Assessment of the impact of ongoing reforms in education and training on adult learning FINAL REPORT

Transcription

1 Assessment of the impact of ongoing reforms in education and training on adult learning FINAL REPORT

2 Contractor: Public Policy and Management Institute (PPMI) Gedimino ave. 50, LT Vilnius, Lithuania Tel Faks info@vpvi.lt Director Haroldas Brožaitis Research team: Haroldas Brožaitis Rimantas Dumčius Simonas Gaušas Laima Užpelkienė Assisted by: Donatas Pocius Laura Jurkuvėnienė Dovilė Rimkutė Dalia Čiupailaitė Mažvydas Jastramskis Service contract No. CONTRACT / GRU- GRUSTU with the European Commission Directorate General for Education and Culture Vilnius, 30 March 2010

3 TABLE OF CONTENTS TABLE OF CONTENTS...1 LIST OF ABBREVIATIONS...2 EXECUTIVE SUMMARY...4 INTRODUCTION THE PROGRESS OF COUNTRIES AND EU POLICY RESPONSES IN THE AREA OF ADULT LEARNING Overview of progress in adult learning and education attainment EU policy progress in the area of adult learning and instruments for stimulating national education and training reforms The development of EU policy initiatives The role of the OMC The role of the EU funding instruments CONCEPTS AND LOGIC OF IMPACT ASSESSMENT THEORIES AND PRACTICE OF ANALYSING EDUCATION AND TRAINING SYSTEMS, THEIR REFORMS AND THEIR IMPACTS Types and areas of education and training reforms Examples of impact assessment of relevant public policy and management reforms KEY STEPS IN IMPACT ASSESSMENT OF EDUCATION AND TRAINING REFORMS: GUIDELINES FOR POLICY MAKERS Deciding scope of impact assessment Intended and unintended consequences Levels of impacts Timing of impact assessment Reconstructing the logic of reform intervention Defining questions for impact assessment and assessing their feasibility Preparing technical specification Choosing methods of data collection and analysis Purpose of impact assessment and selection of methods Timing of impact assessment and selection of methods Accessing data Combining different methods of analysis Developing a research design EXEMPLARY CASE STUDIES ASSESSING IMPACTS OF REFORMS Methodology of case studies Analytical framework for the case studies Data sources Structure of case studies The choice of cases Limitations Case study 1: The establishment of labour market training and counselling system in Lithuania ( ) Case study 2: The Adult Education Initiative in Sweden ( ) Key success factors of the reforms CONCLUSIONS AND RECOMMENDATIONS REFERENCES ANNEX 1: QUANTITATIVE EVIDENCE ON FACTORS INFLUENCING ADULT LEARNING Methodology of quantitative analysis Overview of interrelationships between adult learning and education attainment The influence of policy characteristics on adult participation in learning and educational attainment of the population The influence of contextual variables on adult participation in learning and educational attainment of the population The composition and relative significance of factors explaining adult participation in learning and educational attainment of the population Conclusions

4 LIST OF ABBREVIATIONS AES Adult education survey AT Austria BE Belgium BG Bulgaria Cedefop European Centre for the Development of Vocational Training CVT Continuing vocational training CVTS3 Continuing vocational training survey 3 CY Cyprus CZ Czech Republic DE Germany DK Denmark EE Estonia EEA European Economic Area EFTA European Free Trade Area ES Spain ESF European Social Fund EU European Union EU27 European Union Member States after January 2007 EUR Euro FDI Foreign direct investments FI Finland FR France GDP Gross domestic product GERD Gross domestic expenditure on research and development GR Greece HICP Harmonised indices of consumer prices HR Croatia HU Hungary ICT Information and communication technologies IDI Information and communication technologies development index IE Ireland IS Iceland ISCED International Standard Classification of Education ISCO International Standard Classification of Occupations IT Italy IVET Initial vocational education and training LCS Labour cost survey LFS Labour force survey LLL Lifelong learning LT Lithuania LU Luxembourg LV Latvia MT Malta NACE Classification of Economic Activities in the European Community NL The Netherlands NO Norway NQF National Qualifications Framework OECD Organisation for Economic Co-operation and Development OMC Open Method of Coordination 2

5 PISA PL PPS PT R&D RO SE SI SII SK TR UK VET Programme for International Student Assessment Poland Purchasing power standard Portugal Research and development Romania Sweden Slovenia Summary innovation index Slovakia Turkey United Kingdom Vocational education and training 3

6 EXECUTIVE SUMMARY The report Assessment of the impact of ongoing reforms in education and training on adult learning was prepared by the Public Policy and Management Institute on the basis of service contract No. CONTRACT / GRU-GRUSTU with the European Commission Directorate General for Education and Culture. It explores the methodological aspects of assessing impacts of education and training as well as related reforms on adult learning, developed guidelines for conducting such impact assessments and exemplified their application in conducting two case studies: The establishment of labour market training and counselling system in Lithuania ( ) and The Adult Education Initiative in Sweden ( ). This chapter provides a summary of the main ideas presented in the report. Chapter 1 While some EU (and EFTA-EEA) countries show exemplary performance in several aspects of adult learning, there are wide and persisting disparities across the EU. Adult participation in learning differs substantially between Member States, between low-qualified and highly-qualified, low-skilled and highly-skilled, younger and older adults. The policy responses of the EU and the guidance provided to the Member States have been numerous and increasingly adult learning policy specific. The implementation of EU political initiatives was supported by the Open Method of Co-ordination and EU financing instruments. Depending on domestic political agendas and the motivation of political actors in the Member States EU level initiatives will take a considerable length of time to be translated into action and to yield tangible results. So far most of the positive developments in education and training policy at Member State level in response to EU policies were targeted at the younger population. Reinforced messages from the EU might still be able to convince more domestic political actors that adult learning needs special attention. The best way to convince them is to provide tangible evidence on the benefits of reforms enjoyed by other countries. Chapters 2 and 3 The Open Method of Coordination (OMC) process, which invites voluntary participation, cannot force change upon national institutions, but it should provide a channel of peer expertise for those willing to reform and who are looking for evidence about similar reforms in other Member States. Many reform agendas do not become reforms because of uncertainty. They are seldom supported by hard evidence on the impacts of similar reforms elsewhere. To support the efforts of national reforms the EU institutions have set up a framework of indicators and benchmarks for monitoring progress towards the Lisbon objectives in education and training. But EU-level efforts are unable to produce the sufficiently detailed and reliable evidence, which are needed e.g. in the exchanges between the Directors General from different Member States. They cannot generate reliable evidence on which national reform measures work best, what problems they are able to solve and 4

7 under what circumstances. And even more importantly - no matter how sophisticated EU-level monitoring and evaluation becomes, in the foreseeable future it will only be able to capture very basic policy design characteristics and their outcomes at Member State level. One reason for this is that the European Statistical System would take a long time to develop and apply instruments for the collection of data on indicators of policy and governance in education and training even if these were already agreed between the Member States. Only an in-depth policy analysis at Member State level could provide for country-to-country experience sharing based on hard evidence. However assessments of impacts of education and training reforms on adult learning are rather scarce. Considering that all EU-funded interventions in the area of education and training are monitored and evaluated on a mandatory basis and many of these interventions contribute to national reforms, the lack of evidence about their success is striking. More regular impact assessments of national reforms and their results shared with other interested Member States would enable much faster policy learning, which is needed for the achievement of the Lisbon objectives in education and training. Furthermore, the monitoring and evaluation of EU structural funds should give due attention to assessing the impacts of the education and training reforms they funded and especially the impacts of these reforms on adult learning. Chapter 4 Although the EU institutions provided a wealth of guidance on how to assess the impacts of EU financing instruments, the guidance on how to assess public policy and management reforms, which often go above or beyond financing interventions, has been lacking. The analysis in the Chapter 4 of the report explains the key steps in planning and conducting impact assessment and underlines the areas which require special attention and effort. It reveals that planning and implementing impact assessment after the reform has taken place severely limits the availability of data and thus also the quality of analysis. Ideally impact assessment plans have to be integrated into reform programmes, which would give those willing to study the impacts of reforms a number of advantages. Firstly, the data on reform actions, their outputs, results and impacts should be collected when it is still possible or cost-effective to obtain. Preferably control groups and baselines should be established before the reform. Secondly, the sources of data could be meaningfully triangulated and different methods of analysis could be applied which would complement each other to ensure the strength of the conclusions. Finally, consistent measurements of immediate and intermediate impacts could be utilised in providing assessment of the long term and ultimate impacts of the reform. Absence of prior planning leads to second or third best quality of evidence, which is not as useful and compelling as it could be. Chapter 5 The case studies The establishment of labour market training and counselling system in Lithuania ( ) and The Adult Education Initiative in Sweden ( ) fully support the findings of the Chapter 4. In the case of Sweden the data collection for monitoring and evaluation of reform was integrated into the reform process. The institution responsible for the evaluation of reform impacts financed independent research into the impacts of reform. This created a wealth of information, which could be used 5

8 in assessing the ultimate impacts of the reform in Chapter 5 of this report. In the case of Lithuania the reformers were preoccupied in kick-starting a vital element of any modern welfare state labour market training and a counselling system, the resources were scarce and only very basic data was collected during and after the reform. Subsequently, this reform (as well as parallel reforms) received only very limited attention in academic or applied research. Nevertheless, the case study method proved to be useful in identifying and analysing the impacts of reform. It revealed and illustrated positive and negative as well as immediate and long term impacts. Detailed case studies revealed that there are two success factors which are particularly important for reforms regardless of their geographical location: strong and active political leadership in the relevant institutions responsible for the reform that helps to establish explicit national targets of the reform and maintain its momentum throughout the implementation process; and sufficient public funding of activities of the reform, which increases ownership of the results among relevant institutions and actors. Other success factors seem to depend on the country development level. Annex The Annex to the report provides the results of an attempt to conduct a quantitative analysis on which characteristics embedded in policies explain performance in terms of participation of adults in learning, educational attainment and several other characteristics. The results of analysis are subject to a number of important limitations and therefore should be treated with great caution. Nevertheless they suggest some interesting conclusions, which might be useful in raising the hypotheses of further inquires as well as informing decisions, which have to be made under a large degree of uncertainty. More significantly we identify a number of the following policy variables of the education system which influence adult learning in the group of countries studied during the period for which data was available: stratification (i.e. numbers and types of transitions available to the next schooling level), vocational specificity, flexibility in provision of learning services, school autonomy and spending on education. The reforms in these policy characteristics would be likely to have important influences on adult learning. Statistical analysis has also provided evidence that education and training policy characteristics might not be the only and possibly not even the most significant factors influencing adult learning. Firstly, the characteristics of labour market policies seem to exert strong influences on adult learning. Secondly, there is some evidence that health policies and their outcomes might also be helpful in explaining the variation in the levels of adult learning. This means that any reforms in these policy areas could also affect adult learning. Finally, there are a number of contextual variables (eg. the level of economic development or social cohesion), which might help to explain differences in adult learning and education attainment. Most of the relationships are no doubt two-way relevant policies and context factors not only influence adult learning, but adult learning also exerts positive influences on all areas of life and policy. Statistical analysis revealed that adult participation in formal and non-formal education and training is best explained by models 6

9 involving labour market policies and even more so by models involving a large set of both education, labour market and health policies and context variables. Although the analysis suffers from a number of limitations, it once more underlines the complexity of factors influencing adult learning and the challenges that lie ahead for future inquiries in this field. The table below lists the main questions addressed in each chapter of the report. Chapter Chapter 1 Chapter 2 Chapter 3 Chapter 4 Chapter 5 Annex Main questions addressed in the chapter 1. What is the situation and progress of the EU27 Member States, EFTA-EEA and candidate countries in adult learning? 2. What are the EU policy responses and the main instruments for influencing national education and training reforms? 3. What is impact assessment? 4. Why is the impact assessment beneficial? 5. What are the main types and areas of education and training reforms? 6. What are the examples of impact assessments of relevant public policy and management reforms? 7. What are the preconditions and methodology for a good impact assessment? 8. How the methodology of impact assessment could be applied in practice, when impact assessment was planned after the reform took place? 9. Which policy characteristics and context factors help to explain the different levels of adult learning across countries? 7

10 INTRODUCTION The report Assessment of the impact of ongoing reforms in education and training on adult learning has been prepared by a team of researchers from the Public Policy and Management Institute (further - PPMI) under service contract No. CONTRACT / GRU-GRUSTU with the European Commission Directorate General for Education and Culture (further DG EAC). The work of the research team was supervised by the officials of DG EAC (Unit EAC-B4) and the members of the Focus group on analysing national reforms, who have provided their valuable contributions. The report had the following three objectives: (1) to identify appropriate methods (propose methodology) for assessing the effectiveness of reforms, innovations, trends and developments, by developing a number of key indicators; (2) to exemplify the use of the methodology proposed, by applying it to a particular reform/innovation in a particular country; and (3) to provide a review of key literature for identifying and analysing reforms, innovations, trends or developments to be covered by future studies and a proposal for countries to be analysed. This introductory chapter explains the purpose and structure of the report. Chapter 1 Following the adoption of the Lisbon Strategy in 2000, the EU has made considerable progress on the political level in promoting education and training reforms and in particular those helping to attain the Lisbon target of participation of adult population in lifelong learning. For example, the Commission s 2001 Communication Making a European Area of Lifelong Learning a Reality 1 argued that traditional systems must be transformed to become much more open and flexible, so that learners can have individual learning pathways, suitable to their needs and interests, and thus genuinely take advantage of opportunities throughout their lives. In 2002, a common Education and Training 2010 work programme was launched, which was meant to provide practical support for Member States education and training reforms through the Open Method of Co-ordination. In 2005, following revision of the Lisbon strategy, the Council adopted new Employment Policy Guidelines, which documented the slow progress in reaching the Lisbon target of lifelong learning and specifically urged the expansion and improvement of investment in human capital and the adaptation of education and training systems in response to new competence requirements 2. These are just a few examples of relevant EU policy initiatives. These have been implemented using the Open method of co-ordination and EU financing instruments. Chapter 1 of this report provides an overview of the progress the EU27 Member States, EFTA-EEA and candidate countries have made in raising the levels of adult learning and the EU policy responses. 1 European Commission Communication: Making a European Area of Lifelong Learning a reality, 2 Council Decision 2005/600/EC of 12 July 2005 on guidelines for the employment policies of the Member States. 8

11 Chapters 2-3 Chapter 4 Chapter 5 Conclusions, recommendation s and Annex Nevertheless on the practical level the evidence of which reforms provide the best contribution to achievement of EU policy targets remains relatively scarce and impedes country to country policy learning. This might be attributed to several factors. Firstly, the comparative data on policy and governance in education and training is scarce and therefore opportunities for EU-level comparative analysis are limited. Secondly, the lifelong learning situation varies greatly from country to country. Thus timing and the nature of reforms vary to a great extent. Thirdly, the impacts of these reforms are often under researched. When research is conducted, its results are often accessible only in the national language. Fourthly, available research reports employ a variety of methodological approaches and the quality of their delivery differs. Member States planning reforms face difficulties in accessing robust evidence about similar reforms in other countries. Chapters 2 and 3 of this report introduce the concepts and logic of impact assessment and provide an overview of theories and practice of analysing education and training systems, their reforms and their impacts. The EU institutions have provided considerable methodological guidance for the evaluation of EU structural funds interventions (e.g. in The Tavistock Institute 2003) or for the evaluation of EU activities (in Directorate General for the Budget 2004). However this guidance has been tailored better to measuring the impacts of policy as usual rather than policy reforms. Moreover, reforms, which can affect adult learning, often go above or beyond financing reforms; they entail regulatory changes, restructuring of public service provision and possibly even changes in income-support systems. The current document is meant to address this gap. Chapter 4 contains details of the key steps in planning and conducting impact assessment, including issues to consider when defining the scope and depth of impact assessment and choosing appropriate methods for data collection and analysis. Chapter 5 exemplifies the application of impact assessment methodology and takes into account the literature reviewed in previous chapters. It presents the framework for analysis of two selected reforms from two different EU member states and then reports the results of the following case studies: The establishment of labour market training and counselling system in Lithuania ( ) and The Adult Education Initiative in Sweden ( ). The case studies are based on extensive literature review and in-depth interviews carried out with experts, who have been involved in designing, implementing and (or) studying impacts of selected reforms. The Conclusions and recommendations section of the report summarises the analysis and main findings of all the chapters and the Annex of the report. It also provides several policy recommendations, which are based on the findings. The Annex of the report contains a literature review and quantitative macro-level analysis of which policy characteristics as well as which context factors help to explain the different levels of adult learning across the EU27 Member States, EFTA-EEA and candidate countries. These results have to be treated with caution due to the numerous limitations explained in the methodology section of the Annex. It is expected that the report will be useful for public administration officials, who are responsible for planning and assessing impacts of national reforms in the area of education and training and beyond 9

12 and specifically those, who are concerned with the impacts of such reforms on the levels of adult learning. The report might also be useful for wider audiences, including policy analysts and evaluators, practitioners in public services, private or nongovernmental sectors, who are affected by or interested in the reforms related to adult learning. 10

13 1. THE PROGRESS OF COUNTRIES AND EU POLICY RESPONSES IN THE AREA OF ADULT LEARNING This chapter sets the stage for analysing the impacts of national education and training reforms on adult learning. Firstly, it provides an overview of the progress the EU27 Member States, EFTA-EEA and candidate countries have made in raising the levels of adult learning. Secondly, it reviews the EU policy responses and the main instruments for influencing national education and training reforms: the Open method of coordination and funding programmes Overview of progress in adult learning and education attainment Adult participation in education and training as well as adult education attainment varies greatly across the EU27 Member States, EFTA-EEA and candidate countries. While a number of countries are making good progress in promoting participation of adults in lifelong learning and attaining higher levels of qualification, the disparities in Europe as a whole are immense and call for policy responses Participation in education and training In 2003 lifelong learning was identified by the Education Council as one of the pillars for achieving the Lisbon goals, and the area where progress in the European Union should be monitored against a European reference level, i.e. benchmark. Moreover, the benchmark of increasing participation by adults in lifelong learning to 12.5% among the age group by 2010 has been established as a part of the European Employment Strategy since 2003 (European Commission 2007). Figure 1 below shows that only 9 out of 31 countries for which data is available reach the Lisbon target for participation of adults in lifelong learning. Levels of adult participation in these countries range from 12.8% in Austria to even 32.4% in Sweden. Other countries are still behind the Lisbon target for adult learning. The second group of countries (Poland, Lithuania, Czech Republic, Malta, Italy, Estonia, Luxembourg, Latvia, Belgium, France, Ireland, Germany and Cyprus) demonstrates only moderate results varying from 5.1% in Poland to 8.4% in Cyprus. Bulgaria, Romania, Turkey, Greece, Croatia, Hungary, Slovakia and Portugal form the third group of countries that lag behind in terms of participation of adults in education and training. 11

14 35.0% 30.0% 25.0% 20.0% 15.0% 10.0% 5.0% 0.0% 1.3% BG RO 12.5% Lisbon target 1.3% 1.5% TR 2.1% 2.4% GR HR HU Figure 1: Adult participation in education and training (both formal and non-formal), in 2007 across the EU-27 plus EFTA-EEA and candidate countries* 3.6% 3.9% SK PT 4.4% 5.1% PL LT CZ 5.3% 5.7% 6.0% 6.2% MT 7.0% 7.0% 7.1% 7.2% IT EE LU LV BE FR IE DE CY ES 7.4% 7.6% 7.8% 8.4% 10.4% 12.8% AT SI NL 14.8% 16.6% 18.0% 20.0% UK NO 23.4% 27.0% FI IS DK SE 29.2% 32.4% * Statistics are not available for Liechtenstein and the Former Yugoslav Republic of Macedonia. Source: EUROSTAT, Labour Force Survey (indicator EM051 - Lifelong learning (adult participation in education and training) - Percentage of the population aged participating in education and training over the four weeks prior to the survey) Large differences in adult participation in education and training exist not only between countries, but also within countries. Table 1 below shows very different adult participation rates in education and training (both formal and non-formal) depending on age, fulltime/part-time distinction, occupation, labour status and qualification level attained. The table shows that people aged 55 and over participated in education and training considerably less (overall average of 4.2%) than those aged between 25 and 64 3 (overall average of 9.1%). Participation differences between highly-skilled (those who are working in ISCO 1-3 occupations) and low-skilled (those who are working in ISCO 4-9 occupations) were even larger: on average, 13.8% compared to 5.1%. Similarly, lowqualified (those who had completed vocational, secondary or lower education or training, i.e. ISCED levels 1-4) participated in education and training significantly less than highly-qualified (those who had completed tertiary education, i.e. ISCED levels 5-6): averages 6.8% and 15.9% respectively. However, there were no significant differences in adult learning participation depending on the employment status of a person on average, 9.5% of those employed and 8.7% of those unemployed participated in education and training. Although the latter group have on average lower educational qualifications they are specifically targeted by active labour market policies, which include provision of training. There were almost no differences in participation by full-time/part-time distinction on average, parttime workers participated in education and training only slightly less than those working full-time (averages 9.3% and 9.5% respectively). Therefore the highest differences within countries in participation in education and training exist between highly and low skilled adults. This is further confirmed by the results of the Adult education survey (see Figure 2). European Working Conditions 3 Groups of adults aged between 25 and 64 was selected with a reference to the Lisbon benchmark of adult aged 25-64participation in learning of 12.5% by

15 Survey also confirms that low-skilled and low-qualified workers receive less training than other workers (Lyly-Yrjänäinen, 2008). The European Commission, measuring progress towards the Lisbon objectives in education and training also declared that adults with a high level of education are more than six times as likely to participate in lifelong learning as the low skilled (European Commission, 2007). Table 1: education and training received during previous four weeks (both formal and non-formal, in %), in 2007 across the EU-27 plus EFTA-EEA and candidate countries*, by age group, full-time/part-time distinction, occupation group, labour status and qualification level Countries / Fulltimtime Part- ISCO ISCO Employed Unemployed ISCED ISCED groups AT SI NL NO UK FI IS DK SE Average of 1 st group PL LT CZ MT IT EE LU LV BE FR IE DE CY ES Average of 2 nd group BG * GR HR HU PT RO SK TR Average of 3 rd group Overall average * Statistics are not available for Liechtenstein and the Former Yugoslav Republic of Macedonia. For Bulgaria 2007 data (part-time distinction) is not available, we use 2006 data instead. Source: EUROSTAT, Labour Force Survey (data for variable EDUCW4N calculated by the authors). 1 st group 2 nd group 3 rd group 13

16 Figure 2: Participation and non-participation in education and training in EU-27 by major groups of occupations*, in Total ISCO 1-3 ISCO 4-5 ISCO 6-7 ISCO 8-9 Participation rate in formal and non-formal education and training Participation rate in formal education and training Participation rate in non-formal education and training Rate of persons not participating in education and training * - major groups of occupations are defined in ISCO classification using two dimensions of skill concept - skill level and skill-specialisation. ISCO 1-3 includes the following groups of occupations: legislators, senior officials and managers (ISCO 1), professionals (ISCO 2) and technicians and associate professionals (ISCO 3). ISCO 4-5 includes clerks (ISCO 4) and service workers and shop and market sales workers (ISCO 5). ISCO 6-7 includes skilled agricultural and fishery workers (ISCO 6) and craft and related trades workers (ISCO 7). ISCO 8-9 includes plant and machine operators and assemblers (ISCO 8) and elementary occupations (ISCO 9). For more information see ILO material at: Source: Eurostat, Adult education survey Countries differ significantly according to adult participation in different types of education and training. Figure 3 below shows that the highest difference is in Denmark where 24.8% of adults participate in non-formal education and training, while only 5.4% participate in the formal one. The difference of adult participation by type of education is also substantial for the UK (12.1 percentage points). Meanwhile, it is only in Estonia that the numbers of adults participating in formal and non-formal education and training were both equal to 2.8%. The differences of participation by type of education for other countries ranged from 0.1 percentage points in Italy to 8.2 percentage points in Finland. Figure 3: Adult participation in education and training, by type of education and training, in 2007 across the EU-27 plus EFTA-EEA and candidate countries* (presented from the lowest to the highest difference) 25.0% 20.0% 15.0% Formal education and training Non-formal education and training 10.0% 5.0% 0.0% EE IT LV PL SK HR IE GR BG RO HU LT TR PT DE NL CZ SI MT BE LU CY FR * Statistics are not available for Liechtenstein and the Former Yugoslav Republic of Macedonia. IS ES AT SE FI NO UK DK 14

17 Source: EUROSTAT, Labour Force Survey (data for variables EDUCSTAT and COURATT calculated by the authors) As one might expect, adult participation in non-formal education and training was always higher than adult participation in formal education and training. However, that was not the case in several countries (i.e. Bulgaria, Greece, Croatia, Hungary, Lithuania, Latvia, Poland, Portugal, Romania and Turkey) all of which had lower levels of non-formal compared to levels of formal adult learning in Conversely, most of the countries that had very high adult participation in non-formal education and training belonged to the first group of countries in terms of overall level of adult learning. This might therefore indicate a close link between a country s performance in adult participation in non-formal education and training, and the overall level of adult learning in this particular country. Table 2 below demonstrates that there are significant differences within countries in terms of adult participation in formal and nonformal education and training. Most prominent is again the lowskilled and highly-skilled distinction: low-skilled (ISCO 4-9) participate in formal education and training 3 times and in nonformal education and training 2.8 times less than highly-skilled (ISCO 1-3). A similar tendency is valid for the low-qualified and the highly-qualified. Furthermore, there is a tendency that adults in general, full-time employees, and older workers in particular more often participate in non-formal education and training (many of them on-the-job), while formal learning is more often undertaken by part-time employees or unemployed. Table 2: formal and non-formal education and training received during the previous four weeks (in %), in 2007, overall average results for the EU-27 plus EFTA-EEA and candidate countries*, by age group, full-time/part-time distinction, occupation group, labour status and qualification level Type of education and training Formal education and training** Non-formal education and training*** Fulltime Parttime ISCO 1-3 ISCO 4-9 Employed Unemployed ISCED 1-4 ISCED * Statistics are not available for Liechtenstein and the Former Yugoslav Republic of Macedonia. **55-64 olds distinction is not available for Bulgaria, France, Greece, Croatia, Lithuania, Luxembourg, Romania, Slovenia, Slovakia and Turkey. For Bulgaria (part-time), Cyprus (part-time), Germany (55-64), Estonia (55-64 and unemployed), Hungary (55-64), Luxembourg (full/part-time, ISCO 1-3/4-9, un/employed), Latvia (part-time), Malta (55-64 and part-time), Norway (55-64) and Poland (55-64) 2007 data is not available, we use 2006 or (if not available) 2005 data instead. ***Part-time distinction is not available for Bulgaria, Estonia, Croatia and Romania. For Bulgaria (unemployed), Iceland (unemployed), Malta (part-time) and Romania (unemployed) 2007 data for some variables is not available, we use 2006 data instead. Source: EUROSTAT, Labour Force Survey (data for variables EDUCSTAT and COURATT calculated by the authors) As provided in Figure 4 below, most countries have made at least some progress in adult learning in Sweden and Denmark increased adult participation rates by 14.9 and 10.8 percentage points respectively; meanwhile, the rate for Slovakia decreased by 4.6 percentage points. Variation in other countries ranged between -1 and +7.5 percentage points. 15

18 Figure 4: Absolute change in adult participation in education and training (both formal and non-formal), in across the EU-27 plus EFTA-EEA and candidate countries*, in percentage points SK TR NL UK LV BG CZ RO HR BE PL GR HU PT MT EE IT LU LT IE DE IS NO AT FR CY ES FI * Statistics are not available for Liechtenstein and the Former Yugoslav Republic of Macedonia. For Czech Republic, Croatia, Ireland, Latvia and Slovakia we use data as data for 2001 is not available. Source: EUROSTAT, Labour Force Survey (indicator EM051 - Lifelong learning (adult participation in education and training) - Percentage of the population aged participating in education and training over the four weeks prior to the survey) SI DK SE 14.9 Three groups of countries which were singled out according to the adult learning situation in 2007 are very different in terms of progress in the area. The average increase in the adult learning rate of the first group of countries was equal to 5.7 percentage points. The average progress of the second group was more than two times slower and increased by 2.2 percentage points. Meanwhile the participation of adults in education and training of the third group of countries in 2007 was on average 0.1 percentage points lower than in This reveals a sharp increase of adult learning disparities among the EU-27, EFTA-EEA and EU candidate countries between 2001 and However more recent dynamics reveal a slightly different tendency. Table 3 shows the progress made by different groups of countries (which were grouped according to figure 1) in adult learning in by age, full-time/part-time distinction, occupation, labour status and qualification level. Adult participation in learning in the leading group of countries declined in faster than in the other two groups. The second group of countries showed the best results, while the third group ranked second best during this period. Table 3 also reveals that in all groups of countries the decrease in participation in adult learning of highly-skilled (ISCO 1-3) or highly-qualified (ISCED 5-6) was greater than that of the disadvantaged groups (older workers, low-skilled (ISCO 4-9), lowqualified (ISCED 1-4). Table 3: absolute change in adult participation in education and training (both formal and non-formal, in percentage points), in * across the EU-27 plus EFTA-EEA and candidate countries**, by age group, full-time/part-time distinction, occupation group, labour status and qualification level Groups of countries (figure 1) Fulltime Parttime ISCO 1-3 ISCO 4-9 Employed Unemployed ISCED 1-4 ISCED st group nd group rd group Overall

19 Groups of countries (figure 1) Fulltime Parttime ISCO 1-3 ISCO 4-9 Employed Unemployed ISCED 1-4 ISCED 5-6 average * period was taken due to availability of statistical data. ** Statistics are not available for Liechtenstein and the Former Yugoslav Republic of Macedonia. For Turkey 2005 data (for all indicators), for Bulgaria 2005 data (55-64) and 2007 data (part-time) and for Croatia 2005 data (55-64 and part-time) is not available, we use 2006 instead. For Ireland 2005 and 2006 data (full-time and part-time) is not available, we use 2004 instead. Source: EUROSTAT, Labour Force Survey (data for variables EDUCSTAT and COURATT calculated by the authors) Countries also differ in terms of progress in adult participation by the type of education and training. Figure 5 shows that the highest differences between progress in adult participation in formal and non-formal education and training were in Denmark (13.3 percentage points progress in non-formal and 0.1 fall in formal education), United Kingdom (-13.8 and -1.3 percentage points respectively) and Sweden (-13.8 and -2.8 percentage points respectively). Croatia, Hungary, Estonia, France, and Slovenia, although characterised by slow progress, made similar headway in adult learning for both types of education and training. Differences between progress in adult participation in formal and non-formal education and training for other countries ranged from 0.4 percentage points in Romania to 6.9 percentage points in Iceland. 17

20 Figure 5: Absolute change in adult participation in education and training, by the type of education and training, in across the EU-27 plus EFTA-EEA and candidate countries*, in percentage points (presented from the highest to the lowest difference) DK UK SE IS ES FI SK NL GR NO BE MT IE LT TR DE IT CZ PT LU LV AT CY BG PL RO SI HU FR EE HR -13.8% -13.8% -2.8% -3.1% -1.8% -1.3% -1.4% -1.8% -1.4% -1.6% -0.1% 0.2% 1.3% 3.9% 0.7% 0.5% 0.7% 0.4% 1.2% 1.1% 2.8% -1.8% -0.1% -0.3% 0.2% 1.0% 1.5% -0.7% -0.1% 0.9% 0.3% 0.0% 0.9% -0.4% 0.4% 0.2% 1.0% 0.7% 0.3% 1.4% 0.9% 0.8% 1.3% 0.0% 0.4% -0.2% 0.2% 0.0% 0.4% -0.3% 0.0% 0.2% -0.1% 0.3% 0.1% 0.5% 1.4% 1.5% 0.5% 3.8% 6.1% 13.3% Non-formal education and training Formal education and training -15.0% -10.0% -5.0% 0.0% 5.0% 10.0% 15.0% * Statistics are not available for Liechtenstein and the Former Yugoslav Republic of Macedonia. For Austria, Belgium, Hungary, Italy, Lithuania, Malta, Poland, Portugal and United Kingdom we use data as data for 2003 is not available. For Croatia we use the only available data. Source: EUROSTAT, Labour Force Survey (data for variables EDUCSTAT and COURATT calculated by the authors) In adult participation in different types of education and training by age, full-time/part-time distinction, occupation, labour status and qualification level decreased in all cases except the participation of full-time workers in formal education and training. Table 4 presents average scores for all countries and shows that the highest decrease in adult participation was among part-time workers (both in formal and non-formal learning by 0.3 and 0.6 percentage points respectively), highly-qualified (in non-formal learning by almost 0.5 percentage point) and highly-skilled (in non-formal learning by almost 0.3 percentage point). Other changes were less significant. 18

21 Table 4: absolute change in adult participation in formal and non-formal education and training (in percentage points), in * across the EU-27 plus EFTA-EEA and candidate countries**, by age group, full-time/part-time distinction, occupation group, labour status and qualification level Type of education 55- Fulltimtime loyed loyed Part- ISCO ISCO Emp- Unemp- ISCED ISCED and training 64 Formal education and N/A training Non-formal - education and training³ * period was taken due to availability of statistical data. ** Statistics are not available for Liechtenstein and the Former Yugoslav Republic of Macedonia. Data for this variable is unavailable in most countries, therefore not presented in the table data for TR (all variables), IE (full-time), HR, IE and MT (part-time), MT (ISCO 4-9), EE, LU, LV and MT (unemployed) is not available, we use data available for most recent yea or, if not available, for 2006 or data for Luxembourg (full-time), BG, CY, LU, LV and MT (part-time), LU (ISCO 1-3), LU (ISCO 4-9), EE and LU (unemployed) is not available, we use data available for most recent year. ³ 2005 data for TR (all variables), BG and HR (55-64), IE (full-time), IE and LV (part-time), BG and HR (ISCO 4-9), BG, EE, HR and IS (unemployed) is not available, we use data available for most recent yea or, if not available, for 2006 or data for MT (part-time), BG, IS and RO (unemployed) is not available, we use data available for most recent year. Data for part-time distinction is not available for BG, EE, HR and RO, we exclude these countries from the analysis of this variable. Source: EUROSTAT, Labour Force Survey (data for variables EDUCSTAT and COURATT calculated by the authors) Another dimension of adult participation in education and training is the intensity of education and training, i.e. the mean number of hours spent in education and training. Figure 6 reveals that the intensity of non-formal education and training varies moderately across countries. It is however considerably higher in some countries (Portugal, Greece, Spain, Turkey and France). Interestingly, countries, which have lower levels of participation in adult learning are characterised by a higher intensity of learning (i.e. mean number of hours spent in non-formal education and training) Figure 6: Mean number of hours adults spent on all taught learning activities (non-formal education and training) within the last four weeks, in 2007 across the EU-27 plus EFTA-EEA and candidate countries* SE IS EE UK FI IE 9.1 CY CZ LU NO RO DE LT SI LV BE DK * Statistics are not available for Liechtenstein and the Former Yugoslav Republic of Macedonia. The mean number of hours is derived by multiplying the average number of hours (if the interval is 1-10, it is multiplied by 5, if by 15 and so on ) by number of respondents who participated in non-formal education and training and indicated this interval of hours spent on it. Source: EUROSTAT, Labour Force Survey (data for variable COURLEN calculated by the authors) MT PL BG 13.5 IT SK HR NL AT HU PT GR 24.7 ES TR FR 19

22 Table 5 below shows that older and part-time workers and the lowskilled learned non-formally on average less intensively than others: the mean number hours for older workers was 10.3 compared to 14 of olds, that of part-time workers 10.1 compared to 12.3 of full-time workers; that of low-skilled 10.8 compared to 12.3 of highly-skilled. Unemployed adults constitute a prominent exception the mean numbers of hours was 28.4 compared to 12.1 of those employed. The Low-qualified also participated in non-formal education and training slightly more intensively than highly-qualified: 14.3 and 13.4 respectively. Table 5: mean number of hours adults spent on all taught learning activities (non-formal education and training) within the last four weeks, in 2007 across the EU-27 plus EFTA-EEA and candidate countries*, by age group, fulltime/part-time distinction, occupation group, labour status and qualification level Fulltime Parttime ISCO 1-3 ISCO 4-9 Employed Unemployed ISCED 1-4 ISCED 5-6 Mean no. of hours * Statistics are not available for Liechtenstein and former Yugoslav Republic of Macedonia. For Source: EUROSTAT, Labour Force Survey (data for variable COURLEN calculated by the authors) Figure 7 below suggests that the largest increase in the intensity of non-formal education and training in was made by Bulgaria (+8.1 hours per adult participating in non-formal education and training), Latvia (+7 hours), Slovakia (+6.3 hours), Greece (+5.7 hours) and Croatia (+5.2 hours). Meanwhile Hungary experienced the largest decline in the intensity of adult education and training by hours per adult participating in non-formal education and training. Figure 7: Absolute change in mean number of hours adults spent on all taught learning activities (non-formal education and training) within the last four weeks, in across the EU-27 plus EFTA-EEA and candidate countries* HU DESE RO NLES FR FI LUNO UK SI PTIE CY TR ITCZ IS DK BEEE PL MT LTAT HR GR SKLV * Statistics are not available for Liechtenstein and the Former Yugoslav Republic of Macedonia. For Spain, Croatia and Malta we use data as data for 2004 is not available. For France and Turkey we use the only available data. Source: EUROSTAT, Labour Force Survey (data for variable COURLEN calculated by the authors) BG Table 6 summarises the available information on the intensity of both types of education and training. Similarly to Figure 6, it also reveals the relationship between intensity of non-formal education and training and overall country performance in adult learning on 20

23 average, the lower is performance of a country, the higher is intensity. Table 6: mean instruction hours spent by participant aged on education and training, by the type of education and training, in 2007 across some EU-27 plus EFTA-EEA and candidate countries (presented from the lowest to the highest number of mean instruction hours for both types of education and training) Formal and non-formal education and training Formal education and training Non-formal education and training FR UK BG CZ CY SK SI EE IT GR LT AT HR NO DE SE FI LV BE ES PL PT HU NL N/A N/A 59 Source: EUROSTAT, Adult Education Survey Adult education attainment Countries are less divergent in terms of adult education attainment. This argument holds true especially in the case of upper secondary education attainment. As shown in Figure 8, 20 out of 31 countries for which data is available had 70 or more percent of adults having completed at least upper secondary education. However there are still some countries that need to reach a considerable progress in adult upper secondary education attainment (i.e. Turkey with only 25.8% of adults having at least upper secondary education, Malta with 26.7% and Portugal with 27.5%). Slow progress in education attainment leads to low country performance in overall adult participation in education and training. To illustrate, recent research showed that adults aged are three times more likely to participate in lifelong learning if they have completed at least upper secondary education (European Commission, 2007b). 21

24 Figure 8: Percentage of adults having completed at least upper secondary education (ISCED level 3), in 2007 across the EU-27 plus EFTA-EEA and candidate countries* TR MT PT ES IT GR IS LU IE BE FR CY NL AT UK RO HR DK BG NO HU * Statistics are not provided for Liechtenstein and the Former Yugoslav Republic of Macedonia. Source: EUROSTAT, Labour Force Survey (indicator SO011 - Percentage of the population aged 25 to 64 having completed at least upper secondary education) Figure 9 illustrates that only Portugal and Malta were making considerable progress in adult upper secondary education attainment (increase by 7.3 and 7.4 percentage points respectively in ). Turkey still fails to improve its adult upper secondary education attainment (minor increase of 0.5 percentage point) and still lags markedly behind. Countries that have an adult upper secondary education attainment level lower than 70% made much faster progress (average increase by 7.2 percentage points) than those with a higher level of upper secondary education attainment (average increase by 4.2 percentage points). Interestingly, adult upper secondary education attainment in Norway and Denmark decreased by 7 and 5.2 percentage points respectively in Figure 9: Absolute change of the percentage of adults having completed at least upper secondary education (ISCED level 3), in across the EU-27 plus EFTA- EEA and candidate countries* FI SI DE SE LV PL LT EE SK CZ NO DK TR DE EE AT SK SE LT CZ RO LV HR FR SI PL NL BG PT LU FI MT IS GR IE BE UK CY HU * Statistics are not available for Liechtenstein and the Former Yugoslav Republic of Macedonia. For Croatia we use data as data for 2001 is not available. For Turkey we use the only available data. Source: EUROSTAT, Labour Force Survey (indicator SO011 - Percentage of the population aged 25 to 64 having completed at least upper secondary education) IT ES 22

25 40.0% 35.0% 30.0% 25.0% 20.0% 15.0% 10.0% 5.0% 0.0% TR 9.6% 12.2% 12.3% RO MT 13.5% 13.6% IT PT CZ SK As Figure 8 has shown, most countries have relatively large numbers of adults with at least upper secondary education. This is not the case with tertiary education attainment. Figure 10 indicates that countries are considerably less homogenous according to adult tertiary education attainment. 13 countries have markedly higher numbers of adults who have successfully completed tertiary education (starting with Lithuania with 29% to Finland with 36%). Turkey, Romania, Malta, Italy, Portugal, Czech Republic and Slovakia have the lowest share of adults who have attained tertiary education. Other countries - falling between Croatia (16%) and Luxembourg (26.5%) - show moderate results. Figure 10: Percentage of adults who have successfully completed tertiary education (ISCED levels 5-6), in 2007 across the EU-27 plus EFTA-EEA and candidate countries* 13.7% 14.3% 16.0% 17.6% AT HR 18.0% 18.7% HU PL 22.0% 22.0% GR SI LV 22.3% 22.3% BG 24.7% 26.4% 26.5% 29.0% 29.5% 30.5% 30.6% 30.9% DE FR LU LT ES UK NL IS IE 31.2% 31.3% 31.4% 32.1% 33.2% 33.5% * Statistics are not available for Liechtenstein and the Former Yugoslav Republic of Macedonia. Source: EUROSTAT, Labour Force Survey (data for variable HATLEVEL calculated by the authors) DK SE 34.6% 36.0% BE CY EE NO FI Unlike the tendency in upper secondary education attainment, countries which underperformed in adult tertiary education attainment in 2007 have made less progress than those with better results (see Figure 11). The third group of countries (Turkey, Romania, Malta, Italy, Portugal, Czech Republic and Slovakia) have added on average only 2.9 percentage points to their tertiary education attainment levels in , while the second and first group of countries advanced, on average, by 4.4 and 4.9 percentage points respectively. Slovenia and Luxembourg were the countries which made the best progress (8.8 and 9 percentage points respectively). However greater progress is expected in adult tertiary education attainment. In a strategic framework for European cooperation in education and training (Council 2009) Council concluded that by 2020, the share of year olds with tertiary educational attainment should be at least 40%. The tendency observed above of greater participation among those who attain ISCO 1-3/ISCED 5-6 levels would, therefore, imply that demand for adult learning would grow significantly in parallel to this increase in tertiary qualifications by

26 10.0% 9.0% Figure 11: Absolute change of the percentage of adults who have successfully completed tertiary education (ISCED levels 5-6), in across the EU-27 plus EFTA-EEA and candidate countries* 8.8% 9.0% 8.0% 7.0% 6.0% 5.0% 4.0% 3.0% 2.0% 1.0% 0.0% TR 0.4% 1.1% 1.2% 1.6% 2.1% CZ HR BG NO 2.4% 2.5% RO 2.6% 3.1% 3.2% 3.3% 3.5% 3.6% 3.6% 3.6% DE LV DK AT EE MT IT SK FI FR BE 3.8% 4.1% 4.4% 4.6% HU 4.7% SE UK PT 4.7% 4.8% GR IS 5.7% 5.9% ES 6.4% CY 6.8% 6.8% NL LT IE 6.8% 7.1% * Statistics are not available for Liechtenstein and the former Yugoslav Republic of Macedonia. For Czech Republic, Croatia, Ireland and Latvia we use data as data for 2001 is not available. For Turkey we use the only available data. Source: EUROSTAT, Labour Force Survey (data for variable HATLEVEL calculated by the authors) PL SI LU To summarise the above analysis, the greatest challenge for EU policy in the area of adult learning is the imminent failure to reach the Lisbon target of lifelong learning. While some countries show exemplary performance in several aspects of adult leaning, there are wide and persisting disparities: In adult learning participation between member states, between low-qualified and highly-qualified, low-skilled and highly-skilled, younger and older adults; In adult secondary and tertiary education attainment between member states. In the next chapter we review the recent political initiatives of the EU to address the above challenges EU policy progress in the area of adult learning and instruments for stimulating national education and training reforms EU policy in the area of adult learning has developed intensively over the last several years and its potential in influencing national policies has grown. In this chapter we review the main policy initiatives and the two main instruments for their implementation: 1. The Open method of coordination (OMC); 2. The funding instruments The development of EU policy initiatives As a reaction to the challenges analysed in Chapter 1.2, EU institutions have grown increasingly involved in promoting reforms of adult education and training. The issues and problems related to this topic have attracted a great deal of discussions and the progress at the political level has also become evident. Since the European Year of Lifelong Learning (1996) (European Parliament, Council, 1995, pp ), when the Council first adopted conclusions on the subject, the idea of lifelong learning (LLL) in 24

27 general and adult learning in particular has grown considerably in importance both at the Community and at the national level. Since the adoption of the Lisbon Strategy it has become central to policy debate throughout the European Union and a cornerstone of the reform of the Structural Funds and the Community s employment policy. However, many recent EU policy initiatives in the area of adult learning are still to be translated into action on the national and sub-national levels. Prior to the Lisbon Strategy, adopted in 2000 (European Council, 2000), the adult learning related reforms were hardly mentioned in the Community policy documents. Education (in general) was included as an area of EU competency in the Maastricht Treaty establishing the European Community in 1992 (Treaty on European Union, 1992). The need to put more emphasis on investment in labour market training was already expressed in 1994 Essen European Council for Employment (European Council, 1994). The Heads of State and Government affirmed in 1997 in the preamble to the Amsterdam Treaty (Treaty of Amsterdam establishing the European Community, 1997) their determination "to promote the development of the highest possible level of knowledge for their peoples through a wide access to education and through its continuous updating". In the same year, the European Employment Strategy (European Council, 1997), emphasising the need to create the new culture of adaptability, i.e. a constant renewal of skills within enterprises, was launched. However, none of these processes actually meant sufficient inclusion of adult learning issues in the EU policy agenda. Most of them were too general, directed only at wider or related policy areas and did not call for action in the particular area of adult learning. Therefore, the EU policy developments could not affect the adult learning policies of the Member States (within the limits of the EU supporting actions for member states according to the principle of subsidiarity). Lifelong learning, including adult learning, has been recognised as a key element of the Lisbon strategy adopted in 2000 by the European Council (European Council, 2000). After the Commission s Memorandum on Lifelong Learning in 2000 (European Commission, 2000), followed by extensive consultation with Member States, candidate, EFTA-EEA countries and civil society, a new paradigm of learning emerged placing the learner at the centre of the learning process and emphasising the importance of equal opportunities and the quality and relevance of learning opportunities. The Commission s 2001 Communication Making a European Area of Lifelong Learning a Reality (European Commission, 2001) jointly prepared by the DG for Education and Culture and the DG for Employment argued that traditional systems must be transformed to become much more open and flexible, so that the learners can have individual learning pathways, suitable to their needs and interests, and thus genuinely take advantage of opportunities throughout their lives. The Communication identified the features of a coherent and comprehensive lifelong learning strategy to be the basis for developing and implementing national lifelong learning strategies. The subsequent Council Resolution on lifelong learning in 2002 (Council 2002) invited Member States, within the framework of their responsibilities, to pursue a number of policies and actions, and committed them to developing national lifelong learning strategies. 25

28 In 2001, a common Education and Training 2010 work programme was launched. It covers all learning contexts (formal, non-formal and informal) and levels (pre-primary, primary, secondary, higher, adult education and continuing training) of education and training within a lifelong learning perspective. The programme marks the start of the OMC in the area of education and training (see the next section). However, the generality of this work programme and its suitability to all ages, as well as its lack of coherent, decisive and detailed addressing of adult learning issues meant that most Member States chose to apply it primarily to initial learning (which was easier to influence) and not to adult learning. In 2005, following the revision of the Lisbon Strategy, the Council adopted new Employment Policy Guidelines, which included Guidelines 23 (Expand and improve investment in human capital), and 24 (Adapt education and training systems in response to new competence requirements) (Council, 2005, pp ). Guideline 23 invited Member States to promote lifelong learning through appropriate incentives and cost-sharing mechanisms, and emphasised that low-skilled and older workers should be subject to particular attention. Guideline 24 addressed the education and training quality issues, and also invited the improvement of the definition and transparency of qualifications, their effective recognition and validation of non-formal and informal learning. These Guidelines were also confirmed by the Council for the period of (Council, 2008a, pp ). The revised Lisbon Strategy has also set out the process of developing comprehensive national reform programmes and monitoring their implementation at EU level. This meant that each member state would also receive feedback on its progress in education and training reforms, including those affecting adult learning. In 2006, a Recommendation of the European Parliament and of the Council on key competences for lifelong learning (European Parliament, Council, 2006a, pp ) was adopted. It identified and defined the competences that are necessary for personal fulfilment, active citizenship, social cohesion and employability in a knowledge society. However, most of the practical work related to implementation of this recommendation was focused on children and young people rather than adults (e.g. via promotion of key competences in schools). In adult learning there has been a growing emphasis on employability and skills for the labour market, perhaps due in part to an impetus from the Employment policy guidelines (Council 2005). Despite the political emphasis placed on lifelong and adult learning, the dichotomy between political discourse and reality remained striking. Adult participation in lifelong learning varied widely across the Union, but was unsatisfactory in many Member States. One reason for limited progress in adult learning participation was that new policies have not yet been implemented. No doubt considerable periods of time were necessary for any EU-level initiatives to be translated into action at the member state level and yield results. Another reason could be that most of the positive developments in education and training policy at member state level were targeted at the younger population and that adult learning needed special attention. As a result the European Commission adopted two Communications on Adult Learning: It is never too late to learn in 2006 (European Commission, 2006a) and It is always a good time to learn in 2007 (European Commission, 2007a). 26

29 The 2006 Communication reflected the approach outlined in the Communication on efficiency and equity (European Commission, 2006b) that reforms must <...> be stepped up to ensure high quality education systems which are both efficient and equitable (European Council, 2006). Five key messages for adult learning stakeholders were established in this Communication: 1. To remove barriers to participation; 2. To increase the quality and efficiency of the adult learning sector; 3. To speed up the process of validation and recognition of non-formal and informal learning outcomes; 4. To ensure sufficient investment in adult learning (esp. for older people and migrants); 5. To better monitor the adult learning sector. The 2007 Communication, containing the Action Plan on Adult learning, was aimed at implementing the five key messages established in the 2006 Communication and invited Member States and other stakeholders to develop efficient adult learning systems through participation in a European Action Plan for the sector consisting of actions in the following areas: Analysing the effects of reforms in all sectors of education and training in Member States on adult learning; Improving the quality of provision in the adult learning sector; Increasing the possibilities for adults to achieve a qualification at least one level higher than before; Speeding up the process of validation and recognition of learning outcomes achieved in non-formal and informal learning; Improving the monitoring of adult learning sector. The actions outlined in the 2006 and 2007 Communications gave rise to a number of subsequent conclusions, resolutions and opinions elaborated by other bodies at European and national level. The 22 May 2008 Council Conclusions (Council, 2008b) indicated a list of specific measures to be taken in order to implement the 2007 Action Plan on Adult learning. Among the tasks envisaged for the Commission were the following: To analyse the impact of national education and training reforms in terms of the distribution of funding resources across the various age groups, in line with a lifelong learning approach; To analyse reforms in education and training at national level, especially the development of national qualifications systems in relation to the European Qualifications Framework and credit transfer systems relating to both formal, non-formal and informal learning, with a view to improving adult access to qualifications systems. The brief analysis of the recent developments at the EU and national level has shown that lifelong learning ideas, including adult learning, have been broadly accepted by all stakeholders in all of the countries. Most countries have elaborated their lifelong learning policies; however, their effective implementation still remains the greatest challenge in many countries. Complex, fragmented, incomplete and/or isolated adult learning systems lacking capacity to ensure quality services are often to blame for low participation rates of adults in lifelong learning. Many poorer countries struggle to finance even the initial education needs of their population and 27

30 resources for the lifelong learning needs of adults are severely limited. Often their most important source of financing adult learning policy is EU funding instruments. Furthermore, underdeveloped monitoring and evaluation arrangements in the adult learning sector limit its possibilities for policy learning, lead to failure in demonstrating its benefits and result in weak visibility. Thus, the problems characteristic to adult learning often receive low attention and priority in national reforms of education and training. As education and training lies within the responsibility of individual countries, the translation of EU policy initiatives into actions can be done only by the member state governments, which determine the scope, intensity and timing of adult learning reforms. The EU institutions can only stimulate this process through the OMC and the EU funding instruments. The two mechanisms for EU influence on national reforms are briefly reviewed in subsequent sections The role of the Open Method of Co-ordination (OMC) The first instrument for the EU institutions to influence national reforms in the area of education and training, employment or social policy is the Open Method of Co-ordination. The previous section already introduced major EU actions taken in the framework of OMC. The current section provides a discussion of OMC as policy co-ordination instrument and its ability to affect national policies or their reforms. Since the Lisbon Summit of March 2000 where the OMC was defined as a governance instrument of the Lisbon Strategy, it has become the so-called third way between intergovernmental negotiations and the classical Community Method. The OMC was created in the search for a governance mechanism which would still be using soft law measures, but also ensures a greater influence of EU policy developments on national reforms. Indeed, in politically sensitive areas where EU Member States strive to retain their responsibility, such as education and training, the OMC is the only possible mechanism of EU influence since it is voluntary, respects national diversity and does not foresee the transfer of competencies to the European level. The OMC is a tool for spreading best practice and achieving greater convergence towards the main EU goals. This method, which is designed to help Member States to progressively develop their own policies, involves the following activities (European Council, 2000): Fixing guidelines for the Union combined with specific timetables for achieving the goals which they set in the short, medium and long terms; Establishing, where appropriate, quantitative and qualitative indicators and benchmarks against the best in the world and tailored to the needs of different Member States and sectors as a means of comparing best practice; Translating these European guidelines into national and regional policies by setting specific targets and adopting measures, taking into account national and regional differences; Periodic monitoring (incl. biennial joint interim reports), evaluation and peer review organised as mutual learning processes. 28

31 The OMC in education and training was launched in March 2002 Barcelona European Council with the start of the Education and Training 2010 work programme which established for the first time a solid framework for European cooperation in the field of education and training, based on mutual learning and knowledge diffusion. Recent Council conclusions (Council, 2009, pp. 2-10) introduced the following four new strategic objectives to be addressed by the framework Education and training 2020 : (1) making lifelong learning and mobility a reality; (2) improving the quality and efficiency of education and training; (3) promoting equity, social cohesion and active citizenship; and (4) enhancing creativity and innovation, including entrepreneurship, at all levels of education and training. In this document the Council calls for the effective use of the OMC, which should draw on: the four abovelisted strategic objectives, common reference tools and approaches, peer learning and the exchange of good practice (incl. the dissemination of outcomes), periodic monitoring and reporting, evidence and data from all relevant European agencies, European networks, and international organisations, and making full use of the opportunities available under Community programmes, particularly in the field of lifelong learning. However, implementation of the OMC as well as assessment of its results is extremely challenging for a number of interrelated reasons (Zeitlin 2009): The variety of distinct processes with different institutional characteristics subsumed under the OMC rubric; Their relative novelty, and the frequency of procedural changes to even the most highly institutionalized processes; and The horizontal and vertical complexity of the OMC processes, which typically cut across sectoral policy domains and involve multiple levels of governance (European, national, sub-national) in 27 Member States. To illustrate, indeed, EU Member States and the Commission have been cooperating closely to support national reforms of education and training systems through numerous distinct processes and programmes including the "Education & Training 2010", the European Employment Strategy, National Action Plans for employment and social inclusion and more recently the Lisbon National Reform Programmes, the new Programme for Employment and Social Solidarity, Lifelong Learning Programme and its subprogrammes etc. Furthermore, the new strategic framework for European cooperation in education and training will be changed to include an even wider array of instruments: it will more actively integrate the Lifelong Learning Programme to support the development, testing, transfer and implementation of new approaches and innovation; agencies, networks, expert groups and other international organisations to support research and analysis; reinforce and better link the OMC with the policy dialogue with several external partner countries (European Commission 2008c). Last but not least, the lifelong learning perspective calls for coordination and priority setting across different education and training sectors. Specific policy agendas have been set out, priorities set, and progress of work monitored for schools, VET, higher and adult education. The success of the OMC depends on the ability to assess the progress of Member States towards the agreed policy targets and to generate and share evidence of which policies and their reforms 29

