This online annexe first provides descriptions and explanations to particular variables used in the explanatory model. In the appendix, additional evidence is provided on the data used (A) complete estimation results (B) and results from the sensitivity analysis (C). References to the tables in the appendix are made in the main text of the paper.
Structure of the annexe
1.2 Explanatory variables observed at the individual level
1.3 System characteristics observed at the regional or country level
Table A1: Identified groups of interest among the employed population
Table A3: Descriptive statistics of the EU LFS sample, employed 25-64 years old, by country
Table A4: Estimation results for individual characteristics (b/se/p)
Table A5: Estimation results for the household characteristics (b/se/p)
Table A6: Estimation results for the job characteristics (b/se/p)
Table A7: Estimation results for the employer´s characteristics (b/se/p)
Table A8: Estimation results for the role of system characteristics in AL participation (b/se/p)
Table B1: Descriptive statistics of the explanatory variables used in the models
Table B2: Weighting of household member in computation of the Care index
Table B3: Correlation coefficients among selected system determinants
Table B5: Estimates of the sub-equations of the model
1. Explanatory model
Our theoretical model assumed a regression type of model with one (main) dependent variable and a wide list of explanatory variables. For the sake of better orientation, we organize the explanatory variables in blocks. Namely referring to: characteristics of individuals, their household, job as well as employer observable at the level of individuals; and finally system-level characteristics - referring to policies or their outcomes observable at the national/regional level. Where feasible, explanatory variables are stylised to indicate potential barriers to AL.
The dependent variable for further analysis is the AL participation rate. The definition of AL participation differs between available data sources surveying European households (EU-LFS and AES) (CEDEFOP 2015, p. 31). While the AES inquires about AL participation during the period of 12 months prior to the collection of the survey, EU-LFS asks about AL participation within the last four weeks prior to the surveying period. The surveying period of EU-LFS observations shifts randomly during the whole calendar year (to avoid biases caused by seasonality), with a quarterly data collection and sample components remaining in the sample for up to 4 quarters.
EU-LFS further allows a more precise distinction between learning activities by distinguishing between (i) formal and non-formal learning; and (ii) work-related and not work-related learning. Nevertheless, the information necessary to identify work-related AL was not collected in all 31 European countries. Therefore, we only distinguish between formal learning and non-formal learning activities.
Our dependent variable is collected at the level of individuals and have the form of a dummy variable which indicates whether the surveyed individuals did participate in the particular type of AL activity, during the 4 weeks reference period (1) or not (0).
In reality, some adults participate in both identified types of AL, formal as well as non-formal, during the 4 weeks reference period. Such cases present only 0.62% of the whole sample[1]. The construction of our dependent variables assesses adults separately for formal and non-formal AL.
Table 1: Proportions (in %) of employed adults (25-64 years) in the sample by AL participation status
|
|
Non-formal AL |
Total |
||
|
Not participated |
Participated |
|||
|
Formal AL |
Not participated |
88.38 (1 421 300) |
9.41 (151 300) |
97.79 (1 572 600) |
|
Participated |
1.78 (28 600) |
0.43 (6 900) |
2.21 (35 500) |
|
|
Total |
90.16 (1 449 900) |
9.84 (158 200) |
100.00 (1 608 100) |
|
Note: Shares in %, number of unweighted observations in the parenthesis (rounded to hundreds)
Source: EU-LFS 2016
In absolute terms, the dataset includes over 1.6 million observations of employed adults aged 25-64. Approximately 28.6 thousand adults participated in formal education and more than 151 thousand adults in non-formal learning activities. Approximately, 6,900 individuals combine formal education with non-formal learning activities.
1.2 Explanatory variables observed at the individual level
Although the existing literature on AL participation often considers barriers to AL participation (Section 2), our data limit us to construct variables purely based on available characteristics. However, based on the previous literature, we are able to construct proxies for barriers to AL participation, for example, situational barriers can be captured by the characteristics of the household, like having small children. We also consider individual characteristics (e.g. migrant status, female vs male) and characteristics of the job. Institutional barriers are easily collected from the official statistics such as Eurostat or the World Bank; some institutional barriers were constructed by using EU LFS. Dispositional barriers cannot be identified directly in the EU-LFS data and had to be approximated. For example, the flexibility of the labour market can be captured by considering the share of persons in temporary contracts at the regional level and the dismissal rate at the regional level is a good proxy of the level of employment protection. We believe to provide some information on the dispositional barriers by looking specifically at the disadvantaged groups. A basic overview of the model is displayed in Scheme A (appendix). As follows, we first focus on the construction of individual-level characteristics of households, individuals and their jobs.
a. Household characteristics
For the household related barriers, we construct two indexes grasping the mechanisms behind deciding on the final number of working hours: (1) the Care index; and (2) the Share of non-earners in the household.
