DOI: 10.1111/joes.

12424

ARTICLE

Skill shortages and skill mismatch: A review of

the literature

Giorgio Brunello1 Patricia Wruuck2,*

1Department of Economics and

Management, University of Padova, Abstract
Padova 35123, Italy We review the recent economic literature on skill short-
2European Investment Bank, ages and skill mismatch by paying particular attention
Luxembourg, Luxembourg
to the demand side of skill mismatch, which have not
Correspondence received much attention in the relevant literature. We
Giorgio Brunello, Department of Eco- discuss measurement issues and how skill shortages
nomics and Management, University of
Padova, via del Santo 33, Padova 35123, and mismatch vary both with the business cycle and
Italy. with structural factors. We review the economic costs
Email: [email protected]
for workers and firms and conclude by considering the
* The opinions expressed are the authors’ policy implications, including how responsibilities for
only. The paper should not be reported as skill development can best be shared and what role poli-
representing the views of the European
cies can play to better address skill shortages and mis-
Investment Bank.
matches.

1 INTRODUCTION

Labour markets are currently in a phase of cyclical downturn triggered by the COVID-19 pandemic
and undergoing structural transformation due to globalization, changes in labour market institu-
tions, demographic trends and advancing digital technologies including automation. Against this
background, businesses often report that the limited availability of skills poses an impediment to
corporate investment.
Genuine skill constraints can negatively affect labour productivity and hamper the ability to
innovate and adopt technological developments. For individuals, not having ‘the right skills’ limits
employability prospects and access to quality jobs. For firms, skill shortages mean longer time with
unfilled positions or the recruitment of workers with inadequate skills. For national economies,
persistent skill gaps and mismatches come at economic and social costs.
This paper reviews the recent economic literature on skill shortages and skill mismatch. We pay
particular attention to skill shortages and the demand side of skill mismatch compared to other
reviews (McGuiness et al., 2018; Somers et al., 2019). This area has not been given much attention
in papers on mismatch, which tend to focus on workers rather than on firms. We also discuss
how skill shortages and mismatch vary both with the business cycle and with structural factors

J Econ Surv. 2021;35:1145–1167. wileyonlinelibrary.com/journal/joes © 2021 John Wiley & Sons Ltd. 1145
 14676419, 2021, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joes.12424 by Duale Hochschule BW Mannheim, Wiley Online Library on [10/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1146 BRUNELLO and WRUUCK

(demographic and technological changes, globalization). Whether mismatch is pro-cyclical or

counter-cyclical matters for the labour market consequences of the economic downturn induced
by the pandemic that most countries are experiencing. Although our review considers both Euro-
pean and non-European research, the data we use are mainly for Europe.
We start with a conceptual overview of skill shortages and mismatch and how to measure them.
An issue discussed in the first section is the measurement of job requirements, a demand side vari-
able, that some authors compute using surveys of individuals, which typically collect information
on the supply side (educational attainment, foundation skills). Individuals, however, often do
not have a reliable view of job requirements and may actually have an incentive to inflate them.
Another issue is whether skill shortages stated by employers reflect the lack of suitable candidates
among job seekers or are due instead to the wage and working conditions being offered.
The second section looks at how skill shortages and mismatch are affected by cyclical and struc-
tural factors. Whether mismatch is pro- or counter-cyclical depends on the relative strength of
different factors. In a recession, poor matches are destroyed, skilled workers are willing to accept
unskilled jobs as jobs are scarce and matching efficiency declines, as workers losing their jobs may
not have the skills required to fill the existing vacancies.
Structural factors contributing to skill shortages and mismatch include ‘megatrends’, notably
globalization, digitalization and ageing. In addition, institutional factors shaping labour markets,
skill utilization and formation at the national level can work to reinforce or mitigate skill shortages
and mismatches and are an important factor mediating the impact of structural trends.
The third section discusses the economic costs of skill mismatch and shortages. Both not only
affect individuals and firms but can also reduce average productivity by leading to an inefficient
allocation of resources across firms. The final section considers the policy implications, including
how responsibilities for skill development can best be shared and what role policies can play to
better address skill shortages and mismatches.

2 MEASURING SKILL MISMATCH AND SHORTAGES

2.1 Skill mismatch

Skill is a complex concept and challenging to measure.1 According to a definition provided by

the OECD (2017), skills refer to both cognitive and non-cognitive abilities and to abilities that are
specific to a particular job, occupation or sector (technical skills). Cognitive skills consist in the
ability to understand complex ideas, adapt effectively to the environment, learn from experience,
engage in various forms of reasoning, overcome obstacles by taking thought. They include liter-
acy, numeracy and the ability to solve abstract problems. Non-cognitive skills are characteristics
across multiple domains (social, emotional, behavioural) not included under cognitive skills, such
as work habits, behavioural traits, and physical characteristics. Finally, technical skills are com-
binations of cognitive and non-cognitive skills used to accomplish specific tasks (Margolis, 2014).
These combinations could also include manual skills. While skills are multi-dimensional, their
measurement often focuses on selected dimensions, mainly due to data limitations.
At the macro level, skill mismatch refers to the gap between the (aggregate) supply and demand
for skills, typically with reference to a specific geographical unit (region, country or country
group), and to the fact that observed matches between available workers and available jobs offered
by firms in terms of skills and/or qualifications are sub-optimal. At the micro level, skill mismatch
occurs when workers have a level of skills that is different from what is required for their job.2
 14676419, 2021, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joes.12424 by Duale Hochschule BW Mannheim, Wiley Online Library on [10/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BRUNELLO and WRUUCK 1147

Skill shortages arise when employers are unable to recruit staff with the required skills in the
accessible labour market and at the ongoing rate of pay (Quintini, 2011) and skill surpluses occur
when the supply of certain skills is higher than demand. Finally, the term skill gaps is often used
when the skill levels of the existing workforce are insufficient to meet the requirements of firms
(McGuiness et al., 2017).
Skill mismatch at the macro level can be assessed by comparing the composition of vacancies
by qualification or education with that of the working age population (as a proxy of labour sup-
ply). Instead of vacancies, one can compare the composition of employment as a proxy of labour
demand with that of the population at working age, or the composition of unemployment with
that of the labour force.
Macroeconomic measures of skill mismatch focus on whether or not individuals with a certain
type of qualification or skill are (un)employed but do not assess whether – if they are employed –
their job is in line with qualifications (see McGuiness et al., 2017 for discussion). Skill mismatch
(micro), instead, is typically measured by comparing the skills or qualifications of an employed
worker with the skills or qualifications required by her or his job. If the worker has skills compat-
ible with what the job requires, the pair is a good match. A worker can also be classified as over
(under) skilled with reference to a specific position.
Skill requirements and the quality of matches are identified either (a) by asking employees
whether they have the skills required to do a more demanding job that their current one or
whether training is needed to carry out the job in a satisfactory way (self-reported mismatch) or (b)
by the individual measurement of cognitive skills (literacy, numeracy and problem solving) and
the comparison of attained values with average or median values in the occupation, which are
used as proxies of job requirements (realized-matches mismatch). Both approaches are reviewed
in detail by Mc Guiness et al. (2017), and discussed by Allen and van der Velden (2001), Green and
McIntosh (2007); Hartog (2000); Sala (2011); Quintini (2011).
The importance of how job requirements are defined for measured skill mismatch is shown
by Pellizzari and Fichen (2017), who use the data from the Survey of Adult Skills (PIAAC) and
compare the approach based on self-reported mismatch and the realized matches approach.3 They
find that the large majority of workers are considered to be over-skilled with the former method,
but well matched with the latter method (Figure 1).
Recently, these authors have also developed a measure of skill mismatch, which combines both
methods and consists of two steps. First, the proficiency scores of workers who report to be well
matched to their job are used to create a quantitative scale of the skills required to perform the job
for each (ISCO 1-digit) occupation. Second, minimum and maximum threshold values in this scale
are identified, and workers with scores above the maximum or below the minimum are classified
as over and under-skilled. This approach identifies the level of skills in an occupation with the
literacy test score attained by each adult worker. This seems restrictive but perhaps unavoidable
given the data at hand. A potential problem is that ISCO 1-digit occupations are too broad and can
span a range of tasks, which may require different skills (see McGowan & Andrews, 2015a).