32 help to achieve these targets. The EU institutions have sought to improve monitoring and promote assessment of impacts of reforms on adult learning. Council conclusions of 25 May 2007 established a framework of indicators and benchmarks for monitoring progress towards the Lisbon objectives in education and training (Council 2007). Council conclusions of 22 May 2008 on adult learning (Council, 2008b) stated that progress and monitoring of adult learning should be consistent with the coherent framework of indicators and benchmarks adopted by the Council in May 2007 and be included in future joint progress reports on the Education and Training work programme. The Conclusions adopted specific measures for the period , which urged the Commission and the Member States to analyse the impact of national education and training reforms and to promote and support the exchange of good practice, and mutual learning. The Action Plan on Adult Learning of September 2007 (European Commission 2007a) approved 5 areas of action, which included improved monitoring of adult learning sector and analysis of the effects of reforms in all sectors of education and training in Member States on adult learning. But these EU-level efforts will not produce sufficiently detailed and reliable evidence on which national reform measures work best in stimulating adult learning, what problems they are able to solve and under what circumstances, unless Member States make greater efforts in assessing impacts of their policies and policy reforms and sharing openly their collected evidence with other Member States. In Chapter 3 we argue that impact assessments of education and training and related reforms on adult learning are scarce or inaccessible, which is an obstacle for country-to-country policy learning in the EU. The impact of the OMC on national policy making and policy convergence is challenging to assess, especially in quantitative terms. To cite the Commission, it is, in particular, difficult to show robust evidence of causal linkage between the OMC and national policy outcomes in terms of quantifiable results (European Commission, 2008c) (eg. in relation to raising participation levels as measured by progress in relation to the benchmark of adult participation in lifelong learning). In qualitative terms, the OMC in the education and training area seems to have limited impact on national policies. In the 2008 Communication on an updated strategic framework for European cooperation in education and training (European Commission, 2008d) the European Commission underlined that the progress towards meeting the five benchmarks (incl. adult participation in lifelong learning) set for 2010 has been insufficient. Academics (Humburg, 2008, Lange and Alexiadou, 2007, Gornitzka, 2006a, 2006b, Laffan and Shaw, 2005) argue that although the OMC in education and training has increased the level of institutionalisation and improved the structure of the process by establishing clear objectives and clear deadlines for the objectives to be achieved, it has a number of shortcomings which limit its impact. Firstly, the Community competence in education and training is weak since Maastricht, its competencies remain restricted to quality assurance and mobility issues. Secondly, the Commission s decisional autonomy and capacity is low compared to other areas such as employment policy: the weak treaty basis, no right for the Commission to make recommendations, the European Court of Justice as well as the European Parliament is excluded from the process. Last but not least the OMC exerts limited impact because of its voluntary nature: it does not oblige Member States 30

33 to produce national action programmes, does not allow for peer reviews naming and shaming particular countries and does not force the inclusion of social partners and other stakeholders into the process. However the above assessment of the impact of the OMC is very general and could be very different for individual countries. The impact of the OMC processes on national policies significantly depends on domestic institutional and political conditions (Zeitlin 2009). Equally important for the impact of the OMC, however, is how creatively it is appropriated or adopted by the domestic actors. This is to be assessed for each country concerned. Therefore no clear-cut conclusion on the impact of the OMC on national settings and policies could be made. The OMC process cannot force change upon national institutions, but it provides a channel of peer expertise for those willing to reform and seeking evidence about similar reforms in other Member States. Many reform agendas do not become reforms because of the presence of too many factors which cause uncertainty. The reforms are seldom supported by hard evidence on impacts of similar reforms elsewhere. The OMC could become more useful if exchanges between Member States were more detailed and deep, more open and based on good quality evidence The role of the EU funding instruments The role of the EU structural funds The implementation of the actions outlined in the 2006 Communication and subsequent 2007 Action Plan on Adult learning, as well as in Employment Guidelines 23 and 24, is supported through the use of the European Social Fund (ESF). The EU Structural Funds, in particular the ESF (as one of its priority areas is lifelong learning), has considerable potential to influence national policies in education and training. The Structural Funds (and in particular the ESF) have contributed in many member states to reshaping national policies in education and training. Particularly important was the introduction of support to structures and systems in the period of ESF. Evaluation of ESF funding for the development and modernisation of education and training structures and systems (IDEC SA, 2006) stated that ESF has produced great impact on the administrative machinery, inserting practices of consultation, partnership and networking. However in the area of education and training the political will was considered the most important prerequisite for the successful implementation of the interventions. Community initiative EQUAL has also contributed to development of lifelong learning policies and structures. The interim evaluation of the Community initiative EQUAL (Bernard Brunhes International, ICAS Institute and Economix Research and Consulting, 2006) revealed two main initial and potential impacts for Employment Guidelines 23, 24 and for the programmes of DG Education and Culture: 1) Dissemination of new methodologies for encouraging low skilled workers participation in training; and 2) Use of ICT to increase access to learning for people considered as disengaged learners or facing physical barriers to traditional learning. The actual impacts of Equal are still to be assessed in the final evaluation, which is currently under way. 31

34 In the period ESF substantially increased its budget and broadened its scope to reflect a life-cycle approach to education and training. Figure 12 shows that the bulk of ESF funding in this period is given to three priorities related to adult learning: promoting education and training throughout working life (16.3% of total ESF funding), development of LLL systems (12.4%) and reforming education and training systems (11%). Figure 12 also reveals that convergence countries and regions 4 tend to allocate twice as much resource to reforming their education and training systems than regional competitiveness and employment countries and regions. Out of 470 projects under this priority 362 (77%) were implemented in Finland, 58 (12.3%) in Czech Republic, 17 (3.6%) in Slovakia, 10 (2.1%) both in Denmark and Estonia. France implemented 6, Spain 4 and Germany, Luxembourg and the Netherlands only 1 project. Other countries did not implement any project under this priority. Figure 12: ESF funding per priority (related to adult learning) and objective in , % of total 25.0% 20.0% 15.0% 10.0% 5.0% 0.0% 10.2% 12.8% 15.6% 12.4% Development LLL systems 3.6% 3.0% 3.5% 3.3% Employment and training support for w orkers and companies 13.4% 11.1% 6.5% 11.0% Reforming education and training systems 15.8% 20.1% 10.4% 16.3% Promoting education and training throughout w orking life 7.0% 5.9% 3.3% 5.8% Developing human potential in research and development 1.3% 1.0% 1.5% 1.2% Partnerships, netw orks and initiatives Convergence Multi objectives Regional Competitiveness Total 4.9% 2.2% 0.1% 2.7% Improving institutional capacity 3.7% 2.9% 3.5% 3.3% Technical assistance Source: ESF website at Furthermore, EU Member States are obliged to carry out evaluations on the efficiency and effectiveness of the use of the EU financing instruments. Therefore the EU funding instruments is a vital mechanism through which the EU stimulates policy learning process within Member States. However, as in the case of the OMC, the actual value added of the EU funds to stimulating national reforms depends almost entirely on the political will and implementation capacity of the national actors. The examples of education and training reforms funded by the ESF and their impact assessments are reviewed in Chapter 3 of this report. Since all structural fund interventions in the area of education and training are monitored and evaluated on a mandatory basis and many of these interventions contribute to national reforms, the 4 The relative wealth of a country or region measured by GDP per capita determines whether it falls under the Convergence or Regional Competitiveness objective. There is a higher ESF contribution up to 80% of co-funding for Convergence regions. Convergence objective spending aims to improve job creation and employment opportunities, thus bringing the wealth and employment in a region closer to the EU average. Regional competitiveness and employment objective spending aims to give countries and regions the workforce and labour markets they need to build successful, competitive economies, able to compete globally. 32

35 lack of evidence about their success is striking. The monitoring and evaluation of structural funds should give due attention to assessing the impacts of education and training reforms they funded and especially the impacts of these reforms on adult learning. The role of the Lifelong Learning Programme and its subprogrammes The EU Lifelong Learning Programme (LLP) was established in 2007 and integrated a number of education, vocational training and e- learning programmes, which ended in 2006 together with the programming period. The LLP makes reference to the objectives set out in the Education and Training 2010 work programme. However, the implementation of the new LLP represents rationalisation of the objectives of the previous generation of education and training programmes, supplemented with new ones that reflect the policy developments in education and training outside these programmes. Key programmes, which were present during the period of and were integrated in the LLP in the new period, are: Erasmus ( ); Comenius ( ); Leonardo da Vinci ( ); Grundtvig ( ); LINGUA ( ); E-Learning ( ) and the Action programme to promote bodies active at European level in the field of education and training ( ). The sub-programmes of the LLP have contributed significantly to the enhancement of adult learning in the EEA. For example, the evaluation of the Leonardo II programme showed that this particular initiative enhanced employability, entrepreneurial spirit and job creation, and also provided value-added to the development of a Europe of knowledge (Ernst & Young, Barbier Frinault & Associés, 2003) and bringing about convergence between Member States in policy and practice in the field of VET (ECOTEC, 2008). The Erasmus programme is regarded by many experts to be very successful in promoting habits of mobility and enhancing employability of the future workforce, development of soft skills of exchange students, improvement of transparency and transferability of their qualifications (CHEPS, INCHER-Kassel, ECOTEC Research and Consulting Ltd, 2008). Grundtvig action is generally seen as relevant to the needs of national and European adult education policies and programmes, and sometimes even influential in shaping them, in cases when Grundtvig objectives are more innovative than national policies, especially in convergence countries and regions (ECOTEC Research and Consulting Limited, 2003). Grundtvig action also played a large role in popularising exchange of experience through learning partnerships among adult education institutions, as the number of participating institutions rose fourfold during the period of (ECOTEC, 2008). However, although in relative terms Grundtvig sub-programme s budget increased considerably, in absolute terms it is the smallest among the existing sub-programmes of the LLP. Furthermore, it is estimated that only approx. 10% of all the LLP funding is spent on adult training. The budgets of the programmes were increased significantly as they were integrated in the LLP for the new programming period of with the total earmarked budget of the LLP reaching billion. The decision of the European Parliament and of the Council establishing an action programme in the field of lifelong learning (European Parliament and of the Council 2006) foresees 33

36 that the sectoral sub-programmes will receive the following minimum shares of total support: Comenius (13%); Erasmus (40%), Leonardo da Vinci (25%) and Grundtvig (4%). The increase in funding of sectoral sub-programmes is reviewed in table 7. Table 7: Increase in funding of the LLP sectoral subprogrammes Support in Support in , , billions euro (Calculations by Subprogramme (in %) Increase billions euro authors based on European (ECOTEC, Parliament and of the 2008) Council 2006) Comenius % Erasmus % Leonardo da Vinci % Grundtvig % As is evident from the data in the table above, the most impressive increase in funding is foreseen for the Erasmus programme. This could enable even stronger effects on the internationalisation of the workforce, increases in its mobility and employability. However, the funds allocated to those sub-programmes which have immediate effects on adult learning (Leonardo da Vinci and Grundtvig) also improved substantially. To summarize, although the lifelong learning ideas have been broadly accepted by most stakeholders in most of the countries since the 1990s, it was only in 2006 that considerable political emphasis on adult learning at the European level started. Both the OMC and EU funding instruments (although with limited financial resources for adult learning) have helped to implement the earlier European political initiatives, however specific evidence on their impacts is still lacking. The impacts of EU policy on national reforms are difficult to analyse due to the large diversity of member states, their political systems, and their contexts. These determine the scope, intensity and timing of education and training as well as other national reforms and their impacts on adult learning. 34

37 2. CONCEPTS AND LOGIC OF IMPACT ASSESSMENT In this chapter we present the main concepts of impact assessment, its purpose and benefits and then look in more detail at particular limitations of assessing impacts of public policy and management reforms in education and training sector (and beyond) on adult learning. What is impact assessment? Impact assessment (or evaluation of impact 5 ) is the process, which helps to explore the relationship between the inputs and activities of a particular intervention and its wider impacts on individuals, organisations, policy or society as a whole. Policy impacts and their assessment are the objects of policy analysis applied research discipline, which uses the instruments and methods developed by a wide spectrum of social sciences. Some impact assessments are conducted before the start of intervention and are meant to improve the intervention or to ascertain that the causal theory behind the intervention is correct. Impact assessment before the intervention is mostly based on analysis of development theories, logical models and making comparisons of planned intervention to previous experience. Every piece of EU legislation has to overcome the scrutiny of impact assessment before it is finally approved. Largely similar requirements are also applied in a number of EU member states. The actual impacts of interventions can be measured only after they have been fully implemented. Thus impact assessment after the intervention is a crucial element of policy learning. It provides feedback to policy makers whether their initiative has achieved its objectives and underpins detailed recommendations on how elements of intervention could be improved in the future and (or) used in other policy contexts. Ex-post evaluation, which includes assessment of impact and Community value added, is mandatory for all EU financed or co-financed interventions both at EU and national (or regional) level. 6 Value added compared to previous guidance Detailed guidance on how to perform such evaluation is already provided by the European Commission services in the Guide for the Evaluation of Socio-Economic Development and accompanying sourcebooks and materials (Tavistock Institute 2003) or in the Practical Guide for the Commission services for Evaluating EU Activities (Directorate General for the Budget 2004). Current report is meant to supplement the existing wealth of guidance rather than duplicate previous work. In particular, the report explains the differences in measuring impacts of policy reforms rather than the impacts of policy as usual. It also goes beyond explaining how to assess the impacts of financial interventions or financing reforms. Reforms which might have strong impacts on adult learning also include regulatory changes, restructuring of public service provision and possibly even changes in income-support systems. Very often 5 We use impact assessment and evaluation of impact as synonyms. 6 For example such requirement exists in the EU structural funds regulations. 35

38 reforms are a complex set of measures combining different types of government interventions. Benefits of impact assessment The benefits of policy evaluation including impact assessment are well summarised in the Guide for the Evaluation of Socio-Economic Development (Tavistock Institute, 2003). Impact assessments - to mention just several important aspects - help to: Reduce uncertainty and improve planning and implementation of policies; Choose between alternative policies and ensure the relevance of programmes; Establish the value added of past interventions and provide knowledge on how to maximise the value of future interventions; Capture their unintended consequences or unforeseen obstacles and facilitate policy learning. All the above benefits of impact assessment are in particular true for assessing impacts of reform initiatives. Reforms differ from most policy initiatives because they often affect large constituencies, and reallocate the flow of costs and benefits within society. Thus reforms are often actively resisted by those, who are likely to bear the brunt of the costs and later strive for the redress of their grievances in the next elections. Any reform has to be well prepared, supported and justified by robust evidence, and certainly well communicated by those involved in the planning and implementation of reforms. Evidence from previous impact assessments of similar reforms in other countries is particularly helpful to politicians willing to undertake education and training or other reforms in their home country. Definition of reform Limitations in assessing impacts of reforms There is no single correct definition of reform in theory or practice. Drawing on the work of Pollitt and Bouckaert, we suggest using a broad definition, which encompasses all the deliberate and essential improvements in public management structures, processes and policy content (Pollitt and Bouckaert, 2000), which lead to greater policy efficiency and (or) social justice. Deliberate means existence of clearly identifiable political and (or) legal initiative (strategy, programme, legal act, etc.), setting out the logic of reform (objectives, expected outcomes, timing, resources, etc.). Essential means that changes brought by the improvements are characterised by significant breadth, depth, and intensity. Thus we think that any reforms, which are national or system-wide or covering a substantial component of education and training system are the most obvious targets of the guidance on impact assessment provided in this document. Studying the experience of such reforms provides a special value for policy learning within and across the EU member states and should be prioritised in future research. Tracing the reforms from their systemic-level changes down to impacts at individual, organisational, systemic or societal level is still not widespread in the existing practice of policy analysis and evaluation. There is a body of literature analysing institutional and administrative capacity development, but only a fraction of it has been devoted to studying institutional and policy reforms in the education and training sector (see the review of relevant previous research in Chapter 3.2). This might be attributed to a number of general reasons such as lack of motivation and investment in data collection and analysis, but also to several specific reasons, which 36

39 include the complexity involved in researching the impacts of reforms, including those in the area of education and training. Multifaceted nature of reforms Indirect relationship to ultimate beneficiaries Influence of contextual factors Assessing the impact of education and related reforms on adult learning is constrained by several factors. The first is the multifaceted nature of variables characterising policies. Modernisation of public management and policies requires complex programmes. For example, improvement of flexibility in the provision of education and training is often dependent on decentralisation of the management of providers, strengthening their autonomy, but also designing accountability arrangements based on outputs rather than inputs, engaging social partners in deciding the content of services or their quality assurance, introducing modular programmes, developing the recognition of prior learning, developing competences to implement all the above, etc. A slip in any of the above factors or their improper sequencing might impede the overall progress of reform or limit its impacts. Related to this is the second constraining factor. Reforms of education and training policies do not target directly the ultimate beneficiaries of education and training services. They are usually aimed at systemic and structural improvements of service provision and these improvements are expected to result in better quality or accessibility of services. For example, governance reforms and institutional capacity development would be likely to affect adult learning through better regulation and its enforcement, improved better design, targeting, and more effective and efficient delivery of financial interventions. These of their own turn would mean better provision to the ultimate beneficiaries. The third constraining factor is the importance of contextual factors in determining policy reform outcomes. Reforms often require wellco-ordinated changes in more than one level of governance, agreement between numerous institutional actors, engaging interest groups and overcoming (often conflicting) vested interests. The complexities and limitations in the decision making phase are confounded by even more complex challenges in the implementation phase. The more simplistic is the design of reform (excluding some of its relevant elements), the greater will be the danger for adverse influences from contextual factors, which have not been internalised in the design of reform. But the more complex is the reform programme, the more likely it is to be bogged down by its share of complexity and lack of resources for all relevant interventions. The multifaceted nature of reforms, their indirect influences on adult learning and the significance of contextual factors have important implications on how the reforms could be analysed. They often rule out or impede the use of experimental or comparative analysis. Firstly, reforms most often differ in scope and external environment and it is difficult to control for large numbers and different intervening factors. Secondly, it is very difficult to identify a suitable external counterfactual and any analysis is often confined to comparing a hypothetical reform to no-reform situation. The methods of analysis are discussed in Chapter 4.2 of the report. 37

40 3. THEORIES AND PRACTICE OF ANALYSING EDUCATION AND TRAINING SYSTEMS, THEIR REFORMS AND THEIR IMPACTS This chapter provides an overview of the literature analysing education and training systems, their reforms and their impacts. It categorises the types and areas of education and training reforms and provides concrete examples of impact assessment of reforms presenting their findings and methods used. This chapter provides the basis for the subsequent chapter, which provides guidance on the key steps in planning and implementing a good quality impact assessment of education and training reforms on adult learning Types and areas of education and training reforms Types of education and training reforms are not different from those of other public policies. However academics use different typologies to classify government interventions, including all kinds of public policy and management reforms. For example, Nicholas Barr offers the following typology of forms of government intervention (Barr 1998): Regulation where the state interferes with the market through a large number of new or amended legal acts. This is, however, the most widespread type of intervention as no deliberate or essential reform is done without changing the legal basis (e.g. adoption of a new law on vocational education and training). Finance where the state provides subsidies (or taxes) to prices of specific commodities or affects the incomes of individuals. As most education and training depends on limited financial resources of the state, the sound performance of this type of intervention is crucial for the successful development of a sector (eg. the development of demand-side financing, creation of efficient and equitable costsharing mechanisms). Production or service provision where the state takes over the supply side by producing goods and services itself. This intervention is very different from previous ones (regulation or finance) as the former changes the market mechanism, while the latter only changes market outcomes and leaves the basic mechanism intact. In this type of intervention the state owns capital inputs (school buildings and equipment) and employs the necessary human resources (e.g. teachers). Income transfers where the state changes the income of individuals with the help of transfers tied to specific types of expenditure (eg. vouchers) or untied transfers (e.g. social security benefits). 38

41 Areas that are the most important in reforming adult learning sector The first three types of reforms are obviously the most relevant for the education and training sector. Generic typologies of reforms provide only limited information on how they could affect the levels of adult learning. Education systems are characterised by an increasingly large numbers of policy features. However not all are relevant to the adult learning sector. For the purposes of this report we have selected the following policy characteristics of the education system which are the most important in adult learning sector reform (for more detailed description see chapter 1.3 in Annex 1): Standardisation: the term standardisation refers to the degree to which governments create the conditions (e.g. teachers education, education financing, curricula, etc.) and the control mechanisms (nationwide testing, school-leaving examinations, etc.) to achieve certain nationwide standards of quality in the education provided by different schools (Matějů, Soukup, Basl and Smith 2007). Highly standardised education systems are found, for example, in France, Germany, Italy, Scandinavian countries, the Netherlands and the United Kingdom. Stratification: the term stratification refers to the numbers and types of transitions available to the next schooling level (Allmendinger 1989). Stratification could be regarded as the practice of using testing and tracking on a group of students based on a variety of factors. For example, in the Netherlands where stratification is very high pupils are already grouped according to their abilities into different education tracks during their basic secondary education from the age of 12. At later stages underperforming pupils can be downgraded into less demanding tracks, but the opposite is no longer possible. Other examples of highly stratified education systems are Austria, Belgium, Czech Republic, Germany, Hungary and Liechtenstein. Meanwhile in Scandinavian countries where stratification is very low youngsters do not move to secondary education until they are years old and the entire compulsory period of schooling (primary education, including the first phase of secondary education) takes the form of a single type of education for all children between the ages of 6 and Vocational specificity: the term vocational specificity is the degree to which curricula are designed to prepare students for particular vocations (Matějů, Soukup, Basl and Smith 2007). A greater emphasis on specific vocational skills and a closer link between schools and employers should lead to an easier transition from education to the labour market. Highly vocational specific education systems are, for example, found in Austria, Denmark, France, Germany, Ireland, Italy, Luxembourg, the Netherlands and the United Kingdom. Flexibility in provision of learning services: according to the results of Cedefop and OECD surveys, time constraints were the most important reasons for not participating in learning programmes, as indicated by those adults who would like to participate (Cedefop 2003, 2004; OECD 2003b). This is confirmed by the most recent results of AES (2007), which shows that 50.8% of the respondents, who did not participate in adult learning, indicated this was due to the lack of time due to family responsibilities. Part-time education arrangements, modular learning, credit transfer system and distant learning create opportunities for more flexible learning as well as better conditions 39

42 for learners mobility. Highly flexible education systems operate, for example, in Belgium, Scandinavian countries, Slovenia, the Netherlands, and the United Kingdom. Validation of prior learning: the Cedefop glossary (Cedefop 2008a) defined validation of learning outcomes as confirmation by a competent body that learning outcomes (knowledge, skills and/or competences) acquired by an individual in a formal, non-formal or informal setting have been assessed against predefined criteria and are compliant with the requirements of a validation standard. Validation typically leads to certification. Validation of prior learning and transfer of credit between institutions are one of the main objectives of the development of both sectoral and National Qualifications Frameworks in various countries. Examples of countries where validation is, to varying degrees, a practical reality for individuals are Belgium, Denmark, Estonia, Spain, Finland, France, Ireland, the Netherlands, Norway, Portugal, Romania, Slovenia and the United Kingdom. Information, guidance and counselling: existence of information, guidance and counselling systems ensure smooth functioning of both education and labour markets. They help adults to identify, assess and choose the most appropriate learning and work possibilities. The best practice examples of information, guidance and counselling systems often include Ireland and Scandinavian countries. Governance of institutions: this area basically concerns the mode in which educational institutions are being managed. The most important issue is increased management autonomy, i.e. less input controls on financial and human resource matters. It is a significant institutional driver for the efficiency of the education and training institutions. Ownership and competition are also referred to as the factors leading to efficiency. However evidence regarding the latter factors is contradictory. Countries that are more inspired by the new public management ideas (e.g. the Netherlands, the United Kingdom) are often mentioned as the best practice examples in terms of governance of institutions. Belgium, Czech Republic, Estonia and Sweden are also mentioned among countries with the high autonomy of schools. Financing of education and training: the lack of financial resources is often indicated by individuals as one of the most important reasons for not participating in learning (Cedefop 2004). Therefore, the sustainable increase in education spending seems to be a vital institutional factor for economic growth and achievement of the Lisbon objective to increase the overall level of adult learning. Efficient use of limited public resources for education and training and introduction of cost-sharing mechanisms, hereby stimulating private financing of education and training, are the main themes in this area. Examples of countries with very high total public expenditure on education as % of GDP are Belgium, Cyprus, France, Iceland, Malta, Scandinavian countries and Slovenia. Human resources: human resources in this case include teachers, trainers and the training support staff. Teachers and trainers are the most important, as they are the primary guarantors of quality and equity of adult education and training. Countries increasingly perceive teachers and trainers as the key to making education and training reforms work. Policy measures that 40

43 focus on the quality of teachers and trainers have become one of the priority areas in many countries (Cedefop 2009b). Raising the status of teachers and trainers, setting standards, increasing their qualification levels and keeping initial and continuing training upto-date have become major issues of concern in current policymaking. Typical best practice examples of developing and motivating adult learning human resources are Nordic countries. The above-mentioned areas that are considered the most important in reforming adult learning sector could be grouped according to three main types of reforms: Changes to standardisation, stratification, vocational specificity, information, guidance and counselling and validation of prior learning in education and training systems would fall into the group of regulatory (or legal) reforms; Changes to flexibility in provision of learning services, governance of institutions and human resources in education and training systems would fall into the group of service provision reforms; Changes to efficient use of public resources as well as stimulation of private financing of education and training would fall into the group of financial reforms. Very often reforms include a complex mix of several types of interventions, which are designed to mutually reinforce each other in influencing adult learning. Also reforms important to adult learning are not necessarily limited to the reforms of the education and training sector. Reforms of labour market and health policies as well as those influencing contextual factors such as the macroeconomic environment, social cohesion, innovation and research, and environment and transport might have significant impacts on adult learning (see Annex 1 for the overall discussion on factors influencing adult learning) Examples of impact assessment of relevant public policy and management reforms In this part of the report we present some examples of the previous work done in assessing the impacts of public policy and management reforms on adult learning or other factors, which are closely related to adult learning. There are relatively few previous impact assessment studies or reports on the subject. Table 8 below provides a review of reforms which were followed by an impact assessment. It aims to illustrate immediate and intermediate impacts which were reported in the impact assessment reports, as well as summarize research methods used. The table summarizes evidence on 12 reforms and 20 impact assessment reports, which are not limited to education and training reforms, but also public policy and management reforms in other sectors such as the labour market. In the table, likewise in all the report, a broad definition of reform is used. It comprises intended reforms, major programmes and other instruments, which might have had an impact on adult learning. 41

44 Name, country and start year of the reform VET 2000 in Denmark (2000) Competence reform in Norway (2001) Hartz reforms in Germany ( ) Table 8: Review of the selected public policy and management reforms 7 and research methods used for their impact assessment Scope and type of Intermediate or intervention Immediate impacts reported ultimate impact Research methods used Vocational education and training Reform Higher education Reform Labour market policy, including continuous vocational training Reform Key elements of the reform Introduction of individualised learning pathways for students, student logbook, guidelines for ministerial orders, involvement of teachers in the guidance procedure. Recognition of informal learning in admission to Norwegian higher education institutions. A number of reforms to enable active labour market policies (e.g. reorganisation of employment agencies, provision of benefits depending on person s learning activities). Introduction of training vouchers as new instrument to raise participation level in CVT and to improve quality of CVT while improving the match between training More flexible organisation of learning activities, more autonomy and curriculum planning opportunities for teachers. Cooperation between apprenticeship companies and students improved where logbooks were used (Nieuwenhuis and Shapiro, 2004). Higher education institutions developed new elaborate guidelines and systems for recognition of non-formal learning during the admission process (Opheim and Helland 2006). Helped improve the effectiveness of active labour market policies, especially training programmes (Jacobi and Kluve 2006). Abridged courses of vocational training for unemployed people resulting in reduction of lock-in effect (when unemployed persons under training are not actively searching for the job). No effect or negative effect on the participation rates of low qualified reported No end point impact assessment (on adult learning) is provided (Nieuwenhuis and Shapiro, 2004). Provided better access to higher education for persons with non-formal competences (Opheim and Helland 2006). No end point impact assessment (on adult learning) is provided (Jacobi and Kluve 2006). No end point impact assessment is provided. Surveys, including survey of drop-out students; focus groups with managers and teachers; interviews with apprenticeship companies, local educational councils, teacher unions (Nieuwenhuis and Shapiro, 2004). Desk research (document analysis: local guidelines from 57 of 70 HE institutions in Norway); interviews with key personnel at HE institutions (14 persons from 9 institutions); data from centralised admissions database (Opheim and Helland 2006). Meta-analysis of earlier smaller impact assessment reports and evaluations (Jacobi and Kluve 2006). Meta analysis of evaluation studies, econometrical modelling, secondary data analysis (usage of statistical data), survey, expert interviews, monitoring data analysis, 7 This table includes not only education and training reforms, but also public policy and management reforms in other sectors such as the labour market. 42

45 Name, country and start year of the reform Three-year IVET programmes in Sweden (pilot programme in 1991) Modern Apprenticeships programme in United Kingdom (1995) Adult Education Initiative in Sweden ( ) Program "Lernende Regionen - Förderung von Netzwerken" (Learning Scope and type of intervention Vocational education and training National pilot programme Apprenticeship National programme Upper secondary education for adult National programme Lifelong learning Federal programme Key elements of the reform supply and demand and stimulating competition between training providers. Instead of two-year VET programmes during upper secondary school, three year programmes were introduced, with more emphasis on workplace training. Introduction of new apprenticeship schemes, new types of certificates, new funding measures. A substantial amount of additional state grants for adult learning and the responsibility of municipalities to provide flexible and individualised services for adults Networking approach towards education and training. Building regional Immediate impacts reported persons and other disadvantaged groups. Reduced cooperation between employment offices and training providers. Insignificant effect on reduction of unemployment (Schneider et al, 2006) (Bonin and Schneider, 2006). Reduced take-up of learning opportunities among the members of the target group the unemployed (Kühnlein and Klein, 2003). Raised university enrolment rates in pilot municipalities; did not have any immediate effect on population activity rates (Ekström (2002). Helped introduce new training, attract new learners, provide broader or higher level of training, increased flexibility and effectiveness of training opportunities (Anderson and Metcalf, 1995). Increased participation of unemployed adults with lower levels of education within municipal adult education. Around adults raised their educational level (Stenberg, 2003). Increased participation in training of socially disadvantaged groups less participating in CVT and people with a Intermediate or ultimate impact reported No end point impact assessment (on adult learning) is provided (Ekström (2002). Encouraged positive views of employers towards training (Anderson and Metcalf, 1995). About 25 per cent of the participants moved on to further learning other than labour market training. The reform had a beneficial effect on unemployment incidence, but a negative effect on unemployment duration (Stenberg, 2003). The end-point impact assessment exists (Dobischat et al 2006, Research methods used qualitative interviews, case studies (Schneider et al, 2006). Expert interviews, survey, secondary data analysis (usage of Employment office data) (Kühnlein and Klein, 2003). Survey, secondary data analysis (usage of employment office data) (Bonin and Schneider, 2006). Data from register of pupils, register of higher education, register-based labour market statistics. Data between pilot municipalities and other municipalities were compared (Ekström (2002). Telephone survey of 1500 involved employers (random sample from programme database). For the analytical purposes the data was also divided by sectors (Anderson and Metcalf, 1995). Assessment of counterfactual using statistical analysis and econometric modelling methods (Stenberg, 2003). Case studies, qualitative interviews, biographical interviews, expert 43

46 Name, country and start year of the reform regions promotion of networks) in Germany (2001) Union Learning Fund in United Kingdom (1998) National Framework of Qualifications in Ireland (2003) Scope and type of intervention Workplace learning, National funding instrument Qualifications Awards instrument Key elements of the reform networks among different formal, non-formal education and training institutions, employers organisations aiming to approach regional shortcomings in CET. The position of so-called Union Learning Representative (ULR) was introduced. The ULRs were to be responsible for promotion of learning in the workplace, definition of training needs. New fund (the Union Learning Fund) was introduced to support these activities. Introduction of the National Framework of Qualifications. Immediate impacts reported poor labour market position. Improved transition between education and training and employment. Developed advisory services. Increased transparency of supply of training services in the regions (Forschungskonsortium 2004). Developed network of the regional counselling services. Developed quality standards of the educational process. Increased supply of services for the disadvantaged groups. (Nuissl et al 2006) Successfully introduced ULR positions where the unions were already strong, drew sufficient support for ULRs from the unions, many unions approached employers with learning agreements (Wood and Moore 2005). Helped introduce new training, provided guidance for learners, including disadvantaged learner groups. Encouraged trade unions to adopt written policy commitments to LLL, included specific funding for learning activities (Shaw et al 2006). Partnership development; successful engagement of employers; success in reaching non-traditional learners (York Consulting 2005). All Irish awards from basic education and literacy to post-doctoral level across the education and training Intermediate or ultimate impact reported Emminghaus and Tippelt 2009), but has been inaccessible to research team. Improved overall collaboration between unions and employers (Wood and Moore 2005). Raised awareness of LLL among employers, ensured continuity of training (Shaw et al 2006). Building capacity of ULRs to negotiate workplace learning provision with employers and suppliers; strengthened the capacity of its social partners (employers and trade unions) to influence national policies on workforce development and their associated delivery (York Consulting 2005). No end point impact assessment (on adult learning) is provided Research methods used interviews, observation, surveys, monitoring data analysis, document analysis (Conein et al 2002) Surveys of network managers and partners, expert interviews, documentary analysis, secondary data analysis, case studies (Nuissl et al 2006) Survey of national union representatives and union learning representatives, workplace case studies (Wood and Moore 2005). Survey of employers, trade unions, union learning representatives, learners; case studies of employers (Shaw et al 2006). Surveys (of learners, of ULRs); consultations with key stakeholders; focus groups (York Consulting 2005). Desk research (document analysis, analysis of qualification award data); 44

47 Name, country and start year of the reform Amended legislation of continuous training in Germany (2006) Scope and type of intervention Lifelong learning Legal instrument Key elements of the reform Legislation introduced to stimulate continuing training and to build up a system of lifelong learning. Immediate impacts reported sectors are included in the National Framework of Qualifications. Improved quality of training programmes, qualification awards, standards; improved learning experience (FETAC 2008). New standards in adult learning (although the process is slow) (Collins et al 2009). Change in supply of services of continuing training organisations towards more labour market specific training (Boden et. al. 2007). Intermediate or ultimate impact reported (FETAC 2008). Increased number of progression routes into higher education; better transparency of these routes (Collins et al 2009). No end point impact assessment (on adult learning) is provided. Research methods used focus groups with stakeholders; online survey (160 respondents - all types of stakeholders, including programme providers, teachers, learners, government officials) (FETAC 2008). Desk research (a background paper); reports from key stakeholder bodies in a common template; case studies; consultative forum of stakeholders (focus group) (Collins et al 2009). Expert interviews, group discussions, Peer-Review (Boden et al 2007) Vocational Training Opportunities Scheme in Ireland (1989) RVCC (National System for the Recognition, Validation and Certification of Competences) in Portugal (2000) Secondary and vocational education ESF funded project Lifelong learning, ESF funded project Integration of vocational and general skill development, ability to seek secondary and vocational qualifications simultaneously. Introduction and development of RVCC which helps, firstly, to recognise the knowledge and skills acquired inside and outside the professional sphere, and, secondly, to validate them by registering and certifying so-called key competences. Increased participation of adults with low levels of qualification; increased educational attainment (WRC Social and Economic Consultants Ltd 1994). Increased self-knowledge, self-esteem or self-improvement of learners; increased employment and raised salaries; positive transition from fixedterm to open-ended contracts (CIDEC 2004). Improved certification, redirection of learners to learning activities (Instituto Better progression of learners to further educational course (WRC Social and Economic Consultants Ltd 1994). Motivated learners to pursue formal education: about 13% of the people who answered the questionnaire continued their studies returning to the regular education system; and more than two thirds (65%) of the respondents were thinking of studying further (CIDEC 2004). Improved social performance and Methods were not clearly identified; the findings suggest that admission data or survey of learners could have been used (WRC Social and Economic Consultants Ltd 1994). Desk research; surveys; case studies (CIDEC 2004). Methods were not clearly identified; the findings 45

48 Name, country and start year of the reform Families First programme in Tennessee, the USA, 1998 Source: compiled by the authors Scope and type of intervention Welfare reform programme Key elements of the reform Work with local community organizations to develop educational materials, the training of teachers Immediate impacts reported de Gestão do Fundo Social Europeu, 2004). The Families First programme has encouraged a more collaborative and coherent approach to assistance for adults in need. The programme has strengthened the infrastructure of adult basic education by increasing funding and by providing a reliable, fairly consistent market for adult basic education services. Intermediate or ultimate impact reported employability of participants (Instituto de Gestão do Fundo Social Europeu, 2004). The importance of basic skills education was acknowledged as a vital part of welfare reform efforts. Research methods used suggest survey of participants, as well as certification data from RVCC centres could have been used (Instituto de Gestão do Fundo Social Europeu, 2004). Interviews with key actors, informal conversations with students, desk research. (Connie White, Mary Ziegler, and Beth Bingman, 1998) 46

49 The analysed impact assessment reports cover a variety of reforms: from the establishment of qualification frameworks, recognition of prior learning, to tackling various financing and governance issues in adult learning and generally in education or labour market sectors. There were several reports assessing the impacts of reforms dealing with the recognition of non-formally acquired competences. The reviewed studies used different research methods to assess the impact of the reforms. For example, both Irish National Qualifications Framework impact assessment reports are based on desk research (mostly document analysis), but one of them also uses focus groups and online survey as a tool for data collection, while another case studies and detailed surveys from key stakeholders. All in all desk research, surveys, focus groups and case studies were among the most widely used research methods in impact assessments. In the literature reviewed we encountered only one assessment of counterfactual, which was carried out in evaluating impacts of the Adult Education Initiative in Sweden (Stenberg, 2003). The scope and depth of the research methods applied varied across the reports. Some impact assessments drafted their conclusions from a detailed study of one stakeholder group, while the others based their more general findings on smaller datasets from various stakeholder groups. The primary focus of the reviewed studies was on the immediate impacts of the reforms on the target groups. The analysis of intermediate or ultimate impacts of the reforms was less frequent. The findings of the impact assessment studies have generally indicated positive developments in adult learning, with some hints on things that still need improvement. The reforms were found to be successful in engaging previously excluded groups in learning, creating new learning opportunities, improving the quality of learning both in terms of participant satisfaction and learning outcomes, assisting learners to obtain higher a level of qualification and opening pathways to higher education. The reforms were also reported to have had a positive influence on the employment and activity of the learners, as well as on the attitude of employers towards lifelong learning and on their relationships with trade unions. Among the issues of concern the reviewed impact assessment studies mention are the low completion and high dropout rates of learners, low rates of learners acquiring formal certification, and in some cases slow implementation of the reforms. This supports the importance placed on persistence and follow up of learners highlighted at the Peer Learning Activity on April 2009, in London, on the theme of "one step up". To summarize, the education and related reforms are complex government interventions, which are difficult to assess. We have been able to identify only relatively few examples of studies, which assessed the impacts of reforms on adult learning. The available studies in most cases measure immediate impact and say very little or nothing about intermediate or ultimate impacts of the reforms. Lack of systemic collection of monitoring data of the reforms impedes efforts to assess longer-term impacts on adult learning sector. 47

50 4. KEY STEPS IN IMPACT ASSESSMENT OF EDUCATION AND TRAINING REFORMS: GUIDELINES FOR POLICY MAKERS In this chapter we discuss the key steps in planning impact assessment. Firstly, we look at the issues of defining the scope and depth of impact assessment. Then we provide guidance on building the analytical framework. Finally, we discuss the selection of relevant methods of impact assessment. The guidelines are addressed to policy makers who will outsource impact assessment to professional analysts, but who need the expertise to initiate, give direction to and follow up the process of impact assessment Deciding scope of impact assessment As in any other area resources available to impact assessments are usually scarce and limited. Therefore the ambition of the assignment in terms of the number of evaluation questions has to be commensurate to the resources available. This requires careful planning of the scope of the research, and the breadth and depth of the impact assessment plan. Some impact assessment assignments could be deep aimed at answering detailed questions about one specific type of impact, cross-checking and testing the findings using several data sources and methods. Other assignments could be broad intended to answer more general questions about several types of impacts of the same intervention or several similar interventions. The broad scope of impact assessment allows an explanation of something about many cases rather than explaining a lot about a single case or a small set of cases. It allows generalisations. Depth allows comprehensive, detailed and rich analysis of a single research question or of the impact of a single intervention. As J. Gerring puts it, the demand for analysis, understanding or explanation is, among other things, a call for depth 8. Given limited resources the broad scope of the assessment might therefore limit the depth of analysis and the ability to provide reliable answers based on a sufficient number of observations, the complementarity of sources of data and methods of research. More information about breadth and depth can be found: Gerring, J. Social science methodology a criterial framework. Cambridge University Press, In this chapter we provide answers to the following questions: Why clarity about the scope and depth of impact assessment is important? What are the factors influencing the scope and depth of impact assessment? What are the levels of impacts and levels of their analysis? 8 Gerring, J. Social science methodology a criterial framework. Cambridge University Press, 2001, 106 p. 48

51 Why is it important to select an appropriate level of analysis? Why timing of impact assessment is important, how to select it properly? What are the benefits of planning impact assessment in advance? What is the role of theory in impact assessment? How impact assessment questions are formulated? How the feasibility of answering them is assessed? Which parts of an impact assessment plan have to be defined by the organisation commissioning the impact assessment? Intended and unintended consequences The scope of impact assessment depends on aims and objectives of reform Impact assessment could be designed to assess the unintended consequences of reform The scope of impact assessment is usually defined by the intervention logic of the reform initiative itself. The overall aims and objectives of the initiative correspond to its desired ultimate and intermediate impacts. Most often impact assessments serve as a means of establishing to what extent the aims of the intervention have been achieved. In this case they are implemented or commissioned by the institution responsible for the success of the intervention. This institution has limited flexibility in determining the overall scope of impact assessment, which has to cover each and every impact explicitly sought in the reform initiative. If the reform suffers from inherent weaknesses such as unrealistic or conflicting objectives, so too will the impact assessment be constrained, unless the objectives are agreed and clarified in the impact assessment plan. It is therefore considered good practice to involve evaluators in the early stages of reform planning. Before an intervention, impact assessment helps to ensure the relevance of its design in terms of addressing the right problems with the right measures. It also helps to verify its internal logic - the theory about relationship between its inputs, outputs, results and impacts. In future reforms, therefore, impact assessment should be built in from the start. Impact assessment could also be designed independently of the officially stated intervention logic of the reform initiative. Reforms often have numerous unintended consequences, which are interesting to a much broader variety of actors than those directly involved in the implementation reform or those immediately affected by its measures. In this case planners of impact assessment have greater liberty in designing the scope and depth of impact assessment. The assessment of the impact of education and training and other reforms on adult learning, when quantity, quality and (or) accessibility of adult learning is not among the primary objectives of the reform, would often fall within this category of impact assessments. This could include the wider benefits of adult learning in terms of improved health, social participation, etc Levels of impacts Appropriate levels of impacts and of their analysis have to be considered Assessments could be targeted at different levels of impacts. Any major reform could bring not only substantial changes to the lives of individuals, but also changes in their organisations, broader policy framework and society. On the individual level reform of education institutions and policies might improve e.g. the quality and accessibility of learning, which would lead to greater learning participation, higher qualification levels. This could also yield 49

52 individual benefits on the labour market, which in turn would again improve the motivation of individuals for life-long learning. In this case learning could be considered both as an intermediate and ultimate impact, since increased learning might lead to increased motivation for more learning. On the organisational level reform might improve e.g. the governance of service providers, which could diminish costs, improve the motivation of staff, and the efficiency of service provision. Greater efficiency of provision might enable more investment in quality and accessibility of services, which could lead to further impacts on the individual level. Finally, reforms might trigger further changes and adaptation in a broader policy framework and society. For example, a successful educational financing reform might stimulate the demand for recognition of prior learning and lead to further regulatory and financing initiatives in this area. Or setting up a new network of schools or providers of counselling services might determine the need to define the boundaries and identities of the new service provision systems with respect to other systems, improve their interoperability, harmonise standards, etc. This might lead to a string of regulatory initiatives, which might cause restructuring on the organisational level and have further effects on an individual level. Given the shared complexity of potential inter-relationships between reform and its impacts it might be impossible to trace and assess all the possible impacts. A good planning of impact assessment would thus involve choosing the most relevant scope of analysis. Simultaneous analysis of all levels and types of potential impacts might be unfeasible given resources and time available Timing of impact assessment Proper timing for impact assessment is crucial for its success Impact assessments should be planned in advance A crucial factor defining the scope of impact assessment is the timing of the impact assessment. Only immediate impacts could be measured during or immediately after the intervention. Initial changes brought by the reform often trigger further changes and transformations, which might take long periods to materialise. The ultimate impacts of reform might become measurable only several months or even years after the reform initiative was completed. In practice this causes a conflict between the need of reformers to know and be able to demonstrate the outcomes of reform as soon as possible and the limited ability of researchers to provide reliable answers about the success or failure of reform on the basis of only immediate impacts. Most official statistics, which is often one of the major sources in measuring the ultimate impacts, takes 1-2 years to be collected, tested and officially published. Thus final assessment of ultimate impacts might be feasible only 3-5 years after the reform. The timing of impact assessment has to be chosen carefully depending on the substance and the timing of reform. However, assessment of ultimate impacts might be very difficult if not impossible without previous assessments of immediate and intermediate impacts. These often provide the causal links between the reform intervention and its ultimate impacts. Ideally impact assessment programmes should be long-term and well planned to trace different impacts when they are expected to take place, to utilise appropriate methods of analysis when they can be applied 50

53 and to make use of relevant data, when it becomes available. The collection of evidence for impact assessment ideally should be integrated into the reform plan and relevant data collected throughout the process and after the completion of reform. Tracing immediate and intermediate impacts might be difficult or impossible when too long a time has passed after the reform (e.g. the surveys of beneficiaries are possible only if they still remember and can answer detailed questions about their participation in specific interventions). The best timing for establishment of control groups for counterfactual analysis or making baseline assessments is before the start of reform. This provides the basis for measuring impacts as they appear during and after the reform. It is important to note that the impact assessment process of the reform might be staged before, during and after its lifetime or when it is subsumed into the next reform Reconstructing the logic of reform intervention Impact assessment questions have to be based on theory explaining the relationship between reform and its impacts As we have seen in Chapter 3, there are different theories explaining the different aspects of education and training systems and their impacts. Therefore the next step in planning impact assessment is to reconstruct the theory underpinning the relationship between specific reform and its impact. This is necessary to assess the ability of the reform initiative to reach its intended goals, e.g. increase the quantity and quality of adult learning. A clear identification of the reasons why this should be expected is an important precondition to asking the right impact assessment questions. If the causal theory underlying the reform initiative is not sufficiently well explained in the reform programme, this task has to be undertaken before commissioning an impact assessment. Without this potential external evaluators might provide very different impact assessment plans in their proposals, which will be hard to choose from on the basis of a similar set of criteria. The two main methods of filling the gaps in the logic of reform intervention are the analysis of stakeholder views and the body of scientific knowledge. The most obvious stakeholders are the decision makers in the institutions responsible for designing and delivering the reform. Their views could be obtained via interviews or analysis of reform documents. Once the opinions of key stakeholders have been considered, it might be useful to review the already existing body of scientific knowledge about the reform factors and their influences. The usefulness of theoretical (scientific) knowledge for impact assessment could be defined as follows: Theories help to simplify and understand the workings of a complex empirical reality. They help to identify the most salient factors so that impact assessment might concentrate on their analysis. Theories are often the results of empirical research. Therefore they often explain causal relationships that have been proven elsewhere. Theories are scientific: they define concepts, the main factors and their relationships, which can be applied in explaining similar phenomena in different environments or situations. 51

54 Logical models provide frameworks for analysis An in-depth analysis of theoretical knowledge is normally left to potential evaluators to do in their evaluation proposals or even the initial phases of the assignment. It is very important that potential evaluators are requested to explain their understanding of reform logic, which is usually detailed in a logical model. This model later becomes the framework for analysis. The logical model is a way of demonstrating the expected relationships between the reform actions/inputs and impacts as well as intervening external factors. Logical models are based on a number of assumptions, which help to understand a complex reality. When any assumptions prove to be wrong in reality, logical models have to be amended to take this into account. An example of a logical model for analysing the effects of education and training reforms on adult learning is provided in Chapter Defining questions for impact assessment and assessing their feasibility General impact assessment questions have to be broken down into specific ones; questions irrelevant to decision making should be dropped Impact assessment should include salient external factors Once the issues on the purpose, level, scope, breadth, depth and timing of impact assessment have been carefully considered, the planners would be ready to define specific questions to be answered by impact assessment. Questions would normally be intended to predict or describe and (or) explain the possible impacts depending on the timing of impact assessment. The questions which are designed to gather predictive information are raised when the impact assessment will be carried out before expected changes could be observed. They would normally address the ultimate impacts of reform, when only the immediate impacts or no impacts can be monitored. Questions about facts (description) and explanations as to why those facts occurred (analysis) are raised after intervention and asked if the implemented reform has led to any expected impacts. General questions would normally be split into a series of more detailed ones. Detailed questions would address different relationships in the causal chain between reform and its ultimate impacts, influences of external factors on the success of reform and possibly any unexpected impacts of reform. All impact assessment questions have to aim at obtaining information relevant to decision making. Questions which are irrelevant to decision making would normally be dropped for pragmatic reasons of saving resources. Impact assessments are usually inquiries into long and complex causal chains. The length of the causal chain is determined by the number of factors which explain the connection between the reform and its ultimate impacts. The complexity is determined by the number of external conditions (often called conditional factors ) under which the links between the reform and its ultimate impacts work. The longer the causal chain, the more important external influences are and the more difficult it is to measure precisely the value added of a particular reform. The assessments of the more immediate impacts of reform would be more precise and definitive than those of ultimate impacts. To increase the precision and reliability of assessments of ultimate impacts the relevant external factors 9 have to be studied. Thus questions on the 9 The relevant external factors can be characterized on behalf of chosen theory, body of previous empirical research and by means of exploratory research, which can be integrated in the overall design of impact assessment. 52

55 external influences on reform and its impacts should be compulsory for any impact assessment plan. Analysis of external factors would often offer explanations, as to why the reform has not had all the expected impacts, why its impacts have been reinforced or offer an explanation alternative to reform logic of why the impacts have taken place. Managing external influences and uncertainty is the greatest challenge to any reformer and recommendations based on impact assessments of previous reform could be a valuable source for future decision making. Feasibility of answering questions has to be considered Once impact assessment questions have been identified, the feasibility of answering them has to be considered. A number of factors can make a question difficult or impossible to answer. If certain questions seem unlikely to be answered properly with the time and resources available, it may lead to the decision to drop them entirely or to review their formulation. If the key impact assessment questions do not pass the feasibility test, this may lead to postponing or deciding not to undertake impact assessment. A preliminary analysis of feasibility would have to include at least some of the following issues (Tavistock Institute 2003): Probability that impacts will have taken place by the time the impact assessment starts; Probability that the conclusions of impact assessment will still be needed for decision making when the impact assessment is completed; Possibility that impact assessment questions will be answered by other research projects; Clarity and stability of objectives of impact assessment; Stability of concepts and theories necessary for explaining causal relationship between specific reform and its potential impacts; Absence of political conflict, which would be liable to impede the impact assessment process; Availability of reliable data for impact assessment (about reform factors, immediate, intermediate and ultimate impacts and salient external factors likely to influence reform or its impacts); Possibility of data collection over the established period of time and using the resources available Preparing technical specification Once all of the above issues have been considered and the impact assessment is deemed useful and feasible, the planners of impact assessment would normally be ready to put their plan on paper. If impact assessment is to be carried out by external providers, its technical specification has to form the basis for their detailed research plans (which will be the basis for their selection), but will not substitute them. The technical specification has to explain the background of the assignment, its purpose and the intended usage of its results, the list of questions that need to be answered, the requirements for the level of detail and reliability of answers (setting the basic minimum requirements for methodology of impact assessment) and the timing as well as format of expected outputs. All these elements have to be well explained so that service providers can come up with specific impact assessment plans. 53

56 Analytical framework, research methods and detailed work plan are normally proposed by potential service providers As a rule technical specification also defines the minimum qualification requirements for potential service providers and their research teams. This is necessary for sending a signal about the desired level of quality of the impact assessment and for limiting the costs of preparation and evaluation of research plans. The analytical framework, methods of data collection and analysis as well as detailed work plan are normally left up to competing research teams to propose and justify. They form the basis for assessing the quality of impact assessment plans and for selecting a service provider, who is offering the best value for money. The following chapter provides guidance on the selection of methods for data collection and analysis of reform impacts. It is intended to provide information about the setting of standards for impact assessments and to provide an understanding that more sophisticated research methods also require good quality data, which might be unavailable if its collection is not built into the reform plan Choosing methods of data collection and analysis The choice of appropriate methods, the access to data and certainly the competence of researchers determines the value of impact assessment in terms of generating credible evidence and specific recommendations for policy improvement. In this chapter we review the key issues affecting the selection of methods for impact assessment. Then we present the most commonly used methods, their advantages and disadvantages, and provide some examples of how they have been applied in relevant impact assessments Purpose of impact assessment and selection of methods The choice of data collection and analysis methods depends upon the purpose, object, timing and data availability for the impact assessment. In terms of the purpose and process the impact assessments can be labelled as two broad categories (Coles 2004): 1. Summative assessment and 2. Formative assessment. Summative assessments are accountability tools Formative assessments are meant for policy learning Summative assessment aims at providing feedback to central planners on progress and is mechanical in nature, using targets or performance indicators as accountability tools. Formative assessment is embedded in the reform process itself and offers feedback to a range of interested parties. Indicators and performance measures are less prominent and are used to improve organisational learning. The assessment is seen as a way of communicating thoughts and actions across a range of actors. The complex nature of adult education and its social embeddedness makes summative assessment of limited value. The assessment process might be considered as learning about adult education change itself; therefore, a formative assessment process is essential. A review of previous impact assessments (briefly presented in chapter 3) shows that evaluations of national reforms are rarely singularly summative or singularly formative, but fall within a spectrum of types ranging from one to the other. The two distinct types of impact assessments are summarised in Table 9. 54

57 Type of assessment Table 9: Summary of data categories and collection methods used for different types of impact assessment Purpose of Focus of Necessary data Data collection assessment assessment methods Summative Feedback to the policy makers, accountability Policy targets Official statistics Quantitative Performance indicators of specific programme Monitoring data Predominantly quantitative Formative Feedback to a range of interested parties, policy learning Process of implementation Reasons for success or failure Good or bad practices Source: elaborated by the authors on the basis of Coles All of the above plus data collected by evaluators Quantitative and qualitative methods equally important The impact assessments aimed at policy learning are much more demanding than those aimed solely at ensuring political accountability. They include in depth analysis of causality and an estimation of net effects, revealing and learning from the weak points in policy design and implementation. Just as summative assessments, those aimed at policy learning assignments, are contingent on the availability of good statistics and monitoring data, so too is independent data gathering needed, and the use of a combination of qualitative and quantitative research methods Timing of impact assessment and selection of methods The choice of methods depends on the objective and timing of impact assessment The choice of methods for data collection and analysis depends on the timing and level of impact assessment. The popular data collection and analysis methods in relation to timing and level of impact assessment are summarised in table 10. Table 10: Popular data collection and analysis methods in relation to timing and level of impact assessment Individual level Organisational level Regional, sectoral or national (policy) level Immediate impacts (within 0-6 months) Intermediate impacts (6 months 2 years) Ultimate impacts (more than 2 years) Source: compiled by the authors Survey (acquiring and application of competences), Interview Survey (application of competences and remuneration), Interview Survey (career progress and further motivation to learn), Interview, Statistical data analysis and modelling Survey, Interview Survey, Interview Survey, Interview, Statistical data analysis and modelling Literature review, Interview, Analysis of monitoring data Literature review, Interview Literature review, Interview, Statistical data analysis and modelling If impact assessment has to begin while the reform is still being implemented or has just been completed, it will be focused mostly on immediate impacts. E.g. 6 months after a reform allowing the use of a new learning opportunity is a critical period during which individuals might still remember what they learned and how useful it was. Also during this period individuals have to start using the competences acquired in the process of learning or otherwise the 55