Care index - The Care index quantifies the need for allocating time to carrying activities within the household. In computing this index we take into account the number of children and elderly in the household of the responding (employed) person (See Table 2). When summing the number of household members in the index, individuals are ascribed a different weight based on their age.
Table 2: Weighting of a household member in the computation of the Care index
|
Age of the household member |
Weight of the household member |
|
0-4 |
4 |
|
5-9 |
3 |
|
10-14 |
2 |
|
15-19 |
1 |
|
20-59 |
0 |
|
60-74 |
1 |
|
75-84 |
2 |
|
85+ |
3 |
Weights of household members are ascribed arbitrary, assuming a different intensity of carrying activities based on the age of the household member. The Care index enters the model separately for males and females. First, we explore its association with the number of actual working hours during the reference period. Here we expect, a negative association between the Care index and AL participation, more pronounced in the case of females.
Share of non-earners in the household -The second index constructed out of the information observable for the household is the proportion of non-earners among the household members. The values of the index also enter in the model separately for males and females. Similarly, we first explore its association with the working hours and in the second step with the AL participation. We assume that especially for males, the association with working hours is going to be positive, as more non-earners in the household should press on male members of the household to work more working hours.
Working hours - Here, the sum of actual working hours from all declared jobs is used. The time period for which the actual working hours are inquired is the same as the reference period for which AL participation is indicated.
We assume a negative trade-off between work and formal AL because time is a limited resource and in the case of formal learning, it is not allocable simultaneously to work and AL. This is different for the non-formal AL, while a dominant share of it is being delivered at the workplace. More working hours may, thus, lead to higher participation in non-formal AL.
b. Job characteristics
Out of the information on occupation in the current job, we construct four variables: (1) over-education index; (2) mean probability of computerization; (3) high-skill occupation (dummy); and (4) Supervision (dummy).
Over-education index - To construct the
over-education index, we first create an individual specific variable for years
of schooling. This is an estimate based on the year of graduation of individual
with regard to the highest level
of education attained by this individual, and relative to his/her year of
birth:
HATYEAR denotes the year when the highest level of education
was successfully completed; YEAR the year of the data collection (in our case
2016); AGE the age of the individual respondent; and ENTRANCE_PRIMARY the
official entrance age in compulsory education (country-specific). The subscript
denotes
the individual.
In the next step, we construct the variable median years of
schooling. This variable indicates how long it takes for an individual living in
country 
and enrolled in a particular
educational track 
to graduate from that
educational level. We express the duration in the median years of schooling as
to account for extreme values.
Using a similar approach, we also construct the median years
of schooling for an individual in occupation 
and accounting
for differences across countries:
Subscript o denotes each code from the International Standard Classification of Occupations (ISCO). The EU LFS edition 2017 provides us with ISCO 3-digit.
Finally, the over-education index is the difference between individuals´ median years of schooling, which has been determined by the highest level of education attained, and the occupational median years of schooling observed for all individuals within the occupational group defined at the level of ISCO 3-digit. We then may write:
The over-education index is included in the model separately for individuals working in a low-skilled and high-skilled occupation. This should allow us to explore different dynamics in the association between over-education and AL participation conditional on the complexity of the current job.
Two possible directions of the association between over-education and AL participation can be assumed. Either those who are overeducated in their current job, are overeducated because they have a higher propensity (motivation) to acquire more education for which they are more likely to participate in AL; or in the contrary, underutilizing the already acquired education in the current job leads to a decline in the propensity (motivation) to seek further education and thus participating in AL.
Mean risk of computerization - The variable of the mean probability of computerization is created using the probabilities of computerization estimated in Frey and Osborne (2016). They use the US occupational database of job descriptions - O*NET. Occupation specific probability of computerization is quantified based on an assessment of the tasks and job descriptions from O*NET in the context of the expected advances in machine learning and robotics. Authors report the probability of computerization for the occupations coded in the Standard Occupational Classification (SOC). The SOC classification can be translated into the International Standard Classification of Occupations (ISCO) used in the EU-LFS. When shifting to a different level of detail we use the mean probability of computerization estimated for all the occupations within the concerned occupational group assuming equal weight for each of the sub-occupations.
High-skilled occupation - The high skilled occupation is defined by the occupational group of ISCO 1 to ISCO4. The occupational classification ISCO is coded hierarchically in terms of the complexity of the tasks performed in the occupation. Occupations coded in the group ISCO 1 are the most complex, require the highest level of skills and education; an occupational group of ISCO 9 shelters the least complex occupations. Although the risk of computerization shows a small negative correlation with the occupational complexity, risk of computerization also appears in the most complex occupations.
We assume that for individuals in occupations with higher complexity of tasks, the need for further learning is higher.