2.2 Skill shortages

Since it is possible that employers and managers have more accurate information than employees
do about skill requirements, employer surveys can provide valuable information on skill shortages
from a demand side perspective.4 Examples of employer surveys include the U.K. Employer Skills
Survey, the European Skills and Jobs Survey, the European Union (EU) Business and Consumer
 14676419, 2021, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joes.12424 by Duale Hochschule BW Mannheim, Wiley Online Library on [10/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1148 BRUNELLO and WRUUCK

F I G U R E 1 Comparing self-reported and realized-matches mismatch

Source: Pellizzari & Fichen, 2017
[Colour figure can be viewed at wileyonlinelibrary.com]

Surveys, the Manpower Talent Shortage Survey (which covers more than 40 countries globally)
and the European Continuing Vocational Training Survey, which focuses however on training
issues.
The EIB Investment Survey (EIBIS), for instance, provides information on skills and
investment-related aspects, by regularly assessing corporate investment for training and whether
the availability of staff with the right skills poses an impediment to investment for firms in the
EU.5 Data from the EIBIS show that, since 2016, the limited availability of skills has increasingly
become a concern for firms, topping the list of the most frequently named impediment in a list
of nine obstacles to investment in recent years (Figure 2). On average, 77% of firms reported the
limited availability of skills as an impediment to investment. Moreover, the most recent indicate
that skill shortages remain relevant to firms against the background of the pandemic (EIB, 2020).
When interpreting the results of employer surveys, it is important to connect the responses of
firms to company characteristics and complement these responses with an understanding of the
operating environment, that is, including the cyclical position as well as labour market character-
istics that affect firms’ difficulties to hire (see Savšek, 2018). At the firm level, reported shortages
and problems to fill vacancies could be due to the wage and working conditions being offered,
workforce characteristics or the effectiveness of the recruitment process, rather than to the lack of
suitable candidates among job seekers. Establishments offering a higher average wage relative to
the average wage for the same occupational group in a given area have been found to report fewer
shortages (Directorate General for Internal Policies, European Parliament, 2015; Haskel & Martin,
2001). In addition, Monti and Pellizzari (2016), have shown that the occupations where hiring was
declared to be the most difficult in Italy in 2012 were not the ones where wages increased the most
during the period 2012–2015.
 14676419, 2021, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joes.12424 by Duale Hochschule BW Mannheim, Wiley Online Library on [10/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BRUNELLO and WRUUCK 1149

F I G U R E 2 Share of firms reporting different obstacles to investment, in % (2016–2018).

Notes: The question asked in the questionnaire is: thinking about your investment activities in your country, to
what extent is each of the following an obstacle? The graph adds the share of firms naming the impediment as a
major or minor obstacle. Sample: all firms.
Source: EIBIS 2016–2018
[Colour figure can be viewed at wileyonlinelibrary.com]

In a series of studies, CEDEFOP (2015, 2018), has attempted to separate ‘genuine’ shortages,
defined as recruitment bottlenecks that occur when firms offer competitive starting salaries to
potential recruits, from reported shortages that originate from wage offers below competitive lev-
els. Drawing on the Eurobarometer Flash Survey no. 304, these studies show that while 47% of
interviewed firms report difficulties in recruiting graduates with suitable skills, the total propor-
tion of employers facing genuine shortages is much lower at 34% (see McGuiness et al., 2017).
To disentangle genuine from other shortages, results from employer surveys should be com-
plemented with indirect measures that signal shortages in specific occupations, including price
measures (wage growth), volume measures (employment growth, vacancy rates) and work inten-
sity measures (incidence of overtime). The rationale is that occupations with genuine shortages
should be characterized by faster wage growth, or by higher overtime, than occupations without
genuine shortages.
These indirect measures have been recently used by the Organisation for Economic Co-
operation and Development (OECD) to compute an index of skill shortages in two steps: first, an
occupational indicator is constructed by combining information on hourly wage growth, employ-
ment growth and growth in hours worked by occupation. The relevant data are drawn from large
comparable international surveys, including the Current Population Survey for the United States
and the European Labour Force Survey and the European SILC Survey for Europe. Second, this
indicator is translated into a skill index using the O*Net database, which maps occupations into
bundles of tasks and skills.
Based on this methodology, recent skill shortages are concentrated among content skills (e.g.,
reading comprehension, writing, speaking and active listening), process skills (e.g., critical think-
ing and active learning), complex problem-solving skills and social skills (e.g. instructing, social
 14676419, 2021, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joes.12424 by Duale Hochschule BW Mannheim, Wiley Online Library on [10/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1150 BRUNELLO and WRUUCK

F I G U R E 3 Skill needs across Europe

Source: OECD, 2017
[Colour figure can be viewed at wileyonlinelibrary.com]

perceptiveness) (Figure 3). By country, these shortages are biggest in Finland, Luxembourg, the
Netherlands, Spain and Germany and smallest in Switzerland and Hungary. Surpluses are more
common for some technical skills, including maintenance and repairing (OECD, 2017).

3 CYCLICAL AND STRUCTURAL FACTORS AFFECT SKILL

SHORTAGES AND SKILL MISMATCH

In the medium to long run, the demand for skills is driven by technological and organizational
innovations, demographic changes and changes in the patterns of consumption.6 New sectors
 14676419, 2021, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joes.12424 by Duale Hochschule BW Mannheim, Wiley Online Library on [10/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BRUNELLO and WRUUCK 1151

F I G U R E 4 Recent trend in global talent shortages

Source: Manpower Talent Shortage Survey, 2018. The survey does not cover Russia, Africa (with exception of
South Africa), the Middle East, Indonesia and the Philippines
[Colour figure can be viewed at wileyonlinelibrary.com]

and jobs continuously emerge while others shrink.7 Even within existing occupations, the tasks
performed by workers and the skills needed to carry them out are subject to important changes.
In the short to medium run, skill demand varies with business cycle conditions.
The supply of skills also changes, due for instance to the expansion of higher education,
increased female labour force participation, changes in retirement patterns and migration flows.8
To some extent, labour mobility is a short to medium term supply response to demand fluctuations
(see Boswell et al., 2004; Danish National Bank, 2019). Most changes on the supply side, however,
happen slowly and skill imbalances might occur consequently. While short periods of shortages
and mismatch are to be expected in any dynamic economy, persistently high skill mismatch or
shortages are symptoms of structural problems that can have adverse economic consequences for
individuals, firms and the aggregate economy.

3.1 Cyclical factors

Skill and labour shortages typically increase during economic expansions, when many firms rely
on the outside labour market to recruit the skills required to fill new positions, and decline during
recessions. Whether and how labour supply responds to rapid increases in demand depends on
the distribution of existing skills, the extent of geographical mobility and the signals provided by
wage adjustments.
Globally, the Manpower Talent Shortage Survey indicates that skill shortages have increased
since 2009 (Manpower, 2018, see Figure 4). For Europe, the European Business Survey similarly
indicates an increase in reported average labour shortages as the economy recovered from the
2008 financial crisis followed by the sovereign debt crisis (see Figure 5). This trend, however, is
 14676419, 2021, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joes.12424 by Duale Hochschule BW Mannheim, Wiley Online Library on [10/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1152 BRUNELLO and WRUUCK