58 competences might be lost. The period after 6 months provides a good timing for further panel survey of those who used or did not use learning opportunities. Although individuals, as time passes by, would not be able to remember clearly what and when they learned and how it was relevant to what they do, researchers could track further the application of competences acquired during learning, career paths, and further motivation to learn as well as further learning. Similar research could be conducted at the organisational level - tracking the initial impacts of reform on organisations to the ultimate changes in their management and operation. Organisations could be both the subjects of learning (in terms of adopting new working methods and practices and improving their performance) and the providers of new learning opportunities for the individuals. They could be schools, which focus chiefly on organising and delivering education and training services, or companies, which produce other services or goods, but provide learning opportunities for their staff in order to improve their performance and stay competitive. Approximately 2 years after the reform the evaluators of impacts can start utilising the data which becomes available from the official statistics. E.g. the data from the evaluators own research might be related to data from the official statistical surveys and used for establishing control groups and the assessment of counterfactuals. This quasi-experimental research design might be very useful when the application of experimental research methods could not be applied by evaluators earlier. If official statistical surveys are carried out on a regular basis an over years before and after the reform, their data could be used in various sophisticated methods of statistical analysis and modelling. Most qualitative methods of analysis could be applied at different points in time and for measuring different levels of impacts. They can be applied in combination with quantitative methods of analysis (e.g. in helping to design a good survey and in interpreting its findings). Qualitative methods would also be accessible to researchers when the question of assessment requires deep and rich description and analysis or when quantitative methods could not be applied due to limited availability of data Accessing data Setting up a monitoring database The choice of timing is not the only precondition, which influences data availability for impact assessment. First of all data has to be collected so that it can be used for impact assessment or other purposes. It is therefore vital to collect monitoring data about the reform the resources allocated and disbursed, the activities and beneficiaries, the outputs and results of reform as well as other salient information, which are deemed useful for the assessment of impacts. This data can be collected only when the reform is being implemented. The absence of a monitoring process and database is one of the key obstacles which impede the impact assessment of reforms. A monitoring database has to be designed before the launch of the reform according to what this specific reform wants to achieve. Its designers have to make sure all the relevant indicators measuring its implementation and success are included, clearly defined and explained to those, who will be responsible for collecting and entering data. When a monitoring database is poorly designed it could well be useless for impact assessment. 56

59 Dealing with data protection rules Getting access to disadvantaged groups Early involvement of beneficiaries and stakeholders Detailed individual and organisational level data is subject to confidentiality and data protection rules. They limit access to it only to authorised institutions and individuals, who are obliged to follow strict procedures in handling this data. Research institutions might need to receive such authorisation or to co-operate with already authorised institutions, which could help evaluators in systemising the data, integrating different data sets and providing a research database in an aggregate form which impedes identifying individuals or organisations. Getting access to some stakeholder groups, usually disadvantaged ones, is often critical for the assessment of the impact of a reform on them. Therefore, special measures need to be planned to overcome the problems of access. These can include making links with community gatekeepers so that they can act as local advocates of an evaluation, producing instruments in different languages or in multiple formats (e.g. Braille, audio tapes), employing people from these groups to collect information, assist in running focus groups and explaining the evaluation within their own networks. Beneficiaries and stakeholders are sometimes reluctant to cooperate with evaluators, if they are to get no benefits from it. Such non-cooperation could be overcome or reduced by involving beneficiaries and stakeholders in the design stages of the assessment and in designing and piloting particular measures. Guarantees should be given that all who cooperate will receive feedback. This can take the form of a publicly available report, a letter with an executive summary or an invitation to a feedback meeting once the assessment is complete. The response rate, the access to and willingness of stakeholders and beneficiaries to cooperate are key determinants of data quality and ultimately of the assessment as a whole. Therefore, their information and involvement is normally an integral part of impact assessment Combining different methods of analysis Strengths and limitations of methods have to be considered Triangulation of data sources and methods helps to verify findings All methods of data collection and analysis have both strengths and weaknesses, so they should be chosen and applied in a manner that exploits their virtues and recognises their limitations. This means that it is best to apply methods in combination, since a single method drawing on one source of evidence provides weaker evidence than a combination of methods drawing on multiple sources and perspectives. Triangulation is a general principle for verifying findings based on the use of different sources of data to answer the same research question. Although the quality of analysis depends on a number of factors (and chiefly on the competence of the researcher and the quality of data available) triangulation might always help to avoid one-sided and subjective judgements. As a rule it entails combining several quantitative and qualitative methods of research. The ability to understand and combine sensibly multiple methods is characteristic of highly qualified researchers (Tashakkori and Teddlie 2003). Thus including a triangulation requirement in the technical specification might set a higher standard for impact assessment and help to select better qualified external evaluators. 57

60 Table 11: Strengths of quantitative and qualitative methods of data analysis Quantitative Qualitative Provide aggregate results Help to trace and explain Allow explanatory modelling subtle processes Allow an estimate of the extent Set processes into contexts and scope of change Take into account judgements Permits some degree of and experiences of comparison stakeholders Permit stronger evaluations to Help to explain causal be made of particular patterns interventions Allow the assessment of Allow for trend analyses to be differential impacts for made over time different groups Source: compiled by the authors based on The Evaluation of Socio-Economic Development, the Guide by Tavistock Institute, Qualitative methods help to locate the reform in a context Qualitative methods help to explain causal links between the inputs and outputs of a reform Qualitative methods for data collection play an important role in impact assessment by providing information useful for understanding the processes behind observed results and improve the quality of quantitative data by guiding its collection and providing the basis for interpreting findings. Qualitative methods help to locate the intervention in a wider context of the issue being addressed. Previous research shows that qualitative techniques add value to the evaluation process by establishing the causal links between inputs and outputs. These are often obtained through subjective judgments made by the people involved in the development, implementation and evaluation of reforms. Therefore, where resources allow, data collection methods that are sensitive to context should be used. These include in-depth interviews, focus groups, and literature review. The impact assessment of a reform requires quantitative measures of the impact, to judge if an intervention has worked or not, to estimate the extent of the effects and to explain the occurrence of the effects 10. Qualitative information can also be employed in estimating if an intervention has worked or not 11, but does not allow drawing a conclusion on the extent of the effects. Qualitative information is also useful to explore the reasons why the intervention has succeeded or failed. Although in many cases it is difficult to identify and develop clear criteria and indicators for success or failure, the requirement to elaborate on questions of success and failure is important for policy learning Developing a research design The choice of research design for impact assessment depends on the availability of the data. Counterfactual analysis assesses impacts of reform against a situation without reform The counterfactual analysis is a research design which provides an opportunity to estimate the effect of reform against the situation if the reform had not been introduced. The success of reform (or any other intervention) can be measured best against a situation without the reform or against other intervention alternative to that introduced by the reform. Counterfactual analysis provides an 10 For example employing regression analysis. 11 The biographical interviews and participant observation are useful methods in estimating the impact on individuals and organisations. 58

61 estimate of the circumstances that would have prevailed if a reform had not been introduced. It is carried out by comparing counterfactual outcomes (often referred to as either control or comparison group outcomes) with outcomes measured in the group targeted by the reform. The counterfactual analysis aims to determine whether or not the reform gave rise to impacts that it aimed for. If implemented correctly, this research strategy helps to control for the influence of extraneous factors and provides credible evidence of the reform impact on target groups. However, counterfactual analysis is contingent upon the ability to identify target groups and control groups identical to them before the intervention. Then measurement of the situation of both groups after the intervention has taken place. Furthermore, the comparison groups have to be representative of their general populations. Ideally data series longer and broader than just two observations for two comparison groups would be needed. However, in reality the most common strategy for analysis of impacts is comparing the situation of reform target groups before and after the intervention or even only estimating the situation of reform target groups after the intervention in the absence of pre-reform measurement. This type of research design is extremely weak in estimating other potential influences on the success of reform. Ultimately it is unable to single out the reform impacts from all impacts on its target group. The counterfactual analysis is the preferable research design, because it provides the policy makers with hard evidence of the success or failure. Yet it is hard to implement and, if implemented incorrectly, it can produce misleading results just as inferior research designs. It is very challenging to sample the control group which would be very similar to the target group, but unaffected by reform 12. In order to carry out the counterfactual analysis containing measurement of comparison groups before and after the intervention, precise monitoring data and on-going evaluation, starting from the launch of reform is needed. The evidence shows that this is rarely the case. Another type of counterfactual analysis is an ex post facto research when the comparison groups are compared only after the intervention. Further reading on counterfactual analysis Gerring, J. Social science methodology a criterial framework. Cambridge University Press, Morgan, S.L., Winship, Ch. Counterfactuals and causal inference methods and principles for social research. Cambridge University Press, 2007 Case study allows in-depth contextualised analysis Case study - an intensive study of a single case is another widespread type of research design. The purpose of this type of research, however, is to understand a larger class of cases. It distinguishes from cross-case research in the depth and intensity of a study. The analysis of one or few cases is, however, not representative for the population of all cases. It offers a possibility 12 The difficulty of sampling of comparison group lies in the problem of selectivity. The groups affected by reforms usually are in advance different from those not affected. Some ethical questions are also important to bear in mind while choosing between groups that will get the benefits of reform and that will not. 59

62 of studying a limited number of carefully chosen cases instead of a representative sample of cases (in relation to general population). Although the findings of the case study cannot be generalised to a broader population, it helps to understand how the reform worked and why it might have succeeded or failed, that is, it is useful in generating insight into causal mechanisms. The case study aims at providing an in-depth and comprehensive understanding about a clearly defined issue with regard to its context. Therefore it is often used not as a separate research design, but also as a method complementary to other methods of analysis. In the context of impact assessment the case study analysis would usually mean analysis of several cases, which include all the major properties of the target group and its context that might be useful in explaining impacts. The case study relies on multiple sources of evidence and benefits from the prior development of theoretical propositions. It could be based on any mix of quantitative and qualitative evidence and use different data collection and analysis methods. A comparative cross-case study allows drawing conclusions about similarities and differences of multiple cases. The selection of a case (or cases) is essential - inappropriate selection of cases can produce misleading results. Further reading on case study research design: Gerring, J. Case Study Research Principles and Practises. Cambridge University Press, Documentary analysis / Documentary analysis is a wide group of research methods. It is useful for screening, systematising and analysing documentary evidence produced in the process of reform and after, such as legal acts, policy documents, various administrative records, etc. The documentary analysis allows using the data to analyse the process, reconstruct intervention logic, and generate ideas for interviews. An advantage of this method is that the documents and the written records of reform in general are less likely to be subject to memory shortage compared to interviewed individuals, who might forget the details after a period of time. However, a disadvantage is that documents are hardly comprehensive and might be also subject to selectivity of the evaluator. While analysing documents is important to bear in mind that documents should be treated as data not as references. The design of analytical framework should be based on the scientific knowledge in the field. Further reading on documentary analysis: Prior, L. Using Documents in Social research Introducing Qualitative Methods. Sage Publications,

63 Secondary data analysis Statistical and monitoring data The data which was produced not by evaluators (that is why it is referred to as secondary ), but by previous research in the field, monitoring activities or data collection efforts of statistical agencies is also very important for impact assessments. It is used for establishing counterfactual, for comparing the findings of impact assessment to findings of similar research carried out elsewhere or for other purposes. The most important source of secondary data for the impact assessments is the statistical and monitoring data. Assessment of the impact of ongoing reforms is often built on statistical and monitoring data. Monitoring could be carried out through the regular collection of data at all relevant levels of governance (school, municipality, county, etc.) by using a set of indicators (e.g. enrolment rate, drop-out rate, employment incidence and duration, wage earnings of graduates, etc.). The data sets may also include information on age, gender, income, education, citizenship, place of residence, civil status and family situation. Monitoring data could sometimes be obtained from national, municipal, individual schools or other administrative registers and statistical databases. It also includes annual reports of authorities and other organisations. Impact assessments measuring actual outcomes, such as increased wage earnings, as opposed to opinions or expectations of participants about their future outcomes, is based on statistical data. The financial effect of the reform on the individual level could be assessed by comparing statistical data relative to groups of individuals participating and not participating in the reform activities over a period of time. The table 12 below is an example of the results of such measurement. Table 12: Average wage earnings levels, in thousands of SEK, before and after the reform Total AEI LMT Total AEI LMT Total AEI LMT Before After % change N 20,124 11,405 8,719 20,124 11,405 8,719 14,433 7,913 6,520 Source: Stenberg 2003a. Further reading on secondary data research: Schutt, R.K. Investigating the Social World the process and Practice of Research. Sage Publications, Casley, Dennis and Krishna Kumar The Collection, Analysis, and Use of Monitoring and Evaluation Data. World Bank: Washington DC. Interviews Interviewing as a qualitative research method, it entails asking questions, listening to and recording the answers on a particular issue. Interviews can be structured on different degrees. Most often questions are open-ended and respondents are encouraged to express their own perceptions of the issue. Qualitative interviews can be of different types. They depend on the information the evaluator wants to gather. Interviews could be expert, biographical, semi-structured, unstructured interviews, etc. Qualitative interviews are usually held with the key informants, i.e. individuals who have access to information valuable for the evaluator. Key informants are a source of information which can assist in understanding the context and process of a reform, 61

64 clarifying impacts on individual or organisational levels, particular issues or problems. Further reading on qualitative interview: Lincoln, Y.S., Guba, E.G. Naturalistic inquiry. Sage Publications, Bryman, A., Bell, E.A., Teevan J.J. Social Research Methods. Oxford University Press, 2009 Patton, Michael Quinn How to Use Qualitative Methods in Evaluation. Newbury Park, California: Sage Publications, Inc. Focus groups Focus group interview is held with a small group of people in order to elicit the whole spectrum of views. This method helps to identify trends in the perceptions and opinions expressed by different groups, collect information on the reform process and identify good or bad practises, provide a contextual basis for survey design, help to interpret findings of the survey. A challenge is however to achieve even participation and representation of stakeholders. Focus groups are not an alternative to the interview, the group dynamics allow gathering additional or different types of data compared to those obtained through interviewing. The findings of this research method are not representative of the target population, and they should not be used directly to estimate the impacts of an intervention. This method is rather useful for getting insight into the context of implementation, understanding different views and potential factors influencing implementation and possibly determining impacts of an intervention. Further reading on focus groups: Bloor, M., Frankland, J., Thomas, M., Robson, K. Focus groups in Social research. Sage Publications, Krueger, R.A., Casey, M.A. Focus groups a practical guide for applied research, Sage Publications, 2000 Questionnaire based surveys Questionnaire based surveys are used to measure both quantitative and qualitative impacts of the reform on the target group or other stakeholders. Surveys use micro data and their results can be generalized. Questionnaires usually include (Coles 2004): (a) Personal and demographic background information of the target group; (b) Ratings by the target group of aspects of education and training provision they are using or have used; (c) Judgements by the target group about future prospects for employment or learning progression. Satisfaction surveys Individuals and enterprises are usually surveyed for satisfaction ratings (a satisfaction survey). Satisfaction surveys seek to find out how individuals feel about the intervention and are generally restricted to current experience of a reform programme or immediate effects of it. They generally have a limited time dimension and rarely investigate the long-term effects of the reform. For example Results from the study on individual development under the Adult Education Initiative in Sweden ( ) showed that the majority of 62

65 the learners were very motivated to start studying under the initiative. Around 60% were very eager to begin and a further 20% were quite eager to begin studying. Those studying in programmes organised by the municipality were more motivated in the beginning, in comparison with those studying at a folk high school. In addition, the absolute majority of students (80%) were very satisfied or quite satisfied with the education they received under the AEI (Coles 2004). A typical satisfaction survey, however, can be extended to seek the views of learners about changes that followed after an intervention. For example, a survey can assess the changes of workers as a result of education and training: have become more qualified, more motivated, have been promoted, have received better pay, etc. Surveys are also used to measure change on organisational level (e.g. developing trainers, financing, qualifications) and associated supporting arrangements (e.g. guidance, skills needs analysis). Questionnaires are usually the preferred means of data gathering, but they do not allow the extraction of specific information. The assessment of the impact of a reform at the organisational level is also a difficult task because a number of benefits of learning are not quantifiable. Moreover, surveys of organisations provide information on their own costs and gains rather than wider impacts. Further reading on surveys: Iarossi, G. The power of Survey Design. A User s Guide for managing Surveys, Interpreting results, and Influencing respondents. The World bank, Tourangeau, R., Rips, L.J., Rasinski, K. The psychology of survey response. Cambridge University Press, 2000 Schutt, R.K. Investigating the Social World the process and Practice of Research. Sage Publications, Statistical and mathematical methods based on social and economic theories Impacts can be measured using a variety of mathematical and statistical methods based on social, economic, learning and other theories. They can be applied when there is good quality data available from panel surveys or other methods of collecting statistical data (like aforementioned satisfaction surveys). Selection of particular statistical methods while testing social science theories mostly depends on the type of the distinguished variables (level of measurement) and characteristics of the available data. If variables that measure the participation in reform are treated as nominal (f.e., participation vs not participation, before participation vs after participation) and variables that measure the impact (f.e., satisfaction with particular aspects of reform) are measured on interval/ratio level (with some exceptions for ordinal data), statistical methods of mean comparison could be applied. For example, if data from panel surveys is available (there are two samples of the same individuals surveyed before the reform and after the reform), dependent t-tests for paired samples (repeated measures) would be of great importance. In the aforementioned way counterfactual analysis is also implied, because two dependent samples (before the reform and after the reform) can be compared, looking for statistically significant differences between 63

66 means of the dependent variables (that measure the impact of the reform). Another example of measuring the impact of reforms by statistical methods could be independent samples t-test: comparing means of the dependent variables between two samples of people (that participated and did not participate in the reform), surveyed at one time. It could be useful in cases when data from panel surveys is not available. This method also could be applied when data from panel surveys is available, to crosscheck the results that are obtained by comparing dependent samples. If dependent variables (that measure the impact of reform) are also nominal, simpler statistical methods could be applied: e.g. tables of cross-tabs (like Table 12), measures of association for the categorical data (like Cramer s V). While measuring the reform degree on interval/ratio level, measures of correlation (Spearman s or Pearson s r) or more sophisticated methods, like regression analysis, could be applied. Another statistical approach could be econometric modelling, that uses statistical data in combination with economic theory in order to analyze and test economic relationships between a set of variables. These methods allow quantifying impacts, but examples of these methods being used for assessing the impacts of education and training or related reforms on adult learning are extremely scarce (see chapter 3.2). It is important to stress that mathematical and statistical methods have to be applied together with qualitative methods, which both help to develop framework for analysis on the basis of existing theories and then also to explain and complement the results of analysis. Further reading on statistical and mathematical methods: Field, Andy (2005) Discovering Statistics Using SPSS, Sage Publications. The impact of a reform on society is difficult to quantify The measurement of the long-term impacts of a reform in a broader societal context is a difficult task because of the attribution problem. The links between improved individual performance, improved organisational performance, greater well-being for individuals, their families and their community is hard to establish in quantitative terms and very few studies attempt to establish such links. The analysis leading to evaluating the value-for-money of reforms is problematic due to the difficulties in obtaining financial data, difficulties in comparing the gains of the reform with the situation in the system that existed before the reform was introduced, mainly due to the absence of controllable variables and the consequential lack of a baseline for measurement. The problem of attribution and data gaps however can be solved by a careful planning of impact assessment, the establishment of counterfactual and systematic data collection. 64

67 5. EXEMPLARY CASE STUDIES ASSESSING IMPACTS OF REFORMS This chapter exemplifies the application of impact assessment methodology and takes into account literature reviewed in previous chapters. The first part presents the analytical framework, research methods, structure used for carrying out detailed case studies, as well as limitations faced during the application of this methodology. The second and third parts are dedicated to the exemplary case studies of selected reforms in Sweden and Lithuania. The concluding part summarises and compares the cases Methodology of case studies Analytical framework for the case studies Figure 13 outlines the analytical framework used for the case studies. This framework is based on the input/output 13 model, which is a widely used instrument for the assessment of public management programmes (Pollit and Bouckaert 2000). The analytical framework has two dimensions. It links inputs and activities of reforms in adult learning sector with their immediate, intermediate and ultimate impacts. Inputs of the reform are defined as all types of policy interventions targeted at the achievement of the reform objectives. On the input side the framework contains three key types of reforms in the adult learning sector: regulatory (or legal, including such areas as qualifications and quality), financial and service provision. These interventions are essential for a sound and well-functioning adult learning system. Below we review each of them. Regulatory reforms. Policies and legal frameworks regulate and promote adult education. They give a statutory right to adult education at different levels, integrate all or some aspects of adult learning, regulate public financial support to service providers, establish individual entitlements to educational leave, offer financial incentives to learners, establish a framework for the recognition of prior, non-formal and informal learning, set standards, degree of stratification and vocational specificity of education and training system, improve information, guidance and counselling system etc. Regulatory reforms touch qualifications and quality areas which are very important for adult learning: Qualifications. Reforms of qualifications are of key importance in increasing motivation, access, participation and learning outputs. These include the establishment of national qualification frameworks, improvement of assessment, validation and certification procedures, development of partial qualifications, credit transfer systems etc. Reforms of qualifications create the links between formal, non-formal and 13 However impacts and not outputs are measured in the detailed case studies. 65

68 informal adult education and training, promote learning in multiple settings and labour mobility. Quality. Learning outcomes and the transparency of education depend highly on the quality of adult education and training. It comprises many inter-connected factors, such as the quality of providers, staff and delivery. Reforms in this area include efforts to introduce quality assurance mechanisms, create measurement and monitoring systems, invest in teaching methods and materials adapted to the needs of adults, make learning outcomes explicit, enhance staff qualification, etc. Service provision reforms. Adult education, in contrast to general education, is characterized as being highly fragmented, taking place in different settings (formal, non-formal and informal), organised by a variety of education providers and resulting in numerous outcomes. Therefore, efforts in the area of service provision are necessary to ensure compatibility, transparency and flexibility of adult learning. Moreover, the institutional fragmentation of adult education and training implies a great need for policy co-ordination and efficient and comprehensive monitoring of the whole field of adult education. Financing reforms. Financing is a necessary pre-condition for the implementation of any policy initiative. All governments recognize the role of public investment in adult education and training to ensure sustainability of provision and infrastructure, and equal access to learning for at risk groups. At the same time they seek to broaden financial responsibility for it by promoting cost-sharing mechanisms. Since financing is an ideological, as much as a technical issue, reforms in this area result in a shift in political culture, roles of various stakeholders and governance arrangements. Impacts in this study are defined as immediate, intermediate or ultimate benefits or, in some cases, negative effects developed within the target groups as a result of the reformed public policy interventions. On the impact side our analytical framework suggests the assessment of impacts of ongoing reforms at three different levels: Individual level (impact on the motivation of individuals to learn and actual participation in learning, educational attainment and integration into the labour market); Organisational level (impact on motivation and actual involvement of public providers and/or private sector employers in the financing and/or provision of adult learning); and Policy level (impact on administrative capacity building, institutional development, policy learning, and synergy with other policies). The Policy level is essential for the sustained improvement of adult learning in the long term. It refers to the ability of public administration to learn from previous experience and its capacity to design and deliver services that are most suitable for adult learners. The analytical framework also includes two types of contextual influences: (a) initial circumstances and constraints, i.e. the historical development of adult learning, its political and socioeconomic context, and (b) external factors, i.e. other initiatives of 66

69 the government or parallel reforms in the field of education and training. The analysis of initial conditions and external factors places the reform in a context and helps to identify factors of success or failure. Figure 13: Analytical framework of the detailed case studies Inputs / Activities Immediate / intermediate impacts Ultimate impacts Individual level: influences on motivation to learn and actual participation in learning, educational attainment and integration into labour market Initial policy, problems and needs Reforms in adult learning sector Regulatory Financing Service provision Organisational level: involvement of public providers and/or private sector employers in the financing and/or provision of adult learning The overall increase in participation of adults in education and training and improved educational attainment Policy level: administrative capacity building, institutional development, policy learning, synergy with other policies at local, regional and/or national/central level. Initial circumstances and constraints External factors, i.e. other government initiatives or parallel reforms Source: compiled by the authors Data sources The detailed case studies used the following sources of data: Extensive literature review of previously carried out evaluation or impact assessment reports, analyses, other documents related to the reform; In-depth interviews carried out with experts, who have been involved in designing, implementing and (or) studying impacts of selected reforms; and Statistical data provided in evaluation/impact assessment reports and/or data available in Eurostat or national bureaus of statistics Structure of case studies Both case studies have the same structure. The first part is dedicated to the historical development of adult education in a country, and its general context. It describes the initial circumstances and constraints which influenced the perception of 67

70 problems and needs and the selection of policy reform choices. The second part presents the object of the analysis: problems and needs, objectives and target groups of the reform. The third part is dedicated to inputs and activities of reform and is structured by types of reform intervention, which in both cases include changes in regulation, financing, and service provision of education and training. The fourth part elaborates on the external factors, which competed for influence on adult learning with the analysed reforms. The fifth part presents outcomes of the reform on individual, organisational and policy levels. The final part of the case studies draws conclusions about major achievements and drawbacks of the reform, its success factors and provides recommendations The choice of cases Two national reforms were selected to exemplify the application of impact assessment methodology: Case study 1: The establishment of labour market training and counselling system in Lithuania ( ), Case study 2: The Adult Education Initiative in Sweden ( ). The cases were selected to represent countries, which are different in terms of social and economic development in general and the development of adult learning in particular. A Nordic country (Sweden) and an Eastern European country, whose independence was restored as the Soviet Union fell (Lithuania), reflect the specificity of their respective regions in Europe. The two case studies also reflect adult learning in different stages of development. The Lithuanian reform covers the development of a traditional labour market training system with little learner autonomy and a narrow understanding of learning needs. The Swedish reform is much more complex, focused on the individual learner and reflects a rather holistic and liberal attitude. The first case reflects a top-down and rather reductionist way of implementing the reform to make adult education and training obey labour market needs. On the contrary, the Swedish example evidently reflects a complex approach to reform through cooperation and complexity, with formal, non-formal and informal learning spaces being incorporated into adult education. 14 The two reforms also had different timing the Swedish reform started when the Lithuanian was ending. Lithuania copied a traditional labour market training system from the Nordic countries (chiefly from Denmark), while they were starting to reform theirs. It must be noted that the chief aim of the case studies is to provide examples of assessing immediate, intermediate and ultimate impacts of reforms in adult learning sector. The two case studies are to be treated as specific examples of national reforms and their impact assessment, and their findings can not be generalised at the European level. However, decision makers in different countries might wish to consider the recommendations 14 This paragraph is based on feedback from Dr. Balázs Németh, a member of the Focus Group I for analyzing national reforms in adult learning and providing policy advice to DG EAC. 68

71 provided for both Lithuania and Sweden, when planning reforms which will affect adult learning Limitations Results of the detailed case studies should be interpreted with due consideration of the limits of the case study method. There are two main limitations of the detailed case studies presented in the next two chapters of the report. Firstly, detailed case studies used mainly secondary sources of data (literature and statistics). Sophisticated impact assessment designs (i.e. counterfactual analysis) could not be applied independently due to the late timing of impact assessment and limited resources. Secondly, efforts are made to trace the impact of education and training reform over a rather long period of time: 12 years in the Lithuanian case study and 7 years in the Swedish case study. On the one hand, this is an advantage, since all ultimate impacts of the reforms could be observed. On the other hand, this is a considerable limitation, since the number of intervening variables grows and the possibilities of taking them all into account decreases. A number of data collection gaps made many years ago are mostly impossible to rectify at this point in time Case study 1: The establishment of labour market training and counselling system in Lithuania ( ) General context and historical development of adult education in Lithuania Soviet rule General adult education During the period of the independent Republic of Lithuania ( ) the foundation of educational institutions were laid, adult literacy promoted and educational services developed. The main forms of adult education were evening courses for adults at primary schools, Sunday courses, adult schools and the so-called folk universities. Non-governmental organisations were especially active in promoting the training of specialists and improving their qualifications (in 1939 there were 789 such organisations). Charity and cultural societies in particular were interested in and provided financial support to adult education (The Education System in Lithuania, 2006/2007). In the period of Soviet rule ( ) the foundations of adult education system were taken over and transformed into an ideological structure. During this period a highly centralised and tightly controlled Soviet style education system, designed to service the needs of communist ideology and a centrally planned economy, was firmly established in Lithuania. It was afforded only limited flexibility to adopt unique education policies reflecting language and culture (OECD 2002). All adult education programmes were regulated by the government. They were completely government financed and were free to all adult learners. A system of adult secondary schools was also established during the Soviet period. Under that system, secondary education was compulsory. State enterprises sent their workers either to evening classes or to correspondence courses, characterised by significant face-to-face classroom attendance. In general, the learners were reluctant to attend these schools, and the standards of learning 69

72 were poor. A certificate gained at adult secondary schools was generally regarded as inferior to those gained at ordinary secondary schools at that time (OECD 2002, p. 167). The former Soviet adult education, at least from the eighties, degenerated to second hand quality education mainly for teenagers, who had no interest in attending school. The number of students in adult secondary (evening) schools in Lithuania dropped from in 1985 to in 1990 (Normantas). Vocational training Vocational counselling Non-formal adult education The vocational education and training system in Soviet Lithuania was closely tied to the command economy. An elaborate and centrally planned vocational training system consisted of initial training and continuing training. Most of the initial training was within the responsibility of the Ministry of Education, while a large network of vocational training schools was within the responsibility of the Ministry of Agriculture. Continuing training of specialists (construction, transport, industry, etc.) was within the responsibility of other line ministries, which owned specialised training centres (so-called combinates ). All employees were obliged to attend continuing vocational training in their respective combinates at regular periods, e.g. once in five years. 15 The legacy of Soviet vocational training in Lithuania was outdated curricula, worn-out equipment and training materials, deteriorating facilities and teachers who were ill-prepared for new professions and market economy combined to make vocational education and training largely irrelevant to the developing labour market (OECD, 2002). The system of vocational counselling existed in Soviet Lithuania, but it was targeted only at young people. The so-called vocational orientation rooms existed in all general schools. They were responsible for directing pupils into vocational schools. After the restoration of independence, the system was terminated. Vocational counselling for adults, in a sense that we understand it today, did not exist in the command economy. There was career planning, also centralised, when all graduates were directed to their employment places. Both formal and non-formal programmes made up the country's adult education system. They were provided in people s universities. Museums, libraries, clubs, theatres, and cultural centres were other providers of non-formal, non-school programmes for adults. Adult education programmes had many different purposes, including eliminating illiteracy, raising occupational skill levels, attaining a 100-percent secondary school completion rate, raising ideological and political levels, and encouraging cultural enrichment and constructive use of leisure time (Wells and Goetz 1987). The restructuring of the economy and the resulting unemployment required urgent development of vocational training for the emerging labour market. In 1992 Lithuania s Labour Market Training Service was established to coordinate vocational training for the labour market. The network of former training combinates and their branches, which existed in Lithuania till 1990, was transformed into the first labour market training centres for adults. 15 Interviews with B.Kindurienė and A.Vainienė. 70

73 The object of the analysis Training and counselling services The object of the analysis is the establishment and development of the labour market training and counselling system, as a separate sub-system of adult education in Lithuania. This initiative is considered as a major reform in the area of adult learning, since it featured deliberate and essential changes in service provision structures, policy priorities and the content of adult learning. The reform is framed in the period from 1992, when the Lithuanian Labour Market Training Authority (LMTA) was established, till 1997, when the new Law on Vocational Education and Training (VET) was adopted. By that time the main elements of the new sub-system of adult education were developed and functioning. The new law integrated them (at least formally) into a single VET system. Therefore, all the developments which took place after 1997 are considered as impacts of the reform Problems and needs Economic restructuring Needs of market economy The establishment of the vocational labour market training and counselling system was the result of economic transition from the centrally planned economy to the market economy in Lithuania in The collapse of the Soviet regime brought significant economic restructuring and, as a result, vast numbers of adults became unemployed for the first time in their lives. Officially there was no unemployment in the command economy: jobs, be they productive or not, were created for every person of working age. The emerging market economy and increased unemployment, however, called for vocational counselling services and retraining opportunities for the unemployed. Liberalisation and privatisation of the economy, economic transformation and related hardship made the services of the former centrally managed training centres irrelevant or unsustainable for the demands of the emerging market economy. Their founding ministries considered abolishing them, but the Lithuanian Ministry of Social Security and Labour (MoSSL) came forward with the proposal to take them over and use them as the basis for the new labour market training system Objectives The creation of vocational counselling and training services The establishment and development of the Lithuanian labour market training and counselling system was primarily regarded as a tool for the active labour market policy. 16 Its general aims included support for the unemployed through vocational training, counselling, guidance and analysis of the labour market needs. Therefore, new public services had to be created in order to satisfy the vocational training and counselling needs of adults, especially the unemployed. The training system had to be lean and marketoriented to satisfy the newly emerging and quickly changing needs of the economy undergoing radical transformation. It had to be able to train/retrain large numbers of adults and help them integrate/return to the labour market. The counselling system for adults had to be created from scratch. Adult training and counselling competences were scarce and major investment in 16 Interview with A. Tamulevičius. 71

74 human resources was required. New programmes had to be developed, and their implementation required major investment in the training and counselling infrastructure Target group Unemployed adults The major target groups of the reform were unemployed people, people facing unemployment and all other adult learners and their employers, willing to acquire labour market counselling and training services. The collapse of the state planned economy left many people unemployed, despite their qualifications and length of work experience. This target group was very sensitive. It was facing not only changes in their professional life, but also deteriorating social status and self-esteem, despair and frustration. A distinctive feature of participants in labour market training was their diversity in terms of previous experience, age, nationality, social background, etc. As the economy was undergoing a dramatic shift from industry and agriculture to services, from implementing top-down production plans to taking risk of independent entrepreneurial activity a vast share of the labour force had to reinvent its vocational profile The reform: inputs and activities This chapter reviews the steps which were taken in the period from 1992 till 1997 in order to build the labour market training and counselling system in Lithuania. They include activities in the areas of legislation, governance, financing, qualifications and quality New legislation The Law on the Support of the Unemployed The legal grounds for the involvement of the Ministry of Social Security and Labour (MoSSL) in vocational education and training rested in the Law on the Support of the Unemployed (1990). The involvement of the MoSSL in adult training activities was marked by the institutional turf fighting between the MoSSL and Ministry of Education and Science (MoES). The MoSSL has been criticised by the representatives of the MoES for taking over the activities of their jurisdiction. However, in the context of that time it was a positive decision, since the agenda of the MoES was overloaded with the reforms of general and high education. The existing system of initial vocational training was unable to respond to the increased demand in vocational training of unemployed adults, due to its rigidity and methodological differences. 17 Moreover, initially vocational training schools still had sufficient demand for the training of young people due to the favourable demographic situation and a limited choice of other vocational training possibilities, thus, adults were not seriously considered among their target groups at that time New public management institutions and a network of service providers 17 Interview with dr. V. Dienys. 72

75 Labour Exchange and Labour Market Training Authority Based on the Law on the Support of the Unemployed (1990) the MoSSL was authorised to develop and implement active labour market policy. For this purpose it established two key agencies: the Lithuanian Labour Exchange (LE) in 1990 and the Lithuanian Labour Market Training Authority (LMTA) in The LMTA comprised 7 regional authorities and 11 labour market training centres together with their branches in many town centres covering all of Lithuania. 18 Figure 14: The newly established administrative structure and cycle of labour market training and counselling The Ministry of Social Security and Labour Labour Exchange Labour Market Training Authority 46 territorial branches 7 regional branches 11 labour market training centres Information Counseling Training Unemployed adults Other service providers: vocational schools, private companies Labour market Source: compiled by the authors The foundations for services The service provision initially was organised as follows: first, the LE formed the groups of adults who needed vocational counselling or training. Then, the groups were directed to labour market training centres or other service providers (in case the LMTA centres did not have specific programmes or infrastructure) for training courses. Finally, the LE was looking for employment possibilities for the trained (re-trained) adults. This was the so-called closed model of labour market training. 19 During the reform period the foundations for labour market training and counselling services were laid. This included: 1. The development of vocational counselling programmes for adults, including psychological support; 2. The development of vocational training (re-training) programmes for adults, including different training modules; 3. The training of teachers and counsellors; 4. The development of training infrastructure; 5. The development of service management. 18 Source: 19 Interview with dr. T.Jovaiša. 73

76 Labour market needs and special needs of different adult groups Registers of programmes and providers The involvement of social partners As a result, a range of new labour market training programmes for adults, both formal and non-formal, were developed. The programmes were tied to the needs of the labour market, especially in the emerging service sector: hospitality, catering and tourism services, handicraft, dressmaking and beauty services. 20 The following jobs were needed by the labour market and provided in the labour market training centres: waiter, barman, baker, chef, confectioner, kitchen assistant, hotel receptionist, housekeeper and chambermaid, beautician, manicurist and pedicurist, hairdresser, dressmaker, seamstress, etc. Some of the programmes were adapted to the special needs of disadvantaged social groups: people with disabilities, ex-prisoners, etc. In 1994/1995 the Register of Studies and Training Programmes and the Register of Education, Science and Study Institutions were created. All vocational adult training programmes and service providers have been included into the national registers. The involvement of social partners in both initial vocational education and labour market training did not take place until Earlier due to the high unemployment level, the employers had no shortage of labour and did not immediately acknowledge the need for training or retraining their potential or current employees. As the economy stabilised and started growing employers organisations became involved in the validation of training programmes and the evaluation of graduates achievements. Moreover, the employers made use of the already existing labour market training centres to improve the qualifications of their employees New financing Employment Fund The initial financing for the newly established system of adult vocational training was appropriated from the Employment Fund through the Lithuanian Labour Exchange. The grants to public providers covered the costs of the initial development of infrastructure and programmes for training and counselling of unemployed people. Participants in the labour market training received a study grant in addition to the unemployment benefit Introduction of vocational counselling Focus on psychological support The development of vocational counselling for adults was a priority activity of LMTA. The main target group of this activity - adults experiencing unemployment - were offered new psychological support, counselling and learning opportunities. As a result, the LMTA became the largest organisation engaged in the development of labour market vocational training, information and counselling, with regional branches covering all the country (Euroguidance Lithuania, EU Leonardo da Vinci programme Coordination Support Foundation, 2005). In the reform period the LMTA developed career planning, social adaptation, personal and professional growth programmes. It adapted, standardised and implemented diagnostic methodologies, regularly upgraded counsellors qualifications and developed 20 Interview with A.Vainienė. 21 Interviews with dr.v.dienys and S.Baliukynas. 74

77 vocational counselling programmes for people with special needs. Regional branches of the LMTA employed psychologists who provided the following services: Individual counselling Group counselling Counselling for people with disabilities Counselling for employers and staff managers Counselling for adult training providers. The vocational counselling for adults aimed at recognising personal strengths, interests and capacities. It included career planning and provided information about available education and training. Special counselling programmes for employers and staff managers aimed at improving communication within organisations, stress and change management. Counselling for adult training providers aimed at ensuring the quality of adult training, both formal and informal. Personal outcomes of vocational counselling for adults comprise increased self-confidence, a clearer vision of the future and plans for further education or training. As a result of newly introduced individual or group counselling, the participants developed individual career plans, received information about learning possibilities and found out their own needs and aspirations The first steps in quality assurance The basis for quality inspection During the reform period the LMTA became responsible for the licensing of labour market training programmes and service providers. At that time it carried out a formal quality inspection rather than quality assurance. The inspection was based on legal acts which regulated the organisation of training process. It was carried out at the end of the programmes, so that participants could be interviewed, their final work assessed and exams leading to qualification monitored. The following aspects were included in the assessment: - The adequacy of material resources - The adequacy of human resources (the qualification of teachers) - The organisation of training process (theoretical and practical) - The organisation and implementation of final qualification exams. Evaluators assessed whether or not training infrastructure, the qualification of teachers, the organisation of the training process and the final exams corresponded to the provisions of a training programme and legislation Qualifications Since its establishment in 1992 the LMTA has focused on increasing flexibility and accessibility of vocational training for adults. 23 It developed labour market training programmes which corresponded to the initial vocational training programmes in vocational schools, 22 Interview with L.Kuginienė, the head of the career planning department at Vilnius LMTA. 23 Interview with dr. T.Jovaiša. 75

78 but were shortened to one year or half a year. The new labour market training programmes delivered vocational qualifications, which broadly corresponded to the vocational qualifications provided by the vocational schools. During the reform period the legal and institutional grounds were laid for the development of Lithuania s vocational qualifications and their system. 24 However, the development of qualifications per se was not an issue in this early reform External factors In parallel to labour market training and counselling reform, a number of parallel reforms were taking place in Lithuania. Participation of adults in learning could be directly affected by changes in the general education system for adults, while the reform of initial vocational training could affect learning opportunities mostly indirectly - through regulation targeting both initial vocational education and labour market training systems. The EU accession goal and financial assistance through the EU Phare programme were also influential factors of change both during the reform period and later General adult education reform Transformation of adult education The processes of political, economic and social transformations that started after 1990 had an impact on adult general education in Lithuania. In the period between 1991 and 1993 prerequisites for the development of adult education were created: units for adult education and the training of andragogues were set up in the MoES, the MoSSL and universities. The network of institutions providing adult general education was reformed. The General Concept of Education in Lithuania (1992) laid down the key provisions relating to lifelong learning and a general outline of adult education (The Education System in Lithuania, 2006/2007). In 1992 a Lithuanian Association of Adult Education (LAAE) was founded. It aimed at creating the initial basis at governmental and non-governmental levels for the reform of adult education in Lithuania. The Programme for Educational Reform for the period was adopted. It comprised actions in the fields of legislation, educational programmes, institutional restructuring, research and information (Normantas): The Concept of Adult Education was prepared in A package of documents influencing the provision of formal and non-formal adult education was issued in , while the Law on Non-formal Adult Education was only adopted in The number of adult general education schools has decreased from 89 in 1985 to 28 in They were reorganised into Adult Education Centres. Eight Adult Education Information centres were established in The state programme for teaching Lithuanian language to adults of non-lithuanian origin was developed and its supporting institution, the Lithuanian Language Teaching Centre, was established in Analysis of the Current State of Qualifications in Lithuania, LMTA,

79 Teacher training was a priority step in reforming adult education in Lithuania. Since 1992 the transfer of knowledge in adult education from the countries with long adult education traditions (Denmark, Sweden, Finland and Germany) has taken place through study visits, seminars and meetings. Afterwards, teacher training courses were organised in Lithuania to share the knowledge gained. New adult learning methods and materials were developed: new textbooks, foreign language courses, and distance learning facilities. The adult general education reform resulted in the establishment of a legislative basis for adult education in Lithuania, the adoption of strategic development plans and their harmonisation with the provision of EU regulations and the Commission Memorandum on lifelong learning. However, the major drawback of adult learning in Lithuania was (and still remains) the implementation of non-formal learning, especially, the absence of its validation (Teresevičienė et al., 2006). This was an obstacle in motivating adults to acquire new competences though non-formal learning Initial VET reform Initial VET reform In parallel to the establishment of the labour market training and counselling system under the MoSSL, the reform of initial vocational education and training (VET) was taking place under the MoES. Three stages of the reform were outlined (Tūtlys 2007): During the 1st stage ( ) a new vision of VET was created, VET policy and strategy were formulated and international cooperation was developed. Organisational restructuring of the vocational schools took place. During the 2nd stage ( ) the management of vocational schools was partly decentralized, the teaching and learning curriculum were renewed and social partnership developed. During the 3rd stage (starting in 1997) the main goal was to improve the quality of study programmes and to work out teaching/study standards. To ensure the quality of VET reform, vocational training research and methodological development institutions were established: the Vocational Training Methodology Centre (established in 1996) and the Centre for Vocational Education and Research (established in 1997 at Vytautas Magnus University) EU accession and the Phare programme The EU accession goal The Phare programme Reforms of Lithuania s education and training were stimulated by the goal of accession to the European Union (EU) and supported by foreign assistance. In June 1995 Lithuania signed a treaty with the European Communities and their member states, and in December 1995 it submitted the official membership application to the European Council. This was a pre-condition for international partnership development and financial assistance from the EU. The Phare programme was one of the three pre-accession instruments financed by the EU to assist the transition countries in their preparations for joining the EU. Originally created in 1989, it was reoriented to this purpose in 1993, following the Copenhagen 77

80 Council's invitation to Central and Eastern European countries to apply for membership. A Phare-funded Vocational Education and Training Reform Programme was launched in 1995 under the joint responsibility of MoES, MoSSL and the Ministry of Agriculture. The programme consisted of the following key components: Curriculum development: the development and implementation of new curricula for vocational training in nine selected employment fields, involving 27 vocational schools and labour market training centres across Lithuania; Teacher training: the training of teachers in modern educational and teaching methodology, and the training of managers in modern educational management; Partnerships with the EU vocational institutions: the establishment of working relationships with the European vocational schools and training centres; Up-grading of teaching equipment: the up-grading of learning materials and institutional infrastructure to suit the new curricula. Through the Phare programme Lithuania adopted the approach of contemporary European countries towards the development of adult education and VET. In order to ensure the coherence and sustainability of VET, the Phare programme included all constituent parts of it: initial, higher and labour market vocational training. This decision was based on the idea of building bridges between different sub-systems (Jotautienė, Janiūnaitė, Večkienė 1999). The Phare programme brought about a new methodological approach towards the reform. It was based on the initial piloting of ideas through experimental projects and their subsequent dissemination in a systematic way. The programme funded separate projects, which altogether had to bring about significant qualitative changes. Through the Phare programme Lithuania took over European practices and applied them in the local context. Pilot projects were implemented together with partner institutions in Austria, Denmark, Great Britain, the Netherlands and Finland. Thus, the Phare programme played a significant positive role in developing a conceptual and strategic basis for further VET reform and in supporting pilot projects in curriculum development and teacher training for both initial and labour market vocational training (Laužackas, Tūtlys, Spūdytė 2006). All the above reforms had favourable influences on the levels of adult learning. They either improved the quality (but not quantity) of adult general education or raised the standards of vocational training for adults. But the largest new learning opportunities and further institutional development were brought by the establishment of the labour market training and counselling system The impacts of the reform Although the creation of the labour market training and counselling system took about 5 years to accomplish, the impacts of this reform are still felt now. Tens of thousands of adults receive training and retraining services each year, while the institutional 78

81 legacy of the reform is shaping the new reform, which intends to merge the labour market training system with the initial vocational training system with the view to developing common standards and improving efficiency of provision. However, in this case study we cover the impacts of the reform for up to ten years after the reform Immediate impacts on participation Direct and indirect impact Labour market training centres trained over adults in the period The establishment and activity of the vocational labour market training system had a twofold effect on adult participation in vocational training. First, it had a direct impact on the participation of unemployed adults in vocational labour market training centres through the Labour Exchange. Second, it had an indirect impact on adult participation through the development of vocational training programmes for adults, which had been used by different service providers. The statistics available from the LMTA comprise only those adults who were trained in vocational labour market training centres. During the reform period about 6.3% of all unemployed adults were trained in the newly established labour market training centres. In the period between 1991 and 1997 out of the cumulative total unemployed adults adults participated in training provided by the labour market training centres. 25 The total number of adults trained by the labour market training centres during amounted to More than half of them paid for services individually or were sent by their employers (who covered the bill) rather than the labour exchange. The annual number of participants rose sharply between 1992 and 1995 and then remained almost stable until the end of reform (see the chart below). Figure 15: Adults trained in the labour market training centres in Lithuania in Source: Lithuanian Labour Market Training Authority 25 Data provided by B.Kindurienė, the MoSSL. 79

82 Participation was sensitive to economic situation A distinctive feature of labour market training is its close relationship with the country s economic situation. The volume of training decreases when unemployment levels fall, which could be observed in Lithuania in the period between 1995 and During economic recession unemployment levels rise providing more labour market training clients. However, more than half of the clients of Lithuanian labour market training centres were those, who paid for services themselves. In 1998 Lithuania suffered a huge economic recession due to the economic crisis in Russia the single most important trading partner at the time. The revenues of businesses went down as did the wages. Consequently, although the numbers of unemployed in labour market training increased (as far as the labour exchange could pay for their training), this was by far outweighed by the loss of private clients. With the revival of the economy in subsequent years, the demand and participation of adults in labour market training also grew Impacts on the individual level Labour market impacts The impact assessments of labour market training on the labour market situation of trained adults are available only from the periods after the reform. They are useful in understanding how training during and after the reform affected the lives of trained adults and increased their motivation for further learning. Since 2001 the social and economic effectiveness of vocational labour market training has been measured with the help of questionnaire-based surveys of participants. Data has been analysed by applying statistical methods. Advantages of the method: 1) low cost, 2) easy administration and 3) standardisation allows comparison and summary of results. Drawbacks of the method: 1) low return rate of the questionnaires (in this case about 20%) and 2) low validity of answers is possible. Employment Improved labour market situation The surveys of participants showed that the main personal outcome of labour market training for the unemployed was employment. The effectiveness measurement of 2001 indicated that in the 2-3 month period after participation in labour market training courses about 70% of the unemployed got a job (Lithuanian Labour Market Training Authority 2001). About 80% of those who completed labour market training courses were willing to engage in further learning activities, especially in qualification improvement courses. However, there is no data on actual participation of these adults in further learning. Surveys indicated positive changes in the labour market situation of the graduates of the labour market training programmes in 2003: 34% of them got a job, 5.7% got a better-paid job and 3.6% changed job (Labour and Social Research Institute 2004). About 55% of the trainees said that their economic situation 26 Interview with A.Vainienė. 80

83 improved after the training and 58% believed that positive changes were due to their participation in labour market training courses. The employment rate of the long-term unemployed (often suffering complex social and health problems) after the completion of labour market training courses is significantly lower than the overall employment rate of the participants. In 2003 it was only 16.5% in comparison with the overall rate of 60% (Labour and Social Research Institute 2004). The level of prior education and regional differences were among the main reasons for persistent unemployment. The employment rate of long-term unemployed with higher education qualifications was 8% in comparison with 33% employment rate of long-term unemployed with vocational education qualifications. The employment rate of long-term unemployed in urban areas was 14%, while in rural areas it was only 5%. Non-labour market impacts The analysis of non-labour market impacts of labour market training were often carried out in parallel to analysis of labour market impacts. Two methods were used to assess personal and social non-labour outcomes of vocational labour market training: 1. A questionnaire-based survey of participants (as described earlier) 2. Biographical method. This method was used in a Qualitative Study on Lifelong Learning and its Results (2007). In contrast to quantitative questionnaire-based survey, biographical method is context-specific and helps to establish the relationships between personal and social changes. The drawback of the method is the low level of the generalisation of results. The method was based on 8 semistructured biographical interviews with participants and 8 expert interviews with the teachers and managers of adult training institutions. Positive changes in personal and social lives In addition to the changes in the labour market situation, the graduates of labour market training programmes indicated positive changes in their personal and social lives: increased self-confidence (60% of respondents), increased self-esteem (55%), expanded social circle (52%), increased sense of social security (52%), personal development (41%), etc. (Labour and Social Research Institute 2004). Both long-and short-time unemployed indicated a high satisfaction rate with labour market training. Almost 70% of them said that the courses corresponded to their needs. In contrast to general education, labour market training is targeted towards and, therefore, results in instrumental and practical benefits for participants, such as the acquisition of practical vocational skills. Specific labour market needs were dominant in the motivation structure of the participants (Lithuanian Centre for Adult Education and Information 2007). Likewise, participants in non-formal education and training were most motivated by improved career perspectives (see the table 13 below). 81

84 Table 13: Participants in non-formal education and training by the reason of participation, per cent (in 2005) Reasons for participation % of respondents To do job better and improve career prospects 20.4 To be obliged to participate by the employer 6.9 To be less likely to lose job 8.2 To increase possibilities of getting a job, or changing a 4.6 job To start own business 0.9 To get knowledge/skills useful in my everyday life 11.1 To increase knowledge/skills on a subject that interests 13.3 To meet new people or just for fun 3.1 To obtain a certificate 10.9 Source: Statistics Lithuania Box 1: Reflections of the participants in labour market training about their motives for participation and expected outcomes I left the job and chose this vocation. < > When you learn on your own, you know one method, here you can learn more methods, how to do the same thing better, faster. < > I wanted this qualification. For the beginning, courses are enough and later I will see. Many friends work in this field. I know that construction companies do not employ you if you have no paper. I will see how it goes; maybe I will go to the university. (29 year old man) I have entered Vilnius College, but then left it, did not like the studies themselves. Management boring, and much more difficult than in the secondary [school]. Did not want to lose a year, so came here. Advertisements show that without a qualification you cannot find a job. I learn to be a tile layer. Not difficult. If you want to learn, then you will learn, as everywhere. With a diploma it will be easier to find a job. It is better to have a career rather than be an assistant, to carry sacks. I have plans for the future, to enter higher [education], also in construction. I want to be a builder, these courses is a step towards the qualification which I like. (22 year old man) I wanted to change something, worked for four years as a stockman, only with a secondary [education]. I was simply bored, needed changes. I took the course which they offered at the [labour] exchange. < > So far I like it, and the results are good, not difficult to learn. < > In the future, as long as it s profitable, I plan to work in construction. I think if you want, you will always find a job. (24 year old man) Source: Lithuanian Centre for Adult Education and Information 2007 The possibilities of employment, both in Lithuania and abroad, increased income and the prestige of the profession are the main factors of vocational attainment. The participation of adults in general education, on the contrary, is more motivated by the development of personal competences, the perspective of further learning and changes in social status (Lithuanian Centre for Adult Education and Information 2007) Impacts on the organisational level Increased supply of labour market training LMTA took an active part in providing methodological and financial support for the development of new labour market training programmes. Once developed, such programmes could be copied by any public and private providers free of charge. In 2000 the labour market training system was opened to competition allowing any organisation to be licensed for labour market training provision. As a result, the number of vocational labour market training providers gradually increased (see the table 14 below). 82

85 Table 14: The number of assessed labour market training programmes and providers ( ) Year Number of training providers Number of training programmes Source: the External Quality Assessment of Lithuanian Labour Market Vocational Training 2007 Institutional turf fighting and provider battle The reform was deemed by many stakeholders as having longlasting negative impacts. The establishment of the Lithuanian labour market training and counselling system has been criticised for duplicating the initial VET system and its infrastructure. 27 Both the reform and its outcomes were marked by institutional turf fighting at the Ministerial level and among service providers. Box 2: The provider battle In a number of vocational schools in Lithuania visited by the OECD review team in 2002, serious resentment was expressed about the role taken by the labour market training centres. In one vocational school, a labour market centre was operating upstairs in the same building and running the same classes as the vocational school below, duplicating the considerable teaching resources and equipment needed. Vocational schools felt cheated of the opportunity to earn extra revenue from continuing vocational classes for adults. In one school, this situation was described as [the labour market training centres] stealing our bread. Although a tendering process existed to award this work, vocational schools and colleges were rarely successful in the bidding process and, rightly or wrongly, do not regard the competition as being played on a level playing field. This contest involving two Ministries was regarded as unproductive and extremely wasteful of resources. Source: OECD The criticism, however, has to be taken with caution, having in mind the different specialisations and capacities of vocational schools and labour market training centres. It has been argued that LMTA centres were better equipped for working with adults than vocational schools, due to the shorter training programmes, greater flexibility and closer partnership with employers. 28 Most of the labour market training centres however contracted teachers from vocational schools. Due to the lack of infrastructure, the centres often used the facilities of employers and initial vocational schools. The reform resulted in the development of new training and counselling services, primarily to the unemployed and those at risk of becoming unemployed. The outcome of the reform was the increase in supply and variety of vocational adult training services and closer cooperation with social partners. Changes in financing In 2000 the financing of labour market training was reformed. Direct subsidies from the state were replaced by public tenders through the Labour Exchange (Ministry of social security and 27 Interviews with R.Totoraitis and V.Dienys. 28 Interviews with A.Tamulevičius and A.Vainienė. 83

86 labour, Republic of Lithuania 2000). The newly established system opened the way for other service providers (private training companies, initial vocational schools, NGOs, etc.) to participate in labour market training. This could be considered a logical step in implementing the Law on Vocational Education and Training of 1997, which laid the common foundation for both initial vocational education and labour market training. It has been argued however that tendering procedures after the financing reform 2000 remained more favourable to the entrenched labour market training centres than other service providers. Still, although the centres received about 60-70% of their income from the LE through the tenders, they were forced to live more or less under the market conditions. For individual service providers this meant pressure to look for additional financial resources, mainly through projects funded by EU programmes and the development of training/retraining services for employed adults. Box 3: The example of impact of financing reform on individual service providers When direct subsidies from the state were replaced by public tenders, there was a challenge to find other ways of funding. Thus, project activities were stimulated and partnership with foreign training institutions and social partners was initiated. The centre was looking for new target groups who needed vocational training. Thus, the programmes for people with disabilities were developed and implemented. Cooperation with employers resulted in the organisation of training courses for their employees, especially those working in the hospitality sector. Since 1996 Vilnius Žirmūnai Labour Market Training Centre has participated in 9 PHARE projects; since 1999 it has participated in 10 Leonardo da Vinci mobility projects; since 1997 it has participated in 7 international cooperation projects and since 2006 it has participated in 5 projects co-financed by the ESF. Examples of the projects include: distance learning for adults with disabilities, vocational training for unemployed women, the use of ICT methods for the training of long-term unemployed, mobility project for employees in the hospitality sector, quality management of vocational training, competence building for teachers, the development of qualification systems, to mention but a few. Working under the market conditions posed regular pressure for the institution, but at the same time it made life much more interesting. It increased the work load for staff, but at the same time it gave them a chance to implement new ideas and be rewarded by success. Source: interview with A.Vainienė. The introduction of public tendering brought full financial selfsustainability of labour market training centres, improved capacities to live under market conditions and the ability to adapt to changes. On the systemic level it led to more efficient allocation of resources in the broader vocational training system. Changes in vocational counselling Participation in international programmes Once established, the LMTA has developed cooperation with different professionals and social partners both in Lithuania and abroad. Psychologists who worked as counsellors in regional branches of the LMTA kept regular contacts with specialists from labour exchange, municipalities, trade unions, employers associations, adult training institutions, NGOs and other organisations. During the reform period regional branches of LMTA became wellestablished and, therefore, were able to participate in international projects. Since 1998, when Lithuania joined the Socrates and 84

