Point 5 of the agenda

Transcription

1 EUROPEAN COMMISSION EUROSTAT Directorate F: Social Statistics and Information Society Unit F-4: Education, culture and science statistics Doc. ESTAT/F4/2006-ETS-27-EN Version Point 5 of the agenda Education in the EU-LFS recent developments Meeting of the Education and Training Statistics Working Group Room Quetelet Luxembourg, 6-7 September 2006

2 EDUCATION IN THE EU LABOUR FORCE SURVEY RECENT DEVELOPMENTS SUMMARY This paper aims at informing the ETS WG of the work done the last months and foreseen for the future concerning data on education from the EU-LFS: 1. Regular LFS data collection a) Annual averages b) Quality profile on the structural indicator on lifelong learning c) Updating of the NEACs (National Education Attainment Classifications) d) Conclusions from the Task Force meeting on explanatory notes for the LFS (19-20 June 2006) e) Proposal for a new structure of the LFS educational domain in NewCronos 2. Preparation of the LFS ad-hoc module 2009 "Entry of young people into the labour market" (first Task Force meeting on June 2006). a) Annual averages In order to improve the LFS estimations, Eurostat proposed moving the quarterly data to annual averages in autumn 2006 for two indicators - youth education level and lifelong learning at the same time as an annual LFS domain is created on the Eurostat website. The indicator early school leavers requires additional data (new code for students on holidays) which will be available from 2006 onwards. The annual averages on two structural indicators: lifelong learning and youth education attainment level were calculated and presented to the Employment Committee Indicator Group in July This group will decide whether annual averages should be used for the 2006 structural indicators autumn release during its 11 th /12 th October 2006 meeting. In general, no big differences were noted in comparing spring data and annual averages. Eurostat is now waiting for the validation of these annual averages by the NSIs. They will be published on the Eurostat website in October 2006 or Spring 2007 according to the decision of the Employment Committee Indicator Group. Appropriate footnotes and warnings will be added on Eurostat website to alert users about the changes in reference periods. For more detailed information please refer to annex 1. Annual averages will progressively be introduced for all LFS data presented in NewCronos. The ETS WG members are invited to take note of this information on introduction of annual averages. 2

3 b) Quality profile on the structural indicator on lifelong learning The background document for the quality profile was presented to the Employment Committee Indicator Group and countries (members of LAMAS WG and members of ETS WG) will be asked to provide their comments. A final document is foreseen for December The ETS WG members are asked to take note of the upcoming request for comments about the quality profile on structural indicator on lifelong learning. c) Updating of the NEAC (National Education Attainment Classifications) The last updated versions of the NEACs came to Eurostat in August. Seven missing countries are kindly asked to send their NEACs (or related information) by October. A first comparison of the responses shows that the question is worth more in-depth analyse because the responses were quite complex for some countries but very straightforward for others. As the LFS could be used as a model for the coding of educational level of persons in other social surveys, Eurostat will examine from a general perspective ways in which this variable is collected and coded in different surveys. A short questionnaire about the educational data collection in the surveys is planned to be launched this autumn with the aim of understanding better national situations and trying to harmonise the core variable on educational attainment. This question also arises if we try to evaluate to what extent the ISCED can be used to measure educational attainment (e.g. how to distinguish vocational education, how to determine educational attainment of non-nationals etc.) With the growing mobility of workers, a short but comprehensive NEACs (in English and national languages) available at least for the European countries could help the interviewers and improve the quality of data on education in social surveys. The ETS WG is invited to take note of the proposal to analyse the core variable on educational attainment in social surveys and to comment on it, especially on the possibilities to collaborating on this activity at a national level. 3