F I G U R E 5 Incidence of industrial companies indicating that labour shortages are a factor currently limiting
their production (%), by macro area
Source: Eurostat Business Surveys, several years. Southern Europe: Italy, Spain and Portugal; Eastern Europe:
Czech Republic, Hungary, Latvia, Lithuania, Poland, Slovenia and Slovakia; Northern and Central Europe:
Austria, Belgium, Denmark, France, Germany, Luxembourg, Netherlands, Finland, Sweden and the United
Kingdom
[Colour figure can be viewed at wileyonlinelibrary.com]

considerably less pronounced in Southern Europe, where recovery has been slower than in the
rest of the Continent.
Increasing shortages in Eastern Europe have been accompanied by sustained real earnings
growth, in contrast with developments in the ‘core’ countries of Northern and Central Europe
(see Figure 6), where shortages have also increased but real earnings growth had been slower to
pick up as labour market slack has been more persistent in some countries (see ECB, 2017).
The relationship between skill mismatch and the business cycle is driven by several factors.
In downturns, two factors are at play: on the one hand, mismatch declines because low quality
jobs are destroyed but high-quality matches between firms and employees survive. On the other
hand, mismatch increases either because firms post fewer vacancies and job seekers are willing
to accept less-desirable jobs due to the higher competition they face, or because those who have
lost their jobs do not have the skills required to fill the existing vacancies.
When demand is buoyant and the labour market is tight, employers may be forced to adjust their
hiring standards downward to cope with difficulties in recruiting skilled labour, which increases
the incidence of under-skilling and of lower quality matches (Healy et al., 2015; Livanos & Nunez,
2017).
Which effect prevails during an economic downturn is unclear and may depend on labour mar-
ket institutions and the approach of firms to talent management. Empirical evidence based on
U.S. data suggests that the destruction of low-quality jobs dominates in recessions (Baley et al.,
2018) and that skill mismatch is therefore pro-cyclical. In Europe, stricter employment protec-
tion9 (Belot, 2007) – which reduces involuntary separations in a downturn – suggests that this
destruction may be less important than in the United States, casting doubts on the pro-cyclicality
of skill mismatch in the former area. Consistent with this conjecture, Liu et al., 2012, find that in
 14676419, 2021, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joes.12424 by Duale Hochschule BW Mannheim, Wiley Online Library on [10/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BRUNELLO and WRUUCK 1153

F I G U R E 6 Real hourly earnings growth in manufacturing. By macro area

Source: Eurostat Business Surveys, several years. Southern Europe: Italy, Spain and Portugal; Eastern Europe:
Czech Republic, Hungary, Latvia, Lithuania, Poland, Slovenia and Slovakia; Northern and Central Europe:
Austria, Belgium, Denmark, France, Germany, Luxembourg, Netherlands, Finland, Sweden and the United
Kingdom
[Colour figure can be viewed at wileyonlinelibrary.com]

Norway skill mismatch among college graduates is strongly counter-cyclical and driven both by
over-skilling and by downgrading in the average quality of job matches.10
A reason why economic downturns can increase mismatch is that matching efficiency, defined
as the quality of the matching process involving unemployed workers and unfilled vacancies,
declines.11 Matching efficiency can be improved by active labour market policies (ALMP), includ-
ing re- and up-skilling measures increasing the matching prospects of the long term unemployed.
Escudero (2018), provides evidence that ALMP improved employment outcomes in OECD
countries from 1985 to 2010, particularly for low-skilled individuals. During the Great Recession
of 2008/2009, 11.7% of crisis response stimulus measures in advanced economies were devoted
to active labour market programmes. In high-income countries, training was the most common
policy response, and job training and re-training programmes were implemented by government
to help unemployed or dislocated workers get back into jobs (see Hansen, 2012).
Business cycle effects on matching efficiency are mediated by the strategies and management
practices of firms, which can fill new vacancies either by hiring or by developing the required
skills in-house, using training and upgrading.12 In the United States, for example, the observed
decline of promotion-from-within systems in medium and large firms is bound to increase hiring
challenges substantially, by expanding the range of jobs through which hiring takes place: from
“entry-level” jobs filled by inexperienced school leavers to virtually every position in the organiza-
tion (Cappelli, 2014).13 This shift in human resources management practices can accentuate skill
shortages during expansions.14 It is not clear whether this applies to Europe as well. One indicator
that this may not be the case is that average job tenure has been declining in the United States
(see Farber, 2010) – in line with the reduced importance of promotion-from-within systems – but
has remained more or less constant in Europe (see Cazes & Tonin, 2010).
 14676419, 2021, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joes.12424 by Duale Hochschule BW Mannheim, Wiley Online Library on [10/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1154 BRUNELLO and WRUUCK

3.2 Structural factors

Demographic trends are one factor that can add to skill shortages through its impact on size,
age and composition of the labour force. Demographic change also impacts on the demand for
goods and services, and hence on the demand for the skills necessary to provide them, medical
services and personal care being one example. The combination of shifts in the demand for labour
towards more skilled jobs and population ageing can produce skill shortages and mismatch as
older workers are endowed with partially outdated education and skills that do not match closely
with those required by the process of digitalization of modern economies (European Parliament,
2005).15
Developed economies have a comparative advantage in goods and services that use (high)-
skilled labour intensively and globalization tends to reinforce this.16 Globalization – as a stan-
dalone factor and in combination with technological change – has been associated with increas-
ing polarization of the occupational structure and a decline in middle-skilled jobs, and can pro-
duce skill shortages if labour reallocates slowly from declining to growing sectors where different
skillsets may be needed (see Goos et al., 2014). To what extent this materializes depends on the
intensity of the adjustment shock, on the pool of skills in the workforce, and on the possibilities
to (re)deploy these skills.
Some European economies have been able to offset the employment losses from import com-
peting sectors with job creation in exporting sectors, for instance by penetrating new markets (see
Dauth et al., 2014; Donoso et al., 2014). While the spread of global value chains has been accompa-
nied by offshoring of intermediate inputs and back-office services, mainly affecting middle skilled
workers, a lot of offshoring activity has taken place, for example, as nearshoring from Western to
Eastern Europe, shifting skill demand in both locations but allowing firms to become more pro-
ductive.
Technological change is a factor inducing sectoral dynamics17 – with some sectors gaining
employment and some other sectors shrinking – and changing the demand for skills within occu-
pations and firms. Technological change can produce skill shortages by creating the need for new
skills that are not immediately available in the labour market, until the broad education system
(including employer training) is able to meet the new skill requirements. In addition, firms oper-
ating in rapidly changing markets may wish to hire workers with higher qualifications to ease
labour adaptation in the future (Desjardins & Rubenson, 2011).
Technological progress in the past few decades has been characterized by the spread of infor-
mation and communication technologies (ICT), which have resulted in (a) an acceleration in the
demand for skilled workers that outweighs the available supply (skill biased technical change; see
Acemoglu & Autor, 2011; Katz & Autor et al., 1999);18 (b) the combined reduction in the demand for
routine cognitive and manual tasks and increased reliance of production on no-routine tasks that
cannot be easily automated, which has led to the polarization (hollowing out) of employment – the
contemporaneous increase in the share of low-skilled and high skilled occupations and decline
in the share of medium-skilled jobs (see Autor et al., 2003; Das & Hilgenstock, 2018; Goos et al.,
2009).19
Projections based on the Skills and Job forecast by CEDEFOP for Europe indicate that employ-
ment shares are rising and expected to increase further for professionals, managers and techni-
cians on the one hand and for elementary occupations on the other hand, and to decline for clerks,
craft workers and plant and machine operators, suggesting (further) polarization.20
 14676419, 2021, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joes.12424 by Duale Hochschule BW Mannheim, Wiley Online Library on [10/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BRUNELLO and WRUUCK 1155

F I G U R E 7 Skills premiums in European economies

Source: IMF, 2018
[Colour figure can be viewed at wileyonlinelibrary.com]

There is evidence that the share of middle paying jobs has decreased between 2002 and 2016,
a trend also associated with technological change (European Commission, 2018b). While polar-
ization has affected all countries, its intensity has been higher than average in France, the United
Kingdom, Ireland and Portugal – among others – and lower in Germany, Poland and Denmark
(see also Goos, et al., 2009).
In spite of the fact that recent technological progress has increased the relative demand for skills
and non-routine tasks, technological developments are not necessarily an incessant force creating
demand for higher skills. First, the college wage premium, after increasing in several countries
during the 1980s and the 1990s, has remained relatively stable in the 2000s, at least in the United
States, in spite of a slowdown in the increase in the number of college graduates, and consistent
with a slowdown in the demand for college skills (Cappelli, 2014).21 The maturation of information
technology slowing the demand for higher skills, a levelling off of complementarities between
highly educated labour and new technologies and stronger competition between education groups
for increasingly scarce well-paid jobs have been suggested as factors explaining the flattening skill
premium in the United States. In Europe, there is some indication (see Figure 7) that the skill
premium, that is the ratio of the skilled to unskilled wages, has remained stable or mildly declined
in recent years (IMF, 2018), possibly reflecting increased competition for jobs as the economy
emerged from the 2008 recession.
Second, the recent advances in digital technology suggest that digital innovation and machine
learning may be able to substitute not only routine tasks but also some non-routine tasks that
are typically performed by skilled workers (Brynjolfsson & McAfee, 2014; Frey & Osborne, 2017).
Estimates of the impact of automation on jobs are subject to some uncertainty. Compared to Frye
 14676419, 2021, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joes.12424 by Duale Hochschule BW Mannheim, Wiley Online Library on [10/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1156 BRUNELLO and WRUUCK

