disabilities. 218 While such data is not collected systematically in the Adult Education Survey, transposing the estimates about participation in adult learning from the Survey on Income and Living Conditions to the averages presented in this section would be equivalent to roughly the participation figures of the inactive population (21.73). 219 The average employment rate at EU level among persons with disabilities is also considerably lower (52.0%) than for other adults (76.2%), as a result of which they also are less likely to receive support from an employed to enrol in training. 220

4. Recommendations to increase participation in adult learning in the European Semester

Analysis of the Country Specific Recommendations over the years 2019/2020 shows that all countries received a recommendation related to skills development, mostly related to basic and digital skills. The contents of each recommendation are classified and summarised in Table A6.1. It shows how recommendations also referred to the need for strengthening lifelong learning; and improving the performance, quality labour market relevance, inclusiveness and flexibility of education and training systems addressing skills mismatches. 221  While references to the education system or inclusive education may not explicitly refer to adult learning per se, such recommendations still have a bearing for adult learning, which also covers formal learning In 19 Member States, adult learning was highlighted specifically as an area to focus system-level reforms. The table below compares Member States on their existing level of participation in adult learning (the main benchmark for this study) and categorise these in three groups. The table shows that despite the variety of different adult learning systems, Recommendations single out adult learning in most Member States, both in Member States where participation is above the EU average and below.

Several Country Specific Recommendations refer to specific target groups like low qualified/ skilled, job seekers, inactive people, older workers, people with a migrant background, Roma, and other groups with specific needs. In these cases, the Country Specific Recommendation generally address the need for increasing adult learning and improve coverage of education and training systems as well as the need to strengthen quality and labour market relevance of training offer, on which the ILA initiative could contribute.

Table A6.1: Overview of Country Specific Recommendations 2019/2020 on skills, adult learning and performance of education and training systems