3072 en

EN EN ISSN: 2363-216X
EVIDENCE FROM CEDEFOP’S EUROPEAN SKILLS AND JOBS SURVEY

Skills, qualifications
and jobs in the EU:
the making
of a perfect match?
What has been the impact of the economic crisis
3072 EN – TI-RF-15-003-EN-N – doi:10.2801/606129

on skill mismatch? Is there a cost in getting the
unemployed quickly into any job? Why is skill
mismatch prevalent among the EU workforce?
To answer these and other timely questions on
skill mismatch, in spring 2014 Cedefop carried
out the European skills and jobs (ESJ) survey.
The findings caution that the prolonged
economic downturn is threatening the long-term
potential of the EU human resources. A greater
share of recent job finders has entered into jobs
that need lower qualifications and skills than
their own. The unemployed also run a greater
risk of misplacement into jobs of lower skill
intensity. More than one in five EU employees
has not developed skills since they started a job,
as over one third of EU jobs are characterised
by poor task complexity and lack of continued
learning. Closer stakeholder collaboration and
policy action is needed in the EU to generate not
only more skills but also, crucially, better jobs for
better-matched skills.
and jobs in the EU:
European Centre for the Development
of Vocational Training
Europe 123, 570 01 Thessaloniki (Pylea), GREECE
the making
PO Box 22427, 551 02 Thessaloniki, GREECE
Tel. +30 2310490111, Fax +30 2310490020, E-mail: info@cedefop.europa.eu
of a perfect match?
ISBN 978-92-896-1945-5
PREVIOUS CONTENTS NEXT
and jobs in the EU:
the making of a perfect
match?
Evidence from Cedefop’s
European skills and jobs survey
Cedefop Reference series 103

Luxembourg: Publications Office of the European Union, 2015
Please cite this publication as:

Cedefop (2015). Skills, qualifications and jobs in the EU:
the making of a perfect match? Evidence from Cedefop’s
European skills and jobs survey
Luxembourg: Publications Office.
Cedefop reference series; No 103. http://dx.doi.org/10.2801/606129
A great deal of additional information on the European Union

is available on the Internet.
It can be accessed through the Europa server (http://europa.eu).
Luxembourg:
Publications Office of the European Union, 2015
ISBN: 978-92-896-1945-5
ISSN: 2363-216X
doi:10.2801/606129
Copyright © European Centre for the Development

of Vocational Training (Cedefop), 2015
All rights reserved.
Designed by adam@artdirector.gr
Printed in the European Union
The European Centre for the Development

of Vocational Training (Cedefop) is the European Union's
reference centre for vocational education and training.
We provide information on and analyses of vocational
education and training systems, policies, research and practice.
Cedefop was established in 1975
by Council Regulation (EEC) No 337/75.

Tel. +30 2310490111, Fax +30 2310490020
E-mail: info@cedefop.europa.eu
www.cedefop.europa.eu
Joachim James Calleja, Director

Micheline Scheys, Chair of the Governing Board
Foreword
The central argument of Cedefop’s European skills and jobs (ESJ) survey is that, to get the most out of
its labour force, Europe must improve how it develops and uses skills.
Despite rising education attainment and high unemployment, one employer in three reports difficulty
filling vacancies. But many people work in jobs for which they are overqualified and in which their skills
are underused. Those graduating after 2008 are twice as likely to be overqualified for their first job as
those graduating in the 1990s.
The ESJ survey, the first to look at skill mismatch over time, reveals that most adult employees
constantly learn new things as their job tasks become more varied or change. Skill gaps also vary across
generations. However, despite rising demand for skills and qualifications a sizeable share of jobs in the
EU require only basic skills.
The policy implications of the survey’s findings are profound.
The survey confirms that work-based learning (WBL) can help people find not just a job, but a good
job. However, many students across the EU have limited access to WBL. The survey data strongly
support extending WBL across countries, occupations, sectors, qualification levels and fields of study.
Tackling skill mismatch means helping unemployed people into the right job, not just any job. This
requires a rethink of active labour market policies to help people find work. Making skills more visible
through validation leading to recognised qualifications may help. Career guidance and counselling for
both young people and adults also need strengthening.
The survey also confirms the need to invest in continuing vocational education and training (CVET)
and adult learning to keep up with changes at work. Training provision needs to adapt to take account
of workers’ different learning needs and the increasing importance of problem solving, teamwork and
communication skills at work.
The survey underlines that the supply of good skills depends on demand for them. European skills
policies need to consider how to use the skills we have to their best effect. A key to skill demand is good
jobs. Skill-intensive jobs with opportunities to learn new skills are needed to develop the labour force
and improve productivity and competitiveness. Employers have considerable scope over job design,
including the nature and variety of tasks as well as learning opportunities. However, a significant
proportion of adult employees are in jobs with limited scope for skill development.
There remains much to learn about skill mismatch and how to tackle it. The survey, with its many
insights, provides an excellent beginning.
Joachim James Calleja

Director
Acknowledgements
This report is the outcome of team effort and of own in-

house empirical analysis by Cedefop experts;
Konstantinos Pouliakas together with Giovanni Russo
drafted the report and were responsible for the survey
design, data collection and main empirical analysis.
Valuable inputs were received from Jasper van Loo,
Daniel Scheuregger and Steve Bainbridge. The assistance
of a dedicated network of experts, Ipsos MORI, the
Organisation for Economic Cooperation and Development
(OECD) and Eurofound is gratefully acknowledged.
Special thanks are also extended to other Cedefop
colleagues, who revised the report or contributed with their
ideas at different stages of its preparation.
Table of contents
Foreword 7
Acknowledgements 8
Executive summary 14
CHAPTER 1
Introduction 19
1.1. Skill mismatch and the European economic crisis 19
1.2. Why a Cedefop survey on skill mismatch 20
1.3. The European skills and jobs survey 23
1.4. Terminology used 25
1.5. Aim and structure of report 26
CHAPTER 2
Qualification and skill mismatches in the EU workforce 29
2.1. Qualifications supply and demand in the EU job market 29
2.1.1. Economic and social benefits of education and training 29
2.1.2. Qualification requirements in EU jobs 30
2.2. Qualification mismatches in the EU 33
2.2.1. Incidence of overqualification 33
2.2.2. Determinants of overqualification 35
2.2.3. Underqualification of EU workers 37
2.3. Skill mismatches in the EU 37
2.3.1. Skill mismatch relative to required level 37
2.3.2. Factors related to skill mismatch 39
2.3.3. Skill deficits undermining productivity potential 40
2.3.4. The imperfect relationship between qualification and skill mismatch 44
2.4. The cost of qualification and skill mismatch 45
CHAPTER 3
The legacy of the crisis: un(der)employment and skill mismatch 51
3.1. Graduate entry to the labour market 51
3.2. Return of the unemployed to work 52
3.3. Impact of the economic crisis 56
CHAPTER 4
Tackling skill mismatches among the young with high quality VET 59
4.1. Underskilling at labour market entry 59
4.2. Tackling initial skill mismatches with VET 60
4.3. Promoting all forms of WBL 62
4.4. Importance of key competences for VET 67
4.4.1. Fundamental key competences 67
4.4.2. Digital skills 70
4.4.3. Other competences 70
4.4.4. STEM skills 71
Skills, qualifications and jobs in the EU: the making of a perfect match?
10 Evidence from Cedefop’s European skills and jobs survey
CHAPTER 5
Matching better skills to better jobs 75
5.1. Shifting skill needs in European job markets 75
5.2. Evolution of adult worker skill mismatch 76
5.3. Skill development and adult learning in the EU 80
5.4. Bad-quality jobs can undermine lifelong learning 85
5.5. Helping labour market transitions tackle skill mismatch 90
CHAPTER 6
Conclusions and lessons for policy 94
6.1. No one-size-fits-all policies for skill mismatch 94
6.2. School-to-work transition and skill mismatch 95
6.3. Active labour market policies need greater focus 95
6.4. One-shot solutions will be short-lived 96
6.5. Employment stability leads to continued skill formation 96
6.6. Skill demand must be stimulated in the EU 96
6.7. Stimulating better EU labour market skill matches 97
List of abbeviations 99
References 100
ANNEXES
1. Details of the data collection 106
2. Empirical methodologies used to derive reported findings 114
Table of contents 11
List of boxes, figures and tables
Boxes
1. What is skill mismatch? 14
2. Novel elements and value added of the ESJ survey 24
3. Methodological note 27
4. Skill mismatch in specific occupations 39
5. Heterogeneity of skills across and within occupations and the importance of job design 89
6. Members of the expert group of the ESJ survey 107
Figures
1. Key facts from the ESJ survey 15
2. Incidence of WBL by field of study, EU-28, 2014 16
3. Difference in skill formation and skill needs of adult employees’ current jobs
by prior labour market status, 2014, EU-28 17
4. Skill deficits and labour productivity, 2014, EU-28 17
5. Share of EU jobs with stable or decelerating task complexity by economic sector,
2014, EU-28 18
6. Pillars of the ESJ survey 25
7. Level of qualification needed for jobs, adult employees, 2014, EU-28 31
8. Average incidence of qualification mismatch, adult employees, 2014, EU-28 34
9. Breakdown of overqualification by level of education, adult employees, 2014, EU-28 34
10. Overqualification by population group, % of adult employees, 2014, EU-28 35
11. Overqualification by field of study, adult employees with higher-level qualifications,
2014, EU-28 36
12. Underqualification by age group, adult employees, 2014, EU-28 37
13. Incidence of skill mismatch, adult employees, 2014, EU-28 38
14. ISCO groups most likely to say their skills are higher than required
in current job, adult employees, 2014, EU-28 40
15. Average skill deficit by EU Member State, adult employees, 2014, EU-28 41
16. Worker groups by type of skill mismatch, adult employees, 2014, EU-28 43
17. Impact of current skill mismatch on likelihood of skills obsolescence,
adult employees, 2014, EU-28 47
18. Impact of skill mismatch on job satisfaction, adult employees, 2014, EU-28 47
19. Skill deficits and labour productivity, 2014, EU-28 49
20. Overqualification of first job entrants by graduation cohort, 2014, EU-28 52
21. Difference in shares of underskilled employees at job entry
by prior labour market status, adult employees, 2014, EU-28 53
22. Raw difference in shares of employees with low-ranked skills relative to those required
for their job by past unemployment status and specific type of skill, 2014, EU-28 54
23. Lack of opportunity to attend job interviews by past labour market status and
EU Member State, adult employees (aged 24-65), 2014 54
24. Difference in skill formation and skill needs of adult employees’ current jobs
25. Difference in skill formation within jobs between previously employed and
unemployed workers, 2014, EU-28 56
26. Share of adult employees with few opportunities to find a job suitable
for their skills and qualifications by period of job entry, 2014, EU-28 57
27. Average share of adult employees (aged 24-65) underskilled at the start
of their job by period of job entry, 2014, EU-28 57
28. Raw difference in shares of underskilled employees at job entry
29. Raw difference in shares of employees with low-ranked skills relative
to those required for their job by age group and specific skill type, 2014, EU-28 60
30. Average incidence of WBL versus share of employees with VET as highest
qualification, adult employees with at least upper secondary qualifications, 2014, EU-28 61
31. Skill mismatch groups by VET status, medium-qualified adults, 2014, EU-28 62
32. Proportion of adult workers who completed study involving some workplace learning,
2014, EU28 63
33. Incidence of WBL by field of study, adult employees, 2014, EU-28 64
34. Work-based learning by industry, adult employees, 2014, EU-28 64
35. Work-based learning by occupation group, adult employees, 2014, EU-28 65
36. Transitions to first job by WBL and past labour market status, 2014, EU-28 65
37. Mean labour market outcomes and work-based learning status, 2014, EU-28 66
38. Level of fundamental skills required for the job, adult employees, 2014, EU-28 69
39. Highest level of literacy skills required by occupation, adult employees, 2014, EU-28 69
40. Highest level of numeracy skills required by occupation, adult employees, 2014, EU-28 69
41. Level of ICT skills required by occupation, adult employees, 2014, EU-28 70
42. Importance of transversal skills for job, adult employees, 2014, EU-28 70
43. Share of STEM graduates by industry, adult employees with tertiary level education,
2014, EU-28 72
44. Incidence of qualification and skill mismatch of adult employees
with tertiary level education by STEM status, 2014, EU-28 73
45. Share of adult employees with tertiary level education who think it very likely
that some of their skills will become outdated in the next five years, 2014, EU-28 73
46. Share of jobs with significant rise in the need to learn new things by industry,
adult employees (aged 24-65), 2014, EU-28 75
47. Share of jobs with rising variety of tasks by occupation, adult employees, 2014, EU-28 76
48. Likelihood of skills becoming outdated in next five years, EU-28, 2014 77
49. Likelihood of skills becoming outdated by industry, EU-28, 2014 77
50. Dynamic evolution of skill mismatch between start of current job and
present time, adult employees, EU-28 78
51. Skill mismatch transitions within jobs, adult workers, 2014, EU-28 79
52. Incidence in skills within jobs by occupation, adult workers, 2014, EU-28 81
53. Method of improving/acquiring new skills in job, adult workers, 2014, EU-28 82
54. Types of training completed in last 12 months, adult workers, 2014, EU-28 82
55. Incidence of training completed in last 12 months by type, adult workers, 2014, EU-28 83
56. Reasons for participating in training, adult workers, 2014, EU-28 83
57. Public versus private financing of training, adult employees, 2014, EU-28 84
58. Work place changes supported with training activities offered by employer
by occupation, adult employees, 2014, EU-28 84
59. Share of EU jobs with stable or decelerating task complexity by economic sector,
2014, EU-28 86
60. Share of jobs requiring basic or no ICT skills, adult employees, 2014, EU-28 87
61. Variance of skills importance index within occupational groups, adult workers,
2014, EU-28 90
Table of contents 13
62. Persistence of overskilling between and within jobs, adult workers, 2014, EU-28 91
63. Skill mismatch by type of occupational transition, adult workers, 2014, EU-28 93
64. Skill deficits and lifelong learning, adult workers, 2013-14, EU-28 94
65. Cedefop ESJ and OECD PIACC: employment contract 109
66. Cedefop ESJ and OECD PIACC: employer paid for training 110
67. Cedefop ESJ and Eurofound EWCS: employment contract 110
68. Cedefop ESJ and Eurofound EWCS: overall skill mismatch 110
69. Sample profile: occupation 111
70. Sample profile: industry 112
71. Sample profile: gender and age 112
72. Sample profile: education 112
73. Schematic illustration of estimation methodology of determinants of
skill mismatch transitions in ESJ survey 117
Tables
1. Definitions of skill mismatch 26
2. Mean effect of education level on labour market outcomes, adult employees,
2014, EU-28 29
3. Incidence of adult employees by level of education needed for their job, 2014, EU-28 32
4. Examples of two-digit occupations with highest and lowest rates of overqualification,
5. Worker groups by type of skill mismatch/deficit, adult employees, 2014, EU-28 42
6. Combinations of qualification and skill mismatch 44
7. Qualification versus skill mismatch, adult employees, 2014, EU-28 45
8. Impact of qualification-skill mismatch on labour market outcomes
of tertiary education graduates, adult employees, 2014, EU-28 46
9. Impact of qualification-skill mismatch on gross hourly earnings, adult employees,
2014, EU-28 48
10. Skill mismatch of adult employees by VET status, 2014, EU-28 62
11. Bundles of skills used in ESJ survey 68
12. Evolution of skill mismatch within jobs in relation to skill development and
job complexity, adult employees, 2014, EU-28 80
13. Skill mismatch transitions: previous job and start of current job,
14. Distribution of adult workforce according to skill mismatch transitions between
previous job – start of current job – present job, adult employees, EU-28, 2014 92
15. Overview of questionnaire structure 108
16. Unweighted and weighted national sample sizes in ESJ survey 113
Executive summary
European skills and jobs The ESJ survey, carried out in 2014 in all EU-
survey: identity and rationale

28 Member States, is a state-of-the-art survey
instrument that collected information on the
match of the skills of about 49 000 EU adult
The most common understanding of skill
workers (aged 24 to 65) with the skill needs of
mismatch in the European Union (EU) is one of
their jobs. The data were collected using quota
employers unable to fill vacancies despite high
sampling, following extensive testing, including a
unemployment. However, skill mismatch – a term
pilot survey in January 2014. The fieldwork
not always clearly understood (Box 1) – is more
(mixed mode, online plus telephone interviewing)
pervasive. It is not only a problem for those
was conducted from March to June 2014.
looking for a job, but affects most of the workforce
The new survey provides an innovative
and can hamper both economic productivity and
insight into the dynamics of qualification and skill
individual potential.
mismatch in the EU, focusing on the interplay
Cedefop’s European skills and jobs (ESJ)
between changes in the (cognitive and non-
survey provides insights into the match between
cognitive) skills of employees in their jobs as well
the skills and jobs of adult workers in European
as the changing skill needs and complexities of
countries. The analysis aims to inform a diverse
their jobs. People continually develop their skills,
audience of public vocational education and
while job complexity and skill intensity may
training (VET) and labour market policy-makers
change significantly. Unlike previous studies
and the social partners, and to provide important
focusing on a given point in time, the ESJ survey
lessons for policy-making geared towards
is the first to look at skill mismatch over time. It
tackling the phenomenon of skill mismatch in
also distinguished between qualification level and
European job markets.
skills needed for the job. A job requiring a
particular qualification level, medium or high, may
Box 1. What is skill mismatch?
not be skill-intensive. People with a similar
qualification level may have substantially different
skill levels.
The survey also helps establish the evidence
Employers unable to find the right talent, despite
base to inform policies on initial (formal and work-

offering competitive wages, face skill shortages.
based learning (WBL)) and continuing vocational

Skill gaps arise where the skills required are
education and training (CVET) (non-formal and

unavailable in the workforce, for example, due to
informal training) and on workplace design and

technological advance. Over or underqualification
labour market mobility for mitigating skill

is where individuals take jobs that do not match
mismatch.
their qualifications. People are over or under-
skilled where, whatever their qualification level,
their skills do not match their job.
Figure 1. Key facts from the ESJ survey
44% 27%
Talent unnourished Talent in dead end
Employees whose skills are lower than needed Employees whose skills are higher than
to achieve full productivity in the job and have needed to do the job today and have limited
potential to grow in the job potential to grow in the job
22% 33%
Skill stagnancy Poor jobs
Employees who have not developed their skills Jobs where adult employees need
since starting their job no or only basic information and
communications technology (ICT) skills
25% 21%
Skill underuse Skill gaps
Highly qualified employees Employees whose skills were lower
who are overqualified for their job than needed at the start of their first job
Source: Cedefop ESJ survey.

Skill mismatch during transition There is a strong case for expanding WBL to
to work align training more closely to labour market
needs. Around 40% of adult employees have
WBL can smooth the transition into work but it is completed education or training involving some
not used enough. WBL, but this varies considerably across
The ESJ survey indicates that people whose countries and fields of study. Only about 25% of
studies involved WBL are more likely to go younger (24 to 34 year-old) graduates in
directly from education to their first job and into humanities, languages and arts, economics,
more skill-intensive jobs. Given that skill business and law have participated in WBL,
mismatch is persistent, policies to improve skill compared to 67% of graduates from medicine
matches before or at the start of the working life, and health-related sciences.
such as WBL, can be highly effective.
Figure 2. Incidence of WBL by field of study, EU-28, 2014
Humanities, languages and arts

Other social sciences
Economics, business, law and finance
Maths and stats
Natural sciences
Engineering sciences
Computing sciences
Agriculture and veterinary sciences
Teacher training and education science
Other field
Security, transport or personal services
Medicine and health-related
0% 10% 20% 30% 40% 50% 60% 70%
Active labour market policies programmes end when an unemployed person

and validating learning: finds a job. To ensure the best possible skill
tackling skill mismatches due to

match, the unemployed should be able to
complete their training programmes either before
unemployment starting a job or while working. Increasing WBL
for unemployed adults could also improve
Cedefop’s ESJ survey confirms that unemployed alignment and job prospects and give enterprises
people returning to work are also more likely to a source of new recruits.
enter less skill-intensive jobs that may not Lack of formal certification of informal skills
continuously develop their skills. Skill mismatch may hinder career progression, and job prospects
may, therefore, reinforce the ‘scarring’ effect of (Cedefop, 2015b). Consequently, strategies to
unemployment, where someone out of work for a upgrade unemployed people’s skills may be more
while is more likely to become unemployed again effective if the skills acquired were made more
in the future. It is a ‘hidden’ cost of getting visible by validating learning and certifying skills
unemployed people back to any job. with a recognised qualification (or part of one)
The key to addressing unemployment and linked to the national and European qualifications
skill deficits is getting people into the right job, not frameworks.
just any job. Many active labour market training
Figure 3. Difference in skill formation and skill needs of adult employees’ current jobs by prior
labour market status, 2014, EU-28
Previously LTU Previously employed
100%
80%
60%
40%
20%
0%
Task complexity Skill intensity Changing task complexity Improvement in skills
NB: Percentage of workers with high-ranked scores on each index (i.e. at the top of each scale).
Skill mismatch at work Skill gaps at work vary across generations,

highlighting the challenges of tailoring CVET pro-
Many EU adult workers have skill deficits but vision to the needs of different groups of learners.
not all have a chance to improve Compared to their older colleagues, younger em-
ESJ survey data show that, since they started ployees are more likely to have higher skill gaps
their job, the need to learn new things and the in technical skills (specialised knowledge to per-
variety of their job tasks has significantly form job tasks) and soft skills (communication,
increased for about 53% of adult employees in teamwork, customer-handling and problem solv-
the EU. ing). Older workers are more likely to have skill
Overall, around 26% of EU adult employees gaps in foreign languages, abilities to learn and
have significant skill deficits (their skills are much apply new methods and techniques (including
lower compared to those an average worker new technology) and in digital skills.
needs to be fully proficient in their job) leaving The dynamics of skill mismatch highlight the
scope to improve skills and productivity. But even importance of investment in CVET and adult
though these workers could develop, not learning. Social partners have a key role to play.
everyone has the chance to do so, as 27% are in They are best placed to encourage learning at the
‘dead-end’ jobs. In such jobs, employees have workplace and to arrange work organisation and
higher skills than they need to do their job and working time to ease participation in CVET and
only limited potential to develop. adult learning.
Figure 4. Skill deficits and labour productivity, 2014, EU-28
80
70 LU
IE
FR BE
60 DK
Labour productivity
50 SE ES
AT
IT FI
NL DE
40 UK EL
SI
HR LT
PT
SK EE CZ
30
PL HU LV
20 RO
BG
10
0
10 12 14 16 18 20 22 24
Skill deficit
Source: ESJ survey; World Bank.
40%
0% 10% 20% 30% 40% 50% 60% 70%

30%
100%
20%
80%
Good jobs are crucial to developing good The ESJ survey also confirmed the close link
60% 10%
skills between job stability and people working in jobs

The
20% survey finds that good jobs are needed to where a higher level of skills is needed. Job
40%
develop good skills. Skill-intensive jobs with stability enables workers to cope with complex
0%
UK
opportunities to acquire skills continuously are a workplace changes that place higher demands on
sign of a healthy labour market. Europe needs their skills.task complexity
0%
more jobs that can fully use and develop the skills Skills are Europe’s competitive advantage.
Task complexity Skill intensity Changing Improvement in skills
40%
of its
80 workforce. High-quality jobs can reduce skill mismatch,
70Skill demand is low and stagnant for many ensure skill development and improve
European workers. Cedefop’s survey found that competitiveness. Organisations have a significant
LU
IE
40% 50 of adult employees only need basic literacy degree of control in determining the skill content
60
FR BE 30%
DK
skills to do their job, 58% need only basic of their jobs and so couldSIuse better the skills
Labour productivity
SE ES
AT
FI
numeracy and 33% of jobs in the EU require only available PTby reorganising work EEto CZadapt and
NL IT
DE
40 UK
basic ICT or no ICT skills at all. Over a third of deployPL workers’

EL
HU skills as technology and working

HR LT
SK
20%
30
jobs in sectors such as hotels and restaurants, methods change. Employment stability leads to
LV
20 RO
transport and wholesale and retail have stagnant skill formation and stable careers maximise the
BG
skill 0needs. benefits of skill investments.

10
10%
10 12 14 16 18 20 22 24
Figure 5. Share of EU jobs with stable or decelerating task complexity by economic sector,
Skill deficit
2014, EU-28
0%
IE
% no change or reduction in variety of job tasks since start of job

Financial, insurance or real estate services T
Professional, scientific or technical services
Manufacturing or engineering
Contract
Supply, management or treatment of water
Services relating to education or health
Information technology or communication
Supply of gas or electricity, mining
Administration and support services
Previous labour
market status
Construction or building Not
Agriculture, horticulture, forestry or fishing
Other
Social and personal services
Gender
Retail, sales, shop work or wholesale

Cultural industries (arts, entertainment)
Transportation or storage
Age group
Accommodation, catering or food services

0% 10% 20% 30% 40% 50%
NB: Percentage of adult employees who reported no change or reduction in the variety of job tasks since they started their current job.
Security, tra
Introduction
CHAPTER 1
1.1. Skill mismatch and the part-time (European Commission, 2014a). This
European economic crisis has heightened fears about repeated spells of
short-term unemployment and underemployment
The 2008 global economic recession had a of individuals taking up precarious employment,
significant impact on the wider European with associated scarring effects (European
economy, markedly affecting the employment and Commission, 2014a; Mavromaras et al., 2015a).
social situation of EU Member States. The slow Overall, deteriorating labour market
economic recovery during 2008-14 resulted in prospects in recent years, which have particularly
only a fraction of the 6.7 million jobs lost during affected younger and lower-skilled individuals,
the recession being recovered by mid-2014 have highlighted the need for European policy-
(European Commission, 2014a). The prolonged makers to continue investing in initial and adult
economic slump, which has affected some EU (mainly work-based) learning and in active labour
Member States to a greater extent than others, market policies, to overcome skill mismatches
has led to concerns that much of what was and ensure effective transition back into
originally cyclical unemployment is in danger of employment (Cedefop, 2015c). It has also called
becoming structural (Draghi, 2014; Institute for for making better and targeted use of VET and
Public Policy Research, 2014). employment policies that encourage investment
Skill imbalances, in particular the rising in training and the better matching of job seekers
demand for higher-skilled workers and the with available jobs across Europe.
collapse in demand for lower-skilled or younger In addition to the deep stigma left by the 2008
individuals, have underpinned rising mismatches recession on the European landscape, marked
in EU labour markets (European Commission, long-term challenges also lie ahead of the conti-
2012a; ECB, 2012). The doubling of the rate of nent, including adverse demographic pressures,
long-term unemployment between 2008 and technological innovations and the need to tackle
2013 at EU level has led to concerns about the rising social inequalities and tensions. This situa-
potential skills atrophy of the long-term tion calls for a general rethink of European ‘skills
unemployed (European Commission, 2014a). policies’. Policy-makers have increasingly recog-
Further, during the recession there was a stark nised the need to understand better the root
fall in the numbers of people hired in manual causes of skill mismatch in order to support evi-
occupations and in the manufacturing and dence-based policy-making, as highlighted in the
construction sectors, while, according to the European Commission’s communication New
European vacancy monitor (European skills for new jobs (European Commission, 2008)
Commission, 2014d), most occupations that grew and in one of the flagship initiatives of the Europe
in the post-crisis era required higher levels of skill. 2020 strategy, the Agenda for new skills and jobs
The sustained recession is undermining the (European Commission, 2010). The EU commu-
long-term labour market potential of graduates’ nication Towards a job-rich recovery (European
skills and is distorting incentives for further skill Commission, 2012b) has also made recommen-
development. Graduating in a recession has dations to Member States emphasising the need
persistent negative effects on career prospects to boost job creation in Europe.
and rates of return on education, because many Skills hold the key to higher productivity
individuals accept jobs that do not match or fully growth for Europe and are crucial for overcoming
use their qualifications and skills, resulting in skill declining living standards due to the shrinking
underutilisation (Oreopoulos et al., 2012; Liu et future workforce (European Commission, 2013b).
al., 2012). Also, in the aftermath of the economic Even though Europe today has one of its most
crisis a large proportion of the new jobs created highly qualified cohorts of young graduates
in the European economy were temporary or entering the labour market, high skill mismatches
undermine economic competitiveness and Some of these questions, which have

growth, sustain unemployment and high social underpinned the motivation and design of the
insurance costs and affect social inclusion questionnaire of the new Cedefop survey on skills
(European Commission, 2013a; European and jobs are covered below.
Commission, 2014a). Continued skills
development, which is not met by a
corresponding increase in skill demands, is
Is skill mismatch an issue of generalised
unlikely to mitigate skill mismatches in the EU
concern for all EU Member States? If so, what
labour force on its own. EU countries
types of skill mismatch are of greater concern?
characterised by higher levels of competitiveness
and dynamic labour markets generally do not only Much of the previous academic evidence on
invest in education and training (both initial and skill mismatch has relied on information available
continuing), but also in skills maintenance in the from a limited subset of EU countries, including
workplace (European Commission, 2014a). In those that benefit from the availability of own
this respect, the European Commission (2013a) extensive national data sources (such as the
and the Institute for Public Policy Research Netherlands and the UK). However, estimates of
(2014) have shown that the loss of particular qualification mismatch as derived by Cedefop
types of job since the recession has also resulted (European Commission, 2013a) highlighted that
in qualified individuals having to take up jobs that EU countries may have a differential propensity
fail either to match or fully utilise their skills. to alternative forms of skill mismatch. For
instance, while southern Europe Member States
Why a Cedefop survey on

(Greece, Spain, Cyprus) along with Ireland and
1.2. the UK suffer from high rates of overqualification
skill mismatch in their job markets, so higher-educated
individuals are employed in medium- or low-
Despite the critical nature of skill mismatch for skilled jobs, the Baltic states and several
policy-makers, the existing statistical countries in eastern and central Europe (Bulgaria,
infrastructure and evidence at a pan-European the Czech Republic, Hungary, Austria) have been
level was relatively weak and biased towards prone to skill shortages and skill gaps in their
specific EU countries until a few years ago. It was workforces. Even within countries, skill
also largely biased towards the view of employers mismatches may affect various population groups
on skill shortages and on the transition from differently: the young may be more likely to be
school to work. underskilled while the older suffer mostly from
To address these deficiencies, Cedefop skill obsolescence. There may also be different
initiated in 2008 an ambitious research forms of skill mismatch across individuals
programme that has sought to investigate in belonging to the same group or cluster; some
depth the nature, incidence and causes of skill young individuals may be underskilled and others
mismatch in European labour markets (Cedefop, overskilled, even if they all have tertiary level
2009; 2010a). Cedefop’s research has paid much education.
attention to the education, training and A key source of information on skill mismatch
development strategies that can help mitigate skill at European level is Eurofound’s European
mismatch for both individuals and enterprises working conditions survey (EWCS). Never-
(Cedefop, 2012b; 2015a). It has also investigated theless, the focus of the EWCS on the state of
the causes of skill mismatch among particular working conditions in the EU implies that it cannot
vulnerable groups of the population, such as fully capture critical contextual information
older or ‘silver’ workers and people with a migrant necessary for fully understanding the
background or ethnic minorities (Cedefop, 2010b; phenomenon of skill mismatch, including the skill
2011). Several important issues and questions intensity of jobs or the dynamics of the skill
arose in this research which could not be mismatch process. Further, the main measure of
answered in a convincing manner with the underskilling used by the EWCS, which focuses
existing European data infrastructure. on individuals’ need for training to cope with the
CHAPTER 1
Introduction 21
duties in their job, tends to confound the evidence that could allow for careful evaluation of
measurement of the primary concern (the need the impact of the economic crisis on skill
for further skill development on behalf of mismatch patterns in Europe.
employees) with the overall culture and incidence
of CVET in different EU Member States.
More recently, the OECD survey of adult skills Why do individuals accept jobs that are not a
(PIAAC) marked a significant improvement in the good match with their qualifications and skills?
collection of data, allowing for the international Is the ‘overeducation wage penalty’ a genuine
comparability of adult skills, yet only 17 EU waste of productivity?
countries participated in the first round of the
survey. Further, attempts to measure objectively Although the academic literature has
different types of skill mismatch (Pellizari and highlighted the significant costs associated with
Fichen, 2013) have been criticised (Allen et al., skill mismatch for both individuals and societies,
2014), with some arguing that subjective recent studies that have exploited longitudinal
perceptions of workers can also be reliable data sets have called into question the genuine
indicators of job skills required in relation to own cost of education mismatch (Sloane, 2014). Past
skills (Green, 2013). research based on cross-sectional evidence
These issues justified collecting new data on could not refute the claim that overqualification
skills and skill mismatch that allow for may be an outcome of conscious job choices by
measurement and identification of a number of individuals that reflect their preferences for
different forms of skill mismatch across a wide amenities (including a high salary, good working
range of adult employees in European labour conditions, a positive work-life balance,
markets. challenging work, a pleasant area of work) that
are extraneous to the suitability of their skills with
the job’s skill demands. In this case, an individual
who is mismatched in terms of qualifications and
What has been the impact of the economic
skills may still enjoy a high level of job

downturn on skill mismatch?
satisfaction. The empirical snapshot pointing to

During the recent economic downturn the high levels of overqualification from cross-
demand for individuals with high qualifications sectional data sets may simply reflect the optimal
and in high-skilled jobs rose at the expense of the workings of a free labour market and therefore
lower-skilled. Despite the marked increase in the not merit policy intervention. In contrast,
supply of unemployed labour in depressed job government intervention and investment of public
markets, four in 10 EU employers still report resources in skill matching instruments is
difficulties in finding workers with the right skills warranted if skill mismatch is a ‘forced’ or an
(Cedefop, 2015a). The non-trivial share of firms inefficient outcome that arises because of market
in the EU faced with talent shortages, as reported failures: inadequate supply of suitable jobs in the
in employer surveys, appears to be at odds with economy (vacancy externalities); job-finding
a wealth of studies on growing overqualification externalities (a shrinking probability of recruitment
among employees in several advanced for medium- or lower-skilled individuals due to
economies across the globe (McGuinness, 2006; crowing out from the higher-qualified); imperfect
World Economic Forum, 2014). It was also in guidance and labour market intelligence; social
contrast to the observed trend in the European constraints (such as lack of adequate supply of
labour market, which saw an increasing share of child care facilities); and monopsonistic practices
higher-educated workers finding employment in prevailing in the job market.
positions that are not commensurate with their All this implies that there was a need for
qualifications and skills (European Commission, collection of data on skill mismatch that would
2012a; ILO, 2014). permit sophisticated empirical analysis that could
Contradicting findings and different account for the diverse preferences and
interpretations called for the collection of up-to- constraints individuals face when selecting their
date and comprehensive European-wide job.
High unemployment is public enemy number

Are VET and WBL effective in mitigating skill one for the EU, along with the financial/debt crisis.
gaps and in improving the match of young The economic crisis has resulted in very high
individuals’ skills with their job demands? levels of (youth) unemployment that, despite
evidence of rising levels of economic activity in
The recent economic crisis and the most EU Member States, have remained
historically high levels of youth unemployment persistently stuck above 10%. Many observers
experienced in several EU Member States have and policy-makers have noted that, in the
spurred the implementation of a series of aftermath of the economic crisis, getting people
important EU level initiatives, including the youth quickly back to work is a priority. However, others
employment initiative and youth guarantees as have argued that a rushed return of the
part of it. The European alliance for unemployed to the labour market may result in
apprenticeships and the quality framework for them getting trapped in unproductive and low-paid
traineeships were developed in response to jobs that will only perpetuate a low-pay/no-pay
widespread evidence of the benefits of WBL and cycle. The literature has also shown that there are
of apprenticeship schemes in aiding faster scarring effects associated with unemployment
integration and adaptation of young persons’ (and low pay) on future employment prospects
skills to the labour market needs. (Stewart, 2007). By contrast, there is evidence
The 2009 ad hoc module of the European that the provision of extended social support and
labour force survey (EULFS) provided a data set of unemployment benefits, that can provide
that enabled some analysis of the labour market adequate time for unemployed or inactive
prospects of young adults in the EU according to individuals to search for longer, may enable them
whether their academic orientation was to find jobs that constitute a better match for their
vocational or academic (Cedefop, 2012a). skills and preferences (Tatsiramos, 2014;
However, the lack of quantitative data available European Commission, 2014a).
from Eurostat and other international sources has In general there is limited evidence on the
generally prevented researchers from painting a extent to which skill mismatch, particularly in the
comprehensive picture of the impact of VET and form of skills underutilisation, may have long-term
WBL on different forms of skill mismatch. Further, detrimental consequences similar to the scarring
previous evidence has focused mostly on the effects associated with unemployment; a few
transition of young adults to the job market and studies confirm that this is the case in Australia
on employment rates, without paying much (Mavromaras et. al., 2015a) and in some
attention to the type of jobs in which they find European countries (Baert et al., 2012; Longhi,
employment or to their subsequent skill 2015). For this reason, data that allow exploration
accumulation. of the dynamics between unemployment (and its
Better data were necessary to shed more interlinkage with previous skill mismatch) and
light on the impact of WBL and of VET on the skill current labour market outcomes, including the
mismatch status of individuals making the transition to a well-matched job after an
transition to their first job. The ESJ makes it unemployment episode, were needed at pan-
possible to validate the previously disparate European level.
(mainly qualitative) evidence that has highlighted
the benefits of work-oriented education and
training programmes in terms of labour market
How does skill mismatch evolve in jobs? What
success.
actions do individuals take, formal, non-formal
and informal, to develop their skills in their
jobs? What is the role of the demand for skills
Is finding any job for the unemployed enough? and of job tasks in mitigating skill mismatch?
Is the cost of skill mismatch lower than the cost
of unemployment? Previous literature has tended to adopt static
measures of skill mismatch, focused on
measurement of individuals’ correspondence
CHAPTER 1
Introduction 23
between their skills and the skill demands of their by adopting policy measures that can liberalise
jobs at the time of the survey. This has inhibited these markets, such as the reduction of
full understanding of the dynamics of the employment protection legislation or the removal
matching process between skills and labour of other barriers to mobility (including recognition
market needs: it is widely acknowledged that of qualifications across borders and
people continually develop and adapt their skills establishment of professional rights in different
throughout their careers, while job characteristics occupations and countries). From an EU
also undergo significant changes over time in perspective, the promotion of labour mobility is
terms of complexity and skill intensity of tasks. also a key principle of EU level mobility
Despite this apparent dynamic character of the programmes, such as ‘Your first EURES job’ and
skill matching process in labour markets, previous ERASMUS+. Several researchers have stressed
scans of the incidence of skill mismatch would that skill mismatch affecting individuals can only
tend to reveal that in more than six in 10 jobs in be a temporary phenomenon, as long as they
the EU the skills of individuals are inherently have adequate career or occupational mobility to
‘matched’. Such a statistic could lead to move to better matched jobs in different
misleading policy conclusions, as it erroneously geographic labour markets (Sicherman, 1991;
gave an impression of stability in skill mismatch Robst, 1995).
within jobs. While in flexible and dynamic labour markets
Most surveys on adult education and training it is important to enable efficient allocation of
have also focused on the participation of productive human capital to its best use, several
individuals in training courses, without explicit commentators have cautioned that excessive job
attention being given to the effectiveness of the mobility can also be conducive to skill mismatch
training in terms of improving the fit of the if it is associated with a significant loss of human
acquired skills with the skill needs of jobs. Further, capital, or with non-stable and short-term
inadequate information has been collected in employment relationships that prompt job
conventional European surveys with respect to churning and turnover (European Commission,
the non-formal and informal learning that job 2013a; Panos et al., 2014).
seekers and employees may undertake to Despite the high policy interest on the issue
acquire the necessary skills and experience of labour market mobility, existing European data
required to perform well in a job. sources did not allow for in-depth investigation of
Despite the obvious need for information on the relationship between both occupational and
the continuous development of skills and of jobs geographic mobility and skill mismatch. To tackle
in European labour markets, the availability of this deficit, the Cedefop ESJ survey included two
longitudinal evidence on skill mismatch is limited, questions to capture the extent to which adult
with some EU studies focusing on just a few workers in the EU have engaged in occupational
countries (such as Germany, the Netherlands, the or geographic mobility prior to the start of their
UK) and greater advances made in Australia jobs.
(Mavromaras et al. 2007; 2010; 2012) and the US
The European skills and