and Osborne (2017), who find that 47% of the jobs in the United States are at high risk of being
automated, a recent OECD study (Nedelkoska & Quintini, 2018) estimates that about 14% of jobs in
OECD countries participating to the Survey on Adult Skills (PIAAC) are highly automatable, that
is, have a probability of automation of over 70%.22 This is equivalent to over 66 million workers in
the 32 countries covered by the study. In addition, another 32% of jobs have a risk of between 50%
and 70% of being automated. Richer countries appear to be less at risk than middle income ones
but wide gaps exist between countries of similar wealth, reflecting differences in specialization,
job content and in institutional and organizational structures.
Low skilled jobs are considered to be more automatable. Unfortunately, workers in these jobs
are less likely to invest in training (see Nedelkoska & Quintini, 2018) and often have limited access
to it. A question this prompts is what strategies should be used to encourage training participation
and what type of training can effectively work to allow people to upgrade skills and move to jobs
that are less automatable (see for instance Goerlitz & Tamm, 2017; Schmidpeter & Winter-Ebmer,
2018; Tamm, 2018;).

4 THE COSTS OF SKILL MISMATCH AND SHORTAGES

The economic costs of skill mismatch and skill shortages affect individuals, firms and the overall
economy. Individual costs include lower wages and poorer skill development, and can partly turn
from temporary to permanent. The costs faced by firms comprise lower productivity and the hiring
and training costs associated to increasing job turnover. Aggregate costs include the efficiency
losses – in terms of lower average productivity and higher unemployment – associated with the
sub-optimal allocation of resources (see for instance Garibaldi et al., 2020).

4.1 Effects on individuals

Mismatch affects earnings negatively if it induces individuals to accept a less desirable job.23 This
effect can turn from temporary to permanent if it associates with a scarring effect, for instance
because of human capital depreciation. Summerfield and Theodossiou (2017), use German data
to show that unfavourable economic conditions at graduation decrease the likelihood of a good
job-worker match over a worker’s subsequent career. Oreopoulos et al. (2012), investigate the long-
term effects on earnings of graduating from college during a recession, when the quality of new
labour market matches is typically lower than average (see also Altonji et al., 2016). Although
these authors do not explicitly study the consequences of mismatch, a consequence of recessions
is higher competition for scarcer jobs, which makes individuals accept less desirable jobs. Using
a large longitudinal university–employer–employee dataset from Canada, they find that the cost
of recessions for new graduates in terms of lower wages is substantial and can last as long as ten
years.
The wage costs endured by those graduating in a recession may spill-over to the unskilled
because the search strategies of the former during recessions, which include accepting lower pay
jobs, crowd out the job opportunities of the latter, with consequences for their current and future
earnings. This crowding out effect could also delay the recovery of the labour market, particularly
for low skilled individuals (Arsenau & Epstein, 2014).
Assessing the direct effects of mismatch on individual productivity is difficult because of mea-
surement problems. The existing empirical evidence relies on two approaches, one exploiting the
 14676419, 2021, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joes.12424 by Duale Hochschule BW Mannheim, Wiley Online Library on [10/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BRUNELLO and WRUUCK 1157

idea that in a competitive equilibrium wages always equal marginal productivity, and the other
focusing on the impact of mismatch on individual job satisfaction. Quintini, 2011, uses data from
the European Community Household Panel – a European wide household survey that preceded
the EU–Statistics on Income and Living Conditions (EU–SILC) – to study the effects of qualifi-
cation mismatch on earnings. Adopting an empirical approach that controls for individual unob-
served heterogeneity, she finds that over-qualified individuals earn about three percent less than
individuals who have the same qualification but are well matched, and that under-qualified work-
ers earn about two to three percent more than workers with similar qualifications who are well
matched.
Under the assumption that productivity is always equal to wages, this can be interpreted as evi-
dence that over and under-qualification are associated with a productivity loss and gain, respec-
tively. An alternative view, that does not subscribe to the assumption that wages are continuously
equal to productivity, is that the over and under-qualified are as productive as the well matched but
their earnings reflect either excess supply (over-education) or excess demand (under-education)
for the job.
Studies exploring the impact of mismatch on job satisfaction have argued that, by reducing sat-
isfaction, mismatch can increase absenteeism and/or reduce productivity. Mismatched workers
are more likely not only to be absent from work more often, but also to change jobs more frequently
and invest less in training, with potentially negative consequences on productivity. The existing
evidence indicates that over-qualified workers are less likely to participate in training than well-
matched workers with the same qualifications (Verhaest & Omey, 2006), and that over-skilling
has a negative effect on job satisfaction (Allen & van der Velden, 2001).

4.2 Effects on firms

The direct costs of mismatch for firms are not easy to assess, as they require information on the
share of mismatched employees as well as on firm performance metrics. Moreover, different types
of mismatches may coexist in firms and it is not clear how these interact to affect performance.
While theoretically firms could capitalize on lower wages for over-skilled workers, negative effects
on motivation and absenteeism may negatively impact on productivity (over the longer term).
Tsang and Levin (1985), argue, for instance, that over-education might be linked to counterpro-
ductive behaviour in the workplace. Several studies (see for instance Allen & van der Velden,
2001) have found that over-qualified workers are more mobile than well-matched workers with
the same qualifications. By increasing job turnover, mismatch is costly to firms, that have to incur
additional hiring and training costs. In turn, some of the costs associated with mismatch depend
on the labour market situation and the presence of skill shortages.
The argument behind skill shortages having a negative impact on firms focuses on produc-
tion losses due to unfilled positions or on the recruitment of workers with lower skills than the
job would require (Bennett & McGuinness, 2009). It has also been argued that skill shortages
limit investment and the adoption of new technologies, with negative impact on productivity (e.g.,
Foley et al, 1993). Haskel and Martin (1993), find that the increase in the shortage of skilled labour
in the United Kingdom over the mid-1980s reduced productivity growth by around 0.7% per year.
Similarly, Bennet and McGuinness (2009), after controlling for selection effects as it is often more
productive firms that are more likely to experience skill shortages, find a negative effect of hard-
to-fill and unfilled vacancies on the output per worker in high-tech firms in Northern Ireland.
 14676419, 2021, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joes.12424 by Duale Hochschule BW Mannheim, Wiley Online Library on [10/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1158 BRUNELLO and WRUUCK

Tang and Wang (2005), also find a negative impact of skill shortages on the labour productivity of
small and medium-sized companies.
In addition, Foley et al. (1993), suggest that the shortages of craft workers in the United Kingdom
have acted as a barrier to the use of new technologies and have led to lower productivity. An
indirect effect of skill shortages on productivity via difficulties in technology adoption is also found
by Forth and Mason (2006), in their study on ICT skill shortages in the United Kingdom. Similarly,
Nickell and Nicolitsas (2000), find that skill shortages reduce firms’ investment in R&D, although
the effect is found to be only temporary.
The costs of shortages to firms also depend on their duration. Bellman and Hubler (2014), for
instance, find that skill shortages in German firms are normally short-term phenomena. Healy
et al. (2015), investigate the strategies used by firms to respond to skill shortages using the Aus-
tralian Business Longitudinal Database and find that most firms respond to skill shortages by
improved utilization of their core workforce through longer hours and better pay, while some
firms use peripheral strategies such as temporary employment and outsourcing (see McGuiness
et al., 2017).