87 Leonardo da Vinci programmes of the European Commission, LMTA centres have taken part in it. Box 4: Example of participation in international projects Vilnius branch of the Labour Market Training and Counselling Authority participated in the following international projects: - Leonardo da Vinci programme project Self-Help Group Model of Job Seeking for Students and Graduates ( ); - Socrates programme project Give a Word to Adult Learners ( ); - Leonardo da Vinci programme project Competencies and Skills in Tourism ( ); - Leonardo da Vinci programme project Train&Compete - VET Trainers' Competencies through LifeLong Learning ( ). Source: Vilnius branch of the LMTA, Focus on adults from socially disadvantaged groups One of the impacts of the labour market training reform on adult learning has been the involvement of disadvantaged social groups. Being established under the MoSSL, the labour market training and counselling system has been focused on the social inclusion of people with disabilities, prisoners and ex-prisoners, Roma people, etc. Regional branches of the LMTA have developed a number of counselling activities targeted at adults from these groups. First, they have developed special psychological counselling programmes. Second, they have provided individual and group counselling for the target groups. Finally, they have prepared methodological recommendations for the specialists within training institutions. Box 5: Example of counselling for adults from socially disadvantaged groups in Vilnius branch of the LMTA Vilnius branch of the LMTA participates in the following projects (co-funded by the ESF) for the integration of socially disadvantaged groups: 1) the development of social and professional skills of ex-prisoners; 2) the integration of people with disabilities into the labour market; 3) the development of employment capacities of people from national minorities; and 4) vocational counselling of drop-outs. Source: Vilnius branch of the LMTA, The analysis of the LMTA activity results shows that the participation of people with disabilities in vocational training, and especially counselling, has been increasing gradually in the long term after the reform. Table 15: Number of people with disabilities participating in vocational counselling activities Year Number of participants Source: LMTA. Lack of coordination Despite positive developments in vocational counselling for adults in Lithuania, the major drawback from the very beginning of the reform has been a lack of coordination. 29 The problem was that counselling alone could not solve the educational and employment problems of the target groups. To be effective, it had to be closely linked with the personal and labour market needs of the participants, and offer the most appropriate training possibilities. This was not done and, as a result, adults after the completion of 29 Interview with dr. T.Jovaiša. 85

88 vocational training programmes used to come back to the labour exchange and start the cycle of counselling and training all over again. Moreover, during the reform period a great variety of institutions which provided counselling services both for adults and young people were established. These were regional branches of LMTA and Lithuanian Labour Exchange, career centres at universities and schools, private companies and others. Their activity was not well coordinated, so vocational counselling lacked consistency and coherence. Career guidance information system To ensure the accessibility of career information and counselling services, Lithuania has joined a single European information system, comprising EURES (Search for Jobs in Europe) and PLOTEUS (Training Opportunities in Europe). The MoES has initiated the creation of open vocational information, though the counselling and guidance system AIKOS. This system gives access to information on education, statistics and labour registers, as well as giving relevant information to counselling service providers (Euroguidance Lithuania, EU Leonardo da Vinci programme Coordination Support Foundation 2005). To sum up, reform in the field of vocational counselling has led to the creation and development of new vocational adult counselling services for different target groups. Individual and group counselling methods were elaborated, however, due to the lack of coordination, they were not always effective. Quality assessment After the reform the systematic development of quality assessment of labour market training took place. Three stages of this development were distinguished: preparation (since 1997), implementation (since 2000) and improvement (since 2004) (Kavaliauskienė and Žibėnienė 2007). Box 6: Example of changes in the quality of labour market vocational training service providers In Vilnius region there are about 60 organisations which have a license to provide labour market training services. The changes in the quality of services are obvious, especially after 2001, when a more systematic quality assurance was developed and carried out. Before that the quality supervision was rather formal. The training services were also poorly developed: due to the lack of a systematic approach towards labour market training, methodological basis and qualified teachers, the service providers were working helter-skelter. There is a very great diversity in service providers. For example, vocational training schools and colleges did not need a license for adult training, only a compliance note. After 2000 a lot of small private companies entered the market due to the new tendering procedure. Vocational adult training was their secondary business, so they often subcontracted to other providers. Most of the big companies were also licensed to provide adult training, for example Lithuanian Gas, Lithuanian Post, Lithuanian Railways and all major companies in sewing and energy sectors. They were mostly training their own employees. Source: interview with G.Kaušinienė, the head of the quality assurance department in Vilnius LMTA Focus on general skills Another outcome of quality assurance was the renewal of the training curriculum with a focus on general skills, lifelong learning, the promotion of employability and other competition aspects of human resources. General subjects, such as foreign languages, ICT 86

89 skills, communication and others were included in labour market training programmes. Quality inspection developed by the LMTA during the reform period served as the basis for further development of a systematic quality assessment of labour market training in line with the common quality assurance framework for VET in Europe. The contribution of the EU-funded programmes in this area was significant Impacts on the policy level New legislation The end of the reform was marked by the adoption of the new Law on Vocational Education and Training (VET) in It legitimised the labour market training established during the reform as a subsystem of VET and laid the foundations for a single modern VET in Lithuania (OECD 2002, p. 151). The outcomes of the reform directly fed into the key provisions of the law: VET is defined as 1) initial VET and 2) labour market VET. It is provided by two main types of institutions: 1) vocational schools for VET and general education and 2) labour market training centres only for VET. The system is managed by two institutions: 1) initial VET is within the responsibility of MoES and 2) labour market VET is within the responsibility of MoSSL. The MoES is responsible for the licensing of both types of VET institutions. The involvement of social partners is provided: they shall represent one third of each examination commission and take part in defining the requirements for programmes. VET institutions may create new programmes and engage in commercial activity, provided that the income is used for educational and training purposes. The new law was followed by the White Paper on Vocational Education and Training in It proposed many changes to the VET system, including the involvement of initial VET institutions in the provision of vocational qualifications to adults through evening and distance education courses. The legislative outcome of the reform was the involvement of MoSSL into vocational adult education and training, the establishment of two sub-systems of VET (initial and labour market) within the equal responsibility of two ministries (MoES and MoSSL) and the involvement of initial VET institutions into adult education. Introduction of quality assurance Cooperation between MoES and MoSSL After the adoption of the new Law on VET in 1997 the responsibilities for quality assurance were shared between the MoES (Methodological Centre for Vocational Education and Training) and the MoSSL (Labour Market Training and Consulting Authority). The MoSSL was responsible for the expertise of the programmes and service providers, while the MoES was responsible for the issuing of licenses. After the adoption of amendments to the new VET law in 2007, all responsibility for licensing was transferred to the MoES. The development of a common quality promotion system for both initial and continuous vocational education and training was the result of the project Development and implementation of a common quality assurance system in vocational education and training ( ), which was implemented by the MoES and 87

90 MoSSL and co-financed by the ESF. The project resulted in a model for a quality assurance system for vocational education and training in Lithuania, which was developed in line with the European Common Quality Assurance Framework in VET. Moreover, methodologies for internal and external quality assessment were created and evaluators were trained. A common strategy for vocational counselling The first step towards the development of a coherent counselling system in Lithuania was a common strategy for vocational guidance adopted by the MoES and the MoSSL in The strategy was the result of the European Commission Twinning programme within Phare The draft of the strategy was developed during the project Vocational Training Aimed at Economic and Social Cohesion, which was implemented within the framework of the programme. The task force for the draft strategy included the representatives of MoES and MoSSL, vocational counselling professionals and the EU consultant (Ministry of education and science, Ministry of social security and labour, Republic of Lithuania 2003). In the action plan for the implementation of the vocational counselling strategy (2004) a significant role was assigned to the LMTA. Among other responsibilities, it has been authorised to provide vocational counselling services to all inhabitants in Lithuania, to develop professional standards for vocational counsellors and to develop general regulations for vocational information centres (Ministry of Education and Science, Ministry of Social Security and Labour, Republic of Lithuania 2004). MoES was responsible for the establishment of vocational counselling and orientation centres for young people at general and vocational schools. However, in reality nothing has changed (at least within the LMTA) after the adoption of a single strategy. 30 For example, Vilnius branch of the LMTA was carrying out its daily counselling activities as usual. Moreover, it has been contracted by general schools to provide vocational counselling services for young people. This has been due to the fact that psychologists at general schools were overloaded with other activities, thus, vocational counselling functions have been transferred to the LMTA branch. Introduction of validation of prior learning The first legal provisions for validation of prior learning were adopted after the reform, and it is difficult to trace whether or not early initiatives in this field fed into national policies. The principle of formal recognition of qualifications, irrespective of how they were acquired was established in the White Paper on VET (1998). The Law on Non-formal Adult Education (1998) established the right for non-formal education participants to receive the assessment of knowledge acquired in the system of non-formal adult education and a state recognised certificate. In legislation defining the procedures for the recognition of knowledge and skills were adopted. Finally, a new edition of the Law on Education (2003) set out the key elements of formal certification of competencies acquired through non-formal or informal learning. However, despite the established legal basis, the actual validation 30 Interview with L.Kuginienė, head of the career planning department at Vilnius LMTA. 88

91 of prior learning is not taking place, and this is considered to be a major obstacle for flexibility in adult learning. 31 The development of VET standards Contribution of EU funds National qualifications framework The development of national VET standards, which are the basis for validation of prior learning, started in The process, however, was slow mainly due to financial problems (ECOTEC 2007). The other reason for the slow development of VET standards was poor coordination between the two vocational education and training systems: initial and labour market. 32 Finally, a further obstacle was the content of vocational training programmes. Being based on the example of academic university programmes, VET programmes explicitly defined only theoretical knowledge, while practical skills - the main outcome of vocational training - were not defined. A significant contribution to the development of national VET standards was made by EU funds. Building on the results of the "Framework of Qualifications Standards" project under the Phare 2001 Economic and Social Cohesion Programme, the Methodological Centre for Vocational Education and Training was to develop 100 VET standards under the successive project "Development of the System of VET Standards" ( ), which was co-financed by the ESF (ECOTEC 2007). The project served as the methodological basis for the establishment of the Qualifications Authority by the Government of Lithuania at the end of The legal basis for this was provided by the amendment to the Law on VET introducing the National Qualifications System in In 2009, however, the Government approved new amendments to the Law on VET by which the Qualifications Authority is to be abolished and its functions are to be delegated to the MoES. The process of introducing validation of prior learning and the development of a national qualifications system in Lithuania was long and slow. The direct impact of the reform on national policies in this field could not be traced. The contribution of the EU has played a far more significant role here. It can be concluded from the evidence collected in this report that the reform resulted in the development of institutional capacity of the MoSSL, LMTA and individual organisations to manage, finance, deliver and support adult learning. The system has been reformed and gradually improved on a continuing basis after 1997, which has also triggered development and adaptation in the wider education and training system. Involvement of employers The partnership with employers organisations has been strengthened since At that time the partnership agreement between LMTA and Kaunas Chamber of Commerce, Industry and Crafts was signed. It aimed at increasing the role of social partners in the preparation of formal and non-formal training programmes and their validation. Since 2003 the Association of Lithuanian Chambers of Commerce, Industry and Crafts has been engaged in the examination of graduates and the validation of qualifications. 31 Interview with V.Dienys. 32 Interview with V.Dienys. 89

92 This has been regarded as a positive step towards the improvement of training quality. 33 Employers attitude towards involvement in the process of vocational adult training was influenced by surplus labour supply as well as lack of traditions and experience in organising training for employees. Workers with inadequate qualifications would be dismissed and new higher-skilled workers would be employed without any investments into employees training and qualification improvement. In 2004 Lithuania again started facing a deficit of qualified labour force, and employers paid more and more attention to keeping and training their employees. Though legislation does provide for the participation of the social partners in the vocational training process, this participation is insufficiently significant or effective in practice Long-term impacts on participation of adults in learning Increase in adult learning There is no previous research on the impact of the labour market training and counselling reform on the overall levels of adult learning in Lithuania. Although a number of parallel reforms were taking place during the same period, the labour market training and counselling reform had the greatest and mostly positive influence on the dynamics of adult learning in Lithuania. In the long-term after the reform general adult participation in learning was gradually increasing (see tables below). Table 16: Participation of adults (25-64) in lifelong learning (percentage of adult population) Year % 3,9 2,8 3,6 3,3 4,6 6,6 6,3 5 6,1 5,1 Source: Statistics Lithuania Table 17: Participation of adults in formal and non-formal learning in Lithuania (percentage of adult population) Year % 2,7 1,7 2,3 2,1 2,8 5 5,6 4,4 4,8 Source: EUROSTAT Labour Force Survey (EDUC4WN) No reliable data on participation in non-formal learning in Lithuania Data on adult participation in non-formal learning is available only for the years 2005 and In 2005 only 7.8% of all adults in Lithuania participated in non-formal learning. This was one of the lowest results, in comparison with other EU member states (EUROSTAT, 2005). The other survey shows that in 2006 about 30% of the adult population was engaged in non-formal learning (Department of Statistics, Lithuania, 2006). Such huge differences in data arise due to the limited access to information about nonformal learning and different understanding of what constitutes such learning. The Lithuanian government did not financially support non-formal adult education and, therefore, it had no access to information about it. 35 The reform and subsequent liberalisation of labour market training brought three types of labour market training programmes: 1) formal programmes registered in the MoES register; 2) non-formal programmes registered in the MoES 33 Interview with A.Vainienė. 34 Inga Blaziene, Institute of Labour and Social Research, Lithuania: collective bargaining and continuous vocational training, Interview with dr.v.dienys. 90

93 register; and 3) non-registered programmes, provided by private enterprises. Non-registered programmes were the most numerous; however, data on participation in them was not available. 36 Some data on the participation of employed adults in vocational training is available for the years 1999 and 2005 (see the table 18 below). In that period both a number of private companies providing continuing vocational training and the number of employees participating in such training increased. Table 18: The involvement of private sector employees in adult learning (1999 and 2005) Indicator Enterprises providing any type of continuing vocational training 42.8% 46.4% Employees participating in continuing 9.8% 14.6% vocational training courses by share of employees in all enterprises Hours spent in continuing vocational training courses by share of hours 41.4 h 32.3 h Costs of continuing vocational training courses by share of direct costs per participant Source: Statistics Lithuania LTL LTL Barriers to adult participation in learning Previous research shows that the main obstacle to participation in learning indicated by the respondents is the lack of money for learning (Teresevičienė et al. 2006). Other reasons for nonparticipation were the lack of time, a big work load, baby-sitting and the care for other dependent persons, uncompleted basic education, the lack of information about where to learn, no learning opportunities in the neighbourhood and health problems. Other research showed that the absence of the need to learn was the major barrier to adult learning, followed by inappropriate health or age (see the table 19 below). Table 19: Obstacles to adult participation in formal or nonformal education and training by the type of obstacle (in 2005) Obstacles to participation Per cent of respondents There was no need 36.4 Did not have proper skills 4.5 Education/training was too expensive 15.6 Lack of employer s support 3.7 Difficult to combine work and training 17.6 No time because of family responsibilities 13.3 Training venue was too far 6.4 Not confident with the idea of going back to school 12.8 Health or age was not appropriate 24.3 Other reasons 6.8 Source: Statistics Lithuania Little motivation of long-term unemployed Another survey of the long-term unemployed showed that an absolute majority of them tended to link their unemployment with the unfavourable economic situation rather than a lack of professional competence. Only 1.2% of the long-term unemployed who had been employed before indicated insufficient professional competence as the reason for their dismissal (Labour and Social Research Institute 2001). The long-term unemployed had the 36 Interview with dr.v.dienys. 91

94 lowest motivation to learn and increase their professional competence: only 7% of them had registered at the labour exchange with the aim of acquiring or changing their professional qualification, 37% said they were ready to improve their professional qualification and 38% said they were ready to change it. About a quarter of respondents expressed no opinion about their professional and learning needs, which might be an indication of low perception of the role of professional competence-raising in the labour market. To sum up, the long-term unemployed gave priority to any employment rather than targeted learning Conclusions and recommendations Relevance of the analytical framework Major achievements The case study was based on desk research and interviews with experts. Application of other methods of data collection was limited by the length of time since completion of the reform. Furthermore, the establishment of the labour market training and counselling system as well as parallel reforms received only very limited attention in previous research and thus availability of data from secondary sources was also limited. Nevertheless the case study method proved to be useful in identifying and analysing the impacts of reform on adult learning, but also on vocational training more generally. The case study revealed and illustrated positive and negative as well as immediate and long term impacts. The reform was successful in achieving its objective: it resulted in the establishment of a labour market training and counselling system with adult learning programmes and infrastructure across the country. The following long-term benefits for the target group were achieved: Unemployed adults have a choice of formal and non-formal vocational training opportunities. Adults from socially disadvantaged groups, such as long-term unemployed, people with disabilities, ex-prisoners, ethnic minorities, etc. have a choice of vocational training and counselling programmes adapted to their needs. Participation in labour market training and counselling increases employment opportunities and self-esteem of the participants. The reform also had a number of significant intermediate impacts: The acquisition of nationwide competence of vocational training for adults; The development of a flexible and self-sustainable nationwide system of labour market training centres; The development of a nationwide vocational counselling system adapted to the needs of different target groups; The continuing capacity building of professional adult trainers and counsellors; The foundations for the development of a quality assurance system. Success factors The reform achieved its aims due to the following factors: Active political leadership in the Ministry of Social Security and Labour; Availability of public financing both for investment into infrastructure and procurement of training services; Availability of examples of the functioning of similar systems in foreign countries and access to their experience; 92

95 Motivation of the service providers to participate in international projects for the development of labour market training; Significant support from the EU Phare programme. Drawbacks The establishment of the labour market training and counselling system in Lithuania was carried out in parallel to the reforms in initial vocational education and training, and adult general education. The reforms resulted in the establishment of two parallel structures of vocational education and training with separate training programmes, staff and infrastructure. The decision to establish a sub-system of VET for adults was positive in the context of that time. However, immediately after the new system started functioning it was marked by institutional turf fighting, which could be considered as an important obstacle for further development of both adult learning and initial VET in Lithuania. The conflict between the sub-systems of education is characteristic not only to Lithuania. It has been observed in most OECD countries and highlights the dilemma of policy coordination between the different sub-systems of adult learning (OECD 2001b). The drawbacks occurred due to: Legacy of the former system and weakness of all state institutions and policy making capacity. Depleted resources due to multiple problems of transition -the whole public management system was undergoing a sweeping transformation and learning process. Emerging and unfamiliar problem of unemployment led to focus on serving the relatively short-term labour market needs. Lack of cooperation and coordination between public administration stakeholders. Weak bottom-up input from social partners and education providers. Recommendatio ns Based on the evidence collected in the case study, the following recommendations can be made: Recommendation 1: In the planning phase of the reform explicit targets should be established, so that the actual outcomes of the reform can be compared with them in order to assess the success of the reform. Recommendation 2: Before the commencement of the reform the involvement and coordination of stakeholders, including policy-and decision-makers, should be ensured, so that the effects on in-fighting among actors on the progress of the reform can be prevented. Strong political leadership at Government level should be ensured to steer the reform. Recommendation 3: Sufficient funds should be allocated, so that the progress of the reform will not be hindered by the lack of resources or motivation. Recommendation 4: The collection of monitoring data and independent evaluation should be ensured from the beginning of the reform, so that different aspects of the reform outcomes can be assessed and policy learning can take place. 93

96 5.3. Case study 2: The Adult Education Initiative in Sweden ( ) General context and historical development of adult education in Sweden Long adult learning traditions Established municipal adult education Education for immigrants Education for adults with learning disabilities Non-formal and voluntary adult education and its infrastructure When the Adult Education Initiative (AEI, Kunskapslyftet) was introduced in 1997, Sweden already had a long and wide-ranging tradition of adult education, as well as high participation rates. Adult education in Sweden has comprised municipal adult education, commercial training companies, folk high schools and in-service training (Bostrom, Boudard, Siminou 2001). This means that adults were used to studying both in formal and non-formal settings, so the AEI fitted well with their educational habits. A cornerstone for the implementation of the AEI in Sweden was the existence of an already well-functioning municipal adult education system (komvux). The government's decision was that the reform would be embedded in the existing institutional framework without any dramatic changes. Municipal adult education (komvux) was introduced in 1968 for the benefit of adults lacking the equivalent of basic or upper secondary schooling. It comprises basic adult education, upper secondary adult education and post-secondary training courses. The 288 municipalities at that time in Sweden receive a general state grant for education, which they disburse to the schools according to local decisions. 37 The Swedish for Immigrants (SFI) programme is the oldest single measure for the linguistic education of adult refugees and other immigrants in Sweden. Education in Swedish as a second language for adult immigrants has been carried on from the mid-60's and under different educational forms. SFI provides immigrants with a knowledge of the Swedish language and Swedish society in order to facilitate their integration into society and entry into the labour market. Municipalities have an obligation to offer it to newly arrived adult immigrants. Programmes are organised differently from municipality to municipality. The AEI however did not have a special focus on the needs of immigrants. Municipalities in Sweden have an established network of education for adults with learning disabilities (särvux), which aims at improving students level of independence by combining practical and theoretical studies. This special school form follows the same curriculum as municipal upper secondary for adults and regular upper secondary school, but uses specially-adapted syllabi and timetables. The programmes offered can therefore lead to skills in individual subjects, as well as equivalence to compulsory school or upper secondary vocational training for adults with learning disabilities. Before the introduction of the AEI, Sweden had an established system of non-formal and voluntary adult education, with the supporting infrastructure, called Folkbildning. Folkbildning is a collective name for the activities conducted by the country s folk high schools and study associations in the form of courses, study circles and cultural activities. The first folk high schools in Sweden 37 Based on: and Bostrom, Boudard, Siminou

97 were established in 1868 and today there are 148 schools in the country (105 of the schools are run by popular movements, organisations and associations, while the remaining 43 are run by county councils or regions). Folk high schools are open to all adults over the age of 18 with no special entrance requirements, except for some more specialised courses. All tuition is free of charge and national student aid is available. Folk high schools provide various types of courses, both general and special. General courses can qualify the students for university studies, which means that the folk high schools serve as a bridge between the non-formal and the formal systems. The 148 folk high schools thus constitute an established non-formal parallel adult learning pathway in the Swedish educational system and was therefore a natural part of the AEI. 38 Continuing vocational training (CVT) Long study support tradition Older people s organisations and intergenerational programmes Decentralization and goal orientation Providers of continuing vocational training (CVT) are public institutions (e.g. Komvux, folk high schools), trade unions, employers associations, private enterprises and training companies. Vocational training can be conducted in several different ways. Employees can attend formal (municipal or state provided) or non-formal (provided by companies or training organisations) vocational training. A large part of the vocational training is performed within the company or bought in from external educational companies. Sweden has no coordinated system of CVT neither on national, local or sectoral level. The administration of the CVT is mostly decided by the individual employer and employees, and can hence be conducted in the way they find appropriate (Brunk and Wahman 2009). The Swedish study support also has a long tradition and is an important part of the Swedish welfare system. Part of the stated goal of the study support system is to reduce the impact of social, financial and geographical background by making it easier for disabled and older people to study. All adult students can apply, regardless of their education levels, and can attend full time or part time. The upper age limit for students to receive support is 54 years. Study support also applies as part of the compensation taken into account when calculating a worker s pension benefit (Regeringskansliet, 2009). Sweden has a long history of older people s organisations and intergenerational learning. The first older people s organisations were found at the end of the 1930 s and the beginning of the 1940 s. Currently there are two big associations of organisations for retired people (Swedish National Pensioners' Organisation (PRO) with about members and the Association of Retired Persons in Sweden (SPF) with about members) and 890 small local organisations all over Sweden. The PRO runs a folk high school of its own, where senior citizen can go for courses from all of Sweden. Almost 50% of those over the retirement age are members of older people s organisations. A well-structured organisational network was one of the reasons for the quick expansion of intergenerational programmes and other adult learning initiatives in Sweden the end of the 1990 s (Hatton-Yeo and Ohsako 2000). 38 Sources: and interview with Björn Garefelt, National Council of Adult Education. 95

98 Before the introduction of the AEI, the Swedish education system had undergone a fundamental shift towards the decentralization of responsibility and decision-making powers. In 1991 a full mandate for all teaching staff and undivided responsibility for organising and implementing school activities were transferred from the central authorities to the municipalities and their local school authorities. The principles of school management by objectives with fewer regulations and clearer goals were also laid down. According to the new division of responsibilities within the Swedish education system, the parliament and government control educational activities by defining national goals and guidelines for education. The central and local education authorities together with various education providers are responsible for ensuring that education is organised in compliance with national goals. Within such a framework, local authorities and education providers, on the one hand, have considerable freedom to decide upon the distribution of resources and implementation of activities. On the other hand, they are required to systematically follow up and evaluate educational activities in relation to the set goals. The consequence of decentralization for individual learners has been increased diversity of education and more possibilities to choose between different types of schools and study routes (Bostrom, Boudard, Siminou 2001, p. 60). Childcare and the care of the elderly Women in Sweden have a high employment (and education) ratio due to the well-developed childcare and care of the elderly system, the right to parental leave and financial security during study. Local authorities are obliged to supply pre-school activities or care for schoolchildren taking into account the parents employment or study needs. Elderly care has been extensively developed since 1970s and, consequently, very few elderly people live with their children (Bostrom, Boudard, Siminou 2001, p. 60). The Adult Education Initiative was built on a long-standing solid basis of adult education and its support traditions, an established municipal adult education and a variety of educational paths. The decentralised responsibility and decision-making, strong social networks and a well-developed social infrastructure were favourable conditions for the reform The object of the analysis The Adult Education Initiative is analysed as a major adult education reform in Sweden. The initiative can be defined as a reform, since it carried essential changes in adult education, characterized by significant breadth, depth, and intensity. The Adult Education Initiative (AEI) The AEI was a nation-wide five year programme launched in 1997 and lasted for five years until It comprised both formal and non-formal adult learning sectors. The AEI was the largest ever investment in adult education in Sweden with annually 100,000 extra places in municipal AE and 10,000 extra at folk high schools. It was intended as a complement to traditional labour market training by offering the unemployed education at compulsory or upper secondary levels (Skolverket 1999). 96

99 Problems and needs Increased unemployment and inefficiency of labour market policy Long-term unemployment and low educational levels The AEI emerged largely as the result of high unemployment rates which increased following an economic recession in Sweden in the early 1990s. Between 1990 and 1993, the percentage of officially unemployed rose from 1.7 to 8.2% of the total labour force. In the Swedish context, it was the highest recorded unemployment rate since the depression of the 1930s (Stenberg and Westerlund 2006). Although the number of participants in active labour market programmes remained high, the declared unemployment did not decrease. This gave rise to increasing concerns about the efficiency of traditional labour market policy measures. Two population groups especially suffered from the economic crisis: the long-term unemployed and those with relatively low levels of education. In Sweden in the early 1990s, the relative labour market situation of these groups was influenced by reforms in regular youth education, whereby the two-year programmes in upper secondary school were extended to three years (eligibility to university demands a three-year upper secondary diploma) (Stenberg and Westerlund 2006). Thus, the circumstances of the time provided politically persuasive arguments in favour of additional public expenditure on adult education. By introducing the AEI, the Swedish government offered not only a supplementary alternative to traditional labour market programmes, but also a way of bridging the educational gap of the group with a two-year upper secondary school from the previous system of youth education Objectives Raise educational levels and reduce unemployment The primary goal of the AEI was to reduce unemployment by improving the confidence and employment possibilities of those in weak positions in the labour market (particularly low qualified) and encouraging them to pursue further studies. The overriding aims of the reform were to: Reduce unemployment; Develop adult education; Reduce educational differences in the labour market; Promote economic growth. Exchange of good practice The idea was that evaluation and the exchange of ideas and experience between municipalities would improve the overall quality of adult education by highlighting good ideas and spreading them further. The AEI was to contribute to the reform of adult education both in terms of content and working methods. Over the five year period, a new, reformed system of adult education had to be developed, so that it would be better adapted to the needs of individuals, the labour market and society in general in the future Target group The unemployed without uppersecondary education The AEI was primarily aimed at the unemployed adults (aged 25-55) who lacked the full three-year upper-secondary qualifications, i.e. had not completed secondary level with passes in all subjects or did not have the required knowledge in one or more subjects. A considerable share of the Swedish labour force had a two year upper secondary education as the highest educational attainment 97

100 level, which was usually not enough for applying to higher education. So, the AEI was targeted at removing this obstacle for further learning and thus reducing educational differences in the labour market. The idea was that those adults most in need of education should be given a chance to catch up and add to their stock of knowledge. Improved levels of skill and increased self-confidence would help them to strengthen their position in the labour market. As access to the qualified labour increased, the opportunities for economic growth would increase accordingly. The employed were also eligible to the AEI Hybrid nature of the AEI Employed individuals with low levels of education were also eligible for participation in the AEI, under the condition that their employers filled their vacancies during the study period with a long term unemployed individual. It was particularly important for those working in sectors that are sensitive to economic and structural changes, e.g. health care and manufacturing industry. Thus, the AEI may be considered as a new form of labour market measure as the tradition had been to offer subsidised employment, vocational training, relief work or other forms of work related training. The AEI had a hybrid nature, combining labour market programmes with comprehensive schooling. The AEI emerged as a consequence of economic recession and high unemployment in Sweden. Therefore, its primary aim was to reduce unemployment by increasing the educational level of lowskilled unemployed people. This aim however could be reached only by reforming adult education so that it could better meet the needs of adults and society at large Inputs and activities Financing Reallocation of funds Annual contribution of about 345 million euro Financing the AEI was a political, as much as a technical issue, indicating a shift in the government position on the role and benefits of education and training, the responsibilities of various stakeholders and governance arrangements. The initiative was financed from four different sources: the Ministry of Finance, the Ministry of Labour, the Ministry of Education and Science and the Ministry of Energy. In 1993 a green energy taxation system was introduced in Sweden. A new carbon dioxide (CO2) tax considerably increased total energy taxation, while the revenue generated was used to decrease taxes on labour. Consequently, the Ministry of Labour received additional funds. 39 The AEI was mainly financed by reallocating funds earmarked for passive measures in the Ministry of Labour s chief policy area to active education and training programmes in the Ministry of Education and Science s main area of operation (OECD, 2001b). The AEI was the government s programme, funded solely from its budget. The government invested an average annual contribution of 345 million euro (3 billion SEK) in the programme, which was equivalent to full time study places. Every year the government set aside a fixed amount of money to cover the cost of 39 Interview with Åsa Sohlman 98

101 adult education in municipalities. Extra money was paid for vocational courses, as they are usually more expensive than basic courses. Financing took the form of state grants, while each municipality was individually responsible for the organisation, planning and implementation of training activities. In order to ensure that the planned level of education was undertaken, the government had decided that a municipality must repay transfers if it had educated fewer people than it was supposed to. Municipalities which completed more than their planned share of adult education received extra transfers (Tangerås, 1999, p.15). All the funding for the AEI was from national sources. Financial support Various forms of study assistance were available to those studying under the AEI: The special grant for education and training (UBS) for adults who were unemployed or in work and had not completed their education. The grant was available on condition that the person in question was aged 25-55, was studying at elementary or upper secondary level and fulfilled the criteria for the receipt of unemployment benefit. The grant was equivalent to this benefit and could be paid for a maximum of one year instead of the unemployment benefit. In a case of a person in work, the employer had to fill the vacancy created with an unemployed person. Special adult study assistance (SVUX) was intended for people who did not have full education and who took leave of absence from their job to study at elementary or upper secondary level. The support took the form of a grant and a loan. Those eligible were people who had been in work for at least four years and whose maximum age was 50. Special adult study assistance for the unemployed (SVUXA) was available to those aged who were unemployed and registered job-seekers who intended to study at elementary or upper secondary level, on condition that they had previously been in work for at least three years. The assistance took the form of a grant and a loan. Those above the age of 50 in certain cases could be eligible for SVUX or SVUXA as well. Study funding in the form of a grant and a loan. This was primarily intended for students in higher education, but also for those adults who were studying at elementary of upper secondary level. A new form of study assistance introduced by the AEI, namely UBS, was more generous than the usual student grant and it was paid instead of the unemployment benefit Governance Preparation for the AEI Involvement and cooperation Prior to the launch of the initiative, a "Delegation for the Adult Education Initiative" was set up in the Government Office with representatives from the Ministries of Education, Labour Market, Finance, Industry and Trade as well as the Ministry of Public Administration and a secretariat was attached to the delegation. The Secretariat was responsible for the dialogue with municipalities during the planning of the AEI and its implementation during the first year. According to the report by the Swedish Agency for Administrative Development, which analysed the model for implementing the AEI chosen by the Government, the AEI had in a short time led to a 99

102 widespread involvement, new thinking and development work on adult education issues. Co-operation between municipalities and other adult education co-ordinators increased. Co-operation with the employment office and the trade union organisations on the whole was satisfactory. Co-operation with local industry, however, was not so successful. Management by results Increased responsibility of municipalities National Agency for Education Project-based implementation A distinctive feature of the AEI was management by results. This has been an innovative approach to adult learning. The government had set very clear national targets: to reduce unemployment and promote adult learning. The reform was targeted at second-chance learners, so there was a very clear message from the central government not to fail them. This meant that the needs of individual learners were placed at the centre of the reform; therefore, the traditional education system had to change in order to meet these needs. The AEI was a joint responsibility of the state and the municipalities, based on the idea of decentralization and a principal-agent relationship. 40 The government set national targets, gave generous funding and monitored the initiative, while actors at the municipal level were given freedom to develop adult learning activities in the local context. Each municipality was responsible for its own local organisation, planning and implementation of adult training activities. By means of the state grant the municipalities had to attain national goals as regards the orientation, scope and quality of adult education. They could choose between organising the education themselves, or in conjunction with other municipalities or by purchasing the services of other providers. Municipalities also had the freedom to decide on the management of the AEI: whether to organise it on the basis of already existing institutions or to establish new management bodies. The initiative was managed by a special national delegation with representatives of several ministries and an operative secretariat, the National Agency for Education (Skolverket). General duties of the agency included development, follow-ups, evaluation and supervision in pre-school, school and adult education. In addition, the agency was commissioned to manage and monitor the initiative in the municipalities and to support and encourage their efforts to develop the initiative. All municipal activities within the AEI were carried out on a project basis. Each municipality had to write a project proposal based on its planned adult education activities and the number of places and apply for the state funding. Applications had to be renewed every year with an indication of achieved results. The project proposals (applications) were submitted to the National Agency for Education, which decided on how the funds were to be allocated. The distribution of the state grants for the AEI was based on the following criteria: The level of unemployment and education in each municipality; The aims and scope of the municipality s programme in terms of orientation courses, general courses and vocational courses; 40 Principal-agent relationship is the arrangement that exists when one person or entity (called the agent, in this case local municipalities) acts on behalf of another (called the principal, in this case the national government). 100

103 The municipality s initiatives for the renewal and development of adult education. Supervision of initial projects Partnership and involvement Regular monitoring and evaluation Political supervision and independent evaluation After the first year of the AEI, the National Agency for Education supervised the AEI activities in Linköping municipality. Based on the supervision, recommendations for the municipalities were formulated. Both positive and negative aspects of the initial experience were highlighted. The recommendations were disseminated to all municipalities in Sweden, so they could learn and clearly see the expectations posed by the AEI. 41 This supervision carried out by the National Agency meant a lot for the further development of the AEI both for in Linköping and on the national level. A lot of people working with the AEI projects referred to that supervision in their own development. For Linköping, it gave them the courage to continue the work because the National Agency s report showed that the municipality was on the right path. It gave them self-confidence and reliability in the contacts with partners: companies, labour market organisations, and in contacts with local politicians the Kunskapslyftet gained a high standing position. 42 The involvement and partnership between different actors in the provision of adult education was directly and indirectly promoted by the AEI. First, there was a requirement from the state that all municipalities which applied for the AEI state grant had to submit a plan for cooperation with the employment offices, trade unions and private companies. Thus, new forms of cooperation have been developed, e.g. through the boards of national employment offices, adult education councils and other networks. Second, the municipalities were given freedom to choose between different providers of adult education, both formal and non-formal, both inside and outside the municipality. Thus, a number of new actors were involved in the provision of adult learning, e.g. libraries, social services, immigration administrators and the business community. The evaluation of the AEI was within the responsibility of both the government and the municipalities. During the implementation of the initiative each municipality had to do a self-evaluation and submit a report to the National Agency for Education, while the agency regularly monitored, followed up and evaluated all municipal adult education. After the AEI was finished, universities were engaged in the evaluation of the initiative. The government supported academic studies in the field of the AEI. The Commission for the AEI was appointed by Parliament with the tasks of proposing goals for the national adult education initiative, assessing education and in-service training needs of adults, making recommendations on the division of the responsibility and financing of education for adults and considering measures which would stimulate pedagogical development in adult education. The Commission on the AEI was authorised by the Government to monitor and co-ordinate the national evaluation of the AEI, as well as to develop independent national evaluations of the initiative (OECD 2000). 41 Interview with Eva Nordlund, Director of Education, National Agency for Education 42 Interview with Örjan Albhin, Linköping municipality 101

104 Information activities Communication channels The introduction of the AEI was announced during a campaign in May In communicating information on the AEI the National Agency for Education has used a number of channels, including the following (Bostrom, Boudard, Siminou 2001): a) A website which contained press releases, the agency s newsletters and subjects of topical interest. The AEI internet site ( was created to serve as a meeting place for project manager, teachers, students, employers and other stakeholders of adult education. The site provided information and answers to the questions and was continually updated with news and reports. b) A computer network for schools, the content of which included a link store, meeting places, a notice board, school addresses and a dictionary; c) Dialogue with municipal authorities, by maintaining continual contact with project managers and contact persons. The aim of the dialogue was to contribute to the development of adult education by disseminating examples of good practice at municipal level. d) A newsletter on school information, news and on-going projects, including the AEI. The newsletter was sent out to school administration units of the municipal authorities, school governing boards, school managers and head teachers. e) Field units at 11 locations around the country which dealt with enquiries and provided information. At each field unit there was a contact person for questions on the AEI. f) Invitations to conferences and seminars were held with the aim of disseminating information and networking. Box 7: Information and outreach activities in Linköping Municipality In the beginning of the AEI, Linköping Municipality employed one person to be responsible for information and outreach activities for the AEI: meetings with local actors (trade unions, employers, education providers, etc.) and actual search for adults throughout the city. He had no office, but a bicycle and a telephone. At first, some advertisements in local newspapers were published. At the same time the advertisement on national TV was released. Information about adult learning possibilities in Linköping was collected and a catalogue was published. Trade unions helped to spread the information to their members (in the first year information was especially targeted at the unemployed). Then, public meetings with adults who wanted to study were organised. The first meeting was overcrowded, so more meetings had to be held. Meetings were also held in the workplaces, especially at the companies which intended to reduce their staff, e.g. at Ericsson, which reduced its staff by 1000 workers in Linköping. The start of the AEI was challenging, but very interesting. The combination of state funding and local freedom was very motivating for people at local level to implement new ideas in the field of adult education. Source: interview with Örjan Albhin, adult counsellor in Linköping Municipality National knowledge week In order to enhance the understanding of the needs of adult education and lifelong learning from both individual and society perspectives, the Ministry of Education proposed and supported a national knowledge week (Kunskapsveckan) in During the week municipalities and other providers of adult education 102

105 presented their courses and programmes for the forthcoming year. The activities of the week were targeted at those most in need of learning through debates, exhibitions, reports in newspapers, TV and radio, campaigns in the streets and shopping centres and special announcements aimed at the unemployed. All actors in most of the 230 participating municipalities were coordinated by the AEI coordinators. Incentives for participation The AEI was primarily focused on adults with low level of education and long-term unemployment, who are known to have weak motivation to study. Therefore, active recruitment measures have been encouraged in order to reach and motivate this group. This has been done most effectively by people within the same social context and with the necessary legitimacy, usually trade unions. The economic benefits of learning have been highlighted and more generous study grants were offered. The choice of individualised and flexible forms of learning both in formal and non-formal educational institutions was developed. 43 Box 8 Active recruitment activities in Malmö Municipality Malmö in the south of Sweden is the third largest city with almost inhabitants. One fourth of them have immigrant background. In order to reach the right target group for the AEI, an Infotek was established in the centre of Malmö and run as a joint project of the municipality, the employment office and the University College of Malmö. The key functions of the Infotek were recruitment and counselling. The AEI also purchased visiting services from the Swedish Trade Union Confederation, Swedish Industrial Workers Union and the Commercial Employees Union. The trade union organisations had a number of full-time employees working on recruitment activities for the AEI. The municipality had a large proportion of inhabitants receiving social assistance. In order to motivate people to study, the municipality decided that they could retain their social assistance in addition to study grants and also get textbooks and other learning material cost-free. Source: Bostrom, Boudard, Siminou 2001 (adapted from OECD 1998) Service provision Different implementation Different forms of adult learning Municipalities chose different organisational forms for implementing the initiative (Wass, 1999). Special emphasis was given to the relation between the AEI and the local employment office. Employment offices in different municipalities had different views on what kind of education the AEI should focus on and how the education should be related to the labour market training provided by the local employment offices (OECD 2000). The reform encompassed upper secondary adult education. This education could take a number of different forms: various types of education and training organisers were involved, the aim being to provide flexible solutions where possible. In order to offer a wide variety of courses, municipalities and county authorities, national schools for adults, adult educational associations, folk high schools and training companies were all working together. The traditional general courses offered by the municipal adult education organisations (komvux) were supplemented by a large number of vocational courses. Part of the education could be spent in work placements, and it was also possible to combine upper 43 Interview with Björn Garefelt, National Council of Adult Education 103

106 secondary courses with studies at an elementary level, or with one of the programmes organised by the National Labour Market Board for the unemployed. Special orientation courses were provided for those who felt uncertain about the level of studies which would suit them. Cooperation between municipalities Attention to local and regional needs Cooperation with folk high schools Counselling Focus on the individual Guidance and study plans With very clear objectives set by the state government, each municipality had to find the best way to achieve them. Therefore, municipalities were looking for ways to cooperate and link their resources in order to achieve common goals of adult education. For example, there were agreements between municipalities that the home municipality of an adult learner would pay for the studies in a neighbouring municipality if a certain study programme was not available in the home municipality (the so-called intermunicipal fees were possible). This way, municipalities utilized their educational potential and sectoral strengths (e.g. industry or social care) without duplication and avoided surplus investment in new programme development. Another area of cooperation between municipalities was the analysis of education and training needs in the region. Municipalities conducted common analysis of labour market needs and provided regular guidance about learning opportunities and further employment possibilities. The experience of municipalities shows that adults did not mind moving to other municipalities to study or work. Such cooperation between municipalities has developed into a regular practice. There was cooperation between some folk high schools and study associations when performing certain activities within the AEI. Folk high schools and study associations also cooperated with municipalities on a contractual basis providing services for certain target groups financed by the municipality. Cooperation also often took place between folkbildning organisations and municipalities when it came to e.g. counselling and information services. Common counselling of adults was a prominent form of cooperation between institutions. Employment offices together with the providers of adult education assessed the needs of individual learners and advised on whether a person should get short and specific labour market training or the AEI training to continue to higher education. The intention was for all education which took place under the AEI to be governed in form and content by the needs, wishes and capacities of the individual. The education had to help to improve the individual s opportunities in the labour market and provide an option for further studies. It had to be demand-led, in terms of organisation, planning and the actual courses provided. Every individual had to have ample scope for personal choice in terms of the type of study, its timing and location. There was an emphasis on guidance during the AEI. Qualified guidance both before and during studies was intended to ensure that everyone had an equal chance to make well-founded choices. Guidance resulted in an individual study plan which specified what adults hoped to achieve in their studies and the ways in which their aims could be achieved. The plan was drawn up according to the adult s previous education and on the assessments which were made regarding other skills and knowledge. 104

107 Quality requirements Quality requirements The government had put forward two quality requirements for the municipal authorities with regard to the implementation of the AEI: (i) education shall be tailored to meet each individual's demand; and (ii) education shall increase the participants' employability or provide skills required for higher-level education (Tangerås 1999, p.14). It also had taken two precautions to ensure that quality standards were maintained. First, education had to meet minimum requirements defined by the law. 44 Second, the government had allocated funds for local projects aimed at quality improvement. 45 The implementation of the AEI was marked by political leadership, explicit national targets, generous financial resources, local responsibility and regular supervision and monitoring of results. The promoted partnership and local freedom allowed multiple actors to influence the impacts of the reform. Such sharing of responsibility (or policy ownership) is a pre-condition for the sustainability of the reform results External factors External factors are parallel initiatives of the government in the field of adult education, which might have influenced the participation of adults in education and training. Although their scope was not as big as the AEI, they help to see a broader picture of adult education in Sweden. Gender equality measures Emphasis on ICT learning In line with the AEI, the Swedish government has undertaken a number of measures targeted at the integration of gender perspective into all levels of labour market policy. For the budget years from 1997 to 1999 the National Labour Market Administration (AMS) used up to SEK 30 million (EUR ) each year for project activities with the aim of breaking gender segregated labour market and training its personnel in equal opportunities issues (Bostrom, Boudard, Siminou 2001, p. 52). Since 1998 the AMS has given priority to special development areas, mainstreaming gender in counselling, technology (a guidance programme in technology for women) and the training of more men as schoolteachers. The AEI had also included among its targets an increase in the percentage of women in education and vocational training by providing more guidance. In 1997 the IT Commission of the Swedish Government initiated a programme called SeniorNet. It is an association for people aged over 55 years who wish to acquire and improve ICT-use skills. The two major organisations for retired people (PRO and SPF) collaborated on the programme and a bank sponsored it. The programme aimed at enabling 25%, i.e. more than a million older adults to be active internet-users by the end of the year Another objective of the programme was to build bridges between generations. SeniorNet still operates in cooperation with 44 Education had to satisfy the standards prescribed by Skollagen (SFS 1985:1100), ch.11 and Förordning (SFS 1992:403) om kommunal vuxenutbildning; or Lagen (SFS 1991:1108) om statens skolor för vuxna and Förordning (SFS 1992:601) om statens skolor för vuxna. 45 However, in 1999 no such projects were undertaken (Tangerås 1999, data from Statskontoret (1998) op. cit., p. 16). 105

108 different stakeholders, such as e-commerce companies and banks. 46 As a follow up to the UN international Year of Older Persons in 1999, the Swedish Government started the Elderly project to work with a number of issues, including lifelong learning and contacts between generations. The project funded 10 intergenerational programmes, three of which were IT projects. It also partially funds the SeniorNet. Distance learning Increase in the retirement age In 1998 the Swedish government decided to launch an initiative for the development of new methods of distance learning for adults, based on ICT (Bostrom, Boudard, Siminou 2001, p. 59). A task force was set up to launch projects for the development of new methods and support for flexible learning. In 1999 the Swedish Agency for Distance Education (DISTUM) was set up to support and promote the development and application of ICT based distance education. The agency s activity covers universities and colleges, adult educational associations and folk high schools throughout the country. It allocates funds, follows up, evaluates and serves as a network-based information centre and exchanges with other international bodies dealing with IT-based distance education projects (Bostrom, Boudard, Siminou 2001). However, some studies indicate that most adults prefer education in a social setting with other students and with direct contact with the teacher (Turner, Toft, Witte 2008). In March 2002, however, the Ministry of Education and Science decided to close the agency. In May 2001, the Swedish parliament added a new rule to the Employment Protection Act, which gives all employees the right, but not the obligation, to remain in employment until the age of 67. Several studies have suggested that education is associated with later retirement from the labour market. However, the results of a study based on Swedish register data which encompasses and respectively indicate that adult education has no effect on the timing of the retirement from the labour force (Swedish Institute for Social Research 2008). The underlying view is that inadequate education, particularly among the low skilled, is an important reason for decisions to retire early (rather than re-educate) as it increases the risk of becoming unemployed and decreases the chances of receiving job offers. The reverse relationship, i.e. the impact of increased retirement age on adult education has not been studied. The efforts of the Swedish government to promote adult learning were not limited to the AEI. Although the specific influence of each initiative can hardly be measured, these initiatives show a comprehensive attitude of the government and society at large towards adult education The alternative to the AEI Labour market training Before the introduction of the AEI, labour market training (LMT) was the biggest adult training programme in Sweden. It started on a small scale just after the Second World War and increased considerably during the 1958 recession. In the 1960s adult education became a central part of Swedish education and labour 46 For more information about the activities of SeniorNet see 106

109 market policies, as a result of the active labour market policy development in the 1950s (Rubenson 1992). Since the second half of the 60s labour market training activities have been expanding rapidly with growing numbers of participants. LMT had a very broad supply of programmes which covered most branches on the labour market, especially those with a high demand for workers. In 1992 there were more than participants in LMT, representing about two per cent of the total labour force (Stenberg 2003a, p.3, 2003b, p. 4). A counterfactual to the AEI The AEI, in contrast to LMT, offered a chance for adults to go one step up in education In academic studies LMT was regarded as a counterfactual to the AEI due to their similarities in enrolment procedures and their target groups (Stenberg, 2003a). Both programmes offered financial support in the form of a training grant equal to the size of the unemployment insurance (UI) and the completion of both programmes entitled their participants to an additional period of unemployment benefits. For many individuals, LMT would have been the alternative, if the AEI had not existed. However, a significant difference between the programmes was that the AEI had involved comprehensive education (mainly theoretical), whereas LMT has been orientated towards a given profession (mainly vocational training). In the period of high unemployment the choice of the Swedish government to allocate additional public funds for adult education, as a complement to the traditional labour market training, confirms its two-fold aim: to improve both the labour market position and educational attainment of the target group. This comes as a result of the previous reforms in regular youth education in early 1990s, whereby the two-year programmes in upper secondary school were extended to three years (eligibility to university demands a threeyear upper secondary diploma) (Stenberg and Westerlund, 2006). Thus, a year of studies within the AEI not only offered an alternative to traditional labour market programmes, but also bridged the educational gap for the group with a two-year secondary school from the previous system of youth education. The differences between the AEI and LMT reveal a far-reaching aim of the reform: to promote general adult learning, not only to achieve short-term labour market benefits. The results of both for the individual recipient were compared in the evaluation studies of the AEI and LMT impacts Impacts of the reform Immediate and long-term impacts on participation Immediate participation impacts were traced through the analysis of statistical data The data used in the evaluation study of the AEI impacts was collected from different registers which have been merged by Statistics Sweden. First, all individuals that had been registered in adult education at komvux, sometime during the autumn semester of 1997, were identified. Second, the participants in the AEI were identified by using information on the special grant for education and training (UBS) which was offered by the AEI. A participant in the AEI was thus defined as an individual who had 107

110 been registered in adult education and had received the special grant UBS during the same semester. Of those offered UBS there were eight per cent who enrolled in studies at folk high schools. They were excluded from the study. Due to the lack of individual records of drop out, the notion programme participation actually refers to those who started rather than completed a programme. The data set also included information on age, gender, income, education, citizenship, place of residence, civil status and family situation (Stenberg 2003a). Increase in municipal adult education The AEI covered a major part of the adult population in Sweden and had a considerable economic, social and individual impact. The immediate and the most obvious impact of the AEI was increased adult participation in municipal adult education (komvux). After the introduction of the AEI in 1997, total enrolment in komvux, both full-time and part-time, increased substantially. It peaked in 1998 and remained high during 1999 (see figure 16 below). The number of pupils nearly doubled and the number of course participants more than tripled during the late 1990s and early 2000s. Two years after the introduction of the AEI the number of participants started to decrease continually until the programme ended in A possible reason for the decrease was a steady improvement of the labour market situation (Stenberg 2003b). Figure 16: The number of individuals enrolled in komvux during Source: Stenberg 2003b The main beneficiaries As a result of the AEI, around people raised their educational level to an average level corresponding to one year of upper secondary level. The two major groups of beneficiaries were adults who had not completed full time education and the longterm unemployed. Adults with a two year secondary school diploma (who were a major target group of the AEI) constituted more than 60 % of the participants in the AEI. However, there are indications that many of those with an upper two year secondary school diploma did not complete the third year while enrolled in the AEI (Stenberg 2003d, p.19). Among those educated through the AEI % were of foreign origin Interview with Karin Nilsson, Director of Education, the National Agency for Education 108

111 In the period between 1997 and 1999 participants in the AEI represented 1.2 to 1.5% of the total labour force and exceeded those in the Labour Market Training, which since the 1950s had been the largest labour market programme in Sweden (see figure 17 below). Figure 17: Unemployment, labour market programmes and adult education in Sweden (percentages of labour force) Source: Stenberg and Westerlund 2006 Note: Adult education pertains to the number of individuals enrolled in at least one course, which includes a large fraction attending only short courses. Unemployment and labour market programmes are measured as yearly averages. Differences in participation A comparative study over the period analysed the differences between municipalities concerning the scope of the AEI. The proportion of the population participating in the AEI varied greatly between municipalities. There were no obvious explanations for these variations. One evaluation project (performed by Inregia AB in 2000) focusing on socio-economic aspects of the AEI showed the following (OECD 2000): The unemployment rate in municipalities explained only a small part of the participation differences. The proportion of immigrants among the unemployed had no influence on participation in the AEI. Impact at regional level Long-term participation Explicit aims of the AEI were to promote the prospects of people with a weak position in the labour market and to reduce educational differences. Evidence suggests that individuals in municipalities with overall lower education levels tended to join the AEI and thus reduce regional differences in education (Stenberg 2003a). EUROSTAT Labour Force Survey shows that adult participation level in Sweden after 2002 has been regularly increasing. 109

112 Figure 18: Participation of adults in formal and non-formal learning in Sweden ( ) Source: EUROSTAT Labour Force Survey (EDUC4WN) Figure 19: Participation of adults in formal learning in Sweden ( ) , , , , , ,79 Source: EUROSTAT Labour Force Survey (COURATT) Non-formal adult learning The AEI promoted adult participation in non-formal learning by supporting new student places annually. For the folk high schools this meant an increase of 50 % in relation to the existing organisation. There was cooperation between some folk high schools and some study associations when performing certain activities within the AEI. Folk high schools and study associations could also cooperate with municipalities on a contractual basis performing certain services for certain target groups financed by the municipality. Cooperation also often took place between folkbildning organisations and municipalities when it came to e.g. counselling and information services Interview with Björn Garefelt, National Council of Adult Education. 110

113 Figure 20: Participation of adults in non-formal learning in Sweden ( ) , , , , , ,216 Source: EUROSTAT Labour Force Survey (COURATT) The analysis of statistical data shows that during the reform period ( ) adult education in Sweden increased considerably. Although facing a slight decline in 2002, the participation of adults in both formal and non-formal learning in Sweden was increasing gradually Impacts on the individual level Motivation and satisfaction This section is mainly built on the evidence from previous research which analysed the impact of the AEI on individuals. Two studies performed by Göteborg university in 1999 and 2000 showed the following (OECD 2000): The AEI in particular attracted those adults who have not always succeeded so well earlier in life and school, but who felt a strong desire to start studying. They also developed greater confidence in their own ability to succeed in their studies. Participants often came from the health care area and were usually women. In general, men comprised 32,9%, while women made 67,1% of the AEI participants (Stenberg, 2003b). The AEI had not been so successful in recruiting from amongst the long-term unemployed. Participants were satisfied with their studies and the benefits they derived from them. Particularly satisfied were those who could start a "study career" and continue their studies after a year. The least satisfied were those who went back to being unemployed after studying. Results from the study on individual development under the AEI (performed by Lander R. and Larsson M. in 2000) show that the majority of the adults (who started learning in autumn 1997) were very motivated to start studying under the initiative. Around 60% were very eager to begin and a further 20% were quite eager to begin studying. Those studying in programmes organised by the municipality were more motivated in the beginning, in comparison with those studying at a folk high school. In addition, an absolute majority of students (80%) were very satisfied or quite satisfied with the education they received under the AEI (OECD 2000). 111