Supervision - Finally, we use a dummy for whether the job of the responding person involves supervising duties. Here we assume, that supervising duties are linked with more responsibility, but also a need for additional job-related training. On the other hand, supervising is connected with a more demanding job and, therefore, leaves less time to work-unrelated AL activities.
c. Individual characteristics
In addition to the barriers to AL identified at the individual level and the system determinants identified at the regional or country level, we construct variables measuring the characteristics of individuals. These characteristics are considered important control variables for inclusion in the multi-layered empirical framework. Here we include the “usual suspects” available in the EU-LFS questionnaire, namely: Gender (dummy); Age (continuous); Level of education; Level of urbanization.
Occupation related information is explored in the case of the job-related barriers and was therefore not included in the list of control variables.
d. Employer’s characteristics
Job-related AL significantly differs based on the characteristics of the employer. The EU-LFS data allow us to identify the sector of economic activity as well as constructing a proxy for an employer’s size[2]. Those two variables appear to be the strongest factors explaining the provision of job-related AL by employers (CEDEFOP 2015). Additionally, we have complemented the set of employer’s characteristics by the variable indicating whether the employed person is looking for another job. Since we do control for a relatively rich list of job-related characteristics, we consider “looking for a job” to be an overall assessment of actual working conditions.
1.3 System characteristics observed at the regional or country level
In selecting the relevant system-level determinants of AL participation, we depart from the model outlined in Groenez et al. (2007). Because of the high collinearity between some of the system level determinants, we are first accounting for associations between them in order to obtain unbiased coefficients for each of their direct associations with the dependent variable.
a. Subjective assessment of costs being the main obstacle in AL participation
In between the individual and system level of variables lies the assessment of costs being the main obstacle in AL participation. This variable is constructed based on the data from the AES, collected for 2016 [Data code: trng_aes_179]. Its values present the share of the interviewed persons identifying costs to be the main obstacle in AL participation. [Data code: trng_aes_179]
b. Demography
Here we decided to use only one variable, the regional mean age of the population computed by using the EU-LFS microdata for 2016. It reflects the extent to which society suffers from ageing. As the returns AL decreases with higher age, older societies may invest less in AL than the younger ones (Groenez et al., 2007). This may be especially true for formal AL as this is more often accredited and thus to a higher extent recognized at the labour market.
The originally included variable of the mean number of household members under 25 was later excluded from the model because it was not showing any significant association.
c. Initial education
The shape of the system of initial education to a substantial extent determines AL participation. We grasp the main shape of the system by four variables: a) the years of compulsory schooling; b) age of entrance into lower secondary education; c) the share of students at the upper secondary level in vocational programmes; d) public expenditures on education as a per cent share on GDP.
The years of compulsory schooling are based on the UNESCO UIS.Stat database. It varies at the country level and grasps the international differences in the amount of initial schooling provided by the state. We have considered to include also the average years of schooling averaged at the regional level, but the years of compulsory schooling showed to be explanatory stronger. All the system level variables in the area of initial education are, thus, collected at the level of countries; and refer rather to the design of the educational policies, than to the outcomes of these policies.
A wide list of studies documents a positive association between initial schooling and AL participation at the country level (OECD 2003; Brunello, 2001; Jung & Cervero, 2002; Bassanini et al. 2005; Wöβmann & Schütz, 2006). Given this, we expect to observe a positive association also at the individual level.
The age of entrance into lower secondary education determines the age at which pupils are being selected into specialised educational programmes – educational tracks. When this age is higher, the ratio of skills received by pupils within initial schooling is more in favour of general than specialized skills. When exploring available literature, Groenez et al. (2007) find support for both of the directions of the association between the level of specialization in initial schooling and AL participation, namely: less specialized and more general systems of initial education precondition higher AL participation because of the lack of specialized skills received during initial education (Antikainen, 2006; Brunello, 2006); and more specialised and less general systems of initial education precondition higher AL participation because of specialised skills getting obsolete faster than general skills (Bassanini et al., 2005).
d. Labour market
Among the system determinants in the area of labour market, we use the regional employment rate of the main age group (25-64), based on the EU-LFS; regional share of employed persons aged 25-64 who have a temporary contract type, based on the EU-LFS; regional share of persons who were dismissed or were made redundant in their last job or business, based on the EU-LFS; country-level active labour market policy expenditures on training (LMP type 2) as a share of GDP, based on the Labour Market Policy Database administrated by Eurostat.
The employment rate is strongly positively correlated with regional GDP, for what we have accounted in the model by implementing an association path between the employment rate and regional GDP. The regional employment rate is also endogenous in the model, allowing its association with other labour market system determinants (share of dismissals, the share of temporary contracts, ALMP expenditures on training), public expenditures on education and the number of years in compulsory education.