4.3 Aggregate effects

By distorting the optimal allocation of resources, skill shortages and mismatch are expected to
reduce average productivity. Mavromaras et al. (2007), attempt to quantify the costs of skill mis-
match in terms of Gross Domestic Product (GDP). They proxy the individual productivity loss with
the estimated wage penalty associated with over-skilling, and multiply this penalty by the number
of over-skilled workers by educational attainment level, concluding that the costs of over-skilling
in Australia amount to about 2.6% of GDP in 2005.24
Skill mismatch can also affect wage inequality. Using data for the United States, Slonimczyk
(2009), reports that a substantial part of the observed increase in wage dispersion (11% for men
and 32% for women during 1973–2002) can be attributed to increases in over-qualification rates
and premia.25 Another area where skills mismatch affects aggregate outcomes is the ability of a
labour market to adjust to unemployment shocks. If there is little skills mismatch, then presum-
ably a labour market can readjust relatively quickly, ceteris paribus, as it just has to sort out the
unemployment-vacancy inefficiency. However, if skills mismatch is widespread, then while the
market is trying to match workers with jobs it is also trying to get better matches for skill mis-
matched workers. That sorting, presumably, would take much longer as search costs would be
higher.
The negative relationship between skill mismatch and average productivity operates mainly via
two channels: lower within-firm productivity and a less efficient allocation of labour resources
across firms. McGowan and Andrews (2015a), estimate the association between skill mismatch –
measured using data from the OECD Survey of Adult Skills (PIAAC) – and labour productivity –
measured using industry data for 19 OECD countries – and report that higher skill mismatch is
associated with lower labour productivity because of a less efficient allocation of resources across
firms. Because of mismatch, more productive firms find it difficult to attract skilled labour and
gain market shares at the expense of less productive firms. Using these results, they simulate for
the countries in their sample the counterfactual productivity gains from reducing skill mismatch
to the level associated to best practice. These gains turn out to be sizeable in a number of countries
(above 9% in Italy, Spain and the Czech Republic; between 5% and 9% in Germany, Norway, Great
Britain and Austria).
 14676419, 2021, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joes.12424 by Duale Hochschule BW Mannheim, Wiley Online Library on [10/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BRUNELLO and WRUUCK 1159

The observed differences in skill mismatch across countries are associated with differences in
the institutional and policy environment. According to a recent OECD study, skill mismatch turns
out to be lower in countries with economic institutions that promote the efficient reallocation of
resources – including less cumbersome product and labour market regulations and a bankruptcy
legislation that does not excessively punish business failure – and where housing policies do not
impede residential mobility. Lower mismatch is also associated with greater flexibility in wage
negotiations and higher participation in lifelong learning as well as with higher managerial qual-
ity (see McGowan & Andrews, 2015b). Given that it relies on data from a cross section, however,
this study cannot separate the effects on mismatch of variables characterising the environment
from unobserved country effects. Results should therefore be considered as suggestive, requiring
further examination.

5 POLICIES TO ADDRESS SKILL SHORTAGES AND MISMATCH

Skill development starts from school and continues over the life cycle as individuals and firms
invest in training and on-the-job learning. Under-investment in education can occur if individu-
als do not have the resources to finance their desired education, or if they fail to properly account
for the social benefits and costs of education. Under-investment in training can also occur for
several reasons, including that (a) the private and social returns to training do not coincide, and
workers or employers only consider private returns in their rational decisions. Examples of social
returns are the spill-overs of training on the productivity of other workers and the effects of train-
ing on innovation activities (see Brunello et al., 2007); (b) employers who invest in the training
of employees are often forced to share the returns to their investment either with the employees
themselves by paying higher wages (the hold-up problem), or with other employers, who can hire
the trained workers (the poaching problem).
Policies addressing skills mismatch tend to concentrate on developing initiatives aimed at
enhancing the responsiveness of the education and training system to emerging labour market
needs. This include for instance: (a) steps to reform vocational education and training (VET) edu-
cation in some countries and enhancing the employability of young people – see for instance the
European Youth Employment Initiative;26 (b) forecasting future skill needs and supply by using
occupational models, sectoral or occupational skills councils and commissioning of bespoke qual-
itative and quantitative research projects. The view that skills mismatch is also a function of asym-
metric information between jobseekers, workers and firms has led some countries to improve
career guidance and counselling services (McGuiness et al., 2017).
The concerns that firms express about the limited availability of skills as a factor limiting corpo-
rate investment are reflected in their views about public policy priorities. When asked about the
areas where the public sector should focus in the next three years, 24% of the European respon-
dents to the EIBIS survey in 2017 chose professional training and higher education as a policy pri-
ority. By country, this share tends to be higher where concerns about skills are more pronounced.
Does the responsibility for developing the skills that employers want fall exclusively onto job
seekers and schools or should employers share it? Schools designing curricula are likely to find
it difficult to foresee what skills will be in demand three or five years after the programs start
(Cappelli, 2014). Pupils investing in vocational programs that may fit future employer demands
need to balance the advantages of a rapid school to work transition with the expected costs of skill
obsolescence in the medium to long run and may therefore invest in more academic skills (see
Brunello & Rocco, 2017).
 14676419, 2021, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joes.12424 by Duale Hochschule BW Mannheim, Wiley Online Library on [10/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1160 BRUNELLO and WRUUCK

Along these lines, results from a recent survey on VET show that the choice of VET is often
guided by the expectation of finding a job quickly, while general education students tend to report
their choice as based on the possibility to continue to higher education (Salvatore & Villalba-
Garcia, 2018).27 A tension exists between firms with skill shortages advocating more vocational
education and individuals facing an uncertain professional future and the risks of automation and
technical progress, who try to diversify risks by choosing broader educational curricula. Increas-
ing permeability of VET systems and options for life-long learning might be steps to mitigate these
tensions.
Schools are not particularly well suited to provide work experience that is better acquired in
the workplace using apprentice-like arrangements (Cappelli, 2014). However, implementing well-
designed apprenticeship systems where they are less common and keeping firms involved in train-
ing apprentices – even where systems have a longstanding history – remains a continuous chal-
lenge for many countries (see Wieland, 2015 for Spain; Wieland, 2018 for the Italy; Thies, 2018).
Employers can respond to skill shortages by activating several measures, including training,
recruitment abroad, improved working conditions to attract qualified applicants and internal re-
organizations aimed at increasing the productivity of existing employees. A study by the European
Commission (European Commission, 2014b) shows that the relative importance of each measure
depends on the occupations experiencing shortages. For instance, training and improving work-
ing conditions turn out to be particularly frequent in ICT and sales occupations, respectively.
Recruitment efforts outside the country are instead frequent in the health sector (see Directorate
General for Internal Policies, European Parliament, 2015).
Persistent skill shortages that are not solved by market mechanisms can be addressed by govern-
ment policies, which can try to reduce the under-provision of education or training. In Europe,
for instance, public policies encouraging adult training include co-financing programs targeted
at firms, for example levy-grant schemes, tax deductions, and co-financing programs targeted at
individuals, for example vouchers or individual learning accounts (Brunello et al., 2007).
Skill policies can also provide a way to proactively respond to structural trends such as dig-
italization and the challenges that come with it, notably rising polarization in labour markets
and inequality. To that extent, having policies in place that support a smooth transition between
jobs and the development of qualifications required by the labour market will be increasingly
important – both for firms requiring different skillsets as well as for job seekers in labour mar-
kets. How this can best be achieved against the backdrop of the ongoing technological changes
remains an area for future research. However, since skill development interacts with labour mar-
ket and broader policies (e.g., competition, housing, migration policies), the links across policy
areas should clearly be considered.

