EUROSTAT /
Doc. Eurostat/F/11/DSS/01/3.3EN
Item 3.3The future of life long learning statistics (including CVTS)
MEETING OF THE EUROPEAN DIRECTORS OF SOCIAL STATISTICS
luxembourg, 21 – 22 September 2011
BECHbuilding, room QUETELET
At the occasion of their September 2009 meeting[1], the Directors of Social Statistics (DSS) approved Eurostat's general approach for enhancing the coherence between the three EU statistical surveys on lifelong learning (Adult Education Survey - AES, Continuing Vocational Training Survey - CVTS and Labour Force Survey - LFS).
This strategy was reflected in the adoption of the AES and CVTS Commission Regulations for the 2011 data collections[2] as well as the launch of tests on the improvement of LFS variables on participation in education and training running until March 2012[3].
Three recent initiatives emphasise the need to further streamline lifelong learning statistics:
- In May 2011, it was broadly agreed by the European Statistical System Committee to repeal/modifythe CVTS legal acts[4]; this would imply restricting the CVTS data collection from 2015;
- The modernisation of social statistics[5] suggests an integration of the AES variables in pillar sub-sample(s) in the future;
- The implementation of ISCED 2011 is foreseen to happen at the same timein all EU sources in 2014(see document for item 5.4 on the implementation of this classification); for sample surveys, this will cover participation in formal education in addition to educational attainment.
The aim of this document is to present the impact of these recent developments on the 2009 strategy on lifelong learning statistics and more precisely to;
- Present the scope and planning for the review of the 2015 CVTS and its consecutive waves[6] (section 1).
- Review the scope of changes to be discussed in main household survey variables on participation in education and training in 2012 (section 2);
This corresponds to the tasks foreseen withinthe 2012 Eurostat work programme on lifelong learning statistics and has to be considered in the light of the conclusions of the Council of November 2010 on Vocational and Education Training systems (see section 1).
The Directors of Social Statistics are invited to:
-Comment on the future of CVTS and of educational variables in household surveys (i.e. module approach to prepare the integration of the educational variables into a pillar approach for social surveys).
-Agree on setting-up a task force in charge of proposing a restricted CVT data collection on enterprises and amendments to the legal instruments.
-Take note of the need for an annual indicator on participation of adults in lifelong learning using a reference period of 12 months: the implementation issues and timetable will be discussed in the DSS meeting of September 2012.
1. State of play as regards the future of the CVTS
In September 2009, it was agreedto enhance the coherence between the three EU statistical sources on lifelong learning so that AES would provide detailed information on the participation of individuals in education and training activities every 5 years while LFS would provide annual evolutions for a limited set of indicators and CVTS would complete the AES results each 5 years with a focus on enterprise strategies for employees’ skill developments.
Further integrationof CVTS with existing surveys is however required as suggested inthe recent Eurostat's proposal to the ESS Committee to repeal/modify the CVT legal acts.
The principles and planning of this task related to the 5th CVTS data collection are proposed in sections 1.1 to 1.3 below. Section 1.4 presents the relationship of this initiative with the views of the main stakeholders and the recommendations from the recent rolling review on education statistics[7].
1.1.Revision of the CVTS legal acts
The Continuing Vocational Training Survey (CVTS)is an enterprise survey conducted each 5years. The 2010 data will be delivered in 2012 according to the Regulation No 822/2010. However, its future is questioned within the European Statistical System as for the 2015 and consecutive waves.
More precisely, Eurostat proposes to amendthe framework Regulation 452 / 2008 of 23 April 2008 on education and lifelong learning statistics tocover a restricted data collection on CVT in enterprises and a repeal of current CVTS legal acts.