Based on Groenez and co-authors (2007) we assume that AL participation might be higher in regions with higher employment rate mainly because the workplace generates an additional supply of training (which is true mostly for what we consider here as the non-formal training) (McGivney, 2001); adults are more confident that learning will be utilised in a better job (McGivney, 2001); companies invest more in human capital when confronted with shortages of skilled labour (Gorard & Rees, 2002).
Although we control for individual-level working hours, the regional share of temporary contracts is included to proxy for the employment legislation. Lassnigg (2005) and McGivney (2001) hint that employers are less willing to invest in AL of employees working under part-time or temporary contracts, which might imply a lower regional supply of AL.
The share of dismissals is used as a proxy for employment protection (flexibility). We assume that both sides - employers and employees - are more willing to invest in AL if employment protection is higher. More flexible labour markets allow flexible hiring and dismissal to compensate for potential skills shortages or obsolescence.
Active labour market policy expenditures on training present alternative support to the supply of AL as the programmes financed from these resources might target unemployed as well as employed.
e. Economy
Exploring their empirical associations, we have limited the number of economy-related system determinants of AL to only two, namely: the Gross domestic product (GDP) at current market prices by NUTS 2 regions in euro per inhabitant, based on the Statistics of National Accounts published by Eurostat; and patent applications to the European patent office (EPO), published by Eurostat.
The GDP indicator is always a “central” indicator when considering various aspects of the economy, such as the openness, labour productivity etc. An aspect of special importance, in the context of AL participation, is the innovation intensity of the production. Here we capture this aspect using the indicator of the number of patent applications to the European patent office.
Groenez and co-authors (2007) estimate multiple specifications of explanatory models on AL participation and inequality in AL participation. They identify the level of innovation as one of the key system determinants appearing as significant in all of their model specifications. Its possible association with AL participation is assumed. An overview of the descriptive statistics of the explanatory variables is provided in Table B1 of the Appendix.
2. Appendix
Table A1: Identified groups of interest among the employed population
|
Group of interest |
Age of participants |
Highest education attained |
|
All |
25-64 |
All |
|
Low-educated adults |
25-64 |
≤ ISCED 2 |
|
Low-educated young adults |
20-29 |
≤ ISCED 2 |
|
Migrants |
25-64 |
All |
Source: From the authors.
Table A2: Number of unweighted observations in the (EU-LFS 2016 dataset), by the type of AL activity
|
Groups of interest |
Observations in the sample (Unweighted) |
Population (Weighted) |
Participation rate - Formal AL (%) |
Participation rate - Non-formal AL (%) |
|
Population 25-64 |
2 247 000 |
274 220 800 |
3.16 |
8.43 |
|
Employed adults 25-64 |
1 608 100 |
200 844 000 |
2.58 |
9.91 |
|
Low-educated adults 25-64 |
276 400 |
34 859 600 |
0.89 |
4.23 |
|
Low-educated young adults 20-29 |
31 600 |
4 958 200 |
13.37 |
5.77 |
|
Employed migrants 25-64 |
171 600 |
25 879 800 |
3.35 |
8.20 |
Note: Groups are identified based on the definitions in Table A1
Source: EU-LFS 2016
Table A3: Descriptive statistics of the EU LFS sample, employed 25-64 years old, by country
|
Observations in the sample (Unweighted) |
Population (Weighted) |
Participation rate - Formal AL (%) |
Participation rate - Non-formal AL (%) |
|
|
Austria |
76 394 |
3 641 500 |
3.3 |
12.2 |
|
Belgium |
37 774 |
4 243 400 |
1.5 |
5.6 |
|
Bulgaria |
13 000 |
2 822 200 |
0.9 |
0.4 |
|
Croatia |
10 425 |
1 445 300 |
1.3 |
1.0 |
|
Cyprus |
15 829 |
328 800 |
1.9 |
4.7 |
|
Czech Republic |
16 986 |
4 730 100 |
1.5 |
10.2 |
|
Denmark |
39 348 |
2 323 900 |
5.9 |
24.1 |
|
Estonia |
10 370 |
563 000 |
3.9 |
12.1 |
|
Finland |
10 410 |
2 119 900 |
7.9 |
22.4 |
|
France |
168 895 |
23 599 300 |
1.2 |
20.0 |
|
Germany |
228 657 |
36 815 300 |
2.6 |
5.9 |
|
Greece |
75 192 |
3 472 500 |
1.2 |
2.1 |
|
Hungary |
87 138 |
4 008 300 |
0.7 |
6.1 |
|
Italy |
186 196 |
21 263 700 |
1.1 |
7.8 |
|
Latvia |
15 814 |
800 200 |
1.9 |
5.1 |
|
Lithuania |
25 384 |
1 212 500 |
1.2 |
5.3 |
|
Luxembourg |
13 471 |
245 400 |
1.8 |
16.0 |
|
Malta |
9 171 |
164 800 |
4.2 |
5.2 |
|
Netherlands |
34 920 |
6 970 800 |
6.1 |
14.9 |
|
Norway |
11 478 |
2 221 700 |
4.5 |
16.2 |
|
Poland |
114 362 |
14 764 900 |
1.4 |
2.6 |
|
Portugal |
61 256 |
4 108 800 |
1.8 |
7.4 |
|
Romania |
89 006 |
7 684 100 |
0.1 |
0.9 |
|
Slovak Republic |
35 722 |
2 311 400 |
0.7 |
2.1 |
|
Slovenia |
24 897 |
844 900 |
4.1 |
9.0 |
|
Spain |
34 868 |
17 378 900 |
3.2 |
7.4 |
|
Sweden |
129 942 |
4 214 900 |
5.0 |
25.1 |
|
United Kingdom |
31 190 |
26 543 700 |
3.4 |
13.1 |
|
Total |
1 608 095 |
200 843 900 |
2.6 |
9.9 |
Note: Population of employed, 25-64 years old in the sample, using sampling weights, individuals may combine participation in formal and non-formal AL. Rates in %, numbers rounded to hundreds.