6 CONCLUDING REMARKS

This paper has reviewed the recent economic literature on skill mismatch and shortages. The key
points can be summarized as follows:

1. There are different approaches to measure skill mismatch that produce substantially different
results. Measures relying on self-reported mismatch produce a much lower share of well-
matched individuals than statistical measures that compare individual skills with average
skills in the occupation.
 14676419, 2021, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joes.12424 by Duale Hochschule BW Mannheim, Wiley Online Library on [10/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BRUNELLO and WRUUCK 1161

2. Employers and managers are likely to have more accurate information than employees about
skill requirements. Asking employers not only about impediments to economic activity and
investment behaviour but also about skill-related issues the firm is facing is a valuable com-
plement to the information asked to employees.
3. Indicators of skill shortages derived from employer surveys need to be complemented with
indirect measures of the presence of shortages in specific occupations, including price mea-
sures (wage growth), volume measures (employment growth, vacancy rates) and work inten-
sity measures (incidence of overtime).
4. The existing evidence points to pro-cyclical skill mismatch in the United States and to counter-
cyclical mismatch in Europe. This contrast may be due to the fact that involuntary separations
in a recession are more difficult in Europe than in the United States.
5. Skill shortages declared by firms need not be always genuine. For instance, after the 2008
recession, reported shortages have increased in many European countries, yet real earnings
growth has remained subdued, with the exception of Eastern Europe, suggesting that in some
areas these shortages may be over-estimated.
6. Structurally, the adoption of new technologies creates the demand for new skills that are not
immediately available in the labour market, giving rise to skill shortages until the broad edu-
cation system (including employer training) is able to meet the new skill requirements. The
importance of these shortages and the length of the adjustment process can be exacerbated if
wages and working conditions fail to provide adequate signals of relative scarcity.
7. With the impact of the COVID-19 shock on labour markets still unfolding, it is too early to
assess its effects on skill shortages and mismatch. Yet, three facts point to an increase of skill
mismatch. First, in some countries the use of short-time working or furlough schemes and
dismissal freezes has encouraged firms to hold on to their staff, reducing the immediate elim-
ination of inefficient matches. Second, hiring has stalled overall, with some sectors affected
by the lockdown and other sectors experiencing a surge in demand. Meeting this demand at
a time when job interviews are difficult and training can be done only online has been a chal-
lenge (OECD, 2020). Third, COVID-19 is accelerating the ongoing trend towards digitalization
(European Investment Bank, 2020). To address COVID-19 induced mismatch, governments
should take measures to facilitate the retraining and redeployment of displaced workers and
of those currently on short time working schemes who may not be able to return to their
former jobs due to firm bankruptcies or changes in demand for some product and services
(OECD, 2020).
8. Skill shortages and mismatch are costly to individuals, firms and society because they nega-
tively affect earnings, productivity, innovation and productivity growth. The effects on earn-
ings can be quite persistent.
9. The responsibility for developing the skills that employers want – which includes financing
skill development – should fall both onto job seekers and schools and onto employers. Per-
sistent skill shortages that are not solved by market mechanisms can be addressed to some
extent by government policies. Importantly, the effects of these policies need to be accurately
evaluated.
10. Future research in the area of skill shortages and skill mismatch should try to increase our
ability to measure these economic phenomena in a satisfactory way. In particular, finding
ways to separate genuine from non-genuine shortages and providing adequate measures of
skill requirements are in our view crucial steps for improving our understanding of skill mis-
match and shortages. In addition, the systematic assessment of measures to mitigate skill
mismatches, including a strong focus on impact assessment, would inform policy design and
 14676419, 2021, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joes.12424 by Duale Hochschule BW Mannheim, Wiley Online Library on [10/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1162 BRUNELLO and WRUUCK

support the implementation of effective skill and labour market support measures. Finally, the
use of new analytical techniques such as big data analysis to better capture current and antic-
ipated skill demand could be useful to improve the measurement of mismatch and become
an important tool to help mitigating its effects.

ORCID
Giorgio Brunello https://orcid.org/0000-0002-4830-4199

ENDNOTES
1
Despite the key role skills play in the labour market and economies, there is little agreement in the literature as
to what ‘skills’ are and how they should be defined. See for instance Clarke and Winch (2006).
2
Eurostat, 2016, further classifies ‘micro’ skills mismatch into: (a) vertical mismatch, that is the mismatch
between formal education and job requirements measured against a benchmark; (b) horizontal mismatch, for
example mismatches between the worker’s field of education and job requirements. Mismatches have also been
analysed in terms of over-skilling and under-skilling of workers.
3
The Survey on Adult Skills (PIAAC) collects information on four digits ISCO 2008 occupations.
4
McGuiness and Ortiz (2014), compare the skill gaps as perceived by managers and employees located in the
same firm. Their evidence suggests that employee perceptions may be prone to higher level of subjective bias
than employer subjective measures. The latter can also be biased, however, if managers do not really know the
skills needed to complete a job. Decentralizing decision making at least in part to workers would allow to take
advantage of the information that workers have that managers/firms do not have. Barron et al. (1997), examine to
what extent employer and employee responses to training questions are consistent and find that establishments
report 25% more hours of training than workers do.
5
For further description of the EIBIS, the methodology and the questionnaire see also
http://www.eib.org/en/about/economic-research/surveys-data/about-eibis.htm
6
At the same time, the demand side determinants can be affected by skills supply, for example through production
and adoption of innovation or differences in consumption patterns.
7
A distinctive trait of developed economies is the decline in manufacturing jobs. See IMF (2018). Another trait
is how manufacturing that remains is done. The skill mix of workers in modern manufacturing in developed
countries is likely very different than the skill mix from earlier generations or even of manufacturing workers
in low-income countries.
8
See IMF (2018), for a discussion of recent trends in labour force participation by gender and age, and Batsaikhan
et al. (2018), on migration flows into the European Union.
9
The OECD index of employment protection regulation in 2019 was 0.1 in the United States and 2.6 in France,
Germany and Italy.
10
In a similar fashion, Summerfield (2015), shows that jobs formed in a recessions have relatively more man-
ual tasks, increasing the probability that workers are overqualified. Counter-cyclical mismatch and pro-cyclical
productivity suggest that the former should negatively correlate with the latter. Conditional on business cycle
effects, however, the residual correlation could take either sign, as shown by European Commission (2017).
11
A recent analysis (European Commission, 2018a) examining determinants of matching efficiency in EU labour
markets suggests that sectoral mismatches, measured by unemployment dispersion across sectors, have a big-
ger impact on aggregate matching efficiency during recoveries, whereas skill mismatches matter more during
normal periods.
12
See Stevens (1994); Brunello (1996); Brunello et al. (2007). Using administrative survey data with detailed infor-
mation on hiring and training costs for Swiss firms, Blatter et al. (2016), find evidence that firms can invest in
internal training of unskilled workers and thereby avoid the costs for external hiring.
13
“. . . Hiring may well be more difficult now simply because employers have to do much more hiring these days
because of widespread and substantial declines in employee tenure . . . , which translates into more frequent
vacancies and more hiring to fill them. The decline of life-time employment practices and the associated rise of
lateral hiring have been underway for some time especially in larger organizations. . . ” (Cappelli, 2014).
 14676419, 2021, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joes.12424 by Duale Hochschule BW Mannheim, Wiley Online Library on [10/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BRUNELLO and WRUUCK 1163

14
Business cycle effects are mediated (or enhanced) by worker preferences too. For example, if there is a reluctance
to move to other locations where a better match can be found, then firm behaviour may not matter even if they
change management practices.
15
We do not discuss in this review non-education related reasons for skill mismatch, including for instance those
induced by partners moving to new jobs.
16
It is reasonable to assume that countries with higher GDP per capita tend to be better endowed with skilled
labour in global comparison and that they specialize in production of goods for which they have a comparative
advantage.
17
Sectoral dynamics may refer to shifts across sectors as well as the emergence of new products.
18
During the previous decade, about a million jobs were created in Europe in occupations tightly knit to ICT ser-
vices, such as ICT professionals, technicians and associate professionals. According to CEDEFOP skill forecasts,
a further half a million more jobs are anticipated to be created in the next decade in Europe (CEDEFOP, 2018).
19
A consequence of widespread digitalization is the rising incidence of new online forms of platform employment
or crowd work in the so-called gig economy. According to McKinsey (2015), such platforms could add more than
370 billion to the EU economy and bring more than 5.2 million workers into employment. See CEDEFOP (2018).
20
In a similar fashion, the top growth occupations identified by CEDEFOP (2016b), and the European Vacancy and
Recruitment report (European Commission, 2014a) are in health, ICT, engineering, teaching, administration
and sales (see also Directorate General for Internal Policies, European Parliament, 2015).
21
The observed slowdown in the supply of college graduates may have been even higher if one considers the
possibility that the average quality of graduates has declined because of increased access.
22
Quintini and Nedelkoska build on the methodology developed by Fry and Osborne to estimate automation risk
but go beyond occupation levels in their analysis, using PIAAC data to assess the relationship between job tasks
and the risk of automation.
23
See Iriondo and Perez-Amaral (2013), and Sellami et al. (2017), for recent reviews of the literature on the effects
of mismatch on wages, and Haskel and Martin (2001), on the effect of skill shortages on wages.
24
See Vandeplas and Thum-Thysen (2019), for a recent review of the relationship between skill mismatch and
productivity.
25
Budria and Moro-Egido (2006), also find that mismatch contributes to higher wage differences within education
groups.
26
Policies not discussed here include structural reforms changing labour market institutions. See for instance De
Haan and Parlevliet (2018).
27
Results are based on a survey among 35,000 students at upper secondary level in 28 EU Member States.