(Oreopoulos et al., 2012; Clark et al., 2014). The
new Cedefop survey was developed in response 1.3.
to the scarcity of information on the dynamic jobs survey
evolution of skill mismatch among adults in their
jobs. The Cedefop ESJ survey is a state-of-the-art
survey of adult employees (aged 24 to 65) in all
EU-28 Member States, collecting information on
the match between their skills and the skill needs
Is labour market mobility a silver bullet for
of their jobs. The aim of the survey is to help
mitigating skill mismatch?
inform the development of European policies on
One of the traditional policy remedies initial and continuing vocational education and
suggested to tackle high rates of skill mismatch training (IVET and CVET) and employment
in market economies is to promote labour mobility policies. To do so, it seeks to understand how
individuals’ qualifications and skills are matched The survey asked respondents a series of
(or not) to the changing skill demands and questions designed to assess the extent to which
complexities of their jobs. The survey also looks a selection of their skills are seen to be important
at the extent to which employees’ skills are in their jobs and the extent to which their skills (as
developed and used in their workplace (Figure 6). a whole, but also for 11 specific skills, including
The survey, carried out by the survey foundation, technical and generic skills) are at the
company Ipsos MORI and its network partners in level needed to do their job. The key contribution
the EU-28 Member States between 7 March and of the new survey is that it takes a longitudinal
26 June 2014, relied on quota sampling. In total, perspective, with some of these questions asked
48 676 respondents from different demographic several times, referring to different time periods,
and socioeconomic groups took part either by enabling dynamic analysis of skill mismatch
telephone (9 154 employees) or online (39 522 among EU employees. The survey offers the first
employees). A mixed methodology approach comparable evidence of the dynamic evolution of
ensured that data collected provided a employee skill mismatch across all EU-28
representative sample of the 24 to 65 working Member States.
age population in each of the 28 countries. The survey also collects information on: the
micro/macro-economic motives and constraints
Box 2. Novel elements and value added of the that may influence individuals’ job choice;
ESJ survey determinants of skill accumulation in jobs,
focusing on CVET; the experience of WBL as part
of initial education and training; whether
individuals undertook occupational and/or
• The ESJ survey provides the first comparable
geographic mobility prior to accepting their
evidence on skill mismatch across all EU-28
current job; the changing nature of workers’ tasks
Member States from a longitudinal perspective.
over time.
• It provides measures of qualification and skill
The survey questionnaire was prepared by
mismatch in total for each EU Member State, but
Cedefop experts in collaboration with a network
also for 11 specific cognitive and non-cognitive
of leading experts in skills and skill mismatch, the
skills.
OECD and Eurofound. The questionnaire was
• It collects information on VET and WBL and
translated into the national languages of the EU
examines their effect on skill mismatch.
countries using a strict translation protocol,
• It draws critical contextual information on the
managed by Ipsos MORI. In advance of the
motives, constraints and preferences affecting
survey, extensive cognitive and pilot tests
the job choice of individuals.
examined the content and validity of the survey
• It examines the impact of job mobility
instrument. Further details regarding the data
(occupational, geographic) on skill mismatch.
collection process and the technical
• It looks at the determinants of individuals’ skill
specifications of the survey are provided in the
accumulation in jobs (formal/non-formal/
Annex 1.
informal training).
• It allows comprehensive analysis of the evolution
of skill mismatch by measuring the impact of
dynamic changes in skills and job tasks within
workplaces.
CHAPTER 1
Introduction 25
Figure 6. Pillars of the ESJ survey
25%
Previous labour market status Start of current job
20%
(WBL, VET, student, employed, (circumstances, motives,
short- or long-term unemployed) job mobility) 15%
10%
5%
0%
24-29
Current labour market status
(occupation, industry,
job characteristics)
60%
50%
40%
Skill development Skill needs
(VET, motives and financing (foundation, technical and generic skills; 30%
of training, methods of skill improvement) tasks in job; changing tasks, technologies /
work methods) 20%
10%
0%
AT UK E
Dynamic change in skill mismatch Labour market outcomes Two digit occupa
(previous job start of current job (wages, job insecurity, expected
Elementary o
start of job current post) skill obsolescence, job satisfaction)
ISCO 91:
Cleaners and helpers
ISCO 92:
Agricutural labourer
1.4. Terminology used between aggregate labour

(ISCEDsupply
and demand)
ISCO 93:
or qualitative (differences between individual’s

Low (ISCED 0-2) Medium 3-4)
High (ISCED 5-6) Labourer in miningco
Skill mismatch is often used to describe various skills and job requirements) (Sattinger, 2012).
70%
ISCO 94:
forms
60% of mismatch between supply and demand This report adopts the following terminology of
Food preparation ass
in the labour market (Table 1). Skill mismatch – a skill mismatch concepts, in line with the
definitions presented by Cedefop (2010a) and in
ISCO 95:
difference between the available skills and
accordance with literature convention.
Street and related sa
qualifications and those required by the labour

50% service workers
market
40% – can be quantitative (differences ISCO 96:
Other elementary wo
30%
25
Table 1. Definitions of skill mismatch
Mismatch A situation where industries, occupations, locations or groups with different

levels of education/skill differ in the ratio of unemployment to vacancies.
Skill mismatch A situation where there is a discrepancy between the qualifications and
skills that individuals possess and those needed by the labour market; a
pillar of labour market mismatch.
Overqualification A situation in which an individual has a higher qualification than his/her
current job requires.
Underqualification A situation in which an individual has a lower qualification than his/her
current job requires.
Overskilling (or skill underutilisation) A situation in which an individual is not able to utilise fully his or her skills
and abilities in his/her current job.
Underskilling (or skill gap) A situation in which an individual lacks the skills and abilities necessary to
perform the current job adequately.
Skill shortage A situation where demand for a particular type of skill exceeds the supply
of that skill at the prevailing rate of pay.
Skill deficit A situation where the skills and abilities of individuals are lower than a
given benchmark level of skills.
(Economic) skills obsolescence A situation in which skills previously utilised in a job are no longer required
or have diminished in importance.
Source: Cedefop, 2010a; Cedefop, 2012b; Lazear and Spetlzer, 2012; Green, 2013.
Skill mismatch often manifests at the which arise when the skills available in the
recruitment stage. It is most notable in the form workforce are insufficient to meet the demands of
of higher unemployment rates coinciding with their jobs, for instance, due to technological
vacancy creation that cannot be met explicitly due advances.
to skill deficiencies in the available workforce. Quintini (2011) has shown that an individual’s
When employers have difficulties filling vacancies qualifications mismatch and his/her skill
with the right talent, despite offering competitive mismatch over the career is imperfectly
wages, they face skill shortages. correlated. A sizeable share of employees, who
However, mismatches in skill also arise within may be matched in terms of their qualifications
the workplace and during the tenure of when entering their job, may feel that their skills
employees, as their qualifications and skills may are underutilised during job tenure or may
diverge from the formal education and skill become underskilled over time as the demands
requirements of their jobs over time. Some of their job tasks accelerate due to technological
individuals may be hired in jobs with qualification or organisational progress.
requirements that are different from the ones that
the individuals actually possess, breeding over-
or underqualification. The education qualifications 1.5. Aim and structure of report
of employees may be insufficiently matched to
those needed either to be hired for the job or to The remainder of this report outlines key findings
be able to perform the job to a satisfactory level. and empirical analysis based on Cedefop’s new
People are over- or underskilled where, ESJ survey. Chapter 2 examines the evidence of
whatever their qualification level, their skills are the new survey with regards to the incidence and
not matched to their job. Underskilling is cost of different forms of skill mismatch for
sometimes treated as synonymous to skill gaps, European citizens and economies.
CHAPTER 1
Introduction 27
Chapter 3 devotes attention to the recent employees improve their skills in their jobs. The
economic crisis, in terms of its impact on the skill analysis identifies population groups that are at
mismatch of recent cohorts of graduates and job risk of job insecurity and skill obsolescence, due
finders relative to the pre-crisis era. The analysis to the fact that they experience stagnation in their
focuses on difficulties of reintegrating the skill development. The positive impact of
unemployed into work, most notably the extent to continuing vocational training (CVT) and adult
which their labour market entry is hampered by learning in tackling skill mismatch is outlined.
skill gaps. The chapter also considers the demand side,
To inform policies to combat the high youth investigating the degree to which different skills
unemployment in Europe, Chapter 4 investigates are important for performing jobs in different EU
the magnitude of skill gaps of European citizens job markets and the implications for skill
during their first transition to the labour market mismatch as implied by differences in the nature
(including individuals that made a school-to-work of individuals’ job tasks and the extent to which
transition). The analysis emphasises the role of such tasks are changing over time.
VET and of WBL programmes in facilitating faster Chapter 6 concludes by drawing useful
and better match of job seekers’ skills with labour lessons informing the design of better and more
market needs. effective policies that may mitigate skill mismatch.
Chapter 5 focuses on the dynamic evolution Annex 1 provides further technical details on
of skill mismatch among EU employees during the preparation of the survey, the data collection
their job tenure, looking at the extent to which EU process and the analysis of the new data.
Box 3. Methodological note
The findings in this report show associations between information about the skills and job features of
different (sets) of individual and job characteristics and individuals, analysis of skill mismatch in this report is
the incidence of various measurements of skill mismatch based on subjective evaluation by the individuals
of adult workers in different EU countries. They are themselves. The survey has not engaged in a direct
derived on the basis of a cross-section survey, which assessment of the skills of individuals, nor does it analyse
provides a snapshot of issues at a certain point in time the skill requirements of jobs in an objective fashion, such
but cannot be relied on to derive strong conclusions about as by relying on job descriptions. Nevertheless, all
the causality of different variables. Nevertheless, the necessary efforts were taken during the preparation of
statistical significance of most empirical findings in the the survey to ensure that any reporting biases on the part
report has been validated using multivariate regression of the respondents are minimised. In relation to the
analysis. In most cases discrete choice estimators measurement of skill mismatch, in particular, the survey
(probits, multinomial logit) have been employed, given employs multiple questions that aim to identify the extent
that the dependent variables of interest (qualification of response bias among respondents in a manner that
mismatch, skill mismatch, incidence of WBL, labour was not done in prior surveys.
market status) tend to be binary or categorical variables. Given that the survey was carried out in 28 different
Ordinary least squares (OLS) estimators have been used countries, the issue of cross-cultural comparability is also
in cases of continuous dependent variables, such as skill important. During the cognitive testing phase that
intensity of jobs and improvement in skills over job tenure. preceded the survey, specific attention was paid to
Typically, the statistical significance of the explanatory ensuring that difficult terms such as ‘skills’ and ‘tasks’
variables has been established by first controlling for key were accurately perceived by individuals in their own
exogenous variables (gender, age, immigrant status, level country context and language. Further, it has been
of education, type of contract, occupation, industry) and ensured that the empirical estimates with regards to
subsequently taking into account other subjective and correlations among different variables have been purged
possibly endogenous determinants. of any cross-cultural influences (with the use of country
It is important to acknowledge that although survey- dummy variables in the empirical specifications).
based questionnaires are a versatile tool for extracting
29%
In 2014, qualification mismatch affected,
on average,
29% of the European adult working population
Qualification and skill mismatches

CHAPTER 2
in the EU workforce
2.1. Qualifications supply and that confirm the significant economic and social
demand in the EU job

benefits of higher levels of education and training.
market
Individuals with higher qualifications and skills are
generally found to be more likely to enjoy superior
labour market and social outcomes, such as
2.1.1. Economic and social benefits of better chances of employment, higher wages and
education and training levels of job satisfaction but also elevated levels
In line with the key Europe 2020 education target, of health and social capital (OECD, 2013;
to raise the level of tertiary education attainment Hanushek et al., 2014; European Commission,
of adults (aged 30 to 34) to at least 40% by 2020, 2012a).
European countries have seen rising levels of Cedefop’s ESJ survey confirms that
education attainment in recent decades; this individuals holding a tertiary qualification in the
trend is expected to continue over the next EU earn higher hourly wages, on average, and
decade (Cedefop, 2012d). European and national are more satisfied with their jobs, while those with
education and training policies have placed great a level of education below upper secondary level
emphasis on the skills upgrading of European feel they are in greater danger of losing their jobs
citizens, based on a wealth of empirical evidence in the near future (Table 2).
from both developed and developing countries
Table 2. Mean effect of education level on labour market outcomes, adult employees,
2014, EU-28
Highest level Hourly wage Job satisfaction Job insecurity Skill intensity
of education attainment of jobs
Low (ISCED 0-2) 14.39 6.89 2.73 6.00
Medium (ISCED 3-4) 14.50 6.96 2.60 7.00
High (ISCED 5-6) 17.76 6.98 2.59 7.54
NB: Indices of job satisfaction (How satisfied are you with your job?) and job insecurity (How likely is it to lose your job in the next two years?) defined on
a 0 to10 scale (0 = very dissatisfied/unlikely-10 = very satisfied/very likely). The index of skill intensity is an average of the degree of importance of
several skills (cognitive, digital, technical and generic) in people’s jobs (0 = not important at all-10 = extremely important). Column 1 is the gross
hourly wage from the job (before deductions or credits of tax and national insurance).
In 2014 European employees with tertiary characterised by a higher level of skill intensity.
education enjoyed a higher gross hourly wage Skill-intensive jobs are those that require a higher
premium of around 25% relative to those with degree of key competences (literacy, numeracy,
medium-level qualifications, while the lower- ICT), technical skills (specialised knowledge
educated suffered a 16% wage penalty (1). Part related to field of work) and transversal or soft
of this wage premium is linked to the fact that, on skills (communication, problem solving, planning,
average, individuals with higher qualifications are teamwork, learning to learn). Given that jobs with
more likely to end up employed in jobs higher skill content are associated with higher
(1) Based on the estimation of a mincer-type wage regression, which controls for gender, age, age square, a quadratic term of years of
employer tenure and 28 country dummy variables.
levels of labour productivity (2), it is evident that posts that could be filled instead by individuals
continuing to invest in a higher-skilled workforce holding high-level vocational diplomas (EQF 5).
for the purposes of strengthening the productivity Clerical jobs, for which employers may desire job
of its knowledge-based economy is a key strategy applicants in possession of good ICT skills, or
for raising European competitiveness and growth. jobs in market sales and services, which require
An increasingly highly educated workforce is graduates with high levels of communication
generally expected to elevate the level of demand skills, are examples where a ‘bumping down’
for qualifications by employers, who will adjust process may take place, with high-qualified
their recruitment practices and skill demands over individuals displacing those with lower
time, but also raise the overall level of labour qualifications. Even though the qualification level
productivity in the economy. The availability of a of the latter could be suitable for such jobs, firms
highly educated workforce encourages may prefer those with higher qualifications given
employers to restructure their jobs, making them that a higher education degree can act as a signal
more complex, while it can also support the of higher productivity and skills at the time of
adoption of advanced technology. However, when recruitment. This so-called ‘crowding out’ effect is
an increasing supply of qualified and skilled accentuated at times of economic recession,
individuals is not effectively met by adequate when the competition for a limited pool of jobs is
demand for skills in the job market, qualification fiercer.
mismatches may arise in the medium-term. Even among tertiary education graduates,
Qualification mismatches tend to be associated some individuals may have chosen education
with lower economic and social returns on higher programmes that tend to be more generic in
qualifications (Section 2.4), so they may nature in terms of skills and competences
ultimately undermine the incentives of young acquired (such as humanities and arts), while
Europeans to continue investing in their human others acquire skills of a more specific or
capital. technical nature. The former group may have a
higher likelihood of working in jobs needing
2.1.2. Qualification requirements in EU jobs different qualifications from their own.
Qualification mismatches occur when the formal Respondents in the ESJ survey were asked
qualifications that individuals hold deviate from about the qualifications and skills needed for their
the qualifications that are required to be able to jobs (Figure 7). In 2014, about 33% of adult job
be hired for and to perform the job (Quintini, holders in the EU believed that a tertiary
2011). In recent years there has been increasing education degree is required so that they can
concern about the high and rising numbers of perform their job. Pointing to the need for the EU
European employees who have taken up jobs to continue investing in strong VET systems as a
requiring lower qualifications or skills than their means of preventing skill mismatches, most EU
own (McGuinness, 2006; Leuven and employees (40%) responded that their jobs could
Oosterbeek, 2011; European Commission, be performed with a medium-level qualification,
2013a; European Commission et al., 2014). To largely vocational. Moreover, 21% of adult
the extent that such qualification mismatches workers stated that a low level or no qualifications
indicate an involuntary misallocation of at all suffice for doing their jobs.
qualifications and skills to jobs, they are of Education requirements of jobs vary across
significant policy concern given the waste of countries. More than 40% of employees in
resources that is entailed (Sloane, 2014). Belgium, Denmark, France, Luxembourg, Malta
For instance, without a high quality or highly and Finland reported that a tertiary education
esteemed vocational education system in a degree is needed to perform their jobs. By
country, employers experiencing growth in tertiary contrast, a high share of employees in Spain,
education graduates (at EQF levels 6 to 8) may Portugal and the UK, believe that a low
exhibit a preference for hiring such graduates for qualification or no education diplomas are needed
(2) Although the survey does not contain information on labour productivity, a positive association is found between higher skill intensity
jobs and proxies of labour productivity, such as job satisfaction and hourly earnings.
Dynamic change in skill mismatch Labour market outcomes Two digit occupa
(previous job start of current job (wages, job insecurity, expected
Elementary oc
start of job current post) CHAPTER 2
skill obsolescence, job satisfaction)
ISCO 91:
Cleaners and helpers
ISCO 92:
Figure 7. Level of qualification needed for jobs, adult employees, 2014, EU-28
Agricutural labourer
ISCO 93:
Low (ISCED 0-2) Medium (ISCED 3-4) High (ISCED 5-6) Labourer in miningco
70%
ISCO 94:
Food preparation ass
60%
ISCO 95:
Street and related sal
50% service workers
40% ISCO 96:

Other elementary wo
30%
25
20%
20
10%
0% 15
AT BE BG CY CZ DE DK EE EL ES FI FR HR HU IE IT LT LU LV MT NL PL PT RO SE SI SK UK EU-28
NB: Country averages based on responses to the question: ‘What are the educational qualifications, if any, which someone actually needs to do your job
today?' Sums not equal to 100% exclude missing responses. Overqualified Underqualified
10
Source:
50% Cedefop ESJ survey.
at all for doing their jobs. This should be of competent in their jobs. By contrast, a greater
concern, given that workers in these jobs proportion of workers in manufacturing or 5
perceive them as being of very low productivity, construction require a medium-level qualification.
40%
70%
in the sense that they can be performed by It is often argued that the qualification needed
individuals with a very basic level of education. to get a job is not equal to that needed to perform
0
30%
Table 3 shows that education requirements a job. Particularly in depressed labour markets,
RO CZ MT
also
20% depend on the occupation and industry in employers may inflate recruitment criteria to filter
sly employed
which individuals are employed. Close to eight in the best candidates. They may also deliberately
10 adult workers in the EU employed in seek to hire individuals with higher education as
professional occupations and 56% of managers a means of hedging against greater economic
report a need for a tertiary education qualification uncertainty (Bulmahn and Krakel, 2002) or to
10%
to be able to perform their jobs. A total of 44% of stimulate innovation. This induces individuals to
technicians
0%
UK CZ and
HR ATassociate
SK EE HUprofessionals
IE SI LT ES need
EU-28 CYa
higher qualification, although a similar percentage
SE DKacquire
DE LV higher
competitive
BG EL qualifications
LU FR IT PL toFI bePThired
labour market,
in aBEmore
fostering
RO NL MT 4
(42%) require a medium-level degree. An upper credentialism, even though the jobs for which
secondary or non-tertiary education degree is also they are hired may eventually require lower
n skills High Medium
more frequently required in clerical, service and qualifications and skills than those they possess.
40%
market sales, and building, crafts or related trades To investigate whether there is a discrepancy Talent
jobs. A higher share of employees working as between the qualifications needed for recruitment
plant and machine operators, in occupations purposes and the level of education that is a
Own skills ar
30%
related to the primary economic sector (such as genuine prerequisite for performing the
to do the job
deficit
agriculture or fishing) and in elementary jobs necessary tasks in a job, the survey asked
potential to
report
20%
that a low education level or no respondents to assess both the level of
qualifications at all are needed to be able to do qualifications needed to get their job, if someone
their jobs. Employees working in sectors providing would apply for their job today, and the level of
RO
professional or scientific services, education and qualifications needed to do their job.

health, ICT and finance or insurance are also in Limited evidence is found in the survey to
10% 35%
greater need of a tertiary education degree to be suggest that rising levels of mismatch in EU
24 30%
0% 25%
IE UK LT EE CY ES AT HR SE EL EU-28 CZ BG SI SK FI PL HU LU DE NL IT RO LV DK FR MT PT BE
20%
e start of job
15%
Table 3. Incidence of adult employees by level of education needed for their job, 2014, EU-28
Low or no qualifications Medium High

Occupation group
Managers 10% 29% 56%
Professionals 3% 15% 78%
Technicians and associate professionals 9% 42% 44%
Clerical support 18% 56% 19%
Service and market sales workers 37% 45% 9%
Skilled agricultural, forestry and fishing 53% 32% 8%
Building, crafts or related trades 28% 57% 8%
Plant and machine operators and assemblers 46% 42% 3%
Elementary 65% 23% 2%
Sector of economic activity
Agriculture, forestry and fishing 40% 34% 18%
Industry 20% 45% 28%
Services 21% 37% 36%
Public administration 18% 44% 32%
Other 29% 37% 21%
Contract
Full-time 19% 40% 35%
Part-time 29% 40% 24%
Indefinite/permanent 19% 41% 34%
Fixed term/temporary 25% 37% 31%
Temporary employment agency 40% 37% 15%
Informal (no contract) 41% 33% 19%
Other 28% 28% 32%
Importance of skills in jobs
Basic literacy 30% 45% 18%
Advanced literacy 9% 37% 49%
Basic numeracy 23% 43% 28%
Advanced numeracy 9% 36% 51%
Basic ICT 31% 45% 17%
Moderate ICT 13% 42% 39%
Advanced ICT 6% 32% 57%
Skill intensity index* 6.44 7.35 7.78
NB: * The index of skill intensity is an average of the degree of importance of several skills (basic, technical and generic) in people’s jobs (0: Not important
at all-10: Extremely important). When the sum of the columns is not equal to 100%, the residual is ‘no answer’ or ‘don’t know’ responses.
CHAPTER 2
labour markets are driven by inflated recruitment 2.2. Qualification mismatches

criteria. At a time when the European economy
in the EU
was characterised by disproportionally high
unemployment to vacancy ratios, about 35% of
2.2.1. Incidence of overqualification
EU adult workers stated in the survey that a
As limited job opportunities and financial
tertiary qualification is needed to be hired for their
constraints weigh more heavily on recent job
job, in contrast to the 32.7% who said that such a
seekers, many skilled European citizens have
degree is needed to do the job (3). However, the
recently accepted jobs that do not fully use their
wedge between the need to have a tertiary
qualifications and skills. Several sources have
education degree simply for recruitment purposes
cautioned that, as a result of the economic crisis
(75.6%) as opposed to truly performing one’s job
but also due to an increasing supply of higher-
(71.6%) is greater among tertiary education
qualified individuals in European labour markets,
graduates. Such apparent credentialism
there is a tendency towards higher over-
prevailing among job finders with higher levels of
qualification across Europe (Cedefop, 2014a;
education is greater for younger employees
ILO, 2014). Overqualified workers can be affected
(aged 24 to 29), 73% of whom stated that a
by a depletion of their skills, given that they utilise
higher degree is necessary for recruitment in their
fewer of their competences in their jobs and they
jobs as opposed to 67% who claimed that their
fail to develop them continuously as part of their
level of qualification is truly needed to perform
job tasks.
their job tasks.
Estimates of qualification mismatch based on
These findings confirm that, overall,
the ESJ survey rely on direct comparison of
employers’ qualification requirements when
individuals’ highest qualification level with the
attempting to screen job candidates can
level they report as necessary to do their job at
potentially diverge relative to the genuine skill
the current time (Section 2.1). The survey
needs of a job, particularly when concerning
confirms that a substantial share of the European
workers with limited labour market histories.
workforce is employed in jobs that need a
However, in the remainder of this report the
different (higher or lower) level of qualifications
reported qualification needed to do a job is used
than their own. In 2014, total qualification
as a proxy for the qualification requirements of
mismatch affected, on average, 29% of the
jobs, as it is believed to constitute a summary
European adult working population: 17% were
indicator of the total level of knowledge, skills and
overqualified and 12% underqualified (Figure 8).
competences required by workers to be
About one in four (23.6%) tertiary-educated
competent in their jobs (4).
workers and 16% of medium-qualified employees
in Europe are found to be overqualified for their
jobs. In 2014, 25.2% of highly qualified young
adult employees (aged 24 to 35) in Europe were
overqualified for their jobs, so the incidence of
overqualification is higher among younger
cohorts of EU tertiary education graduates.
The incidence of overqualification in the EU,
as found in the ESJ survey, is very close to the
level detected from the EU labour force survey
(3) A standard t-test of the difference in the means of these two statistics, with or without assumption of equal variances, confirms that the
wedge is statistically significant (t-test = 7.7).
(4) Using the subjective opinions of employees to detect the level of education required in jobs has a number of advantages relative to
alternative measures used in previous literature, such as the fact that they directly refer to an individual’s job and rely on the information
set of the job holder. By contrast, several researchers have adopted an empirical method that detects qualification requirements on
the basis of the mean or modal level of education attainment within broad occupation groups, or categorise jobs according to the broad
level of skills dictated by international occupation classifications (Leuven and Oosterbeek, 2011; European Commission, 2012a). Such
measures tend to be affected by the endogenous influence of labour supply on the distribution of employees within occupations and
neglect the marked heterogeneity of skill requirements within broad occupational groups.
st job Other
ISCO eleme
95:
Street and re
50%
30% service work
25
40% ISCO 96:
20%
34
Other eleme
Evidence from Cedefop’s European skills and jobs survey
20
30%
10%
25
Figure 8. Average incidence of qualification mismatch, adult employees, 2014, EU-28

20%
0% 15
20
10%
Overqualified Underqualified
10
50%
0% 15
40% 5
60% 70% Overqualified Underqualified
10
50%
0
30%
RO C
Previously employed 5
40%
60% 70% 20%
0
30%
RO C
Previously employed 10%
20%
0%
UK CZ HR AT SK EE HU IE SI LT ES EU-28 CY SE DK DE LV BG EL LU FR IT PL FI PT RO NL BE MT
10%Based on comparisons of an individual’s highest level of education qualification with the (self-perceived) level of education qualification actually
NB:
needed to do his/her current job.
ement in skills High Medium
40%
Figure 9. Breakdown of overqualification by level of education, adult employees, 2014, EU-28

0%
UK CZ HR AT SK EE HU IE SI LT ES EU-28 CY SE DK DE LV BG EL LU FR IT PL FI PT RO NL BE MT
Ta
Own
30% High Medium
ement in skills to do
40%
po
Ta
EE CZ 20%
Own
30%
RO to do
BG
10% po
35%
EE CZ 20%
22 24 30%
RO
BG 0% 25%
10% 35%
20%
asks
22 since start of job
24 30%
15%
Temporary employment agency
0% 25%
10%
IE UK LT EE CY ES Informal
AT HR SE EL EU-28 CZ BG SI SK FI PL HU LU DE NL IT RO LV DK FR MT PT BE
20%
5%
Contract
Fixed term/temporary
asks since start of job NB: Based on comparisons of an individual’s highest level of education qualification with the (self-perceived) level of education qualifications actually 15%
0%
needed to do his/herIndefinite/permanent
current job.
Source: Cedefop ESJ survey. 10%
Part-time
(EULFS) using a different method of measure- Czech Republic, Estonia, Croatia, Austria, and
Informal
Full-time
ment; in 2013 this stood at 15.5% (European Slovakia are found to be overqualified. Such
5%
Contract
Commission,
Not working2015b). Nevertheless, there are differences could arise because of the differential
Unemployed
0%
Previous labour
Indefinite/permanent 7.6
market status
notable differences with respect Part-timeto the ranking of method of proxying the qualification requirements
(e.g. child care, family) 7.4
EU Member States. For instance, Full-time the EULFS data of jobs. The empirical method adopted using the
Employed in another job
show that southern Europe countries (Greece, EULFS data, which approximates the level of
7.2
In education or training
Spain, Italy, Portugal), alongFemale with the UK have the education needed in different occupations by the
Unemployed 7.0
labour
highest overqualification rates. In the ESJ survey most frequently observed qualification held by
7.6
market status
Gender
Not working (e.g. child care, family) 6.8
a relatively large proportion ofjob

employees in the employees in the respective occupational groups,
7.4
Male
Previous
6.6
Employed in another 7.2
24-29 6.4
In education or training 7.0
30-34 6.2
group
Female
Gender
6.8
35-44 6.0
40% 50% Male
ge
6.6
0% 15
PREVIOUS CONTENTS NEXT Overqualified Underqualified

10
50%
CHAPTER 2
40% 5
% 70%
is30%affected by the (endogenous) education as Estonia, Ireland, Spain, Cyprus, Lithuania and
0
structure of an economy. What this implies is that the UK, have relatively higher shares of tertiary
RO CZ MT
countries with a high share of tertiary education graduates employed in jobs that demand lower
sly employed
graduates in their adult population are least likely qualifications than their own.
20%
to have high overqualification rates. By contrast,

the
10% qualification requirements identified from the 2.2.2. Determinants of overqualification
ESJ survey are more likely to reflect the views of Figure 10 illustrates that overqualification in the
employees regarding the level of required EU is more prevalent among younger employees,
productivity
0%
UK CZ inHR their jobs. DKfemales,
DE LV BGindividuals
EL LU FR who
IT PL were
FI PToutside
RO NL of
BE the 4
The overqualification rates presented in labour market (such as unemployed or inactive)
AT SK EE HU IE SI LT ES EU-28 CY SE MT
Figure 8 are also affected by the inclusion of prior to accepting their current employment and
individuals with upper secondary education those in jobs with non-standard High
contractual
Medium
employed in jobs that need low-skilled labour. As arrangements (such as part-time, informal, and
n skills
40%
identified in Figure 9, most overqualified temporary agency contracts).

individuals in the Czech Republic, Croatia, Austria Overqualification is also higher among tertiary
Talent
and Slovakia have medium-level qualifications. education graduates, particularly those who
Own skills ar
30%
Such workers believe that they are employed in finished school in the post-crisis period. Tertiary
to do the job
inherently low-skilled jobs, which can imply a loss education graduates from certain fields of study,
deficit
potential t
of labour productivity and a waste of individuals’ including humanities, languages and arts and
skills, particularly in countries with high quality other social sciences (sociology, political
20%
(vocational) education and training systems. sciences, anthropology, psychology) are found to
RO
Nevertheless, much of the public debate in recent be more likely to be overqualified, as opposed to
years has associated overqualification with the university graduates with degrees in medicine or
10% 35%
24
expansion of tertiary education systems in EU health-related sciences, engineering or
countries. Figure 9 illustrates that countries such computing sciences (Figure 11).
30%
0% 25%
Figure 10. Overqualification by population group, % of adult employees, 2014, EU-28

20%
ce start of job
15%
10%
Informal
5%
Contract
Indefinite/permanent 0%
Oversk
Part-time
Full-time
Unemployed
Previous labour
7.6
market status
Not working (e.g. child care, family) 7.4

Employed in another job 7.2
In education or training 7.0
Female
Gender
6.8
Male 6.6
24-29 6.4
30-34 6.2
Age group
35-44 6.0
50% Oversk
45-54
55-65
0% 5% 10% 15% 20% 25% 30% 35%
24-29 24-65

Female 6.8
Gende
Male 6.6
24-29 PREVIOUS CONTENTS NEXT 6.4
30-34 6.2
Age group
35-44 6.0
40% 50%
45-54
55-65
Figure 11. Overqualification by field of study, adult employees with higher-level qualifications,
0% 5% 10% 15% 20% 25% 30% 35%
2014, EU-28
24-29 24-65
Maths and stats
Computing sciences
80
70
Teacher training and education science
60
Labour productivity
Natural sciences 50
40
30
Humanities, languages and arts 20
10
0
10
0% 5% 10% 15% 20% 25% 30% 35% 40% 45%
NB: Sample of individuals with higher-level qualifications (ISCED 5-6).

Rates of overqualification are also Table 4. Examples of two-digit occupations

significantly different depending on job with highest and lowest rates of
characteristics, with overqualification probability overqualification, adult employees,
higher for individuals employed in smaller-sized 2014, EU-28
firms and in less skill-intensive occupations and
jobs (Table 4). Overqualified workers also have
Overqualified to do the job
fewer possibilities for CVET and for career Highest

progression (they are less likely to have been
promoted in their job).
Sales worker 32%
The incidence of overqualification is higher

Food preparation assistant 32%
among individuals employed in elementary and

Labourer in mining, construction, manufacturing 31%
skilled agricultural jobs, while it also Cleaner or helper 28%

disproportionally affects plant and machine
operators and assemblers and market sales
A protective service worker 26%
workers. Adult workers in professional and

Customer services clerk 25%
managerial occupations are the least likely to

Lowest
report that they are overqualified for their jobs.

Legal, social and cultural professional 10%
Relative to those employed in the manufacturing Science and engineering technician 10%
sector, individuals working in wholesale and retail
trade, accommodation or food services,
Production or specialised services manager 9%
transportation and storage, arts and

Science and engineering professional 6%
entertainment, and also in ICT services are more

Health professional 6%
likely to have higher qualifications relative to the Teaching professional 4%

ones needed by their jobs. Source: Cedefop ESJ survey.
CHAPTER 2
Empirical analysis further reveals the about 15% of medium-qualified EU employees

importance of taking into account the job (Figure 12).
preferences and constraints faced by individuals The underqualified are more likely to have
when accepting their jobs, as they are found to been motivated to accept their job because of
be statistically significant correlates of the career concerns or the intrinsic nature of the job,
chances of overqualification among adult as opposed to job security or proximity to their
workers. Part of the reason why individuals are home. They are also less likely to have changed
overqualified in their jobs can be attributed to the occupation before accepting their current job. They
fact that they selected their job because of their are employed mostly in high-skilled occupations
desire for job security and a positive work-life (professionals, associated professionals) or in jobs
balance, or due to its proximity to home. By involving a greater degree of task complexity. They
contrast, adult workers who accepted their job have accumulated significantly more work
because of the intrinsically satisfying nature of experience and work-specific skills as they have
work, the fact that it offered possibilities for career been employed with their current employers for a
progression but also the suitability with their own longer time period.
skills and qualifications are less likely to be
overqualified. Figure 12. Underqualification by age group,
The constraints that individuals face both at adult employees, 2014, EU-28
the micro and at the macro level can influence
their job search and, ultimately, the quality of their
25%
job match. In particular,

Start of workers
current whojobexperienced
greater financial constraints
(circumstances, motives, entry, and
prior to job
20%
those searching for a job job in a slack labour market

with few suitable job opportunities, have
mobility) 15%
significantly higher chances of being over-

qualified. The survey findings further indicate that
10%
a radical change in occupation is unlikely to form

a good strategy for mitigating the chances of 5%
overqualification. Individuals employed in a

completely different occupation relative to the one
they had in the previous job are found to have a
0%
24-29 30-39 40-54 55-65
higher likelihood of overqualification. By contrast,

rent labour market status
those who decided to move home to a different
(occupation, industry,
country before accepting their current job have a These findings highlight that underqualified
job characteristics)
significantly lower chance of overqualification. employees are likely to possess a large stock of
(non-formal or informal) skills to perform their
60%
2.2.3. Underqualification of EU workers jobs.

50% Nevertheless, they lack the qualifications
As shown in Figure 8, about one in five needed for their jobs. This indicates the need for
employees in France, Italy and Portugal reported validation or certification of their acquired, yet
that their own qualification was below the level unrecognised, skills for the purposes of enabling
40%
needed to do their jobs. By contrast, workers in their continued mobility and career progression in
Skill needs
the Baltic states, the Czech Republic andskills;
Croatia the labour market (European Commission,
30%
were the least likely to believe that they /are 2012a; OECD, 2013; Cedefop, 2015b).
(foundation, technical and generic
underqualified for their work jobs. Cedefop’s ESJ

tasks in job; changing tasks, technologies
survey reveals further that underqualification

methods) 20%
affects mostly older individuals (particularly those 2.3. Skill mismatches in the EU
over 40) and employees with low levels of
10%
education (below lower secondary level). In 2014, 2.3.1. Skill mismatch relative to required level
about 39% of all lower-educated workers When ATtheUKstock
EL of
DE knowledge andCY competences
0%
believed that a higher qualification than their own of employed workers deviate from the skill
IE FI ES SK PL HR DK SE HU NL SI CZ IT FR
h is needed to do their job,

Labour while the
market same holds for
outcomes requirements of their jobs, the latter derived from
Two digit occupation groups who say their skills are higher than required...
(wages, job insecurity, expected
skill obsolescence, job satisfaction)
Elementary occupations Sales and service workers
ISCO 91: 45 ISCO 51: 45
the complex set of job tasks they have to perform are lower than what is required and need to be
as part of their daily work, skill mismatch ensues. further developed (5). Some of these questions
Skill mismatch is a major challenge facing were asked several times, referring to different
European economies (Cedefop, 2010a). periods of the individual’s working life.
Evidence from past surveys, including the 2010 Respondents were asked to assess the degree
EWCS, have suggested that more than a half of of mismatch between their skills and those
EU workers’ skills are not well-matched to their needed to do their jobs:
current jobs. Most mismatched adult workers are (a) in their previous job;
overskilled, which means they are capable of (b) at the start of their current job;
handling more complex tasks and their skills are (c) at the time of the survey (6).
underused; a smaller share of employees are
underskilled and they lack the skills needed for Cedefop’s ESJ survey shows that, in 2014,
their jobs. about 45% of employees in the EU-28 Member
25%
To assess the extent to which skill mismatch States experienced skill mismatch: 5% of workers
job 0.30
20%
affected European workers in 2014, the ESJ felt that some of their skills are lower than needed
ves, 0.25
survey asked respondents a series of questions to do their job and 39% believed that they have
designed to evaluate the extent to which some of more skills than needed by their job. Self-reported
15%
0.20
their
10% skills are seen to be important for or underskilling is highest in the Baltic states, but
matched to their jobs, and also the overall degree also in the Czech Republic, Ireland and Finland, 0.15
to5%which their skills are at the level needed to do while more than 40% of workers feel that their
their job. The respondents could either report that skill level is higher than needed to do their job in 0.10
their
0% skills are higher than or matched to what is Germany, Ireland, Greece, Austria and the UK
required 24-29
by their job, or that some
30-39 40-54 of their skills
55-65 (Figure 13). 0.05
Figure 13. Incidence of skill mismatch, adult employees, 2014, EU-28 0.00
Overskilled Underskilled
60%
50%
45%
40%
40%
35%
eneric skills; 30% 30%

echnologies /
25%
20%
20%
10% 15%
10%
0%
AT UK EL DE IE FI ES SK CY PL HR DK SE HU NL SI CZ IT FR BE LU BG PT RO EE LV LT MT EU-28 5%
comes
(5) The measure of underskilling used in the ESJ survey deviates from the measures adopted in the ESWC and PIAAC surveys, which
Two digit occupation groups who say their skills are higher than required... L
asked individuals whether they are in need of training to cope with their job duties. Based on inputs from the ESJ expert group, as well
xpected
as feedback received as part of the cognitive and pilot testing that preceded the data collection, a new phrasing was preferred to
isfaction)
Elementary occupations Sales and service workers Plant and machine operators
capture the incidence of underskilling. The ESJ measure avoids confusing the issue of underskilling with the differential provision of
training
91: in different EU countries.
45 The questionISCOwas 51:
phrased in a neutral tone, so that social desirability
ISCO 81: bias could be mitigated to some
50%
ISCO 45 42
extent.
Cleaners andThis was deemed necessary given that
helpers most service
Personal individuals had a natural tendency to Stationary
workers avoid reporting that they
plant machine had lower skills
operators
than needed to do their jobs.
40%
(6) As92:
part of the cognitive
35testing phase of the ISCO
project,
52:particular attention was 47paid to whether
ISCOthe82:responses of individuals47to these
questions are affected by memory bias. The feedback received was that this was not an issue of significant concern, although caution
ISCO
needs to be exercised in the interpretation.
Agricutural labourer Sales workers Assemblers 30%
ISCO 93: 42 ISCO 53: 36 ISCO 83: 47 20%

High (ISCED 5-6) Labourer in miningconstruction Personal care workers Drivers and mobile plant operators
43 10%
ISCO 94: ISCO 54: 42
CHAPTER 2
2.3.2. Factors related to skill mismatch The constraints individuals faced at the time
Specific groups of the European adult working of their job search, both at the micro and the
population are more likely to consider that their macro level, are an additional factor in
skills exceed or are lower than those needed to determining whether they are employed in jobs
do their jobs. Men are more likely than women to that do not fully exploit their potential.
be overskilled, as are older workers (aged over Respondents whose skills are not matched to
55) and individuals with a higher level of their job are more likely to report that they had
education. The prevalence of overskilling is found financial difficulties at the time of their job search
to be greatest for individuals employed as service than those whose skills are matched (20%
and sales workers, plant and machine operators, compared to 17%, respectively). More of them
and those working in elementary occupations. (45%) reported that there were few job
European workers in these occupation groups opportunities available for people with their skills
are more likely to say that their skills are higher and qualifications at the time that they started
than required (44%, 45% and 47%, respectively), working for their current employer (compared to
as opposed to individuals working in crafts and 40% of those in well-matched jobs). They are also
related trades workers (31%) and as more likely to report that, despite sending many
professionals (35%) (Box 4). job applications, they had few opportunities to
The importance of targeted and good-quality attend job interviews (39% compared to 33% of
guidance and career counselling when choosing respondents not experiencing skill mismatch).
a job is also apparent in the survey findings. Relative to individuals whose skills are well-
Individuals who placed greater emphasis on the matched to what is needed by their job,
fit of the job with their skills and qualifications, as overskilled workers are employed in jobs that
well as the chances provided in terms of work require, on average, a higher degree of task
experience and career development, are less complexity and a higher level of skills to perform
likely to have ended up in a mismatched job. The them. Nevertheless, the latter type of workers
contrary is true for workers whose primary appears to be dissatisfied by the fact that, even
motivation for selecting their job was a desire for though they were in possession of a higher stock
job security. of skills relative to what was needed at the start
of the employment relationship, the extent of
Box 4. Skill mismatch in specific continued development in their skills since then
occupations was limited. By contrast, the complexity of the
tasks they perform, defined by the continual need
to learn new things or to engage in a greater
variety of challenging tasks, decreased since the
The ESJ survey collected information on the
beginning of their job tenure.

respondents’ occupation at the two-digit inter-
Respondents are more likely to consider that

national standard classification of occupations
their skills lag behind those needed to do their job

(ISCO) level (7). Figure 14 shows the proportion of
(underskilled) if employed in higher skilled

respondents in the three broad occupation groups
occupations (managers, professionals) or in

most likely to say that their skills are higher than
sectors of economic activity related to

required, broken down at the two-digit level. There
professional, scientific and technical services and

are several differences within occupation groups,
in the manufacturing and engineering sectors.

most notably among elementary occupations where
Underskilled workers are more likely to have

street and related sales and services workers are
returned to (temporary) jobs following a spell of

significantly more likely than other elementary
inactivity. Given that such individuals may have

workers to say their skills are higher than required
experienced some atrophy of their skills during

(63% compared to 47% overall). Among sales and
the time they remained at distance from the

service workers, personal care workers are the
labour market, some catching up is needed to be

least likely to say their skills are higher than
able to follow the demands of their job.

required (36% compared to 44% overall).
(7) http://www.ilo.org/public/english/bureau/stat/isco/ [accessed 7.8.2015].