Source: EU-LFS 2016
Table A4: Estimation results for individual characteristics (b/se/p)
|
Target group |
Employed |
||||||||
|
All |
Low-educated |
Young and low-educated |
Migrants |
||||||
|
Type of AL |
Formal |
Non-formal |
Formal |
Non-formal |
Formal |
Non-formal |
Formal |
Non-formal |
|
|
Female |
-0.033 |
0.241 |
0.069 |
0.140 |
-0.109 |
-0.175 |
-0.027 |
0.194 |
|
|
0.045 |
0.029 |
0.121 |
0.075 |
0.194 |
0.192 |
0.093 |
0.071 |
||
|
0.457 |
0.000 |
0.570 |
0.061 |
0.575 |
0.362 |
0.773 |
0.006 |
||
|
Age |
-0.099 |
-0.008 |
-0.103 |
-0.016 |
(omitted) |
(omitted) |
-0.095 |
-0.014 |
|
|
0.007 |
0.001 |
0.010 |
0.002 |
|
0.009 |
0.003 |
|||
|
0.000 |
0.000 |
0.000 |
0.000 |
|
|
0.000 |
0.000 |
||
|
Level of education (primary omitted) |
Lower secondary |
-0.029 |
0.276 |
(omitted) |
(omitted) |
(omitted) |
(omitted) |
0.010 |
0.396 |
|
0.105 |
0.039 |
|
|
0.130 |
0.106 |
||||
|
0.785 |
0.000 |
|
|
|
|
0.938 |
0.000 |
||
|
Upper secondary |
0.496 |
0.376 |
(omitted) |
(omitted) |
(omitted) |
(omitted) |
0.488 |
0.411 |
|
|
0.075 |
0.045 |
|
|
0.137 |
0.076 |
||||
|
0.000 |
0.000 |
|
|
0.000 |
0.000 |
||||
|
Tertiary |
0.446 |
0.504 |
(omitted) |
(omitted) |
(omitted) |
(omitted) |
0.343 |
0.561 |
|
|
0.105 |
0.075 |
|
|
0.164 |
0.111 |
||||
|
0.000 |
0.000 |
|
|
0.037 |
0.000 |
||||
|
Degree of urbanization (City omitted) |
Town |
-0.295 |
-0.055 |
-0.019 |
0.029 |
0.113 |
0.135 |
-0.263 |
-0.009 |
|
0.045 |
0.025 |
0.095 |
0.067 |
0.154 |
0.171 |
0.074 |
0.068 |
||
|
0.000 |
0.025 |
0.845 |
0.663 |
0.461 |
0.429 |
0.000 |
0.898 |
||
|
Rural |
-0.428 |
-0.057 |
0.071 |
0.168 |
0.255 |
0.172 |
-0.308 |
0.041 |
|
|
0.051 |
0.026 |
0.121 |
0.063 |
0.207 |
0.205 |
0.099 |
0.067 |
||
|
0.000 |
0.026 |
0.556 |
0.008 |
0.219 |
0.400 |
0.002 |
0.539 |
||
Legend: b/se/p
Source: EU-LFS 2016
Table A5: Estimation results for the household characteristics (b/se/p)
|
Target group |
Employed |
||||||||
|
All |
Low-educated |
Young and |
Migrants |
||||||
|
low-educated |
|||||||||
|
Type of AL |
Formal |
Non-formal |
Formal |
Non-formal |
Formal |
Non-formal |
Formal |
Non-formal |
|
|
Household related barriers |
Female*care index |
-0.092 |
-0.023 |
-0.044 |
-0.026 |
-0.072 |
0.015 |
-0.074 |
-0.028 |
|
0.011 |
0.003 |
0.020 |
0.012 |
0.027 |
0.033 |
0.026 |
0.007 |
||
|
0.000 |
0.000 |
0.027 |
0.032 |
0.009 |
0.647 |
0.005 |
0.000 |
||
|
Male*care index |
-0.071 |
-0.004 |
-0.075 |
-0.014 |
-0.083 |
-0.013 |
-0.066 |
-0.004 |
|
|
0.008 |
0.004 |
0.022 |
0.009 |
0.039 |
0.034 |
0.021 |
0.010 |
||
|
0.000 |
0.292 |
0.001 |
0.115 |
0.034 |
0.706 |
0.001 |
0.694 |
||
|