The review is closely linked to output requirements resulting from the recent outcomes of the Council of November 2010 whose conclusions underline the need for data on Vocational Education and Training (VET) systems in the context of the Copenhagen process and of the important contribution it has to make to the Europe 2020 strategy[8].In particular, the Bruges communiqué of 7 December 2010 on enhanced European Cooperation in Vocational Education and Training for the period 2011-2020[9]states that “EU level policy-making in VET should be based on existing comparable data. To this end, and using the Lifelong Learning programme, Member States should collect relevant and reliable data on VET – including VET mobility – and make these available for Eurostat. MemberStates and the Commission should jointly agree on which data should be made available first”.
1.2.Proposed approach
In the light of Eurostat’s communication on the production method of EU statistics[10] and the discussions on the modernisation of social statistics, it is proposed to review the potential of current EU data sources providing data onVET activities as regards:
- volume and main characteristics of VET activities,
- Individual and enterprise costs and strategies (incl. outcomes and quality).
This requires reviewingthe output provided by CVTS and more generally the outputsneeded as regardsenterprise-provided learning activities in the context of the recent VET policy developments. This will allow considering statistics which:
(a)could be made available from other household surveys (e.g. AES or LFS for volume and characteristics as well as individual costs and strategies) based on the improvements proposed in section 2 below;
(b)could be made available from other enterprise surveyswith slight adjustments and/or quality improvements within the European Statistical System[11](e.g. LCS for enterprise costs, incl. receipts and contributions).
These two steps should obviously provide a consistent framework (e.g. similar scope, frequencies and quality) and take account of variables required in each statistical source to allow consistent breakdowns of outputs.
Other variables- in particular on VET enterprise strategies - could be considered as a module to be attached to other existing data collections carried out on enterpriseswithin the European Statistical System[12].
For the sake of exhaustiveness of the approach, two main statistical components will have to be considered: statistics on adult learning and statistics on initial education of young people. Alternatives to be investigated in that context are summarised in table A below.
More details are presented in annex 1 on adult learning (i.e. on enterprise-provided education and training) and in annex 2 on initial education of young people (i.e. on formal vocational education and apprenticeship).
Table A – Potential alternative sources to CVTS
CVTS / LFS / AES / OtherADULT LEARNING
Volume (participants, hours) / / (1) / (1)
Characteristics (e.g. forms, level) / (1) / /
Enterprisecosts / / (2)
Enterprisestrategies
(incl. difficulties, outcomes) / / (3) / (3) / (3)
INITIAL EDUCATION
Volume (participants) / / (4) / (4)
Enterprise costs / / (2) (5)
Notes to table A:
(1)Breakdowns by individual characteristics only in household surveys;Characteristics of learning partially available in CVTS;
(2)Labour Cost Survey (potential to be further investigated ; guidelines and quality of data to be improved accordingly),
(3)Background data available through household surveys (e.g. through employees in disadvantaged groups); Alternatives in data collections performed within the European Statistical System or by European agencies to be confirmed.
(4)Based on an improvement of the current LFS education variables and/or of the AES (see section 2);
(5)Data available in CVTS 3 (low comparability) and CVTS 4 (optional variables);Data on costs of vocational formal education in education systems is available in the UOE data collection;
It should be outlined that the review of output and potential sources will need to make references to surveys carried out by European Agencies with experience in similar issues such as the European Foundation for the Improvement of Living and Working Conditions (Eurofound) or the European Centre for the Development of Vocational Training (CEDEFOP). This is of particular importance in the context of Europe 2020 flagship ‘an agenda for new skills and jobs’.
1.3 Planning for the revision of the CVTS
Within this context the following planning is proposed for the discussions of proposals and revision of the CVTS legal acts. An important step for 2011 is the discussion of the overall approach by the Directors of Social Statistics in September 2011. This covers the approach for both CVTS and main educational variables in household surveys(as described in section2 for the latter).