4 d) Conclusions from the Task Force meeting on explanatory notes for LFS (19-20 June 2006) One part of the meeting was devoted to the educational variables included in the LFS. After the TF members approved the conclusions of the meeting, a consultation was launched with all Member States (representatives for the LAMAS WG). A summary of the TF discussion: Issues related to the new code 3 for variable EDUCSTAT ("Person in regular education but on holidays") were discussed. As it is problematic to codify the future situation of the respondent, the TF agreed to apply the rule based on the previous situation of the person who is enrolled in education but is on school holidays and/or will continue studying. Questions about the level and field of education of such a person concern the previous educational level/field. For the variable EDUCSTAT, a clarification of the definition of apprentice was proposed, underlining the criterion of the apprenticeship belonging to the educational programme. Eurostat encourages countries to use the more detailed codes of field of education (see "Manual on fields of education and training", Eurostat 1999). For the best coding of the important variable HATLEVEL (educational level attained), use of the NEACs (National Educational Attainment Classifications) was strongly recommended. Some changes in the EU-LFS questionnaire were proposed for the future: - MAINSTAT: to be coherent with the EDUCSTAT variable, to remove apprenticeship from code 1 (work) to code 3 (student), and, consequently, to apply this change in WSTAT1Y (main status 1 year before). - EDUCSTAT: to distinguish 2 separate codes: for students and apprentices. - EDUCLEVL (level of current education): to apply the same, more detailed coding as in HATLEVEL (educational level attained) with the aim to allow distinguishing the participants in ISCED 3C programmes, with labour market destination. - COURPURP, COURFILD, COURWORH (characteristics of participation in current nonformal education): subject to discussion if the questions should be related to the most recent or to the most important learning activity. Comments from the members of the ETS WG on these proposals concerning the explanatory notes for LFS are very welcome before the end of September 2006 (directly to Eurostat or by contacting your colleagues working on the LFS). 4

5 e) Proposition for the new structure of education domain in NewCronos Given: a new approach to distinguishing Eurostat' NewCronos database data related to education from those on labour market characteristics, a wish to present data currently split in different domains (education, labour market, indicators) in a more coherent and user-friendly way and a wider use of the LFS annual data, a new structure of the educational domain in NewCronos is proposed to WG members, with a request for comments. Changes will be made during this autumn. This is also an opportunity to introduce a relatively new vocabulary from the Classification of Learning Activities (CLA) with the aim of familiarising users with this wording and especially with the concepts that lifelong learning starts in babyhood and that adult learning (probably a more appropriate word than education because of its "active" meaning) can include formal or non-formal education and informal learning. A general proposed schema is the following (for details please see annex 2): Lifelong learning: education and adult learning Formal education: educational systems (UOE data) and outputs of education (educational attainment levels, (LFS data); Adult learning: formal, non-formal and informal: LFS (annual data + ad-hoc module 2003), CVTS, and soon AES; Transition from education to work (LFS ad-hoc module 2000 and the forthcoming 2009 module). The members of the ETS WG are asked to send Eurostat comments on the proposed structure of the NewCronos domain "Lifelong learning: education and adult learning". 2. Conclusions of the first TF meeting preparing the LFS ahm 2009 "Entry of young people into the labour market" (15-16 June 2006) This Task Force preparing the EU-LFS ad-hoc module 2009 "Entry of young people into the labour market" is conducted jointly by Eurostat Unit F-2 (Labour market statistics) and by Unit F-4 (Education, science and culture statistics). National delegates come from 9 Member States: Austria, Czech Republic, Germany, Greece, Spain, France, Italy, Finland, United Kingdom, and from Switzerland. European policy makers show a significant interest in the topic of transition from education to work so representatives from DG EAC and DG EMPL, from Cedefop and from ETF also take part in this Task Force. 5

6 The objective of the first meeting of the Task Force LFS ad-hoc module 2009 was to define the aims and the target population of the survey and to make the first selection of topics to be covered. The proposed timetable foresees three TF meetings (June and November 2006, May 2007) and the consultation of the LAMAS and ETS Working Groups before sending a draft Regulation on the LFS ad-hoc module 2009 to the SPC for approval in November The subject of the transition from education to work is of great interest to national and European policy makers. The title of the LFS ad-hoc module 2009 ("Entry of young people into the labour market") suggests a more labour market oriented approach but bearing in mind that education plays an essential role in providing essential skills and is often a starting point for social exclusion. Policy makers are interested in better knowledge of patterns of transition from education to work and how these vary from country to country; what the obstacles to a smooth transition are and which measures can help young people to start an independent life. In many countries there is a problem of identifying the event 'leaving education', then the target population is proposed to be those persons aged years. Using the proxies in the LFS ad-hoc module survey can not be forbidden and therefore the questions asked should be the simplest and clearest possible. The testing of questions should be feasible in some countries (end of 2006/beginning of 2007), thanks to the supplementary budgetary resources available from Eurostat. It was recommended to explore more core LFS variables for analysis of the transition from education to work. Eurostat will analyse data on those topics from the core LFS and present the results of this exercise in the next TF meeting. The TF members (countries delegates as well as representatives of users from the Commission) were asked to provide comments on the draft list of the topics to be covered by the LFS ad-hoc module 2009, mentioning their priorities and remarks concerning the feasibility. The next meeting of the TF is planned for 30 th November/1 st December The draft minutes of the first meeting and all documents related to the work of the TF LFS adhoc module 2009 can be consulted on the Circa site: r/2006_06_15&vm=detailed&sb=title Members of the ETS WG are kindly invited to comment before the end of September on the proposed topics to be covered by the LFS ad-hoc module 2009 (see annex 3) by sending suggestions directly to Eurostat or contacting the colleagues working on the LFS in their countries. 6