REFERENCES
Acemoglu, D., & Autor, D. (2011). Skills, tasks and technologies: Implications for employment and earnings. (NBER
Working Paper no. 16082.) National Bureau of Economic Research.
Allen, J., & van der Velden, R. (2001). Educational mismatches versus skill mismatches: Effects on wages, job
satisfaction, and on-the-job search. Oxford Economic Papers, 53(3),434–452.
Altonji, J., Kahn, L., & Speer, J. (2016). Cashier of consultant? Entry labour market conditions, field of study and
career success. Journal of Labor Economics, 34, S361–S401.
Arsenau, D., & Epstein, B. (2014). The welfare costs of skill mismatch unemployment. Finance and Economics Dis-
cussion Series, Divisions of Research & Statistics and Monetary Affairs. Federal Reserve Board.
Autor, D., Levy, F., & Murnane, R. (2003). The skill content of recent technological change: An empirical explo-
ration. Quarterly Journal of Economics, 118(4), 1279–1333.
Autor, D., Dorn, D., & Hanson, G. (2013). The China syndrome: Local labor market effects of import competition
in the United states. American Economic Review, 103, 2121–2168.
Baley, I., Figueredo, A., & Ulbright, R. (2018). Mismatch cycles, mimeo.
Barron, J., Berger, M., & Black, D. (1997). How well do we measure training? Journal of Labor Economics, 15(3),
507–528.
Batsaikhan, U., Darvas, Z., & Raposo, I. (2018). People on the move: Migration and mobility in the European Union.
Bruegel.
 14676419, 2021, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joes.12424 by Duale Hochschule BW Mannheim, Wiley Online Library on [10/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1164 BRUNELLO and WRUUCK

Bellman, L., & Hubler, O. (2014). Skill shortages in German establishments. IZA Discussion Paper no. 8290. Institute
for the Study of Labor.
Belot, M. (2007). Why is employment protection stricter in Europe than in the US? Economica, 74(295), 397–423.
Bennet, J., & McGuiness, S. (2009). Assessing the impact of skill shortages on the productivity performance of
high-tech firms in Northern Ireland. Applied Economics, 51(6), 727—737.
Blatter, M., Muehlemann, S., Schenker, S., & Wolder, S. (2016). Hiring costs for skilled workers and the supply of
firm-related training. Oxford Economic Papers, 68(1), 238–257.
Boswell, C., Stiller, S., & Straubhaar, T. (2004). Forecasting labour and skill shortages: How can projections better
inform labour migration policies? Paper prepared for the European Commission. Office for Official Publications
of the European Communities.
Brynjolfsson, E., & McAfee, A. (2014). The second machine age: Work, progress, and prosperity in a time of brilliant
technologies. WW Norton & Company.
Brunello, G. (1996). Equilibrium unemployment with internal labour markets. Economica, 63, 19–35.
Brunello, G., Garibaldi, P., & Wasmer, E. (2007). Education and Training in Europe, Oxford University Press.
Brunello, G., & Rocco, L. (2017). The labor market effects of academic and vocational education over the life cycle:
Evidence based on a British cohort. Journal of Human Capital, 11(1), 106–166.
Budria, S., & Moro-Egido, A. (2006). Education, over-education and wage inequality: Evidence for Spain. Studies
on the Spanish Economy, 211, FEDEA.
Bureau of Labor Statistics. (2020). Job openings and labor turnover highlights, January 2020. Bureau of Labor Statis-
tics.
Cappelli, P. (2014). Skill gaps, skill shortages and skill mismatches: Evidence from the US. (NBER Working Paper
20382). National Bureau of Economic Research.
Cazes, S., & Tonin, M. (2010). Employment protection legislation and job stability: A European cross-country anal-
ysis. International Labour Review, 149 (3), 261–285.
CEDEFOP. (2015). Skill shortages and gaps in European enterprises. CEDEFOP.
CEDEFOP. (2016a). Future skills needs in Europe: Critical labour force trends, CEDEFOP.
CEDEFOP. (2016b). Skill shortage and surplus occupations in Europe. Briefing note. CEDEFOP.
CEDEFOP. (2018). Insights into skill shortages and skill mismatch. CEDEFOP.
Clarke, L., & Winch, C., (2006). A European skill framework? But what are skills? Anglo-Saxon versus German
concepts. Journal of Education and Work, 19(3), 225–269.
Das, M., & Hilgenstock, B. (2018). The Exposure to Routinization: Labour Market Consequences for Developed and
Developing Economies, IMF Working Paper, 18, 135.
Dauth, W., Findeisen, S., & Suedekum, J. (2014). The Rise of the east and the far east: German labor markets and
trade integration. Journal of the European Economic Association, 12, 1643–1675.
Danish National Bank, (2019). Intra-EU labour mobility dampens cyclical pressures. Danish National Bank publica-
tion.
De Haan, J., & Parlevliet, J. (2018). Structural reforms moving the economy forward. Springer.
Desjardins, R., & Rubenson, K. (2011). An analysis of skill mismatch using direct measures of skills. (OECD Education
Working Papers no.63) OECD.
Directorate General for Internal Policies, European Parliament. (2015). Labour market shortages in the European
union. European Parliament.
Donoso, V., Martın, V., & Minondo, A. (2014). Do differences in exposure to Chinese imports lead to differences in
local labour market outcomes? An analysis for Spanish provinces. Regional Studies, 49(10), 46–64.
Escudero, V. (2018). Are active labour market policies effective in activating and integrating low-skilled individuals?
An international comparison. IZA Journal of Labor Policy, 7(1), 4.
European Parliament, (2005). Labour Market Shortages in the European Union, Directorate General for Internal
Policies, Policy Department A: Economic and Scientific Policy.
European Central Bank. (2017). Euro Area labour markets and the crisis. (Occasional Paper series no. 138). European
Central Bank.
European Commission. (2014). European vacancy and recruitment report. Publications Office of the European
Union.
European Commission. (2014a). Mapping and analysing bottleneck vacancies in EU labour markets. Overview report.
Publications Office of the European Union.
 14676419, 2021, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joes.12424 by Duale Hochschule BW Mannheim, Wiley Online Library on [10/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BRUNELLO and WRUUCK 1165