114 The evaluation of personal labour market impacts of the AEI relative to Labour Market Training (LMT) was carried out by Anders Stenberg (Stenberg 2003b, 2009). - Paper I studied unemployment incidence and unemployment duration for participants in the AEI relative to LMT. The method used was an instrumental variable adaptation of the symmetrically trimmed least squares estimator. A bivariate probit model was used in order to study unemployment incidence. The results indicate a beneficial effect of the AEI relative to LMT on unemployment incidence (74,9% and 89,7% respectively), but longer duration in unemployment among the AEI participants. People enrolled in LMT programmes were still included in the registers of employment agencies, but not as immediately available for work. Participants in the AEI were no longer registered as unemployed and received a special education grant (UBS) instead of unemployment benefit. - Paper II used annual wage earnings in 1999 to compare the AEI and LMT for individuals who do not continue in education following programme completion. Two separate estimation methods were used, the classical selection model and the method of matching on the propensity score. The effects of the AEI were very heterogeneous. First, only those completing at least one year of studies contributed with positive effects. Second, long term unemployed (Stenberg and Westerlund 2008 in Labour Economics) showed quite large benefits from AEI, which is in line with Stenberg 2009, as those with small prior earnings could increase both number of hours worked and wage levels. The results of both methods indicate that wage earnings of the AEI participants were lower than those of LMT participants. However, it must be stressed that the study did not include those AEI participants who went on to further education and focused only on short-term perspective, therefore, results could be different over a longer period of time. - Paper III used employed data for those enrolled in 1997, as well as 1998 to study the annual wage earnings of 1999 and 2000 and data on attachment to branches of employment before and after the programme. Data on branches of employment indicate less mobility among the AEI participants. This is largely due to a stronger attachment to the public service sector. The analysis of wage earnings of the sample enrolled in 1997, indicate that the effects on wage earnings tended to be more advantageous for the AEI in 2000 rather than 1999, possibly implying a lag in the effects of the programme. Labour market impacts In 2005 another paper evaluated the effects of participation in the AEI on individual labour market impacts, notably employment and annual income, as well as on the labour market equilibrium (Albrecht, Van den Berg, and Vroman 2005). Fixed effect methods allowing for treatment effect heterogeneity were applied for the effects at the individual level. The data was based on a number of matched longitudinal administrative data sets covering the full population of Sweden. For the equilibrium effects, an equilibrium search model with heterogeneous worker skills was applied. This model was calibrated using pre-programme observations. 112

115 Wage earnings Evidence indicates a more negative effect of the AEI on wage earnings relative to LMT, especially for those participants who were born in foreign countries. This may indicate that the AEI was not a successful labour market measure for those who had cultural barriers to overcome (Stenberg, 2003c, p.17). On this evidence, vocational training seems to be a more efficient tool for foreign born and foreign citizens than theoretical schooling. About 25% of the participants in the AEI, who continued in further education, did not experience short-term labour market benefits, such as increased earnings, in contrast to those who moved right to employment after the AEI. However, the evaluation of long-term labour market effects of the AEI might present different impacts. The AEI had different effects relative to LMT between genders, with the AEI being more beneficial for women than men (Stenberg, 2003b, p.15). This has been explained by Martin and Grubb (2001) who reported that formal classroom training helps women to a greater extent than men. The study by Axelsson (1996) also provided observations that men have less confidence in confronting theoretical studies than women do, so they are more likely to choose LMT instead of the AEI (Stenberg 2003d, p. 18). Branch mobility Results on mobility indicate that the AEI participants had a lower probability of changing branch of employment and a relatively stronger attachment to the public service sector (including health), in comparison with LMT. The AEI attracted more participants from the public sector than the LMT, and their share was the biggest among all participants in the AEI (see the table 20 below). Table 20: The fractions of various characteristics among participants in the AEI and LMT Enrolled 1997 Enrolled 1998 % of AEI % of LMT % of AEI % of LMT Manufacturing 9,1 10,2 8,6 12,0 Construction 4,0 5,1 2,7 3,7 Retail 15,0 14,6 17,3 18,7 Private service 6,6 7,2 7,2 9,2 Public service 34,2 24,8 34,6 25,3 No attachment 31,1 38,1 29,6 31,1 Source: based on Stenberg 2003d, p.9 As employment after the AEI generally increased, there is a decrease in the employment of no branch after the reform and an increase in the employment on all branches. The largest changes occurred in the private sector which in the 1997 sample increased its share by about 70%. There is also a large increase in the employment in manufacturing by 50% (see the table 21 below). Table 21: Distribution across branches of employment before and after the AEI Before After Before After Before After No branch Manufacturing Construction Retail Private service Public service Source: Stenberg 2003d, p.11 The AEI participants tended to return to the public service sector to a greater extent than LMT participants, while a relatively higher 113

116 percentage of LMT participants returned to manufacturing branches (Stenberg 2003d, p. 12). Table 22: Percentage of individuals who return to a particular branch of employment AEI LMT AEI LMT Manufacturing 41,0 48,6 40,0 50,5 Construction 41,5 46,3 48,7 47,4 Retail 50,2 47,8 50,9 48,8 Private service 35,6 38,5 32,7 37,9 Public service 73,3 59,1 76,1 66,9 Source: Stenberg, 2003d, p.12 It must be noted that short term labour market effects were almost universally better for those in LMT compared to the AEI (Stenberg 2007). The long term effects however have not been studied, but judging from some evidence, the impacts of general training may catch up (Stenberg 2009). Increased educational levels The National Agency for Education published a report assessing the changes in educational levels after the AEI for the period 1997/1998. It provided the following evidence (OECD 2000): Of the students in municipal adult education (which made 92% of the AEI, the remaining 8% studying at folk high schools), had an educational level lower than three years of upper secondary schooling before their studies. Approximately persons raised their educational level during studies which makes about 20% of all students. Of those with one year of upper secondary schooling, 50% raised their education by at least one level, as well as 33% of those with compulsory and two years of upper secondary schooling. The study showed that lifelong learning was a reality, with the oldest graduate student being 78 years old. Nearly 2,000 of those who increased their level of education were 50 years old or more. Statistical methods have been used to evaluate personal and social non-labour market impacts of education in general. Data from various surveys was used to find correlations between education and other social (and economic) parameters. Further education One of the major measurable non-labour market impacts of the AEI was further education of the participants. Every fourth student having taken part in the AEI moved on to further learning other than labour market training. In comparison with the LMT, the AEI participants went on to further education more frequently, while LMT had a larger percentage going into employment (see the table 23 below). This, of course, would be a reason for them to have gained higher wages (cf. p. 111), but the long-term effects on earnings have not been estimated. Table 23: Reasons for ending unemployment spell by treatment group Mean % of AEI % of LMT duration Censored (observations) 1,666 8,6 9,3 Employment ,0 52,0 114

117 Miscellaneous ,2 17,0 Unknown reason ,4 10,8 Further education ,0 10,9 Total ,0 100,0 Source: Stenberg, 2003a, p. 20 Evidence suggests that adults who came to higher education from the AEI showed the same academic results as those who came from general education. 49 The Interview method was used in this study to find out personal and social impacts of the AEI on its participants. Increased selfesteem Improved parental roles Increased motivation of children Increased motivation of men Social inclusion Removed barriers vs lack of motivation The AEI had a multiple impact on the personal development of the participants. 50 Two major target groups of the AEI were long-term unemployed and low-skilled people. This means that the initiative targeted those who have failed in education and/or labour market. This is a sensitive group, usually with very low self-esteem. Those who had taken the second chance offered by the AEI and raised their educational level, also increased self-esteem ( I can do it! ). An important aspect of adult education was that after attending education courses adults managed to help their children with their studies (became informal assistants to the teachers) and discussed a number of new topics with them. Having acquired the background of general knowledge, adults were more interested in the education of their children. The fact of being able to help one s children with school subjects (e.g. mathematics) was a great source of self-esteem for adults. This aspect of educational impacts was a new discovery. The example that parents showed to their children by participating in the AEI increased the motivation of the children to learn and achieve better results. Children were proud of their parents who, after attending study courses, were able to use computers, speak English or do well in mathematics. This contributed to the personal development of both parents and children. In many industrial regions of Sweden, where a majority of men were involved in low-skilled jobs, upper secondary education was dominated by women. After the introduction of the AEI the example of women in education motivated more men to study. The initiative attracted adults from marginalized groups and subcultures, for example, punks, who were the so-called community outcasts. After attending study courses, the behaviour of these people had changed. They became softer and improved their dress code. Participants expressed the opinion that it was very important for them to be a part of the community, to feel a membership in it. A study on difficulties in recruiting unemployed adults to adult education in the context of the AEI was carried out by S.Paldanius, Linköping University, in Research in the field of adult 49 Interview with Åsa Sohlman 50 Based on interviews with Eva Nordlund and Orjan Albhin 115

118 education has shown that the people who are most difficult to recruit are also those who need education most, namely, individuals who lack upper secondary qualifications. The paper argues that the target group s experiences of earlier education, working life and unemployment are important contextual factors that need to be considered, in line with the interest to recruit them. Reluctance to study, shaped by the previous experience, was the reason why some of the adults targeted by the AEI chose not to participate in the programme, despite barriers to their participation being removed. There were not many barriers at all, but there was a lack of interest or motivation for education. The study revealed that some target groups perceived adult education as the deterioration of their position in life, as it had a connotation of being positioned as a child once again, condemned to boredom or adult day care. The other target groups could not attach value to adult education, as it could not offer firsthand experience (as opposed to reading about something), some distrusted governmental agencies, especially originating from being unemployed, and regarded education as a luxury rather than necessity. Such rationality of the target groups opposed the rationality of recruiters. Almost all recruiting resources during the AEI went to measures based on knowledge about barriers and their removal. However, not a lot of efforts concentrated on creating motives to study Impacts on the organisational level Organisational networks The implementation of the AEI comprised various forms of collaboration and co-ordination between different organisations and actors and, consequently, established long-term cooperation networks in the field of adult learning. Thus, the impact of the reform is not only related to individual performance, but also to the organisational set-up and infrastructure of adult learning as a whole (Abrahamsson, 2004). In 2001 the implementation of the AEI in Umeå municipality was evaluated (Segerholm 2001). The paper described how this national reform was organised and carried out in one municipality. The explanation-oriented approach was used. The way students perceived their education, instruction and learning was analysed in relation to their educational backgrounds, stated interests and needs. The views of students and observed instruction were discussed in relation to educational content, provider/organiser of education, gender, and the students' socioeconomic backgrounds. As this extensive adult education initiative focused on upper secondary education, new patterns for planning educational activities in life became immanent in the municipality. Another paper in 2004 analysed the restructuring of adult education in Gothenburg municipality, as a result of the AEI and other developments (Beach 2004). The study used ethnographic data about changes to the supply of basic and upper secondary adult education, Swedish for immigrants, in particular. The restructuring of adult education had consequences for all education suppliers, but in particular to an adult education company called Studium Ltd. It was created in 2001 and was the largest supplier of adult education, but lost its contracts during tendering and in 2004 was on the brink of bankruptcy. Public funds paying for the conversion of public services to private 116

119 seems to be a consistent element of education restructuring according to international research. New organisations The AEI stimulated cooperation between municipalities, adult education providers and employment offices. Common meetings, common counselling of adults and other activities initiated during the AEI resulted in the establishment of single adult learning centres. After the AEI, in special state funding was given to municipalities to improve adult learning infrastructure, including the adult learning centres. In 2004 the majority of municipalities in Sweden (257 out of 290) had such centres. 51 Box 9: Coordination of adult education in Linköping Municipality At the beginning of the AEI, Linköping municipality prepared the first project proposal (application) itself, but later in 1997 it established a separate organisation to be responsible for the daily work on the AEI project (it was called Kunskapslyftet, the same as the initiative). It had a chief officer who was in charge of planning activities and cooperation with private education providers. The new organisation also employed two study counsellors and two people responsible for the admission of students. A lot of cooperation work was done at the beginning to disseminate information and establish sustainable networks. Thus, the initial aim of the new organisation was the coordination of the AEI activities within the municipality. When the AEI ended in 2002, the municipal executive board decided that Kunskapslyftet organisation should be permanent. Then it was called Kunskapstorget and has been responsible for the coordination of all municipal adult learning in Linköping. In 2009 Kunskapstorget had seven staff members, an office and a webpage. It provides information for adults about learning possibilities by publishing an annual catalogue of municipal adult learning. It also gives counselling and any other personal help for adults on a regular basis. 52 Source: interview with Örjan Albhin, adult counsellor in Linköping Municipality Unconventional solutions Increased local responsibility and cooperation stimulated the development of new adult learning models and methods, flexibility in provision and a variety of learning environments. New forms of activities, unconventional organisational solutions, information technology and new pedagogical approaches were applied. Box 10: New adult education providers in Linköping The AEI has stimulated the establishment of new adult education providers. During the first years of the initiative the following schools were operating in Linköping: the ordinary public schools for adult education; public schools for secondary education, which accepted adult students (in special groups but in the same buildings); educational associations like ABF, Folkuniversitetet, Medborgarskolan; folk high schools like Valla Folkhögskola, Bona; and private schools, both old private schools like LIber Hermods and NTI-school, but also new private schools which on behalf of the municipalities provided adult education on a local basis and also some who started in several municipalities. As the result of cooperation between the Bona Folk High School and the educational association ABF, both supported by the trade unions, a new school was established in the residential area, called Skäggetorp. This area 51 Interview with Åsa Sohlman 52 For more information see /Kunskapstorget/ 117

120 had high numbers of lower educated residents. Local authorities understood that neighbourhood was very important for these people and they would not go to study in other parts of the city. Thus, by establishing a new school in the neighbourhood, the municipality reached a lot of people with low educational levels. Source: interview with Örjan Albhin, adult counsellor in Linköping Municipality Flexibility in learning Focus on individual Learning in any setting The AEI facilitated a shift from traditional adult education to flexible learning in Sweden. 53 The overall idea of the initiative was to offer adults individual study paths which would be the most beneficial to them. This was especially important, since the reform targeted those adults who had a comparatively short educational background, sometimes also with negative school experiences from the past. Therefore, the government aimed to ensure that adults who once had failed as young learners would not fail as adult learners, when they were given a second chance to learn by the AEI. There was a very clear message from the central government to the municipalities not to fail second-chance learners. This meant that traditional education system had to be changed in order to meet the needs of adult learners. The changes in adult education in Sweden were built on two basic principles: flexibility and focus on individual needs of learners. Adults studied according to individual plans and they could choose between different subjects, methods, time and place. For example, the adult learners could choose to study history in a library, mathematics in a classroom with the intense supervision of a teacher and English at home with the help of a computer. As a result of the AEI, the discourse on and practice of adult learning in Sweden has shifted from learning at school to learning in any setting. Increased access to learning due to unconventional solutions resulted in increased motivation for further learning. Box 11: The impacts on unconventional solutions in Linköping Municipality There were a number of companies in Linköping working on an environment project. They had to be certified in order to carry out the work, so they needed to increase their environmental knowledge and skills of the employees. The companies began cooperating with the coordinating office for the AEI in Linköping, and the employees applied for and got access to the general secondary school course in nature studies called Naturkunskap A. The tuition was planned in cooperation with the companies and the school to be suitable for adult learners. The solution was to carry out studies during working hours, so that the employees would not lose salaries, and in the workplace, so that it would be easily accessible. As a result, about people participated in the course. The effect of this was that a lot of people afterwards were interested in taking general courses, e.g. Swedish and English and came to the municipality to enrol or get information. Source: interview with Örjan Albhin, adult counsellor in Linköping Municipality Individual study plans Individual study plans had existed in Swedish adult education before the AEI, however, they were formal and comprised mainly information on study subjects. After the AEI, individual study plans have become more results-based, including educational goals and further perspectives. Moreover, they have been built on the basis of adults prior experience in education and work. 53 Based on interviews with Karin Nilsson and Eva Nordlund, Directors of Education, the National Agency for Education, and Björn Garefelt, National Council of Adult Education 118

121 Validation of prior learning In Sweden the concept of validation (validering) was introduced in 1996 in relation to the AEI (Fejes 2007). It was the first initiative in the field of validating informal and non-formal learning in Sweden and aimed at facilitating the access of the target groups to secondary and/or upper secondary learning. The idea was to try out new methods and models of validation, so that the individual s prior learning would be taken into account, no matter where such learning had taken place. The AEI worked as a catalyst for the development of validation of prior learning in Sweden. Discussions and actions with regard to the validation of prior learning in Sweden have taken place both at local and national levels. During the AEI and later, the issue has been discussed among individual teachers and within municipal adult learning organisations, especially regarding the validation and description of silent or experiential knowledge. 54 Validation of prior learning was a priority for the government, so it was included among the criteria of funding municipal adult learning within the AEI. A number of pilot projects were carried out under the AEI. Most of the initial projects were mainly focused on the recognition of the immigrants vocational competence. More specifically, the focus was on practical knowledge and how one could take this into account when relating it to the Swedish context (Fejes 2007). Six municipalities were involved in trying to find methods and organisations for validation (Andersson et al 2003). Based on the analysis of different validation systems in these municipalities, several forms of organisation were distinguished (see the table 24 below). Thus, the funding and freedom given by the government to the municipalities was a chance to develop and test different models of validation. Table 24: Different types of validation in Sweden Organisation form Starting point for validation Vocations The individual Recognition integrated within the educational system Vocational and educational recognition Individual and educational recognition Validation parallel to the educational system Vocational and trade based recognition Individual and trade based recognition Source: Andersson, Fejes, Ahn Box 12: Validation of prior learning in Linköping Municipality Many municipalities began to use the orientation-courses in a larger perspective, e.g. for validation purposes during the AEI. It was possible for the students to have study allowances during this period. The time during these courses was used for interviewing adults about their education and work experience. Then, the results of the interviews were discussed and used as a basis for individual study plans. Consequently, the study plans could be shortened relative to prior learning or work. Moreover, orientation courses were used to discuss the practical aspects of adult learning, such as how to get to school, how to organise child care, etc. This was a completely new initiative. Source: interview with Örjan Albhin, adult counsellor in Linköping Municipality Distance learning The AEI stimulated distance learning, the use of ICT and learning 54 Silent knowledge is the knowledge acquired as a result of practical experience, not studying. It is so natural that most people find it difficult to describe in words. Silent knowledge poses a particular problem in describing and validating prior learning from experience. 119

122 in different environments. Box 13: Distance learning in Jokkmokk municipality Jokkmokk, a sparsely populated municipality in the north of Sweden with 6500 inhabitants has an industrial life dominated by forestry, hydropower and reindeer husbandry. Unemployment is high and traditional industries have a declining need for labour. By means of IT technology and under the slogan Let s move knowledge, not people! a large part of the education in the AEI was carried out as distance learning. The ambition was to make high quality education accessible to all people in the non-coastal areas of Norrland. Source: Cedefop Panorama 2001 (adapted from OECD 1998) New adult learning methods established during the AEI have fed into regular adult learning system. For example, adults can have flexible learning time and start their studies not only in August, as it used to be, but at any time of the year. The need for individual learning plans is now legally provided. Changing role of teachers Due to the AEI, the role of teaches has changed radically from knowledge provider to coach. The teachers were challenged to work in many different ways with groups varying in size and knowledge. Problem based learning and long distance education, as well as the cooperation within and between municipalities, have had an effect upon the development of the teachers profession. 55 The exchange of knowledge and practise between the different schools started during the AEI has in some way spread new methods between teachers. For example, adult education institutions have the so-called pedagogical café s a couple of times a month, which are used for discussions on topical issues. Teachers are also offered lectures in cooperation with the universities and the adult teacher education institutions. 56 Biographical (Occupational life history) method The development and change in Swedish municipal adult education (komvux) was investigated through occupational life history studies of four teachers in different municipalities who had worked in komvux since the mid 1970s. Individuals narratives of occupational life have been used to trace the changes of the teachers role. Reforms and restructuring on the one hand were coupled with the development and change in the teachers occupational trajectories and their organisational and municipal context on the other hand (Loeb, 2007). The teachers stories show that both teachers and their organisations were pressed by the AEI to develop a new work order, with the emphasis on cooperation, collaboration, individualization and flexibility. Three (out of four) teachers and their organisations experienced a great deal of change during the implementation of the AEI. The need to restructure municipal adult education in order to create new, flexible and individualized learning system was acknowledged and discussed on all municipal levels. The new ideas of individualization and flexibility implied 55 Interview with Karin Nilsson 56 Interview with Orjan Albhin 120

123 tension and struggle for the teachers and students to find structures to attain the goals of the curriculum and the syllabi. Work under market conditions Moreover, there was a question of whether the existing komvux organisation, now one of several providers in the market, would be able to transform and survive. After the first couple of years ( ) the teachers in three (out of four) organisations were dismissed or transferred to comprehensive schools in the municipality because the other providers received a considerable amount of the students (40-60%). For those left in the organisation, work pressure increased, while planning and preparation time decreased. The tendering of adult education during the AEI implied that more work had to be produced while the rates of pay were fixed according to wage agreements for teachers in the public sector. One (out of four) organisations went through less dramatic changes, because the AEI was implemented mostly within the existing institutional framework. Only 20 % of the AEI activities were allocated to other providers with flexible models. This could be due to the close collaboration and discussions between the teachers and municipal politicians. Nevertheless, there was also a pressure for changes in line with the ideas of flexibility and resultorientation (Loeb, 2007). Implementation of new ideas On the other hand, the AEI provided a unique chance for teachers to realize their new ideas, which could not be realized in the old adult learning system. Box 14: Private initiatives of teachers in Linköping Municipality In 1998 two adult teachers in Linköping Municipality applied for the state grant within the AEI framework to start a new private adult education company in the field of nursery schooling, called Moa-Learning Centre. They had a lot of ideas, and believed that they would have more opportunities to implement them in a new company than in the existing institutional framework. The AEI gave them money and freedom to implement their ideas. So, the teachers left their former jobs and established a new school, which was different from the traditional learning environment. The school building was home-like, with small rooms. The learning was organised in smaller or bigger groups, depending on the subject. Validation of prior learning was initiated. The teachers created their own model for the validation of prior learning in the field of nursery teaching and applied it in their school. After the AEI additional funds were allocated to build infrastructure for adult learning. Consequently, the company has expanded to other municipalities and currently operates 4-5 nursery schools for adults. Source: interview with Albhin Örjan, adult counsellor in Linköping Municipality National Agency for Flexible Learning Quality assurance As part of the AEI, the National Agency for Flexible Learning was established in 2002 in order to encourage and stimulate the local authorities to make decisions in the direction Government wanted. The agency has played an important role in supporting the development of flexible learning in municipal adult education, folk high schools, study associations and workplaces (Beach 2004). During the period of the AEI a quality assurance system of adult education was built up at the municipal level. After the reform, the average quality level of the adult learning courses has improved; at least adult education providers have better knowledge of the quality level of services. 121

124 Methods to measure quality impacts of education: - A direct method is to ask participants about their experience. An index ( Vuxenbarometern ) was constructed to measure the quality of the AEI. It was built on questionnaires collected from participants. The drawback of this method is that quality means different things to different people and personal opinions may be difficult to measure in an unambiguous way. - A success rate measure is the share of participants in the AEI who managed to get a job or continued to higher-level education. It indicates the degree to which adult education motivates and provides the foundations for further studies or employment. The method could be used to compare success rates across municipalities or adult learners inside and outside the AEI. - Drop-out rates can also be used to measure the quality of education. Reputation for quality is a determinant of demand for adult education. As might be expected, external education providers, which are not funded by the municipality and have to struggle for participants and income, are motivated to improve quality of education and, consequently, have lower drop-out rates. The drawback of this method is that participation in adult education might be motivated by factors other than quality. (Tangerås, 1999) The following quality assurance measures have been applied on the municipal level after the reform: Student surveys two times per year. The quality indicators of the average grade on 1-5 scale are over 4 and have risen over the last 10 years. All schools are visited once a year and meetings with school managers, teachers and students are held. The schools have to report annually to the local municipalities on their activities and achievement of quality indicators. Municipal adult education institutions are also supervised (one in five years) by the Swedish National Education Authority. Continuity of the AEI The AEI ended as a separate project in 2002, but its activities continued on a regular basis within komvux. The level of financing was gradually reduced over several years from the equivalent of places to places in In the period a new targeted state grant was provided for municipal adult education. The grant was prolonged for yet another period and since 2007 it has been a part of the targeted state grant included in the general grant to municipalities Impacts on the policy level Legislation The AEI culminated in the Education Act and the Municipal Adult Education Ordinance, which aimed to reform and renew Sweden s education system extensively, including adult education. In 2001 the Swedish parliament approved the government s bill Adult Learning and the Development of Adult Education (prop. 2000/01:72). In this Bill the Government presented proposals on goals and strategies for the development of adult learning. The Bill sets out a strategy for support from the state and the municipalities for adult learning based on the needs of the individual. The focus was shifted from teaching in school-like forms 122

125 to more flexible ways of providing support for the individual's learning. The goal of adult education was defined as follows: All adults must be given the opportunity to broaden and deepen their knowledge and skills in order to promote personal development, democracy, equality of opportunity, economic growth and employment, and an equitable distribution of wealth. Strategies In order to reach these goals the following strategies were formulated (Regeringen, 2000): Methods of teaching, learning and working are to be developed corresponding to the individual s need to learn in a knowledge-based society. Counselling and guidance based on the recognition of previously acquired knowledge and skills are cornerstones when the individual sets out on a learning project. A system for financial support should be able to stimulate adults participation in education and skills development. Society, employers and individuals have co-responsibility for catering to the needs of various individuals and various groups as regards both general education and education for special purposes All formal and non-formal education that enjoys the support of society is to be permeated by a democratic view of life and a respect for the equal value of all persons A fundamental co-view of political areas and cooperation between authorities, both sides of the labour market and NGOs is to be striven for in order to realise an individual s education and competence development. Outreach activities, counselling, validation, accessibility and study support were identified as the foundations of the infrastructure for lifelong learning. Validation of prior learning The AEI was soon followed by the bill on Adult Learning and the Future Development of Adult Education (2000/2001:72). In this bill it was emphasised that any resident in Sweden should be able to have their knowledge and competence validated within the framework of municipal adult education, and this validation should be documented. As a result, this bill granted additional funds (until 2008) to municipalities to undertake further development projects in a range of issues, including validation (Cedefop 2007). Subsequently, in 2003, this bill was followed by the bill on Validation (Ds 2003:23). In this bill the government stated that is was too early to regulate validation. More time should be given to pilot projects to gather more experiences, and to discussion before deciding on regulations and passing acts. It was emphasised in the bill that the responsibility for validation should be shared between educational authorities, business sector organisations, the social partners, universities and municipalities and that regional cooperation is especially essential. National commission on validation Right after the end of the AEI and based on the experience of the pilot projects, the government set up the Swedish national commission on validation (Valideringsdelegation) in 2003 for the period of The aim of the commission was to enhance regional cooperation and to explore quality and methodological issues around the topic (Nelissen 2007). In order to achieve its objectives, the commission funded several developmental projects 123

126 focused on trying out different methods of validation, especially focused on validation of vocational knowledge within a few occupations. Another focus was on the recognition of vocational learning in a more broad sense, not only related to specific vocational tasks or educational programmes. The Commission was financed by the Swedish government and had 6.5 million (SEK 60 million) at its disposal until 2007 for its activities. Synergy of education and labour market policies Increased support to nonformal learning The AEI served as a tool for enhancing cooperation between education and labour market policies towards the common goal of adult learning. The reform was mainly financed by reallocating funds earmarked for passive measures in the Ministry of Labour s chief policy area to active education and training programmes in the Ministry of Education and Science s main area of operation (OECD 2001b). There was a consensus that the unemployed who lacked upper secondary school competence needed education in order to acquire a stronger position in the labour market. For this reason they had access to educational opportunities in the first instance, and not unemployment benefits. After the end of the AEI in 2002 the Swedish Government decided that the folk high school should keep resources corresponding to 7000 of the full time student places given to the folk high schools during the AEI. This extra resource was later made permanent and added to the regular state grant of the folk high schools which was thus considerably increased. Furthermore, the folk high schools were asked to take part in extra new activities in cooperation with the Public Employment Service specifically for long-term unemployed in Most folk high schools took an active part in these activities. 57 In 2006 the Swedish Parliament decided on an overall future policy concerning state support to folkbildning (folk high schools and study associations). The decision was to substantially increase (by 12, 6 %) the general state grants to the folk high schools and study associations. All political parties agreed on this at the time. This decision was taken following a state evaluation of the societal and individual effects of folkbildning activities as a whole. The Parliament decision expressed support for folkbildning. No doubt the positive experiences from the AEI and other similar activities contributed to this, even though it is impossible to isolate and measure to what extent the AEI had an impact on the overall decision. Conceptual shift It has been argued that in Sweden, during the introduction of the AEI in 1997, there was a peak of the conceptual transition of lifelong learning from democratic and humanistic dimensions of the concept towards the economic thinking of human capital (Gustavsson 2009). Adult education in Sweden moved from centralized lifelong learning (characterized by uniform education programmes and settings) to individualized adult education, with focus on flexibility, responsiveness to user-demands, ability to combine studies at different educational levels, and encouragement of engagement, commitment and investment. The reform generated a substantial increase in subsidised comprehensive adult education and paved the way for many 57 Interview with Björn Garefelt, National Council of Adult Education. 124

127 subsequent projects. It resulted in the establishment of a new approach towards adult learning in Sweden, the restructuring of adult learning institutions and the formation of long-term cooperation networks. The conceptual and cultural shifts towards individualised and flexible adult education took place. To get a broader and international perspective on Swedish reforms in adult education and lifelong learning, the Commission on the AEI invited three prominent researchers and experts to study developments in Sweden and submit their views on the strengths and weaknesses of the Swedish system for adult learning and education. The researchers examined adult education, especially the AEI, in three municipalities. They also interviewed a number of politicians, including members of the Riksdag's education committee and a large number of civil servants at the central level of the Government office, the central authorities as well as the trade unions. On the basis of their observations and interviews, they wrote a report that provided a historical perspective of Swedish adult education, international views and comparisons. The conclusions of the OECD Thematic Review on Adult Learning: Sweden included the following (OECD 2000): There was a shift from education to learning and a broader perception of competence. Learning takes place not only in formal education but also in everyday life and on the job. A strategy for lifelong learning, covering competence development in working life and learning in everyday life, was lacking. There was also a lack of a strategy for validating knowledge. The AEI was a success in quantitative terms, but systematic evaluation of quality was missing. No clear indicators or criteria for assessing quality were developed. The specially assigned state grant to the municipalities made a substantial contribution to the successful implementation of the AEI. It also enabled adult education to be rapidly expanded Conclusions Relevance of the analytical framework Major achievements of the reform The case study was based on the analysis of previous research findings and expert interviews. In contrast to the labour market training and counselling reform in Lithuania, the AEI in Sweden was well researched and documented. Therefore, the impacts of the reform were exemplified on different levels. Expert interviews in particular proved to be useful in providing concrete examples of the reform activities and their impacts at local level. The major positive impact of the AEI was a chance for adults who had not finished upper secondary education to catch up on education. Many of those adults who had left school when they were came back to their studies and managed to finish them and go even further to higher education. Thus, the AEI increased the possibilities for adults to go "one step up. This had multiple long-term benefits for individual adult learners and society at large: 1. Provided better employment opportunities and reduced unemployment; 125

128 2. Encouraged further education and increased education levels; 3. Increased self-esteem of both adults and their children; 4. Improved parental roles; 5. Increased motivation of children to achieve better results at school. 6. Encouraged more men to study. 7. Reached adults from marginalized social groups and involved them in education and community, 8. Provided more opportunities to make personal choices about learning. Another positive impact of the AEI was that it built the road while at the same time walking on it. Due to the initiative, municipalities developed the capacity to offer a variety, as well as increasing the volume of adult learning courses in cooperation with different stakeholders. The AEI worked as a catalyst for adult learning in Sweden to develop from traditional education to flexible learning. A shift towards individualised learning, flexibility in provision and validation of prior learning and experience were the main success components. Success factors It can be concluded from the evidence collected that the AEI was successful in reaching its goals: reforming the adult learning sector and increasing the educational levels of adults. The following success factors of the reform can be identified: 1. The reform fitted well with the long adult education traditions in Sweden and was approved by social partners. 2. The involvement of social partners contributed to active recruitment measures e.g. through trade unions. 3. The reform was built on the already existing decentralised system of municipal adult education. It took advantage of and developed the existing infrastructure for Lifelong learning (both formal and non-formal) on local, regional and national levels. 4. The reform was steered by strong political leadership and explicit nation targets and driven by generous government funding. The implementation of the reform was entrusted to local actors and allowed them considerable freedom of action. 5. The reform promoted a holistic approach to adult learning and comprised both formal and non-formal education, so adults could choose between alternative pathways. 6. The reform avoided a strictly instrumental one size fits all implementation and allowed a lot of experimentation at local level. 7. The government explicitly encouraged cooperation between different stakeholders and different ministries. 8. The establishment of coordinating institutions at local level. 9. Discussions and pilot testing preceded the adoption of regulations. 10. The reform was focused on the individual needs of adult learners and a more flexible approach towards adult learning. 11. Adult learning was seen as an investment rather than a cost. 12. Removed economic barriers to participation by providing educational grants. Drawbacks In line with the achievements of the reform, the following drawbacks could be identified: 1. Some target groups, particularly people with the lowest education level and a weak position in society, were not motivated by the AEI to participate in adult learning. 126

129 2. No cost-benefit analysis of the AEI has been made, so it is not known whether this huge state investment has paid off for society at large. 3. The initiative came a bit too late with regard to the economic recession in the country (which was at its worst in ). Recommendations Based on the evidence collected in this case study, and especially on the drawbacks of the reform, the following recommendations can be made: Recommendation 1: Along with the collection of monitoring data and evaluation of the reform, cost-benefit analysis should be made before and/or after the reform, so that financial, social and economic costs and benefits for society at large could be estimated. Recommendation 2: The synergy of different policies (education, social protection and inclusion, employment, etc.) should be promoted, so that the most disadvantaged groups could be reached by the reform activities. Recommendation 3: Political leadership, generous funding and autonomy of service providers should be considered as pillars for a successful reform. Advanced governance tools, such as management by results and the involvement of stakeholders, should be considered as the major facilitators of the reform progress Key success factors of the reforms General context and historic development The above case studies analyse reforms that were implemented in very different national contexts. For a better understanding of these differences table 25 below provides a short comparison of the two cases according to the main topics, chosen for our analytical framework (see Figure 13). Table 25: Short comparison of the two cases Topics Similarities Differences Both reforms were started and carried out in the context of deep economic crisis. The object of the analysis Inputs and activities SE already had a well-developed adult learning system, strong social networks and learning culture, while LT could make only very limited use of previous institutional structures, networks and culture; SE had a long partnership tradition and strong local government, while LT inherited a culture of mistrust and strong central government. Both reforms were targeted SE reform was more results-based, at unemployed adults in comparison with LT. It had Both reforms were aimed at explicit national targets, while LT the reduction of reform was more input-based. unemployment and changes in adult learning system. Financing SE reform financed the demand side of adult education, while LT invested more in supply; - SE reform was financed on a project-basis, while LT reform was financed on an institutional budgetbasis. Service provision Both SE and LT were LT created a closed model of learning by doing, but in service provision, which was SE the learning was taking liberalised later, while SE liberalised 127

130 Topics Similarities Differences place more at local level. it from the beginning; SE reform was a joint responsibility of the state and municipalities, while LT reform was centrally planned and implemented with little involvement of local actors; SE reform had a political leadership at the national level and promoted institutional partnership, while LT reform was marked by institutional turf fighting (at least in later stages); SE reform was regularly monitored by results, while LT introduced more systematic monitoring only after the reform. Regulative Both SE and LT had legal provisions for the quality assurance of services; Legal provisions for validation of prior learning was in a rudimentary stage both during SE and LT - reforms; In both SE and LT key legislation followed the reform, thus legitimising the reform outcomes. In both SE and LT parallel In LT parallel initiatives were of a External factors initiatives in adult learning took place. larger-scale than in SE. In both SE and LT personal outcomes of participation in Immediate/intermediate learning were better impacts at individual level employment and increased - self-confidence of the participants. Immediate/intermediate impacts at organisational level Immediate/intermediate impacts at policy level Ultimate impacts on participation in adult learning Availability and quality of data Source: compiled by the authors In both SE and LT reforms resulted in increased institutional flexibility. In both SE and LT reforms resulted in increased (and further increasing) participation of adults in learning. - In SE organisational impacts of the reform occurred mainly on the municipal (local) level, while in LT organisational changes occurred mostly on the central (national) level. In SE the reform shifted adult learning to be steered by demand, while in LT it was still inputfocused; In SE the reform strengthened the cooperation between and within municipalities, while in LT municipal involvement remained weak, but employer involvement increased. In SE increase in participation was significantly higher than in LT. Various impacts of SE reform were well researched, while there were very few studies of the impacts of LT reform. To summarise, detailed case studies revealed that there are two success factors which are particularly important for reforms regardless of their geographical location: Strong and active political leadership in relevant institutions responsible for the reform that help establish explicit national targets of the reform and maintain its momentum throughout the implementation process; and 128

131 Sufficient public funding of activities subject to the reform, which increase ownership of the results among relevant institutions and actors. Other success factors depend on the context of the country. Findings of the detailed case studies allow the argument that in countries with well-developed economies and civil societies the following factors contribute to the success of the reform: Existence of long adult education traditions accepted by all stakeholders (not only social partners), strong institutional partnerships and service provision arrangements at local, regional and national levels; Very active involvement of actors at local and regional levels; and Use of a complex approach which tailors implementation of policy initiatives to individual needs, encompasses all forms of adult learning (formal, non-formal and informal) and actively promotes adult learning as an investment. In the case of less-developed economies which have a relatively weak middle class, underdeveloped or very new structures of government and a passive civil society the following success factors drive the reforms forward: Availability of best practice examples of the functioning of similar systems in other countries and ability to adapt them; Motivation of national and, especially, regional and local stakeholders to participate in reform activities regardless of limited financial resources, workload and related adverse working conditions; and Availability of financial resources supporting reform efforts other than limited national government funding (e.g. EU funding instruments, other international aid); these sources could usefully finance at least learning from the best practice examples. Both reforms emerged as a reaction to economic recession and radical increased unemployment. The findings of the case studies also revealed that each reform, regardless of its geographical location, is very often subject to three drawbacks. Firstly, policymakers failed to choose the best timing for reform. For example, in the case of the Swedish reform the initiative came a bit too late with regard to the country's economic recession (which was at its worst in ). Secondly, policy-makers fail to pay due consideration to all the financing options of the reform. This often results in a less favourable cost/benefit ratio of the reform. Very often cost/benefit analysis is not carried out at all (valid for both case studies). Finally, reforms fail to reach specific target groups which are characterised by their lack of motivation to participate in adult learning. For example, in the Swedish case, people with the lowest education level and a weak position in society were not motivated to participate in learning despite removal of the barriers. The information about the participation of the most vulnerable groups in labour market training and counselling during the Lithuanian reform or immediately after it was not even collected. 129

132 CONCLUSIONS AND RECOMMENDATIONS The report explored the methodological aspects for assessing the impacts of education and training as well as related reforms on adult learning; in addition guidelines for conducting such impact assessments were developed and their application exemplified in conducting two case studies: The establishment of a labour market training and counselling system in Lithuania ( ) and The Adult Education Initiative in Sweden ( ). In this chapter we provide the overall conclusions and recommendations of the report. Uneven progress of Member States in adult learning participation and attainment Adult learning sector needs more attention and better evidence The EU institutions have taken steps to monitor and evaluate adult learning across Member States The greatest challenge for EU policy in the area of adult learning is the imminent failure to reach the Lisbon target of 12.5% participation in lifelong learning (which for 2020 has been raised to 15%). While some EU (and EFTA-EEA) countries show exemplary performance in several aspects of adult learning, there are wide and persisting disparities across the EU. Adult participation in learning differs substantially between Member States, between low-qualified and highly-qualified, low-skilled and highly-skilled, younger and older adults. Secondary and tertiary education attainment of the adult population also differs to a great extent between Member States. But as our examples show, reforms can be successfully introduced in different economic settings and affecting different groups of adult learners. The policy responses of the EU and the guidance provided to the Member States have been numerous and increasingly adult learning policy specific. The implementation of EU political initiatives was supported by the Open Method of Co-ordination and EU financing instruments. Depending on domestic political agendas and the motivation of political actors in the Member States EU level initiatives will take a considerable length of time to be translated into action and to yield tangible results. So far most of the positive developments in education and training policy at Member State level in response to EU policies were targeted at the younger population. Reinforced messages from the EU might still be able to convince more domestic political actors that adult learning needs special attention. The best way to convince them is to provide tangible evidence of the benefits of reforms enjoyed by other countries. The OMC process cannot force change upon national institutions, but it should provide a channel of peer expertise for those willing to reform and who are looking for evidence about similar reforms in other Member States. Many reform agendas do not become reforms because of uncertainty. They are seldom supported by hard evidence on the impacts of similar reforms elsewhere. To stimulate national reforms the EU institutions have set up a framework of indicators and benchmarks for monitoring progress towards the Lisbon objectives in education and training. They have made sure that these are used in national reporting processes under education-related OMC, such as the Education and Training work programme or the Copenhagen process. All EU-funded interventions in the area of education and training are monitored and evaluated on a mandatory basis. The EU has even provided 130

133 some methodological tools (such as the LIME assessment framework for national reform programmes or the Guide for the Evaluation of Socio-Economic Development), which help track the progress of national reforms or guide the evaluation of EU-funded programmes, which often support reforms in the Member States. The EU-level monitoring and evaluation improves general awareness of adult learning problems, but only efforts at Member State level could generate hard evidence about their solutions Assessments of impacts of education and training reforms on adult learning are scarce and difficult to access The above EU-level efforts can help to improve a general awareness of adult learning situation across Member States and capture some of the weaker or stronger points of the national education and training systems. But they are unable to produce the sufficiently detailed and reliable evidence, which is needed e.g. in the exchanges between the Directors General from different Member States. They cannot generate reliable evidence on which national reform measures work best, what problems they are able to solve and under what circumstances. And even more importantly, no matter how sophisticated EU-level monitoring and evaluation become, in the foreseeable future they will only be able to capture very basic policy design characteristics and their outcomes at Member State level. One reason for this is that the European Statistical System would take a long time to develop and apply instruments for the collection of data on indicators of policy and governance in education and training even if these were already agreed between the Member States. Such agreement would take even longer. For example, the OECD is conducting an ambitious biennial survey to measure governance across all policy areas (OECD 2009). This new initiative will definitely provide better opportunities for international comparative analysis of governance. However, not all EU Member States take part in it. Only an in-depth policy analysis at Member State level could provide for country-to-country experience sharing based on hard evidence. Some EU Member States demonstrate an extremely good performance in stimulating adult learning. Although assessments of impacts of education and training reforms on adult learning are rather scarce, some assessments are already available as demonstrated in Chapter 3 and could be readily used. Some impact assessments are inaccessible to wider audiences for other reasons, including language barriers or an unwillingness to share unfavourable results. The EU institutions have to seek ways of supporting an evidence-based exchange of good practice between reform practitioners in the Member States. More regular impact assessments of national reforms and their results shared with other interested Member States would enable much faster policy learning, which is needed for the achievement of the Lisbon objectives in education and training. The EU could attempt moving from peer learning to peer review in the education-related OMC and in particular focus on how to make the national reforms more effective in stimulating adult learning. Recommendation 1: Member States should be invited to carry out impact assessments of their reforms and to share their results with other Member States. The EU institutions should seek innovative and effective ways of supporting evidence-based exchange of good practices between Member States, and provide and facilitate forums for such exchanges. 131

134 Evaluation of the EU structural fund programmes produces few detailed assessments of supported reforms Since all EU-funded interventions in the area of education and training are monitored and evaluated on a mandatory basis and many of these interventions contribute to national reforms, the lack of evidence about their success is striking. The monitoring and evaluation of EU structural funds should give due attention to assessing the impacts of the education and training reforms they funded and especially the impacts of these reforms on adult learning. Recommendation 2: The Managing Authorities in charge of programmes financed by EU structural funds and the relevant Commission s directorates and services should consider commissioning thematic evaluations of EU cofinanced projects, which supported national reforms with potential influence on adult learning. Furthermore support to national reforms from EU structural funds should be monitored and evaluated on a regular basis. Although the EU institutions provided a wealth of guidance on how to assess the impacts of EU financing instruments, the guidance on how to assess public policy and management reforms, which often go above or beyond financing interventions, has been lacking. The analysis in the Chapter 4 of the report explains the key steps in planning and conducting impact assessment and underlines the areas which require special attention and effort. It reveals that planning and implementing impact assessment after the reform has taken place severely limits the availability of data and thus also the quality of analysis. Guidelines for the impact assessment of national reforms Ideally impact assessment plans have to be integrated into reform programmes, which would give those willing to study the impacts of reforms a number of advantages. Firstly, the data on reform actions, their outputs, results and impacts should be collected when it is still possible or cost-effective to obtain them. Preferably control groups and baselines should be established before the reform. Secondly, the sources of data could be meaningfully triangulated and different methods of analysis could be applied which would complement each other to ensure the strength of the conclusions. Finally, consistent measurements of immediate and intermediate impacts could be utilised in providing assessment of the long term and ultimate impacts of the reform. The data could also feed into the reform's own development and evolution throughout its lifetime as well as any future reforms. Absence of prior planning leads to second or third best quality of evidence, which is not as useful and compelling as it could be. Recommendation 3: Member States planning education and training or other reforms with potential impacts on adult learning should be invited to integrate impact assessments into their reform plans. These should include development of monitoring information systems, which have to be designed on the national level according to the intervention logic of particular reforms. Such systems enable the collection of data on the performance of reform 132

135 programmes the only chance to do this is when reforms are being implemented. Control groups and baselines for future impact assessment have to be established before or during the initial stages of reform. Two case studies exemplifying the guidelines for the impact assessment The case studies The establishment of labour market training and counselling system in Lithuania ( ) and The Adult Education Initiative in Sweden ( ) fully support the above conclusions. In the case of Sweden the data collection for monitoring and evaluation of reform was integrated into the reform process. The institution responsible for the evaluation of reform impacts financed independent research into the impacts of reform. This created a wealth of information, which could be used in assessing the ultimate impacts of the reform in Chapter 5 of this report. In the case of Lithuania the reformers were preoccupied in kick-starting a vital element of any modern welfare state labour market training and a counselling system, the resources were scarce and only very basic data was collected during and after the reform. Subsequently, this reform (as well as parallel reforms) received only very limited attention in academic or applied research. Nevertheless, the case study method proved to be useful in identifying and analysing the impacts of the reform. It revealed and illustrated positive and negative as well as immediate and long term impacts. But significantly for adult learning, both reforms showed the positive impact of reform on provision, access and (or) participation in adult learning. Detailed case studies revealed that there are two success factors which are particularly important for reforms regardless of their geographical location: strong and active political leadership in the relevant institutions responsible for the reform that helps to establish explicit national targets of the reform and maintain its momentum throughout the implementation process; and sufficient public funding of the reform activities, which increases ownership of the results among relevant institutions and actors. Other success factors seem to depend on the level of development in a country. In countries with well-developed economies, civil societies and long adult education traditions the success of reform is likely to lie also in the active involvement of actors at local and regional levels, and the use of a complex approach, which tailors implementation of policy initiatives to individual needs, encompasses all forms of adult learning (formal, non-formal and informal) and actively promotes adult learning as an investment. In countries with less-developed economies, a weaker middle class, civil society and younger government institutions the success of reform is likely to lie also in the engagement and non-financial motivation of key stakeholders to participate in reform activities. In addition, unlike the reformers in countries with well-developed economies, civil societies and long adult education traditions, that only have to use their country assets competently to implement a successful reform, the reformers in less prosperous and less developed countries need external best practice and resources to support their efforts. 133

136 Recommendation 4: Reform experience shared between EU Member States in the OMC process should also become readily available to the Candidate Countries, countries taking part in the EU Eastern Partnership as well as other countries, which receive EU external aid. EU financial support to these countries should facilitate policy learning from the more experienced EU Member States and cover some of the reform costs. Quantitative evidence on factors influencing the key indicators of adult learning The Annex to the report provides the results of an attempt to conduct a quantitative analysis on which characteristics embedded in policies explain performance in terms of participation of adults in learning, educational attainment and several other characteristics. The results of the analysis are subject to a number of important limitations and therefore should be treated with great caution. Nevertheless they suggest some interesting conclusions, which might be useful in raising the hypotheses of further inquires as well as informing decisions, which have to be made under a large degree of uncertainty. More significantly, the policy variables of the education system which influence adult learning in the group of countries studied for which data was available are identified. These are variables characterising stratification (the number of school types or distinct educational programmes and the first age of selection in the education system), vocational specificity (private sector involvement in the provision of initial VET and the prevalence of job related non-formal learning), flexibility in provision of learning services (inflow of students and proper opportunities for those aged to participate in formal education and training), school autonomy (esp. in budget allocations within the school) and spending on education (absolute and relative per pupil/student). The reforms in these policy characteristics would be likely to have important influences on adult learning. It is important to stress again the limitations of these findings. It might be that not all the important characteristics of education and training systems have been or could be studied due to huge data gaps. Statistical analysis has also provided evidence that education and training policy characteristics might not be the only and possibly not even the most significant factors influencing adult learning. Firstly, the characteristics of labour market policies and more specifically the flexicurity policy mix or its individual elements seem to exert strong influences on adult learning. Secondly, there is some evidence that health policies and their outcomes might also be helpful in explaining the variation in the levels of adult learning. This means that any reforms in these policy areas could also affect adult learning. Finally, there are a number of contextual variables, which might help to explain differences in adult learning and education attainment. The most significant and widespread link is between adult learning and a set of research and innovation variables, especially ICT development. The high level of economic development, social cohesion, and efficient use of resources also seem to form a favourable context for adult learning and education attainment. Most of these relationships are no doubt two-way, since adult learning also exerts positive influences on all areas of life and policy. 134

137 Multiple regression modelling which helped to measure the relative significance of variables revealed that adult participation in education and training (both formal and non-formal) is best explained by models involving labour market policies (incl. intervening variables balance of labour supply and demand and labour market mobility) and even more so by models involving a large set of both education, labour market and health policies and context variables. Although this analysis suffers from a number of limitations, it once more underlines the complexity of factors influencing adult learning and the challenges that lie ahead for future inquiries in this field. Recommendation 5: Future inquiries should consider that education and training policies might not be the only and possibly not even the most significant factors influencing adult learning; other policies and contextual influences should be given due attention when assessing the impacts of reforms on adult learning. Since many policies and especially policy reforms are both dependent on adult learning and influence the availability of learning opportunities, their evaluation should include the assessment of their impacts on adult learning as a horizontal policy priority for the EU. The complexity of factors influencing adult learning and the challenges that lie ahead for future inquiries in this field would merit the financing for fundamental research from the EU 7th Framework Programme. 135

138 REFERENCES 1. Abrahamsson, K. School to Work Transition, Labour Market Flexibility and Lifelong Learning in the Swedish Context, Albrecht, J. W., G. J. Van den Berg, S. B. Vroman, The Knowledge Lift: The Swedish Adult Education Program That Aimed to Eliminate Low Worker Skill Levels. IZA Discussion Paper No. 1503, February Allmendinger, J., Educational Systems and Labour Market Outcomes. European Sociological Review, Vol. 5, No. 3, 1989, pp Andersen, T., J. H. Haahr, M. E. Hansen and M. Holm- Pedersen. Job Mobility in the European Union: Optimising its Social and Economic Benefits. Danish Technological Institute, Anderson, T. and H. Metcalf. Modern Apprenticeship Employers: Evaluation Study. National Institute of Economic and Social Research, research Report RR417, Available from the Internet: pdf [retrieved on ]. 6. Andersson, P., A. Fejes, S-e. Ahn, Recognition of prior vocational learning in Sweden, Studies in the Education of Adults, 36(1), Arum, R., Do Private Schools Force Public Schools to Compete?. American Sociological Review, Vol. 61, No. 1, 1996, pp Auer, P. Flexibility and security: labour market policy in Austria, Denmark, Ireland and the Netherlands, in Schmid Gunter, Bernard Gazier (eds.), Dynamics of Full Employment: Social Integration Through Transitional Labour Markets, Cheltenham, Northampton: Edward Elgar, Anhang Lernende Regionen. Instrumente und Zusatzauswertungen aus dem Projekt "Lernende Regionen" Available from the Internet: ende_regionen/anhang/index.html 10. Baldridge, J. V., D. V. Curtis, G. Ecker and G. L. Riley, Diversity in Higher Education: Professional Autonomy. The Journal of Higher Education, Vol. 48, No. 4, 1977, pp Barr, N.. The economics of the welfare State, 2nd ed. Oxford: Oxford University Press, Beach, D., The Public Costs of the Re-structuring of Adult Education: A Case in Point from Sweden Becker, G. S., Health as human capital: synthesis and extensions, Oxford Economic Papers, Vol. 59, pp , Bernard Brunhes International, ICAS Institute and Economix Research and Consulting, EU-wide evaluation of the Community initiative EQUAL : Final report (volume 1). 29 July, Bjørnåvold J., Making Learning Visible. Identification, assessment and recognition of non-formal learning in Europe. Cedefop,

139 16. Blanchard, O. and J. J. Wolfers, The Role of Shocks and Institutions in the Rise of European Unemployment: The Agrregate Evidence, The Economic Journal, Vol. 110, No. 462, March Bloom, D. E. and D. Canning, The Health and Wealth of Nations, Science, Vol. 287, pp , Boden, G., Hillmann, K., von Horadam, M., Koch, S. Evaluation der Angebote der Weiterbildung nach Weiterbildungsgesetz NRW Qualitätsentwicklung in der Weiterbildung, Available from the Internet:, bericht_evaluation.pdf 19. Bonin, H., Schneider, H. Wirksamkeit der Förderung der beruflichen Weiterbildung vor und nach den Hartz- Reformen. Forschungsinstitut zur Zukunft der Arbeit, Institute for the Study of Labor Available from the Internet: ftp://ftp.iza.org/dps/dp2069.pdf 20. Bostrom A. K., E. Boudard, P. Siminou, Lifelong Learning in Sweden: the Extent to Which Vocational Education and Training Policy is Nurturing Lifelong Learning in Sweden. Cedefop Panorama, Breen, R., Explaining Cross-national Variation in Youth Unemployment: Market and Institutional Factors. European Sociological Review, Vol. 21, No. 2, Breen, R. D. F. Hannan and R. O Leary, Returns to Education: Taking Account of Employers Perceptions and Use of Educational Credentials. European Sociological Review, Vol. 11, No. 1, Brunk, T. and M. Wahman, Sweden: Collective Bargaining and Continuous Vocational Training Available from the Internet: e q.htm [retrieved on ] 24. Bukodi, E. and P. Róbert, Occupational Mobility in Europe. Report. Dublin: European Foundation for the Improvement of Living and Working Conditions, Cedefop, Individual Learning Accounts. Cedefop Panorama series; 163. Luxembourg: Office for Official Publications of the European Communities, 2009a. 26. Cedefop, Continuity, consolidation and change: Towards a European era of vocational education and training. Cedefop Reference series; 73. Luxembourg: Office for Official Publications of the European Communities, 2009b. 27. Cedefop. Lifelong learning: citizens views. Luxembourg: Office for Official Publications of the European Communities, Cedefop, Lifelong learning: citizens views in close-up. Findings from a dedicated Eurobarometer survey. Luxembourg: Office for Official Publications of the European Communities, Cedefop. National VET Systems Training Village. Sweden thematic overview Available from Internet: tionalvet/ [retrieved on ]. 30. Cedefop, Sharing the costs of vocational education and training: An analysis of schemes in the newer EU Member States. Cedefop Panorama series; 182. Luxembourg: Publications Office of the European Union, Cedefop, Terminology of Europe education and training policy: A selection of 100 key terms. Luxembourg: Office for Official Publications of the European Communities, 2008a. 137

140 32. Cedefop, Validation of non-formal and informal learning in Europe: a snapshot Luxembourg: Office for Official Publications of the European Communities, 2008b. 33. CHEPS, INCHER-Kassel, ECOTEC Research and Consulting Ltd. The Impact of ERASMUS on European Higher Education: Quality, Openness and Internationalisation Available from the Internet: pdf [retrieved on ]. 34. CIDEC Interdisciplinary Centre for Economic Studies. The Impact of the Recognition and Certification of Lifelonglearned Competences: Update. Lisboa: Direcção-Geral de Formação Vocacional (DGFV), May Coles, M., Evaluating the impact of reforms of vocational education and training: examples of practice. In: Descy, P. and M. Tessaring (eds). Evaluation of systems and programmes Third report on vocational training research in Europe: background report. Luxembourg: Office for Official Publications of the European Communities, 2004 (Cedefop Reference series, 57). 36. Collins, T. et al. Framework Implementation and Impact Study, Available from the Internet: c [retrieved on ]. 37. Conein, S., Ambos, I., Nuissl, E. Lernende Regionen wissenschaftlich begleitet in Nuissl, E., Schiersmann, Ch., Siebert, H. (Hrsg.) Literatur und Forschungsreport Weiterbildung Nr. 50 Dezember 2002 W. Bertelsmann Verlag p, Connie White, Mary Ziegler, and Beth Bingman Families First: Implications of Welfare Reform for Tennessee Adult Basic Education Council, Council conclusions on a strategic framework for European cooperation in education and training (ET 2020). Brussels: Council of the EU (OJ C 119 of , p. 2-10). 40. Council, Council Resolution of 27 June 2002 on lifelong learning (OJ C 163, ), Council, Council Decision 2005/600/EC of 12 July 2005 on guidelines for the employment policies of the Member States (OJ L 205, ), Council, Council conclusions of 25 May 2007 on a coherent framework of indicators and benchmarks for monitoring progress towards the Lisbon objectives in education and training (2007/C 311/10), Council, Council Decision 2008/618/EC of 15 July 2008 on guidelines for the employment policies of the Member States (OJ L 198, ), 2008a. 44. Council, Council conclusions of 22 May 2008 on adult learning (OJ C 140, ), 2008b. 45. Council, Council conclusions of 12 May 2009 on a strategic framework for European cooperation in education and training ( ET 2020 ) (OJ C 119, ), Delors, J., Learning: the Treasure Within. a Report to UNESCO of the International Commission on Education for the Twenty-first Century, Deutsches Institut für Erwachsenenbildung, Deutsches Institut für Wirtschaftsforschung Berlin, Universität Duisburg-Essen, Ludwig-Maximilians-Universität München Ausgewählte Ergebnisse der wissenschaftlichen Begleitung 138