7 Annex 1: Information on the transition to annual averages for LFS education indicators 1. TRANSITION TO ANNUAL AVERAGES FOR EU LFS RESULTS The main advantages of the transition to annual averages are: - The coverage of all the quarters/weeks of the year: Such reference period reflects the situation of the whole year and avoids problems of comparability among countries when using a unique quarter which can not be the most representative in all countries. - The increase of the sample size behind the estimation. This is especially the case of the LFS education based indicators. The EU LFS started its transition to a continuous quarterly survey in The last Member States implementing such structure were Cyprus in 2004 and Germany in The quarterly datasets start to be available at Eurostat progressively as follows: Table 1 - Availability of quarterly EU LFS data * 1998 CZ, EL, ES, PT, SK, FI 1999 BE, DK, IE, IT, HU, SI, UK, RO 2000 EE, NL, PL, BG 2001 SE 2002 LV, LT, MT 2003 FR, LU* 2004 CY, AT 2005 DE * Year when most variables are delivered on a quarterly basis. LU: annual continuous sample but not providing quarterly results. The transition to annual averages for all LFS based indicators is now possible. An annual domain will be created along with the domain presenting quarterly data 1. The annual database will be built as the concatenation of annual averages - as soon as quarterly data are available and spring data beforehand. In order to avoid breaks in series for EU aggregates, the EU data will integrate the annual averages for individual countries in 3 steps: in 2000 for the most countries available that year, in 2003 for the 5 / 6 countries available from 2001 to 2003 (according to the indicator) and in 2005 for Germany and Cyprus. Spring data will be used for those countries not having a complete set of quarterly data in between those 3 dates. 2. ANNUAL AVERAGES FOR EU LFS EDUCATION BASED INDICATORS The transition to annual averages will be already possible for the following indicators: 1 Annual averages are already used for the main indicators (including employment and unemployment rates). Missing quarters are estimated using national LFS data or by interpolation. Such procedures can not be applied to all LFS based indicators. 7

8 - Lifelong learning - Youth education attainment level For the indicator early school leavers, the transition to annual averages and more precise definition (regular education only) requires the use of a new code available in the LFS from 2006 onwards: This is to avoid considering early school leavers students on holidays: the LFS questions prior to 2006 checks the participation in the last 4 weeks which creates obvious problems for quarter 3. Once the results of the 2006 survey are available 2, Eurostat will check the correlation of the results on 'students on holidays' with other variables / codes which could be used for previous years for that quarter and will propose annual averages a priori from 2000, as for the other education based indicators. Principle: Spring data will be used until 1999 for all countries. In case of seasonal quarters (IE, UK), the annual average covers March of the year Y until February of the year Y+1. Averages for continuous half year surveys are made of the average of the two results (LV, LT, HR). For EU results, the introduction of country annual averages is implemented when quarterly data exist or can be estimated. The EU data will be flagged accordingly. - in 2000 (most countries: see the annexes 1.1 and 1.2 for each indicator). - in 2003 (along with the main break due to the new LFS questionnaire on participation in education and training) for 5 / 6 countries. - In 2005: for the two last countries (Germany and Cyprus). For country data, it is proposed to show according to the annual averages as soon as they are available from 2000 but NSI still needs to be consulted on this aspect (mainly Malta and Sweden for which data could be revised for ). Atypical quarters: lifelong learning Atypical data are detected graphically in case of break in the seasonality of quarterly results. This happens in few cases, especially when the national education questionnaire was changed during the year. The most frequent corrections are needed to smooth the effect of the 2003 LFS ad-hoc module on lifelong learning which over-estimated the participation as measured in other quarters (multiple questions have a positive impact on respondents memory). In case of atypical quarterly data, corrections are applied on the basis of the following or preceding quarterly data so that the seasonality of the following or preceding year is respected. When such estimation provides an annual estimation similar to the one obtained by suppression of the atypical quarter (average on three quarters), the atypical quarter is simply suppressed (no estimation to guarantee consistency with other breakdowns). Atypical quarters: youth education attainment level There is much less corrections needed for this indicator. In absence of seasonality, corrections are simply calculated on the basis of the spring data, applying the same ratio on 2 Quarter 3 data will be available by the end of 2006, too late to propose annual averages for early school leavers for the 2007 spring Council. 8