35%
generic skills; 30% 30%

technologies /
25%
) 20%
40 Evidence from Cedefop’s European skills and jobs survey 20%
10% 15%
Figure AT14.UKISCO
EL DEgroups
IE FI most
ES SK likely
CY PL to
HR say
DK their
SE HU skills
NL SI are
CZ higher than
LU required
BG PT RO in
EE current job,
10%
0%
adult employees, 2014, EU-28

IT FR BE LV LT MT EU-28 5%
0%
tcomes Two digit occupation groups who say their skills are higher than required...
xpected
tisfaction)
Elementary occupations Sales and service workers Plant and machine operators
50%
ISCO 91: 45 ISCO 51: 45 ISCO 81: 42
Cleaners and helpers Personal service workers Stationary plant machine operators
40%
ISCO 92: 35 ISCO 52: 47 ISCO 82: 47
ISCO 93: 42 ISCO 53: 36 ISCO 83: 47 20%

High (ISCED 5-6) Labourer in miningconstruction Personal care workers Drivers and mobile plant operators
43 10%
ISCO 94: ISCO 54: 42
Food preparation assistant Protective service worker
0%
ISCO 95: 70
Street and related sales and
service workers
%f
ISCO 96: 51 DE
Other elementary workers FR
UK
Source: Cedefop ESJ survey. SE
Underskilled employees are also less likely to feel may be biased. Although some would argue that
IT
that their job has offered them opportunities to individual workers possess the best information
25 EL
improve their skills while the complexity and on the competences required to be proficient in
CZ
PL
difficulty of their tasks has continually increased their jobs, it is likely that their opinions will diverge
20 NL
since the time they started their current period of from those of their managers or from what could
DK
employment. be considered a given benchmark of optimal

HU
Another significant finding of the survey is that performance in the long term. Deviation of the
15 ES
E SI SK UK EU-28 AT
there is a significant degree of persistence in skill skill levels from such an optimal threshold would
BE
mismatch: being overskilled (underskilled) even then constitute a measure of their skill deficit (8).
Underqualified IE
in a previous job significantly increases the To assess the extent to which EU employees
10 SK
chances of being overskilled (underskilled) in the understate the degree to which their skills have
FI
current job. Being in a matched job in the past scope to be developed further in their jobs, even
PT
5 EE
significantly reduces the chances of becoming though they may feel that they have ‘surplus
RO
underskilled in the future (see Section 5.2 for skills’ that are ineffectively utilised at present, the
LT
further
RO discussion). DKsurvey
PT PL asked respondents a novel
FI IT question.
0 CY
Individuals were asked to rate, on a 0 to 100

CZ MT EE BG LT SI LV SK ES IE HU HR AT EL EU-28 BE SE UK FR DE CY NL LU SI
BG
2.3.3. Skill deficits undermining productivity scale, the extent to which their skills deviate from
LV
potential the level of skills necessary to carry out their job

LU
Overskilling is associated with lower levels of as well as possible. This self-reported measure
MT
labour productivity and waste of employee skills, of the distance between the skills needed to do
HR
making it an issue of policy concern. However, the job in the best possible way and workers’ own
0%
the incidence of skill mismatch reported above is skills is a unique new measure of skill mismatch
RO NL BE MT 44%
based on the self-perceptions of individuals and 27% 29%
at work (9). In this case the benchmark is defined 100%
80%
(8) Respecting convention, the definition of a skill deficit as described by Green (2013), the assessment of the skills of individuals relative
to a given benchmark, is adopted here.
igh Medium 60%
(9) The self-reported measure of a skills deficit in the ESJ survey was initially piloted in a small-scale online survey on skill obsolescence
carried out in 2011 in four EU countries (Cedefop, 2012b). It was also tested in the pilot phase of the ESJ survey, with no significant
40%
difficulties found in the ability of individuals to respond to such a question. In future waves of the survey an additional question may
Talent unnourished Talent in dead end Talent in stand still 20%
Own skills are lower than needed Own skills are higher than needed to Own skills are matched to those
to do the job today or higher skill do the job today and equal/lower skill needed to do the job today and 0%
deficit than average – deficit than average – equal/lower skill deficit than average –
potential to grow skills in job limited potential to grow skills in job limited potential to grow skills in job
50%
45%
PREVIOUS CONTENTS NEXT 40%

40%
CHAPTER 2
41
35%
Qualification and skill mismatches in the EU workforce
skills; 30% 30%
ogies /
25%
20%
as the level of skills needed so that the employee some skills deficits with regards to achieving
20%
can achieve the best possible performance level higher productivity dividends from their job but
10%
in his/her job. Such a measure differs relative to may be broadly overskilled in relation to the
15%
the
0%conventional construct that has been used in execution of their daily job tasks. Workers tend to
10%
the literature, as described in Section 2.3.1, which be in possession of a range of diverse skills but
AT UK EL DE IE FI ES SK CY PL HR DK SE HU NL SI CZ IT FR BE LU BG PT RO EE LV LT MT EU-28
typically asks employees to evaluate their skills not all of them will be useful for the organisation
5%
in relation to those needed to carry out are

their job than required...
at a given point in time (Ployhart and Moliterno,
0%
es
duties, regardless of how well these tasks are 2011; Ployhart et al., 2014). The talent
Two digit occupation groups who say their skills higher LT LV
carried out. management

workers approach focuses on the potential
d
Elementary occupations Sales and service Plant and machine operators
The measure of 45 skill deficit used here of workers – some of workers’ skills will become
on)
attempts to detect the extent to which individual useful to the organisation

Stationary plant machine operators in the
at some point
50%
ISCO 91: ISCO 51: 45 ISCO 81: 42
workers have space to grow in their jobs. It is a future – rather than focusing on the contemporary
Cleaners and helpers Personal service workers
measure
ISCO 92: that is more 35closely aligned to the52:notion skill (mis)match in the 82: job. It also takes 47 into
40%
47
of individuals being lifelong learners in modern account that the optimal productivity potential of
ISCO ISCO
societies, as espoused by the European jobs36may change ISCO

within 83:organisations over 47 time,
education and training strategy 2020.Personal It is not at given that organisations mayplant
alter the way in
ISCO 93: 42 ISCO 53: 20%
odds with the claim that about 40% of EU workers which they approach the market, change product
h (ISCED 5-6) Labourer in miningconstruction care workers Drivers and mobile operators
are overskilled for their jobs, given ISCOthat

54: the lines 42
or revise their production process altogether,
43 10%
ISCO 94:
contemporary presence of overskilling (the with implications for the subsequent productivity
Food preparation assistant Protective service worker
potential
ISCO 95: to do more demanding 70 job duties) and frontier of jobs.
0%
dynamic underskilling (the need to further Respondents were asked to think about the
Technical
develop one’s skills) is at the heart of talent level of skills needed to do their job as well as
Street and related sales and
management (Cappelli, 2008;

51 Cappelli and Keller, possible. They then rated their own level of skills
service workers
% few opportu
2014). Workers might have, simultaneously, relative to this benchmark on a scale from 0 to
ISCO 96: DE
Other elementary workers FR
Figure 15. Average skill deficit by EU Member State, adult employees, 2014, EU-28
UK
SE
IT
25 EL
CZ
PL
20 NL
DK
HU
15 ES
K UK EU-28 AT
BE
nderqualified IE
10 SK
FI
PT
5 EE
RO
LT
0 CY
RO CZ MT EE BG LT SI LV SK ES IE HU HR AT DK PT PL EL EU-28 BE SE UK FR DE FI IT CY NL LU SI
BG
NB: Skill deficits are derived as the difference between the level of skills needed by an employee to do his/her job as well as possible (equal to a score of LV
100) and his/her own assessed skill level relative to that benchmark. The average EU employee who rated his/her skills as 82 on the 0-100 scale LU
therefore has a skill deficit equal to 18. The larger the skill deficit, the greater the productivity loss for a given worker and the need for continuing MT
training and skills upgrading.
HR
0%
L BE MT 44% 27% 29%

also be asked, which will request from the respondent to report what he/she thinks would be his supervisor’s/rater’s assessment of
his/her skills gap (if any). This could lead to better comparative results, given that in the field of performance assessment the latter
formulation appears to yield a higher correlation with a rater’s true assessment than the own self-reported measure of performance by
an employee (Schoorman and Mayer, 2008; Schat and Frone, 2011). It was not possible to try this additional formulation in the existing
100%
survey due to space and time constraints.

80%
Medium 60%
40%
100, where 0 means that they need to develop all their own skill level at or below 75, which
of their skills (they are still at the start of the corresponds to the highest quartile of the
process of becoming fully proficient in their jobs) distribution of the skill deficit variable. These
and 100 that they have all of the skills needed employees can be characterised as suffering
(they have achieved the highest level of from significant skill deficits (on average, they
proficiency in their job). The skills deficit of have a skill deficit of 38%). Only one in seven
employees is inversely related to their self- (14%) rated themselves as 100 on the scale,
reported relative measure of skills in the survey: meaning they have all the skills they need to do
skills deficit = 100 – individual’s assessed skill their job as well as possible; most respondents
level. A smaller skill deficit implies that workers’ acknowledge that they need to develop their skills
skills are relatively closer to being fully productive, to some extent.
while the opposite case signifies that employees Figure 15 indicates that employees in
have ample scope to develop their potential Bulgaria, the Czech Republic, Malta, Romania
productivity in their jobs via continuing training or and the Baltic states suffer from the largest skill
other forms of skills upgrading. deficits among EU Member States. At the
In the pooled EU-28 sample, the mean opposite side of the spectrum, workers in
assessed skill level of adult employees on a 0 to Germany, Italy, Cyprus, the Netherlands and
100 scale where a score of 100 corresponds to Finland, are more likely to believe that their skills
the optimal benchmark of performance, is 82 and are close to the optimal performance threshold of
the median is 85, so the responses were skewed their jobs.
towards the higher end of the scale. As a On average, younger employees with fewer
consequence the average skill deficit in the EU years of employer tenure, as well as those who
workforce is equal to 18%. This implies that the experienced a previous spell of unemployment
stock of skills in the EU workforce should be before commencing their current job, are more
developed by about one fifth, should the EU wish susceptible to greater skill deficits. Individuals
to reach the maximum productivity potential of its employed in higher-skilled occupations
stock of employed human resources. Roughly (technicians and associate professionals,
one in four (26%) workers in the EU assessed professionals) experience larger skill deficits, as
Table 5. Worker groups by type of skill mismatch/deficit, adult employees, 2014, EU-28
Current skill mismatch

Skill deficit Underskilled Overskilled Matched skills
Above average Underskilled with Overskilled with Matched with
skill deficit growth potential growth potential growth potential
% current skill mismatch 74% 32% 47%
group
% total workforce (4%) (12%) (26%)
Below average Underskilled with limited Overskilled with limited Matched with limited
skill deficit growth potential growth potential growth potential
% current skill mismatch 26% 68% 53%
group
% total workforce (2%) (27%) (29%)
NB: The average skill deficit, used as a threshold to categorise employees according to whether they fall below or above it, is the predicted values based
on an OLS regression with dependent variable the skill deficit of respondents and independent variables as follows: age, gender, level of education,
native status, attitudes to learning, previous labour market status, occupation, industry, size of firm, change in job role since start of job, multisite
workplace, part-time contract, temporary contract, log(hours), years of employer tenure, public sector, index of task complexity in job, index of skill
intensity in job.
15
K UK EU-28 AT
BE
nderqualified IE
10 PREVIOUS CONTENTS NEXT SK
FI
CHAPTER 2
PT
5 EE
RO
LT
0 CY
Figure 16. Worker groups by type of skill mismatch, adult employees, 2014, EU-28
RO CZ MT EE BG LT SI LV SK ES IE HU HR AT DK PT PL EL EU-28 BE SE UK FR DE FI IT CY NL LU SI
BG
LV
LU
MT
HR
0% 1
L BE MT 44% 27% 29% 100%
80%
Medium 60%
40%
potential to grow skills in job limited potential to grow skills in job limited potential to grow skills in job
0.80
do those in non-permanent jobs and single site job’s upper threshold of skill requirements. Figure
0.70
workplaces. Adult workers in the manufacturing 16 illustrates that the skill deficits of about 44% of
0.60
35%
and
30% engineering sector report higher skill deficits the EU adult workforce lie above the average skill
0.50
relative to those in the wholesale and retail trade, deficit as defined for individuals with comparable
0.40
transportation or storage and other service- individual and job characteristics in their own
25% 0.30
related industries (finance and insurance, country. For this group of workers – the
T PT BE
education and health, cultural industries). ‘unnourished talent’ – there is additional scope to
20% 0.20
Adopting a talent management approach to improve their skills and productivity in their jobs.
0.10
15%
identifying whether individuals’ skills are matched By contrast, 27% of adult workers in the EU are
0.00
or5%not to their jobs provides a more nuanced in ‘dead-end’ jobs, defined as jobs where their
10% EL
picture of the incidence of skill mismatch in EU skills are not only ineffectively utilised but also do 60%
labour markets. For instance, when taking into not require continuing improvement as they are
account Overskilled
the scope that individual employees closerTalent
to the optimal Talent
productivity threshold than
0%
Matched skills Underskilled Talent 50%
acknowledge to have to augment their skills in the average worker. Finally, 29% of EU
in dead end in standstill unnourished
their jobs, it becomes evident that almost one third employees are in a ‘standstill’ situation: their skills
40%
of
7.6 overskilled employees (32%) – those who are matched to what is needed to do the job and
reported that their skills are higher than needed to are close to the average skill deficit.
30%
do their jobs at present – have considerable scope Adult workers in dead-end jobs are more
7.4
to grow as professionals in their jobs (Table 5). likely to be young, male and in possession of a
7.2 20%
Similarly,
7.0
47% of those with matched skills in their higher level of education. They are more prone to
current job believe that their skills lag behind the having selected their current employment in
10%
level
6.6 needed to be fully proficient in their jobs. For search of job security or because it offered a good
6.8
about one quarter of workers who declared that work-life balance, as opposed to focusing on
0%
their skills need to be developed to do their jobs, other job amenities, such as the possibility to
6.4 201
the scope for further skill accumulation is limited. accumulate work experience, career prospects or
6.2
6.0
Following the underlying logic described intrinsic satisfaction. Talent in dead-end jobs had
above, it is possible to identify three distinct greater
in dead chances inofstandstill
encountering financial
Overskilled Matched skills Underskilled Talent Talent Talent 60%
groups of employees in EU job markets according difficulties in the period preceding the start of their
end unnourished
to the degree to which their skills and, employment and also experienced limited
50%
subsequently, their productivity, can be opportunities to find jobs suitable for their own
35%
augmented in their jobs or are matched to the skills and qualifications. They are more likely to
40%
24-65 30%
20%
be employed in part-time, low-skilled or their levels of skills and vice versa. Such
elementary jobs. Employees working in industries discrepancy can arise from heterogeneity of skills
such as wholesale and retail trade, and abilities between individuals with the same
accommodation and food service, transportation credentials. Evidence from the OECD PIAAC
and storage and other service industries (such as survey, for instance, has highlighted significant
cultural, social and personal services) are more variance in the assessed level of foundation skills
likely to be in this group relative to those in between individuals with the same level of
manufacturing and engineering. Individuals in education, which accounts for a significant part of
dead-end jobs experience limited possibilities for their differences in labour market outcomes
skill development and training and their job tasks (OECD, 2013). Alternatively, individuals
have not evolved markedly over time. mismatched by qualifications may have adequate
The typical characteristics of the group of skills required for their jobs due to the
unnourished talent are female and prime- and accumulation of informal skills and work
older-aged workers. Their underlying motivation experience acquired in the labour market or in their
when accepting their jobs was the need to job over time.
accumulate more work experience and the career A population group of significant policy
prospects offered. Even though their jobs are concern, given the marked waste in terms of lost
inherently skill-intensive and subject to dynamic potential productivity, are the ‘genuinely
changes in technologies and in task complexity overqualified’: individuals employed in jobs that
over time, the degree to which such workers have require both lower qualifications and skills than
improved their skills since the start of their tenure their own. Some workers may be forced to accept
is deemed by them to be insufficient. a job below their qualification level because of a
Individuals who engaged in a significant poorly performing labour market or because of
change in occupation before starting their current other individual constraints (such as family
job generally have higher chances of not being in obligations or financial constraints). Genuinely
jobs with a close match between their skills and mismatched workers (Table 6) are of particular
the skills needed. Individuals with matched skills policy concern, given the involuntary nature of their
and limited potential to grow their skills in their mismatch and the associated loss of productivity
jobs – a standstill situation – tend to have lower due to the underutilisation of their skills (Green and
levels of education, more years of tenure with Zhu, 2010). They tend to be locked in a low-skill-
their current employer and be less likely to face intensity equilibrium, given that their jobs are
financial constraints or limited job opportunities at characterised by low skill content. Provision of
the time of entry into their current jobs. career guidance and counselling to enable such
workers find alternative jobs or career paths suited
2.3.4. The imperfect relationship between for their skills and qualifications is necessary, as
qualification and skill mismatch are policies to support a higher level of skill
While workers can be well-matched in terms of utilisation within their workplaces.
their formal qualifications, relative to those By contrast, policy-makers could be
required by employers at the time of recruitment, concerned less about the misallocation of
they may nevertheless become mismatched in education resources for workers who may have
Table 6. Combinations of qualification and skill mismatch
Genuinely mismatched Individuals characterised by a mismatch in their qualifications and in their level of skills
relative to the job requirements
Apparently mismatched Individuals characterised by a mismatch in their qualifications but are not mismatched in
their level of skills relative to the job requirements
Perfectly matched Individuals whose qualifications and level of skills are matched to the job requirements
Source: Cedefop, 2010a; European Commission, 2013a.

CHAPTER 2
Table 7.Qualification versus skill mismatch, adult employees, 2014, EU-28
Skill mismatch
Qualification mismatch Underskilled Matched skills Overskilled
Underqualified 7.7% 63.3% 28.9%
Matched qualifications 5.2% 58.3% 36.5%
Overqualified 4.4% 35.7% 59.9%
Talent unnourished Talent in standstill Talent in dead end
Underqualified 47% 34% 19%
Matched qualifications 44% 31% 25%
Overqualified 41% 18% 41%
accepted jobs below their qualification level for to certify and validate their skills, to the extent that
personal reasons, own preferences (such as a the absence of formal certification of their informal
wish to stay closer to a particular locality due to skills may inhibit their further labour market
job or environmental amenities) or because of mobility and career progression. At the same
their inherently lower abilities (Mavromaras et al., time, about 47% of such workers (9% of total EU
2010; Quintini, 2011; European Commission, employees) are in need of continuous skill
2013a). In this case, overqualified individuals may development to be able to perform to the
not suffer from simultaneous underutilisation of maximum level of productivity in their jobs.
their competences and may, instead, enjoy a high Ample space for continued growth in skills is
level of job satisfaction, rendering them also the case for about four in 10 employees in
‘apparently mismatched’. the EU whose qualifications are at the right level
The ESJ data confirm that most overqualified (matched) for their jobs (28% of the total sample
workers (six of 10) have surplus skills relative to of EU adult employees). However, the remaining
those needed for their job or are in dead-end and two quarters of this group (corresponding to 35%
static jobs (so are ‘genuinely overqualified’) of the whole EU adult workforce) have skills that
(Table 7). About 9% of the EU workforce is are adequate for the tasks required by their jobs.
genuinely overqualified. However, four of 10
overqualified employees are in jobs where their
level of skills is matched to or is lower than 2.4. The cost of qualification
required, or have considerable potential to grow and skill mismatch
their skills. This group of ‘apparently overqualified’
workers corresponds to about 6% of the EU Qualification and skill mismatches entail
labour force. Even though such workers have a significant economic and social costs for
higher qualification relative to what is needed for individuals. Past literature, focused on a subset
their job (for example a tertiary education of EU countries and specific non-EU countries
graduate employed in a job in need of a higher- (including Australia, US), has shown that
level vocational degree) they are still in need of overqualified workers are more likely to receive
continuous skill development to be able to lower mean wages and to have lower job
perform it. satisfaction relative to similarly-educated workers
Similarly, most EU employees who entered (McGuinness, 2006). The evidence on the wage
into their jobs with lower credentials than required consequences of overskilling is, by contrast, less
by their job today nevertheless have the proper clear-cut. While overskilled workers have been
or higher skills required to carry out their job found to be subject to small yet significant wage
tasks. For this group of mismatched workers penalties in the Australian labour market, the
(about 6% of the total EU workforce) it is crucial effect of overskilling on wages was found to be
Table 8. Impact of qualification-skill mismatch on labour market outcomes of tertiary education

graduates, adult employees, 2014, EU-28
Overqualified Matched qualifications

Overskilled Matched skills Overskilled Underskilled Matched skills
or underskilled
Gross hourly wage 17.64 14.96 18.32 16.67 18.56
Job satisfaction 6.17 7.08 6.91 6.74 7.35
Job insecurity 3.02 2.81 2.51 3.1 2.35
Skill obsolescence 3.82 4.05 3.74 4.78 3.92
NB: The scales of job satisfaction, job insecurity and skill obsolescence are measured on a 0 to 10 scale, were 0 = very unlikely/very dissatisfied and
10 = very likely/very satisfied.
statistically insignificant in a sample of OECD their skills are likely or very likely to become
countries surveyed as part of the recent PIAAC outdated in the next five years, 20% of matched
exercise (OECD, 2013). or overskilled workers stated that this is true.
According to the ESJ survey, tertiary Employees who suffer from significantly larger
education graduates who find a job that is a skill deficits in their jobs are more worried about
perfect match to both their level of qualifications their inability to keep up with the rising skill
and skills enjoy the best labour market outcomes demands of their jobs.
relative to other mismatched population groups Not having the right skills to be fully proficient
(Table 8). They receive the highest mean earnings in one’s jobs, or skills and qualifications that are
and are the most satisfied with their jobs. They below what is needed to perform the tasks and
generally feel that it is very unlikely that they will duties in a job, is also associated with significantly
lose their job in the next year. Compared to those higher levels of job insecurity (10). All groups of
who are perfectly matched, overqualified tertiary mismatched employees, whether they suffer from
education graduates receive lower gross qualification or skill mismatches, or their skills are
earnings, although having higher skills than in deficit relative to the optimal productivity thresh-
needed by the job is associated with higher mean old of the jobs, rank their levels of satisfaction
wages. The genuinely overqualified (overqualified significantly lower relative to comparable matched
and overskilled) suffer from the highest levels of employees (Figure 18) (11), (12).
job dissatisfaction, while the underskilled are more Earning functions reveal that the genuinely
likely to worry about their skills becoming outdated overqualified suffer from a 22% wage penalty
in the foreseeable future. relative to perfectly matched employees
EU workers employed in jobs demanding (matched in both qualifications and skills), despite
higher qualifications or skills than their own are having the same level of education (Table 9). By
significantly more likely to have greater levels of contrast, the apparently overqualified have 18%
perceived skill obsolescence, a consequence of lower gross hourly wages, on average. The lower
the fact that these individuals are employed in wage penalties of overqualified workers whose
jobs of higher than average skill intensity (Figure skills are matched to their job imply that
17). While 33% of underskilled employees think employers adjust wages according to individuals’
(10) The survey showed that respondents in Greece and Cyprus had the greatest levels of job insecurity in 2014 (32% and 28% said it
was likely that they would lose their job in the next year, compared to 13%, on average, in the overall sample of EU employees), while
respondents in western Europe were the most confident about keeping their jobs.
(11) All raw findings reported in this section have been validated in a multivariate regression framework that examines the independent
influence of qualification and skill mismatch on the skills obsolescence/job insecurity/job satisfaction scales taking into account several
individual and job characteristics as well as the skill intensity of jobs.
(12) For example, seven in 10 respondents (73%) who said their skills are matched to what is required by their job were satisfied with it
compared to around 60% of respondents who thought their skills are either above or below what is required. Just over one in 10
respondents who thought they are overskilled or underskilled said that they are dissatisfied with their job (14% and 12% respectively),
compared to one in 20 who reported that their skills are matched (6%).
Medium
BE MT 44% 27%
29% 60%
100%
40%
80%
Medium CHAPTER 2
47
60%
Qualification and skill mismatches in the EU workforce
40%
Talent unnourished
potential to grow skills in job
Talent in dead end
limited potential to grow skills in job
Talent in stand still
limited potential to grow skills in job 20%
Figure 17. Impact of –current skill mismatch on likelihood ofequal/lower

skills skill
obsolescence,
deficit than averageadult
0.80
potentialemployees,
to grow skills in2014, EU-28limited potential to grow skills in job
deficit than average deficit than average – –
job limited potential to grow skills in job 0.70
0.60
35% 0.80
0.50
30% 0.70
0.40
25% 0.60
35% 0.30
T PT BE 0.50
20% 0.20
30%
0.40
0.10
15%
25% 0.30
PT BE
0.00
10%
20% 0.20 EL
5% 0.10
15% 60%
0.00
0%
10% EL
Overskilled Matched skills Underskilled Talent Talent Talent 50%
5% in dead end in standstill unnourished
60%
NB: 40%
0% Average share of respondents who said that it is likely or very likely (i.e. a score above seven in a 0 to10 scale) that several of their skills will become
outdated in the next five years. 50%
7.6
Overskilled Matched skills Underskilled Talent Talent Talent
Figure 18. Impact of skill mismatch on job satisfaction, adult employees, 2014, EU-28
7.4
40%
7.2 20%
7.6
7.0 30%
7.4 10%
6.8
7.2 20%
6.6
0%
7.0
6.4 20
10%
6.8
6.2
6.6
6.0 0%
6.4 Overskilled Matched skills Underskilled Talent Talent Talent 60% 201
6.2
50%
6.0
% 35% Overskilled Matched skills Underskilled Talent Talent Talent 60%
in dead end in standstill unnourished 40%
NB: Average job satisfaction score on a 0 to10 scale, where 0 = very dissatisfied and 10 = very satisfied. 50%
24-65 Source: Cedefop ESJ survey. 30%
35%
productivity and skills (OECD, 2013). This is also contrast, the wage premiums of underqualified
40%
evident by the fact that underqualified workers, workers partially reflect their employment in more
20%
who nevertheless possess the required skills to skill-intensive jobs, with the residual wage
24-65 30%
do their jobs, are more likely to enjoy positive difference arising due to their higher stock of
10%
relative wage premiums. Underqualification is (non-formal and informal) skills and abilities. Even
20%
generally associated with higher wages among individuals employed in jobs with
0%
compared to those of workers who are well- comparable skill intensity, significant wage
AT BE
10%
matched in their qualifications and skills, although differentials (in the order of 8 to 10%) continue to
the70effect of underskilling isLU not statistically be observed between individuals mismatched in
80 0%
50%
significant, on average, once the (typically higher) their qualifications and the perfectly matched.
AT BE
IE
skill
50intensity of their jobs is taken into account. SE Part of ATthe significant wage penalty of
FR BE 40%
60 DK
70A significant part of the LU wage penalties of DE overqualified tertiary graduates also arises
Labour productivity
80 ES 30%
FI 50%
overqualified workers, especially those with FRbecause ofPTtheirHR own job preferences, EE CZsince these
NL IT
40 UK IE SI
BE EL 20%
medium-level education, can be attributed to individuals areHUless likely to select jobs on the
LT
60 SK
40%
30 DK
being in industries, occupations and NLjobs ITwithFI basis of suitability with their own skills and
Labour productivity
PL 10%
50 SE ES LV
20
AT RO 30%
lower
10 skill content (Mavromaras et al., 2015b). By qualifications HR (Mavromaras LT et EE al., 2015b).
DE BG
40 UK SI 0%
EL 20%
PT
SK
CZ Technica
300
PL HU 10%
LV
20 10 12 14 16 18 20 22 RO 24
BG
Skill deficit 0%
10
Technica
0% 45%
0
7.6
30%
7.4
7.2 20%
7.0
48 6.8Evidence from Cedefop’s European skills and jobs survey
10%
6.6
0%
6.4
Table 9. Impact of qualification-skill mismatch on gross hourly earnings, adult employees,

6.2
2014, EU-28 Matched skills

6.0
Overskilled Underskilled Talent Talent Talent 60%
in dead
Job requires upper secondary end
degree in standstill
Job requires unnourished
tertiary education degree
50%
Upper secondary graduate Matched qualification Underqualified
30% 35%
Overskilled 14.56 17.46 40%
Matched skills 13.76 16.51

24-29 24-65 30%
Underskilled 14.95 19.09
Tertiary education graduate Overqualified Matched qualification 20%
Overskilled 17.64 18.32

10%
Matched skills 14.58 18.56
Underskilled 17.69 16.67 0%
Instead, they place a higher premium on jobs with secondary graduate in a job that is
more70 security, a trend which isLUexpected to have commensurate to IEhis/her level of qualification.
80
50%
been60 accentuated in recent years as many Skill mismatches do not only impose costs on
FR BE
individuals will have sought for job security to individuals ATand enterprises, but also have
40%
DK
Labour productivity
cope with the high levels of economic uncertainty DE negative consequences for economies at macro
50 SE ES 30%
FI
caused by the economic crisis. These findings level. Figure 19 illustrates thatEE CZthere is a
NL IT
40 UK SI
EL LT 20%
also highlight the importance of high quality significant negative rawLVcorrelation, in the order
PT HR
SK
30
career guidance and counselling in terms of its of -0.75, between the incidence of BG skill deficits
PL HU 10%
20 RO
potential to inform better the motives and job that affect the employee populations in EU
0%
choices of individuals. The overqualified also face countries and the cross-country variance in levels
10
greater financial constraints relative to the of labour productivity. Overall, the correlation
0
10 12 14 16 18 20 22 24
perfectly matched, an additional factor that implies that EU Member States that can lower the
Skill deficit
accounts for their lower average wages. average level of skill deficit in their workforce (for
40% 45%
Even though the survey highlights significant example by means of additional or better quality
productivity loss associated with the CVET) can potentially enjoy significant benefits
underutilisation of the skills of higher educated in enhanced labour productivity. The strong
graduates, higher levels of education and skills negative relation shown in the chart confirms that
among individuals are valued by employers in the the measure of skill deficit collected as part of the
labour market. For instance, while overqualified ESJ survey is a good indicator of labour
tertiary education graduates (those in a job productivity, and explains more than half of its
requiring only an upper secondary qualification or variance across EU countries. It highlights that
less) receive, on average, an hourly gross wage EU Member States that fail to explore fully the
equal to EUR 15, this is higher than the potential of their skilled workforce will be
respective mean wage of matched upper hampered by significant productivity losses and
secondary graduates in a comparable job (which can suffer from reduced ability to explore
is equal to EUR 13.7). This wage difference is efficiency gains in the production process or to
greater when the overqualified tertiary graduate engage in innovation (OECD, 2013).
also has surplus skills relative to what is needed
by the job (EUR 17.6). Similarly, an underqualified
upper secondary graduate (employed in a job
requiring tertiary qualifications) earns on average,
less than well-matched workers in similar jobs
(tertiary graduates in jobs in need of a tertiary
education level). However, the earnings of the
former are greater than those of an upper
40%
24-65 30%
20%
CHAPTER 2
10%
0%
Figure 19. Skill deficits and labour productivity, 2014, EU-28

AT BE
80
50%
70 LU
IE
FR BE 40%
60 DK
Labour productivity
50 SE ES
AT 30%
IT FI
NL DE
40 UK SI
EL
HR LT 20%
PT EE CZ
SK
30
PL HU 10%
LV
20 RO
BG
0%
10
Technica
0
10 12 14 16 18 20 22 24
Skill deficit
% 45% NB: Labour productivity is calculated as GDP PPP (estimates) divided by employment and working hours. Skill deficits are derived as the difference between
the level of skills needed by an employee to do his/her job as well as possible (equal to a score of 100) and his/her own assessed skill level relative
to that benchmark. Country averages of skill deficits are displayed in the chart.
Source: Cedefop ESJ survey; World Bank (development data group).
42%
42% of EU workers had few opportunities
to find jobs
suitable for their skills and qualifications
when searching for a job
38% 52%
Pre-crisis Post-crisis
labour market labour market
entrants entrants
(2000-07) (2011-14)
The legacy of the crisis:

CHAPTER 3
un(der)employment and skill mismatch
3.1. Graduate entry to the such as the offer of jobs with undesirable job
labour market conditions, inefficient talent management and
recruitment practices and geographic
The intense job destruction resulting from the unsuitability.
2008 economic crisis, and the general rise in the However, Cedefop (2015a) has also shown
duration of unemployment, exacerbated concerns that genuine shortages in the supply of skilled
about deskilling affecting millions of unemployed labour can be a constraint for Europe’s most
European citizens. Rising levels of skill mismatch innovative and dynamic enterprises, which
were also fuelled by limited job creation, which compete in international markets. Skill shortages
compromised the job prospects of lower- are more likely to have constrained the
educated individuals while prompting higher productivity potential of a subset of EU
levels of graduate overqualification, coupled with economies during the economic downturn
significant restructuring in crucial economic (Germany, Poland, Slovakia, UK), considering
sectors of the European economy (including that low or falling unemployment rates have
construction and manufacturing). Evidence of a coincided with rising difficulties to fill vacancies by
small rise in European structural unemployment, employers (European Commission, 2015b).
as also reflected by the outward shift in the EU However, in several other EU countries (Bulgaria,
Beveridge curve, is said to illustrate rising Greece, Italy) there is growing divergence in the
matching inefficiencies in European labour propensity of employers and individuals to find
markets. Part of these matching frictions is suitable skills and jobs. This signifies the
related to a growing imbalance between demand existence of deeper or structural matching
for and supply of skilled labour (ECB, 2012; barriers in these economies, which tend to be
European Commission, 2014a). two-sided: educated workers are failing to find
Even during times of high unemployment and employment that is suitable for their skills and
low vacancy creation in many EU labour markets, qualifications while employers are failing to find
about four in 10 establishments in the EU claimed labour with desired skills.
that they could not find labour with the right skills In addition to concerns about skill shortages,
(Eurofound, 2013). Employers often claim that the ESJ survey cautions that a greater share of
they encounter difficulties in finding not only the most recent cohorts of EU graduates has
individuals from particular professions (such as accepted jobs that need lower qualifications than
skilled trades workers, metal machinery and their own. A raw comparison of the perceived
sheet workers, engineers, health and ICT qualification match between individuals who
professionals, accountants and managers) but graduated post 2008, after the onset of the global
also for ‘job ready’ graduates and/or candidates financial crisis, with earlier cohorts of graduates,
who possess the right mix of technical and soft reveals that the former are more likely to have
skills (European Commission, 2014d; Manpower, entered into their first jobs as overqualified
2014). Cedefop (2015a) has shown that a workers relative to the latter (Figure 20) (13). The
significant part of such difficulties is not genuinely probability of overqualification has risen from
related to an absence of skills on behalf of job about 0.15 to 0.17 in the pre-crisis period to 0.28
applicants but reflects numerous other influences, for recent (post-2008) graduates.
(13) The analysis refers to individuals who have graduated from at least an upper secondary level of education, although the main empirical
conclusion, that a higher share of recent graduates have accepted jobs requiring lower qualifications than their own, holds even when
restricted to tertiary education graduates only.
Figure 20. Overqualification of first job 3.2. Return of the unemployed

entrants by graduation cohort,
to work
2014, EU-28
Across Europe, the financial crisis and sluggish
economic recovery have caused unemployment
% First job entrants Probability of overqualification 90%
to increase to over 10% (Eurostat, 2014),

0.30 30%
affecting the overall composition and quality of

Europe’s labour supply. ESJ survey data show
0.25 25% 80%
that 18% of adult employees in Europe were
Mean % VET
unemployed or not working before starting their
0.20 20%
current job. About 20% of employees who began

70%
their current employment in the post-crisis period

0.15 15%
(2011-14) were previously unemployed as

opposed to 13% during the years of the crisis
0.10 10% 60%
(2008-10) and 10% in the pre-crisis era. Recent

labour market entrants in Greece and Spain are
0.05 5%
the most likely to have been unemployed before

50%
starting their current job (40% and 32%,

0.00 0%
respectively), illustrating the impact of the deep

1991-2000 2001-2007 2008-2014
Underskilled
recession in these countries. In Greece, those

NB: Sample of overqualified individuals with above upper secondary
who were previously unemployed Previously

are theemployed
most
qualifications whose current job is their first; overqualified workers
likely to have been unemployed for more than a

have a higher level of education attainment relative to the self-
reported level actually needed to do the job. Cohort of graduation is Previously unemployed
year (58% compared to 42% for the EU).

45%determined by the question: ‘In what year did you achieve your Talent in
Reintegration of the unemployed back into

highest level of education?’ The probability of overqualification is
40%based on estimates of a probit regression that controls for individual
the labour market is a significant challenge and

and job characteristics to ensure that comparable individuals are
highlights the need to invest in the prevention of

35%contrasted across cohorts. Talent un
skill mismatches not only at the time of their job

30%
25% The increasing probability of overqualification (re-)entry but also during their jobs, after they
among recent labour market entrants is have returned to the labour market. Given that the
Talent in
statistically significant and holds even after taking unemployed, and in particular those unemployed
20%
into
15% account several observable factors that could for more than one year before starting their
confound the divergent overqualification rates current job (LTU), are likely to have experienced
between separate graduate cohorts. For some skills atrophy (or lack of further skill
10%
example, the estimated probability is robust even development) during the time they were outside
accepting that the average graduate may have the labour market, they are more susceptible to
LV LT MT EU-28 5%
different
LT LV demographic andIE socioeconomic having
DE BGlower skills
HU than
GR BEneeded
PT DK by SI their jobs
NL and
0%
characteristics that vary depending on the period will consequently be in need of continuous
EE SK MT CZ CY FI RO LU SE IT HR FR PL AT ES UK EU-28
of graduation (14). Such observable (yet non- learning so that their skills can catch Unemployed
up with
exhaustive) factors in the ESJ survey include the changing job requirements.
e operators Other
years of employer tenure of an individual The ESJ survey confirms that, on average in
50%
42
employee, age at start of the first job, gender, the the EU-28, previously unemployed individuals
erators
type of contract offered, career progression since returning to the labour market, particularly those
40%
the
30% start of the job, as well as the overall skill in long-term unemployment, are more likely to
47
intensity of the job (captured by the type of suffer from underskilling at the time of job entry,
occupation and industry of the employee as well relative to those moving from one job to another
47 20%
as the complexity of job tasks). (Figure 21). The chances of previously

erators
10%
(14) It is acknowledged that comparisons

Teamwork of skill mismatches of different cohorts Learning
of graduates who madeLiteracy
the transition to their firstICT
job at
0%
distinct time intervals are likely to be hampered
languagesby varying population
solvingcompositions and economic circumstances over time. Memory
Technical Communication Foreign Customer Problem Planning Numeracy
bias may also affect the responses to the survey, particularly in relation to questions about the respondents’ skill gaps at the time of
entry into the job.
% few opportunities to attend job interviews E LTU
DE
FR
UK
Mean
70%
0.15 15%
0.10 10% 60%
CHAPTER 3
0.05 5%
50%
0%
0.00 0%
Figure 21. Difference in shares of underskilled employees at job entry by prior labour market
1991-2000 2001-2007 2008-2014
status, adult employees, 2014, EU-28

erskilled
Previously unemployed Previously employed

45% Talent in standst
40%
35% Talent unnourishe

30%
25%
Talent in dead en
20%
15%
10%
MT EU-28 5%
0%
LT LV EE SK MT CZ CY IE FI RO LU SE IT HR FR DE BG PL AT HU GR BE PT DK SI ES UK NL EU-28
NB: Responses to the question: ‘When you started your job with your current employer, to what extent would you say that some of your skills were lower
ators than required to do the job at that time and needed further development?’ Other Unemployed
Source:
unemployed persons reporting that they were that a past episode of unemployment manifests
underskilled when they started their job are in higher raw skill gaps in technical and generic
40%
highest in the Baltic states, the Czech Republic, skills, which employers also tend to identify as the
7
Malta, Slovakia and lowest in Spain, the primary reasons for the difficulties they face in
30%
Netherlands, Slovenia and the UK. filling their open vacancies (Manpower, 2014).
7
The ESJ survey also asked respondents to This poses a significant challenge for training
20%
rank how well matched a set of their specific skills programmes geared towards mitigation of skill
10%
is0%to the level that is needed to do their jobs. The mismatches of the unemployed (for example as
selected skills covered a broad
Foreign range
Customerof part ofLearning
active labour
Planning market policies or other ICT
adult
cognitive, digital and transversal skills, identified learning courses), as it implies that they need to
Technical Communication Teamwork Problem Literacy Numeracy
languages solving
in the literature as aggregate groupings that strike a balance between the endowment of a
effectively summarise or capture the variance of right blend of job-specific and soft skills to prepare
a wider range of different tasks performed by job seekers better for their return to the world of
DE
individuals as part of their jobs (Mañé, 2013; work (Cedefop, 2015c).