Table B – Planning for the revision of the CVTS
1 / Nov 2010 – May 2011(completed) / Rolling review of education and lifelong learning statistics, preliminary consultation of Commission services on lifelong learning statistics (see section1.4)
2 / May 2011
(completed) / First discussion at the Education and Training Statistics WG meeting on the approach and planning for the revision of the CVTS
3 / September 2011 / Decision by the Directors of Social Statistics on the approach and planning for the revision of the CVTS (incl. creation of a Task Force –see section1.5)
4 / Dec 2011 – June 2012 / Preparation of revised legislation with the support of theTask Force and consultations of the Education and Training Statistics WG, taking into account the results from CVTS4 and key policy needs in continuing vocational training.
5 / Before the end of 2012 / Discussion within the DSS, BSDG[13] and the ESSC on the CVTS and lifelong learning revised legislations
1.4. Consultation of users and data producers
The consultation of users already started among European Commission services which are already of the opinion that:
- Complementarities of CVTS and AES/LFS should be further clarified; Significant improvements are required as regards quality and coherence of results on participation rates; The production of annual, accurate and comparable ratespreferably using a 12-month reference period is considered as a priority[14]; The availability of breakdowns focussing on VET[15] are of particular importance.
- An appropriate balance between statistical needs, burden on respondents, relevance and quality of the information should be foundfor information related toenterprise costs and strategies on VET;
- Future of CVTS should be assessed on the basis of a detailed proposal by Eurostat;
The latter supposes listing main statistical indicators required for VET policies and shedding light on potentialalternative sources for each of them (incl. CVTS);This will cover an assessment of differences on definitions, scope (coverage), frequencies and data transmission delays as well asonaccuracy and comparability; Level and precision of breakdowns inall sources will have to be displayed accordingly, in particular:
- Availability of outputs by participants and hours of training (e.g. for costs and strategies).
- Availability of information by size and sector (e.g. enterprise and household surveys generally use slightly different concepts - ‘enterprises’ versus ‘local units’).
Theproposal should allow selecting potential sourcesfor discussion by the Task Force and the relevant Working Groups from December 2011 until June 2012, in the light of the outcomes of the recent rolling review on education and lifelong learning statistics[16] (partner and user surveys - see section 1.5)as well as Member States’ proposals for negative priorities in 2013.
The discussion on data sources on continuing vocational training should also take into account the benefit a survey dedicated to the theme (such as the CVTS) is likely to have with regards to flexibility (to respond to changing policy needs in the future) and coherence/harmonised data collection.
It should be outlined that requirements for statistics on skill assessments and skill needs are discussed in parallel to allow considering the wider scope of the Europe 2020 flagship initiative entitled 'an agenda for new skills and jobs'. This means taking account of the future of PIAAC[17]as well as the CEDEFOP pilot survey on enterprise skill needs and other related data collections at EU Level[18].
The sustainability of a module attached to an existing enterprise survey and dealing with VET enterprise strategies, compared to a stand-alone survey, will therefore need to be discussed in the light of policy requests for information on enterprise strategies on skill assessment and skill needs.
Other variables - in particular on - could be considered as a module to be attached to other existing data collections carried out on enterprises within the European Statistical System
1.5. Proposed mandate of the task force
As indicated in step 4 of the planning above, it is proposed that the DSS sets a task force up to support the revision of CVTS. The task force would:
a)Provide recommendations on the provision of indicators on CVT in a coherent framework aiming to reduce further the overlapping of sources and burden on enterprises;
b)Advise on the related improvement of sources and define a restricted data collection on CVT (module) focussing on enterprises' strategies(and on enterprises’ costs if LCS is not a suitable alternative);
c)Propose the technical elements for revised legal acts (including setting up minimum quality requirements for outputs);
d)Assess the sustainability of the proposals for lifelong learning and VET statistics, based in particular on the first outcomes from the recent CVTS 4.
The task force would meet three times in December 2011, March 2012 and June 2012. It would be composed of representatives from about 10MemberStates as well as main users. A call for volunteers would be launched after the DSS have agreed upon setting-up such a task force in their September meeting.