9 the corresponding quarter to be corrected. Estimations have been performed only in few cases where the quarterly data proved to be inconsistent with the current spring series. The list of corrections and reference periods selected for each country is presented in the annexes 1.1 and ANNUAL AVERAGES FOR THE INDICATOR YOUTH EDUCATION LEVEL The impact of the annual average should reflect the occurrence of new degrees obtained outside the spring season for young people 3. EU results are therefore slightly higher based on annual averages. The discrepancy slightly increases with the introduction of quarterly data, from 2000 to The increase is about 0.4 points, between 0.2 and 0.5 percentages points for both EU25 and EU15 aggregates. For New Member States, the increase is a bit higher: +0.8 points in The maximum impact on EU results is an increase of 0.5 points in 2004 and 2005, 0.5 points for EU15 in 2004, 0.9 points for NMS10 in 2002 and The impact is slightly bigger on females. Table 2 Youth education attainment level, EU results Annual average EU EU NMS Spring EU EU NMS Difference EU EU NMS The inclusion of new populations in additional quarters may however balance this effect (demographic factor). 9

10 Table 3 Youth education attainment level, EU25 results by gender Females Spring Annual average Difference Males Spring Annual average Difference For two countries (Latvia and Finland), annual averages results are however lower than spring data, respectively 1.9 and 1.4 in For 13 countries, the differences are between 0.4 and +0.5 points. 10 countries have differences between +0.6 and 1.5. Three countries have higher differences, Estonia, Lithuania and Malta. It is also in these three countries along with Belgium, Denmark, Latvia and Cyprus that differences between quarterly results are highest within the year 2005, more than 3 percentage points. For all these countries the small sample size and the sample design could explain the differences (further investigations are required). In other countries, differences within quarterly results are lower than 2 points except in Sweden and Finland, with respectively a range of 2.5 and 2.7 in

11 Table 4 Youth education attainment level, country data spring 2005 difference spring - annual average range (difference min-max of the 4 quarters) Difference spring - annual average on min mean max mean range (difference minmax of the 4 quarters) EU EU NMS BE CZ DK DE (1) EE EL ES FR IE IT CY (1) LV LT LU HU MT NL AT PL PT SI SK FI SE UK Note: data in bold are provisional. The same breaks as for spring series apply. 11

12 A ranking of 2005 spring or annual average results does not change significantly the country order for this indicator (table 5). Table 5 Youth education attainment level - Country classification, 2005 Spring Annual averages SK 91.8 SK 91.5 CZ 91.2 SI 90.6 PL 91.1 CZ 90.3 SI 90.5 PL 90.0 NMS NMS LT 87.8 LT 87.8 SE 87.5 SE 87.5 AT 85.9 AT 85.9 IE 85.8 IE 85.8 EL 84.1 EL 84.1 HU 83.4 HU 83.4 FI 83.4 FI 83.4 FR 82.6 FR 82.8 EE 82.6 LV 81.8 BE 81.8 EE 80.9 CY (1) 80.4 CY (1) 80.7 LV 79.9 BE 80.3 UK 77.7 UK 77.1 EU EU DK 77.1 DK 76.0 NL 75.6 NL 74.6 EU EU IT 73.6 IT 72.9 DE (1) 71.5 LU 71.1 LU 71.1 DE (1) 71.0 ES 61.8 ES 61.3 MT 53.7 PT 48.4 PT 49.0 MT