141 des Programms Lernende Regionen Förderung von Netzwerken, Ding, W., S. F. Lehrer, J. N. Rosenquist and J. Audrain- McGovern, The impact of poor health on education: new evidence using genetic markers, National Bureau of Economic Research, Working paper No , Directorate General for the Budget. Evaluating EU Activities: A Practical Guide for the Commission services, Luxembourg: Office for Official Publications of the European Communities, Dobischat, R., Nuissl, E., Tippelt, R., Hagen, K. Regionale Bildungsnetze. Ergebnisse zur Halbzeit des Programms Lernende Regionen Förderung von Netzwerken, Docteur, E. and H. Oxley, Health-care Systems: Lessons from the Reform Experience. Economics Department Working Papers No Paris: OECD, Dohmen, D. et al. Current trends in the demand-led financing of further training in Europe a synopsis. Berlin: Institute for Education and Socio-Economic Research and Consulting, Dooren, W. V. et al, Institutional drivers of efficiency in the public sector. OECD, GOV/PGC(2007)16/ANN, Dronkers, J. and P. Robert, Differences in Scholastic Achievement of Public, Private Government-Dependent, and Private Independent Schools: A Cross-National Analysis. Educational Policy, Vol. 22, No. 4, Duvekot et al (eds.), Managing European Diversity in Lifelong Learning: the many perspectives of the Valuation of Prior Learning in the European workplace. HAN University, Foundation EC-VPL, Hogeschool van Amsterdam, ECOTEC Research and Consulting Limited, The Interim Evaluation of the Grundtvig Action of the SOCRATES Programme ( ) covering the years Available from the Internet: ation/2004/grundtvig/grundintrep_en.pdf [retrieved on ]. 57. ECOTEC, A European inventory on validation of non-formal and informal learning: Lithuania ECOTEC, Joint Report on the Final Evaluation of Socrates II, Leonardo da Vinci II and elearning Available from the Internet: ing/2007/joint/joint_en.pdf [retrieved on ]. 59. Ekström, E. The value of a third year in upper secondary vocational education - Evidence from a piloting scheme, Available for the Internet: [retrieved on ]. 60. Emminghaus, Ch., Tippelt R. (Hrsg.) L ebenslanges Lernen in regionalen Netzwerken verwirklichen. Abschließende Ergebnisse zum Programm 'Lernende Regionen - Förderung von Netzwerken'. Bertelsmann Verlag Bielefeld, Ernst & Young, Barbier Frinault & Associés. External Interim Evaluation of the Leonardo da Vinci II programme. Summary and Conclusions Available from the Internet: ing/2003/leoii_interim/leoiiintsum_en.pdf [retrieved on ]. 139

142 62. Esping-Andersen, G. Who is Harmed by Labour Market Regulations? Quantitative Evidence, in Gosta Esping- Andersen, Marino Regini (eds.), Why Deregulate Labour Markets?, Oxford, New York: Oxford University Press, Estavez-Abe, M., Iversen, T., Soskice, D. Social Protection and the Formation of Skills: A Reinterpretation of the welfare State, in Peter A. Hall, David Soskice (eds.), Varieties of Capitalism: the Institutional Foundations of Comparative Advantage, Oxford University Press, Euroguidance Lithuania, EU Leonardo da Vinci programme Coordination Support Foundation, Education and Career Guidance in Lithuania European Association for Education of Adults, Adult education trends and issues in Europe. Brussels: EAEA, European Commission, A memorandum on lifelong learning. Commission Staff Working Paper, SEC(2000) 1832, European Commission, Making a European Area of Lifelong Learning a reality. Communication, COM(2001) 678 final, European Commission, DG Employment, Social Affairs & Equal Opportunities, General and national evaluations of the European Social Fund. 69. European Commission, Adult learning: It is never too late to learn. Communication, COM(2006) 614 final, 2006a. 70. European Commission, Efficiency and equity in European education and training systems. Communication, COM(2006) 481 final, 2006b. 71. European Commission, Evaluation of the open method of coordination for social protection and social inclusion: a synthesis of replies by Member States and other actors to an evaluation questionnaire on the Open Method of Coordination in the fields of social inclusion and adequate and sustainable pensions. Commission Staff Working Document, SEC(2006) 345, 2006c. 72. European Commission, Action Plan on Adult learning: It is always a good time to learn. Communication, COM(2007) 558 final, 2007a. 73. European Commission, Progress towards the Lisbon objectives in education and training: indicators and benchmarks. Commission Staff Working Document, 2007b. 74. European Commission, Draft 2008 joint progress report of the Council and the Commission on the implementation of the "Education & Training 2010 Work Programme", {COM(2007) 703 final}, 2008a. 75. European Commission, Proposal for a recommendation of the European Parliament and of the Council on the establishment of a European Quality Assurance Reference Framework for Vocational Education and Training Impact Assessment. Commission Staff Working Document, SEC(2008) 440, 2008b. 76. European Commission, An updated strategic framework for European cooperation in education and training Impact Assessment. Commission Staff Working Document, SEC(2008) 3047, 2008c. 77. European Commission, An updated strategic framework for European cooperation in education and training. Communication, COM(2008) 865 final, 2008d. 78. European Council, Presidency conclusions, Essen, 9-10 December,

143 79. European Council, Presidency conclusions, Luxembourg, November European Council, Presidency conclusions, Lisbon, March, European Council, Presidency conclusions, Brussels, March, European Guidance and Counselling Research Forum, Overview of selected European guidance systems Available from the Internet: [retrieved on ]. 83. European Parliament, Council, Decision No 2493/95/EC of the European Parliament and of the Council of 23 October 1995 establishing 1996 as the 'European year of lifelong learning' (OJ L 256, ), European Parliament, Council, Recommendation of the European Parliament and of the Council 2006/962/EC of 18 December 2006 on key competences for lifelong learning (OJ L 394, ), 2006a. 85. European Parliament, Council, Decision No 1720/2006/EC of the European Parliament and of the Council of 15 November 2006 establishing an action programme in the field of lifelong learning (OJ L 327, ), 2006b. 86. EURYBASE Eurydice database on education systems in Europe. Country descriptions for Austria, Belgium, Bulgaria, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Liechtenstein, Lithuania, Malta, Netherlands, Norway, Poland, Portugal, Romania, Slovakia, Spain, Sweden, Turkey and United Kingdom. Brussels: European Commission, Directorate-General for Education and Culture, Eurydice, Non-Vocational Adult Education in Europe: Executive Summary of National Information on Eurybase, Working Document, January EURYDICE - National summary sheets on education systems in Europe and ongoing reforms. Country descriptions for Austria, Belgium, Bulgaria, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Liechtenstein, Lithuania, Luxembourg, Malta, Netherlands, Norway, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, Sweden, Turkey and United Kingdom. Brussels: European Commission, Directorate-General for Education and Culture, FETAC. Implementation and Impact of the National Framework of Qualifications in the Further Education and Training Sector, Available from the Internet: in_doc30sep08rd.pdf [retrieved on ]. 90. Fagerberg J., D. C. Mowery and R. R. Nelson (eds.), The Oxford Handbook of Innovation. Oxford: Oxford university Press, Fejes, A., Validation of Non-formal and Informal Learning. Statements and Comments. A paper presented during a peer review on 8-9 November 2007 in France Fragoulis, H., J. R. Masson and V. Klenha, Improving Opportunities for Adult Learning in the Acceding and Candidate Countries of Central and Eastern Europe. European Journal of Education, Vol. 39, No. 1, 2004, pp

144 93. Gasskov, V. Government interventions in private financing of training. International Labour Organisation, Gornitzka, Å., The Open Method of Coordination as practice - A watershed in European education policy?. Arena - Centre for European Studies, University of Oslo, working paper No. 16, December Gornitzka, Å., The Open Method of Coordination in European Education and Research Policy: Animating a Label. Arena - Centre for European Studies, University of Oslo, paper to be presented at European Union Center of Excellence and School of Education, University of Wisconsin, Friday 13th October Gustavsson, B., Swedish Adult Education in a Critical Perspective. In Lifelong Learning in Europe, 2/ Häkkinen, U. and I. Joumard, Cross-country analysis of efficiency in OECD health care sectors: options for research, OECD, Economics department working papers, No. 554, Hannan, D. F., D. Raffe and E. Smyth, Cross National Research on School to Work Transitions. Paris: OECD, Hansen, M. N., Rational Action Theory and Educational Attainment. Changes in the Impact of Economic Resources. European Sociological Review, Vol. 24 No. 1, 2008, pp Hargroves, K. and M. H. Smith (eds.), The natural advantage of nations business opportunities, innovation and governance in the 21st century. The Natural Edge Project, Earthscan, London, Hatton-Yeo, A. and T. Ohsako (eds.), Intergenerational Programmes: Public Policy and Research Implications. An international Perspective. UNESCO, Howell, D. R. Introduction, in David R. Howell (ed.), Fighting Unemployment: The Limits of Free Market Orthodoxy, Oxford, New York: Oxford University Press, Humburg, M., The open method of coordination and European integration: The example of European education policy, Berlin working paper on European Integration No. 8, July IDEC SA, Evaluation of ESF contribution to employment, inclusion and education & training policies through the support to systems and structures: Final report (version 1). September, International Monetary Fund. Unemployment and Labour Market Institutions: Why Reforms Pay Off, World Economic Outlook, Washington, DC: IMF, Instituto de Gestão do Fundo Social Europeu. National System for the Recognition, Validation and Certification of Competences (RVCC), Available from the Internet: [retrieved on ] Jacobi, L. and J. Kluve. Before and After the Hartz Reforms: The Performance of Active Labour Market Policy in Germany, Available from the Internet: [retrieved on ] Joumard, I., C. André, C. Nicq and O. Chantal, Health Status Determinants: Lifestyle, Environment, Health Care Resources and Efficiency. Economics Department Working Papers No Paris: OECD,

145 109. Joseph, G., O. Pierrard and H. R. Sneessens, Job Turnover, Unemployment and Labour Market Institutions, IZA Discussion Paper, No. 835, July Jotautienė, M., B. Janiūnaitė, N. Večkienė, PHARE programos diegimo, modernizuojant Lietuvos profesinio rengimo sistemą, metodologiniai aspektai. In Socialiniai mokslai, 1999 No.3 (20) Kalleberg, A. L., Nonstandard Employment Relations: Part- Time, Temporary and Contract Work. Annual Review of Sociology, Vol. 26, 2000, pp Kavaliauskienė, E., G. Žibėnienė, The External Quality Assessment of Lithuanian Labour Market Vocational Training Keogh, H., The State and Development of Adult Learning and Education in Europe, North America and Israel. Regional Synthesis Report. Hamburg: UNESCO Institute for Lifelong Learning, Kerckhoff, A. C., Education and Social Stratification Processes in Comparative Perspective. Sociology of Education, Vol. 74, Extra Issue: Current of Thought: Sociology of Education at the Dawn of the 21st Century, 2001, pp Knott, J. H. and A. A. Payne, The Impact of State Governance Structures on Management and Performance of Public Organisations: A Study of Higher Education Institutions. Journal of Policy Analysis and Management, Vol. 23, No. 1, Kühnlein, G., Klein, B. Mehr Eigenverantwortung, mehr Markt, mehr Effizienz? Erfahrungen bei der Neuausrichtung der beruflichen Weiterbildung, Available from the Internet: df 117. Labour and Social Research Institute, The Survey of Long- Term Unemployed [in Lithuania] Labour and Social Research Institute, The Study of Social and Economic Effectiveness of Vocational Labour Market Training [in Lithuania] Laffan, B. and C. Shaw, Classifying and Mapping OMC in different policy areas. University College Dublin, July Lange, B. and N. Alexiadou, New Forms of European Union Governance in the Education Sector? A Preliminary Analysis of the Open Method of Coordination. European Education Policy Research Journal, Volume 6, Number 4, Laužackas, R., V. Tūtlys, I. Spūdytė, Analysis of the Current State of Qualifications in Lithuania. Lithuanian Labour Market Training Authority, Libert, I., Improving Opportunities for Adult Learning - the Construction Industry in France. European Journal of Education, Vol. 39, No. 1, 2004, Lipinska, P., E. Schmid and M. Tessaring, Zooming in on Reassessing vocational education and training. Cedefop, Lithuanian Centre for Adult Education and Information, Qualitative Study on Lifelong Learning and its Results Lithuanian Labour Market Training Authority. Activity report Loeb, I. H. Development and Change in Swedish Municipal Adult Education: occupational life history studies and four 143

146 genealogies of context in Policy Futures in Education, Volume 5, Number 4, Lyly-Yrjänäinen, M., Who needs up-skilling? Low-skilled and low-qualified workers in the European Union. European Foundation for the Improvement of Living and Working Conditions, Mandl, U., A. Dierx and F. Ilzkovitz, The Effectiveness and Efficiency of Public Spending. Economic Papers 301, February Brussels: Directorate-General for Economic and Financial Affairs, Martinaitis, Ž. Europietiški vaistai Lietuvos darbo rinkai: neskanūs ar netinkami?, in Vitalis Nakrošis (ed.), Europos Sąjungos Lisabonos strategijos įgyvendinimas Lietuvoje: atvirojo koordinavimo metodo taikymo analizė, Vilnius:Vilniaus universiteto leidykla, Martinez, M. C., Understanding State Higher Education Systems: Applying a New Framework. The Journal of Higher Education, Vol. 73, No. 3, 2002, pp Masso, J. and R. Emets. Macro-level Labour Market Flexibility in the Baltic States, in Tiiu Paas, Raul Emets (eds.), Labour Market Flexibility, Flexicurity and Employment: Lessons of the Baltic States, New York: Nova Science Publishers, Masson, J. R. (The European Training Foundation), Financing vocational education and training in the EU new member states and candidate countries: recent trends and challenges. Luxembourg: Office for Official Publications of the European Communities, Matějů P., M. L. Smith, P. Soukup and J. Basl, Determination of College Expectations in OECD Countries: The Role of Individual and Structural Factors. Czech Sociological Review, No. 6, 2007; 134. Milton, J., K. E. Watkins, S. S. Studdard and M. Burch, The Ever Widening Gyre: Factors Affecting Change in Adult Education Graduate Programs in the United States. Adult Education Quarterly, Vol. 54, No. 23, 2003, pp Mijs, J, J, B. and H. G. van de Werfhorst, Educational Differentiation and Inequality: The Netherlands in Comparative Perspective, 2008; 136. Ministry of education and science, Ministry of social security and labour, Republic of Lithuania, Strategy of Vocational Guidance Ministry of Education and Science, Ministry of Social Security and Labour, Republic of Lithuania, The Action plan for the Implementation of the Strategy for Vocational Guidance Ministry of social security and labour, Republic of Lithuania. Decree of the Minister of social security and labour No. 103 Regarding the amendment of the decree of Minister of social security and labour, dated 10 October 2000, No December Morgan, J. and A. Mourougane, What Can Changes in Structural Factors Tell us About Unemployment in Europe, Scottish Journal of Political Economy, Vol. 52, No. 1, February Müller, W., Education and Youth Integration into European Labour Markets. International Journal of Comparative Sociology, Vol. 46, No. 5-6, 2005, pp Müller, W. and K. Wolfgang, Social Selection in Educational Systems in Europe. European Sociological Review, Vol. 9, No. 1, 1993, pp

147 142. Müller, W. and Y. Shavit, The Institutional Embeddedness of the Stratification Process. In: Y. Shavit and W. Müller (eds.), From School to Work. A Comparative Study of Educational Qualifications and Occupational Destinations. Oxford: Clarendon Press, 1998, p Myers K. and P. de Broucker, Too Many Left Behind: Canada s Adult Education and Training System. Research report W 34, Canadian Policy Research Networks Inc., National Reports on the Development and State of the Art of Adult Learning and Education for the 6th International Conference on Adult Education (CONFINTEA VI): Austria, Belgium, Croatia, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Latvia, Lithuania, Netherlands, Norway, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, Sweden, The former Yugoslav Republic of Macedonia, Turkey, United Kingdom Nelissen, E., European Inventory on Validation of Informal and Non-Formal Learning. Sweden NCVER (National Centre for Vocational Education Research). Funding and financing vocational education and training: research readings. Adelaide: NCVER, Nieuwenhuis, L. and H. Shapiro, Evaluating systems change policies in VET: learning from Danish and Dutch cases. In: Descy, P. and M. Tessaring (eds). Evaluation of systems and programmes Third report on vocational training research in Europe: background report. Luxembourg: Office for Official Publications of the European Communities, 2004 (Cedefop Reference series, 57) Normantas, E. Adult education in Lithuania three years after the beginning of its reform. Available from the Internet: [retrieved ] Nuissl, E., Dobischat, R., Hagen, K., Tippelt, R. Regionale Bildungsnetze. Ergebnisse zur Halbzeit des Programms Lernende Regionen Foerderung von Netzwerken in. Nolda, S., Nuissl, R., Tippelt, R. (Hg.) Theorie und Praxis der Erwachsenenbildung, OECD - Thematic Reviews on Adult Learning. Background report for Austria, Denmark, Finland, Germany, Netherlands, Norway, Poland, Portugal, Spain, Sweden, Switzerland, United Kingdom. Paris: OECD, OECD - Thematic Review on Recognition of Non-formal and Informal Learning. Background report for Austria, Belgium, Czech Republic, Denmark, Germany, Greece, Hungary, Iceland, Ireland, Italy, Netherlands, Norway, Slovenia, Spain, Switzerland, United Kingdom. Paris: OECD, OECD. Thematic Review on Adult Learning: Sweden. OECD background report. Paris: OECD publishing, OECD. Economics and finance of lifelong learning. Paris: OECD publishing, 2001a OECD. Thematic Review on Adult Learning: Sweden, country note. Paris: OECD publishing, 2001b OECD. Reviews of National Policies for Education. Lithuania. Paris: OECD publishing, OECD. Beyond rhetoric: adult learning policies and practices. Paris: OECD publishing, 2003a OECD. Upgrading workers skills and competencies (Chapter 5). In OECD Employment Outlook Paris: OECD publishing, 2003b. 145

148 158. OECD. Promoting Adult Learning. Paris: OECD publishing, OECD. Government at a Glance Paris: OECD publishing, OECD/Statistics Canada, Learning a Living: First Results of the Adult Literacy and Life Skills Survey, Paris: OECD, Opheim, V. and H. Helland, Evaluation of the Competence Reform in Norway: Access to Higher Education Based on Non-formal Learning. In: OECD, Higher Education Management and Policy, Vol. 18, No. 2, Otero, M. S. et al, First evaluation of Europass. A Final Report to the Directorate General Education and Culture of the European Commission, ECOTEC, Parkes, D. (The European Training Foundation). Financing vocational education and training: priorities and mechanisms in South-Eastern Europe. Luxembourg: Office for Official Publications of the European Communities, Pfeffer, F. T., Persistent Inequality in Educational Attainment and its Institutional Context. European Sociological Review, Vol. 24, No. 5, 2008, pp Pollitt C. and G. Bouckaert, Public Management Reform: A Comparative Analysis. Oxford: Oxford University Press, 2000; 166. Pont, B., Improving Access to and Participation in Adult Learning in OECD Countries. European Journal of Education, Vol. 39, No. 1, Radaelli, C. M. and A.C.M. Meuwese, Better Regulation in the European Union: The political economy of impact assessment Radaelli, C. M. and F. De Francesco, Regulatory impact assessment in Cave et al (eds.) The Oxford Handbook of Regulation Regeringen, Regeringens proposition 2000/01:72 Vuxnas lärande och utvecklingen av vuxenutbildningen Available from the Internet: f [retrieved on ] Regeringskansliet, The Swedish study support system Available from the Internet: [retrieved on ] Reischmann, J., International and Comparative Adult Education: A German Perspective. PAACE Journal of Lifelong Learning, Vol. 13, pp , Rodrigues, M. J. (ed.), Europe, Globalization and the Lisbon Agenda. Cheltenham: Edward Elgar Publishing, Ryan, Paul, The School-to-Work Transition: A Cross- National Perspective. Journal of Economic Literature, Vol. 39, No. 1, 2001, pp Rubenson, K., Adult Education Policy in Sweden, in Adult Learners, Education and Training ed. by Richard Edwards, Sandy Sieminsky and David Zeldin, Saar, E., M. Unt and I. Kogan, Transition from Educational System to Labour Market in the European Union: A Comparison between New and Old Members. International Journal of Comparative Sociology, Vol. 49, No. 31, 2008, pp Scarpetta, S. Assessing the Role of Labour Market Policies and Institutional Settings on Unemployment, OECD Economic Studies, No. 26, 1996/1. 146

149 177. Scarpetta, S., Labour Market Reforms and Unemployment: Lessons from Experience of the OECD Countries, Inter- American Development Bank, Office of the Chief Economist Working Paper No. 382, Washington DC, Scherer, S., Patterns of Labour Market Entry Long Wait or Career Instability? An Empirical Comparison of Italy, Great Britain and West Germany. European Sociological Review, Vol. 21, No. 5, Schimank, U., A Comparative Perspective on Changes in University Governance in Europe, Schlotter, M., G, Schwerda and L. Wößmann, The Future of European Education and Training Systems: Key Challenges and Their Implications. Analytical report, European Expert Network on Economics of Education (EENEE), Schneider, H., Brenke, K., Kaiser, L., Steinwede, J., Jesske, B., Uhlendorff, A. Evaluation der Maßnahmen zur Umsetzung der Vorschläge der Hartz-Kommission, Segerholm, C. The Swedish Adult Education Initiative in a Municipal Context. An evaluation. Paper presented at the ECER-conference in Lille, September 5-8, Shaw, N. et al. Evaluation of the Union Learning Fund ( ). York Consulting Ltd, Research Report RR789, Available from the Internet: df [retrieved on ] Social and Cultural Planning Office, Public Sector Performance: An international comparison of education, health care, law and order and public administration. Hague: Social and Cultural Planning Office, Shavit, Y. and H.-P. Blossfeld (eds.), Persistent Inequality. Changing Educational Attainment in Thirteen Countries. Boulder, CO: Westview Press, Skolverket, The Adult Education Initiative. An overall investment in adult education Stenberg, A. Comprehensive Education for the Unemployed - Evaluating the Effects on Unemployment of the Adult Education Initiative in Sweden. Umeå University, 2003a Stenberg, A. An Evaluation of Adult Education Initiative Relative Labour Market Training. Umeå University, 2003b Stenberg, A. Short Run Effects on Wage Earnings of the Adult Education Initiative in Sweden. Umeå University, 2003c Stenberg, A. The Adult Education Initiative in Sweden - Second Year Effects on Wage Earnings and the Influence on Branch Mobility. Umeå University, 2003d Stenberg, A. Comprehensive Education or Vocational Training for the Unemployed?. International Journal of Manpower, 28(1) Stenberg, A. Upgrading the Low Skilled: Is Public Provision of Formal Education a Sensible Policy?. SOFI Working Paper 2009/ Stenberg, A. and O. Westerlund, Does comprehensive education work for the long-term unemployed? Labour Economics, Strauss, J. and D. Thomas, Health, nutrition and economic development, Journal of Economic Literature, Vol. 36, pp , Swedish Institute for Social Research Can Adult Education Delay Retirement from the Labour Market?. Working Paper 6 /

150 196. Tangerås, T.P., Efficiency Problems with the Adult Education Initiative a Rudimentary Analysis, Tashakkori, A., Teddlie, C. Handbook of Mixed Methods in Social and Behavioural Research. SAGE, Tavistock Institute, The Evaluation of Socio-economic Development: The Guide Teresevičienė, M., V. Zuzevičiūtė, R. Kuncaitis, A. Rutkienė. Suaugusiųjų mokymasis Lietuvoje: aprėptis, poreikiai ir pasiūla. Andragogas-praktikas, The Education System in Lithuania. Eurydice paper, 2006/ Treaty of Amsterdam establishing the European Community (OJ C 340, ) Treaty on European Union - Title II: Provisions amending the Treaty establishing the European Economic Community with a view to establishing the European Community - Article G - Article G D 25 (OJ C 191, ) Turner, J. A., C. Toft, H. A. Witte, Lifelong Learning for an Ageing Population: Lessons from Scandinavia Tūtlys, V., The Vocational Education and Training System in Lithuania. Presentation 12 November 2007, Kaunas UNESCO; Titmus, C. et al. Terminology of adult education = Terminologia de la educación de adultos = Terminologie de l éducation des adultes. Paris: Unesco, Wells, R. L., D. N. Goetz, Adult Education in the Union of Soviet Socialist Republics, Werquin, P., Propositions for a typology of recognition systems. OECD, EDU/EDPC(2007)25, Wilkinson, R. and M. Marmot (eds.), Social determinants of health: the solid facts, 2 nd edition, World Health Organisation, Wolbers, M. H. J., Patterns of Labour Market Entry A Comparative Perspective on School-to-Work Transitions in 11 European Countries. Acta Sociologica, Vol. 50, No. 3, 2007, pp Wood, H. and S. Moore. An Evaluation of the UK Union Learning Fund - its Impact on Unions and Employers. Working Lives Research Institute, London Metropolitan University: Final report, No: VP/2003/012 {REF /EMPL/F/4/ARC(2004)D/14341}, June Available from the Internet: [retrieved on ] World health organisation, Health, Economic Growth, and Poverty Reduction, The report of working group I of the Commission on macroeconomics and health, WRC Social and Economic Consultants Ltd. Report to the Minister of Education on an Evaluation of the Vocational Training Opportunities Scheme, Available from the Internet: [retrieved on ] York Consulting, TUC Learning Services: Evaluation of Building Opportunities in Workplace Training Final Report, May Available at: [retrieved on ] Zajda J. (ed.). International Handbook on Globalisation, Education and Policy Research. London: Kluwer, Zeitlin, J., The Open Method of Co-ordination and reform of national social and employment policies: influences, mechanisms, effects, to appear in Heidenreich, Martin and 148

151 Jonathan Zeitlin (eds.) Changing European Employment and Welfare Regimes: the Influence of the Open Method of Co-ordination on National Reforms, Routledge/EUI Studies in the Political Economy of Welfare, London: Routledge, forthcoming

152 ANNEX 1: QUANTITATIVE EVIDENCE ON FACTORS INFLUENCING ADULT LEARNING The first chapter of this annex provides a short overview of the progress of the EU27 Member States, EFTA-EEA and candidate countries (in adult learning and education attainment and explores important interrelationships between the two factors. The second and third chapter outline the results of statistical analysis regarding the influence of policy as well as contextual factors on adult learning and education attainment. The fourth and final chapter determine the most significant factors that influence adult participation in learning and educational attainment of the population. Where relevant the analysis explores the influences on learning participation and attainment of lower-qualified adults (i.e. those working in ISCO 4-9 occupations). When the results of the analysis for this particular group do not differ substantially from the results for all adults aged 25-64, they are not provided in the report. 1. Methodology of quantitative analysis This chapter provides the methodology of the quantitative analysis. We start with presenting our research questions and framework for analysis. Then we present our dependent, independent and context variables and how they will be measured. Finally we present our research methods, which are used to answer our research questions Analytical framework The main empirical problem for our research is the slow progress of the European Union and its close partners from the EEA-EFTA and candidate countries in achieving the Lisbon benchmark of adult participation in learning of 12.5% by Moreover the adult learning participation disparities among the countries concerned are immense and there are only a few examples of the countries lagging behind catching up with the leading ones. Large disparities also remain in terms of overall educational attainment of the adult population (see chapter 1.1 for more details). As a vital element of a competitive economy providing more and better jobs the participation of the adult population in lifelong learning has been high on public policy agenda especially since the adoption of the Lisbon strategy both at European and national levels (see also chapter 1.2 for more details). A number of reform efforts and new policy initiatives in adult learning, a broader education system and outside of it were undertaken to boost the investment in human capital and improve its efficiency. The main aim of our inquiry in this part of the report is to understand whether and to what extent policies (including their reforms) implemented over the past decade have been responsible for the achieved progress in adult learning and educational attainment. There is no comparable information across the 150

153 countries analysed about their policy reforms. Instead we analyse their selected education, labour market and health care policy characteristics, which are covered by the official statistics, and correlate changes in these characteristics to changes in adult learning and educational attainment as well as to changes in the general social and economic context. Although changes in the policy characteristics most likely stand for reforms, we cannot be sure about their properties and we are unable to pursue analysis here at this level of detail. Our analysis might reveal, however, which policy characteristics for the whole group of countries are associated with higher adult learning participation or educational attainment. Our main hypothesis is that the level of adult learning in different countries depends on two major groups of factors: Policy characteristics, and General social and economic context, in which policies are being implemented. Many policies might influence adult learning directly or indirectly. For our analysis we have selected education, labour market and health policies. They arguably have the most immediate and important effects on adult learning compared to other policies. Individuals have limited time and other resources (money, health, etc.), which they allocate according to their individual circumstances and preferences. Individual choices obviously have opportunity costs. Once individuals decide how much of their free time they are willing to exchange for work income, they limit time available for learning. If individuals have poor health, their working and learning abilities could be constrained. Once individuals have more income than they need to meet their immediate needs, some of it can be saved and invested in learning, better health or even to generate more income. We have selected 4 broad groups of factors for our analysis of social and economic context. First of all a stable macroeconomic environment and healthy public finances might be vital for availability of funding for adult learning and related policies. Secondly, higher levels of social cohesion might be important for the distribution of learning opportunities among different social groups, especially those with lower income and other disadvantages. Thirdly, the extent to which the economy is innovative and oriented at producing goods and services, which require more sophisticated knowledge and skills, could be both cause and effect of greater adult participation in learning and educational attainment. Finally, the quality of infrastructure and efficient use of energy resources could also have significant influences on our dependent variables. The analytical framework for this part of the report is summarised in figure

154 Figure 1: Analytical framework Policy characteristics Education Labour market Health Adult learning and educational attainment Social and economic context: Macroeconomic environment; Social cohesion; Innovation and research; Environment and transport. Source: compiled by the authors The scope of our analysis will inevitably be limited by the availability of evidence. Most internationally comparable statistical data, which would include many of the EU, EFTA-EEA and candidate countries, and which could be used in tracking potential reform outcomes, is only available for 1997 and (or) later years Dependent variables In this report we use the following two main dependent variables: participation of adults in education and training and adult education attainment level. Participation of adults in education and training is our main dependent variable, which measures the level of adult learning, both formal and non-formal. We operationalise this variable as the percentage of the population aged who received some education and training (both formal and non-formal) during the previous four weeks (this is a variable EDUC4WN in the EUROSTAT Labour Force Survey and a structural variable EM051 presented in the EUROSTAT statistics database). Participation of adults in education and training is a general variable. It includes two different types of learning: formal and non-formal: Participation of adults in formal education and training, i.e. the percentage of the population aged who have been students or apprentices (including those on holidays) in regular education during the previous four weeks (variable EDUCSTAT in EUROSTAT Labour Force Survey); and Participation of adults in non-formal education and training, i.e. the percentage of the population aged who have attended any courses, seminars, conferences or received private lessons or instructions outside the regular education system (i.e. taught learning activities, job-related or for personal purposes) within the last four weeks (a variable COURATT in EUROSTAT Labour Force Survey). Both variables indicated above show the levels of participation of adults in learning. They do not reveal fully the accessibility of this education and training. Therefore, we complement our main 152

155 dependent variable, participation of adults in education and training, with the additional one on intensity of learning. Intensity of education and training shows the quantity of education and training received by those who participate in learning. We operationalise this indicator as the number of hours spent on all taught (non-formal) learning activities within the last four weeks (variable COURLEN in EUROSTAT Labour Force Survey). Data on the intensity of formal education and training is available for a small number of countries and only for 2007, the only reference year to date of Adult Education Survey. Therefore, in our analysis we mainly use the variable for non-formal education and training. Adult education attainment level variable measures the competence level (not the quality) of the corresponding population. Learning is a life-cycle process, thus achievement at one level builds the foundation for the attainment of skills and competences at the next level. Levels attained in initial education, therefore, have a significant impact on whether an individual participates or not in adult education and training. Adults with the lowest education attainment are the least likely to participate in adult education and training. This is the double-edged sword that exacerbates educational inequalities due to the so-called Matthew effect whereby individuals with the highest levels of initial education and training are the most likely to participate in adult education and training (Eurydice, 2007). We operationalise adult education attainment level as the percentage of the population aged having completed at least upper secondary education (variable SO011 presented in EUROSTAT statistics database). Most countries made significant progress and considerably increased the general education attainment level of the population, therefore, we use an additional variable, which is the percentage of the population aged25-64 having completed at least tertiary education (variable HATLEVEL in EUROSTAT Labour Force Survey). It could possibly better measure the progress in adult education attainment across the countries concerned. But, we use both indicators to better measure adult education attainment level of population Independent variables Education policies Education policies have the most visible and obvious effects on adult learning. They directly regulate, finance and provide learning opportunities. However, policies of different countries are not equally well adapted to the particular needs of adult learners, which results in considerable differences in terms of accessibility and quality of learning supply. Education systems are characterised by an increasingly large number of groups of policy features. For the purposes of this report we have selected the following most important policy characteristics of an education system: 1. Standardisation; 2. Stratification; 3. Vocational specificity; 4. Flexibility in provision of learning services; 5. Validation of prior learning; 6. Information, guidance and counselling; 7. Governance of institutions; 153

156 8. Financing of education and training; and 9. Human resources. Each of the selected policy characteristics is described below Standardisation Standardisation refers to the degree to which governments create conditions (e.g. teachers education, education financing, curricula, etc.) and the control mechanisms (nationwide testing, schoolleaving examinations, etc.) to achieve certain nationwide standards of quality in the education provided by different schools (Matějů, Soukup, Basl and Smith 2007). Although one may notice a tendency that standardized educational provisions appear in centralised public administration systems, this is not always true. Educational provisions may be standardized and their coherence may be ensured by a greater emphasis on coordination incentives: by establishing coordinating bodies, responsible for the uniformity and comparability of structures, institutions, curricula, and the leaving certificates of various education providers. For example, in Germany the Standing Conference of State Education Ministers ensures that academic and professional training meets the same standards throughout the country, which means that the degree of standardisation in Germany is high. In the standardized systems employers can rely on information given by (standardized) certificates and there is no need to screen, as well as train individuals entering the labour force (Allmendinger 1989). Therefore, the organisation of the schooling and vocational system has meaningful and far-reaching impact on the labour market entry, matching between supply and demand for qualifications, as well as employment mobility Stratification Stratification refers to the numbers and types of transitions available to the next schooling level. In a stratified system the programme offerings in the types of secondary schools are associated with different degrees of access to opportunities for additional, more advanced schooling. Stratified educational systems are characterized as embracing a tight linkage between the education system and differentiated occupational structure. In non-stratified systems, conversely, the coupling between the education system and occupational structure is loose (Allmendinger 1989). Academics argue that students in education systems with high differentiation within secondary education are less likely to invest in higher education to distinguish themselves from other students in the labour market. Previous analysis shows that in the Netherlands and Germany, for example, students with secondary vocational schooling perform relatively well in the labour market (Muller and Shavit, 1998). Mijs and Werfhorst similarly argue that if selection occurs at an early age, a smaller group of students will continue to tertiary education (Mijs & Werfhorst, 2008). Therefore, it appears that in countries with high levels of stratification schooling is a better predictor for the level of job attainment than in countries with a low degree of stratification. Consequently, in highly stratified educational systems the transition from education to the labour market is easier and earlier, which means that students spend less time in formal education and training and achieve lower education attainment. 154

157 Significant differences among countries may be noticed according to the integration level of their education systems. For example, in the Netherlands pupils are already grouped according to their abilities from the age of 12 into different education tracks for their basic secondary education. At later stages underperforming pupils can be downgraded into less demanding tracks, but the opposite is no longer possible. Thus, the Netherlands education system is categorized as being highly stratified. Systems in which pupils are grouped according to their abilities at an early stage of life tend to be less flexible in providing opportunities for those from less demanding education systems to enter additional, more advanced schooling. In the Scandinavian countries, as well as in Portugal, Poland, Estonia, Latvia, Slovenia,, youngsters, on the other hand,, do not move to secondary education until they are years old and the entire compulsory period of schooling (primary education, including the first phase of secondary education) takes the form of a single type of education for all children between the ages of 6 and The choice of a uniform provision for all pupils was prompted largely by a desire to provide equal opportunities and education of a high standard for as many pupils as possible (Social and Cultural Planning Office 2004). Furthermore, highly stratified education systems often result not only in lower (formal) adult learning and adult tertiary education attainment, but also in a lower equality of educational opportunity. Academics (Mijs and Werfhorst, 2008) argue that, for example, the low degree of stratification in Norway and Sweden is motivated by social democratic principles for reducing social inequality. The later the age of selection, the more the selection of lower class students will be based on ability rather than social origins. Consequently, social inequality between classes should be lower. If students were to make important educational decisions at an early age, lower class students would more often end up in lower levels of education as their parents would, at early age, to a larger extent determine these choices (i.e. life-course hypothesis ; see Shavit and Blossfeld, 1993) Vocational specificity Vocational specificity is the degree to which curricula are designed to prepare students for particular vocations (Matějů, Soukup, Basl and Smith 2007). Among the main dimensions of educational systems, vocational specificity may be an equally important factor, as it presumably contributes to shaping the structure of labour markets, as well as affecting an employee s demands while advancing in the labour market. In systems with high vocational specificity lower occupational mobility rates are observed, as the qualification-occupation match is high (Bukodi and Róbert, 2007). Vice versa, willingness to change occupation reliably signifies the need to choose another education track. Moreover, a high degree of vocational specificity can very often also indicate a high degree of system stratification, because schools providing training for specific occupations usually co-exist with schools preparing for further, more academic types of education at a higher level. In other words, high vocational specificity goes hand in hand with high stratification, usually within the so-called dual system of secondary education, such as that in Germany (Matějů, Soukup, Basl and Smith 2007). 155

158 Educational systems vary in the extent to which they offer curricula that are designed to prepare students for particular vocations and award credentials that are vocation specific. E. Bukodi and P. Róbert assume that those countries in which vocational education is truly specific are those in which a high proportion of pupils are in vocational schools and classify countries according to this criterion (Bukodi and Róbert 2007). However, according to R. Breen, those countries in which vocational education is truly specific are only a limited subset of those in which a high proportion of pupils are in vocational schools. The nature of vocational education differs greatly from country to country, while in some countries vocational schools are aimed more at providing education for low ability pupils and have a limited success in teaching occupationally useful skills. While measuring a country s vocational specificity R. Breen puts a strong emphasis on the clarity of the institutional link between vocational training and employers: to what extent employers are involved in determining what is taught and how much of that training involves extended periods of time actually working with the employer. In training systems where the degree of vocational specificity is high, certificates set a strong informal signal to employers and recruitment for positions available takes place directly and rapidly (Breen 2005). Educational systems that provide standardized, specific vocational qualifications, as well as having highly stratified arrangements, are generally described as being the most responsive to the demands of the labour market and providing clear educational signals to the employers while making labour market allocation decisions. In addition, a greater emphasis on specific skills and a closer link between schools and employers leads to an easier transition from education to the labour market Flexibility in provision of learning services According to the results of Eurobarometer and OECD surveys, time constraints were the most important reasons for not participating in learning programmes, as indicated by those adults who would like to participate (Cedefop 2003, 2004; OECD 2003b). A certain correlation between adults enrolment rates in full-time and parttime education and the role of part-time education was also found (Social and Cultural Planning Office 2004). Modular learning is another feature which enhances the flexibility and transferability of the education system, and ensures learners mobility. On the other hand, modular learning may pose a risk to the educational system s coherence. Therefore, in order to keep the educational system standardised and coherent enough, modularisation must be coupled with significant incentives to ensure standardized educational qualifications. For example, the vocational training system in Britain has been characterized as complicated partly because of its modular structure. Recent reforms helped to achieve clearer organisation, as well as greater standardisation in certification (Scherer 2005). Another example may be Spain where the National Institute for Qualifications is responsible for defining, drawing up and updating the National Catalogue on Vocational Qualifications and the corresponding Modular Catalogue for Vocational Training (EURYBASE, The education system in Spain, 2007/2008; Pont 2004). Another feature that signifies the flexibility of the education system, as well as better conditions for learners mobility, is the 156

159 existence of a credit transfer system. An education system with a highly elaborated credit transfer system provides the possibility to combine units from different programmes and, thus, ensures greater flexibility and responsiveness of the system to the needs of individual learners. The existence of a credit transfer system or any kind of system that allows the transfer of credits or the awarding of partial qualifications offers more possibilities for organising the recognition of non-formal and informal learning (Werquin 2007) Validation of prior non-formal and informal learning The recognition of prior learning and transfer of credit between institutions was one of the main objectives for the development of both sectoral and National Qualifications Frameworks in various countries. There is a wide understanding that individuals who had limited access to formal education and training or those who learned skills predominantly in the workplace are often disadvantaged in gaining access to further training or in securing employment. Workers with few or no formal qualifications are least likely to secure rewarding work opportunities. The formal recognition of workers competencies improves their employment and career prospects, as well as access to further learning. Today there is a wide range of practices of prior-learning recognition in countries. Some countries have long standing, largescale, fully institutionalized recognition systems, while others have systems that are at rudimentary stage of development (e.g. we use the classification of countries according to the development of validation systems provided in Cedefop 2008b) Information, guidance and counselling Information, guidance and counselling systems ensure the smooth functioning of both education and labour markets. They help adults to identify, assess and choose the most appropriate learning and work opportunities. Countries differ in their strategies for providing information to the public. Some of them have elaborate systems of guidance and counselling, while others are in the early stages of development. Ireland may serve as an example of good practice, where Local Adult Learning Boards coordinate and promote comprehensive information services (Pont 2004). Countries also vary according to the degree to which their guidance systems are decentralised and locally established. Some countries have information desks in every municipality (for example, Norway, Sweden, and Denmark). In addition, some countries develop selfservice on the Internet and by telephone in order to enhance services accessibility for the users. Countries also differ in the scope of target groups addressed by guidance and counselling services. For example, in Sweden employment offices provide a wide range of information and counselling to adults, mainly aimed at the unemployed. In Finland, conversely, all citizens whether they are students, at work, unemployed or outside the labour market are entitled to counselling and guidance at all ages (European Guidance and Counselling Research Forum, 2007) Governance of institutions The way in which educational institutions are being managed is another significant factor that has an impact on the performance of 157

160 education systems. OECD analysis has recognized increased management autonomy, i.e. less input controls on financial and human resource matters, as a significant institutional driver for the efficiency of public sector organisations. Autonomy is considered as a necessary condition for motivating managers to improve their performance by rewarding them for accomplishing goals or achieving expected results (Dooren et al 2008). J.H.Kott and A.A.Payne in their analysis of the impact of the state governance structures on management and performance of higher education institutions draw similar conclusion and argue that productivity and resources are higher at universities with a state-wide board that is more decentralized and has fewer regulatory powers (Knott and Payne 2004). It is also important to mention that some scholars tend to pay significant attention to ownership and competition as the factors leading to efficiency. However, the impact of these factors is contradictory and misleading, as there are as many studies showing that private ownership is not a guarantee of efficiency, as many studies suggesting the opposite (Dronkers and Robert 2008; Dooren et al 2008). Generally, a higher degree of deregulation and autonomy is taking place in countries that are more inspired by new public management ideas (for example, the United Kingdom, the Netherlands). Significant reforms to introduce more managerial self-governance by regulating the staffing, authority, and the tasks of those positions that make up central university leadership were carried out in Austria. At the other extreme of the continuum there are countries that follow a more conservative approach towards deregulation of their educational sector. Germany would be the best example here (Schimank 2005) Financing of education and training Literature often refers to education and training as a growth enhancing asset and a key pillar of the Lisbon Strategy. Furthermore, lack of financing is often indicated by individuals as one of the most important demand side reasons for nonparticipation in learning (Cedefop 2004). Therefore, a sustainable increase in education spending seems to be a vital institutional factor for economic growth and achievement of the Lisbon objective to increase the overall level of adult learning. However, the literature also shows that there is no clear link between spending on education and education outcomes (Mandl, Dierx and Ilzkovitz 2008). Authors have tried to find a link between PISA scores in the area of reading and the money spent on education. Results revealed that although countries spent roughly the same shares of GDP on education, their performance in PISA was very different. Interpretation of these results could be twofold: either other, non-monetary determinants of education performance play the major role or education spending in some countries is inefficient and leads to the poor education performance. Reasons for the inefficiency of education spending could be different within and across countries. The financial resources spent on education may be disproportionate for different sectors of education and training. The lack of resources for vocational or nonformal education and training may have a negative effect on country performance in adult learning. The lack of cost-sharing 158

161 mechanisms, which could motivate adults (and esp. disadvantaged groups of adults) to participate in education and training and stimulate private investment in learning, might be another reason for the lower impact of financing of education and training on adult learning and education attainment Human resources Human resources here include teachers, trainers and the training support staff. Teachers and trainers are the most important, as they are the primary guarantors of quality and equity of adult education and training. Countries increasingly perceive teachers and trainers as the key to making education and training reforms work. Policy measures that focus on the quality of teachers and trainers have become one of the priority areas in many countries (Cedefop 2009b). Raising the status of teachers and trainers, setting standards, increasing their qualification levels and keeping initial and continuing training up-to-date have become major issues of concern in current policy-making. Adult participation in non-formal education and training is usually considerably higher than in formal education and training. A large part of non-formal education and training is provided by employers. Therefore, other factors, such as the availability of training support staff within companies and the quality of providers, have considerable influence on the level of participation of adults in non-formal education and training. In the above sections we have outlined the most important policy characteristics of the education system. On the basis of literature and statistical data review we have tried to operationalise each policy characteristic. We present the results in Table 1 below. However, some aspects of these policy characteristics (e.g. the availability and quality of cost-sharing mechanisms, the quality of training providers, etc.) could not be operationalised due to the lack of data. Table 1: Indicators for measuring the impact of education polices on adult learning 58 No. Policy characteristic Measurement indicators Standardisation (degree to which the quality of education meets the same standards nationwide in terms of teachers training, school budgets, curricula, and the uniformity of schoolleaving examinations) Stratification (numbers and types of transitions available to the next schooling level) Status (implemented, in progress, limited progress) of the development of national qualification frameworks (NQFs), status on 12 December 2008, Cedefop; The extent to which standards-based external examinations exist in the system, PISA Number of school types or distinct educational programmes available to 15-year-olds, PISA 2003 and 2006; First age of selection in the education system, PISA 2003 and 2006; Proportion of 15-year-olds enrolled in programmes that give access to 58 Note that PISA 2003 and 2006 data is used to measure the impact of some particular policy characteristic. Standardisation or stratification of the education system and other education policy characteristics can be traced by looking at the general education level. This usually signifies a general pattern of education system in a particular country. PISA data on general education policy characteristics is comparable and ready available across the countries concerned compared to still fragmented and scarce data on adult learning policy characteristics. We use indicators to measure the general impact of education policy characteristics on adult learning and not a specific impact (i.e. impact on adult learning levels in certain age groups). 159

162 No. Policy characteristic Measurement indicators vocational studies at the next programme level or direct access to the labour market, PISA 2003 and 2006; Percentage of students in schools where the principal reported that within the school (between and/or within classes) there was ability grouping for all subjects, PISA 2006; Percentage of students in schools where the principal reported the students' academic records as a "prerequisite" or a "high priority" for admittance at their school, PISA 2003 and 2006; Entrants at theoretical starting age in ISCED level 5 as % of all persons of the corresponding age group, 2003 and 2006, UIS/OECD/EUROSTAT. The percentage of full-time and part-time upper secondary and postsecondary non-tertiary students enrolled in public and private vocational education programme, , UIS/OECD/EUROSTAT; The percentage of companies providing initial vocational training, 2005, Vocational specificity (extent to which educational systems offer curricula that are designed to prepare students for particular vocations and award credentials that are vocationally specific) Flexibility in provision of learning services (part-time education, modular learning, credit transfer systems, etc.) Validation of prior non-formal and informal learning Information, guidance and counselling Governance of institutions Financing of education and training, (including levels of investment in education and training, spending on human resources). EUROSTAT (CVTS3); The percentage of apprentices from total the number of employees in companies with 10 or more employees, all NACE branches except agriculture, fishing, public administration, private households and extraterritorial organisations, 2000 and 2004, EUROSTAT (LCS); Percentage of students in schools where the principal reported that business and industry have a considerable influence on the curriculum, PISA 2006; Existence of part-time or combined school and workplace courses at the lower secondary vocational (ISCED 2) and/or upper secondary vocational (ISCED 3) level, 2002/2003, EURYDICE; Purpose of the most recent taught learning activity (variable COURPURP), % of answer "Mostly job related (professional)", , EUROSTAT (LFS). Part-time students as % of all students, all ISCED levels combined, , UIS/OECD/EUROSTAT; Inflow of students (ISCED 5-6) from EU-27, EEA and Candidate countries - as % of all students in the country, , UIS/OECD/EUROSTAT; Education or training (non-formal) received during previous four weeks (variable COURATT), , EUROSTAT (LFS); Did the most recent taught learning activity (non-formal) take place during paid working hour? % of answers "Only during paid working hours" and "Mostly during paid working hours" (variable COURWORH), , EUROSTAT (LFS); Student or apprentice in regular (formal) education during the last four weeks, % of those aged who answered "Has been a student or apprentice" and "Student on holiday"(variable COURWORH), , EUROSTAT (LFS). Status (high degree of implementation, medium level of development, low level of activity) of the development of validation of non-formal and informal learning, status at the end of 2007, Cedefop. Percentage of students in schools where the principal reported that career guidance is formally scheduled into students time at school (compulsory), PISA 2006; Percentage of students in schools where the principal reported that they have one or more specific career guidance counsellors employed at the school and/or have one or more specific career guidance counsellors who regularly visit the school, PISA Percentage of students in schools where the principal reported that only school has considerable responsibility for the following aspects of school policy and management: selecting teachers for hire; dismissing teachers; establishing teachers' starting salaries; determining teachers' salary increases; formulating the school budget; deciding on budget allocations within the school; establishing student disciplinary policies; establishing student assessment policies; approving students for admission to the school; choosing which textbooks are used; determining course content; deciding which courses are offered, PISA 2000, 2003 and Share of students in government-dependent private and governmentindependent private schools, PISA 2000, 2003 and Total public expenditure on education as % of GDP, for all levels of education combined, , UIS/OECD/EUROSTAT; Expenditure on educational institutions from private sources as % of GDP, for all levels of education combined, , UIS/OECD/EUROSTAT; Total public expenditure on education as % of GDP, at secondary level of education (ISCED 2-4), , UIS/OECD/EUROSTAT; Total public expenditure on education as % of GDP, at tertiary level of education (ISCED 5-6), , UIS/OECD/EUROSTAT; Annual expenditure on public and private educational institutions per pupil/student in EUR PPS, for all levels of education combined, based on full-time equivalents, , UIS/OECD/EUROSTAT; 160

163 No. Policy characteristic Measurement indicators Annual expenditure on public and private educational institutions per pupil in EUR PPS, at secondary level of education (ISCED 2-4), based on fulltime equivalents, , UIS/OECD/EUROSTAT; Annual expenditure on public and private educational institutions per student in EUR PPS, at tertiary level of education (ISCED 5-6), based on full-time equivalents, , UIS/OECD/EUROSTAT; Ratio of annual expenditure per student at the tertiary level of education to the annual expenditure per pupil at primary level, in public educational institutions, based on full-time equivalents, , UIS/OECD/EUROSTAT; Vocational training costs as % of total labour cost, C_TO_O_NOT_L - All NACE branches except agriculture, fishing, public administration, private households and extra-territorial organisations, 2004, EUROSTAT (LCS); Direct cost of CVT courses per participant, in Purchasing Power Standard (PPS), 1999 and 2005, EUROSTAT (CVTS2 and CVTS3). Ratio of students to teachers (ISCED 3), , UIS/OECD/EUROSTAT; Teachers aged 50+ teaching in public and private at ISCED levels as % of total teachers teaching in ISCED levels 2-3, , UIS/OECD/EUROSTAT; Teachers/academic staff aged 50+ in ISCED level as % of all teachers/academic staff in ISCED level 5-6, , UIS/OECD/EUROSTAT; 9. Human resources Distribution of non-formal education and training activities by provider - employer, 2007, EUROSTAT (AES); Distribution of non-formal education and training activities by provider - formal education institution, 2007, EUROSTAT (AES); Percentage of training enterprises having a specific person or unit responsible for training, 2005, EUROSTAT (CVTS3); Percentage of students in schools where the principal reported that achievement data are used in evaluation of teachers' performance, PISA Source: compiled by the authors on the basis of literature and statistical data review Labour market policies Education and labour market systems and achievement are closely interrelated. On the one hand, education outcomes are used as inputs in the labour market and higher educational attainment most often leads to faster labour market entry and better quality jobs (for example, Muller, 2005). On the other hand, learning is increasingly taking place in the workplace, and higher individual s labour market achievements (higher ISCO occupational group) lead to the higher participation in learning. A number of links between education and labour market systems suggests that adult learning is influenced considerably by labour market policies. In this research we consider the direct and indirect influence of the following labour market policy characteristics on adult learning: Employment protection; Income support for the unemployed; Investment in labour market training and services. Below we discuss in detail the likely influence of these characteristics on our adult learning variables Employment protection The direct influence of employment protection on adult learning has not been previously examined according to our knowledge. Thus, a more elaborate theoretical explanation for such a relationship is missing. One possible explanation is that higher employment protection is associated with less flexible working arrangements, which might be an obstacle for a combination of part-time work and part-time study or a longer term training leave for a full-time study. 161

164 Employment protection is likely to influence adult learning indirectly by affecting: 1. Balance between labour supply and demand (unemployment, employment and activity rates), which then influences learning or training preferences of labour market participants. 2. Job mobility, which also influences learning or training preferences of labour market participants. With regard to the first type of influence, there is disagreement between scholars whether or not higher employment protection results in higher or lower unemployment rates. The first argument suggests that employment protection could lower the possibilities for companies to adapt quickly to changing market demand, which in the long term would reduce competitiveness (Wilthagen and Tros 2004). Employers would also be averse to hiring large numbers of new staff in time of economic success if they had difficulty in disposing of it in periods of downturn (Scarpetta 1998, Morgan and Mourougane 2005). In both cases the result would be higher unemployment. The counter argument suggests that higher employment protection, would, first of all, be likely to lead to greater cooperation between employers and employees and higher investment in training of employees (Estavez-Abe, Iversen and Soskice 2001). This would result in greater productivity and competitiveness. Secondly, although higher employment protection would limit job creation in the boom times, this would also result in less job loss in an economic downturn (Joseph, Pierard and Sneessens 2003). Previous empirical research provides little if any support to the hypothesis that strict employment protection would lead to higher unemployment and lower employment. A review of 15 previous research studies on the subject published between 1996 and 2007 showed that the results of only 2 studies (Scarpetta 1996, International Monetary Fund 2003) provided evidence of correlation between these variables, while other studies found no such relationship (Martinaitis 2008). Still the relationship is worth considering again with the latest data and a wider sample. Assuming that employment protection level does affect the balance between labour supply and demand, the second part of the logic seems to be supported by previous research. The Cedefop study (Cedefop, 2009a) covering the EU member states provides evidence that participation of the labour force from ISCO4-9 occupations in learning is highly correlated to population activity rates and somewhat inversely correlated to unemployment levels. Firstly, a labour force with vocational qualifications takes part in learning more often than the remaining, inactive, working-age population. Secondly, employers facing the need for new qualifications can choose between recruiting new staff and investing in the qualifications of the existing staff or choosing the most appropriate mixture of both. Recruiting might be cheaper and preferable to training when there is a surplus supply of labour with the required qualifications. However, at times of low supply and high demand, the training of existing staff might be preferable to recruiting. 162