FR
Table 11 for details). The survey highlights that reintegration of the

UK
IT Figure 22 contrasts the mean shares of unemployed is a significant challenge for Europe,
SE
individuals in the EU with skill gaps in their current given that they are likely to possess specific
characteristics relative to other segments of the
EL
job
PL according to whether they were unemployed
CZ
prior to finding their jobs or not. It is clear that a adult working population. For instance, they are Humaniti
spell of unemployment is not associated with more likely to be lower educated, non-natives and
NL
greater cognitive and digital skill gaps (in literacy, are generally faced with considerable financial
DK
numeracy, ICT and foreign language skills) for constraints while searching for employment. They
HU Economics, bus
are more likely to suffer from skills obsolescence

ES
those who returned to employment, relative to

AT
those who did not experience a period out of and increased risk of stigmatisation (such as
BE
work (15). Nevertheless, the ESJ data highlight having fewer opportunities to attend job
IE
SK
FI
(15EE) Part of the reason for this is that a greater share of previously unemployed workers are employed in elementary jobs as well as in
PT Agriculture a
RO clerical and sales occupations and it is more likely for individuals to think that their level of skills is higher than required in lower-skilled
Teacher training
occupations.
LT
CY Security, transpo
NL LU SI
BG Medic
LV
LU
30%
Previously unemployed Previously employed
25%
45% PREVIOUS CONTENTS NEXT Talent in
Talent in
20%
40%
15%
35% Talent unn
30%
10%
Figure 22. Raw difference in shares of employees with low-ranked skills relative to those
V LT MT EU-28 25%
5%
Talent in
LT LV required
EE SK MT for
CZ their
CY IE job
FI by
RO past
LU SEunemployment
IT HR FR DE BGstatus
PL AT and specific
BE PT type
DK SIofESskill, 2014,
0%
20%
EU-28
HU GR UK NL EU-28
15%
operators 10% Other Unemployed
50%
V 42
LT MT EU-28 5%
erators
40%
0%
47 LT LV EE SK MT CZ CY IE FI RO LU SE IT HR FR DE BG PL AT HU GR BE PT DK SI ES UK NL EU-28
30%
operators Other Unemployed
47 20%
rators 50%
42
erators 10%
40%
47
0%
30% Technical Communication Teamwork Foreign Customer Problem Learning Planning Literacy Numeracy ICT
languages solving
47 20%
NB: Share of respondents below the lower quartile of the distribution of responses to the question: ‘How would you best describe your skills in relation to
rators what
% fewisopportunities
required to todoattend
your job
job?’ (corresponding to scores below 6 in a 0 to 10 scale, where 0 means that the level of skills is a lot lowerEthan required
interviews LTU
10% and 10 a lot higher than required).
DE
FR
Figure 23. Lack of opportunity to attend job interviews by pastPlanning

labour Literacy
market Numeracy status and ICT EU
UK
0%
Member State, adult employees (aged 24-65), 2014

SE Technical Communication Teamwork Foreign Customer Problem Learning
languages solving
IT
EL
CZ % few opportunities to attend job interviews E LTU
PL
DE H
NL
FR
DK
UK
HU
SE Econom
ES
IT
AT
EL
BE
CZ
IE
PL
SK
H
NL
DK FI
PT
HU Agric
Econom
EE
ES Teacher t
RO
AT
LT
BE
CY
IE Security,
CY NL LU SI
SK
BG
FI
LV
PT Agric
LU
EE
MT Teacher t
RO
HR
LT
CY 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Security,A
CY NL LU SI
%
BG
LV Supply of
100%
LU
MT Wa
80%
HR
60%0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% A
%
40%Responses to the question: ‘Did any of the following circumstances apply before you started working for your current employer? Despite sending
NB: Re
d still 20%
many job applications, I had few opportunities to attend job interviews; E = employed in another job before starting
Previously LTU current job; LTU = unemployed
Previously employed
Supply of
100%for more than one year before starting current job.’ Acco
d to those
Source:
0% Cedefop ESJ survey. Wa
oday and 80% Task complexity Skill intensity Changing task complexity Improvement in skills
an average –
skills in job 60%
Raw difference Adjusted difference Financia
40%
0.80 Re
50%
40%
7
CHAPTER 3
30%
7 20%
10%
interviews, despite sending many applications) in jobs with lower job complexity and skill intensity
(FigureTechnical
23). The previously unemployed would relativeLearning
tothose who were Numeracy
previously in
0%
benefit from better career guidance and employment (Figure 24). The disadvantaged
Communication Teamwork Foreign Customer Problem Planning Literacy ICT
languages solving
counselling, given that they are more likely to starting point of the previously LTU who return
have selected their mismatched job for reasons back to work is expected to perpetuate over time,
ofFR job security, as opposed to its suitability for given that the skill formation of individuals over
DE
their skills and competences. They also have a their lifespan is just as much a function of their
lower propensity to geographic mobility, such as past investments in human capital as it is
UK
moving between EU countries or even within dependent on the nature of skill needs and of
SE
regions of their own country. work tasks in their jobs (Desjardins, 2014).
IT
EL
PL The greater probability of underskilling among Figure 24 illustrates that, on average in the
CZ
workers who experienced a prior spell of EU, the skills of the previously unemployed
Humaniti
unemployment, mainly LTU, should be a concern returning to a new job are less likely to improve
NL
for
HU two reasons. during their job tenure, relative to those who were
DK
First, the LTU who have managed to find a previously employed in other jobs or graduates
Economics, bus
ES
job are likely to possess particular characteristics entering their first job. Statistically significant ‘skill
AT
(such as better employer networks) or better skills formation penalties’ for the previous unemployed
BE
that
SK distinguish them from the remaining total pool are found in nine EU Member States, most
IE
ofFI LTU individuals. It is therefore reasonable to notably in Greece, the Netherlands, and Slovenia,
expect that the latter group is likely to be but also in Bulgaria, Germany, Spain and the UK
characterised by significantly higher skill deficits (Figure 25). The finding is robust even when
PT Agriculture a
EE
that will inhibit their reintegration to the labour accounting for the different levels of skill
Teacher training
RO
market, posing even more significant difficulties mismatch between the two groups when they
LT
forSI public policy. started their new jobs, changes in the

CY Security, transpo
Second, according to the ESJ survey data, technologies and tasks required by their jobs over
NL LU
Medic
EU workers who experienced a spell of long- time, and whether they experienced any within-
BG
LV
lasting unemployment are also more likely to be job progression (such as promotion) (16).
LU
MT
0% 24. Difference
Figure in skill formation and skill 50%
needs of 60%
adult employees’ current jobs
90%by prior
HR
labour market status, 2014, EU-28

10% 20% 30% 40% 70% 80% 100% Adminis

Supply of gas/ele
100%
Water sup
80%
M
60%
40%
Retail, sa
l 20%
Accomodatio
se
d 0%
age –
job NB: % of employees at the top of each scale: above three in the scale of task complexity; above seven in the scales of skill intensity, changing task
complexity and skill improvement (17). Raw difference Adjusted difference Financial, insur
Source:
0.80 Cedefop ESJ survey.
Professional, sc
(16) The finding holds when employing different econometric methodologies (e.g. OLS, matching estimators, instrumental variables) and
0.70
model specifications that take into account a number of historical variables that preceded their differential labour market status. However,
Services rel
there are several changes that may have taken place over time in people’s jobs that have not been fully captured by the ESJ survey.
0.60
There is also considerable cross-country heterogeneity in the estimated effect. For details of the estimation methodology, see Annex 2.
0.50
) The scales in Figure 24 are derived as a Cronbach’s alpha scale coefficient based on the following items: Task complexity (zero to four
(170.40
S
scale) = how often, if at all, does your job involve the following? (a) responding to non-routine situations during the course of your daily
work; (b) learning new things; (c) choosing yourself the way you do your work. Skill intensity (0 to 10 scale) = how important are the
following skills for doing your work? (a) literacy; (b) numeracy; (c) ICT; (d) technical; (e) communication; (f) learning to learn; (g) teamwork;
0.30
(h) foreign language; (i) customer handling; (j) problem solving; (k) planning and organisation. Changing task complexity (0 to 10 scale)
63% 62%
0.20
= have the following increased, decreased or remained the same since you started your job with your current employer? (a) the variety
of tasks; (b) the difficulty of tasks; (c) the need to learn new things. Improvement in skills (0 to 10 scale) = compared to when you started
0.10
51
your job with your current employer, would you say your skills have now improved, worsened or stayed the same?
0.00
EL SI NL PT UK BG HU ES DE
60% 30%
% 100%
80%
Supply of
Wa
60%
40%
Re
d still 20%
Figure 25. Difference in skill formation within jobs between previously employed and
Accom
to those
unemployed workers, 2014, EU-28

day and 0%
n average –
skills in job
Raw difference Adjusted difference Financia
0.80
Professio
0.70
Serv
0.60
0.50
0.40
0.30
63% 62
0.20
0.10
0.00
EL SI NL PT UK BG HU ES DE
NB: Bars: difference in the skill improvement scale between previously employed and previously unemployed workers. Triangles: difference in skill
60%improvement scale adjusted for age, gender, native status, highest level of30%
education, skill mismatch status in previous job, employer tenure and
employer tenure squared. Responses to question: ‘Compared to when you started your job with your current employer, would you say your skills have
Talent 50%improved, worsened or stayed the same?’ 25%
Impact of the economic

nnourished
3.3. had few opportunities for job interviews (despite

crisis
40% 20%
sending many applications). Respondents in

T
Bulgaria, Greece, Cyprus, Hungary and Romania

30% 15%
were
10% most likely among all Member States to say
% ISC
Evidence from the ESJ survey indicates that the

rise in the share of overqualified young labour that they had few suitable job opportunities before
% ISC
20%
market entrants has been exacerbated by the accepting their current job. In the aftermath of the
% ISC
economic crisis. In the current economic climate, economic crisis a greater proportion (52%) of job
10% 5%
competition for jobs is particularly high and so it finders encountered unsuitable job opportunities
is possible 2011-14 people 2008-10
that felt that they 2000-07
needed to in relation2011-14
to their own 2008-10 qualifications2000-07
and skills
100%
0% 0%
accept jobs for which they were not ideally suited, before they accepted their current employment.
80%
particularly those who faced difficult personal As shown in Figure 26, thoseEmployed
who found their
situations. To try to understand the extent to current job after 2011 were more likely to receive
Education and training in another job 60%
which challenging personal circumstances might a limited set of suitable job offers relative to those
Talent 60%
hired before the crisis. The higher probability of

40%
impact upon people’s decisions to accept

nnourished
unsuitable jobs, the ESJ survey asked recent job finders finding few suitable jobs is
50% 20%
respondents about their situation before they statistically significant and remains so even if
started their current job in relation to family other important individual and job characteristics
40% 0%
obligations, whether they owned their main are included, which could affect their overall set
residence or had to pay a mortgage, and whether of available job opportunities. For example,
30%
they
20% had considerable financial difficulties before females and individuals who experienced a
10
starting their job with their current employer. It previous spell of unemployment are more likely
also asked respondents to assess the to have been confronted with a limited availability
8
macroeconomic landscape affecting them; for of suitable job possibilities. The same holds for
10%
example, whether there were few job those faced with particular constraints that may
6
IEhaveIT LTinhibited
LU LV MTtheir NL PLability
PT RO toSE search
SI SK UKmore
opportunities for people with their skills and
0%
4
qualifications and whether they had opportunities extensively for a suitable job, including family
AT BE BG CY CZ DE DK EE EL ES FI FR HR HU EU-28
to attend job interviews. obligations, financial constraints, home-

2
ownership or labour market immobility.

24-29 55-65
On average in the EU, 42% of respondents

50%
felt in 2014 that they had few opportunities to find Financial distress and declining household
0
jobs suitable for their skills and qualifications at incomes at the time of job entry have also
40%
the time of their job search. Over a third (35%) weighed more heavily on the job choices of
30%
20%
CZ
10%
RO
0%
Technical Communi- Teamwork Foreign Customer Problem Learning Planning and Literacy Numeracy ICT
CHAPTER 3
recent labour market entrants, forcing individuals indicated in the survey that their skills were lower
to accept mismatched jobs. Around one in five than needed at the time of entry to their job,
respondents (19%) faced considerable financial relative to previous cohorts. Part of the smaller
difficulties before entering into their current job. probability of underskilling among recent
Employees in eastern and southern Europe were graduates can be attributed to the fact that the
more likely to face financial difficulty than those jobs created post-2011 are more likely to be
in northern and western Europe (30% and 23% precarious or of lower quality. However, even
compared to 16% and 13%, respectively). taking into account differences in skill intensity
Employees from Germany and Luxembourg were and in other job characteristics (type of contract,
least likely to have faced such adverse labour part-time job, small firm, type of industry and
market circumstances. Generally, overqualified occupation) between the graduate cohorts, no
workers are more likely to say that they had statistically significant differences in rates of
financial difficulties prior to finding their job. underskilling are found.
The ESJ findings fail to support the argument
that higher recent qualification mismatches can
be attributed to deteriorating skills quality of
recent graduates. As shown in Figure 27, a
smaller share of recent graduates and job finders
Figure 26. Share of adult employees with few Figure 27. Average share of adult employees
opportunities to find a job suitable (aged 24-65) underskilled at the
for their skills and qualifications by start of their job by period of job
period of job entry, 2014, EU-28 entry, 2014, EU-28
660% 3
30%
t 50% 25%
hed
40% 20%
30% 15%
20% 10%
10% 5%
0% 0%
2011-14 2008-10 2000-07 2000-07 2008-10 2011-14
Source: Cedefop ESJ survey. NB: Share of adult employees with positive responses to the question:
‘When you started your job with your current employer, overall, how
Employed in another job
would you best describe your skills in relation to what was required
t to do your job at that time? Some of my skills were lower than
hed required by my job and needed to be further developed. The period
of job entry is determined by answers to the question: ‘How many
years in total have you been working for your current employer?’
31%
31% of EU employees
who made a transition from school to their first job
had lower skills than needed
when they started their job
LU
MT
HR
0% 10% 20% 30% PREVIOUS
40% CONTENTS
50% NEXT
60% 70% 80% 90% 100% Administ

Supply of gas/ele
100%
Water sup
80%
Tackling skill mismatches among the

M
CHAPTER 4
60%
young with high quality VET

40%
Retail, sal
l 20%
Accomodatio
e
d 0%
ge –
job
Underskilling at labour
Raw difference Adjusted difference Financial, insur
4.1. The ESJ data show, as expected on the basis

market entry
0.80
of their limited work experience, and given that

Professional, sci
0.70
education and training systems cannot prepare

Services rel
Following the outbreak of the financial crisis, young people fully for any job they take on entry
0.60
several Member States (including Bulgaria, to the job market, that levels of underskilling for
0.50
S
Greece, Spain, Italy and Cyprus) experienced young people (in particular, females and the low-
0.40
markedly higher rates of youth unemployment qualified) at the start of their first job tend to be
0.30
and of young people who were neither in higher than of the average employee. A total of
63% 62%
0.20
employment nor in education or training (NEETs). 31% of respondents whose current job is their
This
0.00 raised concerns about the ability of their first and 27% of those who were in education or
0.10
51
education ELand trainingSI systems to

NL
integrate PT
young training before
UK BG
they started
HU
theirEScurrent job
DE
said
individuals effectively into the labour market. that their skills were lower than required when
Much attention has been paid to their weaker they started (Figure 28). This is significantly
60% 30%
VET
50% systems (relative to countries with dual higher than the average incidence of underskilling
training systems) and the low incidence of at job entry for the whole sample (22%) and
t 25%
apprenticeships and WBL in these countries confirms that a significant share of people making
hed
(Cedefop, 2012a; Eichhorst et al., 2012). Some the transition from school to the labour market do
40% 20%
have argued that the weaker investment in not

15% possess the level of maturity in their skills
Took plac
graduates’ skills in these countries may be related required by their jobs.

30%
to why employers report difficulties in finding the

% ISCO 1 Man
10% Figure 29 indicates that younger employees

right talent, particularly why they report technical are more likely to give a lower ranking to their
% ISCO 2 Prof
20%
and
10% soft skill deficits among young graduates technical skills (specialised knowledge to perform
% ISCO 3 Asso
(McKinsey Centre for Government, 2012). job tasks and handle technical equipment)
5%
Raw difference
Figure 28.2011-14 in shares 2000-07
of underskilled employees at job entry2008-10
by prior labour market
100%
0% 0%
status, 2014, EU-28

2008-10 2011-14 2000-07 80%
60%
Education and training Employed in another job
t 60%
40%
hed
50% 20%
40% 0%
30%
10
20%
8
10%
6
0%
AT BE BG CY CZ DE DK EE EL ES FI FR HR HU IE IT LT LU LV MT NL PL PT RO SE SI SK UK EU-28 4
NB: Responses to question: ‘When you started your job with your current employer, to what extent would you say that some of your skills were lower than 2
24-29 55-65
required to do the job at that time and needed further development?’
50%
Source: Cedefop ESJ survey. 0
40% Improve
30%
20%
40% 0%

30%
60
10
Evidence from Cedefop’s European skills and jobs survey
20%
8
10%
6
Figure
0%
29. Raw difference in shares of employees with low-ranked skills relative to those
required for their job by age group and specific skill type, 2014, EU-28
AT BE BG CY CZ DE DK EE EL ES FI FR HR HU IE IT LT LU LV MT NL PL PT RO SE SI SK UK EU-28 4
2
24-29 55-65
50%
0
40%
30%
20%
CZ
10%
RO
0%
Technical Communi- Teamwork Foreign Customer Problem Learning Planning and Literacy Numeracy ICT
cation language handling solving to learn organisation
24
NB: Share of respondents below the lower quartile of the distribution of responses to the question: ‘How would you best describe your skills in relation to
what is required to do your job?’ (corresponding to scores below six in a 0 to10 scale, scale where zero means that the level of skills is a lot lower
than required and 10 a lot higher than required).
Tackling initial skill

relative to older workers, a greater share of whom 4.2.

report that their skills are higher than required.
mismatches with VET
The younger cohort is also more likely to exhibit
higher levels of underskilling in soft skills, such as The higher skill gaps of younger workers relative
communication, teamwork, customer handling to the adult workforce implies that significant
and problem solving. Those older are more likely investment in their further skills development is
to be underskilled in foreign languages, learning required to attain a better match between their
to learn (learning and applying new methods and skills and those needed in the labour market. This
techniques in your job; adapting to new can be addressed partly by improving their
technology, equipment or materials; engaging in education and training but, crucially, via
own learning) and in digital skills. continuous learning on the job.
The gap between the skills of younger VET is a key pillar of European education and
workers and those needed by their jobs during the training systems, although EU Member States
school-to-work transition also vary considerably differ in the attractiveness and emphasis placed
across the EU, reflecting differences in the quality on this sector as opposed to academically-
and responsiveness of education and training oriented streams. The contribution of VET,
systems but also in technologies and level of skills particularly work-based and apprenticeship-type
demanded in entry level jobs across distinct EU schemes, to fight youth unemployment, to ensure
economies. New labour market entrants from a better match between training and labour
Slovakia and the Baltic states experienced the market needs and to ease transitions to
highest skill gaps in contrast to those from employment is now more widely recognised at a
Luxembourg, Austria, Slovenia and Sweden. In pan-European level. As an indication of the
Slovakia, the proportion of respondents who were urgency of reforms in this sector, 22 country-
previously in education or training and who said specific recommendations adopted within the
that their skills were lower than required at the 2014 European semester exercise were related
start of their job is more than twice as high as the to VET (European Commission, 2014b).
EU average (56% compared to 27%). About seven in 10 respondents to the
Cedefop ESJ survey, holding at least an upper
secondary education (ISCED level 3 and above),
described their highest qualification as vocational
(70%) (18). The share of employees with a VET
(18) A vocational degree was defined in the survey as one that is ‘designed for acquiring knowledge, skills and competences closely linked
to a particular job or trade’.
CHAPTER 4
Figure 30. Average incidence of WBL versus share of employees with VET as highest
qualification, adult employees with at least upper secondary qualifications,
2014, EU-28
90%
DE 70
CZ 60
LU
80% FI
HU 50
PT
BG
Mean % VET
NL 40
EL PL
SI
HR CY
70% EU-28 DK
SE ES LV 30
LT IT AT EE
MT SK 20
BE FR RO
14
60% IE 10
SI
0
UK
50%
0% 10% 20% 30% 40% 50% 60% 70% 80%
Mean % WBL
NB: Employees with above upper secondary qualifications (ISCED 3-5); lines: EU-28 average (41% WBL; 70% VET); positive responses to questions: ‘Overall
would you describe your highest qualification as vocational i.e. designed for acquiring knowledge, skills and competences closely linked No VET VET
to a particular
job or trade?’ and ‘Did your study take place only within an education institution (e.g. a school, college or university) or did it involve some learning in
employed a workplace (e.g. apprenticeships, internships, other forms of WBL)?’
31% ISCO 3 Technicians
Source: Cedefop
Talent in ESJ survey.
standstill 25%
degree is highest in countries such as Czech 22% of medium-qualified employees that did not
ISC
Republic, Germany and Luxembourg and lowest gain skills closely linked to a particular43% job or
in Ireland, Slovenia and the UK (Figure 30) (19). trade as part of their studies were unemployed
ISCO
Talent unnourished
The ESJ survey shows that VET is before finding their first job, as opposed to 13%
46%
instrumental in ensuring excellence and inclusion of graduates

26% with a VET degree. A smaller
ISC
inTalent
education
in dead end and training. Medium-qualified proportion of VET
29% graduates are overqualified in
graduates with vocational qualifications are more their jobs and are more likely to be employed in
ISCO 8 P
likely to select their

0% jobs because10%
they are closely 20% jobs that need their specific40% qualifications.
ISC
suited to their skills and qualifications, as Although there is no significant evidence that the
30% 50%
opposed to reasons such as job security, skills of VET graduates as a whole lag behind
proximity to home, pay and benefits. As Table 10 those required by their jobs at the time of
shows, a VET degree can also ensure smoother recruitment, raw levels of underskilling among
Above 40%
transition from education to work and, in younger VET graduates tend to be higher
Between 30-40%
particular, entry into jobs that are well matched to (although this difference is not statistically
NL EU-28 Below 30%
their qualifications and skills. For these reasons, significant). A much lower share of individuals
ISCO 3 Technicians
VET graduates generally enjoy higher average with a VET qualification exhibits gaps in terms of
nemployed ISC
levels of satisfaction in their jobs and are less their specialised knowledge to perform their job
likely to fear losing their jobs in the near future. duties and to handle specialised technical
ISCO
Table 10 also shows that a significantly equipment. Nevertheless, a greater proportion is

smaller share of individuals with a VET degree (at found to have lower foreign language, literacy ISC
above upper secondary level) experienced an and numeracy skills than needed by their jobs
intermittent spell of unemployment during their compared to non-VET graduates.
ISCO 8 P
transition from school to their first job. A total of

ISC
(19) Respondents employed in industries such as services relating to education or health, construction or building, and manufacturing or
engineering are the most likely to describe their highest qualification as a vocational qualification (78%, 73% and 73% respectively).
The same holds for those working as craft and related trades workers (77%) and as technicians and associate professionals (75%).
ICT
By contrast, individuals working in financial, insurance or real estate services, cultural industries (arts, entertainment or recreation)
and retail sales, shop work or wholesale are the least likely to be VET graduates.
LTU
ISCO 3 Technicians
ISC
Table 10. Skill mismatch of adult employees by VET status, 2014, EU-28
Unemployment before first job Overqualified Underskilled at start of job

24-29 24-65 24-29 24-65 24-29 First job
non-VET 22% 22% 29% 22% 24% 28%
VET 13% 14% 15% 22% 26% 30%
NB: Sample of individuals with medium-level qualifications (ISCED 3 to 4).
4.3. Promoting all forms of WBL

Taking into account a number of relevant

correlated factors, VET graduates are also found
to have significantly smaller skill deficits in their Strengthening WBL (apprenticeships, intern-
jobs and are more likely to be closer to the full ships) and VET systems is a key element of the
productivity threshold. This is evident in Figure EU’s strategy to combat high youth un-
31, which illustrates that, on average, individuals employment, as confirmed by the new set of
with a medium-level VET qualification are more medium-term (2015-20) priorities in VET
likely to be in jobs where their skills are a closer (European Commission and Latvian Presidency
match to their current and potential skill of the Council of the European Union, 2015). The
requirements. rethinking education communication (2012a)
Overall, the findings of the ESJ survey stressed the need to invest in building world-class
provide strong evidence that vocational skills and VET systems and increase participation in WBL.
qualifications are a valuable route for mitigating The European alliance for apprenticeships, ‘youth
youth unemployment and qualification mismatch, guarantee’ and ‘youth employment initiative’ – all
90%
DE 70
in particular increasing rates of overqualification, launched in 2013 – confirmed the crucial role of
in European labour markets. However, the slightly HU VET in increasing the employability of young
CZ 60
LU
greater skill gaps of VET graduates at the start of people. Learning in the workplace is also an
80% FI
50
their (first) jobs are a continued reminder ofSI theCY EL effective PL way to retrain and upskill the
PT
BG
Mean % VET
NL 40
need for reform and the pursuit of quality ES LV unemployed and adults.
70% HR
assurance in the VET sector, in line with the CedefopEE (2015b) has reported on the
EU-28 DK 30
SE
revised medium-term priorities of the Bruges significant progress made at EU Member State
LT IT AT
MT 20
communiqué process (Cedefop, IE 2015b). Such level SIsince 2010, in terms of implementation of
SK
BE FR RO
60%
reforms must ensure that the skills of recent national policy initiatives aimed at meeting one of
10
students enrolled in VET programmesUK remain the most important strategic objectives of the
0
relevant0% and aligned to labour market needs. Bruges communiqué, the

60% promotion
70%of WBL 80%as
50%
10% 20% 30% 40% 50%
Figure 31. Skill mismatch groups by VET status, medium-qualified adults, 2014, EU-28
Mean % WBL
No VET VET
reviously employed 31% ISCO 3 Tec

Talent in standstill 25%
43%
Talent unnourished 46%
26%
Talent in dead end 29%
0% 10% 20% 30% 40% 50%

NB: Sample of individuals with medium-level qualifications (ISCED 3 to 4).
Above 40%
Between 30-40%
ES UK NL EU-28 Below 30%
employed 31% ISCO 3 Technicians
Talent in standstill 25%
ISCO
43% ISCO
Talent unnourished
CHAPTER 4
46%
ISC
26%
Talent in dead end 29% ISCO 8 P
Figure 32. Proportion of adult 10%

workers who completed study involving some workplace learning,
ISC
2014, EU28
0% 20% 30% 40% 50%
Above 40%
Between 30-40%
NL EU-28 Below 30%
ISCO 3 Technicians
nemployed ISCO
ISCO
ISC
ISCO 8 P
ISC
ICT
LTU
ISCO 3 Technicians
ISCO
ISCO
NB: All respondents who have completed upper secondary, post-secondary or tertiary education.
Source: Cedefop ESJ survey. ISC
part of IVET and OtherCVET. To find out more about Realschule. In contrast, respondents in Portugal
vocational learning andfinance

its influence on skill (91%), Lithuania (87%) but also Ireland (77%),
social sciences ISCO 8 P
mismatch, respondents with post-upper Malta (73%) and the UK (72%) have
Economics, business, law and
ISC
secondary qualifications were asked if their predominantly learned in a school-based setting.

Maths and stats
studies took place only in an education institution, Substantial differences in the incidence of
Natural sciences
such as a school, college or university, or if it WBL are also evident between different education
involved some WBL, such as an apprenticeship, programmes or broad fields of study. Over 67%
Computing sciences Problem-
internship, or some other form of WBL.
Agriculture and veterinary sciences of recent graduates (aged 24 to 34) from
Despite
training increasing
and education attention to and reforms medicine and health-related sciences acquired
Tea
involving WBL in EU Member Other fieldStates, most (56%) some experience in a workplace as part of their
Teacher science
adult employees in the EU (with at least an upper studies. At the other end of the spectrum, only
Commun
secondary level of education) responded that 25% did so among those who studied humanities,
L
they completed their studies only within an languages and arts, economics, business and
education institution (a school, college or law or other social sciences (Figure 33).
0% 10% 20% 30% 40% 50% 60% 70% Planning and organ
university). By contrast, 41% of adult workers said There is also a difference by industry when
that their studies involved some learning in a looking at the share of adult workers who had a
Te
% Took place only within an education institution % Involved some learning in a workplace
workplace in the broad sense. Stark differences WBL 60%experience as part of their education. Figure
in WBL exist between Skilledthe EU-28 (Figure 32). 34
55%shows that respondents who work in finance,
100% Administration and support services 36% Customer h
Over two thirds of adult workers in Germany have insurance or real estate, cultural industries or
agricultural 42%
Foreign la
completed some form of WBL, where the two-tier professional, scientific or technical services are
employed
Supply of gas/electricity or mining/quarrying 60% 39%
system emphasises vocational qualifications for the most likely to have completed38% study that took
Water supply/management/treatment 59%
individuals Manufacturing
attending Hauptschule
or engineering and place only within an education 44%
54% institution (67%,
Cultural industries
Construction or building 52% 45%
Servi
Retail, sales, shop work or wholesale 58% 39%
Accomm
Accomodation, catering or food services 53% 46%
Financia
racy ICT
64
E LTU
eracy ICT Evidence from Cedefop’s European skills and jobs survey ISCO 3 Tech
E LTU
Figure 33. Incidence of WBL by field of study, adult employees, 2014, EU-28
ISCO 3 Tec

Humanities, languages andstats
Maths and arts
Other socialsciences
Natural sciences
Economics, business, law and
Engineering finance
sciences
Maths sciences
Computing and stats P
Naturalsciences
Engineering
Teacher training and education sciences
science
Computing sciences
Other field CP
Agriculture
Security, andorveterinary
transport sciences
personal services
Teacher training andand
Medicine education science
health-related
Other field 0% 10% 20% 30% 40% 50% 60% 70% Planning an
NB: All respondents who have completed upper secondary, post-secondary or tertiary education.
Medicine
Source: Cedefop and health-related
ESJ survey. % Took place only within an education institution % Involved some learning in a workplace
Figure 34. Work-based learning by industry, adult employees, 2014, EU-28

90% 100% Administration and support services
0% 10% 20% 60% 30% 40% 50% 36% 60% 70% Planning
Cusa
Skilled agricultural 55% 42%
Fo
eviously employed % Took place only within an education institution % Involved some learning in a workplace
90% 100% Administration and support services 60% 36% Cu
Water supply/management/treatment 59% 38%
F
Manufacturing or engineering 54% 44%
reviously employed Cultural ind
Manufacturing or engineering 54% 44% A
Accomodation, catering or food services 53% 46% Cultural in
Construction or building 52% 45% F
Transportation or storage 54% 42%
ent in skills
ICT services 60% 38%
Adjusted difference
Financial, insurance or real estate services 67% 31%
Transportation or storage 54% 42% A
ment in skills Professional, scientific or technical services 62% 36%
Services relating to education or health 51% 47%
Adjusted difference Financial, insurance or real estate services 67% 31%
Cultural industries 67% 31%
Professional, scientific or technical services 62% 36% Pr
Social and personal services 48% 48%
Cultural industries 67% 31% Suppl

63% 62% 60% P
Social and personal services 56%
48% 48%
51% 52% 52%

NB: All respondents who have completed upper secondary, post-secondary or tertiary education. Sums 49% Supply, ma
47% 47%not equal to 100% exclude
48% missing responses.
Source: Cedefop ESJ survey. 45% 44% 45% Supp
DE 63% 62% Informati
67% and 62% 51%respectively). This is likely to reflect that involved some workplace learning (48%,
60% 38%
56% 35% 36% 36%
the greater level of theoretical knowledge 47% and 46% respectively). This may be
52% 52%
48%
49% Supply, m
required to enter employment in these industries. because roles in these industries tend to involve
47% 47%
45% 44% 45%
Informat
In contrast, those working in the social and more customer-facing interaction and therefore
DE
38%
35% 36% 36% 35%
personal services, services relating to education individuals are likely to need greater levels of
or health or in accommodation, catering or food practical experience before they start working in
30%
Took place only within an education institution Involved some learning in a workplace
services are more likely to have completed study the industry.

25%
35%
% ISCO 1 Managers % ISCO 4 Admin and clerical % ISCO 7 Craft and trades
20%
30%
% ISCO 2 Professionals % ISCO 5 Service and sales % ISCO 8 Plant and assemblers
% ISCO 3 Associate professionals % ISCO 6 Skilled agricultural % ISCO 9 Elementary 15%
25%
10%
Accomodation, catering or food services 53% 46% Cultural industries
Construction or building 52% 45% Financia
Transportation or storage 54% 42% Servi
ills Re
ICT services 60%
38% Accomm
Accomodation, catering or food services 53% 46% A
difference Financial, insurance or real estate services 67% 31% Financia
CHAPTER 4
65
Agricultu
62% Tackling skill mismatches among the young with
36%high quality VET
Transportation or storage 54% 42%
ills Professional, scientific or technical services Re
Ad
difference Financial, insurance or real estate services 67% 31%
Agricultu
Work-based learning
or health by occupation group,
51% adult employees, 2014, EU-28
Figure 35. relating
62% 36% Professio
Social and personal services 48% 48%
Services to education 47%
Cultural industries 67% 31% Supply of gas

63% 62% Profession
Social 60%
and personal services 48% 48%
56%
51% 52% 52%
48%
49% Supply, managem
47% 47%
45% 44% 45% Supply of gas
DE 63% 62% 60%
Information tech
38%
56% 35% 36% 36%
51% 52% 52%
48%
49% Supply, manageme
47% 47%
45% 44% 45%
DE Information tech
38%
35% 36% 36% 35%
30%
25%
35%
20%
30%
% ISCO 2 Professionals % ISCO 5 Service and sales % ISCO 8 Plant and assemblers
25%
NB: %
All ISCO 1 Managers
respondents % ISCO 4orAdmin
who have completed upper secondary, post-secondary and
tertiary clerical
education. % ISCO 7 Craft and trades
10%
Source: Cedefop ESJ survey. Noand
WBLassemblers
WBL 20%
% ISCO 2 Professionals % ISCO 5 Service and sales % ISCO 8 Plant
Figure 36. Transitions to first job by WBL and past labour market status, 2014, EU-28
100% 5%
82%
-07 77%
80% 0%
10%
No WBL WBL Un
nother job 60%
100% 5%
82%
-07 77%
40%
80% 0%
Humanities andUn
ed
20%
60% 15%
nother job 9% 6% 5%
0%
40% Social s
Education and training Unemployment Inactive Humanities and ed
20% 15%
9% 6% 5%
No WBL WBL
100% Social s
Education and training Unemployment Inactive
8 Sample of medium- and higher-qualified workers, whose current job is first job; categories do not sum to 100% because the graph excludes the
NB: No WBL WBL
previously self-employed and ‘don’t know’ responses.
106
Source: Cedefop ESJ survey. Health-related s
84 The ESJ findings confirm a relationship As with the possession of a VET qualification,
between how individuals acquire skills and the a WBL spell during one’s studies is positively
UK EU-28
skill
62 needs of their respective occupations. As associated with the likelihood of individuals Health-related s
shown in Figure 35, managers, professionals and making a direct transition from school to their first
55-65
clerical support workers are most likely to have job, as opposed to having experienced an
40
completed studies that took place only within an intermittent spell of unemployment or inactivity
UK EU-28
educational setting. Those who are employed as (Figure 36). On average, 82% of adults entered
Improve skills in job Skill intensity Job satisfaction Job insecurity Anticipated skills obsolescence
2 Accommodation,
55-65
craft and trades workers, plant operators and into their first job directly after education and
Cultural indus
assemblers
Improveand
skills inskilled agricultural
Skill intensityworkers are training if they Job had engaged in some form of
0
most likely to have completed studies that WBL, as opposed to 77% of those from a purely
job Job satisfaction insecurity Anticipated skills obsolescence Retail, sales
Accommodation,
involved some learning in a workplace. As before, school-based education. A total of 15% of the
this may be due to the higher level of practical latter group also entered into their first job after
CulturalAgricul
indus
and technical skills required by these latter having spent some time in the unemployment
ICT Retail, sales
Soc
occupations, while the former are in greater need pool, while only 9% of WBL graduates did so.
for theoretical or academic knowledge and This positive relationship between WBL and
Agricult
competences. the probability of making a direct school-to-first-

Information tech
ICT
Soc
Man
Administra
Information tech
Supply, managem
Man
Services relat
Fo
eviously employed
Manufacturing or engineering 54% 44%
Cultural ind

A
job transition holds even after taking into account suitable for their own skills and 42% qualifications.
F
that VET and WBL graduates may have entered They are also found to have higher (raw) levels
Transportation or storage 54%
ent in skills
their first jobs younger (due to a shorter cycle of of job satisfaction, are more content with their job
djusted difference
studies), or may have
estatebeen
servicesmore likely to security and believe that their skills are more
encounter adverse financial
services and family durable and have smaller chances36% of becoming
Financial, insurance or real 67% 31%
A
circumstances prior to job entry,

or healthrelative to those obsolete in the medium-term. 47%
Professional, scientific or technical 62%
with an academic/school-based education. It is Apart from aiding smoother integration into

Services relating to education 51%
also not affected by the fact that the two groups the labour market, individuals whose studies
of individuals withanda different academic orientation involved some WBL are also48%significantly more
Pr
Social personal services
may be characterised by a differential aptitude for likely be in jobs that form a good match for their
48%
learning and/or have different job preferences. level of qualifications and which entail greater
Given that some 60% individuals undertaking WBL levels of task complexity and skill intensity. Partly
Suppl
63% 62%
may have 51% not yet completed formal education, due to the greater skill demands of52% their (first and
56%
implying that some of the employees at the time current) jobs, 47% new labour 45%market entrants with a
52%
49% Supply, ma
of the survey may be in a job as part of their history of WBL are also more likely 38%to
47% 48%
45% 44%
Informatio
education programme, the analysis has used only acknowledge that some of their skills were lower
DE
35% 36% 36%
individuals on indefinite contracts. Since most than needed at the start of their first
WBL (apprenticeship) contracts that are part of employment, compared to comparable
formal education programmes are temporary, graduates who followed a purely school-based
restricting the analysis to include only permanent education. The higher skill gaps of those who did
35%
contract Tookworkers ensures that institution

the relationship some learning in some
a workplace may also arise
30%
between WBL and speedier entry to the labour because they may have acquired an
place only within an education Involved learning in a workplace
market
% ISCOof individuals after school is not affected informational advantage with regards totrades
the true
25%
by % the incidence of any temporary WBL nature of the skills demanded in8their workplace.
1 Managers % ISCO 4 Admin and clerical % ISCO 7 Craft and
20%
transitions. Empirical analysis using the ESJ data This would result in greater % ISCO 9self-awareness
ISCO 2 Professionals % ISCO 5 Service and sales % ISCO Plant and assemblers
confirms that the positive empirical relationship of about the true extent to which their skills are
% ISCO 3 Associate professionals % ISCO 6 Skilled agricultural Elementary 15%
WBL is robust even when restricting the analysis matched to their job tasks. Further, individuals
10%
to permanent contract holders. who followed separate forms of learning as part