13 2.2 ANNUAL AVERAGES FOR THE INDICATOR LIFELONG LEARNING The impact of the annual average reflects the lower rates in quarter 3 along with specific patterns in quarters 1 and 4 in some countries. The difference between the spring series and annual averages increases with the introduction of quarterly data, from 2000 to 2005, i.e. a decrease between 0.2 and 0.7 percentages points The maximum impact on EU results is a decrease of -0.7 points in 2005, -0.8 points for EU15 in 2005, 0.7 points for NMS10 in The impact is slightly bigger on females Table 6 Lifelong learning, EU results Annual average EU25 7,5 7,5 7,6 9,0 9,9 10,3 EU15 8,0 8,0 8,1 9,8 10,7 11,3 NMS10 4,8 4,7 4,7 4,9 5,4 5,3 Spring EU25 7,9 7,8 7,9 9,2 10,3 11,0 EU15 8,4 8,3 8,5 9,9 11,1 12,1 NMS10 5,2 5,2 5,0 5,6 6,0 5,5 Difference EU25-0,4-0,3-0,3-0,2-0,4-0,7 EU15-0,4-0,3-0,4-0,1-0,4-0,8 NMS10-0,4-0,5-0,3-0,7-0,6-0,2 Table 7 Lifelong learning, EU25 results by gender Females Spring 8,4 8,4 8,5 10,0 11,1 11,9 Annual average 8,0 8,0 8,2 9,7 10,6 11,1 Difference -0,4-0,4-0,3-0,3-0,5-0,8 Males Spring 7,4 7,2 7,2 8,5 9,4 10,1 Annual average 7,1 6,9 6,9 8,3 9,1 9,5 Difference -0,3-0,3-0,3-0,2-0,3-0,6 13

14 The differences between spring and annual averages results in 2005 vary between -2.4 points for Slovenia, -2.3 for Finland and -0.3 points in Cyprus, -0.4 points in Latvia. For other countries, annual averages provides results lower than 1 point, except in Belgium, Spain, Sweden, Ireland and the United Kingdom where the difference vary between -1 and -1.7 points. As for quarterly variations, 13 countries (lifelong learning rate between 2 and 8%) have differences between minimum and maximum of quarterly results (range) within the year 2005 of less than 2.5 percentage points. In 6 countries, the range is between 2 and 5 p.p.. Highest variations generally occur in countries with higher rates, except for the Netherlands and the United Kingdom (high rate, low variation), and for Belgium and Cyprus (lower rates but high variations). 14

15 Table 8 Lifelong learning, country data spring 2005 difference spring - annual average range (difference min-max of the 4 quarters) Difference spring - annual average on min mean max mean range (difference minmax of the 4 quarters) EU EU NMS BE CZ DK DE (1) EE EL ES FR IE IT CY (1) LV LT LU HU MT NL AT PL PT SI SK FI SE UK Note: Data in bold are provisional. The same breaks as for spring series apply (mainly in 2003) 15

16 A ranking of 2005 spring or annual average results does not change significantly the country order for this indicator (table 9). Table 9 Lifelong learning - Country classification, 2005 Spring Annual averages SE 34,7 SE 32,1 UK 29,1 UK 28,0 DK 27,6 DK 27,4 IS 26,6 IS 25,7 FI 24,8 FI 22,5 NO 19,4 NO 17,8 SI 17,8 NL 15,9 NL 16,6 SI 15,3 AT 13,8 AT 12,9 EU15 12,1 EU-15 11,3 ES 12,1 ES 10,5 EU25 11 EU-25 10,3 BE 10 LU 8,5 LU 8,5 BE 8,3 DE 8,2 LV 7,9 IE 8 DE 7,7 FR 7,6 FR 7,1 LV 7,6 IE 6,9 LT 6,3 LT 6,0 IT 6,2 EE 5,9 CZ 5,9 CY 5,9 EE 5,9 IT 5,8 MT 5,8 CZ 5,6 CY 5,6 MT 5,3 NMS10 5,5 NMS10 5,3 PL 5 PL 4,9 SK 5 SK 4,6 PT 4,6 PT 4,1 HU 4,2 HU 3,9 HR 2,3 HR 2,1 EL 1,8 EL 1,9 RO 1,6 RO 1,6 BG 1,1 BG 1,3 16