165 Employment protection could further influence adult learning through job mobility, since higher protection implies the higher costs of dismissing workers and discourages new recruiting (Muller, 2005). Greater mobility is likely to be associated with a greater need for investment in certain competences and greater participation in learning. In the circumstances of a less well protected labour market, employees would be likely to concentrate on more general competences, which have broader applicability in the labour market. Also, more frequent change of jobs would require the acquisition of new competences more often. However, greater mobility is likely to reduce the incentives for employers to invest in the training of their employees due to the poaching problem and the likely loss of investment. Employers would concentrate on only the training most necessary to provide their employees with the required company-specific skills. This need would be greater in less protected labour markets due to the higher turn-over of staff. The likely casualty of greater mobility of staff would be sector-specific skills, which would be neglected both by employers and employees. Overall we expect that, all conditions being equal, greater job mobility would still result in more adult learning. Previous research suggests that job mobility is also positively associated with higher employment rates and lower long-term unemployment rates (Andersen, Haahr, Hansen and Holm- Pedersen, 2008). Thus, job mobility could affect adult learning not only directly, but also indirectly, through its effects on the balance between labour supply and demand. To check these influences we borrow the occupational mobility index, employment mobility index, job-to-job mobility index and average job mobility index from Andersen, Haahr, Hansen and Holm-Pedersen (2008). We partly measure employment protection on the basis of data from the World Bank Doing Business annual survey. Beginning with 2004 it compares countries in terms of employment protection and provides useful data for international comparative research. It measures the rigidity of employment index, which is the average of 3 sub-indices: a difficulty of hiring index, a rigidity of hours index and a difficulty of firing index, and firing costs (see box 1 below for the composition of these indices). These indices might measure legal regulation rather than actual labour market rigidity (or flexibility), so we include additional indicators from Eurostat on the share of temporary employees in the total number of employees, part-time employment as a percentage of the total employment, and population in employment having a second job. Yet another possible element of employment protection could be the level of safety at work. Box 1: Composition of rigidity of employment index and firing cost indicator World Bank s Doing Business examines the difficulty of hiring, non-standard work schedules and paid annual leave, and the costs and rules governing redundancy termination. The rigidity of employment index is the average of 3 sub-indices: a difficulty of hiring index, a rigidity of hours index and a difficulty of firing index. The difficulty of hiring index measures (i) whether fixed term contracts are prohibited for permanent tasks; (ii) the maximum cumulative duration of fixed term contracts; and (iii) the ratio of the minimum wage for a trainee or first time employee to the average value added per worker. The rigidity of hours index has 5 components: (i) whether night work is unrestricted; (ii) whether weekend work is unrestricted; (iii) whether the work week can consist of 5.5 days; (iv) whether the workweek can extend to 50 hours or more (including overtime) for 2 months a year to respond to a seasonal increase in production; and (v) whether paid annual vacation is 21 working days or fewer. The difficulty of firing index has 8 components: (i) whether redundancy is disallowed as a basis for 163

166 terminating workers; (ii) whether the employer needs to notify a third party (such as a government agency) to terminate 1 redundant worker; (iii) whether the employer needs to notify a third party to terminate a group of 25 redundant workers; (iv) whether the employer needs approval from a third party to terminate 1 redundant worker; (v) whether the employer needs approval from a third party to terminate a group of 25 redundant workers; (vi) whether the law requires the employer to reassign or retrain a worker before making the worker redundant; (vii) whether priority rules apply for redundancies; and (viii) whether priority rules apply for reemployment. The firing cost indicator measures the cost of advance notice requirements, severance payments and penalties due when terminating a redundant worker, expressed in weeks of salary. Source: World Bank Doing Business, annual survey, available at Fixed-term contracts allow employers to meet their short term needs for additional labour, while avoiding the costs of dismissing unnecessary employees, once the needs are meet. Fixed-term contracts might also serve as trial periods for new staff: well performing employees could be offered permanent contracts at the end of fixed term contracts. The more widespread fixed-term contracts are, the lower is employment protection and the greater is the flexibility for employers in hiring and firing staff. This results in greater job mobility and, subsequently, adult learning (although the investment of employers in training their employees portable competences would be lower). The availability of part-time working arrangements might be particularly important to adults, who wish to participate in parttime education or to strike a better balance between career and other obligations, such as taking care of dependent persons: children, elderly or disabled. Part-time work arrangements also allow the attainment of a desired balance between leisure time and work. People can choose to work in more than one work place (substituting some of their leisure time for a second job and higher income) or, alternatively, work less and have more leisure. Parttime working arrangements could also be important for people, who cannot work full-time for health reasons. The more widespread part-time work is in the economy, the more flexible its labour market is likely to be. Greater flexibility is likely to lead to lower unemployment levels and higher participation in learning. The level of safety at work, measured by the index of the number of serious or fatal accidents at work per 100 thousand persons in employment, refers to a completely different aspect of employment protection. It is unrelated to labour market rigidity or flexibility, but rather reflects the quality of regulation or its enforcement, employment conditions and the attitudes of employers towards the value of their employees. Low levels of safety at work could also be the direct result of underinvestment in training of employees, revealing their lack of competences and attitudes needed to perform one or another job. Therefore, the levels of safety at work should be inversely correlated with the level of adult participation in learning, especially of blue-collar workers (ISCO 4-9 occupation groups) Income support for unemployed Greater income support level provides a better safety net for redundant employees and allows for a more selective job search, helping to prevent occupational downgrading, and providing the time necessary for upgrading or acquiring new qualifications. This helps to ensure that those, who had a chance to acquire new competences under the protection of the safety net, would not become unemployed any time soon (Scarpetta 1996, Howell 2005). However, proponents of neoclassical theories maintain that higher 164

167 income support leads to higher unemployment. Firstly, they point out that generous unemployment benefits reduce incentives to look for a job (Blanchard and Wolfers 2000). Proponents of flexicurity disagree that long term duration of benefits rather than their size reduces these incentives (Auer 2002). The motivation properties of unemployment benefits would also depend on how they are administered. The payments should also be contingent on the progress of unemployed people in improving their employability and on the take-up of offered jobs, which correspond to their qualification level and perhaps quality of their previous job. Secondly, generous benefits strengthen the position of employees in bargaining with employers, which leads to higher prices of labour (Esping-Andersen 2000). This eventually drives less productive labour to unemployment (Howell 2005). Higher income support level during unemployment, coupled with greater investment in labour market training and services might serve as a substitute to a higher employment protection level. It might help to maintain an acceptable level of social and employment security, while allowing market forces to balance out irregularities between labour supply and demand, and to ensure the efficient allocation of human resources in the economy. We would expect that greater and longer income support for the unemployed together with lower employment protection and greater job mobility would lead to higher levels of adult learning Investment in labour market training and services Greater investment in labour market training and services influences the levels of adult learning directly, by helping to find suitable and necessary training and to buy it, on the condition that investment is effective and efficient. Labour market training allows the faster adaptation of redundant employees to the needs of the labour market largely by means of upgrading or acquiring new qualifications. Labour market training services support the career planning process, the selection of necessary training and eventual job search to match the qualifications acquired. Both major components of active labour market policy reduce the uncertainty and the costs of employers in recruiting new staff. Indirectly it might also contribute to higher levels of adult learning through lower unemployment and higher employment rates, as well as greater job or occupational mobility. Firstly, an active labour market policy improves employment security for individuals as it helps them return to the labour market (Auer 2002). Secondly, they improve labour market flexibility, the quality of the labour force and productivity through faster and better adaptation to changing labour market needs (Masso, Emets 2006) thus supporting mobility of labour and efficient allocation of resources in the economy. Finally, labour market policy reduces the adverse effects of unemployment benefits on motivation, because participation in policy measures is most often (if not always) one of the key conditions for receiving benefits. The relationships between labour market policy characteristics are demonstrated in Figure 2. The specific indicators we have selected to measure the impact of labour market policies on adult learning are provided in Table

168 Figure 2: The influence of labour market policies on adult learning levels Investment in labour market training and services Income support for the unemployed Employment protection Adult learning Balance of labour supply and demand Labour market mobility Source: compiled by the authors on the basis of literature review. Table 2: Indicators for measuring the impact of labour market policies on adult learning No. Policy characteristic Measurement indicators Difficulty of hiring index, World Bank Doing Business rankings Rigidity of hours index, World Bank Doing Business rankings Difficulty of firing index, World Bank Doing Business rankings Rigidity of employment index, World Bank Doing Business rankings Firing costs, World Bank Doing Business rankings Temporary employees as a percentage of the total number of 1. Employment protection employees, EUROSTAT Part-time employment as a percentage of the total employment, EUROSTAT Population in employment having a second job, EUROSTAT Index of the number of serious accidents at work per 100 thousand persons in employment (1998=100), EUROSTAT Index of the number of fatal accidents at work per 100 thousand persons in employment (1998=100), EUROSTAT 2. Income support for Income support levels for unemployed (social protection benefits - unemployed unemployment function (PPS per inhabitant)), EUROSTAT Investment in labour LMP expenditure on Labour market services (% from GDP), 3. market training and EUROSTAT services LMP expenditure on Training (% from GDP), EUROSTAT Employment rate (%), EUROSTAT Employment rate of older workers (%), EUROSTAT Unemployment rate, yearly averages, EUROSTAT Average exit age from the labour force, weighted by the probability of withdrawal from the labour market, EUROSTAT Population activity rate (%, 15 to 64 years), EUROSTAT 4. Balance of labour supply and demand 5. Labour market mobility Unemployment rates by highest level of education attained (%) - Pre-primary, primary and lower secondary education - levels 0-2 (ISCED 1997), EUROSTAT Unemployment rates by highest level of education attained (%) - Upper secondary and post-secondary non-tertiary education - levels 3-4 (ISCED 1997), EUROSTAT Unemployment rates by highest level of education attained (%) - Tertiary education - levels 5-6 (ISCED 1997), EUROSTAT Occupational mobility index (Andersen, Haahr, Hansen and Holm- Pedersen 2008) Employment mobility index (Andersen, Haahr, Hansen and Holm- Pedersen 2008) Job-to-job mobility index (Andersen, Haahr, Hansen and Holm- Pedersen 2008) Average job mobility index (Andersen, Haahr, Hansen and Holm- Pedersen 2008) 166

169 Source: compiled by the authors on the basis of literature and statistical data review Health policies Health policies influence both working and learning opportunities indirectly. The quality of life in general and participation in education and labour market in particular depend on the health condition of the population. The more time people spend on sick leave, the less they work (Bloom and Canning 2000) or participate in learning (Becker 2007). Better health leads both to higher labour productivity (for example, WHO 2002; Strauss and Thomas 1998; Bloom and Canning 2000) and higher educational attainment (for example, OECD/Statistics Canada 2005; Ding et al 2006), as well as longer working or learning lives (Becker 2007). The determinants of health include a wide range of factors, many of which are beyond individual control: income and social status, education, physical environment, social support networks, genetics, gender and health services (for example, Wilkinson and Marmot 2003). Therefore, health policy reforms, which improve health services and, consequently, the health of the adult population, also have a positive influence on the labour market participation and learning of adults. Health care system, like the education system, could be similarly characterised by an increasingly large number of groups of policy features. Therefore, we have also selected the following most important policy characteristics: service delivery; healthy workforce; medical products, vaccines and technologies; financing of the health system; health information system; governance. We do not describe in detail policy characteristics of the health care system, as they are not directly related to adult learning and education attainment as are education and labour market policies. For the same reason we use only general indicators to measure the outcomes of health care system policies and their indirect impact on adult learning and education attainment. Policy characteristics of the health care system are described in brief and operationalised in Table 3 below. Table 3: Indicators for measuring the impact of health care policies on adult learning No. Policy characteristics Measurement indicators 1. Service delivery (personal and non-personal health interventions) Health workforce Consumer Powerhouse; 2. (sufficiency, distribution, competence, responsiveness and productivity of staff) 3. Medical products, vaccines and technologies EUROSTAT; Financing of health system 4. (funds for health; incentives for providers and users to be efficient) EUROSTAT; Health information system 5. (information on health determinants, health system performance and health status) EUROSTAT; Governance 6. (oversight, coalition-building, regulation, attention to systemdesign and accountability) Source: compiled by the authors on the basis of literature and statistical data review. The Euro Health Consumer Index (total score), , Health Life expectancy in absolute value at 65 females, , Life expectancy in absolute value at 65 males, , Life expectancy in absolute value at birth females, , Life expectancy in absolute value at birth males, , EUROSTAT. Education, labour market and health policies vary to a considerable extent among EU members, candidate and EFTA-EEA countries. Although there are some common tendencies (such as the rise in national qualification frameworks, systems for recognition of prior learning and decline in employment protection legislation), they 167

170 seem to display a great deal of heterogeneity and path dependency. Evidence from prior research suggests that different policy characteristics could sometimes breed rather similar labour market outcomes, as negative influences of some institutions are compensated by positive influences of others (for example Bukodi and Robert 2007, Saar, Unt and Kogan 2008). The same most probably applies to learning outcomes. Thus, there is hardly one single path to a successful reform of the above policies in terms of their adult learning outcomes. Policy properties, as well as socioeconomic context of a specific country, have to be studied in order to trace the impact of reforms on adult learning and educational attainment Context variables Contextual variables are factors, other than education, labour market and health policies or their outcomes, that could possibly have significant influence on adult learning and education attainment. However, social and economic context factors could influence not only adult learning and education attainment. They could well be correlated to education, labour market and health policies and their outcomes. Moreover the causality is likely to be two-way. In general, social and economic context variables reveal the relationship between adult learning and the country s progress towards the European Union s strategic goal, set at the Lisbon European Council in 2000: to become the most competitive and dynamic knowledge-based economy in the world capable of sustainable economic growth with more and better jobs and greater social cohesion. In this report we consider the following contextual variables: Macro-economic environment; Social cohesion; Innovation and research; Environment and transport. The indicators for the measurement of the impact of contextual variables on adult learning and education attainment have been selected from the commonly agreed structural indicators that are used to assess the progress of EU Member States in achieving Lisbon goals. The selected internationally comparable indicators are provided in Table 4 and are briefly discussed below Macro-economic environment Macro-economic environment refers to the performance of the overall economy. Factors such as GDP, labour productivity, inflation rate, etc. are indicators for the analysis of economic background conditions of the country, in which adult learning takes place. Significant differences in adult learning and education attainment exist in countries with different economic development. Countries that are characterised, for example, by a stable macro-economic environment, low level of inflation and sound budgetary policies are more conducive to economic growth, lifelong learning and educational attainment. On the one hand, economic growth creates a more favourable environment for business development, more and better jobs and a demand for a qualified labour force. On the other hand, education is beneficial for labour productivity and is a key component of workplace and business development and 168

171 knowledge-based economic growth. Evidence shows that learning by adults can enhance their employability and help them achieve higher earnings, as well as producing non-economic benefits. Participation in lifelong learning can also contribute to national economic capacity, by raising the skills, understanding and motivation of the workforce Social cohesion Social cohesion refers to the participation of citizens in societal life and the reduction of social divisions, associated with material conditions, relationships, inclusion and equality. The debate over education and social cohesion is very old, and more often highlights the influence of education on social cohesion than vice versa. The European Commission s Memorandum on Lifelong Learning (2001) identifies social cohesion as a major objective of education and training. The EU has recognised the key role which adult learning can play in meeting the goals of the Lisbon Strategy, by fostering social cohesion, providing citizens with the skills required to find new jobs and helping Europe to better respond to the challenges of globalisation. There is significant evidence of a positive relationship between adult learning and education attainment and social cohesion. Researchers have shown that this association is more broadly positive in respect to active citizenship and social connections, social participation, and socio-economic equality. Studies of the Nordic countries suggest that this positive relationship is strongest where policy is aimed at securing high participation rates among the most vulnerable groups. Even though participation rates are unequal, they do not contribute to greater inequality because the gap between the advantaged and disadvantaged is relatively low. Adult education researchers have repeatedly found that participation in education and training varies across the adult population. Participation rates are lowest among certain groups: early school leavers, people outside the labour market, older adults, etc. There are also associations between gender and participation, for example, there are differences in overall participation rates between men and women, which might be a product of such factors, as employment status and child care Innovation and research Expenditure on research and development (R&D) is a key indicator of government and private sector efforts to obtain the competitive advantage in science and technology. This consists of the total expenditure on R&D by all resident companies, research institutes, university and government laboratories, etc. ICT development level includes indicators related to ICT access, use and skills, such as households with a computer, the number of Internet users and literacy levels. The most advanced countries in ICT are mainly high-income countries, while less prosperous countries remain at the lower end of the index with limited access to ICT infrastructure, including fixed and mobile telephony, Internet and broadband. The Summary Innovation Index provides an overview of the relative national innovation performance. The connection between innovation and research and adult learning and education attainment is also two-way. On the one hand, in the context of the global economy and the digital age there is a 169

172 No Contextual variables Macroeconomic environment Social cohesion growing need for a higher skills level, the pressure for all citizens to adapt to technological change and future skills needs, which implies the growing need for learning. Adults are required to continuously develop broader ICT and ICT related skills to actively participate in a modern society throughout life. On the other hand, scientific advances and new information and communication technologies modernise teaching and learning methods, and provide better access to learning through a more flexible delivery of education and training. Moreover, new technological solutions may play an important role in assessment and documentation of people s competences, vocational counselling and guidance Environment and transport Environment and transport variables refer to sustainability indicators. Sustainability has been commonly defined as the development that meets the needs of the present without compromising the ability of future generations to meet their own needs. This means a balance between economic, social and environmental priorities in the decision making process. Since the notion of sustainable development is closely related with new values and attitudes, it demands continuous learning and reinforcement of learning. Therefore, adult learning is one of the pillars for sustainable development, which requires the reeducation of the workforce from the current focus on economic value to a multi-dimensional, interdependent, values-based focus on environmental, social and economic issues. Thus, education contributes to sustainable development, as it enables people to acquire the values, knowledge and skills required for a more sustainable future. Moreover, education and learning opportunities can help people to find new solutions to the social, economic and environmental issues affecting their lives. Table 4: Indicators for measuring the impact of social and economic context on adult learning Measurement indicators GDP per capita in Purchasing Power Standards (PPS); Labour productivity - GDP in Purchasing Power Standards (PPS) per person employed relative to EU-27 (EU-27 = 100); Labour productivity per hour worked - GDP in Purchasing Power Standards (EU- 15 = 100); Inflation rate - Annual average rate of change in Harmonized Indices of Consumer Prices (HICPs); Real unit labour cost growth - Growth rate of the ratio: compensation per employee in current prices divided by GDP in current prices per total employment; Public balance - Net borrowing/lending of consolidated general government sector as a percentage of GDP; General government debt - General government consolidated gross debt as a percentage of GDP; Comparative price levels - Comparative price levels of final consumption by private households including indirect taxes; Market Integration - Foreign Direct Investment (FDI) intensity - Average value of inward and outward FDI flows divided by GDP; Business investment - Gross fixed capital formation by the private sector as a percentage of GDP; Implicit tax rate on labour - Ratio of taxes and social security contributions on employed labour income to total compensation of employees; Tax wedge (in %) on labour cost. Inequality of income distribution - Income quintile share ratio; At-risk-of-poverty rate before social transfers (in %); At-risk-of-poverty rate after social transfers (in %); At-persistent-risk-of-poverty rate (in %); Long-term unemployment rate - Long-term unemployed (12 months and more) as a percentage of the total active population; Early school leavers - percentage of the population aged with at most 170

173 No. Contextual variables Measurement indicators lower secondary education and not in further education or training; Jobless households children - Share of persons aged 0-17 who are living in households where no-one works; Jobless households by gender - Share of persons aged who are living in households where no-one works; Tax rate on low wage earners: Unemployment trap - % tax rate on low wage earners by marginal effective tax rates on employment incomes; Single person without children; Percentage of private households with 0, 1, 2 or with 3 and more children; Social protection expenditure as % of GDP and in PPS per head (excluding expenditure on employment policies); Social protection benefits as % of GDP and in PPS per head (excluding unemployment benefits). Gross domestic expenditure on R&D (GERD) - Percentage of GDP; ICT development index (IDI); Level of internet access households - Percentage of households who have Internet access at home; ICT expenditure by type of product; Information Technology Expenditure - Percentage of GDP; Broadband penetration rate - Number of broadband access lines per Innovation inhabitants; and research High-tech exports - Exports of high technology products as a share of total exports; Innovation expenditure as % of turnover for all enterprises; Total R&D personnel and Researchers as % of labour force and total employment; Employment in high/medium technology manufacturing and knowledge intensive services as % of total employment; Summary innovation index (SII). Energy intensity of the economy - Gross inland consumption of energy divided by GDP (at constant prices, 1995=100) - kgoe (kilogram of oil equivalent) per 1000 euro; Resource productivity - Ratio of gross domestic product to domestic material consumption; Environmental protection expenditure in Europe - indicators: Euro per capita 4. Environment and % of GDP; and transport Car share of inland passenger transport; Density of paved road network; Density of railroad network; Passenger cars per 1000 inhabitants; Volume of passenger transport relative to GDP; Volume of passenger transport relative to GDP - Index of inland passenger transport volume relative to GDP (2000=100). Source: compiled by the authors on the basis of statistical data review Research methods The Literature review has first helped us to define our dependent, independent and context variables and their potential relationships. We use statistical analysis to relate policy differences of the EU, EFTA-EEA and candidate countries to adult learning. This analysis has helped to reveal, which policy characteristics are the most conducive to adult learning and educational attainment. The contextual influences have been controlled for. This chapter describes the most important methodological aspects of the correlation and regression analysis. Spearman s correlation coefficients were calculated, and the correlations of each pair of variables are provided in the tables of this report. The Spearman method was chosen because it produced an output almost identical to that received from using the Pearson method and because few outliers were present in the data. The correlations marked with an asterisk are statistically significant to the significance level indicated. N shows the number of observations in the data set. 171

174 The correlation coefficient can take values from 1 to +1. A value of +1 shows that the variables are perfectly linear related by an increasing relationship, a value of 1 shows that the variables are perfectly linear related by a decreasing relationship, and a value of 0 shows that the variables are not linearly related to each other. A very strong correlation is considered if the correlation coefficient is greater than 0,7, a strong correlation if the correlation coefficient ranges from 0,7 to 0,5, a medium correlation if the correlation coefficient ranges from 0,5 to 0,3, and a weak correlation if the correlation coefficient is less than 0,3. Multiple regression modelling with time series variables was carried out using un-transformed and transformed variables. Transformed variable Adult participation in education and training during previous four weeks (both formal and non-formal) has been used to avoid a break in series. In chapter 2.4 we provide two key measures indicating the explanatory power of the tested models and the relative predictive power of independent variables selected in the process of modelling. R square provides a measure of how well future outcomes are likely to be predicted by the model. R square values vary from 0 to 1. The beta coefficient ( B standardized ) is the ratio of the relative predictive power of the independent variables. All independent variables that might be expected to have an influence on the dependent variables were identified during the literature review and correlation coefficient analysis. Typical limitations of multiple regression modelling: Linearity. It is assumed that the relationship between variables is linear. In practice this assumption can virtually never be confirmed. Causality. The major conceptual limitation of all regression techniques is that one can only ascertain relationships, but never be sure about underlying causal mechanism. Number of observations. Large number of variables in regression requires a large number of observations. The number of observations in our data set is relatively low, thus some regressions produce results with low level of significance. The results of statistical analysis accord well with the desk research carried out by the project team. To illustrate the findings of quantitative analysis we provide a number of policy reform examples from different countries, which illustrate how policy changes might have affected adult participation in learning and educational attainment. 2. Overview of interrelationships between adult learning and education attainment This chapter aims to explore the important interrelationships between variables on adult learning and educational attainment. Special focus is given to disadvantaged groups in terms of adult learning (esp. low-skilled population). 172

175 2.1. Non-formal education and training and overall country performance in adult learning Table 5 below shows that the influence of non-formal education and training on the overall level of adult learning is considerably stronger than in the case of formal education and training. Indeed, statistics (see Figure 3 in chapter 1.1) show that the share of adults participating in non-formal education and training is usually higher than the share of those enrolled in formal education and training. Figure 3 clearly shows the tendency that the higher the country ranks in terms of the overall level of adult learning, the larger is the share of adults participating in non-formal education and training. The importance of non-formal education and training could be justified on the basis of the following arguments: Adults tend to have low demand for formal education and training as they usually already have one formal education degree and the skills necessary for the acquisition of additional knowledge, when needed. The correlation results confirm that adults with tertiary education participate in non-formal education and training more often than those who have at least upper secondary education; Adults, who have spent many years in formal education and training and invested considerable resources in it, are eager to enter the labour market and utilize the acquired knowledge to earn their first salaries; Non-formal education and training, if it is well-developed in a country (eg. where the validation of non-formal and informal learning is a practical reality for individuals), is better equipped to serve adults, since it provides a possibility to select up-to-date courses which perfectly match their current professional needs and offer flexible training without interrupting their job responsibilities Tertiary education attainment and participation Results show that adult tertiary education attainment significantly correlates with adult participation in formal education and training and even more with adult participation in non-formal education and training. On the one hand, more active involvement of adults in formal education and training is likely to increase their educational attainment. This is less likely in the case of adult participation in non-formal education and training, especially if the system for validating non-formal and informal learning is not well established in the country. On the other hand, adults who have higher educational attainment more often participate in non-formal education and training than in the formal one. It is likely that the returns to formal education, when an adult reaches a certain educational attainment level, tend to decrease and the demand for learning tends to be more selective. The interrelationship between adult tertiary education attainment and adult learning could therefore be interpreted as follows: formal education and training increases education attainment and, as a result, higher educational attainment leads to a more active participation in non-formal education and training Participation and intensity of learning Statistical analysis has revealed a positive relationship between participation in and the intensity of adult learning: more active participation of adults in both formal and non-formal education and training leads to the higher intensity of this education and training 173

176 Adult participation in education and training (both formal and non-formal) Adult participation in formal education and training Adult participation in non-formal education and training Intensity of adult education and training (nonformal) Adult upper secondary education attainment Adult tertiary education (it is even more so for the low-skilled adults). However, statistics show (see figure 6 and table 6 in chapter 1.1) that countries ranking high in terms of adult participation in learning tend to have low average intensity of education and training. The answer could be related to educational attainment. High performing countries have significant and constantly increasing population of those who participate in learning and especially high-qualified adults. The latter are more selective and well informed for knowledge acquisition than low-qualified. Thus high-qualified adults spend less time to a acquire knowledge they need. However, higher qualification leads to more education and training as the demand for knowledge constantly rises. Following this logic, intensity should increase together with qualification. Therefore the interrelationship between participation in adult learning and the intensity of this learning is not that clear and should be further analysed in future studies. Table 5: Interrelationships between adult learning and education attainment Intensity of Adult Adult Adult adult participation in Adult upper Adult participation participation education education and secondary tertiary in formal in non-formal and training (both educational educational education education and training formal and attainment attainment and training training (nonformal) non-formal) Spearman Correlation 1,526**,991**,600**,227**,602** N Spearman 1,876**,625**,200**,356** Correlation N Spearman 1,607** 0,061,642** Correlation N Spearman 1 -,274** -,268** Correlation N Spearman -,274** 1,260** Correlation N Spearman Correlation -,268**,260** 1 attainment N * - Correlation is significant at 0.05 level (2-tailed). ** - Correlation is significant at 0.01 level (2-tailed). Coefficients equal or above 0.300** are marked by dark shading. Source: EUROSTAT, calculations by the authors Although adult participation in non-formal education and training seems to have the highest impact on the overall level of adult learning in a country, the formal education and training and educational attainment are also important. The relationships between these variables are always two-way, and the most important linkage or dependent variable cannot be clearly determined. In the following chapters all these dependent variables are used to determine their possibly important linkages with policy characteristics as well as policy context. 174

177 3. The influence of policy characteristics on adult participation in learning and educational attainment of the population This chapter presents the results of the statistical analysis of the impact of education, labour market and health care policies on adult learning and educational attainment Education policies Chapter 1.3 of this Annex has singled out and described the following nine policy characteristics of the education system: standardisation, stratification, vocational specificity, flexibility in provision of learning services, validation of prior learning, information, guidance and counselling, the governance of institutions, financing of education and training, and human resources. In this chapter with the help of the statistical analysis (using Spearman s correlation) we have identified a limited number of indicators characterising these policies, which have a significant influence on adult learning and education attainment. Correlations seem to be found for indicators of the following 6 out of 9 policy characteristics: 1. Stratification; 2. Vocational specificity; 3. Flexibility in provision of learning services; 4. The governance of institutions; 5. Financing of education and training; and 6. Human resources. Each policy characteristic of the education system which have a significant influence on adult learning and educational attainment is shortly analysed below Stratification The statistical analysis has revealed that only two indicators characterising the degree of stratification of education system, i.e. the number of school types or distinct education programmes available for 15-year-olds and the first age of selection in the education system, have a significant influence on adult learning and educational attainment. Results in Table 6 show two significant correlations. Firstly, the higher the number of school types or distinct education programmes, the less adults are likely to participate in formal training (weak correlation) and fewer of them pursue tertiary education (medium correlation). Indeed, the results of the statistical analysis have confirmed a methodological assumption that in the education systems with a high differentiation within secondary education (i.e. a high number of school types or distinct educational programmes and the early age of selection in the education system) students are less likely to invest in higher education to distinguish themselves from other students in the labour market. In highly stratified educational systems the transition from education to the labour market is easier and earlier, which means that students spend less time in formal education and training and less adults attain tertiary education. 175

178 Dependent variables Indicators of stratification Number of school types Spearman or distinct Correlation educational programmes available to 15-year-olds (STRAT1) First age of selection in the education system (START2) Secondly, the later is the year of selection in the education system, the higher is the level of adult learning, both formal and nonformal (small/medium correlation) and adult tertiary education attainment (medium correlation) in a country. The results of the statistical analysis have confirmed that the earlier is the selection in education system, the lower is the equality of education opportunities and, consequently, educational attainment. If students were to make important educational decisions at an early age, lower class students would more often end up in the lower levels of education, since their choices would be largely influenced by the parents. Table 6: The influence of stratification on adult learning and educational attainment Adult participation in education and training (both formal and non-formal) Adult participation in formal education and training Adult participation in non-formal education and training Intensity of adult education and training (nonformal) Adult upper secondary educational attainment Adult tertiary educational attainment -,355* -,408** -,321* ,515** N Spearman Correlation,528**,469**,519** ,565** N * - Correlation is significant at 0.05 level (2-tailed). ** - Correlation is significant at 0.01 level (2-tailed). Coefficients equal or above 0.300** are marked by dark shading. Source: EUROSTAT, calculations by the authors Vocational specificity The methodological chapter of the Annex has revealed that stratified education systems seem to be also vocational specific. Therefore, it would be reasonable to expect a correlation between indicators characterising vocational specificity and our dependent variables. However, the results of the statistical analysis presented in Table 7 reveal only two significant correlations between vocational specificity indicators and adult learning and educational attainment. Firstly, there is a medium correlation between the percentage of enterprises providing initial VET 59 and adult learning: the higher the percentage of enterprises providing initial VET, the larger the share of adults participating in non-formal education and training (correlation with participation of low-skilled adults in education and training is even higher 0.726**). Secondly, the prevalence of job related non-formal education and training seems to have a medium positive correlation with adult learning and a weak correlation with adult upper secondary education attainment (this correlation is lower in the case of low-skilled adults, especially with low-skilled participation in formal education and training where it is 0,618**). The results confirm that the better are the institutional links between vocational education and training and the labour market (characterised by the extensive involvement of enterprises in the provision of training and the prevalence of job related education and training), the better is a country s 59 Note: all types of size of enterprises are included in the sample for this indicator. 176

179 Dependent variables Indicators of vocational specificity The percentage Spearman of enterprises Correlation providing initial vocational training (SPEC1) Purpose of the most recent taught learning activity (variable COURPURP), % of answer "Mostly job related (professional)" (SPEC2) performance in adult learning. Moreover, the enterprises which have strong institutional links with formal education and training institutions seem to be actively involved in financing and/or providing non-formal education and training (or vice versa). Highly vocational specific education systems tend to have a high qualification-occupation match. As a result, lower occupational mobility rates are observed in this type of countries. The more people are vocationally specialised, the less they move between jobs, the more selectively and intensively they study. This possibly explains medium correlations between the prevalence of job related education and training and the variables of adult learning. Table 7: The influence of vocational specificity on adult learning and education attainment Adult participation in education and training (both formal and nonformal) Adult participation in formal education and training Adult participation in non-formal education and training Intensity of adult education and training (nonformal) Adult upper secondary education attainment Adult tertiary education attainment,674**,416*,648** N Spearman Correlation,653**,691**,637**,652**,317** N * - Correlation is significant at 0.05 level (2-tailed). ** - Correlation is significant at 0.01 level (2-tailed). Coefficients equal or above 0.300** are marked by dark shading. Source: EUROSTAT, calculations by the authors The results of the adult education survey support the statistical relationship between the prevalence of job related education and training and the overall level of adult learning. In 2007 the most frequently mentioned reason for participation in non-formal education and training was to do a job better and/or improve career prospects on average 43.7% of EU-27 adults indicated that reason. In some of the countries job related motivation was especially high. For example, 80% of adults in Estonia, 77% in Croatia, 75% in Latvia, 72% in Norway, 68% in Germany and Hungary, 67% in Austria and Poland, 66% in the Netherlands and Lithuania, 65% in Finland, 64% in Belgium, 63% in Slovakia and 62% in Sweden. However not only job related motives encourage adults to learn. Even 31.9% of adults indicated that they learn non-formally in order to increase their knowledge/skills of an interesting subject. Therefore self-realization is also a very important incentive to learn Flexibility of provision Statistical analysis has revealed that four indicators characterising flexibility in provision of training services significantly influence the dependent variables (see Table 8). Firstly, flexible formal education and training arrangements have a moderate influence on adult participation in learning: the higher the share of part-time students in the country, the higher is the level of adult learning and tertiary education attainment. Secondly, EU policies promoting the mobility of students seem to have a positive effect on adult learning and 177

180 attainment: the higher the number of inflow students, the better is a country s performance in non-formal adult learning and tertiary education attainment. Thirdly, better opportunities for reconciliation between work and studies seem to be important both for adult learning and adult tertiary education attainment: the higher the share of those who study during paid working hours, the higher is the level of adult learning and tertiary education attainment (this is less valid in case of formal education and training of low-skilled where correlation is 0,304**). Last but not least, the countries which provide better opportunities to study in formal education and training for those aged 35 and older seem to perform better in adult learning and education attainment. Therefore, it might be concluded that flexible formal education and training arrangements (e.g. modular training), policies fostering higher mobility of students, better reconciliation between work and studies and better formal education opportunities for those aged have an increasingly important impact on the level of adult learning and education attainment. Table 8: The influence of flexibility in provision of learning services on adult learning and education attainment Dependent Adult Intensity variables participation Adult Adult of adult Adult upper Adult in education participation participation education secondary tertiary and training in formal in non-formal and education education (both formal education education and training attainment attainment Indicators of flexibility and nonformalformal) and training training (non- in provision of learning services Spearman FLEX1¹ Correlation,352**,205**,283** -,200*,256**,318** N Spearman FLEX2¹ Correlation,513**,172*,578** ,434** N Spearman FLEX3¹ Correlation,440**,385**,430** ,451** N Spearman FLEX4¹ Correlation,726**,934**,635** -,201*,177*,519** N ¹ FLEX1 - Part-time students as % of all students, all ISCED levels combined; FLEX2 - Inflow of students (ISCED 5-6) from EU-27, EEA and Candidate countries - as % of all students in the country; FLEX3 - Did the most recent taught learning activity (non-formal) take place during paid working hour? % of answers "Only during paid working hours" and "Mostly during paid working hours" (variable COURWORH); FLEX4 - Student or apprentice in regular (formal) education during the last four weeks, % of those aged who answered "Has been a student or apprentice" and "Student on holiday" (variable EDUCSTAT). * - Correlation is significant at 0.05 level (2-tailed). ** - Correlation is significant at 0.01 level (2-tailed). Coefficients equal or above 0.300** are marked by dark shading. Source: EUROSTAT, calculations by the authors Flexibility is named as a key principle in enabling participation in adult education and training (Keogh, 2009). Indeed, results of the Adult education survey show that flexibility in provision is of utmost importance for adults when they make choices whether to participate in learning. Among the types of obstacles to participation in training reported as the most important in 2007, lack of time because of family responsibilities (22.4%) and the fact that training conflicts with work schedule (21.7%) were most frequently mentioned by adults in EU-27. In Cyprus, for example, difficulties in reconciling learning and family life were reported by almost 48% of adults. Difficulties in reconciling learning and work were especially relevant in the United Kingdom (27%) and Lithuania and Slovenia (approx. 26%). Reconciliation of training and work was especially relevant for the highly -qualified: even 29.3% of EU-27 high-qualified adults selected this obstacle as the most important. Furthermore, 5.8% of EU-27 adults indicated that the most important obstacle to participating in training was that 178

181 there was no training offered in the area.. Therefore other flexible forms of training provision (eg. distant learning, personalised learning) could have a positive long-term impact for the overall country performance in adult learning. Confintea VI synthesis report states that personalised courses where the learner receives face-to-face, correspondence and/or online learning support are the ultimate in flexibility. According to this report, documents that distance learning is on the increase in the majority of countries (Keogh, 2009) Governance of institutions School autonomy in deciding on various aspects of school policy and management seems to have a significant influence on the overall level of adult learning and adult upper secondary education attainment (see Table 9). The most important aspect of school autonomy is its role in deciding on budget allocations within the school: the higher is the school s financial autonomy, the higher is the level of adult learning (medium correlation) and tertiary education attainment (weak correlation). Other aspects of school autonomy (selecting which teachers to hire, establishing and determining the increase of their salaries, deciding on learning material and course content) have a moderate, but still a significant influence either on the overall level of adult learning and/or adult upper secondary education attainment. Table 9: The influence of the governance of institutions on adult learning and education attainment Dependent Adult Intensity variables participation Adult Adult of adult Adult upper Adult in education participation participation in education secondary tertiary and training in formal non-formal and education education (both formal education education and training attainment attainment Indicators of the and nonformalformal) and training training (non- governance of institutions Spearman GOVERNANCE1¹ Correlation,415**...,337** N Spearman GOVERNANCE2¹ Correlation,295*...,355** N Spearman GOVERNANCE3¹ Correlation,348*...,431** N Spearman GOVERNANCE4¹ Correlation,566** ,455** N Spearman GOVERNANCE5¹ Correlation,351**... -,257* N Spearman GOVERNANCE6¹ Correlation,397** N Spearman GOVERNANCE7¹ Correlation,351** N ¹ Percentage of students in schools where the principal reported that only school has considerable responsibility for the following aspects of school policy and management, PISA 2006: GOVERNANCE1 selecting teachers for hire; GOVERNANCE2 - establishing teachers' starting salaries; GOVERNANCE3 determining teachers' salary increases; GOVERNANCE4 - deciding on budget allocations within the school; GOVERNANCE5 - choosing which textbooks are used; GOVERNANCE6 determining course content; GOVERNANCE7 - deciding which courses are offered. * - Correlation is significant at 0.05 level (2-tailed). ** - Correlation is significant at 0.01 level (2-tailed). Coefficients equal or above 0.300** are marked by dark shading. Source: EUROSTAT, calculations by the authors 179

182 Financing of education and training Statistical analysis shows that both absolute (defined as % of GDP) and relative (defined as EUR PPS per pupil/student) spending on education and training has significant influence 60 on the overall level of adult learning (esp. non-formal) and adult tertiary education attainment (see Table 10). However, it is likely that a more efficient use of funds available for education and training would have an increasingly positive impact on both adult learning and education attainment in a country. Furthermore, private sector investment in education and training has a considerable impact on a country s performance in adult learning. The results of the analysis show that the more companies spend on vocational training, either in absolute amounts or per participant in continuing VET courses, the higher is the overall level of adult learning in a country (absolute spending is even more significant for low-skilled participation in education and training correlation with Finance 1 indicator was 0,639**). Therefore, it is likely that the development of cost-sharing mechanisms to finance education and training could have a considerable long-term impact on a country s performance in adult learning. Table 10: The influence of financing of education and training on adult learning and education attainment Dependent Adult Intensity variables participation Adult Adult of adult Adult upper Adult in education participation participation education secondary tertiary and training in formal in non-formal and education education (both formal education education and training attainment attainment Indicators of financing and nonformalformal) and training training (non- of education and training Spearman FINANCE1¹ Correlation,601** 0.096,726** 0.271,170*,564** N Spearman FINANCE2¹ Correlation,424** 0.058,539** 0.14,249**,484** N Spearman FINANCE3¹ Correlation,663**,203*,839**,499** 0.141,667** N Spearman FINANCE4¹ Correlation,607**,184*,601** ,560** N Spearman FINANCE5¹ Correlation,455** 0.166,602** ,502** N Spearman FINANCE6¹ Correlation,519** 0.066,447** ,518** N Spearman FINANCE7¹ Correlation,518** 0.157,535* N Spearman FINANCE8¹ Correlation,496** -0.01,422* N ¹ FINANCE1 - Total public expenditure on education as % of GDP, for all levels of education combined; FINANCE2 Total public expenditure on education as % of GDP, at secondary level of education (ISCED 2-4); FINANCE3 - Total public expenditure on education as % of GDP, at tertiary level of education (ISCED 5-6); FINANCE4 - Annual expenditure on public and private educational institutions per pupil/student in EUR PPS, for all levels of education combined, based on full-time equivalents; 60 Absolute spending for all levels of education combined (Finance 1) and for ISCED 2-4 level (Finance 2) has less significant impact on low-skilled respective correlations are 0,530** and 0,315**. However absolute spending on tertiary level of education (Finance 3) is even more significant for the intensity of low-skilled education and training (correlation is 0,565**). 180

183 FINANCE5 - Annual expenditure on public and private educational institutions per pupil in EUR PPS, at secondary level of education (ISCED 2-4), based on full-time equivalents; FINANCE6 - Annual expenditure on public and private educational institutions per student in EUR PPS, at tertiary level of education (ISCED 5-6), based on full-time equivalents; FINANCE7 - Vocational training costs as % of total labour cost, C_TO_O_NOT_L - All NACE branches except agriculture, fishing, public administration, private households and extra-territorial organisations; FINANCE8 - Direct cost of CVT courses per participant, in Purchasing Power Standard (PPS). * - Correlation is significant at 0.05 level (2-tailed). ** - Correlation is significant at 0.01 level (2-tailed). Coefficients equal or above 0.300** are marked by dark shading. Source: EUROSTAT, calculations by the authors Correlation results are consistent with the outcomes of Adult education survey. According to this survey, in % percent of adults living in EU-27 stated that the most important obstacle to participate in training was the fact that training was too expensive or that they could not afford it. It was the third most important obstacle identified by adults after reconciliation of training with work and family life. Cost of training was especially relevant for adults living in new EU member states or candidate countries and those with medium qualifications (ISCED 5-6). For example, 43% of adults in Bulgaria, 35% in Poland, 31.5% in Croatia, 29% in Lithuania, 28,5% in Estonia, Latvia and Slovenia could not afford training because of its cost or difficult financial situation in the family. 18.8% of medium qualified adults shared the same opinion. In addition, 5.1% of EU-27 adults indicated that they lack employer support in financing of training activities. The share of adults with such opinion was the highest in Czech Republic and Norway (approx. 14%). Again, lack of employer support was most frequently mentioned by the medium-qualified (6.8%). Adult literacy and life skills survey also confirm the importance of employer financing for participation in learning: according to the survey, employer financing plays a central role in supporting opportunities to engage in lifelong learning in all countries; furthermore, the levels of worker engagement in literacy and numeracy practices on the job are strongly associated with the likelihood of benefiting from employer-sponsored adult education and training (Learning a living, 2005). Consequently, the importance of employer support justifies the argument that costsharing mechanisms to finance education and training could have a positive long-term impact on a country s performance in adult learning Human resources The ratio of students to teachers and the age of teachers are the only human resource indicators which have an impact on adult learning and education attainment (see Table 11). The ratio of students to teachers seems to affect upper secondary education attainment: countries with a high ratio tend to have relatively more students and, as a result, larger share of them attaining upper secondary education. Additionally, it seems that the higher the number of students in a class, the less time the teacher has for individual work, so fewer students are motivated to learn and they spend less time learning in later years of their lives. Another interesting result is the link between age of the teachers and adult learning and education attainment. It seems that elder, more experienced teachers have a larger impact on adult learning (esp. non-formal) and tertiary education attainment. 181

184 Dependent variables Table 11: The influence of human resources on adult learning and education attainment Adult participation in education and training (both formal and nonformal) Adult participation in formal education and training Adult participation in non-formal education and training Intensity of adult education and training (nonformal) Adult upper secondary education attainment Adult tertiary education attainment Indicators of human resources The ratio of Spearman students to Correlation ,301**,317** 0 teachers (ISCED 3) (HUMAN1) N Teachers aged >50 teaching in Spearman public and Correlation,454**,305**,420** 0.007,176*,414** private at ISCED levels as % of total teachers teaching in N ISCED levels 2-3 (HUMAN2) * - Correlation is significant at 0.05 level (2-tailed). ** - Correlation is significant at 0.01 level (2-tailed). Coefficients equal or above 0.300** are marked by dark shading. Source: EUROSTAT, calculations by the authors To summarise, the most influential policy characteristics of the education system are stratification (the number of school types or distinct educational programmes and the first age of selection in the education system), vocational specificity (private sector involvement in the provision of initial VET and the prevalence of job related non-formal learning), flexibility in provision of learning services (inflow of students and proper opportunities for those aged to participate in formal education and training), school autonomy (esp. in budget allocations within the school) and spending on education (absolute and relative per pupil/student) Examples of reforms in flexibility of learning provision and learning financing The examples of education policy reforms cover the period from 1998 to The statistical analysis has already shown that one of the most important factors influencing adult learning is spending on education and training as well as flexibility in provision of learning services. Moreover, a common Education and Training 2010 work programme and subsequent EU policy documents call for reform and investment (public and private) in education and improved access to lifelong learning for all. Therefore, the examples focus on the reforms aimed at financing of education and training and flexibility in provision of learning services. The examples from different countries were selected on of the following criteria: Spain (made the biggest progress of all Mediterranean countries, represents the Mediterranean tradition and the 15 older EU Member States); Slovenia (made the biggest progress of all transition states of Central-Eastern Europe, represents transition states of Central-Eastern Europe and the 12 newer EU Member States); Austria (made the biggest progress of all Western European countries, represents Western European countries and the 15 older EU Member States); Norway (represents the Nordic tradition and the EFTA-EEA countries). 182

185 Spain 61 A number of education reforms took place in Spain before 1998, the most significant among them being the decentralisation of education ( ) and the development of vocational training and adult learning ( ). As a result, school autonomy, flexibility in provision of learning and stronger local partnerships were achieved and an integrated national vocational training system was established. The Tripartite Agreement on Continuous Training was reached between the employers associations, trade union organisations and the Government, which was renewed three times ( and ). A new process of reform started in with the passing of three laws: the Organic Law on Universities (2001), the Law on Qualifications and Vocational Training (2002) and the Law on the Quality of Education (2002). Universities were given economic and financial autonomy, maximum flexibility in developing their teaching and mobility policies, and external quality evaluation mechanisms were created. Measures in vocational training were directed towards both employed and unemployed, and those continuing in education. They introduced devices and mechanisms that prioritise ongoing learning/training and created the National System of Qualification and VET. The basis for the recognition of informal and non-formal learning, and for the validation and accreditation of apprenticeships was established. The provision of various forms of training (initial vocational training, in-service training in undertakings and on-the-job vocational training) strengthened open and flexible training and increased learning opportunities for adults. In 2004 the government of Spain proposed a new VET model, with the aim of strengthening its effectiveness through the improvement of quality, allowing its decentralisation by giving the management to regional governments and improving institutional administration. In this frame a new Training for Employment Model was created. This new system integrated Occupational and Ongoing Training, prioritising access to training during work life. The new model was destined towards all workers employed or unemployed, including those who do not receive a salary, prioritising groups with greater difficulties of integration (women, over 45 year old, employees of small or medium sized businesses, people with disabilities and workers with minimal qualifications). In line with the reforms in education and due to increased immigration, the Spanish government has carried out immigration and social inclusion policies. The first National Plan of Action for Social Inclusion, adopted in 2001, aimed to integrate immigrants and other groups at risk of social exclusion into the labour market and to promote access of the immigration population to public services, including education and training. 61 EURYDICE papers: Key topics in education in Europe. Financing and Management of Resources in Compulsory Education, Volume 2, Spain; Vocational Guidance Education in Full-Time Compulsory Education, Spain, school year 2007/2008; Structures of Education, Vocational Training and Adult Education Systems in Europe, Spain, 2003; CEDEFOP paper National ReferNet reports on progress in the policy priority areas for Vocational Education and Training, Spain,

186 Slovenia 62 Although general education reform was the first priority of the Slovenian government, in 1996 the Parliament adopted 6 new educational acts laying down the foundation for wide-ranging structural reforms of vocational education and training. The following fundamental innovations were implemented: 1) the creation of the dual system, characterised by alternation of inschool education and in-firm practical training (apprenticeship); 2) the modernisation of curricula; 3) the introduction of transition options to become master craftsman and the necessary vocational upgrading after finishing vocational school, secondary technical and professional school or general secondary school; 4) the introduction of post-secondary vocational schools (colleges), which offer two year profession-oriented courses and have a strong practical orientation. With adopting the National Professional Qualification Act in 2000, Slovenia enacted a system of assessment and certification of non-formal and informal learning and work experiences for the needs of the labour market. The decentralization of vocational education and training policy making to a regional level and the involvement of social partners has enabled schools to become stronger actors in the education market, with more flexibility and responsibility. Alongside gradual reforms, Slovenia used pre-accession instruments both to stimulate the reforms and to achieve immediate results in adult education and training. For example, in ,000 unemployed persons from industries affected by restructuring were trained in basic ICT skills and received recognised certificates, as the result of a Phare programme. Slovenia has also used ESF to modernize VET through the programme reform, improvement of the competences of teachers, increasing ICT use and infrastructure, reducing the education deficit of specific groups (adults, school leavers, unemployed). Funds in the National Budget have been redirected from other sources to increase the participation of adults in formal education. Legislation on income tax included tax relieves for investment in learning and knowledge. Initiatives were also undertaken to stimulate private investment in education by directly subsidising individuals ( voucher system ). Slovenia has managed to dramatically increase the number of computers in schools and provide access to the Internet to all schools, by committing publicly-funded schools to raise up to 50% for the purchase of mutually agreed ICT hardware. However, the efficiency of education spending in Slovenia remains weak in a multiple outcome model (comprising primary pupil-teacher ratios, secondary school enrolment rates and the proportions of university graduates relative to the school-age population 63 ). 62 Investing in and financing VET in an efficient and sustainable way, country report, Slovenia; Vocational Education and Training reform in Slovenia, European Training Foundation report, 1997; Implementing Lifelong Learning Strategies in Europe: Progress report on the follow-up to the 2002 Council resolution, 2003; The Development and State of the Art of Adult Learning and Education, national report of Slovenia, Todd Mattina and Victoria Gunnarsson Budget Rigidity and Expenditure Efficiency in Slovenia,

187 Austria 64 Major reforms in education in Austria have been carried out since 1990 with the aim of making schools autonomous in terms of management, funding and operational expenditure. Since 1993 schools have been legally empowered to issue their own curricular regulations autonomously, i.e. to develop their own profile within a given framework. Since 1996 all Austrian schools can, in a limited way, control the funds allocated to them by the school authorities, which makes occupation-oriented, project-based forms of education (e.g. company practice ) easier to implement. Provisions governing school autonomy at prevocational schools enable a flexible response to the vocational interests of pupils and the respective demands of the particular region. As a consequence, Austria has created a diversity of education providers and the opportunities provided by second-chance education, which are considered to be the strengths of the qualification and skills strategy pursued by the federal and provincial governments. In 2002 Austria extended options for training allowances for employers and depreciations for employees as a form of indirect financing or incentive financing by the State. Measures stimulating private investment from enterprises, households, and individuals in the field of VET are mainly present in the dual training system in the form of financial incentives for apprentices and for training enterprises. The introduction of a grant for company-based training has enhanced the companies willingness to provide training and the number of available apprenticeship posts has increased. Other incentives have also been introduced for apprentices and training enterprises (e.g. apprenticeship training premium, relief for non-wage labour costs, etc). Furthermore, a nationwide network comprising a total of 18 apprentices advisors has been set up to establish direct contacts with the companies. In the period from 2004 till 2006, more than 13,000 companies were contacted and about 5,700 commitments for admission of apprentices obtained. In the adult learning sector, different systems of individual promotions have been established in various provinces, e.g. a comprehensive model of the educational account, the reimbursement of course fees (from 50% to 80%) for certain target groups, the introduction of the educational voucher in In the period from 2002 to 2006, some 290,000 educational vouchers were requested Norway 65 Norway reached the EU benchmark of lifelong learning for 2010 in By that time it had built an education system characterised by a high degree of local autonomy and a focus on cooperation between authorities, social partners and providers of education. 64 EURYDICE paper Key Topics in Education in Europe. Austria ; OECD paper Thematic Review on Adult Learning Austria, Background Report 2003; Structures of Education, Vocational Training and Adult Education Systems in Europe, Austria, 2003 ; Investing in and financing VET in an efficient and sustainable way, Country Report, Austria; Austrian National Report 2007 on the Implementation of the Education and Training 2010 Work Programme ; VET Policy Report, Austria, 2008; National Report on the Development and State of the Art of Adult Learning and Education (ALE) Austria, Prepared for the UNESCO 6th International Conference on Adult Education (CONFINTEA VI). 65 Summary of National ReferNet report on progress in the policy priority areas for Vocational Education and Training. Norway, 2008; Thematic Review on Adult Learning. Norway, country note, OECD

188 In 1999 the Competence Reform, or the national lifelong learning strategy, was launched. It aimed to improve access to formal education and training for second chance learners through the establishment of a national framework for validation of non-formal and informal learning and flexible training programmes. It was initiated by the Norwegian Confederation of Trade Unions (LO) and based on a tripartite agreement between LO, the Confederation of Norwegian Enterprise (the main employer organisation) and the government. The main elements of the reform were: (i) flexible learning: educational programmes developed and adapted to the needs of adults; (ii) framework conditions for the individual: right to leave of absence for educational purposes, subsistence funding, educational funding scheme, improved opportunities for combining work and education; (iii) a competence-building programme: help in producing innovation and development in the field of continuing education and training, (iv) documentation and assessment of nonformal and informal learning in working life and in relation to the education system, (v) "Popular enlightenment" and democratic participation; (vi) a new chance: primary, lower secondary and upper secondary education for adults; (vii) structural changes to the public education system: new organisation of schools and higher education, encouraged co-operation between providers of education in developing educational models, network development, reviewed legislation, rules and agreements; and (viii) motivation and information. The reform gave adults a statutory right to upper secondary and primary and lower secondary education, access to higher education based on assessment of non-formal competencies, including work practice. From 2001 adult employees were given the legal right to study leave, study financing arrangements have been strengthened to better meet adults needs during periods of further training, and systems of validation of prior learning have been tested and established in each county Labour market policies The EU, EFTA-EEA and candidate countries faced a variety of labour market problems over the past decade. They ranged from a high level of unemployment to a high shortage of skills. Sometimes the unemployment and the shortage of skills co-existed, reflecting a wide qualitative mismatch between the supply and demand of skills. Adult learning was frequently perceived as the main cure for different types of mismatch and it was included in almost any labour market policy mix. The countries also searched for the underlying causes of the labour market imbalances. This often pointed to both the lack of responsiveness of the education system to labour market needs and the lack of flexibility in labour markets, preventing the efficient allocation of skills in the economy and the failure to send proper signals to the education system and individual learners. As many European industries declined under global competitive pressures or had to move constantly into higher value-added production, the labour market also required structural adjustments. Millions of skilled employees in the middle of their careers had either to acquire new qualifications or face deteriorating labour conditions or long-term unemployment. In some countries generous unemployment benefits have been perceived as discouraging labour market adaptation, while in others they have been the key element of it, providing the unemployed with more time to acquire new qualifications and allowing them to find more productive and sustainable jobs instead of temporary low-skill 186

189 Indicators of labour market policies Difficulty of hiring index Dependent variables Spearman Correlation alternatives. Still most countries have perceived the need to revisit the balance between labour market flexibility and employment security, looking into employment protection legislation, unemployment benefits, active labour market policies and their complementarity. Below we test the following statements provided in the methodology chapter: Lower employment protection has a positive influence on adult learning and educational attainment, either directly or through the increased mobility of employees in the labour market and a better balance between labour supply and demand; Higher unemployment benefits have a positive influence on adult learning and educational attainment; Higher expenditure on labour market training and services has a positive influence on adult learning and educational attainment The direct influence of labour market policies on adult learning and education attainment Our correlation analysis (see results in Table 12) shows that while overall employment protection measures demonstrate only a weak correlation with adult learning, the high rigidity of hiring, hours, employment and low safety at work in certain cases could have a negative influence (especially for the low-skilled participation ) in adult learning. On the contrary, the greater the share of the labour force in part-time employment demonstrating labour market flexibility has a medium to strong influence on adult learning and especially on their participation in non-formal learning. In addition, more widespread part-time working has a positive influence on tertiary education attainment, possibly via a combination of parttime work and part-time studies. Higher unemployment benefits indeed have some positive influence: low to medium on adult learning, especially non-formal, and low on tertiary educational attainment. Higher expenditure on labour market training and services also has a low to medium correlation with adult learning, demonstrating especially a positive influence on participation in non-formal learning. Greater participation in adult learning in general and non-formal learning in particular seems to be closely associated with the entire flexicurity mix: high income security ensured by high unemployment benefits, high employment security generated by active labour market policy (the latter is especially relevant for low-skilled) and low employment rigidity ensured by liberal labour market regulation. We will also test this conclusion by looking into the indirect influences of labour market policies. Table 12: The direct influence of labour market policies on adult learning and education attainment, data for the EU-27 plus EFTA-EEA and candidate countries Adult participation in education and training (both formal and non-formal) Adult participation in formal education and training Adult participation in non-formal education and training Intensity of adult education and training (nonformal) Adult upper secondary education attainment Adult tertiary education attainment -0,240** -0,312** -0,251** 0,004-0,240** -0,149 N