No WBL WBL
The ESJ reveals 77% some additional benefits of of their formal education may also have different
100% 5%
WBL for the labour market situation and wellbeing (unobserved) traits, abilities, job networks or
82%
2000-07 80% 0%
of adult employees (Figure 37). On average, other characteristics correlated with their skill
respondents who completed studies that involved mismatch status at the time of job entry.
yed in another job 60%
some WBL had to devote less time searching for However, more extensive data explicitly focused
a job before finding their current employment and on the school-to-work transition of WBL
40%
Humanities
had more opportunities to find a job that is graduates would be required to decouple the
20% 15%
9% 6% 5%
Figure 37. Mean labour market outcomes andUnemployment

work-based learning status, 2014,Inactive
EU-28
0% S
Education and training
No WBL WBL
10
6 Health-re
SI SK UK EU-28 4
2
9 55-65
0
Improve skills in job Skill intensity Job satisfaction Job insecurity Anticipated skills obsolescence
Accommo
NB: Sample of medium- and higher-qualified workers; indices measured on 0 to 10 scale.
Source: Cedefop ESJ survey. Cultura
Reta
CHAPTER 4
exact source of their skill mismatches relative to 4.4. Importance of key

those following different learning tracks.
competences for VET
The data highlight that graduates whose
studies involved some workplace learning are 4.4.1. Fundamental key competences
more likely to enjoy a virtuous cycle in their (first) The importance of key competences (either
jobs. This may be because of the elimination of fundamental or transversal) (22), continues to grow
information asymmetries regarding their abilities in EU labour markets, as also revealed by the
or the establishment of greater trust between the ESJ survey. To understand how important
employer and employee (20). The fact that their fundamental skills are for the ability of adult
jobs are more conducive to learning and have employees to carry out their job tasks in
higher skill content feeds into their greater European labour markets, respondents were
propensity to have improved their skills in their asked to describe the highest level of literacy,
jobs (whether they are still in their first jobs or are numeracy and ICT skills they require. Literacy
in subsequent jobs), relative to those whose and numeracy skills were defined at two levels;
education took place solely within an education basic and advanced. ICT skills were defined at
institution. three levels; basic, moderate and advanced. The
The findings point to several directions in respondents were also asked to rank the
which WBL can be expanded further and on importance of a further list of transversal and
which policy-makers can concentrate their efforts technical skills required for their jobs. The list of
to succeed in meeting one of the key goals of the skills surveyed was selected on the basis that
revised medium-term priorities for VET (Riga they comprise aggregate bundles of a wide list of
conclusions), to promote WBL in all its forms. tasks performed by individuals as part of their
First, the data highlight that more progress is jobs (Mañé, 2013). A list of 11 broad groupings of
required in terms of making WBL a more skills were included in the questionnaire and
attractive option for males, who are found to be defined as shown in Table 11.
less likely than females to participate in studies About half the respondents (49%) in the
involving some learning in a workplace (21). survey said that they need advanced literacy
Despite greater attention paid to the merits of skills to be able to perform their jobs (Figure 38).
WBL in recent years, proliferation of WBL can It is interesting that a high level of advanced
only be achieved if it is integrated further in fields literacy skills is required by the occupation groups
without a WBL tradition, particularly in tertiary anticipated to grow the most in terms of
education. Related to this is the need for the employment (Cedefop, 2012d). About three
attractiveness and relevance of such types of quarters of individuals working as managers and
learning to be extended to non-conventional professionals and over a half of clerical support
occupations and industries. The survey findings workers and technicians and associate
also confirm that significant efforts need to be professionals need advanced literacy skills for
made to expand WBL to private sector and their jobs, according to the survey (Figure 39).
smaller-sized establishments, an objective that Most of the respondents (58%)
requires significantly more involvement and acknowledged that they need to be able to carry
commitment by employers and other social out calculations using decimals, percentages or
partners. fractions and to be able to understand tables and
(20) Unfortunately, it is not possible to ascertain from the survey whether the respondents found their first jobs at the same employer where
their spell of WBL took place.
(21) The fact that males are less likely to participate in WBL than females holds even after taking into account that the two sexes tend to
study different fields (e.g. females are more likely to enrol in the fields of humanities and arts, while males in engineering) or enter into
different industries and occupations.
(22) As defined in the corresponding recommendation from 2006, the eight key competences consist of communication in the mother
tongue, communication in foreign languages, mathematical competence and basic competences in science and technology, digital
competence, learning to learn, social and civic competences, sense of initiative and entrepreneurship, cultural awareness and
expression. The recommendation also refers to fundamental skills of language, literacy, numeracy and ICT as essential foundation for
learning. Learning to learn, social and civic competences, initiative-taking and entrepreneurship, and cultural awareness and expression
are considered ‘transversal key competences’ (Council conclusions 2010 on competences supporting lifelong learning and the ‘new
skills for new jobs’ initiative).
Table 11. Bundles of skills used in ESJ survey
Skills Definitions
Fundamental
Basic literacy Reading manuals, procedures, letters or memos.
Advanced literacy Writing long documents such as long reports, handbooks, articles or books.
Basic numeracy Calculations using decimals, percentages or fractions, understanding tables and graphs.
Advanced numeracy Calculations using advanced mathematical or statistical procedures.
Basic ICT Using a PC, tablet or mobile device for email, internet browsing.
Moderate ICT Word-processing, using or creating documents and/or spreadsheets.
Advanced ICT Developing software, applications or programming;
use computer syntax or statistical analysis packages.
Foreign language skills Using a language other than your mother tongue to perform job duties.
Transversal
Communication skills Sharing information with co-workers/clients.
Teaching and instructing people.
Making speeches or presentations.
Teamwork skills Cooperating and interacting with co-workers.
Dealing and negotiating with people.
Customer handling skills Selling a product/service.
Dealing with people.
Counselling, advising or caring for customers or clients.
Problem solving skills Thinking of solutions to problems.
Spotting and working out the cause of problems.
Learning skills Learning and applying new methods and techniques in your job.
Adapting to new technology, equipment or materials.
Engaging in own learning.
Planning and Setting up plans and managing duties according to plans.
organisation skills Planning the activities of others.
Delegating tasks.
Organising own or other’s work time.
Technical Specialist knowledge needed to perform job duties.
Knowledge of particular products or services.
Ability of operating specialised technical equipment.
graphs as part of their daily job tasks. Slightly numeracy skills if they work in particular
fewer than one in three adult employees (29%), industries, such as construction or building (69%
mainly individuals working in managerial managers and 62% professionals) or financial,
positions (50%) and as professionals (43%), but insurance or real estate services (59% managers
also one in four workers in clerical jobs (26%), and 61% professionals). Advanced numeracy
also need to be competent enough to carry skills are also required in the supply of gas or
advanced mathematical calculations or statistical electricity, mining or quarrying sector (as stated
procedures in their job (Figure 40). Managers and by 63% of managers and 49% of professionals,
professionals are more likely to need advanced respectively).
CHAPTER 4
Figure 38. Level of fundamental skills required for the job, adult employees, 2014, EU-28
Skill not required at all Advanced Basic Moderate
E 70
60 58
52
49 Skill not required at all Advanced Basic Moderate
50
E 70 ISCO 3 Technician
40
40
60 58 ISC
30 52 Skill not required at all Advanced 29 Basic Moderate
E 49
70
50 ISCO
20 19 ISCO 3 Technician
14 14 40 58
60
40 12
10 52 9 IS
49 29
50
30 IS
0 40 ISC
ISCO 3 Technician
40
20 19 ISCO 8
14 14 ICT Literacy Numeracy
IS
12 29
80% NB:
10 See Table 11 for definitions of the levels of skills.
30 9 IS
Source: Cedefop ESJ survey. % Basic literacy skills % Advanced literacy skills % Literacy skills not required ISCIS
200 19
Figure 39. Highest level of literacy skills required by occupation, adult employees, 2014, EU-28
14 14 ISCO 8
ICTISCO 1 Managers 23 Literacy 73 12 Numeracy 3 80%
VET 10 9
80% ISCO 2 Professionals 20 75 3 70% IS
IS
0 % Basic literacy skills % Advanced literacy skills % Literacy skills not required
ICT professionals
ISCO 3 Technicians and associate 35 Literacy 59 Numeracy 5 ISCO 8
60%
ISCO 1 Managers 23 73 3 80%
80%
VET ISCO 4 Clerical support workers 41 52 6 IS
% Basic 50%
ISCO 2 Professionals 20literacy skills % Advanced literacy skills
75 % Literacy skills not required
3 70%
% ISCO 5 Service and sales workers 55 29 14
ISCO 3 Technicians and associateISCOprofessionals
1 Managers 23 35 73 59 53 40%
80%
46% VET 60%
ISCO 6 Skilled agricultural 57 17 25
ISCOsupport
ISCO 4 Clerical 2 Professionals
workers 20 41 75 52 63 30%
70%
50%
ISCO 7 Craft and trade workers 57 23 17
% ISCO 3 Technicians
ISCO 5and associate
Service professionals
and sales workers 35 55 59 29 14 5 20%
60%
40%
46% ISCO 8 Plant and machine operators 65 19 14
ISCO 4ISCO
Clerical support
6 Skilled workers
agricultural 41 57 17 52 25 6 10%
50%
ISCO 9 Elementary occupations 57 13 27 30%
% 50% ISCO 5 Service
ISCO sales workers
7 Craft and trade 5557 2329 1714 0%
40%
20%
46% EE CZ
ISCO
ISCO 8 Plant and6 machine
Skilled agricultural
operators % Basic numeracy skills 57 %
65 Advanced numeracy skills 17% Numeracy 25 required
19 skills not 14
30%
10%
ISCO 97 Elementary
ISCO Craft and
ISCOtrade workers
1occupations
Managers 46
57
57 13 23
50 27 17
3
50% 20%
0%
-40% ISCO ESJ
Source: Cedefop 8 Plant and
survey. machine
ISCO operators
2 Professionals 50
65
43
19 14
7 EE CZ
Information tech
Figure 40. Highest level of numeracy skills required by occupation, adult employees, 2014, EU-28
% Basic numeracy skills % Advanced numeracy skills % Numeracy skills not required 10%
ISCO 3 Technicians ISCOand9 Elementary occupations
associate professionals 5657 13 35 27 8
50% 0%
Financia
ISCO 1 Managers 46 50 3
ISCO 4 Clerical support workers 62 26 10 Supply
EE CZ of
0-40% ISCO 2 Professionals % Basic numeracy skills
50 % Advanced numeracy skills % Numeracy
43 skills not required
7
ISCO 5 Service and sales workers 65 14 19
Information tec
Professio
ISCO 3 Technicians and associate ISCOprofessionals
1 Managers 46 56 5035 83 Financi
Acco
0-40% ISCOsupport
ISCO 4 Clerical 2 Professionals
workers 50 62 43 26 107 Supplytec
oA
Information
ISCO 3 Technicians
ISCO 5and associate
and sales workers 56 65 1435 19 8 Professi
Financi
ISCO 8 Plant and machine operators 67 13 19
ISCO 4ISCOClerical support
6 Skilled workers
agricultural 5762 12 26 29 10
Acco
Supply Re
o
ISCO 9 Elementary occupations 53 7 37
ISCO 5 Service
ISCO sales workers
7 Craft and trade 65 1421 1915
Cultural industries
Professi
64
ISCO Acco
ISCO 8 Plant and6 machine
operators % Basic ICT 57 67
% Moderate ICT 12
% Advanced ICT % ICT skills29not19
13 required
R
Serv
ISCO 97 Elementary
ISCO Craft and trade workers
occupations 53 64 7 21 37 15
ISCO 1 Managers 10 67 21 2 Cultural industrie
ISCO 2 Professionals 11 61 25 2 WaR
% Basic ICT % Moderate ICT % Advanced ICT % ICT skills not required
ISCO 9 Elementary occupations 53 7 37 Ser
Cultural industrie
ISCO 3 Technicians and associate professionals 14 57 25 5
Source: Cedefop ESJ survey. ISCO 1 Managers
10 67 21 2
ISCO 4 Clerical support workers %15Basic ICT % Moderate ICT 71 % Advanced ICT % ICT skills not10required
4
W
ISCO 2 Professionals 11 61 25 2 Ser
ISCO 5 Service and
ISCO sales workers
1 Managers 10 32 67 37 4 2421 2 Start
with curren
ISCO 6ISCO
2 Professionals 11
32 23
61
2 42
25 2 W
elow 30%
ISCO 3 Technicians and associate professionals 56 35 8
tween 30-40% ISCO 2 Professionals 50 43 7
ISCO 4 Clerical support workers 62 26 10
Inform
low 30%
ISCO 3 Technicians and associate professionalsPREVIOUS CONTENTS 56
NEXT
35 8
Skills, qualifications and jobs 6in Skilled

the EU: agricultural
the making of a perfect match?
ISCO 4 Clerical support workers 62 26 10 S
ISCO 57 12 29
ISCO 7 Craft and trade workers
Figure 41. ISCOLevel of machine

ICT skills required by occupation, adult 67 employees, 2014, EU-28
64 21 15
Cultural
8 Plant and operators 13 19
ISCO 9 Elementary occupations % Basic ICT % 53

Moderate ICT 7
% Advanced ICT 37 skills not required
% ICT Cultural
ISCO 1 Managers 10 67 21 2
% Basic ICT % Moderate ICT % Advanced ICT % ICT skills not required
ISCO 2 Professionals 11 61 25 2
ISCO 1 Managers 10 67 21 2
ISCO 2 Professionals 11 61 25 2
ISCO 4 Clerical support workers 15 71 10 4
ISCO 5 Service and sales workers 32 37 4 24
ISCO 4 Clerical support workers 15 71 10 4
32 23 2 42
wit
ISCO 6 Skilled agricultural
ISCO 5 Service and sales workers 32 37 4 24
ISCO 7 Craft and trade workers 26 28 7 37
with
ISCO 6 Skilled agricultural 32 23 2 42
ISCO 8 Plant and machine operators 35 22 5 36
ISCO 7 Craft and trade workers 26 28 7 37
ISCO 9 Elementary occupations 25 15 2 55
ISCO 8 Plant and machine operators 35 22 5 36 U
Figure 42. Importance of transversal skills for job, adult employees, 2014, EU-28
% Very important % Moderately important % Not important % Skills not required U
Problem-solving skills 79 14 3 2
% Very important % Moderately important % Not important % Skills not required
Teamwork skills 78 15 4 1
Problem-solving skills 79 14 3 2
Communication skills 77 16 4 2
Teamwork skills 78 15 4 1
Learning skills 74 18 4 2
60% 70% Planning and organisation skills 71 18 6 3

Learning skills 74 18 4 2
Technical skills 67 21 8 3
earning
60%in a workplace
70% Planning and organisation skills 71 18 6 3
Customer handling skills 61 17 11 9

6% 67 21 8 3
Technical skills
arning in a workplace
Foreign language skills 30
61
25 26
17 11
18
9
6% Customer handling skills
39%
8% Foreign language skills 30 25 26 18

39%
Social and personal services O
%
Cultural
NB: % of industries (arts,(48entertainment
all respondents 676). Responsesor recreation)
to the question: ‘On a scale from 0 to 10, where zero means not at all important, five means moderately
important and 10 means essential,
Social how important
and personal are the following skills for doing your job?’. Responses in the interval 7 to 10 of the importance
services
scale haveServices relating
been classified to education
as ‘very or four
important’, health
to six ‘moderately important’ and zero to three ‘not important’; ‘skill not required’ is a separate
39% Cultural industries (arts, entertainment or recreation) Talent
category. ‘Don’t know/no
Accommodation, answer’ notorshown.
catering food services
4.4.2. Digital skills tasks. Even 42% of adults employed in

39% Financial, insurance or real estate services
Talent
About seven
Financial,in 10 EUor realadult workers
services need to elementary jobs need some level of ICT skills to
Accommodation, catering or food services 90%
Retail, sales, shop work or wholesale 100%
possess at least a moderate level of ICT skills to be able to do their jobs in contemporary labour
38%
insurance estate
Administration and support services 80%
be able to do their jobs (52% moderate fishinglevel and markets (Figure 41).
31% 90%
14% advanced). More than a fifthOther of those in

38% Agriculture, horticulture, forestry or
36% 70%
4.4.3. Other competences

managerial andhorticulture,
professional
forestry or roles
fishing or as
80%
31%
technicians and associate professionals need Adult employees were asked to rank the
Agriculture, 60%
36% Transportation or storage 70%
advanced ICT skills for their jobs (21%, 25% and importance of a further list of transversal or
31%
Other
technical skills required for doing their jobs. For

Professional, scientific or technical services 50%
25%, respectively). Around seven in 10 (71%)

60%
clerical support workers require at least a more than three in four EU employees, such skills
31% Construction or building 40%
are deemed to be very or moderately important

moderate level of ICTConstruction skills to perform

or building their job
Supply of gas or electricity, mining or quarrying
30%
Manufacturing or engineering 40%
52% Supply of gas or electricity, mining or quarrying 20%
49% Supply, management or treatment of water or steam 30%
Information technology or communication services 10%
52% 38% 20%
49% Supply, management or treatment of water or steam
CHAPTER 4
for doing their job (23). Figure 42 shows that Following this distinction, STEM graduates
problem solving and teamwork skills, closely have been identified within the ESJ survey by
followed by communication skills, are considered retaining tertiary education graduates whose field
to be very important by adult employees for doing of study is either engineering, maths and stats,
their jobs (79%, 78% and 77%, respectively). natural sciences and computing sciences. Around
Foreign language skills are considered by 31% of adult workers are found in the ESJ
adult EU workers as being of least importance for sample to have graduated from a STEM degree.
their jobs among the set of key competences About half (52%) of adult workers employed in the
quoted in the questionnaire: 26% stated that such broad industrial sector are STEM graduates,
skills are not important and 18% not required. while about a quarter (26%) are employed in
Nevertheless, more than half of EU employees services and a fifth (21%) in public administration.
(55%) require foreign language skills for their Most (59%) adult employees in the ICT sector are
jobs, and the need for proficiency in foreign graduates from STEM subjects as are about half
language skills is found to be highest in particular of workers in manufacturing (52%) and
occupations, such as managers, science and construction (48%) sectors. About 48% of adult
engineering professionals, and ICT technicians or employees in the professional, scientific and
ICT professionals. technical services have completed a field related
to STEM (Figure 43).
4.4.4. STEM skills As a further indication of the demand for
STEM skills are defined as those skills ‘expected STEM graduates in EU labour markets, STEM
to be held by people with a tertiary-education graduates are more likely to be employed in high-
level degree in the subjects of science, skilled occupations, in particular as professionals
technology, engineering and maths’ (EU skills and technicians and associate professionals
panorama, 2014). These skills include ‘numeracy (such as ICT technician, or business and
and the ability to generate, understand and administration technician). About 22% of tertiary
analyse empirical data including critical analysis; education graduates are employed as science
an understanding of scientific and mathematical and engineering professionals, ICT professionals
principles; the ability to apply a systematic and or as science and engineering or ICT technicians.
critical assessment of complex problems with an Shortages of STEM skills in advanced
emphasis on solving them and applying the economies are frequently reported. For example,
theoretical knowledge of the subject to practical the Manpower talent shortage surveys
problems; the ability to communicate scientific (Manpower, 2014) always place skilled trades
issues to stakeholders and others; ingenuity, workers, engineers and technicians at the top
logical reasoning and practical intelligence’ (ibid, with regards to the professions in which
2014). employers have most difficulties filling their jobs.
The understanding and scope of STEM skills The extent to which such recruitment bottlenecks
varies widely from country to country. Supply is reflect the inability of employers to offer an
relatively clearly identified in terms of qualifica- attractive job package (in terms of salary and
tions achieved in STEM subjects, although working conditions) rather than a shortage of
definitions of STEM subjects can vary. For skills per se is an issue of contention (Cedefop,
example, medicine, structural engineering and 2015a).
sports science are not included in some Although the ESJ survey suggests that, on
definitions. ‘Core’ STEM subjects typically average, a greater share of STEM graduates had
include: mathematics; chemistry; computer lower skills than needed when they started their
science; biology; physics; architecture; and, first jobs (34% of first job entrants were
general, civil, electrical, electronics, underskilled, compared to 30% from other fields),
communications, mechanical, and chemical this difference is not statistically significant when
engineering. other important correlated factors are taken into
(23) It is likely that customer handling, foreign language skills and technical skills are not rated as important as the remaining set of skills
because they tend to be more specific to particular occupations. Service and sales workers and managers are more likely to say that
customer handling skills are important to their current job (76% and 75%, respectively).
Learning skills PREVIOUS CONTENTS 74NEXT 18 4 2
60% 70% Planning and organisation skills 71 18 6 3
Technical skills 67 21 8 3
arning in a workplace
Customer handling skills 61 17 11 9
43. Share
FigureForeign languageof STEM graduates
30 by industry, adult
25 employees with tertiary level education,
6%
2014, EU-28
skills 26 18
39% O
%
Cultural industries (arts, entertainment or recreation)
39% Talent
100%
Financial, insurance or real estate services
Retail, sales, shop work or wholesale 90%
38%
Administration and support services 80%
31%
36% 70%
Other
60%
31%
Construction or building 40%

30%
52% 20%
38%
0% 10% 20% 30% 40% 50% 60% 70%
0%
NB: STEM graduates are graduates with a tertiary education degree in the fields of engineering, maths and statistics, natural sciences and computing
science. Other fields STEM
Source:
account (gender, age, 33% occupation, industry). that some of their skills will become outdated in
However, the ESJ suggests that STEM graduates the next five years, as opposed to 15% of
30%
rkplace
with tertiary level qualifications are employed in26% graduates from humanities and education (Figure
30%
25% 28%
jobs
20% that are well-matched to their level of 45). The ESJ data indicate that concerted efforts
raft and trades 25%
qualifications. Specifically, there is a significantly are required by education and training providers
ant and assemblers
lower chance that they are overqualified relative so that STEM graduates possess the right set of
ementary 15% 18%
to
10%tertiary graduates from other fields (Figure key competences and skills to enable them
ISCO 3 Te
44) (24). continuously to adapt and improve their skills

No WBL WBL
Even though STEM graduates are more likely throughout their working lives. However, the
5%
to0%be employed in jobs that are matched to their greater skill deficits of STEM graduates also
qualifications, they are nonetheless
of first job found to have Overqualified
suggest that employers needSkill to deficit
invest in their
relatively higher skill deficits, implying that they continued skill development, should they wish to
Underskilled at start
have room to develop their skills further to reach exploit the full potential of their skills and ensure
full proficiency. This is likely to be a consequence that they are up to date with the rapidly changing
of the demand for high level skills in their jobs, nature of their job tasks.
Humanities and education
with 54% of STEM tertiary graduates stating that

5%
they experienced recent changes in the

Social sciences
technologies used at their work in comparison to

ive
45% of graduates from non-STEM fields. Around

Other
20% of STEM graduates think that it is very likely
No WBL WBL
80%
STEM
(24) Other fields include economics, business and law, agricultural studies, teacher and education sciences, humanities, arts, medicine
and health-related sciences and other social sciences.
60%
Health-related sciences
tage of responses
0% 5% 10% 15% 20% 25% 40%

Administration and supportorservices 60%
80%
Transportation storage
Professional, scientific or technical services 50% 5
70%
Other
Construction or building 40%
60%
CHAPTER 4
73
Transportation
mining oror storage
Tackling skill mismatches among the young with high quality VET
Supply of gas or electricity, quarrying
30%
50% 3
Professional, scientific or technical
Manufacturing services
or engineering
Construction
Supply, management or treatment of waterororbuilding
steam 20%
40%
9%
Figure 44. Incidence of qualification

or engineering and skill mismatch of adult employees with tertiary level
30%
38% 31
education by STEM status,

steam 2014, EU-28
Manufacturing
0% 10% 20% 30% 40% 50% 60% 70%
0%
20%
9% Supply, management or treatment of water or Unde
at sta
Information technology or communication services Other fields STEM 10%
38% 1
35%
0% 10% 20% 30% 40% 50% 60% 70%
0%
30% 33% Unde
30% at sta
Other fields STEM
25% 28%
35% 26%
trades 25%
20%
assemblers 30% 33%
y 15% 30% 18%
25% 28% ISCO 3 Technicia
26%
trades 10% 25%
WBL 20%
assemblers IS
5%
y 15% 18%
ISCO 3 Technicia
ISC
0%
10%
WBL Underskilled at start of first job Overqualified Skill deficit
IS
Source:
Figure 45.andShare
educationof adult employees with tertiary level education who think it very likely that
ISCI
0%
some of their skills will become outdated in the next five years, 2014, EU-28
Humanities
Underskilled at start of first job Overqualified Skill deficit
ISCO
Social sciences
I
IS
Other ISCO
WBL
Social sciences
80% IS
STEM
WBL Other
60%
Persantage of responses
80%
STEM
0% 5% 10% 15% 20% 25% 40%
60%
Increasing need for learning Persantage of responses 20%
0% 5% 10% 15% 20% 25% 40%
Other
bsolescence NB: Percentage of respondents with scores above eight in the 0 to 10 skill obsolescence scale.
Accommodation, 0%
Source: Cedefop ESJcatering
survey. or food services
Increasing need for learning 20% Lea
Cultural industries (arts, entertainment) with
Other
bsolescence
Accommodation, Transportation
catering or foodorservices
storage 0%
50% Lea
CulturalAgriculture,
industries (arts, entertainment)
horticulture, forestry with
Retail, sales,
Socialshop
andwork or wholesale
personal services
40%
Transportation or building
Construction or storage
50%
Agriculture,
Information technologyhorticulture, forestry
or communication 30%
Social and personal
Manufacturing services
or engineering 40%
Construction or services
building 20%
Administration and support
Information
Supply, technology
management or communication
or treatment of water 30%
10%
Manufacturing
Services relating or engineering
to education or health
20%
Administration
Professional, scientific and support services
or technical 0%
Supply, management
Financial, insurance ororreal
treatment of water
estate services T
10%
most
Services relating
Supply of gastooreducation ormining
electricity, health
Professional, scientific or technical services 0% 10% 20% 30% 40% 50% 60% 0%
Financial, insurance or real estate services T
most
21%
One in five EU workers
think it very likely that some of their skills
will become
outdated in the next five years
20%
Information technology or communication servicesPREVIOUS CONTENTS NEXT 10%
38% 1
0% 10% 20% 30% 40% 50% 60% 70%
0%
Unde
at sta
Other fields STEM
35%
Matching better skills to better jobs

CHAPTER 5
30% 33%
30%
25% 28%
26%
rades 25%
20%
assemblers
y 15% 18%
Shifting skill needs in

ISCO 3 Technician
5.1. competence-development throughout life.

European job markets
10%
Lifelong learning, CVET and validation of non-

WBL
IS
formal and informal learning are necessary

5%
The findingsUnderskilled
of the at
ESJ survey emphasise the ingredients, as in the European Commission’s
ISC
0%
high level of skill demand that exists in EU job strategy for rethinking education and training
start of first job Overqualified Skill deficit
markets. It also confirms the need for EU policy- (European Commission, 2012a).
makers to intensify their efforts to strengthen the European workplaces are in a constant state
of flux and employees in today’s labour markets
IS
integration of key competences in VET curricula

and provide more effective opportunities to are often requested to adjust their skills in
ISCO 8
acquire Social
or develop
sciences those skills through IVET. response to new technologies and working
However, coping with the dynamic challenges of methods, as well as developing new skills that will
IS
competitive job Other markets, such as rapid allow them to cope with the challenges of
technological progress, changing professional increasingly customer-oriented, global value
WBL
requirements, globalised markets and a shrinking production chains reliant on high skill demand
workforce, requires extensive participation by (Timmer et al., 2014). Across Europe, 47% of EU
80%
STEM
adults in continuous learning and a high workers have seen the technologies they use
commitment
Health-related to investment in skills upgrading and change since the time they started their current
60%
sciences
Figure 46. Share of 0%
jobs with significant rise in the
10% need to learn
15%new things by
20%industry, adult
employees (aged 24-65), 2014, EU-28
5% 25% 40%
Increasing need for learning 20%
Other
bsolescence
Accommodation, catering or food services 0%
Lea
Cultural industries (arts, entertainment) with
50%
Agriculture, horticulture, forestry
40%
Construction or building
Information technology or communication 30%
20%
Supply, management or treatment of water
10%
Financial, insurance or real estate services Tr
most
0% 10% 20% 30% 40% 50% 60%
NB: Responses to the question: ‘On a scale from 0 to 10, where 0 means it has decreased a lot, five has stayed the same and 10 has increased a lot, has
the need to learn new things increased, decreased or stayed the same since you started your job with your current employer?’ The figure depicts
share of respondents with a score above the median of the distribution (equal to a score of eight).
Figure 47. Share of jobs with rising variety of tasks by occupation, adult employees,
2014, EU-28
c Moderate % Increased % Remained the same % Decreased
80%
ISCO 2 Professionals 73 24 3 70%
ISCO 3 Technicians and associate professionals 73 24 3 60%
ISCO 4 Clerical support workers 71 25 4 50%

40%
30%
20%
10%
y skills not required
0%
NB: Responses to the question: ‘On a scale from 0 to 10, where 0 means it has decreased a lot, five has stayed the same and 10 has increased a lot, has
3 80%the variety of tasks increased, decreased or stayed the same since you started your job with your current employer?’ Increased depicts share of
respondents with a score above five.
3 70%
60%
job, while 43% experienced changes in their high demand for learning is also evident in certain
5 60%
working methods and practices, according to the advanced manufacturing sectors (such as supply
6
50%
ESJ survey (25). An additional 29% saw changes of natural resources). Apart from individuals in
50%
in the products or services produced within their elementary jobs, most adult workers have seen
14 40%
40%
workplace, while 26% have recently experienced the need to learn new things and the variety of
25 30%
changes in the intensity of contact they have with tasks in their jobs increase since their current
30%
17
clients or customers. employment started (Figure 47).

20% 20%
10% Some 51% of adult employees are in jobs

14
10%
involving the need always or usually to learn new

5.2. Evolution
IT LV EU-28 HU PLof
FI adult worker
27
things EE(19% always;

LT RO 32% SK HRusually),
EL PT ESwhileCY IE44%
skill mismatch
0% 0%
sometimes have to learn new things. A total of

CZ SI NL FR UK BE DK DE SE AT LU BG MT
y skills not required
seven in 10 EU adult workers admit to having

considerable flexibility in how to perform their job As a consequence of the dynamically evolving
3
% Unlikely % Moderately likely % Likely % DK
tasks, so their skills need to be resilient and 39skill needs of jobs 29 described above, 29 about a
7 60%
adaptable to beinsurance
able to orcope withservices
such autonomy. quarter (26%) of adult 26 employees in 24 the EU
Information technology or communication services 4
Most employees (72%) always or usually have to labour market think it is moderately likely, and one
8 Financial, real estate 46 4 50%
work as part of a team. For more than half EU in five (21%) think it is very likely, that several of
10 Supply of gas/electricity or mining/quarrying 46 27 23 4
employees (53%) the need to learn new things their skills will become outdated in the next five
40%
Professional, scientific or technical services 48 26 23 4
and the variety of thecatering

job tasks
or food has significantly years. Employees in Estonia and Romania (42%
19
30%
increased since Administration and support services their job,

the time they started and
47 39%, respectively) 26 are the most likely to say
29
Accomodation, services 52 22 21 6
illustrating the dynamic evolution of skill demand they will suffer from very25high levels of 20perceived
21 6 20%
15
in modern economies and the extent to which EU skill obsolescence, in contrast to those from Malta
Construction or building 49 6
employees continuously have to cope with a

10%
and Bulgaria (Figure 48).

19
Retail, sales, shop work or wholesale 47 27 20 6
changing working(arts, pace. More than three in 10 23 respondents working in

7 0%
Cultural industries entertainment or recreation)
Figure 46 illustrates that the need for information technology or communication
52 21 5
continuousServices learning is most prominent in the services think it is very likely 24 to see some 19 of their
Manufacturing or engineering 47 27 20 6
skills not required
financial and professional services industries but skills become outdated in the foreseeable future
relating to education or health 52 4
21 2 Transportation or storage 49 25 20 7
ived training paid by employer
90
Water supply/management/treatment
(25) For those with long job tenure, the question was restricted to whether the changes in their workplace have taken place within the past
25 2 47 28 20 6
five years that they have been in their job.

Agriculture, forestry or fishing 50 26 18 6 80
25 5
Social and personal services 52 24 16 7
70
10 4
24 60
Start of job
Present job
CHAPTER 5
Moderate % Increased % Remained the same % Decreased
80%
(Figure 49). This is reasonable given that the term are highest among respondents 24employed
Moderate ISCO 2 Professionals % Increased 73 % Remained the same 24% Decreased
3 70%
capacity to store, communicate and compute as professionals as well as technicians and

80%
ISCO 3 Technicians and associate
ISCOprofessionals
1 Managers 7380 18 32 60%
informationISCO has4 Clerical

advanced rapidly
workers in the past 25 associated professionals, with 23% stating that43it
70%
years withISCO
the5andintroduction of the internet, digital is very
62 73likely; in contrast, those employed as
ISCOsupport
2 Professionals 71 73 25
24 50%
technology and open source computing. Skill skilled agricultural workers and plant and
41
ISCO 3 Technicians Service and sales
associate workers
professionals 35 24 3 60% 38
40%
obsolescence is also high in financial, insurance machine operators and assemblers are the least
ISCO 4ISCO 6 Skilled
Clerical agricultural
support workers 62 71 34 25 42
or real estate services salesand in professional, likely

6562 to worry about such a contingency.
50%
30%
scientificISCO
and technical services. Concerns about Even though individuals anticipate their skills
ISCO 7 Craft and trade workers 4
ISCO 5 Service 35
32 33 38
40%
their skills becoming outdated over the medium- becoming obsolete due to evolving skill demands
20%
8 Plant and6 machine
ISCO operators
Skilled agricultural 58 62 3834 32
30%
Figure 48. Likelihood of skills becoming outdated in next five years, EU-28, 2014
10%
ISCO
ISCO 97 Elementary occupations
Craft and trade workers 48 65 47 32 53
required 20%
ISCO 8 Plant and machine operators 58 38 3 0%
3 80% BG RO
10%
t required
3 70% 0%
60% BG RO
53 80%
60%
63
50%
70%
50%
60%
14 5 60% 40%
40%
6
50%
50% 30%
30%
1714 40%
40%
20% 20% 17
14 13
30%
30%
10% 10%
17
20% 20% 17
0% 0%
14 13
EE CZ SI LT RO SK HR EL PT ES CY IE NL FR UK BE IT LV EU-28 HU PL FI DK DE SE AT LU BG MT FR SE
required 10% 10%
NB: Percentage of adult employees who think it is moderately or very likely that several of their skills will become outdated in the next five years.
3 0% Cedefop ESJ survey.
Source: 0%
Figure Likelihood
49.technology of skillsservices
becoming outdated
39 by industry, EU-28,
29 2014
EE CZ SI LT RO SK HR EL PT ES CY IE NL FR UK BE IT LV EU-28 HU PL FI DK DE SE AT LU BG MT FR SE
t required
7 60%
Information or communication 29 4
8 3 Financial, insurance or real estate services 46% Unlikely 26 likely 24 4
% Moderately % Likely % DK 50%
10 7 Supply of gas/electricity or mining/quarrying 46 27 23 4 60%
Information technology or communication services 39 29 29 4 40%
9 8 Professional, scientific or technical services 48 26 23 4
Financial, insurance or real estate services 46 26 24 4 50%
30%
Accomodation, catering or food services 52 22 21 6
40%
Administration and support services 47 26 21 6 20%
1915 Professional, scientific or technical services 48 26 23 4
Construction or building 49 25 20 6 30%
Accomodation, catering or food services 52 22 21 6 10%
19
Administration and support services 47 26 21 6 20%
15 0%
Cultural industries (arts, entertainment or recreation) 52 23 21 5
Construction or building 49 25 20 6
19
10%
Manufacturing or engineering 47 27 20 6 S
required
Services relating to education or health 52 24 19 4 0%
Cultural industries (arts, entertainment or recreation) 52 23 21 5
2 Transportation or storage 49
Manufacturing or engineering 47 2725 20
20 76
S
paid by employer
90%
t required Water supply/management/treatment 47 28 20 6
2 Services relating to education or health 52 24 19 4
Agriculture, forestry or fishing 50 26 18 6 80%
52 Transportation or storage 49 25 20 7
by employer
Social and personal services 90%

52 24 16 7
Water supply/management/treatment 47 28 20 6 70%
42
paid training
10
Agriculture, forestry or fishing 50 26 18 6 80%
5 60%
Start of jobSocial and personal services
received
52 Present job 24 16 7
with current employer 70%
training
10 4
50%
receivedwho
NB: Percentage of adult employees who think that several of their skills will become outdated in the next five years.
Source: Cedefop 60%
StartESJ
ofsurvey.
job
employees
Present job 40%

with current employer Underskilled (14.2%) 96% improvement 50%
in skills since start job
loyees%who
30%
0%
Underskilled Matched (64.7%)
Underskilled (14.2%)
40%
96% improvement
0% 0%
EE CZ SI LT RO SK HR EL PT ES CY IE NL FR UK BE IT LV EU-28 HU PL FI DK DE SE AT LU BG MT
skills not required
3
7 78 Evidence from Cedefop’s European skills and jobs survey 60%
Information technology or communication services 39 29 29 4
8 Financial, insurance or real estate services 46 26 24 4 50%
and/or lack of continuous learning, the ESJ data job have remained in the same state throughout
40%
Professional, scientific or technical services 48 26 23 4
show thatAccomodation,
the skill mismatch status of individuals their
52 tenure (Figure 50) (27 22). A total of seven in 10
19
30%
is characterised by persistence
and support (services
). On average, workers with matched 26skills at the start 21 of their
29
catering or food services 21 6
26
six in 10 adult workers Construction or building entered a

in the EU who current employment have matched skills even
Administration 47 6 20%
15
mismatched job remained in the same state of after several years of working in the same job.
49 25 20 6
10%
mismatch after several years of being in it. Those There is less persistence for employees who
19
who entered jobs that do not fully exploit their enter into their jobs with lower skills than needed,
7 0%
Cultural industries (arts, entertainment or recreation)
skills, most of which are ordead-end
engineering jobs (as given that such jobs tend27 to be the most 20 dynamic
52 23 21 5
defined inServicesSection relating2.3.3), areor more

health likely to and,52 as a consequence, individuals develop their
Manufacturing 47 6
skills not required
remain stuck in the same situation of skill skills at a fast(er) pace relative to other states of
to education 24 19 4
mismatch. A total of eight in 10 EU employees skill mismatch.

21 2 Transportation or storage 49 25 20 7
who were overskilled at the start of their current The entry point to the labour market or to a
% employees who received training paid by employer

90
25 2 Water supply/management/treatment 47 28 20 6
Figure 50. Dynamic evolution of skill mismatch between start of current job and present time,
Agriculture, forestry or fishing 50 26 18 6 80
adult employees, EU-28

25 5
Social and personal services 52 24 16 7
70
10 4
24 60
Start of job
Present job
with current employer
50
40
Underskilled (14.2%) 96% improvement
6
in skills since start job 30
Underskilled Matched (64.7%)

85% increase
Skills not required in variety of job tasks
Overskilled (21.1%)
14 3 2
15 4 1
Underskilled (3.4%) 87% improvement
16 4 2
ISCO 3
18 4 2
Matched Matched (70.3%)
78% increase
6 3 Overskilled (26.3%) in variety of job tasks
8 3
11 9 Underskilled (1.75%) 73% improvement

18
Overskilled Matched (17.6%)
67% increase
Overskilled (80.1%) in variety of job tasks
Financ
NB: Share of EU employees with positive answers to the questions: ‘Overall, how would you describe your skills in relation to what is required to do your Profess
job?’ And
Talent ‘When
in dead endyou started
Talentyour job with your current
in standstill employer, overall, how would you best describe your level of skills in relation to what was
Talent unnourished
required to do your job at that time?’ (a) My skills are higher than required by my job; (b) my skills are matched to what is needed by my job; (c) some
100%of my skills are lower than required by my job and need to be further developed. Supply
90% Ser
(26) It is not possible with the cross-sectional data at hand to decouple accurately the extent to which this persistence reflects state
34%
dependence or simply unobserved 44%heterogeneity among workers who are in mismatched jobs. Estimates based on panel data available
Inform
80%
in Australia show that the probability of overqualification or overskilling is strongly related to the past incidence of skill mismatch, after
controlling for
56%unobserved individual effects (Mavromaras et al., 2012).
(27) A slightly smaller yet significant degree of persistence in overskilling (74%) is found for first job entrants.
70%
9%
60%
Agricu
50%
71 18 6 3 Overskilled (26.3%)
67 21 PREVIOUS
8 3 CONTENTS NEXT
CHAPTER 5
79
61 17 11 9 Underskilled (1.75%)
Matching better skills to better jobs
25 26 18
Overskilled Matched (17.6%)
job matters. Entry into skill-intensive jobs is Figure 51. Skill mismatch transitions within
Overskilled (80.1%)
crucial to avoiding getting trapped into a jobs, adult workers, 2014, EU-28
mismatched state and learning conducive forms
of work, that can ensure the continuous provision Talent in dead end Talent in standstill Talent unnourished
of non-formal or informal learning for adult

workers, are critical for avoiding the stagnation of
100%
their skills within dead-end jobs.

90%
The persistence in the skill mismatch of

34%
individuals nevertheless hides a significant

80% 44%
degree of dynamics in the evolution of individuals’ 70% 56%
skills and in the complexity of tasks within their

jobs. One in four adult workers who entered a job
9%
60%
with matched skills became overskilled after

several years of job tenure. Similarly, even within
50%
apparently ‘stable’ job-worker matches, whereby

40%
employees remain consistently matched over

38%
time, there are dynamic changes taking place in

30% 57%
31%
the skill needs of their jobs and in their skill

development. For instance, 87% of individuals
20%
who remained matched over time improved their

10% 18%
skills
20% since starting40%
their job50%
and for 60% one in70%
three
13%
in this group (33%) the improvement was very

% 10% 30%
0%
significant. At the same time,Other 78% experienced

Underskilled Matched Overskilled
at start of job at start of job at start of job
an increasing variety in their job tasks over time.

fields STEM
Individuals who remained consistently

overskilled since the time they entered their job Evidence shows that more than half the
% Skills improved % Skills remained same
experienced a significantly smaller 28% degree of workers (57%) who enter their jobs with higher
skills improvement (73%; for 23% the skills than needed are likely to get trapped in a
ISCO 1 Managers 83
26%
improvement was very significant) relative to dead-end situation, ISCOwith limited potential for
25%
further growth in their skills. However, a third of

2 Professionals 82
those with matched skills. They are also in

comparatively less challenging jobs, with 67% these (34%) still have some space to develop
18%
ISCO 3 Technicians and associate professionals 81
experiencing an increasing variety in the tasks their skills to reach maximum productivity in their
they perform. By contrast, workers who made the jobs. By contrast, a small proportion of workers
ISCO 4 Clerical support workers 76
transition from an initial state of overskilling to who enterISCO their jobs andwithsales
matched
workers skills or lower
matched skills in their current job saw their job skills than needed end up in a dead-end setting.
5 Service 71
tasks become markedly more varied (with 81% Instead, most such ISCO 6employees are characterised
Overqualified Skill deficit
experiencing a rise in task variety). by skill deficits, rendering the further provision of
Skilled agricultural 71
In contrast to employees who were initially continuous ISCO vocational

7 Craft and tradetraining critical for
enhancing their labour productivity (Figure 51).
workers 75
matched or overskilled, individuals who entered

their jobs with lower skills than needed These findings show that the evolution of the
ISCO 8 Plant and machine operators 69
(underskilled) are less likely to get stuck in the skill mismatch status of individuals depends
same state and more likely to experience a crucially on the interplay between opportunities to
ISCO 9 Elementary occupations 55
continuous (co-)evolution in their skills and in the develop their skills in their jobs and the skill
skill needs of their jobs. Some 96% experienced intensity of their job tasks (Table 12). In some
skill development since they started their job, with cases
80% individuals need to ‘keep moving forward
50% experiencing very significant improvement: in order just to stand still’, as with those whose
this is aided by the fact that underskilled workers skills remain consistently matched to their job
needs. For these workers their 63% level of skill
69%
are also more likely to be in dynamic jobs that

60%
development is merely sufficient to compensate

entail increasing complexity of job tasks over time

59%
(as in the 85% that saw an increase in the variety for40%

the constant increase in the complexity of their
job tasks.
10% 15% 20% 25%
of their tasks).
Increasing need for learning 20%
0%
Table 12. Evolution of skill mismatch within jobs in relation to skill development and job
complexity, adult employees, 2014, EU-28
Skill mismatch transitions within job
Changes matched matched overskilled underskilled

since start of job at start of job at start at start at start of job
   
overskilled underskilled overskilled matched at
at present time at present time at present time at present time
(ref: matched (ref: matched (ref: overskilled (ref: underskilled
 matched)  matched)  matched)  underskilled)
Improvement in skills    
Increase in task complexity

(variety, difficulty,    
need to learn new things)
Change in technologies,    
work methods, etc.
NB:  = positive effect;  = negative effect;  = no significant effect. Estimated probability patterns of skill mismatch transitions based on multinomial
logit regressions; to compare ‘like with like’ the reference category includes individuals with the same skill mismatch state at the start of their jobs;
the regressions control for the level of job complexity and skill intensity of a job, changes in job roles (e.g. promotion) since the start of a job and other
individual and job characteristics of the respondents. See Annex 2 for details.
Other individuals are more likely to move to a The above patterns make it clear that two
state of overskilling, from an original state of underlying forces determine whether individuals
matched skills, because their skill accumulation can tilt away from skill mismatch: the extent of
was not accompanied by a corresponding increase their continuous investment in skills and the
in job complexity. The overskilled are more likely to degree of job complexity. Depending on the
experience changes to the technologies and direction of these forces, adult workers can either
working methods they use, but this appears to be gravitate towards a ‘virtuous’ scenario, involving
associated with a process of deskilling or of skills positive feedback between job complexity and
being replaced by technology (Autor et al., 2003; skill development or a ‘vicious’ one, in which
Freeman, 2015), which is potentially linked to the workers’ jobs and skills remain stagnant over
routinisation of their jobs (Cedefop, 2010c). time. The remainder of the chapter focuses on the
Remaining in a persistent state of overskilling is extent to which adult workers in the EU develop
associated with jobs not characterised by their skills via their investment in continuous non-
increasing task complexity and with no further skill formal and informal learning and on factors that
development taking place over time. are correlated with their chances of being
When technological changes in jobs are employed in skill-intensive jobs.
associated with increasing task complexity, yet
individuals’ skills fail to keep up with the rapid
pace of change, it is more likely that employees 5.3. Skill development and adult
will become underskilled, even if their skills were learning in the EU
originally matched to their job requirements. By
contrast, an individual’s progression from an initial Cedefop’s ESJ survey highlights that
skill gap to having matched or surplus skills some considerable skills accumulation takes place in
years later on is associated with rapid skills European workplaces. About three quarters of EU
accumulation that is not accompanied by skill- employees (76%) have improved their skills since
augmenting jobs. starting their current jobs. Younger individuals, or
40%
38% S
30% 57%
31% PREVIOUS CONTENTS NEXT Retail, sal
Cultural ind
CHAPTER 5
20%
10% 18%
13% Accommodatio
70%
0%
Underskilled Matched Overskilled
Figure 52. Improvement in skills within jobs by occupation, adult workers, 2014, EU-28
at start of job at start of job at start of job
STEM
% Skills improved % Skills remained same % Skills worsened % Cannot compare
ISCO 2 Professionals 82 15 22
Technicians
ISCO 3 Technicians and associate professionals 81 15 22

Servic
ISCO 4 Clerical support workers 76 20 22 Buildin
ISCO 5 Service and sales workers 71 25 22 Plant and mach
Skilled agric
.4
NB: Share of EU employees with positive answers to question: ‘Compared to when you started your job with your current employer, would you say that
your skills have now improved, worsened or stayed the same?’
Source:
80%Cedefop ESJ survey.
those
60% entering their first jobs after63%
69%
completing becoming obsolete in future years, greater
education, are more likely to experience positive attention is required by policy-makers to designing
.2
growth in their skills as they accumulate age-friendly skills policies, which take into account
59%
additional years of work experience and job- the specific needs and circumstances of ‘silver
51%
25% 40%
specific know-how. workers’ (Cedefop, 2010b).