17 Annex 1.1: Youth education attainment level - Selection of reference periods and estimation/correction list Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 BE S CZ S DK S DE S S S S S S EE S EL S ES S FR S. S... S... S... IE S. IT S E CY S. S... S... S.. S S LV S H H H H LT S H H H H LU S. S... S... S... HU. S MT S. S... S... S.. NL S AT S PL S. E PT S SI S SK S FI S SE S. S... S... S.. E UK S. E BG R HR R R R R R R R R R H H H H H H H H H H RO S TR R IS S... S... S... S.. NO S CH S... S... S... S... S... S... S... S.. Notes: Quarterly data is available S Spring data is published (no quarterly estimate for annual average) E Quarterly data is not available and estimated H Annual average computed on the basis of the two surveys (referenced as Q2 and Q4) R Quarterly data is not available: the closest quarterly result is used for EU totals only Note: Corrections are performed for Italy 2000Q1 (suppression), Poland 2000Q1 (estimation), UK 2001Q1 (suppression) and Sweden 2005Q1 (suppression). The 'suppression' method is selected when it provide similar results to the estimation method. Estimations are performed using previous or following quarterly data and applying the ratio of spring results between the current and previous or following year. Additional corrections may be required after validation with National Statistical Institutes (especially for the publication of annual averages for Malta 2002 and Sweden ). 17

18 Annex 1.2: Lifelong learning - Selection of reference periods and estimation/correction list Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 BE S CZ R R R R R R R R DK S DE (1) S S. S... S... S... S.. EE S EL S E ES S FR S S S S... IE R R R R R R R R R. E IT S E CY (1) S S S S S S LV R R R R R R R R LT S H H H H LU. S... S... S... S.. HU. S.. E MT. S... S... S.. NL. S.. E AT S... S... S... S... PL R R R R PT. S.. SI SK R R R R R R R R FI. S.. E SE S S S S UK S E E E E BG R R R R HR R R R R R R R R H H H H H H H H H H RO. S.. TR (1) IS S S S S NO S S S CH S S S S S S S S Quarterly data is available S Spring data is published (no quarterly estimate for annual average) E Quarterly data is not available and estimated H Annual average computed on the basis of the two surveys (referenced as Q2 and Q4) R Quarterly data is not available: the closest quarterly result is used for EU totals only Note: Corrections are performed for certain quarters: Italy 2000Q1 (suppression), Denmark 2003Q2 (suppression), Greece 2003Q2 (suppression), Finland 2003Q2 (suppression), Ireland 2003Q2 (estimation), Hungary 2003Q2 (estimation), UK 2001Q1 (suppression), UK 2003 Q1/Q2 UK 2004 Q1/Q2 (estimation) and Sweden 2005Q1 (suppression). The 'suppression' method is selected when it provide similar results to the estimation method. Estimations are performed using previous or following quarterly data and applying the seasonality of the previous or following year. Additional corrections may be required after validation with National Statistical Institutes (especially for the publication of annual averages for Malta 2002 and Sweden but also for breaks in series or impact of the 2003 ad-hoc module for 2003Q2 or 2003Q1 for Austria ). 18

19 Annexe 2: Proposition of the new structure of the NewCronos domain "Lifelong learning: education and adult learning" 19

20 Annex 3: Topics proposed to be covered by the LFS ahm 2009 o Parental educational level and occupation o Country of birth of parents o Impact of family issues (children) on the entry to the labour market o Description of the process o Studies, employment, unemployment, inactivity spells o Education as a second best option (to avoid failure on the labour market) o Characteristics of initial (formal) vocational education and training o Start of study o Month when highest education level achieved o Reason of leaving studies prematuraly o Formal education undertaken but not completed (reasons) o Traineeship o Certificated practical knowledge (possibilities to do it) o Influence of non-formal education on the labour market participation o Work experience during studies o Obstacles to find a job o Job search helping systems o Individual expenditure for the search of job o Time to get a job o First job (definition?, its characteristics) o How many jobs? o Subsidized jobs (but question of reliability of this information coming from individual) o Active Labour Market Policy (methods met to help labour market integration) o Entrepreneurship (wish to establishment as independent and obstacles for it) o Mismatch of skills o Required in-company training at the start of the first job o Instable employment (precarious jobs) o Situation of the special needs people (disabled) o Discrimination (experienced in looking for a job) o Mobility (inside, outside of the country) 20