European Commission, (2017). Skill mismatch and productivity in the EU. Discussion Paper 100. European Com-
mission.
European Commission. (2018). Labour market and wage developments in Europe 2018. Annual Review, European
Commission.
European Commission. (2018a). Employment and social developments in Europe, Brussels.
European Investment Bank. (2020). EIBIS 2020 – EU Overview. European Investment Bank.
Farber, H. S. (2010). Job loss and the decline in job security in the United States. In K. G. Abraham , J. R. Spletzer,
& M. Harper (Eds.), Labor in the new economy (pp. 223–262). University of Chicago Press.
Foley, P., Watts H., & Wilson, B. (1993). New Technologies, skill shortages and training strategies. In P. Swann (Ed.),
New technologies and the firm: Innovation and competition (pp. 131–152). Routledge.
Forth, J., & Mason, G. (2006). Do ICT skill shortages hamper firms’ performance? In Evidence from UK benchmark-
ing surveys. Discussion Paper No. 281. National Institute of Economic and Social Research.
Frey, C. B., & Osborne, M. A. (2017). The future of employment: How susceptible are jobs to computerisation?
Technological Forecasting and Social Change, 114, 254–280.
Garibaldi, P., Gomes, P., & Sopraseuth, T. (2020). Output costs of education and skill mismatch. (Discussion paper
no. 12974). IZA Institute of Labor Economics.
Goerlitz, K., & Tamm, M. (2017). Information, Financial aid and training participation: Evidence from a randomized
field experiment. Labour Economics, 47, 138–148.
Goos, M., Manning, A., & Salomons, A. (2009). Job polarization in Europe. The American Economic Review Papers
and Proceedings, 99(2), 58–63.
Goos, M., Manning, A., & Salomons, A. (2014). Explaining job polarization: Routine-biased technological change
and offshoring. The American Economic Review, 104(8), 2509–2526.
Green, F., & McIntosh, S. (2007). Is there a genuine under-utilization of skills amongst the over-qualified? Applied
Economics, 39(4-6), 427–439.
Hansen, (2012). Preliminary assessment of training and retraining programmes implemented in response to the great
recession, Employment Working Paper No. 122, ILO, International Labor Organization.
Hartog, J. (2000). Over-education and earnings: Where are we, where should we go? Economics of Education Review,
19(2), 131–147.
Haskel, J., & Martin, C. (2001). Technology, wages, and skill shortages: Evidence from UK micro data. Oxford Eco-
nomic Papers, 53(4), 642–658.
Haskel, J., & Martin, C. (1993). Do skill shortages reduce productivity? Theory and evidence from the United King-
dom. The Economic Journal, 103, 386–394.
Healy, J., Mavromaras, K., & Sloane, P. (2015). Adjusting to skill shortages in Australian SMEs. Applied Economics,
47(24), 2470–2487, May.
Iriondo, I., & Perez-Amaral, T. (2013). The effect of educational mismatch on wages using European panel data.
(Queen Mary Working Paper no. 700). Queen Mary University of London.
Katz, L., and Autor, D. (1999). Changes in the wage structure and earnings inequality. In O. Ashenfelter & D. Card
(Eds.), Handbook of labor economics, (Vol. 3, pp. 1463–1568). Elsevier Science.
International Monetary Fund. (2018). World Economic Outlook, International Monetary Fund.
Liu, K., Salvanes, K., & Sorensen, K. (2012). Good skills in bad times: Cyclical skill mismatch and the long-term effect
of graduating in a recession. (IZA Discussion Paper no. 6820). Institute for the Study of Labor.
Livanos, I., & Nunez, I. (2017). Re-thinking under-skilling: Evidence from the first European skills and jobs survey,
CEDEFOP/IZA workshop on skills and skills mismatch. Emerald Publishing.
Manpower. (2018). Manpower talent shortage survey. Manpower Group.
Margolis, D. (2014). Defining and measuring technical, cognitive and non-cognitive skills. The World Bank.
Mavromaras, K., McGuinness, S., & Wooden M. (2007). Overskilling in the Australian labour market. The Aus-
tralian Economic Review, 40, 307–312.
McGowan, M. A., & Andrews, D. (2015a). Labour market mismatch and labour productivity: Evidence from PIAAC
data. The future of productivity: Main background papers. OECD.
McGowan, M. A., & Andrews, D. (2015b). Skill mismatch and public policy in OECD countries. The future of pro-
ductivity: Main background papers. OECD.
McGuiness, S., Pouliakas, K., & Redmond, P. (2017). How useful is the concept of skill mismatch?. ILO.
 14676419, 2021, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joes.12424 by Duale Hochschule BW Mannheim, Wiley Online Library on [10/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1166 BRUNELLO and WRUUCK

McGuiness, S., & Ortiz, L. (2014). Who should we ask? Employer and employee perceptions of skill gaps within firms.
ESRI Working Paper n.482. Economic and Social Research Institute.
McGuiness, S., Pouliakis, K. & Redmond, P. (2018). Skill Mismatch: Concepts, Measurement and Policy
Approaches, Journal of Economic Surveys, 32(4), 985–1005.
McKinsey Global Institute, (2015). A labour market that works: connecting talent with opportunity in the digital age.
McKinsey & Company.
Monti, P., & Pellizzari, M. (2016). Skill mismatch and labour shortages in the Italian labour market. Policy Brief, n.
2, Bocconi University.
Nedelkoska, L., & Quintini, G. (2018). Automation, skills use and training. OECD Social, Employment and Migration
Working Papers No. 38. Technical Report 38, 2018. OECD.
Nickell, S., & Nicolitsas D. (2000). Human capital, investment and innovation: What are the connections. In R.
Barrell, G. Mason, & M. O’Mahoney (Eds.) Productivity, innovation and economic performance (pp. 268–280).
Cambridge University Press.
OECD. (2017). Getting skills right: Skills for jobs indicators, OECD.
OECD. (2020). Skill measures to mobilise the workforce during the COVID-19 crisis. OECD.
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, 4(1), 1–29.
Pellizzari, M., & Fichen, A. (2017). A new measure of skill mismatch: Theory and evidence from PIAAC. IZA Journal
of Labor Economics, 6(1), 1–30.
Pouliakas, M., & Sekmokas, M. (2018). Automation, Skills demand and adult learning. In European Investment
Bank, Investing in Europe’s future: The role of education and skills. European Investment Bank.
Quintini, G. (2011). Right for the job: Over-qualified or under-skilled? OECD SocialEmployment and Migration Work-
ing Papers no.120. OECD.
Sala, G. (2011). Approaches to skill mismatch. Revista de Sociologia, 94(4), 1026–1045.
Salvatore, L., & Villalba-Garcia, E. (2018). Investing in Europe’s future education and skills: Public opinions on
vocational education and training. In Investing in Europe’s future: The role of education and skills. European
Investment Bank.
Savšek, S. (2018). What are the main obstacles to hiring after recessions in Europe? ECB Working Paper Series no.
2153. European Central Bank.
Schmidpeter, B., & Winter-Ebmer, R. (2018). How do automation and off-shorability influence unemployment dura-
tion and subsequent job quality? (Economics Working Paper 2018-08). Department of Economics, University of
Linz.
Sellami, S., Verhaest, D., Nonneman, W., & Van Trier, W. (2017). The impact of educational mismatches on wages:
The influence of measurement error and unobserved heterogeneity. B.E. Journal of Economic Analysis & Policy,
17(1). https://doi.org/10.1515/bejeap-2016-0055.
Slonimczyk, F. (2009). Skill mismatch and wage inequality in the US. Open Access Dissertations, University of
Massachusetts.
Somers, M., Cabus, S., Groot, W., & van den Brink, H. (2019). Horizontal Mismatch between Employment and Field
of Education: Evidence from a Systematic Literature Review, Journal of Economic Surveys, 33 (2), 567–603.
Stevens, M. (1994). A theoretical model of on-the-job training with imperfect competition. Oxford Economic Papers,
46 (4), 537–562.
Summerfield, F. (2015). Matching skill and tasks: Cyclical fluctuations in the over-qualification of new hires. Univer-
sity of Aberdeen.
Summerfield, F., & Theodossiou, I. (2017). The effects of macroeconomic conditions at graduation on over-
education. Economic Inquiry, 55(3), 1370–1387.
Tang, J., & Wang, W. (2005). Product market competition, skill shortages and productivity: Evidence from Canadian
manufacturing firms. Journal of Productivity Analysis, 23, 317–339.
Tamm, M. (2018). Training and changes in job tasks. (IZA Discussion Paper no.11787). Institute for the Study of
Labor.
Thies, L. (2018). What is there to learn from the German apprenticeship system? In Investing in Europe’s future: The
role of education and skills (pp. 42—26). European Investment Bank.
Tsang, M. C., & Levin, H. M. (1985). The economics of over-education. Economics of Education Review, 4(2), 93–104.
 14676419, 2021, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joes.12424 by Duale Hochschule BW Mannheim, Wiley Online Library on [10/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BRUNELLO and WRUUCK 1167

Vandeplas, A., & Thum-Thysen, A. (2019). Skill mismatch and productivity in the EU. (Discussion Paper 100). Euro-
pean Commission.
Verhaest, D., & Omey, E. (2006). The Impact of Over-education and its Measurement. Social Indicators Research,
77, 419–448.
Wieland, C. (2015). Apprenticeship training in Spain – a cost effective model for firms?. Bertelsmann Foundation.
Wieland, C. (2018). Apprenticeship training in Italy – a cost effective model for firms?. Bertelsmann Foundation.

How to cite this article: Giorgio Brunello, Patricia Wruuck. Skill shortages and skill
mismatch: a review of the literature. J Econ Surv. 2021;35:1145–1167.
https://doi.org/10.1111/joes.12424