40%
Nevertheless, more than one in five (22%) EU Individuals employed in semi-skilled and low-
0
Density
employees fail to develop their skills in their skilled occupations are also found to be less likely
r learning 20%
current jobs. Skill development is less likely to to improve their skills in their jobs. A total of 43%
.4
occur among males, those who returned to the job of individuals working in elementary occupations
0%
market with after experiencing a spell of anderror

28% ofwithplant
aid of and machine operatives
on the job have
Learned by interacting Attended training Learned at work through Learned by yourself (e.g. Supervisor taught you
unemployment, individuals who entered their jobs not improved their skills since they started their
colleagues at work courses trial and manuals, books,
with higher education and skills than required by current job (Figure 52) (29). Adults who work in the
videos or online materials)
.2
job ( ) and older employees. Even after taking
the50% 28
accommodation, catering or food services and
into account the additional years of employer transportation and storage industries are most
tenure
40% and the original match of the (higher stock likely to say that their skills have remained the
of) skills of older workers to their jobs, those aged same (28% and 26% respectively). Employees in
0
42%
over 40 years old are less likely to improve their the supply, management, treatment of
skills in their jobs relative to employees aged 24 water/steam are more likely to say that their skills
30% 34%
32% G
to 39.
20% Considering that older-aged workers are have worsened (4% compared to an overall
equally or more concerned about their skills average of 2%).19%
35%
10%
(28) However, the slower growth in the skills of individuals who were overskilled at the start of their jobs is subdued if they are employed
in0%
jobs that involve a high level of task complexity.
(29) Elementary workers (e.g. labourers, packers, goods handling and storage staff) tend to perform routine and repetitive tasks as part of
30%
a production chaincourses
Training or process, which mayTraining
attended leave less
whileopportunity
performing to learn or courses
Training developattended
new techniques.
mostly Prior research
I have by Cedefop
not undergone
(2010c) also suggests elementary workers may also be constrained by structural and social inequality; that is, they may be lower
mostly or only during work hours your regular job or only outside of work hours any training
educated and have limited access to education/training. This may arise either due to lack of personal resources or because of biased
25%
stay overskilled within jobs
company training policies which fail to invest in their skills, due to the comparatively high level of investment required to improve their
skillsets and the low value added of their posts to the firm.
60%
20%
15%
ISCO 5and
ISCO 3 Technicians Service and sales
associate workers
professionals 71 81 25 15 22 22 Plant a
cit Ski
ISCO 4ISCO 6 Skilled
Clerical agricultural
support workers 7176 25 20 22 22
ISCO 7 Craft and
and trade
sales workers Plant a
ISCO 5 Service workers 75
71 2522 2112
cit 82 Evidence from Cedefop’s European skills and jobs survey

ISCO 8 Plant and6 machine operators 69 26 2222 Ski
ISCO Skilled agricultural 71 25
ISCO
Craft and trade workers 55 75 39 22 4 112
ISCOMethod
Figure 53. 8 Plant andof improving/acquiring
machine operators new skills in69job, adult workers, 2014, EU-28
26 22
80% ISCO 9 Elementary occupations 55 39 4 2
69%
60% 63%
59%
80%
51%
25% 40%
69% 40%
60%
Density
63%
59%
g need for learning 20%
51%
25% 40%
40%
0%
Density
Learned by interacting Attended training Learned at work through Learned by yourself (e.g. Supervisor taught you
g need for learning 20% with colleagues at work courses trial and error with aid of manuals, books, on the job
NB: Share
0% of EU employees with improved skills and positive answers to question: ‘Since you started your job with your current employer, have you done
any of the following to improve or acquire new skills?’
50% Learned by interacting Attended training Learned at work through Learned by yourself (e.g. Supervisor taught you
with colleagues at work courses trial and error with aid of manuals, books, on the job
40% 54. Types of
Figure training completed in last 12 months, adult workers, 2014, EU-28
42%
50% 34%
30%
32%
40% 42%
20%
19%
30% 34% 35
10% 32%
20% 30
0%
19%
Training courses attended Training while performing Training courses attended mostly I have not undergone 35
10% mostly or only during work hours your regular job or only outside of work hours any training
25
within jobs
50% 60% 0% 30
Training courses attended Training while performing Training courses attended mostly I have not undergone 20
stay overskilled
mostly or only during work hours your regular job or only outside of work hours any training
25
jobs
NB: Share of EU employees with positive answers to question: ‘In the last 12 months have you undergone any of the following types of training in your
Mean %within
50% 60% 15
current job?’
Source: Cedefop ESJ survey. Mean % stay overskilled 20
Provision of training is an important pillar of accommodation, catering or food services;

10
European lifelong learning policies aimed at construction; and agriculture) and 55% of those
enabling adults at risk of skill obsolescence to in elementary jobs, did not receive any training,
15
upgrade their skills or to reskill (Cedefop, 2013). either on or off the job, in the previous year. Older
5
The ESJ survey highlights that a significant workers, those employed in smaller-sized, private
10
majority (63%) of adult EU employees who have sector firms and individuals returning to the labour
0
improved their skills in their jobs (including those market after being unemployed are less likely to
5
in their first jobs, recent job finders, and those have received training. The same holds for lower-
more resilient to technological and workplace qualified individuals. This training bias holds
changes) have done so by attending training independently of the fact that the lower educated
0
courses (Figure 53). are more likely to be employed in lower-skilled

Nevertheless, about 32% of EU employees occupations and industries, or in jobs with lower
did not undergo any training between 2013 and skill content. The fact that training is biased
2014 (Figure 54). About 45% of adult employees against adults with lower education and with a
in specific economic sectors (including history of unemployment is a cause for concern,
CHAPTER 5
Figure 55. Incidence of training completed in last 12 months by type, adult workers, 2014, EU-28
creased Training during work hours (EU mean: 61%)
80% 74
2 72 73 100%
70 71
69 69
70% 65 67
3 65
creased Training during61work
62 hours
62 (EU 63 61%)
62 mean: 62
60 61 61
3 80%
60% 80%
56 74 100%
2 54 72 73
52 70 71
49 50 50 69 69
4 70% 49 67
3 50% 65 65
61 62 62 62 63 62
41 60 61 61 60%
3
3 60% 38 80%
40% 56
52 54
2 49 49 50 50
4 50%
30% 40%
3 41 60%
3
40%
20% 38
3
2
30% 20%
10% 40%
5
3
20%
0%
3 BG RO EL LV ES LU HU HR PT SI PL AT IT NL LT DE MT CY IE BE EE SE FR CZ SK UK DK FI 0%
20%
10%
5
Training outside work hours (EU mean: 31%)
0%
60%
BG RO EL LV ES LU HU HR PT SI PL AT IT NL LT DE MT CY IE BE EE SE FR CZ SK UK DK 53
FI 0%
49 49
50% 45 47
Training outside work hours (EU mean: 31%) 43
35
40
60%
40%
34 35 35 36 53
31 31 32 49 30
30 49
50%
30% 28 28 29 29 45 47
26 27 27
25 43
22 40 35
20 25
40%
20% 17 35 35 36
17 34
13 31 31 32 30
29 30 20
30% 27 28 28 29
10% 26 27
25
22 25
20 15
20%
0% 17 17 LU
G MT 13
FR SE UK BE FI DK HU CZ SK LV LT IE BG EE DE IT AT RO NL PL PT LU SI HR MT CY EL ES 20
10% 10
NB: Share of EU employees with positive answers to question: ‘In the last 12 months have you undergone any of the following types of training in your
15
% DK 0%current job?’ 5 LU
G MT Source: Cedefop
FR SE ESJ UKsurvey.
BE FI DK HU CZ Training SK LV motivation
LT IE BG in the
EEEU DE
(values:
IT weighted,
AT RO rounded)
NL PL PT LU SI HR MT CY EL ES
Figure 56. Reasons for participating in training, adult workers, 2014, EU-28
60% 10
4 0
4 13
50% 57% 5
% DK
4 Training motivation in the EU (values: weighted, rounded)
52%
60%
40%
4 0
4
42% 13
4 50%
30%
6 57% Cedefo
4 52% 32% What ki
6 40%
20%
4 if any, d
42%
6
30%
10%
6 Cedefo
6 32% 14% Austria What ki
6 20%
0%
5 if any, d
6
6 10% Stay up to date Improve performance Mandatory Career progression Personal reasons Slovakia
6 14% Austria
4 0%
5
7 Spain
paid by employer
6 90% Stay up to date ImproveCZperformance Mandatory Career progression Personal reasons Slovakia
6
NL DK FI 0%
4 NB: Share
6 80% of EU employees who participated in training in the last 12 months with positive answers to question: ‘What were your reasons for doing this
DE EE SE
7 training?’ SK SI Spain
UK
Source:
90%Cedefop ESJ survey.
employer
7
70% CZ
PL
AT BE
paid bytraining
6 FR LT
NL LU DK FI 0%
LV HU IT, IE MT
6 80% PT EE HR
60% DE SE Cedefop
received
SK SI
UK
7 ES Who paid for this
CY
70% RO
PL
BG AT BE
ing
50% LT FR
40% Training motivation in the EU (values: weighted, rounded)
60%
29 23 44 42% 0
30%
2124 64 13
50% 57% PREVIOUS CONTENTS NEXT 32%
2123 64 20% 52%
40%
2023 64
10% 42%
20 66 30% 14%
21 Austria
0% 32%
2121 65 20%
Figure 57. Public versus private financing of training, adult employees, 2014, EU-28
20 6 Stay up to date Improve performance Mandatory Career progression Personal reasons Slovakia
20 6
10%
19
20 46 14% Austria
20 75 0% Spain
21
training paid by employer
90%
CZ
20
20 66 Stay up to date Improve performance DK Mandatory Career progression Personal reasons
NL FI Slovakia
18 80%
19 46 DE EE SE
SK SI
UK
16
20 77 Spain
70% PL
AT BE
FR
by employer
90% LT
LU CZ MT
20 6 LV HU IT, IE
NL PT HRDK FI
60% Cedefop
received
18 6 80%
DE EE SE ES
SK SI CY Who paid
whopaid
RO BG UK
16 7 50% Your emp
70% PL
AT BE
training
LT FR
LU LV HU MT
IT, IE
employees
40% PT HR
60% Cedefop
received
vement EL
ES Who paid
Austria
e start job 30% RO BG CY
50% Your emp
% employees who%
0% 5% 10% 15% 20% 25% 30%
40% % employees who received training paid by public funds Slovakia

ease
vement EL
NB: Sample of employees who received training in the last 12 months; lines: EU-28 averages.
ejob tasks
start job 30%Cedefop ESJ survey.
Source:
Austria
Spain
Figure 58. Work place

ISCO 1 changes supported with training
training paidactivities
by public funds offered by employer by
0% 5% 10% Training for
15%all changes Training
20% for some changes
25% No training at
30%all
occupation, adult employees, 2014, EU-28

% employees who received Slovakia
Managers 24 48 26
ease
vement
job tasks ISCO 2 Professionals 18 52 29
e start job Training for all changes Training for some changes No training at all Spain
ISCO 3 Technicians and associate professionals 22 50 27 Cedefop
ISCO 4 ClericalISCO 1 Managers
support workers 2224 43 48 34 26 What kin
ease do you h
vement ISCO
ISCO 5 Service and2sales
Professionals
workers 18 26 52 41 3029
job tasks
e start job
ISCO 3 Technicians and associate
ISCO 6 Skilledprofessionals
agricultural 22 27 35 50 35 27 Cedefop
Cedefo
ISCO
ISCO47Clerical
Craft andsupport
trade workers 22
What kin
ESJ surve
21 4043 3634
ease do you h
ISCO
ISCO 5 Service
8 Plant and salesoperators
and machine workers 2628 41 30
vement
job tasks 38 33
e start job ISCO ISCO 6 Skilledoccupations
9 Elementary agricultural 20 27 33 35 44 35
EWC
Cedefo
ISCO 7 Craft and trade workers 21 40 36 ESJ surve
ease
vement ISCO 8 Plant and machine operators 28 % no change
38 or reduction in variety of job tasks
33since start of job
ejob tasks
start job ISCO or
9 Elementary occupations 20 33 44
EWC
Financial, insurance real estate services

NB: Share of EU employees who experienced recent changes in their workplace with positive answers to question: ‘Were any of these changes supported
ease % no change or reduction in variety of job tasks since start of job Cedefop
by trainingManufacturing
activities paidorforengineering
by your employer?’
job tasks Source: Cedefop ESJ survey. Overall, h
as it perpetuates theirorservices
disadvantaged labour adults from southern Europe Member States
Supply, management
Financial, insurance ororreal
treatment of water
estate services in relatio
market position over ortime, sustaining inequalities (Greece, Spain, Cyprus, Malta) engage, instead,
Services
Professional, relating tooreducation
scientific technical health
in incomes and in job prospects. in training outside of working hours (Figure 55).
Cedefo
Cedefop
Information technology
Manufacturingor communication
engineering ESJ surve
Significant differences are also observed in Confirming the findings of the latest wave of
Overall, h
Supply of gasorortreatment
Supply, management electricity,ofmining
water in relatio
training participation across EU Member States, the AES survey, job-related and employer-
EWC
which mirror the cross-country

or building patterns observed sponsored training activities are found to
Administration
Services relating toand support
education services
or health
Cedefo
inAgriculture,
the Supply
latest waveelectricity,
of the adult education survey comprise the lion’s share of adult learning. The
Information Construction
technology or communication
ESJ surve
(AES) carried out by Eurostat (Figure 55). ESJ survey indicates that most employees (57%)
horticulture,
of gas or forestry or mining
fishing % Som
Employees in Denmark, Finland, the UK, and engaged in training courses to stay up to date
and
Administration and support services Other EWC
also the Czech Republic and Slovakia, are the with changing skill needs of their job or to perform
% My
% My
most likely to receive training

or fishing during work hours their jobs better (52%) (Figure 56); mandatory
SocialConstruction
and personalor services
building
in Cultural
contrast to (arts,
adult workers from Bulgaria, training ranked third (42%). Approximately a third
Retail, horticulture,
Agriculture, sales, shop work or wholesale
forestry % Som
Greece and Romania. A high percentage of of respondents who underwent training did so to
Unweigh
and
industries entertainment)
Other % My
Transportation
Social and personalorservices
storage % My
Accommodation,
Retail, sales,catering or food
shop work services
or wholesale
Unweigh
Cultural industries (arts, entertainment) ISC
0% 10% 20% 30% 40%
CHAPTER 5
promote their career prospects. 14% of adult significantly with additional years of tenure with
workers pursued training for personal and non- the same employer, which also explains why a
job-related reasons. large percentage of underqualified workers admit
A total of 73% of employees underwent to having skills that are matched to or higher
training with full or partial financial support from relative to what is needed by their jobs. It is
their employer, while 11% financed it themselves evident that a substantial part of skill accumulation
or via a government programme (30). Significant of employees takes place informally, rendering the
differences exist between Member States with recognition and validation of such acquired, yet
regards to who pays for training. Employees in unrecognised, skills crucial for enabling their
the south of Europe and Balkans (Bulgaria, continued labour market progression and mobility
Greece, Spain, Cyprus, Malta and Romania) are (OECD, 2013; Cedefop, 2014b).
the least likely to have training funded by their
employers, a void that appears to be filled by the
public sector (Figure 57). Employers in Greece, 5.4. Bad-quality jobs can
in particular, are unlikely to pay for training (33%). undermine lifelong learning
Around a quarter of respondents in Greece paid
for training themselves or underwent training that Good skills lead to good jobs, but bad-quality jobs
was paid for by the government or another public can ultimately undermine lifelong learning and the
sector employer (26% and 23%, respectively). In development of better skills over the working life.
comparison, almost nine of 10 respondents in the The empirical analysis based on the ESJ
Czech Republic were likely to have undergone highlights the close relationship between the
training that was paid for by their employer (86%). degree to which adult employees improve their
The survey also asked if any changes in skills in their jobs and the skill intensity and job
technologies, products and working methods were complexity of such jobs. Individuals employed in
accompanied by training paid for by the employer. professions with high autonomy and learning at
Among those who experienced recent changes in work are found to be more likely to improve their
their job, around one in five EU employees skills since they started their job.
received training to deal with all changes (22%), Despite high and rising levels of skill demand
under half received training for some of the affecting the majority of the EU adult workforce,
changes (45%) and three of 10 received no as described in Sections 4.4 and 5.1, more than
training (31%). More than four in 10 respondents a third of jobs in some economic sectors (hotels
employed in elementary occupations received no and restaurants, transport, cultural industries and
training at all (44%), despite the fact that their jobs wholesale or retail trades), are still characterised
were affected by technological or other workplace by low need for continuous learning or for staff
changes (Figure 58) (31). Females, individuals who that can carry out a wide-ranging set of tasks
were unemployed prior to starting their job, and (Figure 59). Overall, 22% of respondents stated
employees in smaller-sized, private sector firms that the variety of their job tasks has remained the
are also found to have lower chances of having same or has fallen. Respondents working in
received training supported by their employers. elementary occupations are the most likely to say
Despite the importance of non-formal training, that the variety of their tasks has remained the
about seven in 10 survey respondents said that same or has fallen since starting their job with
they improved or acquired new skills in their job their current employer (43%). Elementary
via informal methods, such as learning by workers are also the most likely to say that the
interacting with colleagues at work. A significant need to learn new things in their current job has
proportion did so by learning at work through trial either stayed the same or decreased.
and error (59%). Skill formation tends to increase A sizeable share of EU jobs only requires a
(30) However, part of the employer-sponsored training could have been financed via a government-funded programme (e.g. ESF), although
respondents may not have been aware of it.
(31) Plant and machine operators and assemblers are the most likely to receive training for all the changes (28%), which may be due to
the physical risks involved in this type of occupation, resulting in a higher or mandatory provision of health and safety training.
ement ISCO 2 Professionals 18 52 29
start job
ISCO 3 Technicians and associate professionals 22 50 27 Cedefop
ISCO 4 Clerical support workers PREVIOUS22 CONTENTS NEXT 43 34
What kind
ease do you ha
ob tasks
Cedefo
ISCO 7 Craft and trade workers 21 40 36 ESJ surve
ISCO 8 Plant and machine operators
Share
Figure 59. ISCO of EUoccupations
jobs with stable
20 or decelerating
33 task complexity by economic sector,
ement 28 38 33
start job
2014, EU-28
9 Elementary 44
EWC
ease % no change or reduction in variety of job tasks since start of job

ob tasks
Financial, insurance or real estate services
Manufacturing or engineering Cedefop

Overall, h
Supply, management or treatment of water in relation
Cedefo
Information technology or communication
ESJ surve
Administration and support services EWC
Construction or building
Agriculture, horticulture, forestry or fishing % Som

and
Other % My s
Social and personal services % My s

Unweigh
Cultural industries (arts, entertainment)

ISC
0% 10% 20% 30% 40%
NB: Percentage of adult employees who reported no change or reduction in the variety of job tasks since they started their current job.
basic level of cognitive or digital skills. and are less likely to have experienced career
Cannot compare Managers
ISCO 6 Skil
Respondents working as service and sales progression in their jobs. More than half of all
Professionals
workers, craft and trade workers, plant and workers in elementary occupations (55%) and a
14 12
machine operators and in elementary significant share of skilled agricultural workers

Clerical support ISCO 8 P
occupations are more likely to require basic (42%), crafts and related trades workers (37%)
15 22
Technicians and associate professionals
literacyService
skills.
andAmarket
quartersalesof respondents employed and plant and machine operators (36%) do not
15 22
as skilled agricultural workers or in elementary require ICT skills in their jobs at all (Figure 60).
workers
occupations report that their current jobs do not For comparison, fewer than two in 10 workers in
20 22 Building, crafts or a related trades
require literacy skillsand(25% and 27% respectively) clerical and professional jobs require either basic
and one third say the same about the required or no ICT skills to do their jobs.
5 22 Plant and machine operators assembly
Unweigh
numeracy skills (37% and 29% respectively).

Skilled agricultural, forestry and fishing
While there are significant differences in the
Despite the EU’s ambition to improve importance of skills attached to the jobs of
5 22
competitiveness by investing in the digital individuals working in different sectors of

Elementary Ag
economy (32), a third of EU workers (33%) are economic 40%activity

50% and 60%
occupation
70% groups, the
22 11
employed in jobs that require only basic digital survey also reveals significant heterogeneity in
0% 10% 20% 30% 80% 90%
22 Su
skills or no skills at all, putting them at a the level of skill demand within broad clusters of
disadvantage in terms of their working and occupations Skilled
or industries. To obtain a clearer
4 2
everyday life. Individuals employed in jobs picture of the factors associated with better
High-skilled non-manual
requiring only basic or no ICT skills are observed chances of individuals working in jobs with higher
.4
to have lower gross hourly wages, on average, skill content, it is necessary first to strip out the
A
(32) http://ec.europa.eu/digital-agenda/en/news/digital-agenda-investment-digital-economy-holds-key-europes-future-prosperity-says-
commission [accessed 10.8.2015].
.2
Fin
Pro
40% Cultural in
0
ity
Agriculture, horticulture, forestry or fishing % Some of my
and need to
Other % My skills are
Social and personal services % My skills are
CHAPTER 5
87
Unweighted sam
Cultural industries (arts, entertainment) Matching better skills to better jobs
Figure 60. Share of jobs requiring basic or no ICT skills, adult employees, 2014, EU-28
ISCO 3 Tech
0% 10% 20% 30% 40%
ompare Managers
ISCO 6 Skilled agric
Professionals
12
Clerical support ISCO 8 Plant and
22
22
Service and market sales workers
22 Building, crafts or a related trades
22 Plant and machine operators and assembly

Unweighted sam
22
Adm
Elementary Agriculture
11
Su
0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
22 Supply, ma
influence of the individuals’High-skilled

occupation and sector offers further evidence of the tight link that exists
4 2
from the variance in the skill intensity of jobs. This between the continuous skill development of
Skilled non-manual
allows examination of the purged effect of a workers and the skill intensity of their jobs.
Reta
.4
further set of individual and job characteristics in Employers are also more likely to request a higher
Accommo
the likelihood that individuals are employed in level of qualification as a prerequisite to hiring
jobs that demand a high level of skills (33), (34). individuals in more skill-intensive jobs.
Regardless of the occupation and sector of It was mentioned in Section 2.2 that the
.2
Financial,
employment, individuals with lower levels of motives underlying the job choices of adults can
Professiona
education or with poor attitudes to learning are be important drivers of their likelihood of ending
Service
less likely to be in jobs with high skill demand. Skill up in a mismatched job. This illustrates the critical
Cultural industries
requirements are also influenced by the quality of role of interventions in career guidance and
0
Density
individuals’ labour market transitions: obtaining a counselling for tackling skill mismatch. Focusing
Skilled manual Elementary
job after undergoing a spell of unemployment on the relationship between individuals’ motives
.4
bears a negative association with a job’s skill and the degree of importance of skills in their
t you
content, compared to individuals who engaged in jobs, it is confirmed that individuals who selected
a direct job-to-job transition. Workers who were their jobs because of suitability with their own
Unweighted sam
long-term unemployed prior to starting their skills and qualifications, and to a lesser extent on
.2
current job face an even larger penalty in terms of the basis of higher monetary benefits and
M
GENDER
skill content. Entering a first job directly from promotion prospects, are more likely to be found
Fem
education and training is also associated with jobs in jobs with high skill needs.
34 to
of lower skill0 content compared 5to the previously 10 0The formation of skills through non-formal
0
employed, which is reasonable given that and informal learning is an investment by firms,
5 10 AGE 40 to
employers are more likely to trust high-skilled jobs in many respects akin to other forms of
Skills importance index 55 to
Graphs by isco (4 categories)
to individuals with a given level of work investments organisations undertake to put in

experience. The higher level of skill demand found place new productive infrastructure, capital
for individuals who benefited from upward vertical equipment and technological innovations.
career mobility (such as promotion) in their job However, investment in skillsUKhas a defining
35%
Unweighted sam
AT
30%
( ) In the remainder of this section, the terms ‘skill demand’, ‘skill needs’, ‘skill intensity’, ‘skill importance’ and ‘skill content’ are used
EL
33
Less than primar
interchangeably as descriptors of an summary index of the importance of a set of cognitive, digital and generic skills, as described in
e
Section
25% 4.4.
IE
(34) See Annex 2 for a description of the empirical methodology used to derive the main conclusions discussed in this section.
overskilled within jobs
ES
DE Upper secondary a
BE, FI PL
SE, HU
20% HR
DK NL
SI IT CY
FR
SK
CZ
characteristic, not shared by other types of professionals are employed in larger-sized firms
financial or capital investments, in that it is while SMEs rely more on service and market
embodied in human beings. Consequently, to be sales workers. Nevertheless, the ESJ data reveal
able to capture the returns from such investments that, after taking into account their differential
in skills, firms and individuals have to benefit from occupational/sectoral distribution, individuals in
stable employment relationships of sufficient SMEs require a higher level and spectrum of
duration that can allow them to amortise their skills to do their jobs compared to comparable
original outlays. workers in large establishments.
The ESJ highlights that there is a significant Another tool that employers can use to
relationship between the stability of adult stimulate or craft the level of skills needed by their
employees’ jobs and the extent to which their jobs staff is the task complexity of their jobs (38). There
are of high skill intensity. Individuals in non- is a significant degree of heterogeneity in skill
temporary contracts are more likely to be needs in jobs that is not captured by occupational
employed in jobs of increased skill content (35). classifications (Figure 61). Generally, employers
The analysis suggests that this positive competing in the same industry or occupational
correlation is mediated by the positive association markets have significant scope to influence the
between permanent jobs and those that are level of skills needed by their employees to
characterised by higher levels of increasing job perform their jobs. Using data from the OECD
complexity (36) or workplaces that entail a PIAAC survey, Pouliakas and Russo (2015) have
significant degree of dynamic change in shown that more complex jobs (offering greater
technologies and working practices (37). This autonomy and requiring solution of complex
suggests that it is not the permanent nature of a problems) tend to have higher cognitive (literacy
contract per se that is associated with more skill- and numeracy) skill requirements. Organisations
intensive jobs, but that employment stability can that employ smarter job design principles may
act as a means of coping with uncertainty and enjoy positive productivity premiums as a
dynamic workplace changes making higher consequence (Box 5). Such positive dividends will
demands on the skills of workers. also arise because jobs with higher skill content
A finding of particular interest is the negative are more likely to attract higher educated
relationship between workplace size and skill individuals, with positive attitudes to learning,
content: employees in smaller-sized firms are during the recruitment and selection stage (ibid,
more likely to believe that their jobs need a higher 2015). Task complexity is consistently and
level of skills to be performed (a higher level of robustly linked to the skill content of jobs. Given
skill intensity) compared to those in larger-sized that jobs involving a greater need to execute a
workplaces. It is well-reported that small and variety of tasks, or to learn new things, are
medium-sized enterprises (SMEs) provide less intrinsically dependent on possessing a high level
CVT to their staff relative to larger-sized of skills among the workforce, it follows from the
organisations (Cedefop, 2010d), which partly ESJ that they are also associated with lower
reflects the skewed distribution of occupations in chances that individuals are or will transit to a
the latter; for example, a greater share of state of overskilling.
(35) The data cannot give a clear-cut answer as to whether individuals in permanent contract jobs are more likely to progress to jobs/posts
that require a higher level of skills or whether employers are more likely to offer indefinite contract posts for positions of higher skill
content and productivity when recruiting labour (externally or internally). The data also do not account for differences in unobserved
characteristics of individuals who take up different types of contracts (risk aversion, time discount rates), even though the preferences
of the two groups for what is perceived to be an optimal job bundle can differ markedly (Pouliakas and Theodossiou, 2009).
(36) Increase in job complexity is a scale based on the score on three items in the ESJ questionnaire: change in the variety of tasks, in the
difficulty of the tasks, and in the need to learn new things. It ranges from 0 to 10; higher values denote increases in task complexity,
low values (close to zero) denote a reduction in task complexity.
(37) The index of organisational change gauges the degree of change that has taken place in the organisation during the five years preceding
the interview (or since the beginning of employment for those with less than five years of tenure). It is based on answers to four items:
changes in the technology used, in working methods and practices, in the product or service produced, and in the amount of contact
with clients or costumers. It ranges from one to four; higher values are associated with higher degree of organisational change.
(38) The job complexity scale is based on the score on three items representing job tasks (Q 13): responding to non-routine situations,
learning new things, and choosing yourself the way in which you do your work (autonomy). The scale ranges from 0 to 10, higher
values denoting higher degree of job complexity.
CHAPTER 5
Box 5. Heterogeneity of skills across and within occupations and the importance of job design
US data have confirmed that most wage variation takes then become jobs (Korpi and Tåhlin, 2009). The way
place within detailed (five-digit) occupational groups and organisations make use of workers’ skills is affected by
that establishments’ characteristics are just as important various contextual factors, some of which include the
as occupations in explaining wage variation (Lane, et al., technology adopted and the way the production process
2007; Barth et al., 2014). Wage dispersion across is organised (Bresnahan et al., 2002) (39).
establishments, above and beyond wage dispersion Management studies and organisational
across different occupations, has also contributed to the psychologists link job design to skill utilisation by
rise of wage inequality in Germany (Card et al., 2013). assuming two useful extreme job design configurations:
This evidence suggests a need to understand better complex jobs (performing non-routine operations,
what drives the difference in wages between jobs in the solving complex problems, possessing a degree of
same occupation group and within the same job title. autonomy in the way job tasks can be performed) versus
Many have attributed such variance in wages to different narrowly defined jobs (designed so that workers perform
personnel or human resource management (HRM) simple and repetitive tasks and use only a limited
practices that firms employ (recruitment, training, career number of skills). The first configuration is often termed
progression, job design, employee empowerment) and high-performance working practices, while the second
to their complementarity with the skills of the workforce case can correspond to the more traditional tayloristic
(Cedefop, 2012c). or fordist mode of work organisation. These two polar
One component of HRM is job design, which may cases represent the two extremes on a scale that
provide at least part of the answer. Jobs within the same subsumes the varied spectrum of possible ways in
job title may turn out to be very heterogeneous with which work can be organised (acknowledging that job
respect to the skills needed to attain proficient complexity comes in degrees). The actual positioning of
performance (see Figure 61). In fact, the above- an organisation on the complexity scale will be
mentioned evidence from Germany and the US suggests influenced by an organisations’ culture, business and
that wage variation across jobs, regardless of product strategies (Boxall and Purcell, 2011; Sung and
occupation, is linked to the way organisations design Ashton, 2015); job design is intimately linked to the
jobs and utilise skills (Spitz-Oener 2006; Autor, 2013; relative importance of human capital as a source of
Autor and Handel, 2013). competitive advantage, on which the success of the
To explain the relationship between job tasks and organisation rests (Takeuchi et al., 2007; Messersmith
skill mismatch, the task approach assumes that et al., 2011; Ployhart and Moliterno, 2011; Jiang et al.,
proficiency in more complex jobs requires workers with 2012; Jiang et al., 2013; Patel et al., 2013; Nyberg et
a wide spectrum of skills. Jobs can be designed in al., 2014; Ployhart and Hale, 2014; Ployhart et al., 2014;
various ways and this will have an impact on the way Fulmer and Ployhart, 2014). The relevance of human
workers will have to deploy their skills to be productive: capital as a source of competitive advantage is
some organisations will choose to bundle a variety of particularly significant for firms adopting advanced
different skills into complex jobs while other companies manufacturing techniques, and for organisations relying
will choose to break down production into precisely on product and process innovations, and on product or
described and rather narrowly defined tasks service customisation (Ichniowski et al., 2000).
encompassing only a limited number of skills, which
The analysis provides strong support to the

notion that good and stable jobs (non-temporary,
full-time, involving complex tasks within dynamic
organisations) are also jobs with higher skill
content that entail a smaller likelihood of
overskilling among adult workers.
(39) The way production is organised is influenced by the industry in which organisation operates and by its size. However, there remains
substantial variability in the way organisations craft jobs (simple versus complex set ups) even within industry and size class.
15 22
Service and market sales workers
20 22 PREVIOUS CONTENTS NEXT

Building, crafts or a related trades
Skills,
andqualifications and jobs in theassembly
EU: the making of a perfect match?
5 22 Plant machine operators and
Unweigh
5 22
Elementary Ag
Figure 61. Variance of skills importance

10% index
20% within
30%occupational groups,
60% adult workers,
22 11
2014, EU-28
0% 40% 50% 70% 80% 90%
22 Su
4 2
High-skilled Skilled non-manual
.4
A
.2
Fin
Pro
40% Cultural in
0
Density
Skilled manual Elementary

.4
or taught you
Unweigh
the job
.2
GENDER
0
0 5 10 0 5 10 AGE
Skills importance index

% Graphs by isco (4 categories)
5.5. Helping labour market

62). The probability of individuals making a
35%
30%transitions tackle skill

transition to or remaining stuck in jobs that
Unweigh
mismatch
UK
underutilise their skills varies ELacross Member

AT
States. Employees in Greece and the UK are

Less than
characterised by the highest persistence in skill

ndergone
The ESJ survey points out that the chances of

ning IE
underutilisation,DE followed by Germany, Spain,

25%
Mean % stay overskilled within jobs
exit from a persistent status of skill mismatch are SE, HUAustria and
ES
Poland. In these countries there are

Upper sec
BE, FI
greater
20% for individuals who change jobs, as
PL
NL high chances HRthat individuals will remain

opposed to relying on the more gradual overskilled both when changing jobs and
DK
adaptation of skills to skill needs taking place

SI IT CY
throughout their current employment.

FR
within jobs. Skill mismatch is a persistent

SK
Overall, 57% of EU employees did not

15% CZ
phenomenon and it is difficult for individuals LU, PTto change mismatch status between their previous
BG
change
10%
their skill mismatch status both within and and the start of the current job (they either
RO
between jobs. But theLV chances LT of transition to a remained matched or mismatched), but
EE
different
5% skill (mis)match state MT are greater when approximately 34% did so (40). Slightly more than
individuals engage in job mobility. a half of all respondents who were either
While 14% of EU employees remained in a overskilled or underskilled in their previous job
perpetual0% state of overskilling or skill moved to15% a different skill mismatch state (with
0%
underutilisation after changing to a new job, more approximately two thirds making the transition to
5% 10% 20% 25%
than 20% did so during their job tenure (Figure matched skills) (Table 13). For comparison, 65%
Mean % stay overskilled between jobs
(40) The remaining 9% of the adult employee sample comprises missing cases.
you
Unweighted sam
.2
M
PREVIOUS CONTENTS NEXT GENDER
Fem
CHAPTER 5
Matching better skills to better jobs 91 34 to
0
0 5 10 0 5 10 AGE 40 to
Skills importance index 55 to

Graphs by isco (4 categories)
Figure 62. Persistence of overskilling between and within jobs, adult workers, 2014, EU-28
35%
UK
Unweighted sam
AT
30% EL
Less than primary
IE
25%
Mean % stay overskilled within jobs
ES
DE Upper secondary a
BE, FI PL
SE, HU
20% HR
DK NL
SI IT CY
FR
SK
15% CZ
BG
LU, PT
10% RO
EE
LV LT
5% MT
0%
0% 5% 10% 15% 20% 25%
Mean % stay overskilled between jobs
NB: Based on responses of individuals who said their skills were higher than needed by their previous job, at the start of their current job and at present;
lines correspond to EU-28 averages.
Table 13. Skill mismatch transitions: previous job and start of current job, adult employees,
2014, EU-28
Start of current job My skills My skills are matched Some of my skills are
are higher than needed to what is needed lower than needed
Previous job to do my job to do my job to do my job
My skills are higher
46.4% 36.4% 17.2%
than needed to do my job
My skills are matched to

19.6% 61.8% 18.6%
what is needed to do my job
Some of my skills are lower

15.9% 37.3% 46.8%
than needed to do my job
of EU adult employees were ‘stuck’ in the same remaining overskilled (given that the overskilled
skill mismatch state between the start of their are more inclined to move to a new job in which
current job and the time of the survey. their skills are either below or matched to the level
Transitions from one job to another are needed in their new jobs). By contrast, adult
generally associated with greater chances of workers are more likely to move from being
individuals moving to or remaining in a state of underskilled to having matched or more skills as
underskilling; they have lower skills than needed they accumulate additional years of tenure within
to do their new job, and fewer chances of their jobs (Table 13).
Table 14. Distribution of adult workforce according to skill mismatch transitions between
previous job – start of current job – present job, adult employees, EU-28, 2014
Skill mismatch transition Transition from previous job Transition from start
to start of new job of new job to present
Matched-matched 32% 35%
Matched-overskilled 10% 13%
Matched-underskilled 10% 2%
Overskilled-overskilled 13% 21%
Overskilled-underskilled 5% 0.5%
Overskilled-matched 10% 5%
Underskilled-underskilled 5% 3%
Underskilled-matched 4% 14%
Underskilled-overskilled 2% 5%
Missing values 10% 2%
The above transitions suggest that EU and match their own skills and qualifications and
national mobility programmes have considerable greater financial distress are also more likely to
scope to contribute to better reallocation of have selected a completely different occupation
mismatched individuals to jobs that offer a better prior to commencing their new employment.
match with their skills. Stimulating individuals’ ‘Radical movers’, those who move to a
labour market mobility can typically take the form different occupation requiring a very different
of either movement across occupations or skills set (especially in the initial period of
mobility across different localities (41). employment), have greater chances of eventually
According to the ESJ survey, one in eight making a transition to a state of overskilling. As
(12%) respondents said that their previous job shown in Figure 63, 29% of adult workers who
was exactly the same as their current job, while moved to a completely different occupation ended
two in five (43%) said that their occupation is up in dead-end jobs, presumably because such
different, and three in 10 (31%) said it is similar. a move is also associated with a considerable
Individuals who embarked on a significant loss of occupation-specific skills (Panos et al.,
occupational change (a move to a different 2014), as opposed to 24% of those who remained
occupation relative to previous) are more likely to in exactly the same occupation.
be lower-educated, self-employed prior to These findings caution that some people
accepting the new job (potentially because of faced with disadvantaged individual or labour
prior failure or difficulty in the survival of a market circumstances may feel they have little
business) and to be employed in jobs that require choice but to change occupation, even if it is not
lower levels of (technical) skills. Individuals who a suitable fit for their skills. Effective policy
faced greater difficulties finding a job that can programmes that seek to stimulate occupational
(41) The ESJ survey enquired about these two types of labour market mobility. Respondents were asked the questions: ‘Were you working
in the same occupation in your previous job as in your current job?’ (a) Exactly the same; (b) similar; (c) different; (d) my current job
is my first job; (e) don’t know. And ‘Did you have to move home when you started your job with your current employer?’ (a) Yes, to a
different country; (b) yes, but within the same region of the country in which you live; (c) yes, to another region of the country in which
you live; (d) don’t know. Due to time limitations, it was not possible to retrieve the exact occupation code of respondents in their
previous job to compare with their current one; the survey relied, instead, on their subjective evaluation. The findings described in this
section are dependent on the respondents’ own evaluation of what constitutes a ‘different’ occupation to the one they held in their
previous job. With regards to the question on geographic mobility, examples of regions were offered to the respondents that correspond
to the conventional NACE 2 level, as defined by Eurostat.
CHAPTER 5
Figure 63. Skill mismatch by type of occupational transition, adult workers, 2014, EU-28
Talent in dead end Talent in standstill Talent unnourished

74
Before
73 100%
(misma
prefere
80% 42% 43% 44%
60%
34% 29% 26%

40%
20%
27% 29%
24%
DK FI 0%
Exactly the same occupation Similar occupation Different occupation

53 Before
mobility must ensure that they are accompanied On average, younger, high-educated males
49 LLL Fitted values (misma
by targeted vocational training and guidance DK are more likely to have moved to a different EU
35 prefere
services and social security support that can country prior to commencing their current job,
prolong and smooth the transition of these motivated by the need to advance their career
30 SE
individuals to better-matched professions. prospects. For instance, 20% of younger

FI
25
The ESJ survey confirms other evidence that respondents aged 24 to 40 moved home
intra-EU mobility is NLvery limited (European compared to 15% of those over 40. By contrast,
20
Commission, 2015a, pp. 87-122). Around one in AT geographic mobility is less common among
FR
five (17%) of respondents had to move home; 9% females,ES older and lower-educated individuals
UK
15 LU
moved to a different region, 6% moved but within and those who experiencedCZ previous spells of
SI EE
EL ES
the same region, andCY 2% had DE to move to a unemployment or inactivity; these are more likely
10
IT PT MT
different country. Across Europe, respondents PLin to HUbe vulnerable to economic shocks and to suffer
BE
IE
5
the countries most severely impacted by the from skill gaps. Policies that aim RO to support
LV
EL LT
prolonged economic15 crisis, Ireland, 17 Greece, 19individuals in finding work in other

23 EU countries,
HR SK BG
0
France and Italy, were the most likely to say that including those falling under the auspices of
13 21 25
they had moved home before they started their EURES, will need to pay greater attention to
Skill deficit
job. these target groups.