190 Indicators of labour market policies Rigidity of hours index Dependent variables Spearman Correlation Adult participation in education and training (both formal and non-formal) Adult participation in formal education and training Adult participation in non-formal education and training Intensity of adult education and training (nonformal) Adult upper secondary education attainment Adult tertiary education attainment -0,352** -0,344** -0,345**,289** 0,104-0,185* N Spearman Rigidity of -0,287** -0,322** -0,302** 0,18-0,104-0,154 Correlation employment index N Index of the number of serious Spearman accidents at work Correlation -0,06-0,342** -0,061-0,403** -0,164* 0,062 per 100 thousand persons in employment N Temporary Spearman employment as a 0,237** 0,157** 0,376** 0,236** 0,044 0,053 Correlation percentage of the total employment N Part-time Spearman 0,679** 0,166** 0,711** -0,073-0,043 0,462** employment as a Correlation percentage of the N total employment Population in employment having a second Spearman Correlation 0,150* 0,046 0,154 0,319** -0,023 0,031 job N Social protection benefits (unemployment function) - PPS per inhabitant Spearman Correlation 0,495** 0,118 0,634** 0,271** -0,223** 0,434** N LMP expenditure Spearman on labour market Correlation 0,422** 0,089 0,631** 0,056-0,101 0,300** services N Spearman LMP expenditure 0,469** -0,185* 0,577** 0,177 0,049 0,309** Correlation on training N * - Correlation is significant at 0.05 level (2-tailed). ** - Correlation is significant at 0.01 level (2-tailed). Coefficients equal or above 0.300** are marked by dark shading. Sources: EUROSTAT, World Bank, calculations by the authors The influence of labour market policies on the balance of labour supply and demand The first potential indirect link between labour market policies and adult learning and educational attainment comes through the influence of these policies on the balance of labour market supply and demand. We have expected that liberal labour market regulation, higher unemployment benefits and higher expenditure on active labour market policies would mean lower unemployment rates and higher population activity rates. Correlations in table 13 provide some support to these hypotheses. The rigidity of hours indeed has some negative influence on employment and activity rates, while higher unemployment benefits and higher expenditure on active labour market policies have a positive influence. The share of the labour force in part-time employment has a positive influence on almost all labour market indicators, especially on employment and activity rates (medium correlation) and those of the low-qualified. The higher rigidity of employment and the higher firing costs in some cases could lead to higher unemployment rates of the labour force with tertiary qualifications. Strict employment protection legislation prevents the 188

191 Indicators of labour market policies Balance of labour supply and demand displacement of the lower qualified by the higher qualified individuals. Table 13: The correlation between labour market policies and the balance of labour supply and demand, data for the EU-27 plus EFTA-EEA and candidate countries Unemployment rates, yearly averages Employme nt rate (%) Employme nt rate of older workers (%) Populati on activity rate (%, 15 to 64 years) Unemployme nt rates by highest level of education attained (%) ISCED levels 0-2 Unemployme Unemployme nt rates by nt rates by highest level highest level of education of education attained (%) attained (%) - ISCED - ISCED levels 3-4 levels 5-6 Difficulty of hiring index Rigidity of hours index Rigidity of employment Spearman Correlation 0,195* -0,164-0,032-0,162-0,127 0,167 0,345** N Spearman 0,256* -0,357** -0,289** 0,362* 0,262** 0,230* 0,224* Correlation * * N Spearman Correlation 0,176-0,194* -0,117-0,186* -0,041 0,253** 0,387** N Firing costs (weeks of wages) Social protection benefits (unemployme nt function) PPS per inhabitant LMP expenditure on labour market services Spearman Correlation 0,196* -0,133-0,065-0,093 0,084 0,211* 0,305** N Spearman Correlation - 0,260* * 0,420** 0,121 0,348* * -0,212** -0,184** -0,045 N Spearman Correlation -0,125 0,383** -0,002 0,393* * -0,046-0,073-0,165 N LMP Spearman 0,510* -0,162* 0,474** 0,232** -0,12-0,199* 0,032 expenditure Correlation * on training N Part-time - Spearman 0,607* employment 0,484* 0,667** 0,416** -0,467** -0,345** -0,274** as a Correlation * * percentage of the total N employment * - Correlation is significant at 0.05 level (2-tailed). ** - Correlation is significant at 0.01 level (2-tailed). Coefficients equal or above 0.300** are marked by dark shading. Sources: EUROSTAT, World Bank, calculations by the authors The influence of the balance of labour supply and demand on adult learning and education attainment The previous chapter has shown the link between the flexicurity policy mix and labour market indicators. This chapter demonstrates how the balance of labour supply and demand is related to adult learning and educational attainment. In the methodology chapter we have expected that the periods of shortage of labour (higher employment and activity rates and lower unemployment) would stimulate greater investment in adult learning, leading to both higher participation and attainment. The results in table 14 support our hypothesis. Employment and population activity rates have a strong correlation with overall adult participation in learning (population activity rate slightly less 189

192 Indicators of balance of labour supply and demand Employment rate (%) Employment rate of older workers Dependent variables Spearman Correlation correlates with the low-skilled participation in education and training) and especially non-formal learning, while unemployment rates have low to medium inverse correlations with the same indicators. The Adult literacy and life skills survey confirms this relationship: adults with higher levels of skill of the types measured in the survey tend to be more employable than adults with low skills; the findings of this survey also show that low skilled adults are more likely than medium to highly skilled adults to experience unemployment in half of the countries and to experience labour force inactivity for six or more months in all countries except Bermuda (Learning a living, 2005). Adult tertiary education attainment has medium correlations with employment and activity rates, as well as average exit age from the labour force. This might point to a two-way relationship whereby the tertiary education attainment level of the population is both effect and, even more likely, the cause of high population employment and activity rates. Higher educational attainment is also associated with better health and longevity leading to later exit from the labour market. Table 14: The correlation between the balance of labour supply and demand and adult learning and education attainment, data for the EU-27 plus EFTA-EEA and candidate countries Adult participation in education and training (both formal and nonformal) Adult participation in formal education and training Adult participation in non-formal education and training Intensity of adult education and training (nonformal) Adult upper secondary education attainment Adult tertiary education attainment 0,765** 0,347** 0,803** -0,139 0,225** 0,600** N Spearman Correlation 0,443** 0,401** 0,471** -0,297** 0,199** 0,601** (%) N Spearman Unemployment -0,458** -0,169** -0,503** 0,162 0,102-0,228** Correlation rates, yearly averages (%) N Average exit age from the labour force (weighted by the probability of withdrawal from the labour market) Population activity rate (%, 15 to 64 years) Unemployment rates by highest level of education attained (%) ISCED levels Spearman Correlation 0,370** 0,379** 0,232* -0,252** 0,011 0,509** N Spearman Correlation 0,753** 0,314** 0,786** -0,131 0,365** 0,627** N Spearman Correlation -0,239** -0,082-0,330** 0,012 0,469** -0, N Unemployment rates by highest level of education attained (%) ISCED levels Spearman Correlation -0,331** -0,136* -0,382** 0,257** 0,112-0,158* 3-4 N Unemployment rates by highest Spearman level of education Correlation -0,403** -0,05-0,373** 0,286** -0,160** -0,095 attained (%) ISCED levels 5-6 N * - Correlation is significant at 0.05 level (2-tailed). 190

193 ** - Correlation is significant at 0.01 level (2-tailed). Coefficients equal or above 0.300** are marked by dark shading. Source: EUROSTAT, calculations by the authors Indicators of labour market policies Rigidity of hours index LMP expenditure on labour market services The influence of labour market policies on labour market mobility The second potential indirect link between labour market policies and adult learning and educational attainment comes through the influence of these policies on the mobility of the labour force. We expected that liberal labour market regulation, higher unemployment benefits and higher expenditure on active labour market policies would lead to higher mobility of the labour force. The correlation results in table 15 provide only a very limited support to our hypothesis. While the logic seems to hold for the employment mobility index, which has low inverse correlation with the rigidity of hour index, a low direct correlation with expenditure on labour market services and a medium correlation with part-time employment. However, the first two coefficients are significant only at 0.05 level (2-tailed). It is very important to note that these correlations are based on a very small sample and therefore it is impossible to verify or falsify our hypothesis. Table 15: The correlation between labour market policies and labour market mobility indexes, data for the EU-27 plus EFTA-EEA and candidate countries Labour market mobility indexes Spearman Correlation Occupational mobility index Employment mobility index Job-to-job mobility index Average job mobility index -0,015-0,470* -0,288-0,209 N Spearman Correlation -0,05 0,454* 0,039 0,059 N Spearman Part-time employment as a Correlation 0,207 0,534** 0,296 0,312 percentage of the total employment N * - Correlation is significant at 0.05 level (2-tailed). ** - Correlation is significant at 0.01 level (2-tailed). Coefficients equal or above 0.300** are marked by dark shading. Source: EUROSTAT, calculations by the authors The influence of labour market mobility on the balance of labour supply and demand The previous chapter has explored the link between flexicurity policy mix and labour market mobility. This chapter demonstrates how the mobility of labour in its own turn is related to the balance of labour supply and demand. In the methodology chapter we have expected that higher mobility of labour would raise employment and activity rates and decrease unemployment, which as we have proved in chapter would stimulate greater investment in adult learning, leading to both higher participation and tertiary education attainment. The correlation results provided in table 16 support our hypothesis. Indeed, all the mobility indices are related both to higher employment (especially that of older workers) and activity rates (medium to high correlation), while the employment mobility index 191

194 Balance of labour supply and demand Labour market mobility indexes is also related to lower unemployment. Thus, the mobility of labour could have a positive indirect influence adult learning, through the influence on labour market supply and demand. However, as discussed previously, correlations involving mobility indices have to be treated cautiously due to small samples. Table 16: The correlation between labour market mobility indexes and the balance of labour supply and demand, data for the EU-27 plus EFTA-EEA and candidate countries Unemployment rates, yearly averages Employment rate (%) Employment rate of older workers (%) Average exit age from the labour force (weighted by the probability of withdrawal from the labour market) Population activity rate (%, 15 to 64 years) Unemployment rates by highest level of education attained (%) - ISCED levels 3-4 Spearman Employment -0,511** 0,698** 0,515** 0,388 0,625** -0,498* Correlation mobility index N Spearman Job-to-job -0,475* 0,491* 0,687** 0,539* 0,439* -0,264 Correlation mobility index N Spearman Average job -0,429* 0,555** 0,766** 0,587* 0,516** -0,417* Correlation mobility index N Spearman Occupational -0,121 0,211 0,509** 0,403 0,212-0,199 Correlation mobility index N * - Correlation is significant at 0.05 level (2-tailed). ** - Correlation is significant at 0.01 level (2-tailed). Coefficients equal or above 0.300** are marked by dark shading. Source: EUROSTAT, calculations by the authors The influence of labour market mobility on adult learning and education attainment The previous chapter has explored the influence of labour market mobility on the balance of labour supply and demand and demonstrated the possibility of indirect influence on adult learning and educational attainment. This chapter will explore the direct links between the mobility of labour and adult leaning and educational attainment. In the methodology chapter we have expected that higher mobility of labour would lead to higher adult participation in learning and higher education attainment. The results provided in table 17 give some support to our hypothesis. Employment mobility is correlated with adult participation in learning (especially non-formal). This might reflect that labour market transitions have demanded new skills, which were more frequently obtained non-formally. Job-to-job mobility and average job mobility indices have a positive relationship with tertiary education attainment (medium correlation). This might also reflect the function of tertiary education as an insurance against unemployment in less protected, more liberal labour markets. At the same time the job-to-job mobility index is inversely correlated with the intensity of adult participation in non-formal learning (medium correlation). This might reflect that lower employment protection and higher mobility of labour discouraged the investment of employers in the portable skills of their employees (especially of the highly-skilled as job-to-job mobility index correlation with the intensity of education and training of lowskilled is insignificant). As expected, they would still train their employees, but training would be limited to a minimum of company-specific skills. In protected labour markets only established highly-skilled workers, who have made careers in their 192

195 Dependent variables Labour market mobility indexes Employment mobility index Job-to-job mobility index Average job mobility index Spearman Correlation companies, would be entitled to more intensive job-related nonformal training. However, as discussed previously, correlations involving mobility indices have to be treated cautiously due to small samples. Table 17: The correlation between labour market mobility indexes and adult learning and education attainment, data for the EU-27 plus EFTA-EEA and candidate countries Adult participation in education and training (both formal and nonformal) Adult participation in formal education and training Adult participation in non-formal education and training Intensity of adult education and training (nonformal) Adult upper secondary education attainment Adult tertiary education attainment 0,681** 0,427* 0,686** -0,217-0,167 0,416* N Spearman Correlation 0,464* 0,420* 0,349-0,563** 0,245 0,669** N Spearman Correlation 0,436* 0,414* 0,362-0,448* 0,012 0,578** N * - Correlation is significant at 0.05 level (2-tailed). ** - Correlation is significant at 0.01 level (2-tailed). Coefficients equal or above 0.300** are marked by dark shading. Source: EUROSTAT, calculations by the authors The above statistical analysis has provided support to most of our hypotheses about the influence of labour market policies on adult learning and educational attainment. The flexicurity policy mix or its individual elements (lower rigidity of hours, more liberal parttime employment, higher unemployment benefits and expenditure on active labour market policies) seem to have correlated with our dependent variables directly or through intervening variables in most of the EU, EFTA-EEA and candidate countries most of the time between 1997 and Examples of flexicurity and other related labour market reforms The term flexicurity, first coined in the 1990s, defines a labour market regime which combines flexible labour market (including dismissal) rules, substantial social benefits and active labour market policies (including lifelong learning promotion). The model of flexicurity reforms was developed as a reaction to numerous problems facing European welfare states. They included aging population, low population activity rates and high structural unemployment. The introduction and success of flexicurity reforms in Denmark in 2003 soon led to widespread labour market reforms in many other EU Member States. The policy preferences supported by the EU institutions also played a significant role in popularising such reforms. Labour market reforms, directed at flexicurity, were seen by the EC as a suitable way of achieving one of the main objectives posed by Lisbon Strategy: to reach sustainable growth with more and better jobs (full employment, labour productivity, and social cohesion). Improving the responsiveness of European labour markets was also defined to be crucial to responding to the challenges stemming from the combined impact of globalisation and ageing European societies. The idea of flexicurity is emphasized in a number of EC policy documents, such as the 2006 Annual Progress Report on Growth and Jobs, the 2007 Communication Towards Common 193

196 Principles of Flexicurity: More and better jobs through flexibility and security and the 2007 Green Paper Modernising labour law to meet the challenges of the 21st century. Some European countries have tried to adopt the whole flexicurity package either by developing their own policy or copying the Danish experience. Some other countries have been selective and adopted only individual elements of flexicurity, which seemed to be useful in their particular situation. In this chapter we overview examples of labour market reforms with flexicurity elements in four very different EU Member States: Denmark, where the reforms were first introduced and proved to be very successful, Germany, which had a different approach towards labour market reforms and adopted only elements of the flexicurity package, Italy, which, despite its worse starting position, implemented the labour market and lifelong learning reforms quite successfully, and, finally, Bulgaria, in which the reforms struggled to achieve the sought results Denmark 66 The most recent important labour market reforms in Denmark were implemented in The reforms were designed in such a way that Denmark could maintain its social protection traditions, while at the same time remaining competitive internationally. The reforms targeted Denmark's need to address its challenging demographic profile a rapidly aging population, increased life expectancy, lower birth rates, and a shrinking workforce. The unemployment levels in Denmark were very low, and the main problem was to find suitable workers for vacant positions, and not to decrease unemployment levels. In order to maintain economic competitiveness without jeopardising social cohesion, the reform commission adopted the idea of "flexicurity" (a combination of a flexible labour market, substantial social benefits, and active labour market policies). With regard to greater flexibility, the reform addressed labour market withdrawal by increasing the country's retirement age from 65 to 67 years old and increasing the voluntary early retirement age from 60 to 62 years old. Generous benefits, almost as high as the wages themselves (90% of the wage for low-income groups), were envisaged to workers who lose their jobs, but these workers were also required to aggressively search for a new position. If the unemployed were unable to find a new job, their requalification was financed. One of the main target groups of the reform was the immigrant population (as they had serious unemployment problems, as opposed to the general population). The immigrants were guided into privately run training programmes, and the 66 Sources: Global Envision. Flexicurity Denmark Style. Available from the Internet: Torben M. Andersen, Michael Svarer. Flexicurity Labour Market Performance in Denmark. Available from the Internet: Jon Kvist, Lisbeth Pedersen. Danish Labour Market Activation Policies. Available from the Internet: Tomas Bredgaard, Flemming Larsen. Comparing Flexicurity in Denmark and Japan. Available from the Internet: 194

197 training companies were not fully paid if the trainees were unable to find a job within 13 weeks of completion of the programme. Persons with disabilities or social and personal obstacles to enter the labour market were also considered as a priority group subject for the activation policies Germany 67 The labour market policies in Germany were for a long time protective and put emphasis on labour productivity rather than mobility. This actually had a high price in terms of unemployment, and work time reductions at existing working places did not create a sufficient number of jobs. A number of flexibility elements were introduced into German labour market policy in order to solve the problem of large scale unemployment, but the measures taken were again more targeted at stabilising existing jobs rather than raising mobility between job; this was the German labour market approach. It was thought that introducing the Danish flexicurity model in Germany would actually undermine the fundamental principles of the German labour market constitution, and this was not regarded as an option. The most significant set of reforms tackling unemployment (the socalled Hartz reforms) was adopted during They aimed at, (a) improving labour market services and policy measures in terms of effectiveness and efficiency, (b) activating the unemployed by enforcing the so-called principle of rights and duties (Fördern und Fordern), and, (c) fostering employment demand by deregulating the labour market. Each of these aims had their own policies. The (a) aim was envisaged to be achieved by reorganising local employment agencies (introducing results-based accountability, extension of services provided, orientation to customers, outsourcing possibilities), introducing the quasi markets (competition among public and private employment service providers), targeting of active measures, and evaluation of effectiveness. The (b) goal was to be achieved by reorganising the unemployment benefits system (tying larger benefits to active seeking of job, sanctioning those unwilling to accept jobs), a New policy mix (business start-up subsidies, wage subsidies for disabled and older workers, and introduction of minijobs and midijobs (generating income below a certain threshold and thus fully or partially exempt from social security contributions). In order to achieve the (c) goal, restrictions on synchronisation, reassignment, fixed-term contracts and the maximum duration of temporary employment were abolished, but the restrictive dismissal regulations were left virtually untouched. The reforms have actually been criticised for a number of aspects in the measures taken and were very unpopular in Germany. The critics point to the fact that unemployed people who have savings or who have taken out life insurance are not eligible for unemployment benefits until that money is used up. They are 67 Sources: Kurt Vogler Ludwig. Flexicurity in Germany. European Employment Observatory discussion paper. November Available from Internet: Euractiv.com: Flexicurity: can it work for France and Germany? Available from Internet: Lena Jacobi, Jochen Kluve. Before and After the Hartz Reforms: The Performance of Active Labour Market Policy in Germany. IZA Discussion Paper No April Available from Internet: 195

198 also not entitled to payments if they have close relatives who could support them. This lowers the level of social protection and forces those unemployed people who have made provision for their old age in the form of life insurance or savings to give those up, leaving them at a minimum social welfare level when they reach retirement age. In contrast to the former scheme, unemployed people can also be forced to take any job, irrespective of whether it provides subsistence and whether the job is suited to the unemployed person's professional skills and qualifications Italy 68 The Italian labour market was for a long time considered to be among the most rigid in Europe. At the beginning of the 1990s, the Italian labour market was characterised by a centralised wage bargaining system providing automatic wage indexation, a wide diffusion of undeclared work (especially in the South), a social protection system not providing universal safety nets, but granting a high level of security to workers employed in large firms, an inefficient employment services system, and a very limited provision of active policies and lifelong learning activities. Since the mid-1990s, a number of reforms regarding each of the four flexicurity pillars (flexibility of contractual arrangements, unemployment benefits, active labour market policies, and lifelong learning), that were envisaged to tackle the aforementioned labour market problems, have been initiated in Italy. The first systematic labour market reform started in 1997, when the so-called Treu package was adopted. It introduced temporary help agencies, internship contracts, and a first partial lift of constraints for hiring workers on fixed-term contracts. These constraints were additionally lifted in 2001, when a new law replaced the rigid inventory of admissibility clauses provided by the former regulation with a vague rule, according to which it is possible to hire fixedterm employees for technical, productive, organisational or substitution reasons. In 2003, a so-called Biagi law introduced new atypical contractual arrangements, such as the job on call, the staff leasing and the job sharing contracts, as well as job insertion contract to enhance the employability of younger workers. However, according to the World Bank s Doing Business index, the difficulty of hiring and firing, as well as the rigidity of employment remained the same in Italy in the past few years. The reforms in the active labour market measures included the creation of employment centres (supposed to play an active role in local labour markets, providing a number of services to both employed and unemployed) instead of offices in 1997, and the abolishment of the public monopoly in employment services supply in A number of reforms were also present in lifelong learning, e.g. introduction of inter-professional funds managed by social partners as a source of lifelong learning financing in 2004, and the most important lifelong learning reform in This reform made the provision of adequate and widespread training supply the responsibility of central and local governments and provision of 68 Source: Flexicurity in the Italian Labour Market: Evaluation of Implemented Policy Measures. State of the Art Report and Quantitative analysis Report. Fondazione Giacomo Brodolini, Available from the Internet: Evaluation.pdf. 196

199 personalised training courses obligatory by training providers. The opportunity for unpaid training leave was also introduced. Currently, the measures introduced by the 2000 and 2004 reforms mostly apply to persons with occupational problems. Notwithstanding numerous developments in other components of flexicurity, there was no significant reform in the Italian unemployment benefits system (in 2007 only slight changes in the amount and length of ordinary unemployment benefits were introduced), and it continues to be limited because of strict entitlement conditions Bulgaria 69 The main challenges for the Bulgarian economy, which its labour market reforms market had to tackle, were almost universal among the countries of the CEE: ageing population, significant emigration, high unemployment and low activity rates, and as a result a lack of competent employees in the labour market. The most important labour market reforms in Bulgaria during the transition period took place in 2001, when a new Labour Code was adopted. These legislative changes were aimed at both liberalising the termination of employment at the employer's initiative, and the protection of employees. The Labour Code reform introduced the possibility to end employment relationships without any explanation, if an employer offered generous (not less than four months salary) compensation. Employees who resigned received only minimum unemployment cash benefits. However, the reform also made it more difficult to terminate contracts with the most socially vulnerable groups of employees. At the same time, labour market policies in Bulgaria shifted from passive (which proved to be ineffective in decreasing unemployment and increasing activity) to active. The new Employment Protection Act was put into force since Before the adoption of the Employment Protection Act, active and passive labour market policies were funded from the same source. After the adoption of this act, however, separate funds, managed by different public institutions, were established for payment of unemployment benefits and for active labour market measures. The active labour market policies included increasing the flexibility of different training programmes. The priority groups targeted by the employment support programmes were disadvantaged groups, e.g. long-term unemployed youth, single mothers, etc. Along with national programmes, regional programmes targeted at less developed regions were also introduced Health care policies Health care policies have only a very indirect (non linear) impact on adult learning and education attainment, either through the overall quality of the health care system (i.e. the Euro Health Consumer Index) or through the outcomes of the health care system. Results in Table 18 indicate that the general index measuring the quality of health care systems has only a very weak influence on the intensity of non-formal adult learning: the higher 69 Source: Sandrine Cazes, Alena Nešporová. Flexicurity a relevant approach in Central and Eastern Europe. International Labour Organisation,

200 the quality of health care in a country - the lower the intensity. Indeed, Figure 6 (see chapter 1.1) shows that countries ranking high in terms of adult learning have low adult learning intensity and vice-versa. The results indicate that these countries tend to have well developed economies (eg. Finland, Denmark, Sweden) and high quality health care systems. However, life expectancy, which is the main outcome of health care systems, has a notable influence on adult learning (esp. nonformal) and education attainment. The cause and effect relationship here is unclear: either adult learning has positive effects on health and its major outcome or better health leads to more learning (esp. non-formal). Academic literature in the field tends to support the former relationship. Interestingly, the higher the life expectancy, the lower is adult upper secondary education attainment. Most likely, countries with high life expectancy rates have larger numbers of senior adults who do not have upper secondary education. However, the countries which are characterised by high life expectancy also have strong economies with a large share of university graduates, which is a necessary asset to sustain their economic growth. Table 18: The influence of health care settings and outcomes on adult learning and education attainment Dependent Adult Intensity variables participation Adult Adult of adult Adult upper Adult in education participation participation education secondary tertiary and training in formal in non-formal and education education (both formal education education and training attainment attainment Indicators of and nonformalformal) and training training (non- health care outcomes The Euro Health Spearman Consumer Index Correlation ,232* ,302** / total score N Life expectancy Spearman in absolute value Correlation,560** 0.179,586** ,461**,389** at 65 - females N Life expectancy Spearman in absolute value Correlation,461** 0.091,510** ,501**,319** at 65 - males N Life expectancy Spearman in absolute value Correlation,484** 0.096,533** ,485**,308** at birth - females N Life expectancy Spearman in absolute value Correlation,453** 0.142,502** ,515**,278* at birth - males N * - Correlation is significant at 0.05 level (2-tailed). ** - Correlation is significant at 0.01 level (2-tailed). Coefficients equal or above 0.300** are marked by dark shading. Source: EUROSTAT, calculations by the authors Correlation results could be further strengthened by the opinion of adults themselves. The Adult education survey shows that in % of adults indicated that health or age was the most important obstacle to participation in training. This obstacle was especially relevant for adults in well developed economies. For example, about 17% of adults in the Netherlands, Sweden and Norway, 16% in Belgium and 11% in France stated that health or age is the most important barrier which prevents them from participating in training. Health or age was also more important for low-qualified adults than those who attained a higher level of qualification. Health or age as the most important reason for not participating in education and training was indicated by 8.3% adults with ISCED 0-2 level of education, compared to 5.8% and 2.8% of those with ISCED 3-4 and ISCED 5-6 level qualifications respectively. 198

201 The Adult literacy and life skills survey also confirms the statistical relationship between health and participation in education and training. The respondents who answered the health related questions of the survey in a way that led to their being classified as healthy not only represent the largest percentage of adults but were also the ones with noticeably higher average literacy and numeracy scores compared to the least healthy group of adults those most likely to report experiencing emotional and physical adversities limiting their activities and capping their accomplishments (Learning a living, 2005). Consequently it could be argued that health policies have an impact on the overall level of adult learning. The impact of health policies is likely to be even stronger in the older EU member states and on the lower qualified part of the population. 4. The influence of contextual variables on adult participation in learning and educational attainment of the population This chapter presents the results of a statistical analysis of the impact of general social and economic context indicators on the level of adult learning and education attainment. The influence of four groups - macro-economic, social cohesion, innovation and research and environment and transport of variables is discussed The influence of the macro-economic environment Statistical analysis has revealed that macro-economic environment indicators have the most influence on adult participation in nonformal education and training, and adult tertiary education attainment. In contrast, they have no impact on adult participation in formal education and training and very limited impact on the intensity of non-formal adult education and training and adult upper secondary education attainment (see the Table 19 below). The most influential macro-economic indicator is GDP per capita in Purchasing Power Standards, which has a strong impact on adult participation in non-formal education and training and some impact on the intensity of adult education and their tertiary education attainment. Likewise, the comparative price levels of final consumption by private households and inflation rates have a significant impact on non-formal learning and somewhat more medium impact on tertiary education attainment. Labour productivity also has a positive relationship with adult learning and attainment, although it is more likely to be the effect of adult learning than vice versa. Public balance and general government debt also seem to have significant influence sound management of public finances and more effective use of public resources as opposed to constantly increasing government debt lead to higher investment in education and training and, therefore, higher level of adult learning and better education attainment. FDI intensity, on the other hand, may be a result and not the cause of higher adult tertiary education attainment. These results lead to a conclusion that adults in countries with a higher level of economic development tend to be more engaged in non-formal education and training and complete higher education, 199

202 and, consequently, be more productive and sustain high economic growth. Table 19: The influence of macro-economic environment on adult learning and education attainment Dependent Adult Intensity of Adult Adult variables participation in adult Adult upper participation in participation in Adult tertiary education and education secondary formal non-formal education training (both and training education education and education and attainment, formal and (nonformal), attainment, training, training, Indicators of macroeconomic environment non-formal), Spearman MACRO1¹ Correlation,749**,188**,843**,331** -,212**,431** N Spearman MACRO2¹ Correlation,525** 0,082,696**,182* -,231**,363** N Spearman MACRO3¹ Correlation -,442** -,131* -,636** -0,131-0,033 -,322** N MACRO4¹ Spearman Correlation,539**,227**,434** -0,09 0,087,602** N MACRO5¹ Spearman Correlation -0,099 -,151** -0,053,220** -,486** -,145* N Spearman MACRO6¹ Correlation,770**,236**,818**,186* -,190**,566** N MACRO7¹ Spearman Correlation,220** 0,051,168* 0,084-0,067,304** N ¹ MACRO1 - GDP per capita in Purchasing Power Standards (PPS); MACRO2 - Labour productivity per hour worked - GDP in Purchasing Power Standards; MACRO3 - Inflation rate - Annual average rate of change in Harmonized Indices of Consumer Prices (HICPs); MACRO4 - Public balance - Net borrowing/lending of consolidated general government sector as a percentage of GDP; MACRO5 - General government debt - General government consolidated gross debt as a percentage of GDP; MACRO6 - Comparative price levels - Comparative price levels of final consumption by private households including indirect taxes; MACRO7 - Market Integration - Foreign Direct Investment (FDI) intensity - Average value of inward and outward FDI flows divided by GDP; * - Correlation is significant at 0.05 level (2-tailed). ** - Correlation is significant at 0.01 level (2-tailed). Coefficients equal or above 0.300** are marked by dark shading. Source: EUROSTAT, calculations by the authors 4.2. The influence of social cohesion Social cohesion variables are slightly less influential than macroeconomic ones, but their influence is more widespread. Social cohesion influences not only adult participation in non-formal education and tertiary education attainment, but also, to a larger extent than in the case of macro-economic variables, upper secondary education attainment of adults (see the Table 20 below). Statistical analysis has shown that the persistent risk-of poverty rate and the percentage of early-school leavers have a strong negative influence on adult learning and, especially, on adult upper secondary education attainment. Likewise, the inequality of income distribution, the risk-of-poverty rate after social transfers, longterm unemployment rates and jobless households have a negative impact to some extent on adult learning and education attainment. Expenditure on social protection and social protection benefits, on the contrary, has a positive influence on adult learning. This means that countries with generous and increasingly inclusive social policies tend to have higher levels of adult non-formal learning and higher adult tertiary education attainment. 200

203 Results also provide significant evidence that the number of children in private households is also an influential factor for adult learning and education attainment. Statistical analysis suggests that adults from households with no children tend to be more engaged in non-formal learning and attain tertiary education, in contrast to those who have 1 or more children. Statistical findings extend the previous conclusion: adults in countries with a higher level of economic development and greater social cohesion tend to be more engaged in non-formal education and training and attain upper secondary and tertiary education. Table 20: The influence of social cohesion on adult learning and education attainment Dependent Adult Intensity of Adult Adult variables participation in adult Adult upper participation participation in Adult tertiary education and education secondary in formal non-formal education training (both and training education education and education and attainment, formal and (nonformal), attainment, training, training, Indicators of social non-formal), cohesion Spearman SOCIAL1¹ Correlation -,459** -0,008 -,597** 0,071 -,384** -,193** N Spearman SOCIAL2¹ Correlation,257**,253**,221* -,206* 0,098,369** N Spearman SOCIAL3¹ Correlation -,495** -0,059 -,596** 0,063 -,419** -,227** N Spearman SOCIAL4¹ Correlation -,667** -0,083-0,738 0,738 -,787** -,570** N Spearman SOCIAL5¹ Correlation -,583** -,228** -,651**,226** 0,11 -,425** N Spearman SOCIAL6¹ Correlation -,378** -,258** -,338** -0,013 -,737** -,250** N Spearman SOCIAL7¹ Correlation -,154* -0,083 -,310** -0,09,154* 0,026 N Spearman SOCIAL8¹ Correlation,463**,261*,443** 0,082-0,005,371** N Spearman SOCIAL9¹ Correlation -,505** -0,113 -,547** 0,051 0,038 -,440** N Spearman SOCIAL10¹ Correlation -,332** -,260* -,252* 0,087-0,099 -,401** N Spearman SOCIAL11¹ Correlation,522** 0,071,640**,332** -0,078,269** N Spearman SOCIAL12¹ Correlation,646**,229**,759** 0,187-0,115,425** N Spearman SOCIAL13¹ Correlation,548** 0,079,646**,320** -0,047,291** N Spearman SOCIAL14¹ Correlation,653**,232**,763** 0,183-0,106,431** N ¹ SOCIAL1 - Inequality of income distribution - Income quintile share ratio; SOCIAL2 - % At-risk-of-poverty rate before social transfers; SOCIAL3 - % At-risk-of-poverty rate after social transfers; SOCIAL4 - % At-persistent-risk-of-poverty rate; SOCIAL5 - Long-term unemployment rate - Long-term unemployed (12 months and more) as a percentage of the total active population; SOCIAL6 - Early school leavers - Percentage of the population aged with at most lower secondary education and not in further education or training; SOCIAL7 - Jobless households - children - Share of persons aged 0-17 who are living in households where noone works; SOCIAL8 - Percentage of private households with 0 children; SOCIAL9 - Percentage of private households with 1 child; SOCIAL10 - Percentage of private households with 2 children; 201

204 SOCIAL11 - Expenditure on social protection - % of GDP; SOCIAL12 - Expenditure on social protection - PPS per inhabitant; SOCIAL13 - Social protection benefits - % of GDP; SOCIAL14 - Social protection benefits - PPS per inhabitant. * - Correlation is significant at 0.05 level (2-tailed). ** - Correlation is significant at 0.01 level (2-tailed). Coefficients equal or above 0.300** are marked by dark shading. Source: EUROSTAT, calculations by the authors Dependent variables Indicators of innovation and research Spearman INNO1¹ Correlation INNO2¹ INNO3¹ INNO4¹ INNO5¹ INNO6¹ INNO7¹ INNO8¹ 4.3. The influence of innovation and research The link between innovation and research variables and adult learning and education attainment is much stronger than that of the social cohesion and even macro-economic environment, and it is also very widespread.. In contrast to other context variables, the context of innovation and research is increasingly interrelated with not only non-formal, but also formal adult learning (see the Table 21 below). Statistical analysis indicates that adult learning has the strongest relationship with gross domestic expenditure on R&D, ICT development index, level of Internet access by households, total R&D personnel and researchers, employment in high and medium technology and knowledge-intensive sectors, and a summary innovation index. Most likely, innovation and research and adult learning as well as tertiary education attainment are mutually dependent, since they are all features of the learning society. For example, higher participation of adults in formal education and training increasingly points to the level of innovation and research in a country and the latter, consequently, further increases the demand for learning. The following conclusion can be drawn: countries with higher spending on innovation and research tend to have higher level of adult participation in both formal and non-formal education and training, as well as higher tertiary education attainment (and vice versa). Table 21: The influence of innovation and research on adult learning and education attainment Adult participation in education and training (both formal and nonformal), Adult participation in formal education and training, Adult participation in non-formal education and training, Intensity of adult education and training (nonformal), Adult upper secondary education attainment, Adult tertiary education attainment, ,764**,203**,804** 0,141 0,047,451** N Spearman Correlation,777**,708**,867** -0,047 0,056,636** N Spearman Correlation,824**,616**,799** -0,057 0,022,610** N Spearman Correlation,640**,412**,674** -,238*,359**,478** N Spearman Correlation,493**,492**,529** 0,033 0,134,482** N Spearman Correlation,333** 0,08,487**,237* -,221**,153* N Spearman Correlation,526* 0,124,557** 0,019 0,219 0,236 N Spearman Correlation,768**,299**,808**,209* 0,09,605** 202

205 Indicators of innovation and research Dependent variables Adult participation in education and training (both formal and nonformal), Adult participation in formal education and training, Adult participation in non-formal education and training, Intensity of adult education and training (nonformal), Adult upper secondary education attainment, Adult tertiary education attainment, N Spearman,775**,196**,822** -0,143,121*,543** INNO9¹ Correlation N Spearman,726**,429**,729** -0,162 0,069,599** INNO10¹ Correlation N ¹INNO1 - Gross domestic expenditure on R&D (GERD) - Percentage of GDP; INNO2 - ICT development index (IDI); INNO3 - Level of Internet access - households - Percentage of households who have Internet access at home; INNO4 - ICT expenditure by type of product; Information Technology Expenditure - Percentage of GDP; INNO5 - Broadband penetration rate - Number of broadband access lines per 100 inhabitants; INNO6 - High-tech exports - Exports of high technology products as a share of total exports; INNO7 - Innovation expenditure as % of turnover for all enterprises; INNO8 - Total R&D personnel and Researchers as % of labour force and total employment; INNO9 - Annual data on employment in high and medium technology and knowledge-intensive sectors; INNO10 - Summary innovation index (SII). * - Correlation is significant at 0.05 level (2-tailed). ** - Correlation is significant at 0.01 level (2-tailed). Coefficients equal or above 0.300** are marked by dark shading. Source: EUROSTAT, calculations by the authors 4.4. The influence of environment and transport Finally, statistical analysis of environment and transport variables and adult education has revealed some statistically significant relationships, though some of them might be accidental and difficult to explain. In general, these variables indicate the sustainability level of a country and provide some insights into the external conditions of adult learning (see Table 22). Energy intensity and resource productivity have the strongest relationship with adult learning. Countries that, for example, are highly economically productive, with mild, temperate weather, demographic patterns of work places close to home, and use fuel efficient vehicles, support carpools, mass transportation or walk or ride bicycles, have a considerably lower energy intensity and higher level of non-formal adult learning than countries that are economically unproductive, with extreme weather conditions requiring heating and cooling, long commutes, and extensive use of generally poor fuel economy vehicles. Likewise, countries with high resource productivity, which is a key concept used in sustainability measurement, have higher adult participation in nonformal education and training, and to some extent higher adult tertiary education attainment than countries with low resource productivity. Resource productivity is closely linked to the use of innovation, technological development, labour productivity and competitiveness of the country (Hargroves and Smith 2005), therefore, adult education and training is more likely to be the cause for resource productivity, than its effect. The car share of inland passenger transport has a significant positive impact on adult learning and adult tertiary education attainment. This indicator includes transport by road, rail and inland waterways, and might simply mean greater mobility of a country s population. To conclude, adults in countries with a higher level of sustainability and higher mobility tend to be more engaged in non-formal education and training and attain tertiary education. 203

206 Dependent variables Indicators of environment and transport Spearman ENV1¹ Correlation Table 22: The influence of environment and transport on adult learning and education attainment Adult participation in education and training (both formal and nonformal), Adult participation in formal education and training, Adult participation in non-formal education and training, Intensity of adult education and training (nonformal), Adult upper secondary education attainment, Adult tertiary education attainment, ,484** -,125* -,597** -,341**,282** -,211** N Spearman,573** 0,01,743** 0,17 -,224**,342** ENV2¹ Correlation N Spearman,559**,207**,596** 0,046-0,036,472** Correlation ENV3¹ N N ¹ENV1 - Energy intensity of the economy - Gross inland consumption of energy divided by GDP (at constant prices, 1995=100) - kgoe (kilogram of oil equivalent) per 1000 euro; ENV2 - Resource productivity - Ratio of gross domestic product to domestic material consumption; ENV3 - Car share of inland passenger transport; * - Correlation is significant at 0.05 level (2-tailed). ** - Correlation is significant at 0.01 level (2-tailed). Coefficients equal or above 0.300** are marked by dark shading. Source: EUROSTAT, calculations by the authors To summarise, there seems to be a two-way relationship between adult learning and educational attainment, and contextual variables. The most significant and widespread links are between adult learning and research and innovation, especially ICT development. On one hand, a high level of economic development, great social cohesion, sustainability and mobility form a favourable context for adult learning and education attainment. On the other hand, increasing economic prosperity, social cohesion, sustainability and mobility constantly demand knowledge, innovation and research which could be only developed with higher adult learning and education attainment. 5. The composition and relative significance of factors explaining adult participation in learning and educational attainment of the population Chapters 3 and 4 of Annex 1 provided an overview of how education, labour market, health policies as well as contextual variables directly or indirectly (through intervening variables, i.e. variables characterising balance of labour supply and demand and labour market mobility) exerted individual influences on adult learning. The current chapter provides the results of multiple regression modelling showing, which independent variables have the best explanatory power for our dependent variables and what is their relative significance. The modelling was carried out using un-transformed variables. R square provided in tables below provides a measure of how well future outcomes are likely to be predicted by the model. R square values vary from 0 to 1. The beta coefficient provided in tables below ( B standardized ) is the ratio of the relative predictive power of the independent variables. All independent variables that might be expected to have an influence on the dependent variables were identified during the literature review and correlation analysis. We have tested several models for each of our dependent variables. 204

207 Model Education policies The first set of models includes separate education or labour market policy characteristics. We have excluded health policies from the first set of models because they were measured by a single aggregate variable. The second set adds intervening labour market variables (i.e. variables characterising balance of labour supply and demand and labour market mobility) to the main labour market policy characteristics. The third set of models combines all the most important policy characteristics (education, labour market and health) and intervening labour market variables. The fourth set only adds the most important context variables Models explaining adult participation in education and training (both formal and non-formal) Two models (model 4 and model 6) described in table 23 exceed our R Square medium relevance benchmark of 0.5. These models have the best overall predictive power although they demonstrate different results in terms of the best composition of independent variables predicting adult participation in education and training. The model which initially included all the relevant policies and intervening variables (model 6) is better than the other one with a much shorter list of initially entered variables (model 4). The better model demonstrates that adult participation in learning is higher where the employment rate is higher (this variable has the highest predictive power). The less predictive model shows that participation in adult learning is higher where the rigidity of hours is lower and labour market expenditure on training is higher. All in all independent variables characterising labour market policies seem to be more important in explaining overall participation in adult learning than those characterising education policies. The overall results of modelling provided in table 23 point to the importance of the employment rate in explaining adult participation in learning. Table 23: Models explaining adult participation in education and training (both formal and non-formal) Labour market Labour market Labour Labour market policies and policies and market policies and intervening without intervening without policies intervening mobility mobility All policies and intervening R Square FINANCE Rigidity of hours index LMP expenditure on Training Employment rate (%) FINANCE4 - Annual expenditure on public and private educational institutions per pupil/student in EUR PPS, for all levels of education combined, based on full-time equivalents. Source: EUROSTAT, calculations by the authors 5.2. Models explaining adult participation in formal education and training Models 6-8 designed to explain adult participation in formal learning (see table 24) fall just short of our R Square medium relevance benchmark of 0.5. the almost identical result is mainly caused by one variable which dominates in all three models the employment rate of older workers. The results of this modelling suggest that adult participation in adult learning is mainly facilitated through better employment opportunities for older workers, i.e. one of the adult groups which participates least in 205

208 Model Education policies formal adult education and training. Furthermore, other indicators in models 3, 5, 8 shows that adult participation in formal learning is further facilitated either by the availability of part-time working opportunities or by a lower rigidity of working hours, enabling e.g. a combination of work and part time or evening studies. This also shows that it is easier to combine employment and professionally related studies, which are often in the interest of employer. Table 24: Models explaining adult participation in formal education and training Labour Labour Labour Labour All policies market market All market market and policies and policies and variables policies policies intervening intervening intervening All variables R Square STRAT Part-time employment as a percentage of the total employment Rigidity of hours index Employment rate of older workers (%) STRAT2 - First age of selection in the education system. Source: EUROSTAT, calculations by the authors Model Education policies 5.3. Models explaining adult participation in non-formal education and training Models 4 and 5 are the best at explaining adult participation in non-formal learning (see table 25). In both models the single best variable explaining higher adult participation in non-formal learning is higher resource productivity. Model 5 combining resource productivity and employment rate indicators however demonstrates slightly better predictive power. However, resource productivity is closely linked with the use of innovation, technological development, labour productivity and competitiveness of the country and, therefore, adult education and training is more likely to be the cause for resource productivity, than its effect. Models, which are just below the 0.5 benchmark, also include annual expenditure on public and private educational institutions per pupil/student (the only relevant variable in model 1), labour market expenditure on labour market services (less relevant variable in model 3) and employment mobility index (the only relevant variable in models 6 and 8) variables. Although this points to certain importance of education/labour market services, financing and labour market flexibility for adult participation in nonformal learning, the overall result is hardly satisfactory and there is a need to search for other sets of entry variables or other explanatory models (perhaps non-linear). Table 25: Models explaining adult participation in nonformal education and training Labour Labour Labour market market Labour Labour market All policies policies and policies and market market policies and intervening intervening policies policies and intervening without without intervening mobility mobility All variables R Square FINANCE Employment rate (%)

209 Model Labour Labour Labour market market Labour Labour market All policies policies and policies and market market policies and intervening intervening policies policies and intervening without without intervening mobility mobility Education policies All variables LMP expenditure on Labour market services Employment mobility index Resource productivity - Ratio of gross domestic product to domestic material consumption FINANCE4 - Annual expenditure on public and private educational institutions per pupil/student in EUR PPS, for all levels of education combined, based on full-time equivalents. Source: EUROSTAT, calculations by the authors Model 5.4. Models explaining intensity of non-formal adult education and training All models explaining the intensity of adult non-formal education and training provided in table 26 are less than satisfactory they all fall short of the R Square low relevance benchmark of 0.3. The relatively poor result suggests the need to search for other sets of entry variables or models (perhaps non-linear) to explain the intensity of non-formal adult education and training. Table 26: Models explaining intensity of adult education and training (non-formal) Labour Labour Labour market All policies market market policies and and policies policies intervening intervening Education policies All variables R Square COURPURP Rigidity of hours index Population activity and inactivity Annual averages Activity rate (15 to 64 years) Expenditure on social protection - % of GDP Population in employment having a second job COURPURP Purpose of the most recent taught learning activity, % of answer Mostly job related (professional). Source: EUROSTAT, calculations by the authors 5.5. Models explaining adult upper secondary education attainment The effort aiming to explain adult upper secondary education attainment has resulted in two models, which exceed our R Square high (0.7) or medium (0.5) relevance benchmark. Both models were performed with all context variables and as a result propose that leaving school early is the single most important factor explaining adult upper secondary education attainment. The better model 9 also reveals the importance of adult learning being job related the more training is associated with the work place the higher is educational attainment. The less predictive model 7 includes another context variable general government debt as a share of GDP, which, as could be expected, is inversely correlated with our dependent variable. Other policy characteristics seem to 207

210 Model Education institutions have no significant relevance in explaining adult upper secondary education attainment. The reason for this might be that education attainment at this level depends on policy characteristics, which have not been included in our models. Table 27: Models explaining adult upper secondary education attainment Labour Labour All policies market market and instituti institutio intervenin ons ns g Labour market policies and intervenin g without mobility Labour market policies and intervenin g without mobility Labour market policies and intervenin g Labour market policies and intervenin g All policies and intervenin g R Square insignifican t Indicator Indicator Indicator Indicator Indicator Indicator 1: COURPURP - Purpose of the most recent taught learning activity, % of answer "Mostly job related (professional)"; Indicator 2: Inequality of income distribution - Income quintile share ratio; Indicator 3: Income support levels for unemployed - social protection benefits unemployment function - Purchasing Power Standard per inhabitant; Indicator 4: Early school leavers - Percentage of the population aged with at most lower secondary education and not in further education or training; Indicator 5: General government debt general government consolidated gross debt as a percentage of GDP. Source: EUROSTAT, calculations by the authors Models explaining adult tertiary education attainment Three models described in table 28 exceed our R Square high relevance benchmark of 0.7. Although the models have high predictive power, they all single out different variables which seem to be the most relevant for adult tertiary education attainment. Model 1 demonstrates that the number of teachers aged 50 and over teaching in public and private schools at ISCED levels 2-3 is the only significant factor explaining high adult tertiary attainment. Model 10 singles out the three most important variables: income support levels for the unemployed, job-to-job mobility and, to a lesser extent, the employment rate. Meanwhile model 12, which includes all variables characterising both policies and context variables, shows that the most significant factors for high adult tertiary education attainment is the ability to reconcile work and non-formal learning and absence of children in a family. Overall the results of these models reveal that none of the policies or context factors have greater explanatory power with regard to adult tertiary education attainment. However, number of teachers aged over 50 could be regarded as most influential educational factor explaining adult tertiary education attainment. Table 28: Models explaining adult tertiary education attainment LMP² LMP² LMP² LMP² and and and and LMP² LMP² LMP² All All Model Education interveninninninning interve- interve- interve- interve- interve- interve- and and and LMP² LMP² variablebles varia- policies without without without without ning ning ning mobility mobility mobility mobility R Square Indicator 1¹ Indicator 2¹ Indicator 3¹ Indicator

211 LMP² LMP² LMP² LMP² and and and and LMP² LMP² LMP² All All Model Education interveninninninning interve- interve- interve- interve- interve- interve- and and and LMP² LMP² variablebles varia- policies without without without without ning ning ning mobility mobility mobility mobility 4¹ Indicator 5¹ Indicator 6¹ Indicator 7¹ Indicator 8¹ ¹Indicator 1: Teachers aged >50 teaching in public and private schools at ISCED levels as % of total teachers teaching in ISCED levels 2-3; Indicator 2: Employment rate (%); Indicator 3: Income support levels for unemployed - social protection benefits unemployment function - Purchasing Power Standard per inhabitant; Indicator 4: Job-to-job mobility index; Indicator 5: Public balance - Net borrowing/lending of consolidated general government sector as a percentage of GDP; Indicator 6: Resource productivity - Ratio of gross domestic product to domestic material consumption; Indicator 7: COURWORH - Did the most recent taught learning activity (non-formal) take place during paid working hour? % of answers "Only during paid working hours" and "Mostly during paid working hours"; Indicator 8: Percentage of private households with 0 children. ² LMP labour market policies Source: EUROSTAT, calculations by the authors Regression results show that models combining all the most important policy characteristics (education, labour market and health), intervening labour market variables and context variables best explain adult participation in education and training (both formal and non-formal) and educational attainment. Models comprised of labour market policies and intervening labour market variables are second best in explaining the above-mentioned dependent variables. Furthermore, these models have the best predictive power in explaining adult participation in non-formal education and training. In contrast, models involving only education policies seem to be unsatisfactory in explaining our dependent variables. The primary reason for this could be very large differences in education policies among the countries concerned (the smaller is R square, the larger are the differences between countries). To illustrate, figure 3 shows that in different countries relative education expenditure has a very different impact on adult participation in education and training (both formal and non-formal). For example, in Cyprus and the Netherlands there is a strong positive relationship between relative education expenditure and the overall level of adult learning, while in Greece and Sweden this relationship is negative. Although this financing variable seemed significant for adult learning (there is medium correlation between these two variables), regression results reveal that it has low explanatory power. This reveals the fact that the relationship between relative education expenditure and overall level of adult learning is influenced by other factors which are specific for one or a number of countries. 209

212 Figure 3: Illustration of differences between countries in terms of the relationship between relative education expenditure and adult participation in education and training (both formal and non-formal) Source: UNESCO Institute of Statistics/OECD/Eurostat and EUROSTAT-Labour Force Survey (data for variables EDUC4WN calculated by the authors), authors calculations However models involving only education policies are good in explaining adult tertiary education attainment. To illustrate, figure 4 reveals that in most countries a number of teachers aged 50 and over has significant influence on adult tertiary education attainment. Almost all the countries are positioned nearby the overall average tendency of these two variables. 210

213 Figure 4: Illustration of similarities between countries in terms of the relationship between the number of teachers aged over 50 and upper secondary educational attainment of adults Source: UNESCO Institute of Statistics/OECD/Eurostat and EUROSTAT-Labour Force Survey (data for variables EDUC4WN calculated by the authors), authors calculations 6. Conclusions As a vital element of a competitive economy providing more and better jobs the participation of the adult population in lifelong learning and improvement of adult qualification levels has been particularly high on public policy agenda especially since the adoption of the Lisbon strategy both at European and national levels. EU Member States, EEA-EFTA and candidate countries have launched a number of reform efforts and new policy initiatives in adult learning, the broader education system and outside of it to boost the investment in human capital and improve its efficiency. The main question for this report has been the effectiveness of these efforts. The literature review helped us to identify the relevant policy characteristics (as well as indicators for their measurement), which have been implemented over the past decade. The correlation analysis and multiple regression modelling enabled us to relate the changes in policy characteristics to adult learning and participation. In this chapter we summarise the results of quantitative analysis. Analysis shows that participation in adult learning and education attainment are closely interrelated. Quantitative analysis confirms 211

214 that non-formal education and training is the most popular form of learning, which strongly determines the overall volume of adult learning. The larger is the share of adults participating in nonformal education and training, the higher the country ranks in terms of the overall volume of adult learning. Furthermore, analysis suggests that greater participation of adults in formal learning leads to higher adult tertiary education attainment and, as a result, higher education attainment leads to higher participation of adults in non-formal education and training. Results partly confirm our initial hypothesis that the policy characteristics of an education system influence adult learning and education attainment. Analysis revealed that the most influential policy characteristics of the education system are stratification (the number of school types or distinct educational programmes and the first age of selection in the education system), vocational specificity (private sector involvement in the provision of initial VET and the prevalence of job related non-formal learning), flexibility in provision of learning services (inflow of students and proper opportunities for those aged to participate in formal education and training), school autonomy (esp. in budget allocations within the school) and spending on education (absolute and relative per pupil/student). Statistical analysis has provided support to most of our hypotheses about the influence of labour market policy characteristics on adult learning and educational attainment. The flexicurity policy mix or its individual elements (lower rigidity of hours, more liberal parttime employment, higher unemployment benefits and expenditure on active labour market policies) seem to have correlated with our dependent variables directly or through intervening variables in most of the EU, EFTA-EEA and candidate countries most of the time between 1997 and Our hypothesis of the mutual interrelationship between health care and adult learning was also confirmed. Statistical analysis revealed that life expectancy, which is the main outcome of health care systems, has notable implications on adult learning (esp. nonformal) and education attainment. The implications of health policies on adult learning are likely to be even stronger in older EU member states with well-developed economies and in low-skilled populations. However the cause and effect relationship here remains unclear: either adult learning has positive effects on health and life expectancy or better health leads to more (and longer) learning (esp. non-formal). Contextual variables (i.e. factors, other than education, labour market and health policies or their immediate outcomes), have significant influences on adult learning and education attainment. In this report we have considered the following four sets of contextual variables: macro-economic environment, social cohesion, innovation and research, and environment and transport. The relationships between adult learning and educational attainment, and many contextual variables could be two-way. The most significant and widespread link is between adult learning and a set of research and innovation variables, especially ICT development. The high level of economic development, social cohesion, and efficient use of resources form a favourable context for adult learning and education attainment. On the other hand, increasing economic efficiency and social cohesion is hardly possible without improving accessibility to learning, learning 212

215 motivation and raising the educational attainment of the population. We have attempted to measure the relative significance of variables explaining adult learning and educational attainment with the help of multiple regression modelling. Adult participation in education and training (both formal and non-formal) is best explained by models involving labour market policies (incl. intervening variables balance of labour supply and demand and labour market mobility) and even more so by models involving a large set of both education, labour market and health policies and context variables. Education policies seem to only have high predictive power in explaining adult tertiary education attainment. These outcomes of multiple regression modelling seem to contradict some of the results of the literature review and correlation analysis. The primary reason for this could be large differences between countries in education and other policies. These differences significantly hinder the quantitative search for common denominators influencing adult learning in the majority of the countries. Furthermore, the limited number of variables and their short time series complicate this search even more. Therefore there is large room for improvement in the quantitative analysis of factors which have a significant impact on adult learning participation and educational attainment. 213

216 Public Policy and Management Institute Gedimino ave. 50 LT Vilnius, Lithuania Tel Fax info@vpvi.lt