Male*non earners in the HH |
0.163 |
-0.305 |
-0.294 |
-0.277 |
-0.651 |
-0.457 |
0.183 |
-0.327 |
|
|
0.082 |
0.050 |
0.238 |
0.116 |
0.406 |
0.446 |
0.229 |
0.135 |
||
|
0.047 |
0.000 |
0.217 |
0.017 |
0.108 |
0.306 |
0.425 |
0.015 |
||
|
Female*non earners in the HH |
0.377 |
-0.167 |
-0.053 |
-0.458 |
0.204 |
-0.269 |
0.660 |
-0.083 |
|
|
0.077 |
0.046 |
0.263 |
0.143 |
0.543 |
0.607 |
0.235 |
0.112 |
||
|
0.000 |
0.000 |
0.840 |
0.001 |
0.707 |
0.657 |
0.005 |
0.461 |
||
|
Working hours |
-0.014 |
0.010 |
-0.008 |
0.009 |
-0.008 |
0.005 |
-0.018 |
0.007 |
|
|
0.002 |
0.001 |
0.003 |
0.001 |
0.003 |
0.006 |
0.002 |
0.001 |
||
|
0.000 |
0.000 |
0.001 |
0.000 |
0.019 |
0.432 |
0.000 |
0.000 |
||
Legend: b/se/p
Source: EU-LFS 2016
Table A6: Estimation results for the job characteristics (b/se/p)
|
Target group |
Employed |
||||||||
|
All |
Low-educated |
Young and |
Migrants |
||||||
|
low-educated |
|||||||||
|
Type of AL |
Formal |
Non-formal |
Formal |
Non-formal |
Formal |
Non-formal |
Formal |
Non-formal |
|
|
Job related barriers |
Low-skilled*Over-education |
0.088 |
0.051 |
0.059 |
0.008 |
0.093 |
0.037 |
0.085 |
0.060 |
|
0.012 |
0.009 |
0.019 |
0.016 |
0.031 |
0.023 |
0.015 |
0.013 |
||
|
0.000 |
0.000 |
0.001 |
0.610 |
0.003 |
0.116 |
0.000 |
0.000 |
||
|
High-skilled*Over-education |
0.001 |
0.039 |
0.077 |
0.020 |
0.030 |
-0.005 |
0.014 |
0.031 |
|
|
0.010 |
0.006 |
0.033 |
0.019 |
0.042 |
0.044 |
0.016 |
0.010 |
||
|
0.880 |
0.000 |
0.018 |
0.294 |
0.473 |
0.907 |
0.377 |
0.002 |
||
|
Risk of computerisation |
-0.268 |
-0.283 |
-0.550 |
-0.087 |
0.056 |
0.355 |
-0.011 |
-0.373 |
|
|
0.055 |
0.024 |
0.179 |
0.095 |
0.265 |
0.300 |
0.117 |
0.072 |
||
|
0.000 |
0.000 |
0.002 |
0.359 |
0.832 |
0.237 |
0.926 |
0.000 |
||
|
High skilled occupation (dummy) |
0.314 |
0.446 |
0.519 |
0.568 |
0.531 |
0.352 |
0.421 |
0.520 |
|
|
0.053 |
0.026 |
0.152 |
0.084 |
0.203 |
0.258 |
0.102 |
0.057 |
||
|
0.000 |
0.000 |
0.001 |
0.000 |
0.009 |
0.173 |
0.000 |
0.000 |
||
|
Supervision (dummy) |
-0.077 |
0.264 |
-0.103 |
0.406 |
-0.510 |
0.230 |
-0.088 |
0.316 |
|
|
0.034 |
0.017 |
0.173 |
0.060 |
0.279 |
0.264 |
0.094 |
0.060 |
||
|
0.022 |
0.000 |
0.553 |
0.000 |
0.067 |
0.385 |
0.348 |
0.000 |
||
Legend: b/se/p
Source: EU-LFS 2016
Table A7: Estimation results for the employer´s characteristics (b/se/p)
|
Target group |
Employed |
|||||||||
|
All |
Low-educated |
Young and low-educated |
Migrants |
|||||||
|
Type of AL |
Formal |
Non-formal |
Formal |
Non-formal |
Formal |
Non-formal |
Formal |
Non-formal |
||
|
Looking for a job |
0.304 |
0.257 |
0.503 |
0.466 |