Cedefop PIACC
What kind of employment contract, What kind of employment contract
if any, do you have in your current job? do you have? Is that...
Austria 89 9 2 Austria 91 6 3
asons Slovakia 79 16 5 Slovakia 82 12 6
Spain 78 17 5 Spain 74 17 9
0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100%
% An indefinite/permanent contract % A fixed-term/temporary contract % Other
Cedefop PIACC
Who paid for this training (e.g. tuition, registration, fees)? Did an employer or prospective employer pay for tuition
Conclusions and lessons for policy

CHAPTER 6
6.1. No one-size-fits-all policies paths suited for their skills and qualifications is
for skill mismatch necessary in this case, as are policies to support
higher skill utilisation within their workplaces.
Qualifications and skill mismatches imply a Changing jobs is also found to be a potential
collective waste of talent and resources, with remedy, although the ESJ survey cautions that
potentially significant economic and social significant occupational transitions Talentmay be
consequences. Better management of skills and associated with higher subsequent skill
Talent in dead end Talent in standstill unnourished
human resources can lead to economic benefits, mismatches due to the loss of job-specific human
74 100%
72 73
as well as benefits in workers’ wellbeing. But capital.

71
policy
80% instruments will vary
42% depending on the type 43% Some EU employees have also 44% entered jobs
of qualification-skill mismatch situation (42). with lower education credentials than the ones
A population group of significant policy required at present time, but are nevertheless in
concern, given the marked waste in terms of lost
60%
possession of the proper or higher skills required
potential productivity, are individuals in dead-end to carry out their job tasks. This group of
jobs,
40% which require lower 34%skills than their own and individuals face another challenge. 26% For them it is
which offer limited opportunities for further crucial to certify and validate their learning, to the
29%
growth. This group appears to be locked in a low- extent that the absence of formal certification of
skill-intensity equilibrium, given that their jobs are their informal skills may inhibit their further career
characterised by low skill24% content. The provision progression.
20%
However. the ESJ survey shows that most

29%
of career guidance and counselling to enable

27%
such workers to find alternative jobs or career EU workers are employed in jobs with sufficient
SK UK DK FI 0%
Figure 64. Skill deficits and lifelong learning, adult workers, 2013-14, EU-28
Exactly the same occupation Similar occupation Different occupation
53
49 49 LLL Fitted values
47
35
DK
30 SE
FI
25
20
FR
NL
UK AT
15 LU
SI EE
MT CY EL ES ES
10 CZ
IT PT MT
CY BE
DE IE
5 LV
PL HU LT
EL
HR SK BG RO
0
13 15 17 19 21 23 25
Skill deficit
NB: Lifelong learning refers to persons aged 25 to 64 who stated that they received education or training in the four weeks preceding the survey; 2013
data.
Source: Cedefop ESJ survey; Eurostat.
Cedefop PIACC
(42) Desjardins
if any,(2014)
do youreaches
have in at a similar
currentpolicy
job? conclusion, with a different categorisation of Is
thethat...
skill mismatch groups: low-skill match
(individuals with low skills in low-skilled jobs), high-skill match (individuals with high skills in high-skilled jobs), low-skill mismatch
your do you have?
(individuals with low skills in high-skilled jobs) and high-skill mismatch (individuals with high skills in low-skilled jobs).
14% Austria 89 9 2 Austria 91 6 3
rsonal reasons Slovakia 79 16 5 Slovakia 82 12 6

CHAPTER 6
scope to enjoy further productivity dividends as market. It is also more likely to place individuals
long as continuous skill development is provided. on a virtuous path, characterised by high-skill
Policy efforts to address skill mismatch for this requirements in their jobs and opportunities for
group would require continued investment in their further skill accumulation. But the ESJ survey
skills and support to adult learning to enable them suggests that European policy-makers wishing to
to cope with the rapid evolution of their job tasks. support WBL should pay greater attention to
As shown in Figure 64, there is negative fields of study as well as occupations and
correlation between EU Member States with the industries where WBL is not yet common.
smallest incidence of adults participating in
lifelong learning in 2013 (the European
Commission’s main indicator of lifelong learning) 6.3. Active labour market
and the magnitude of skill deficits affecting their policies need greater focus
adult workforces a year later (correlation
coefficient = - 0.41; p-value = 0.03). Recent skill mismatch policies in Member States
have emphasised the need to retrain unemployed
School-to-work transition
workers, in particular those with long spells of
6.2. joblessness, focusing on augmenting and
and skill mismatch matching their skills with labour market needs in
emerging or predominant national industries
Concerted efforts are needed by EU policy- (Cedefop, 2015c). The findings of the ESJ survey
makers to prepare new labour market entrants indicate that a strategy focused solely on skill
adequately, particularly lower-skilled vulnerable upgrading may be less effective in the current
graduates, to make an effective transition into the economic environment, given that current
world of work. Given the marked persistence of unemployment includes a greater pool of skilled
skill mismatch, policies that target the and experienced individuals actively seeking to
establishment of better skill matches at the (re-)enter a depressed job market. For this
beginning of individuals’ labour market histories reason, active labour market policies must pay
should enjoy greater efficiency gains relative to greater attention to the quality and suitability of
those that seek to address the problem ex post. the jobs that the unemployed accept as part of
Improving the match of individuals’ skills with their transition back into employment (Tatsiramos,
their job needs at an early stage of their working 2010). Public employment services (PES) play a
lives is dependent on strengthening the channels key role in ensuring that the much-needed return
of communication between the worlds of to employment growth does not come at the
education and work, the provision of customised expense of lower-quality skill matches, which will
career guidance and counselling, the promotion ultimately compromise or prolong the job crisis in
of job mobility, the provision of better intelligence coming years (World Economic Forum, 2014).
on labour market opportunities and support to job New innovative methods and ICT tools have also
networks (World Economic Forum, 2014). It is been recently adopted by several Member States
also critically reliant on the timing of entry into the attempting to ensure a better degree of matching
labour market, given that finding a job in a slack between the skills of job applicants and of
job market may undermine an individual’s long- enterprises (Cedefop, 2015c). Utilising the full set
term career prospects. of European tools for education and training for
The ESJ survey therefore calls for the further the purposes of active labour market policies
strengthening of EU and national programmes (including accreditation/recognition of skills of the
aimed at delivering specialised training modules, unemployed, competence-directed training) is a
assistance to jobseekers, and matching and strategy adopted by few EU countries at present,
placement services via ICT tools aimed at but which has the potential to improve the overall
younger-aged people, such as ‘Your first EURES effectiveness of employment services (ibid,
job’. 2015a).
WBL, in particular, offers a beneficial strategy The ESJ survey also highlights the
for aiding a speedier transition into the job importance of PES and career guidance services
focusing on the promotion of workers’ skills, 6.5. Employment stability leads

aspirations and interests when supporting the
to continued skill formation
reintegration of unemployed workers. Workers
who base their career decisions (accepting a job Excessive focus on numerical flexibility could be
offer) on having their own skills and qualifications counterproductive as it undermines skill
match those needed by their jobs are generally formation. Non-standard and precarious forms of
more likely to end up in skill-intensive jobs which employment contracts can foster labour market
ensure continued skill development. segmentation. Several Member States suffer
Rapid economic restructuring in turbulent from a dual labour market where such contracts
economies may require some degree of are accepted involuntarily and associated with
occupation change for those who wish to return fewer training opportunities, limited social security
to the labour market. The ESJ survey rights and lower job satisfaction than permanent
emphasises that individuals embarking on a contracts (Pouliakas and Theodossiou, 2010).
radical change in occupation would greatly Young people are more likely to have precarious
benefit from accompanying training during the employment terms and conditions, a share that
transition. Significant readjustment in the match has increased since 2007, while they also have a
of their skills with the requirements of their new higher propensity to be employed in part-time
jobs is also required. jobs (European Commission, 2014a).
Temporary contracts can be effective
One-shot solutions will be

stepping stones into regular and secure
6.4. employment, particularly in some Member State
short-lived apprenticeship schemes. Most job creation in the
post-crisis era has been driven in temporary jobs.
The ESJ survey cautions the importance of The ESJ survey suggests that, in addition to the
preventing skill mismatch in a dynamic framework risk of lower earnings and job insecurity,
that acknowledges the possibility of skill gain and temporary labour contracts may come at a
skill loss by workers and changing job content significant long-term cost, as they are associated
(Desjardins, 2014). One-shot solutions to improve with jobs that demand a lower level of skills. A
match between the skills of adults and those focus on flexibility also shortens time on the job,
needed in the labour market, including the which is crucial for the development of skills.
acquisition of additional qualifications or a formal Employer overreliance on non-stable contracts,
certification of the (informal) skills of employees, which may be necessary in coping with erratic
may be short-lived as they will be insufficient to production fluctuations, must be balanced against
sustain a continuous skills balance over the the need to ensure continued learning and
duration of individuals’ job tenure. Continuous utilisation of individuals’ skills over the long-term.
investment in training and other forms of on-the-
job learning are needed to maintain an adequate
balance between increasing job demands 6.6. Skill demand must be
(spurred by workplace innovation and changes in stimulated in the EU
the complexity of jobs) and workers’ skills.
Contrary to the received wisdom that SMEs Even though one in four European workers
have a limited capacity to provide enough acknowledge that they suffer from skill gaps when
training, jobs in SMEs tend to require broader use entering a new job, the ESJ survey highlights that
of skills (jobs are more skill-intensive) and a non-trivial share of employees are in dead-end
workers in SMEs experience faster skills jobs that underutilise their skills. Such jobs tend
development. The ESJ survey confirms the to be associated with lower-paid and precarious
centrality of SMEs in the skill formation process; work and can undermine the value and further
they might invest less in training, but they are adaptation of the skills of individuals with a high
comparatively better at setting up effective on- level of education. Further, the ESJ has also
the-job learning systems. shown that tackling the skill deficits of
unnourished talent requires greater commitment
CHAPTER 6
to their continuing learning, which is intrinsically Aligning industrial policy with substantial
dependent on raising the bar with regards to the institutional transformation in VET systems to
skill intensity of their jobs. Even among the group meet the needs of industry is acknowledged as a
of individuals who are in ‘standstill’, a high level pertinent strategy for mitigating skill mismatch, as
of skill needs in their jobs is necessary to balance evidenced by the experiences of the south-east
their continuous learning and to avoid their Asian tiger economies in previous decades.
transition to overskilling and associated Industrial activity in the EU has become
demotivation over time. increasingly integrated into global value chains
It is important for EU policies to stimulate the over past decades, resulting in increasing
level of skills required and used in jobs. employment of highly skilled labour (Bruegel,
Employers competing in the same industry or 2013; Timmer et al., 2014) (43). However, the ESJ
occupation markets have significant scope to survey identifies a number of economic sectors
influence the level of skills needed by their (including administration of support services;
employees to perform their jobs, in light of their retail and sales; accommodation, catering and
marked heterogeneity in managerial practices food services; transportation and storage; cultural
and job design (Bloom and van Reenen, 2007; industries and social and personal services),
2011; Pouliakas and Russo, 2015). Promoting mostly reliant on service provision, which are
high-skill work and organisational practices that susceptible to skills underutilisation. In these
foster work complexity is an important course of sectors greater policy efforts are required to
action, given that jobs entailing greater employee revamp skill demand.
autonomy, learning and task discretion are Governments, as large providers, funders
generally associated with higher skills needed to and procurers of goods and services to the
perform them. But the exchange of best practices market economy, have multiple policy levers for
on job and workplace design is unlikely to be influencing employers’ demand for skills.
sufficient, given that employers are typically Governments can support local enterprise
bound by varying sets of norms, capabilities, partnerships that seek to bring together local skill
cultures, historical trajectories and idiosyncratic strategies with wider economic development
‘social architectures’ influenced by their priorities. They can also legislate on mandatory
immediate industry or environment (Green, qualification or skill standards in occupations as
2013). Organisational learning with regards to well as adjusting the lower floor of minimum
best practices is also likely to be slow and wages in accordance with the conditions of the
incremental. labour market to prevent overreliance on low-
Directing investment expenditure to projects skilled immigrant labour.
with high capital-skill complementarity (such as
via the European fund for strategic investments)
is also necessary for raising the level of skill 6.7. Stimulating better EU
demand. Supporting businesses in innovation of labour market skill matches
high value-added products, services and
business models via EU modernisation During the recent economic crisis there has been
investments in industry and research and a growing trend in the EU of higher-educated
development is a requisite for shifting EU firms individuals taking up jobs requiring lower skills
towards a high-skills equilibrium, in which they than they possess (European Commission,
can compete on the basis of higher-end product 2012a; ILO, 2014). Searching for a job in a
market strategies. Financing investment projects depressed labour market, when individuals are
with a long-term horizon will induce employers to confronted with financial constraints or
adopt a strategic approach to talent management inadequate labour market opportunities,
and opt out of low-skill, low-cost production contributes to individuals accepting mismatched
strategies. jobs. At times of declining average incomes in
(43) http://www.bruegel.org/publications/publication-detail/view/795-manufacturing-europes-future/ [accessed 12.8.2015].

many European households, the provision of Given the subdued level of employment
social insurance and social security benefits creation in the EU economy relative to the pre-
could potentially provide adequate breathing crisis era, it is evident that stimulation of
space to individuals to seek and find jobs that are aggregate demand is necessary to avoid
a better match with their skills. unsuitable and wasteful skill-job matches, which
The ability to compete on the basis of skills is can have a lasting and damaging effect on the
dependent on significant investment on behalf of health of the EU labour market. A commitment by
both workers and organisations with a long-term governments to the goal of full employment and
horizon. The prolonged economic crisis is to the stability of industrial relations is a key
jeopardising these investments because the ingredient that can serve as collateral to support
incentives and rate of return on human capital is the decisions of all labour market agents
undermined by increased uncertainty in finding (individuals, firms and VET providers) in investing
and sustaining a good job. Willingness to invest in better and stronger skills for the EU’s future.
in improving job quality on the part of employers
is also compromised by the greater threat of
bankruptcy and need for business survival in the
current economic climate.
List of abbreviations
AES adult education survey

CVET continuing vocational education and training
CVT continuing vocational training
EULFS European labour force survey
ESJ European skills and jobs
EU European Union
EWCS European working conditions survey
HRM human resource management
ICT information and communications technology
ISCO international standard classification of occupations
IVET initial vocational education and training
LTU unemployed for more than one year before starting current job
NACE statistical classification of economic activities in the European Community
OECD Organisation for Economic Cooperation and Development
OLS ordinary least squares
PIAAC survey of adult skills
SMEs small and medium-sized enterprises
VET vocational education and training
WBL work-based learning
NEETs individuals neither in employment nor in education or training
References
(URLs accessed 5.8.2015)
Allen, J.P.; Levels, M.; van der Velden, R.K.W. (2014). Skill mismatch and use in developed countries:
evidence from the PIAAC study. RM/13/061. Maastricht: Maastricht University.
http://digitalarchive.maastrichtuniversity.nl/fedora/get/guid:f5190876-00ca-4769-b5e3-
5a961b83cd31/ASSET1
Autor, D.H. (2013). The ‘task approach’ to labour markets: an overview. Journal for labour market
research Vol. 46, No 3, pp. 185-199. http://doi.org/10.1007/s12651-013-0128-z
Autor, D.H.; Handel, M.J. (2013). Putting tasks to the test: human capital, job tasks, and wages. Journal
of labor economics, Vol. 31, No 2, pp. S59-S96. http://doi.org/10.1086/669332
Autor, D.H.; Levy F.; Murnane J.R. (2003). The skill content of recent technological change: an empirical
exploration. The quarterly journal of economics, Vol. 118, No 4, pp. 1279-1333.
http://doi.org/10.1162/003355303322552801
Baert, S.; Cokkx, B.; Verhaest, D. (2012). Overeducation at the start of the career: stepping stone or
trap? IZA discussion paper; No 6562. http://ftp.iza.org/dp6562.pdf
Barth, E. et al. (2014). It’s where you work: increases in earnings dispersion across establishments and
individuals in the US. IZA discussion paper series; No 8437. http://ftp.iza.org/dp8437.pdf
Bloom, N.; van Reenen, J.V. (2007). Measuring and explaining management practices across firms and
countries. The quarterly journal of economics, Vol. 122, No 4, pp. 1351-1408.
http://doi.org/10.1162/qjec.2007.122.4.1351
Bloom, N.; Van Reenen, J.V. (2011). Human resource management and productivity. In: Orley, A.; David,
C. (eds). Handbook of labor economics. Amsterdam: Elsevier, Vol. 4, Part 2, pp. 1697-1767.
http://doi.org/10.1016/S0169-7218(11)02417-8
Boxall, P.; Purcell, J. (2011). Strategy and human resource management. New York: Palgrave Macillian.
Bresnahan, T.F.; Brynjolfsson, E.; Hitt, L.H. (2002). Information technology, workplace organisation, and
the demand for skilled labour: firm-level evidence. The quarterly journal of economics Vol. 117, No
1, pp. 339-376. http://doi.org/10.1162/003355302753399526
Bulmahn, G.; Kräkel, M. (2002). Overeducated workers as an insurance device. Labour, Vol. 16, No 2,
pp. 383-402. http://doi.org/10.1111/1467-9914.00200
Cappelli, P. (2008). Talent on demand: managing talent in an age of uncertainty. Boston: Harvard
Business School Press.
Cappelli, P.; Keller, J.R. (2014). Talent management: conceptual approaches and practical challenges.
Annual review of organisational psychology and organisational behavior, Vol. 1, pp. 305-331.
http://doi.org/10.1146/annurev-orgpsych-031413-091314
Card, D.; Heining, J.; Kline, P. (2013). Workplace heterogeneity and the rise of west-German wage
inequality. The quarterly journal of economics, Vol. 128, No 3, pp. 967-1015.
http://doi.org/10.1093/qje/qjt006
Cedefop (2009). Skill mismatch: identifying priorities for future research: Cedefop research arena
(Cedra). Thessaloniki: Cedefop. Cedefop working paper; No 3.
http://www.cedefop.europa.eu/etv/Upload/Information_resources/Bookshop/537/6103_en.pdf
Cedefop (2010a). The skill matching challenge: analysing skill mismatch and policy implications.
Luxembourg: Publications Office. http://www.cedefop.europa.eu/EN/Files/3056_en.pdf
Cedefop (2010b). The right skills for silver workers: an empirical analysis. Luxembourg: Publications
Office. Cedefop research paper; No 8. http://www.cedefop.europa.eu/EN/Files/5508_en.pdf
Cedefop (2010c). Labour market polarisation and elementary occupations in Europe: blip or long-term
trend? Luxembourg: Publications Office. Cedefop research paper; No 9.
www.cedefop.europa.eu/files/5509_en.pdf
References 101
Cedefop (2010d). Employer-provided vocational training in Europe: evaluation and interpretation of the
third continuing vocational training survey. Luxembourg: Publications Office. Cedefop research
paper; No 2. http://www.cedefop.europa.eu/EN/Files/5502_en.pdf
Cedefop (2011). Migrants, minorities, mismatch? Skill mismatch among migrants and ethnic minorities
in Europe. Luxembourg: Publications Office. Cedefop research paper; No 16.
http://www.cedefop.europa.eu/EN/Files/5516_en.pdf
Cedefop (2012a). From education to working life: the labour market outcomes of VET. Luxembourg:
Publications Office. http://www.cedefop.europa.eu/files/3063_en.pdf
Cedefop (2012b). Preventing skills obsolescence: rapid labour market changes leave too many workers
at risk of losing their skills. Briefing note; July 2012. http://www.cedefop.europa.eu/files/9070_en.pdf
Cedefop (2012c). Skill mismatch: the role of the enterprise. Luxembourg: Publications Office. Cedefop
research paper; No 21. http://www.cedefop.europa.eu/EN/Files/5521_en.pdf
Cedefop (2012d). Future skills supply and demand in Europe: forecast 2012. Luxembourg: Publications
Office. Cedefop research paper; No 26. http://www.cedefop.europa.eu/EN/Files/5526_en.pdf
Cedefop (2013). Benefits of vocational education and training in Europe for people, organisations and
countries. Luxembourg: Publications Office. http://www.cedefop.europa.eu/files/4121_en.pdf
Cedefop (2014a). Skill mismatch: more than meets the eye. Briefing note; March 2014.
http://www.cedefop.europa.eu/files/9087_en.pdf
Cedefop (2014b). Use of validation by enterprises for human resource and career development
purposes. Luxembourg: Publications Office. Cedefop reference series; No 96.
Cedefop (2015a). Skill shortages and skill gaps in European enterprises: Striking a balance between
vocational education and training and the labour market. Cedefop reference series; No102.
http://dx.doi.org/10.2801/042499
Cedefop (2015b). Stronger VET for better lives: Cedefop’s monitoring report on vocational education
and training policies 2010-14. Luxembourg: Publications Office. Cedefop reference series; No 98.
Cedefop (2015c). Tackling unemployment while addressing skill mismatch: lessons learned from policies
and practices in European Union countries. Cedefop research paper; No 46.
http://dx.doi.org/10.2801/648140
Clark, B.; Joubert, C.; Maurel, A. (2014). The career prospects of overeducated Americans. NBER
working paper No 20167. http://doi.org/10.3386/w20167
Desjardins, R. (2014). Rewards to skill supply, skill demand and skill match-mismatch: studies using
the adult literacy and lifeskills survey. Lund economic studies; No 176. Lund: Media Tryck.
http://lup.lub.lu.se/luur/download?func=downloadFile&recordOId=4350627&fileOId=4350629
Draghi, M. (2014). Unemployment in the euro area. Speech delivered at the Annual central bank
symposium in Jackson Hole, 22 August 2014.
www.ecb.europa.eu/press/key/date/2014/html/sp140822.en.html
Eichhorst, W. et al. (2012). A roadmap to vocational education and training systems around the world.
IZA discussion paper series; No 7110. http://ftp.iza.org/dp7110.pdf
EU skills panorama (2014). STEM skills analytical highlight .
http://euskillspanorama.cedefop.europa.eu/App_Controls/Documents/ShowDocument.aspx?docum
entid=150&
Eurofound (2013). Third European company survey: first findings. Luxembourg: Publications Office.
http://www.eurofound.europa.eu/publications/htmlfiles/ef1386.htm
European Central Bank (2012). Euro area labour markets and the crisis: structural issues report.
https://www.ecb.europa.eu/pub/pdf/other/euroarealabourmarketsandthecrisis201210en.pdf
European Commission (2008). New skills for new jobs: anticipating and matching labour market needs:
communication from the Commission to the European Parliament, the Council, the European
Economic and Social Committee and the Committee of the Regions. COM (2008) 868 final.
ec.europa.eu/social/BlobServlet?docId=1496&langId=en
European Commission (2010). An agenda for new skills and jobs: a European contribution towards full
employment: communication from the Commission to the European Parliament, the Council, the
European Economic and Social Committee and the Committee of the Regions. COM (2010) 682
final. http://eur-lex.europa.eu/resource.html?uri=cellar:776df18f-542f-48b8-9627-
88aac6d3ede0.0003.03/DOC_1&format=PDF
European Commission (2012a). Rethinking education: investing in skills for better economic and
social outcomes: communication from the Commission to the European Parliament, the Council,
the European Economic and Social Committee and the Committee of the Regions. COM (2012)
669 final. www.cedefop.europa.eu/files/com669_en.pdf
European Commission (2012b). Towards a job-rich recovery: communication from the Commission to
the European Parliament, the Council, the European Economic and Social Committee and the
Committee of the Regions. COM (2012) 173 final.
http://ec.europa.eu/health/workforce/docs/communication_towards_job_rich_recovery_en.pdf
European Commission (2013a). The skill mismatch challenge in Europe, Employment and Social
Developments in Europe 2012. Commission staff working document. SWD(2013) 2 final.
http://register.consilium.europa.eu/doc/srv?l=EN&f=ST%205571%202013%20ADD%209
European Commission (2013b). Growth potential of EU human resources and policy implications for
future economic growth. Luxembourg: Publications Office. http://dx.doi.org/10.2767/79370
European Commission (2014a). Employment and social developments in Europe 2014. Luxembourg:
Publications Office. http://ec.europa.eu/social/BlobServlet?docId=13404&langId=en
European Commission (2014b). Education and training monitor 2014. Luxembourg: Publications
Office. http://ec.europa.eu/education/library/publications/monitor14_en.pdf
European Commission (2014c). Gender inequalities in the school-to-work transitions in Europe: short
statistical report No 4. http://ec.europa.eu/justice/gender-equality/files/documents/
140502_gender_equality_workforce_ssr4_en.pdf
European Commission (2014d). European vacancy and recruitment report 2014. Luxembourg:
Publications Office. http://ec.europa.eu/social/BlobServlet?docId=11897&langId=en
European Commission (2014e). Is unemployment structural or cyclical? Main features of job matching
in the EU after the crisis. European economy, economic papers; No 527.
http://ec.europa.eu/economy_finance/publications/economic_paper/2014/pdf/ecp527_en.pdf
European Commission (2015a). Labour market and wage developments in Europe 2015.
https://webgate.ec.europa.eu/emplcms/social/BlobServlet?docId=14290&langId=en
European Commission (2015b). Supporting skills development and matching in the EU. Employment
and social developments in Europe 2015. Chapter 5 (jointly with Cedefop and Eurofound).
European Commission; Joint Research Centre; Institute for the Protection and Security of the Citizen
(2014). Occupational mismatch in Europe: understanding overeducation and overskilling for
policy-making. Luxembourg: Publications Office.
https://ec.europa.eu/jrc/sites/default/files/occupational_mismatch_in_europe.pdf
European Commission; Latvian Presidency of the Council of the European Union (2015). Riga
conclusions 2015: on a new set of medium-term deliverables in the field of VET for the period
2015-20, as a result of the review of short-term deliverables defined in the 2010 Bruges
communiqué. https://eu2015.lv/images/notikumi/VET_RigaConclusions_2015.pdf
European Commission; EACEA; Eurydice (2015). Adult education and training in Europe: widening
access to learning opportunities. Eurydice report. Luxembourg: Publications Office.
http://eacea.ec.europa.eu/education/eurydice/documents/thematic_reports/179en.pdf
Freeman, R.B (2015). Who owns the robots rules the world. IZA world of labor 2015; No 5.
http://doi.org/10.15185/izawol.5
References 103
Fulmer, I.S.; Ployhart, R.E. (2014). Our most important asset: a multidisciplinary/multilevel
review of human capital valuation for research and practice. Journal of management, Vol. 40,
No 1, pp. 161-192. http://dx.doi.org/10.1177/0149206313511271
Green, F. (2013). Skills and skilled work: an economic and social analysis. Oxford: Oxford University
Press.
Green, F.; Zhu, Y. (2010). Overqualification, job dissatisfaction, and increasing dispersion in the
returns to graduate education. Oxford economic papers, Vol. 62, No 4, pp. 740-763.
http://doi.org/10.1093/oep/gpq002
Hanushek et al. (2013). Returns to skills around the world: evidence from PIAAC. OECD education
working paper; No 101.
http://www.oecd.org/officialdocuments/publicdisplaydocumentpdf/?cote=EDU/WKP(2013)16&docL
anguage=En
Ichniowski, C. et al. (2000). The American workplace: skills, compensation, and employment
involvement. Cambridge: Cambridge University Press.
ILO (2014). Skill mismatch in Europe: statistics brief. http://www.ilo.org/wcmsp5/groups/public/—-
dgreports/—-stat/documents/publication/wcms_315623.pdf
Institute for Public Policy Research (2014). European jobs and skills: a comprehensive review 2014.
http://www.ippr.org/files/publications/pdf/European-jobs-and-skills_Apr2014.pdf?noredirect=1
Jiang, K.; Takeuchi, R.; Lepak, D.P. (2013). Where do we go from here? New perspectives on the
black box in strategic human resource management research. Journal of management studies,
Vol. 50, No 8, pp. 1448-1480. http://doi.org/10.1111/joms.12057
Jiang, K. et al. (2012). How does human resource management influence organisational outcomes? A
meta-analytic investigation of mediating mechanisms. Academy of management journal, Vol. 55,
No 6, pp. 1264-1294. http://doi.org/10.5465/amj.2011.0088
Korpi, T.; Tåhlin, M. (2009). A tale of two distinctions: the significance of job requirements and informal
workplace training for the training gap.
Lane, J. I.; Salmon, L.A.; Spletzer, J.R. (2007). Establishment wage differentials. Monthly labor
review, Vol. 130, No 4, pp. 3-18. http://www.bls.gov/opub/mlr/2007/04/mlr200704.pdf
Lazear, E.P.; Spletzer, J.R. (2012). The US labour market: status quo or a new normal? Proceedings
of the Economic policy symposium, Jackson Hole, Federal Reserve Bank of Kansas, pp. 405-451.
Leuven, E.; Oosterbeek, H. (2011). Overeducation and mismatch in the labor market. IZA discussion
paper series; No 5523. http://ftp.iza.org/dp5523.pdf
Liu, K.; Salvanes, K.J.; Sorensen, E.O. (2012). Good skills in bad times: cyclical skill mismatch and
the long-term effects of graduating in a recession. IZA discussion paper series; No 6820.
http://ftp.iza.org/dp6820.pdf
Longhi, S. (2015). Do the unemployed accept jobs too quickly? A comparison with employed job
seekers. IZA discussion paper series; No 9112. http://ftp.iza.org/dp9112.pdf
Mañé, F. (2013). Using the job requirements approach and matched employer-employee data to
investigate the content of individuals’ human capital. In: Green, F.; Keese, M. (eds). Job tasks,
work skills and the labour market. Paris: OECD Publishing.
Manpower (2014). 2014 talent shortage survey.
http://www.manpowergroup.com/wps/wcm/connect/ec2b6e68-bc26-4e5a-8493-
78a9b53c5ab8/2014+Talent+Shortage+Infographic-Final.pdf?MOD=AJPERES
Mavromaras, K.; McGuinness, S.; Wooden, M. (2007). Overskilling in the Australian labour market.
Australian economic review, Vol. 40, No 3, pp. 307-312. http://doi.org/10.1111/j.1467-
8462.2007.00468.x
Mavromaras, K.; McGuinness, S.; Fok, K. (2012). Overskilling dynamics and education pathways.
Economics of education review, Vol. 3, No 5, pp. 619-28.
Mavromaras, K.; Sloane, P.; Wei, Z. (2015a). The scarring effects of unemployment, low pay and
skills underutilisation in Australia compared. Applied economics, Vol. 47, No 23, pp. 2413-2429.
http://doi.org/10.1080/00036846.2015.1008762
Mavromaras, K.; McGuiness, S.; Pouliakas, K. (2015b). Decomposing earnings penalties of

overeducated workers: evidence from the Cedefop European skills and jobs (ESJ) survey.
Working paper presented at the Cedefop-IZA workshop on skills and skill mismatch, Thessaloniki,
29 and 30 October 2015.
http://www.iza.org/conference_files/2015_Skill_Mismatch/pouliakas_k5607.pdf
Mavromaras, K. et al. (2010). Job mismatches and labour market outcomes: panel evidence on
Australian university graduates. IZA discussion paper series; No 5083.
http://ftp.iza.org/dp5083.pdf
McGuinness, S. (2006). Overeducation in the labour market. Journal of economic surveys, Vol. 20,
No 3, p. 387-418. http://dx.doi.org/10.1111/j.0950-0804.2006.00284.x
McKinsey Centre for Government (2012). Education to employment: designing a system that works.
http://mckinseyonsociety.com/downloads/reports/Education/Education-to-Employment_FINAL.pdf
Messersmith, J. G. et al. (2011). Unlocking the black box: exploring the link between high-
performance work systems and performance. Journal of applied psychology, Vol. 96, No 6, pp.
1105-1118. http://dx.doi.org/0.1037/a0024710
Nyberg, A. J. et al. (2012). Resource-based perspectives on unit-level human capital: a review and
integration. Journal of management, Vol. 40, No 1, pp. 316-346.
http://dx.doi.org/10.1177/0149206312458703
OECD (2013). OECD Skills outlook 2013: first results from the survey of adult skills. Paris: OECD
Publishing. http://dx.doi.org/10.1787/9789264204256-en
Oreopoulos, P.; von Wachter, T.; Heisz, A. (2012). The short- and long-term career effects of
graduating in a recession. American economic journal: applied economics, Vol. 4, No 1, pp. 1-29.
http://dx.doi.org/10.1257/app.4.1.1
Panos, G.A.; Pouliakas, K.; Zangelidis, A. (2014). Multiple job holding, skill diversification and mobility.
Industrial relations: a journal of economy and society, Vol. 53, No 2, pp. 223-272.
http://dx.doi.org/10.1111/irel.12055
Patel, P.C.; Messersmith, J.G.; Lepak D.P. (2013). Walking the tightrope: an assessment of the
relationship between high-performance work systems and organisational ambidexterity. Academy
of management journal, Vol. 56, No 5, pp. 1420-1442. http://dx.doi.org/10.5465/amj.2011.0255
Pellizzari, M.; Fichen, A. (2013). A new measure of skills mismatch: theory and evidence from the
survey of adult skills (PIAAC). Paris: OECD Publishing. OECD social, employment and migration
working papers; No 153. http://dx.doi.org/10.1787/5k3tpt04lcnt-en
Ployhart, R.E.; Hale, D. (2014). The fascinating psychological microfoundations of strategy and
competitive advantage. Annual review of organizational psychology and organizational behavior,
Vol. 1, No 1, pp. 145-172. http://dx.doi.org/10.1146/annurev-orgpsych-031413-091312
Ployhart, R.E.; Moliterno, T.P. (2011). Emergence of the human capital resource: a multilevel model.
Academy of management review, Vol. 36, No 1, pp. 127-150.
http://is.vsfs.cz/el/6410/zima2012/NA_HRM/um/3904027/2011_Ployhart_Moliterno_Emergenceof
HumanCapital_MultilevelModel127.full_1_.pdf
Ployhart, R.E. et al. (2014). Human capital is dead; long live human capital resources! Journal of
management, Vol. 40, No 2, pp. 371-398. http://dx.doi.org/10.1177/0149206313512152
Pouliakas, K.; Theodossiou, I. (2009). Measuring the utility cost of temporary employment contracts
before adaptation: a conjoint analysis approach. Economica, Vol. 77, No 308, pp. 688-709.
http://dx.doi.org/10.1111/j.1468-0335.2009.00786.x
Pouliakas, K.; Theodossiou, I. (2010). Differences in the job satisfaction of high-paid and low-paid
workers across Europe. International labour review, Vol. 149, No 1, pp. 1-29.
http://dx.doi.org/10.1111/j.1564-913X.2010.00073.x
Pouliakas, K.; Russo, G. (2015). Heterogeneity of skill needs in relation to job tasks: evidence from
the OECD PIAAC survey, IZA discussion paper series; No 9392.
http://www.iza.org/en/webcontent/publications/papers/viewAbstract?dp_id=9392
References 105
Quintini, G. (2011). Overqualified or underskilled: a review of existing literature. Paris: OECD

Publishing. OECD social, employment and migration working papers; No
121.http://dx.doi.org/10.1787/5kg58j9d7b6d-en
Robst, J. (1995). Career mobility, job match, and overeducation. Eastern economic journal, Vol. 21,
No 4, pp. 539-550.
http://college.holycross.edu/RePEc/eej/Archive/Volume21/V21N4P539_550.pdf
Sattinger, M. (2012). Qualitative mismatches. Foundations and trends in microeconomics, Vol. 8, No
1-2, pp. 1-168. http://dx.doi.org/10.1561/0700000052
Schat, A.; Frone, M.R. (2011). Exposure to psychological aggression at work and job performance:
the mediating role of job attitudes and personal health. Work and stress, Vol. 25, No 1, pp. 23-40.
http://dx.doi.org/10.1080/02678373.2011.563133
Schoorman, F.D.; Mayer, R.C (2008). The value of common perspectives in self-reported appraisals:
you get what you ask for. Organizational research methods, Vol. 11, No 1, pp. 148-59.
http://dx.doi.org/10.1177/1094428107307168
Sicherman, N. (1991). Overeducation in the labour market. Journal of labor economics, Vol. 9, No 2,
pp. 101-122.
Sloane, P. (2014). Overeducation, skill mismatches, and labour outcomes for college graduates. IZA
world of labor; No 88. http://wol.iza.org/articles/overeducation-skill-mismatches-and-labor-market-
outcomes-for-college-graduates-1.pdf
Spitz-Oener, A. (2006). Technical change, job tasks, and rising educational demands: looking outside
the wage structure. Journal of labor economics Vol. 24, No 2, pp. 235-270.
http://dx.doi.org/10.1086/499972
Stewart, M.B. (2007). The interrelated dynamics of unemployment and low-wage employment.
Journal of applied econometrics, Vol. 22, No 3, pp. 511-531. http://dx.doi.org/10.1002/jae.922
Sung, J.; Ashton, D.N. (2015). Skills in business: the role of business strategy, sectoral skills
development and skills policy. London: Sage.
Takeuchi, R.; D.P. Lepak, et al. (2007). An empirical examination of the mechanisms mediating
between high-performance work systems and the performance of Japanese organisations.
Journal of applied psychology, Vol. 92, No 4, pp. 1069-1083.
Tatsiramos, K. (2014). Unemployment benefits and job match quality. IZA world of labor; No 44.
http://wol.iza.org/articles/unemployment-benefits-and-job-match-quality.pdf
Timmer, M. P. et al. (2014). Slicing up global value chains, Journal of economic perspectives, Vol. 28,
No 2, pp. 99-118.
World Economic Forum (2014). Matching skills and labour market needs: building social partnerships
for better skills and better jobs.
http://www3.weforum.org/docs/GAC/2014/WEF_GAC_Employment_MatchingSkillsLabourMarket
_Report_2014.pdf
Details of the data collection