The task force’s proposal would in particular take account of the following aspectsalready highlighted in the report to the Council and European Parliament on the implementation of Regulation (EC) No 1552/2005 and in the rolling review on education statistics:
(1)Burden on enterprises: the length and complexity of the questionnaire have meantlow response rates from enterprises in some countries,as a number oflocal departments need to be contacted in order to cover all the topics of the questionnaire. The use of alternative sources should be considered for certain variables.
(2)Reliability of the volume related data collected: many enterprises do not have precise records of the content and characteristics of the training provided in a given year. Household surveys such as the AES could provide similar data.
(3)Coverage of the survey:as the cost and burden of enlarging the CVTS to include small enterprises and public sectors is regarded as far too large by participating countries, the potential of alternative sources covering all or more enterprises and sectors will need to be analysed.
(4)Coherence of sources: further harmonisation of CVTS concepts with AES, LFS and business surveys would be needed to streamline lifelong learning statistics and set-up a coherent framework.
(5)Timeliness: the transmission of complete CVTS micro-data delays the publication of EU results because of the large number of checksthat have to be performed before dissemination. The flexibility and stability of the IT tool as well as the quality of the CVTS technical documentation will need to be further assessed. The review should include the potential for an improvement of the CVTS dissemination timetable (especially for detailed tables and anonymised micro-data).
2.Approach and scope of changes for main educational variables in household surveys
The previous section on the future of CVTS outlined the need for further streamlining educational variables in household surveys for a coherent framework on lifelong learning statistics.All changes foreseen in household survey variables will obviously have to be discussed within the modernisation of social statistics at Working group levels and DSS level.
This likely means considering all adaptation of education variables in a flexible and modular approach to allow an integration into a pillar subsample(s)when required. Section 2.1 below addresses the issue of the overall approach for the revision of education variables in household surveys.
Section 2.2 addresses more specific issues on the two sets of educational variableswhich will need to be discussed already in 2012:
- the scope of the revision for participation in education and training:harmonisation of concepts and further integration of educational variables in household surveys (current AES and LFS) to provide more accurate and comparable data on 'lifelong learning' and a coherent framework with the data collection on CVT in enterprises as proposed in section 1;
-the implementation of ISCED 2011 for educational attainment data as described in item 5.4 of the agenda.
2.1 Overall approach for revision of education variables in household surveys
The LFS and the AES are the two major surveys for education results.If AES variables shall be integrated into pillar sub-sample(s), certain aspects of the current LFS education variables and of the AES will need to be discussed such as:
- The distribution of educational variables in modules[19];
- The identification of the few core educational variables required for any module,
- The frequency of information for core and module educational variables,
- The sample size (sub-sample(s) of the pillar), age groups (e.g. coverage of 18-24) as well as the selection process of learning activities (for detailed information[20]).
Table C below distributes educational variables in three modules and provides a framework suitable to follow implementation progress over the time:
- Current status: LFS data correspond to annual variables in modules 1 and 2;
AES data corresponds to the set of the 3modules each 5 years.
- Medium term:Modules are defined and their implementation is discussed according to progress on the modernisation of social statistics;
- Long term:Modules are implemented in a standard manner in all Member States.
The main features to be discussed in 2012 are highlighted in table C below: use of a reference period of12 months in an annual survey, periodicity of variables under the categorytri-annualand characteristic "enterprise provided" for all main results on participation in education and training.
Table C – Education variables in household surveys: main variables and frequencies
ANNUAL / Levelorientation / Participation in last 4weeks & 12 months / Participation in last12 months
Characteristics I:
level, orientation or form
enterprise provided / Characteristics I:
level, orientation or form
enterprise provided
TRI-ANNUAL / Programmes not completed (drop outs)
Validation of competences / Characteristics II
field, purpose, volume, teaching method, certification / Difficulties for participation in education and training
Field, Apprentices / Characteristics IIIsatisfaction, outcome, quality / Factors influencing participation[21]
ICT, languages, cultural participation
Learning mobility (1) / Informal learning
in last 12 months
Note: (1): see item 3.2 on new policy needs in the field of learning mobility.