-0.598 |
0.390 |
0.171 |
0.266 |
||
|
0.053 |
0.030 |
0.200 |
0.098 |
0.218 |
0.264 |
0.095 |
0.080 |
|||
|
0.000 |
0.000 |
0.012 |
0.000 |
0.006 |
0.140 |
0.070 |
0.001 |
|||
|
Economic sector (Public services omitted) |
Agriculture |
-0.765 |
-0.624 |
-1.183 |
-0.880 |
-0.526 |
-0.770 |
-0.973 |
-0.849 |
|
|
0.122 |
0.063 |
0.243 |
0.136 |
0.387 |
0.468 |
0.310 |
0.308 |
|||
|
0.000 |
0.000 |
0.000 |
0.000 |
0.174 |
0.100 |
0.002 |
0.006 |
|||
|
Industry |
-0.820 |
-0.357 |
-0.731 |
-0.426 |
-0.381 |
-0.445 |
-1.002 |
-0.436 |
||
|
0.042 |
0.044 |
0.147 |
0.075 |
0.171 |
0.188 |
0.102 |
0.072 |
|||
|
0.000 |
0.000 |
0.000 |
0.000 |
0.026 |
0.018 |
0.000 |
0.000 |
|||
|
Construction |
-0.974 |
-0.585 |
-1.071 |
-0.641 |
-0.755 |
-0.463 |
-1.552 |
-0.742 |
||
|
0.067 |
0.047 |
0.209 |
0.090 |
0.239 |
0.327 |
0.234 |
0.145 |
|||
|
0.000 |
0.000 |
0.000 |
0.000 |
0.002 |
0.157 |
0.000 |
0.000 |
|||
|
Private services |
-0.543 |
-0.338 |
-0.792 |
-0.480 |
-0.347 |
-0.561 |
-0.636 |
-0.402 |
||
|
0.034 |
0.033 |
0.108 |
0.058 |
0.149 |
0.178 |
0.094 |
0.067 |
|||
|
0.000 |
0.000 |
0.000 |
0.000 |
0.019 |
0.002 |
0.000 |
0.000 |
|||
|
Number of employees in the local unit (over 50 omitted) |
Less than 10 |
-0.115 |
-0.147 |
-0.139 |
-0.337 |
0.147 |
0.097 |
-0.163 |
-0.218 |
|
|
0.025 |
0.016 |
0.141 |
0.047 |
0.194 |
0.170 |
0.097 |
0.044 |
|||
|
0.000 |
0.000 |
0.323 |
0.000 |
0.451 |
0.567 |
0.091 |
0.000 |
|||
|
11 to 20 |
-0.045 |
-0.075 |
0.219 |
-0.015 |
0.464 |
0.093 |
0.026 |
-0.080 |
||
|
0.029 |
0.023 |
0.141 |
0.070 |
0.250 |
0.231 |
0.092 |
0.070 |
|||
|
0.125 |
0.001 |
0.121 |
0.834 |
0.063 |
0.687 |
0.778 |
0.252 |
|||
|
21 to 50 |
-0.118 |
-0.009 |
0.028 |
-0.005 |
0.250 |
0.216 |
-0.185 |
-0.126 |
||
|
0.040 |
0.018 |
0.133 |
0.054 |
0.167 |
0.188 |
0.131 |
0.072 |
|||
|
0.003 |
0.607 |
0.832 |
0.930 |
0.133 |
0.251 |
0.157 |
0.080 |
|||
Legend: b/se/p
Source: EU-LFS 2016
Table A8: Estimation results for the role of system characteristics in AL participation (b/se/p)
|
Target group |
Employed |
||||||||
|
All |
Low educated |
Young and low-educated |
Migrants |
||||||
|
Type of AL |
Formal |
Non-formal |
Formal |
Non-formal |
Formal |
Non-formal |
Formal |
Non-formal |
|
|
Costs of AL |
0.017 |
0.047 |
-0.084 |
0.102 |
-0.148 |
0.132 |
-0.006 |
0.062 |
|
|
0.013 |
0.016 |
0.023 |
0.022 |
0.040 |
0.031 |
0.016 |
0.021 |
||
|
0.172 |
0.003 |
0.000 |
0.000 |
0.000 |
0.000 |
0.725 |
0.003 |
||
|
Demography |
Mean age of the |
-0.033 |
0.031 |
-0.042 |
0.048 |
-0.019 |
0.112 |
-0.014 |
0.028 |
|
regional population |
0.014 |
0.018 |
0.024 |
0.026 |
0.036 |