ANNEX 1
The ESJ survey is a Cedefop funded project. vided a representative sample of the 24 to 65
Cedefop commissioned Ipsos MORI, a survey working age population in each of the EU-28
company, to conduct a European-wide employee countries. Most data were collected using Ipsos’
survey on skill mismatch which will help to and partner agencies online panels because of
understand better the extent of skills mismatch time and the cost effectiveness of this method of
and contribute towards development of VET data collection. The mixed mode approach en-
policies across Europe. The global objective of sured that the survey could benefit from the ad-
this research was to deliver a data set which vantages of an online survey in terms of cost and
enables detailed empirical analysis of skill time required for fieldwork, while avoiding the
mismatch among European workers. In problem of lack of coverage among some sec-
response to the global objective, and after tions of the target population by conducting some
consideration of the key concept of skill of the fieldwork by telephone. Non-coverage bias
mismatch, a survey instrument was designed to arises when not all members of the target survey
capture valid and reliable data that would allow population have a chance of being included in the
Cedefop to: sample. Online approaches can suffer from cov-
(a) quantify the incidence and speed of skills ob- erage bias because, in some countries, internet
solescence and skill mismatch experienced penetration is still quite low, particularly among
by employees aged between 24 and 65 older people. A mixed mode approach meant that
years; in countries with low internet penetration, tele-
(b) identify the drivers of skills obsolescence and phone interviews could ensure that the proportion
skill mismatch among European employees; of the population without access to the internet
(c) allow for the examination of the relationship had a greater chance of inclusion.
between the process of skills obsolescence
and skill mismatch and various demographic,
Questionnaire design and

socioeconomic and workplace characteris-
A1.2.
translation
tics, as well as differences among EU coun-
tries;
(d) explore the capacity of continuous vocational
training and of various human resource and A1.2.1. Questionnaire design
organisational practices for mitigating skills The questionnaire used in this study was
obsolescence and skill mismatch. predominately designed by Cedefop and their
expert group (Box 6). The initial questionnaire
had been previously piloted by Cedefop in 2011
A1.1. Methodology overview and was further refined by Ipsos and Cedefop
collaboratively. Ipsos provided additional input
The survey was carried out by Ipsos and its drawing on their experience developing
network partners in the EU-28 Member States questionnaires that limit mode effects in mixed
between 7th March and 26 June 2014. In total, mode surveys prior to the cognitive and pilot
48 676 respondents from different demographic interviews phase.
groups took part either by telephone or online. On The design of the questionnaire reflected
average, the questionnaire length for the online several important considerations:
interviews was 20 minutes, and 31 minutes for (a) a need for the survey to work over the
the telephone interviews. telephone and online;
The main reason for using mixed methodol- (b) a need to restrict the questionnaire to 25
ogy was to ensure that the data collected pro- minutes;
ANNEX 1
Details of the data collection 107
Box 6. Members of the expert group of the After the cognitive testing, Ipsos made
ESJ survey recommendations for changes to the
questionnaire/translations. The master
questionnaire was updated and back-translations
were completed for the five countries not included
• Dr G. Quintini (OECD)
in the cognitive test. The existing translations
• Dr M. Curtarelli (Eurofound)
were updated for the other countries and
• Prof. K. Mavromaras (NILS, Flinders University,
proofread. Written instructions were provided to
S. Australia)
the translators outlining the meaning of key terms
• Dr S. McGuinness (ESRI, IE)
and consistency checks made to ensure high
• Prof. A. de Grip (ROA, NL)
quality translations. The main challenges of the
• Prof. M. Tahlin (SOFI, Stockholm University, SE)
translation process were ensuring consistency of
• Prof. W. Gijselaars (University of Maastricht, NL)
terminology throughout the questionnaire, and
• Prof. F. Mañé (Rovira and Virgili University, ES)
that appropriate terms were used for the key
• Prof. K. Mayhew (SKOPE, University of Oxford, UK)
concepts such as ‘skills’ and ‘workplace’.
• Prof. F. Rycx (Universite Libre de Bruxelles, BE)
Translators were briefed by the research
• Dr D. Verhaest (HUBrussel and Ghent University,
team via teleconference. This helped ensure that
BE)
the translation process was focused on creating
• Prof. T. Zwick (Ludwig-Maximilians University, DE)
equivalent meaning across the languages rather
• Prof. I. Theodossiou (University of Aberdeen,
than on literal translations. After the pilot, further
Scotland)
changes were made to the source questionnaire
and the changes were implemented by linguists
(c) that questions were unambiguous and that at Language Connect, who had worked on the
their meaning was consistently interpreted by earlier translations.
the survey audience; The final translations were further checked by
(d) a need to collect information to measure skill linguists and researchers in Ipsos offices and by
mismatch accurately; Cedefop. Particular emphasis was placed on
(e) a need to ascertain the occupation and ensuring that the translations of the education,
industry of the respondent, so that it could be NACE and ISCO codes were sensitive to local
classified under ISCO and NACE. markets. The translations were then finalised and
There were three stages to the questionnaire signed-off by Cedefop prior to the start of main
design and translation: one before the cognitive stage of fieldwork.
testing; one after the cognitive testing and before The survey was finally carried out in the
the pilot; and one after the pilot before the official national languages of the EU Member
mainstage fieldwork. A summary of these stages States.
can be found in the following.
A1.2.3. Cognitive pilot survey
A1.2.2. Translation A cognitive pilot survey was conducted in 24
Translation ran concurrently with the countries: Belgium, the Czech Republic, Denmark,
questionnaire design in three stages. In the first Germany, Estonia, Ireland, Greece, Spain, France,
stage, the English version of the cognitive testing Croatia, Italy, Cyprus, Lithuania, Hungary, Malta,
questionnaire was signed off by Cedefop and the Netherlands, Austria, Poland, Portugal,
then sent to the local agencies for translation. The Romania, Slovakia, Finland, Sweden and the UK.
local translations were back-translated into In each country, up to 10 cognitive interviews were
English by Language Connect, a translation conducted. Quotas were set on age, level of
company. The English back-translations were education, occupation, and industry to ensure a
reviewed against the master questionnaire by the good mix of respondents. The length of time
central coordination team in London and respondents had spent in their current job was
discrepancies were queried with the local teams. also covered to ensure that all sections of the
The finalised versions of the questionnaire were questionnaire were tested. Fieldwork took place
then sent for sign off by Cedefop. between the 26th of June and the 9th of July 2014.
The interviews were conducted face to face in included within one-digit categories; the use of
most cases, with the interviewer sitting next to the 11-point scales and the labelling of a mid-point to
respondent as the respondent completed the make them easier to use; the adding of
survey online. The interviewer observed the introductory sentences to the questions on skills
respondent, and asked follow-up questions on the to try to soften the language and reduce social
key questions answered by the respondent; the desirability bias (i.e. underreporting of skills that
interviewer paid attention not only to the answers are rated as unimportant and lower than
the respondent gave, but also how they behaved required); and limitation of the cognitive effort
when completing the survey. For example, they required by the respondents to answer questions
watched for looks of confusion or delays in about their jobs and skills at several periods of
responding which might indicate problems with their career.
comprehension or the layout of the script. As a
result of the cognitive test, the questionnaire was A1.2.4. Main pilot survey
adapted for the pilot to take into account In total, 30 pilot interviews were conducted in each
difficulties encountered by the respondents. The country. In countries using a mixed mode
most important of these included difficulty in approach, the interviews were split between
correctly identifying a broad occupation group online and telephone modes. Ipsos collated and
(ISCO), answering questions on a seven-point analysed feedback from the telephone
scale or at repeated period in time and reluctance interviewers and both online and telephone
to rate skills as unimportant or to rate their level respondents. In addition to the survey questions,
of skills as lower than required to do their job. The the pilot also asked the respondents to assess the
test also highlighted country specific issues such difficulty of some of the questions and provided
as problems with translation of words and the opportunity to provide feedback on the survey
phrases, and issues with the formatting and experience. In response to this feedback, the
functionality of the online script. length of the questionnaire was significantly
Following the cognitive test, the questionnaire reduced so that the duration of interviews did not
was adapted for the pilot, with several changes: exceed 25 minutes. The questionnaire was also
inclusion of four-digit examples of the ISCO restructured so that it respects a chonological
occupations to illustrate the types of jobs that are sequence of events: current job, start of current
Table 15. Overview of questionnaire structure
Content of section
Selection criteria Questions on age, gender and employment status to confirm that the respondent
is eligible to take part
Current job Questions including occupation, industry, type of organisation, changes in role
since starting job, contract type
Education attainment Questions on the respondents’ highest level, type and subject of education and
assessment of whether it was needed to get their current job
Skills and job demands Questions on the type, level and importance of skills respondents require to do
their current job
Situation since start of current job Questions about respondents’ development of skills and the demands of their
job since they started their job
Training and learning Questions on type and reasons for training, and attitudes to learning
Situation before current job Questions on employment status before starting current job, reasons for
accepting current job, and level of skills in current job
Background information Demographics
74
Before s
73 100%
(mismat
preferen
80% 42% 43% 44%
ANNEX 1
60%
34% 29% 26%
job, before entry to current job, previous job. The statistical analysis. While quota sampling lacks
40%
overall skills mismatch question was adapted to the statistical robustness of random probability
try to minimise social desirability bias which might sampling, quota sampling 29%can achieve
20%
have impacted on the24% proportion of respondents representativeness by using quotas and weights
27%
reporting levels of skills below those required for which align the sample with the population on key
their job. Questions were also redesigned to Similarcontrol variables. This Differentmethod guarantees
DK FI 0%
account for the different data collection representativeness on these control variables
Exactly the same occupation occupation occupation
methodologies. The main reasons for doing this and makes it likely on other variables that are
was to reduce the cognitive burden placed on correlated with them. For the main LLL survey flexible
53 Before s
respondents, reduce the length of the telephone quotas were set on age, gender, occupation and
49 Fitted values (mismat
survey, and ensure accurate coding of the ISCODK industry. The quotas were based on 2012
35 preferen
codes. Eurostat data from the labour force surveys (44).

To evaluate the extent to which the quota
30 SE
A1.2.5. Structure of final questionnaire sampling used to collect the ESJ data are biased
FI
25
The final questionnaire was split into eight relative to other surveys that employ a random
sections. An overview NL of each section can be probability sampling approach, a few of the
20
found in Table 15. Cedefop ESJ estimates were compared with

FR
estimates from the OECD EE survey on adults skills

UK AT
15 LU
A1.2.6. Sampling approach and Eurofound’s 2010CZ EWCS. These

SI
EL ES ES
The survey employed aCY quota approach BEbecause comparisons are limited to questions that were
10
PT MT
of5 its advantages in terms of time and cost when identical or similar in both surveys. In most cases
IT
DE IE
compared to random probability sampling and the itHRis validated that small differences ROexist between
LV
PL HU LT
EL
fact13that it allowed15for the collection 17 of greater 19the different surveys, also taking into account
SK BG
0
sample sizes, which can aid country-specific Skillreasonable levels of sampling error.
21 23 25
deficit
Figure 65. Cedefop ESJ and OECD PIACC: employment contract

Cedefop PIACC
sons Slovakia 79 16 5 Slovakia 82 12 6
Spain 78 17 5 Spain 74 17 9
0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100%
Base: Cedefop: all respondents in Austria (1 000), Slovakia (1 019) and Spain (4 009);
PIACC: all employees aged 24-65 excluding those with ‘apprenticeship contracts’ and ‘don’t know' / refused in Austria (2 690), Slovakia (2 478)
Cedefop
and Spain (2 485). PIACC
Who paid
Source: for this
Cedefop ESJtraining (e.g.OECD
survey and tuition, registration, fees)?
PIACC. Did an employer or prospective employer pay for tuition
Your employer paid/your employer paid part of the cost or registration, exam fees, expenses for books or other costs
(44) The main differences between quota and random probability sampling are at the stage of drawing the sample. Random probability
resulting from you participation in this activity
sampling uses systematic selection procedures which ensure that each member of the survey population has a known and non-zero
probability of selection, and a lot of effort is made to contact and persuade those selected to take part77through processes such as
repeat contact, reissues and incentives, which help to ensure a good response rate and therefore 69 limit non-response bias. Selection
Austria
probabilities, however, cannot be calculated for quota samples. Quota sampling ensures that the achieved sample structure will match
that of the population for the variables on which quotas are set. The advantage of probability sampling is that is based on statistical
30%
theory and so can demonstrate why samples are free of selection bias and why it is possible
63 to calculate standard errors and confidence
85
Slovakia
intervals. In contrast, with a quota sample it is not possible to use deductive methods to demonstrate a lack of bias or to calculate
standard errors or patterns of variable error. Nonetheless, the quality of a quota sample 66 can be assessed by comparing survey
estimates from a quota sample with those from random probability surveys.
Spain 68
ining at all
0% 20% 40% 60% 80% 100%
%Cedefop %PIAAC
Cedefop PIACC
Skill deficit
sonal reasons Slovakia if any, do you have in your
79
current job? 16 5 Slovakia do you have? Is that... 82 12 6
Cedefop PIACC
Skills,
Spainqualifications
What kindandofjobs in the EU: the
14% Austria 89 making of a perfect match? 9 2 Austria 91 6 3
employment 78 contract, 17 5 Spain What kind of employment
74 contract 17 9
sonal reasons Slovakia 0% 20% 40%79 60% 80% 16 100%
5 Slovakia 0% 20% 40%82 60% 80% 12 100%
6
14% Austria 89 9 2 Austria 91 6 3
Figure 66. Cedefop ESJ and OECD PIACC: employer paid for training
Spain 78 17 5 Spain 74 17 9
rsonal reasons Slovakia 79 16 5 Slovakia 82 12 6
Cedefop 0% 20% 40% 60% 80% 100% PIACC 0% 20% 40% 60% 80% 100%
Spain
Your employer paid/your employer 78 paid part of the cost 17 5 Spain
or registration,
% An indefinite/permanent contract exam% fees, 74 for books
expenses
A fixed-term/temporary 17 % Other
or other costs
contract 9
0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100%
Cedefop PIACC 77
Austria Did an contract
employer or prospective
69 employer paycontract
for tuition % Other
Who paid for this training (e.g. tuition, registration, fees)?
% An indefinite/permanent % A fixed-term/temporary
30% Your employer paid/your employer paid part of the cost or registration, exam fees, expenses for books or other costs
resulting from you participation in this activity 85
Slovakia 63
Cedefop PIACC
Who paid for this training (e.g. tuition, registration, fees)? Did an employer or prospective
66 employer
77 pay for tuition
Austria
Your employer paid/your employer paid part of the cost
Spain or registration, exam fees, 68
expenses
69 for books or other costs
No training at all
30% resulting from you participation in this activity
0% 20% 40% 60% 80% 85 100%
26 Slovakia 63
77 %Cedefop %PIAAC
29 Austria
66 69
Base: Cedefop: all respondents who have undergone training in Austria (728), Slovakia (747) and Spain (2 851);
68
No 30%
27 training at all Spain
Cedefop EWCS
PIACC: all employees aged 24-65 who have undergone training in Austria (1 576), Slovakia (1 043) and Spain (1 390). 85
Slovakia
What
Source:kind of
0%employment
Cedefop contract,
ESJ survey and OECD20%if any,
PIACC. 40% What kind of employment
60% 63 contract do you 80%have? 100%
4 26
Spain 67. Cedefop ESJ and Eurofound EWCS: employment contract

Figure
do you have in your current job? %Cedefop %PIAAC
3029 66
68
No training at all
27 Cedefop
Cedefop 0% 20% 40% EWCS 60% 80% 100%
26 84 11 1 3 1
ESJ survey
What kind of employment contract, if any, What kind of employment contract do you have?
4 %Cedefop %PIAAC
29 do you have in your current job?
33
30
27 EWCS
Cedefop 83 EWCS 11 1 4 1
4
Cedefop
What kind of employment contract, if any, 84 What kind of employment contract do you have? 11 1 3 1
ESJyou
do survey
have in your current job?
0% 20% 40% 60% 80% 100%
30since start of job
ks
33
% An indefinite/ % A fixed/ % A temporary employment % I do not have % Other
EWCS
Cedefop permanent contract temporary contract 83 agency contract a formal contract 1111 1 4 1
84 1 3 1
ESJ survey
33 Cedefop 0% 20% 40% EWCS 60% 80% 100%

ks since start of job Which of the following alternatives would best describe
Overall,
EWCShow would you best describe your skills 83
% An indefinite/ % A fixed/ % Askills
your temporary employment
in your own work? % I do not have 11 1 Other
% 4 1
in relation to what is required to
permanent contract
do your job?
temporary contract agency contract a formal contract
Base: EWCS employees in EU-28 aged 24-65
20% (27 842), Cedefop all 40%
respondents (48 676).
Cedefop 0%5 60% 80% 100%
ks since start of job Source:
Cedefop Eurofound EWCS 2010 and Cedefop ESJ survey.54 EWCS
41
ESJ survey
Figure 68. Cedefop ESJ and Eurofound EWCS: 56 overall skill mismatch
% An indefinite/ % A fixed/ % A temporary
Which of the employment
following alternatives % I do not
would have
best describe % Other
Overall, how would you best describe your skills
permanent contract temporary contract agency contract a formal contract
in relation to what is
12 required to do your job? your skills in your own work? 31
EWCS
Cedefop
Cedefop 0% 5 EWCS
20% 54 40% 60% 41
80% 100%
Overall,
ESJ surveyhow would you best describe your skills Which of the following alternatives would best describe
% Some of
in relation to my skills
what is are lower than
required to dowhat
yourisjob?
required by my job your%skills
I needinfurther
your training to cope with my duties
own work?
and need to be further developed % My present skills correspond well with my duties
12 56 31
% EWCS
My skills are matched to what is required by my job % I have the skills to cope with more demanding duties
Cedefop
% My skills are5 higher than required by my job 54 41
ESJ survey 0% 20% 40% 60% 80% 100%
% Some of my skills are lower than what is required by my job % I need further training%toUnweighted
cope with my
ESJduties
sample31 % Eurostat
Unweighted
EWCS sample 12 profile vs Eurostat data: Occupation 56
% My skills are matched to what is required by my job %7 I have the skills to cope with more demanding duties
ISCO 1 Managers 5
% My skills0%
are higher than required20%
by my job 40% 60% 80% 100%
21
% Some of my skills are lowerISCO than2 Professionals
what is required by my job % I need further training to cope with my duties 20
and need to be further developed % My present skills correspond well16with
% Unweighted ESJ my duties
40%
Unweighted sample profile
ISCO 3 Technicians vs Eurostat
and associate data: Occupation
professionals 17 sample % Eurostat
% My skills are matched to what is required by my job % I have the skills to cope with more demanding duties
% My skills are higher
ISCOthan required
4 Clerical by my
support job
workers 7 11
21
ISCO 1 Managers 5
Base: EWCS employees in EU-28 aged 24-65 (27
ISCO 5 Service and sales workers 762), Cedefop all respondents excluding ‘don’t know’ (48 214).
14
16 21
Source: Eurofound EWCS profile
2010 andISCO 2 Professionals
Unweighted sample vsCedefop
EurostatESJdata:
survey.Occupation
1
% Unweighted ESJ sample 20 % Eurostat
ISCO 6 Skilled agricultural, forestry and fishery workers 1 16
40%
ISCO 1 Managers 77
21
ISCO 8 Plant and machine operators ISCOand
2 Professionals
assemblers 7
14 20
ISCO 5 Service and sales workers 8 16
16
Cedefop PIACC
Your employer paid/your employer paid part of the cost PREVIOUS CONTENTS
or registration,NEXT
exam fees, expenses for books or other costs
ANNEX 1
77
Austria 69
30%
Weighting, statistical
85
A1.3. when the relationship of the various combinations

Slovakia 63
reliability, interpreting the

of those characteristics is not known. The
66
data
weightings60% were capped so that respondents with
Spain 68
ining at all
a weighting larger than four were capped at four

0% 20% 40% 80% 100%
to reduce the impact that the weighting had on

%Cedefop %PIAAC
Weighting is an important aspect of a quota the weighting efficiency, and to reduce the bias
sampling approach. The variables that the data that would result from having a few respondents
are weighted on are selected based on variables
Cedefop EWCS
in the data who are heavily unweighted. Prior to
What kind of employment contract, if any, What kind of employment contract do you have?
that one would expect to be closely correlated capping the weightings, 1.2% of respondents had
do you have in your current job?
with questions on skill mismatch. been assigned a weighting larger than four.
For the Cedefop ESJ survey, the following 84 While the reported results are weighted 1 3by
variables
ESJ survey were used in the weighting scheme:
occupation, industry, education, age and gender,
Cedefop 11 1
(a) age and gender (with age bands 24 to 39, 40 the unweighted data are broadly in line with the
to 54 and 55 to 65); population profile and are outlined in11the charts
(b) education (with three bands being low – for occupation and industry below. The difference
EWCS 83 1 4 1
ISCED 1997 levels 1 and 2; medium – between the official Eurostat data and the sample
ISCED 3 and 4; and high – ISCED 5 and 6); profile occurs because flexible quotas were used
0% 20% 40% 60% 80% 100%
(c) economic activity (NACE % Asection level, during fieldwork,

employmentmeaning that for a couple of
start of job
collapsed into weighting classes: agriculture, occupation and industry quotas, the number of
% An indefinite/ fixed/ % A temporary % I do not have % Other
permanent contract temporary contract agency contract a formal contract
industry, services, public administration and interviews achieved are higher or lower than the
other);
Cedefop
Eurostat data.
EWCS
(d) occupation (ISCO 1 digit, collapsed into However, forwork?
education, the sample profile is
Overall, how would you best describe your skills Which of the following alternatives would best describe
weighting classes: managers and different from the Eurostat data in some
in relation to what is required to do your job? your skills in your own
professionals; technicians and associate countries, with people with higher levels of
professionals; skilled/semi-skilled; and other). education being underrepresented in some
Cedefop 5 54 41
The weightings were calculated using the countries and over-represented in others. To
ESJ survey
LFSEWCS population 12 counts of employees aged 24 to correct for this, we weighted 31the data by
65 based on an up-to-date Eurostat data set education, though it was necessary to cap the
56
released 0% in 2013. Rim weighting was used: this large weights; this means80%that in Italy,100% in
uses a of
form of are
regression
lower than whatanalysis
is required bytomyweight particular, thetraining
weighted datamy still
dutiesunderrepresent
20% 40% 60%
according to various demographic characteristics people

% I have with primary education.
% Some my skills job % I need further to cope with
% My skills are matched to what is required by my job the skills to cope with more demanding duties
Figure 69. Sample profile: occupation

% My skills are higher than required by my job
Unweighted sample profile vs Eurostat data: Occupation % Unweighted ESJ sample % Eurostat
ISCO 1 Managers 7
5
21
ISCO 2 Professionals 20
16
40%
21
14
ISCO 5 Service and sales workers 16
ISCO 6 Skilled agricultural, forestry and fishery workers 1

1
7
ISCO 7 Craft and trade workers 11
7
ISCO 8 Plant and machine operators and assemblers 8
5
Unclassified occupation 1
1
0% 5% 10% 15% 20% 25%

Base: All respondents (48 676).
Source: Cedeop ESJ survey and Eurostat.
Unweighted sample profile vs Eurostat data: Industry % Unweighted ESJ sample % Eurostat
12
Administration and support services 13
2
Agriculture, horticulture, forestry or fishing 2
2
5 16
ISCO 3 Technicians
ISCO 5and associate
and sales workers 14
40% 16 17 21
ISCO 2 Professionals 20 21
ISCO 6 Skilled agricultural, ISCOforestry
4 Clerical
andsupport workers
fishery workers 1 11
1 16
ISCO 3 Technicians and associate professionals
40% ISCO 5 Service PREVIOUS CONTENTS NEXT 14 17
ISCO 7 Craft and
and sales
trade workers
workers 7
11 16
21
ISCO 4 Clerical support workers
Skills, 8qualifications and jobs in the and
EU:and
the making of a perfect match?
1 11
ISCO 6 Skilled
Plantagricultural, forestry fishery workers
112
7
Evidence from Cedefop’s ISCOEuropean
5 Serviceskills
and and
salesjobs survey
ISCO and machine operators assemblers 1 8 14
workers 7 16
ISCO 7 Craft and trade
ISCO 9 Elementary occupations workers 5
10 11
ISCO 6 Skilled agricultural, forestry and fishery workers 1
1 7
ISCO 8 Plant and machine operators and assemblers
Unclassified occupation 1 8
1 7
ISCO 7 Craft and trade workers 5 11
Figure 70. Sample profile: Unclassified industry

10
ISCO 8 Plant and machine operators and assemblers 0% 5% 7
8
10% 15% 20% 25%
occupation 1
1
5
Unweighted sample profile vs Eurostat data: Industry 0% 1 5% 10% % Unweighted
15% ESJ sample
20% % Eurostat
25%
Unclassified occupation 1
12
Administration and support services 13
0%
Unweighted sample profile vs Eurostat data: Industry 5% 10% 15%
% Unweighted ESJ sample20% 25%
% Eurostat
2
Agriculture, horticulture, forestry or fishing 2
2 12
80% 90% UnweightedAdministration
sample profileand vs
support services
Eurostat 1
data: Industry 13 Unweighted ESJ sample
% % Eurostat
1
Supply, management
Agriculture, or treatment
horticulture, forestry oroffishing
water 1 2
2
12 14
Manufacturing
Administration
Supply of gas and or engineering
support services
or electricity, mining 2 13 17
1
80% 90% 2 6
Agriculture,
Supply, management Construction
horticulture, forestryoror
or treatment building
of fishing
water 1
2 6
1
2 11
Retail, sales,
Supply shoporwork
of gas
Manufacturing ororengineering
wholesale
electricity, mining 1 13 14 17
80% 90%
1 3
Accommodation,
Supply, managementcatering or food
or treatment
Construction or services
of water
building 1 4 6
6
6 11 14
Transportation
Manufacturing or or storage
engineering
Retail, sales, shop work or wholesale 6 13 17
ICTorservices
services 3 66
Construction
Accommodation, catering or food building 3 4 6
66 11
Financial, insurance or real estate
Retail, sales,Transportation
shop work oror services
wholesale
storage 4 13
6
3 7
Professional, scientific
Accommodation, or technical
catering ICT services
or food services
services 44 6
3
16
Services
Financial, relating
insurance to real
education
estateoror
Transportation
or health
storage
services 666 21
4
CulturalProfessional,
industries (arts, entertainment, recreation)
ICT services 2 6 7
scientific or technical services 1 3 4
Financial, Social and
insurance personal
or education
real services
estateor
services 66 16
Services relating to health 2 4 21
7
CulturalProfessional, scientific
industries (arts, or technical
entertainment, services
recreation) 0% 1
2 4 5% 10% 15% 20% 25%
Services (48
relating to education 16
Base: All respondents Social and
676). personal or health
services 2
6 21
Source:
CulturalCedefop ESJ(arts,
industries survey and Eurostat.recreation)
entertainment, 1
2
0% 5% 10% 15% 20% 25%
Figure 71. Sample profile: gender and age

Unweighted sampleSocial
profile
and vs Eurostat
personal data: Gender and age
services 6 % Unweighted ESJ sample % Eurostat
2
0% 5% 10% 15% 20% 56 25%

Male 54
Unweighted sample profile vs Eurostat data: Gender and age
GENDER % Unweighted ESJ sample % Eurostat
44
Female 46
56
Male
Unweighted34sample 43 ESJ sample
% Unweighted 54
% Eurostat
GENDER to 39 profile vs Eurostat data: Gender and age 40
44
Female 44 46
10 AGE 40 to 54 44 56
Male 54
GENDER 43
34 to 39 13 40
55Female
to 65 16
44
46
AGE 44
10 40 to 54 44
43
34 to 39 0% 10% 20% 30% 40%
40
50% 60%
13
55 to 65 16
AGE 44
10 40 to 54 44
0% 10% 13 20% 30% 40% 50% 60%
Unweighted 55 to 65 profile vs Eurostat data: Education
sample 16 % Unweighted ESJ sample % Eurostat
Source: Cedefop ESJ survey 0%and Eurostat. 10% 20% 30% 40% 50% 60%
Figure 72. Sample profile: education

13
Less than primary,
Unweighted sampleprimary and vs
profile lower secondary
Eurostat data: Education % Unweighted ESJ sample % Eurostat
17
Unweighted sample 13 % Unweighted ESJ42sample % Eurostat

Less than
Upper primary,
secondary postprofile
andprimary vs Eurostat
and lower
secondary secondarydata: Education
non-tertiary 17 49
13
Less than primary, primary and lower secondary 42 46
Upper secondary and post secondary non-tertiary
Tertiary 17
35 49
42
Upper secondary and post secondary non-tertiary 0% 10% 20% 30% 40% 46 50% 60%
Tertiary 35
49
46
Tertiary 0% 10% 20% 30% 40% 50% 60%
35
0% 10% 20% 30% 40% 50% 60%

Source: Cedefop ESJ survey and Eurostat.
25%
25%
ANNEX 1
A1.4. Sample sizes: EU and by

country
Table 16 specifies the final unweighted and
weighted national sample sizes for the EU-28
area.
Table 16. Unweighted and weighted national

sample sizes in ESJ survey
Unweighted EU-28 weighted
sample size sample size
European Union total 48 676 48 676
Austria 1 000 927
Belgium 1 502 1 065
Bulgaria 1 000 719
Croatia 1 004 315
Cyprus 500 85
Czech Republic 1 506 1 123
Denmark 1 000 610
Estonia 1 001 149
Finland 2 004 556
France 4 011 6 217
Germany 4 013 9 428
Greece 2 037 652
Hungary 1 500 983
Ireland 1 004 418
Italy 3 016 4 797
Latvia 1 004 211
Lithuania 1 010 308
Luxembourg 500 61
Malta 500 38
Netherlands 1 502 1 722
Poland 4 017 3 341
Portugal 1 503 987
Romania 1 502 1 785
Slovakia 1 019 549
Slovenia 1 010 211
Spain 4 009 3 902
Sweden 1 001 1 090
United Kingdom 4 001 6 423

Empirical methodologies used to derive

ANNEX 2
reported findings
A2.1. Discrete choice empirical A2.2. Derivation of a new

estimation of qualification categorisation of skill
and skill mismatch mismatch among EU
workers
Probit regressions have been estimated to
identify the determinants of the probability that an To derive a new categorisation of the skill
individual employee is affected by a particular mismatch status of EU adult workers that relies
type of vertical mismatch (a binary outcome, e.g. on talent management and adopts a lifelong
the individual is either mismatched or not). The learning perspective, two separate questions
probability of an individual being either over- or from the ESJ survey were combined. The first
underqualified is modelled on the basis of a latent question measures the skill mismatch status of
variable model, where it is assumed that an workers: the correspondence of their skills to
individual’s unobserved propensity to be in a what is needed to do their job at the current time.
particular state of qualification mismatch is The second question measures the lag between
determined by the equation: the skills of respondents and the level of skills
needed so they can perform with maximum
mit* = β0 + xitβ + Tt + Ci + uit, m=1[m* > 0] (1) proficiency in their job, at the best level possible.
Prior to combining the two questions,
where m, individual i’s mismatch status in time respondents in the survey were categorised
period t, takes the value one if m* > 0 and zero if according to whether or not they suffer from a larger
m* ≤ 0. X is a vector of explanatory variables, than average skill deficit. An average skill deficit has
including demographic, socioeconomic and job- been estimated on the basis of an OLS regression,
related characteristics of the respondents, while which computes the conditional mean skill deficit of
u, the disturbance term, is assumed to follow a workers in different countries on the basis of their
normal distribution and to be independent of x. age, gender, level of education, native status, prior
The estimated parameters, β, denote the effect labour market status, years of employer tenure,
of each explanatory variable on the probability of characteristics of the job (part-time work, type of
a positive response i.e. P(y = 1|x). In all contract, sector, size of workplace, hours of work,
regressions, country-specific conditions (e.g. multisite organisation) and industry and occupation.
differences in labour market institutions) are A comparison of each respondent’s skill deficit with
taken into account via the inclusion in the the average implied by his/her characteristics is
empirical specification of country (C) dummies. subsequently used to categorise the individuals in
The reported estimates are interpreted in terms the sample according to whether they have higher
of the change in the probability of an individual or lower than average skill deficits.
being vertically mismatched in relation to a Following this, individuals have been
particular individual or job characteristic (e.g. categorised according to whether they are ‘talent
gender). unnourished’, ‘talent in dead end’ or ‘talent in
standstill’ as follows:
(a) talent unnourished: individuals have greater
than average skill deficit, regardless of
whether their skills are above, below or
matched to what is needed to do the current
job tasks;
ANNEX 2
Empirical methodologies used to derive reported findings 115
(b) talent in dead end: individuals have an equal treated); and individuals who moved into their
or smaller than average skill deficit and have current job directly from another, but who are
stated that they have higher skills than characterised by a similar probability of being
needed to perform their current job tasks; unemployed on the basis of their observed
(c) talent in standstill: individuals have an equal characteristics as the former group (the untreated
or smaller than average skill deficit and have or counterfactual group).
stated that they have matched skills to what The estimated average treatment effect on
is needed to perform their current job tasks. the treated reveals that there is a statistically
significant difference in the improvement of skills
scores between the two comparable groups of
A2.3. Matching estimators of individuals who differ on the basis of their prior
effect of prior labour market status.
unemployment status on
skill formation
A2.4. Derivation and
The empirical analysis undertaken to investigate determinants of skill
the impact of a prior spell of unemployment on intensity index
the subsequent skill formation of individuals has
followed a so-called matching approach. The The indicator used to measure the skill content of
methodology first estimated the probability that jobs is derived from self-reported accounts of the
individuals were unemployed before entry into perceived importance of a range of cognitive,
their current job as a function of a number of digital and transversal skills. The skill content
explanatory variables that preceded or are index is derived from the orthogonal latent
exogenous to their actual status of joblessness, dimensions of the set of skills (the factors on
such as their gender, their native status, highest which a wider set of skills can be reduced, Mañé,
level of education attainment, age at the time they 2013).
began their new job and skill mismatch status in The skills surveyed in the questionnaire are:
the job they had prior to becoming unemployed. literacy, numeracy, ICT, technical skills,
In addition, in the spirit of Oswald’s theory communication skills, teamwork skills, foreign
underpinning the wage curve (Oswald, 1991), the language skills, customer handling skills, problem
likelihood of experiencing a previous spell of solving, learning skills, and planning and
unemployment was linked to whether individuals organisation skills (46).
owned their main residence (either outright or These 11 skills have been additively
with a mortgage) before finding their new job and collapsed into an index of skill requirements or
to if they (or their family) were faced with family skill content of jobs. However, the indicator is
obligations prior to accepting the new job (45). imperfect in that it is based on a small set of skills
At the second stage, and on the basis of the (even though skills spanning different skills
estimated probability of individuals being domains were used). It is built on subjective
unemployed before entering into their current job statements of the importance of these skills (scale
conditional on specific characteristics, the 0 to 10, where zero is ‘not important at all’ and 10
programme compared improvement in the skills is 'essential'). The index performs well (Cronbach
between two comparable groups of individuals: alpha 0.84) across countries (the country with the
those who actually experienced an lowest reliability indicator is Croatia, alpha 0.75)
unemployment spell prior to job entry (the and occupations (the index’s performance is
(45) In a secondary analysis, additional explanatory variables that are not strictly predetermined have been also added in the first step
regression of the matching estimation, such as whether the respondents live with a partner/spouse, with children or with their parents,
if they moved home or changed occupation before starting their job with their current employer and if they (and their family) were
faced with considerable financial constraints at that time.
(46) The robustness of the results has also been verified by excluding foreign language and costumer communication skills from the derived
index. The reason for doing so is that the Anglo Saxon countries naturally score much lower that other countries on the importance of
foreign language skills, while these skills are also of a lesser generic nature and more specific to particular occupations relative to the
remaining set of skills used in the index.
lower in the first three ISCO categories 0.76, A2.5. Multinomial logit
0.74, and 0.74; respectively).
regressions of skill
mismatch transitions
The empirical analysis is based on the
following regression:
Analysis of dynamic changes in skill mismatch

Model 1: status has taken place by either estimating a
Si = β0 + β1JCi + β2ΔJCi + β3ΔWPi + β4FCi + probit regression for each possible state of a skill
+ β5Xi + εi mismatch transition (e.g. matched to overskilled;
underskilled to matched) or by undertaking
where S denotes a job’s skill content, JC denotes multinomial logit regressions. In the spirit of a
a vector of job characteristics (job complexity, matching-type analysis, the reference value has
team work), ΔJC denotes the change in job been defined in each instance to correspond to
complexity, WP denotes workplace dynamism, FC the respective initial state: if interested in the
denotes a vector of firm characteristics (type of changes in individuals who were matched at the
organisation, private or public, establishment start of their current job and became underskilled
size, number of employees, full-time, type of over time, the reference category would be
contract, tenure, education requirements to be [matched-matched, matched-overskilled] and not
hired in the job and the education requirement the whole remaining sample (which also includes
needed to do the job) and X is a vector of individuals who were overskilled or underskilled
additional controls (country, industry and at start of their job). The reason for doing this is
occupation dummies.). To investigate the impact to ensure that the comparison/reference group
of personal characteristics the above model has has similar characteristics at the time of entry to
been augmented with a set of variables (denoted the job and to minimise the potential
by PC) including: age (in log), highest education heterogeneity that is likely to exist in the
qualification, an indicator variable capturing circumstances of matched, overskilled and
whether the current job is the first job, the previous underskilled people (Figure 73).
labour market position (employed, self-employed, For robustness, the analysis was also
unemployed, long-term unemployed, out of the replicated by including only those cases where
labour force, in education), current mismatch the reference category remains unchanged over
status, initial mismatch status, and mismatch time (matched-matched, overskilled-overskilled,
status in the previous job, an index of the degree underskilled-underskilled).
of occupational change (if individuals held a job in
the same occupation, in a similar occupation, or
in a completely different occupation), if the new
job involved relocation, the reasons that made
them accept the job, family composition, whether
they were born in the country, personal constraints
(family obligations) and job search effort.
The final model reads:
Model 2:
Si = β0 + β1JCi + β2ΔJCi + β3ΔWPi + β4FCi +
+ β5PCi + β6 Xi + εi
Regressions are carried out using appropriate

weights and the standard errors reported are
clustered by country.
The definition of qualifications 117
Figure 73. Schematic illustration of estimation methodology of determinants of skill mismatch

transitions in ESJ survey
rished
Before start of current job
(mismatch in previous job,
preferences, constraints, mobility) Matched at start of job Underskilled today
Change in skills, change in job tasks b coefficient
Before start of current job Matched at start of job Matched (or overskilled)
values (mismatch in previous job, today
preferences, constraints, mobility)
25
6 3
2 6
100%
Other
Access to the Cedefop ESJ survey
The microdata of Cedefop’s ESJ survey will be

made available to the general public and
researchers in spring 2016.
Additional results and key indicators from the ESJ

survey will be available from Cedefop’s website
(http://www.cedefop.europa.eu) and the European
skills panorama web portal
(http://skillspanorama.cedefop.europa.eu).
For further information about the survey, including

background documentation, interested readers
may contact the responsible project managers of
the ESJ survey:
Mr Konstantinos Pouliakas:
Konstantinos.Pouliakas@cedefop.europa.eu
Mr Giovanni Russo:
Giovanni.Russo@cedefop.europa.eu
EN EN ISSN: 2363-216X
EVIDENCE FROM CEDEFOP’S EUROPEAN SKILLS AND JOBS SURVEY

and jobs in the EU:
the making
of a perfect match?
What has been the impact of the economic crisis
3072 EN – TI-RF-15-003-EN-N – doi:10.2801/606129

on skill mismatch? Is there a cost in getting the
unemployed quickly into any job? Why is skill
mismatch prevalent among the EU workforce?
To answer these and other timely questions on
skill mismatch, in spring 2014 Cedefop carried
out the European skills and jobs (ESJ) survey.
The findings caution that the prolonged
economic downturn is threatening the long-term
potential of the EU human resources. A greater
share of recent job finders has entered into jobs
that need lower qualifications and skills than
their own. The unemployed also run a greater
risk of misplacement into jobs of lower skill
intensity. More than one in five EU employees
has not developed skills since they started a job,
as over one third of EU jobs are characterised
by poor task complexity and lack of continued
learning. Closer stakeholder collaboration and
policy action is needed in the EU to generate not
only more skills but also, crucially, better jobs for
better-matched skills.
and jobs in the EU:
European Centre for the Development
of Vocational Training
the making
Tel. +30 2310490111, Fax +30 2310490020, E-mail: info@cedefop.europa.eu
of a perfect match?
ISBN 978-92-896-1945-5

3072 en

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

3072 en

Uploaded by

Copyright:

Available Formats

EN EN ISSN: 2363-216X

EVIDENCE FROM CEDEFOP’S EUROPEAN SKILLS AND JOBS SURVEY

3072 EN – TI-RF-15-003-EN-N – doi:10.2801/606129

Europe 123, 570 01 Thessaloniki (Pylea), GREECE

Cedefop Reference series 103

Please cite this publication as:

A great deal of additional information on the European Union

Copyright © European Centre for the Development

The European Centre for the Development

Europe 123, 570 01 Thessaloniki (Pylea), GREECE

Joachim James Calleja, Director

Joachim James Calleja

This report is the outcome of team effort and of own in-

List of boxes, figures and tables

survey: identity and rationale

base to inform policies on initial (formal and work-

based learning (WBL)) and continuing vocational

education and training (CVET) (non-formal and

informal training) and on workplace design and

labour market mobility for mitigating skill

Figure 1. Key facts from the ESJ survey

Source: Cedefop ESJ survey.

Figure 2. Incidence of WBL by field of study, EU-28, 2014

Humanities, languages and arts

Source: Cedefop ESJ survey.

Active labour market policies programmes end when an unemployed person

tackling skill mismatches due to

Skill mismatch at work Skill gaps at work vary across generations,

Figure 4. Skill deficits and labour productivity, 2014, EU-28

PREVIOUS CONTENTS NEXT

skills between job stability and people working in jobs

basic ICT or no ICT skills at all. Over a third of deployPL workers’

HU skills as technology and working

skill 0needs. benefits of skill investments.

% no change or reduction in variety of job tasks since start of job

Retail, sales, shop work or wholesale

Accommodation, catering or food services

undermine economic competitiveness and Some of these questions, which have

Why a Cedefop survey on

skills may still enjoy a high level of job

satisfaction. The empirical snapshot pointing to

High unemployment is public enemy number

The European skills and

Figure 6. Pillars of the ESJ survey

1.4. Terminology used between aggregate labour

or qualitative (differences between individual’s

qualifications and those required by the labour

Table 1. Definitions of skill mismatch

Mismatch A situation where industries, occupations, locations or groups with different

Box 3. Methodological note

Qualification and skill mismatches

demand in the EU job

40% ISCO 96:

professional or scientific services, education and qualifications needed to do their job.

Low or no qualifications Medium High

labour markets are driven by inflated recruitment 2.2. Qualification mismatches

Figure 8. Average incidence of qualification mismatch, adult employees, 2014, EU-28

Figure 9. Breakdown of overqualification by level of education, adult employees, 2014, EU-28

Not working (e.g. child care, family) 6.8

a relatively large proportion ofjob

PREVIOUS CONTENTS NEXT Overqualified Underqualified

to have high overqualification rates. By contrast,

identified in Figure 9, most overqualified temporary agency contracts).