0.029 |
0.017 |
0.028 |
|
|
|
0.019 |
0.094 |
0.087 |
0.059 |
0.595 |
0.000 |
0.414 |
0.313 |
|
|
Initial education |
Years of compulsory |
-0.145 |
-0.350 |
-0.029 |
-0.326 |
0.064 |
-0.418 |
-0.254 |
-0.224 |
|
schooling |
0.049 |
0.064 |
0.069 |
0.094 |
0.112 |
0.104 |
0.044 |
0.065 |
|
|
|
0.003 |
0.000 |
0.672 |
0.000 |
0.569 |
0.000 |
0.000 |
0.001 |
|
|
Entrance age into |
0.249 |
0.009 |
0.397 |
0.107 |
0.302 |
0.043 |
0.185 |
0.067 |
|
|
lower secondary education |
0.047 |
0.058 |
0.103 |
0.098 |
0.139 |
0.105 |
0.057 |
0.080 |
|
|
|
0.000 |
0.873 |
0.000 |
0.275 |
0.030 |
0.684 |
0.001 |
0.397 |
|
|
Share of students in |
0.008 |
0.005 |
-0.004 |
0.003 |
-0.022 |
-0.008 |
0.013 |
0.000 |
|
|
vocational programmes |
0.004 |
0.005 |
0.006 |
0.007 |
0.009 |
0.008 |
0.005 |
0.005 |
|
|
|
0.059 |
0.350 |
0.504 |
0.708 |
0.017 |
0.321 |
0.004 |
0.956 |
|
|
Government expenditure on |
-0.002 |
0.655 |
0.452 |
0.919 |
0.617 |
0.514 |
0.572 |
0.719 |
|
|
education as a % of GDP |
0.166 |
0.143 |
0.271 |
0.228 |
0.481 |
0.292 |
0.181 |
0.229 |
|
|
|
0.988 |
0.000 |
0.095 |
0.000 |
0.200 |
0.078 |
0.002 |
0.002 |
|
|
Labour market |
Employment rate |
0.004 |
0.006 |
0.005 |
0.011 |
-0.011 |
0.023 |
0.001 |
0.023 |
|
0.004 |
0.005 |
0.008 |
0.009 |
0.011 |
0.011 |
0.006 |
0.007 |
||
|
0.325 |
0.234 |
0.538 |
0.213 |
0.299 |
0.032 |
0.826 |
0.001 |
||
|
Share of dismissals |
0.003 |
0.005 |
-0.023 |
0.003 |
-0.018 |
-0.017 |
-0.006 |
0.004 |
|
|
0.004 |
0.005 |
0.007 |
0.007 |
0.011 |
0.012 |
0.005 |
0.007 |
||
|
0.542 |
0.336 |
0.001 |
0.644 |
0.111 |
0.147 |
0.251 |
0.549 |
||
|
Share of temporary contracts |
0.009 |
-0.024 |
0.044 |
-0.009 |
0.038 |
0.008 |
0.019 |
-0.022 |
|
|
0.010 |
0.009 |
0.018 |
0.012 |
0.029 |
0.027 |
0.014 |
0.013 |
||
|
0.368 |
0.009 |
0.013 |
0.469 |
0.179 |
0.766 |
0.172 |
0.088 |
||
|
Active Labour Market Policy expenditure on Training |
0.050 |
0.101 |
-0.205 |
0.146 |
-0.046 |
0.076 |
-0.196 |
0.126 |
|
|
0.035 |
0.034 |
0.067 |
0.036 |
0.107 |
0.059 |
0.048 |
0.044 |
||
|
0.151 |
0.003 |
0.002 |
0.000 |
0.669 |
0.196 |
0.000 |
0.004 |
||
|
Economy |
Regional GDP |
0.417 |
-0.033 |
0.523 |
-0.346 |
0.150 |
0.311 |
-0.090 |
-0.262 |
|
0.163 |
0.143 |
0.259 |
0.188 |
0.431 |
0.306 |
0.154 |
0.198 |
||
|
0.010 |
0.816 |
0.043 |
0.065 |
0.728 |
0.310 |
0.556 |
0.186 |
||
|
Number of patent |
0.079 |
0.287 |
0.081 |
0.313 |
0.125 |
0.307 |
0.176 |
0.210 |
|
|
applications |
0.052 |
0.070 |
0.062 |
0.105 |
0.100 |
0.115 |
0.042 |
0.082 |
|
|
|
0.124 |
0.000 |
0.188 |
0.003 |
0.213 |
0.008 |
0.000 |
0.010 |
|
Legend: b/se/p
Source: EU-LFS 2016