Page 120 - ePaper
P. 120
Employment and Social Developments in Europe 2014
health is thus an important precondition for policies, which cannot all be presented The aim of the regression is to find evi-
maintaining the available human capital. and analysed here. We decided to focus dence for the human capital deprecia-
on the importance of work intensity on tion phenomenon, such as the impact of
3.3. Using human capital skills performance, on women and labour skills and work intensity on PIAAC perfor-
segmentation and skill mismatches, as mance, i.e. the scores in literacy, numer-
86
Research suggests that the availabil- potential key explanatory variables ( ). acy, and problem-solving. Tables 3 to 5
90
ity of human capital does not gener- show the results of the regressions ( )
ate benefits, notably economic ones, if Skill proficiency is higher among those with the respective coefficients together
81
it is left idle or under-utilised ( ) with who are active on the labour market — a with the standard error of estimation.
the quality of the national institutional finding which may be part of a vicious The higher the coefficient relative to the
and policy frameworks being important circle: the inactive part of the workforce standard error, the greater its statistical
82
in this respect ( ). Examples of poor suffers from skill depreciation and has significance as indicated by the stars in
human capital usage are reflected in evi - lower participation in education and the respective third column ( ).
91
dence of outcomes such as high rates of training, further worsening their pros-
unemployment and inactivity, especially pects to find a job. Such a result reveals Results for the socio-demographic controls:
of women, migrants and young people; that human capital capacity extends well
high levels of (early) retirement; part- beyond the ‘stock’ accumulated during • Age has a clear, negative impact
time work or skill-mismatch. On the other formal education and develops through across all three disciplines. The older
hand, institutions and policies such as: use at work. people are, the poorer they perform in
active labour market policies; financial literacy, numeracy and problem solv-
incentives to work through tax and social 3.3.1. Work intensity, ing. Though cohort effects may play a
welfare systems; retirement policy of the use of skills and skills role, this is an unsustainable situation
increasing retirement age and extend- performances in view of workforce ageing, which
ing working life, can all improve the uti- calls for stronger investment in their
lisation of human capital and thereby The phenomenon of skill usage and its work-related qualifications and skills.
indirectly support further investment in impact on maintaining human capital
human capital. can be investigated through a series of • Women score less favourably in
regression analyses using PIAAC micro- all disciplines.
Moreover, use of skills in and out- data. We build a new variable, which
side work is the best way to maintain takes into account the full working his- • As expected, higher educational
83
and even increase them ( ), as being tory of individuals. ‘Work intensity’ can be attainment leads to better scores in
employed generally corresponds to bet- proxied as the total number of years a all disciplines.
ter skills (see Chart A.1 in Annex). Recent person is paid to work, relative to his or
87
PIAAC data allow us to shed more light her age ( ). The so-defined ‘work inten- • Being foreign-born strongly reduces
on the importance of using skills at work. sity’ variable has been classified into the chance of achieving high scores
The data not only show that the potential quintiles ( ). We include two variables in all disciplines.
88
of highly skilled adults is not exploited reflecting the use at work of those skills
84
to the same extent in all countries ( ), which may be particularly relevant: the Of particular relevance for our analysis
but also stresses the role of a person’s frequency of ‘solving complex problems’ are the impact of work intensity and
89
employment status on skill usage and (of any nature) and ICT experience ( ). the use of skills. A number of results
maintenance ( ). We then control for a number of core stand out.
85
socio-demographic variables: educa-
Good utilisation of human capital cov- tion (highest educational attainment • Not being exposed to ICT in one’s
ers a broad range of problems and achieved), age, gender, and ‘foreign working environment strongly reduces
born’ — a dummy variable reflecting the scores in all three disciplines. The
( ) Knowledge and skill are workers’ capabilities where the respondent was born. same is true for a low frequency of
81
for performing various tasks, and they can ‘solving complex problems’ at work.
be used differently. Therefore Acemoglu and
Autor (2011) argue for adding additional ( ) European Commission publishes extensively
86
variables in the models to distinguish about activation problems and policies. See • The longer someone has been working
between skills (availability) and their use e.g. European Commission (2014g, 2012c,
in order to better understand the labour 2012d, 2012e). for pay, the higher their relative per-
market trends, and the impact of technology 87 formance in numeracy, literacy and,
on employment and earnings. Also, within ( ) A more accurate definition of ‘work intensity’
the companies, the value of its human would have been the number of years one has to a lesser extent, problem solving.
capital depends on its potential to contribute worked for pay, relative to the time elapsed
to the competitive advantage of the firm. since finishing formal education. However, as
See Lepak and Snell (1999 in Baron (2011). many people start working long before they
( ) See review of human capital policies in the finish formal education, we dropped that idea 90
82
because so-defined work intensity would often
EU in Heckman and Jacobs (2009). have exceeded 100 %. We have no information ( ) The number of valid cases per country
83
( ) Reder (1994). on the work history after finishing education. involved may be too small (ranging from
around 2.500 to around 6.000). The last
84
( ) In CZ, PL, NL and Flanders (BE) highly skilled ( ) Dividing the population into five equal classes column shows the international average for
88
individuals who are out of the labour force with respect to people’s ‘work intensity’: countries where the results are most reliable,
represented more than 2 % of the total adult Class 1: Work intensity (WI) > 60 %; due to a higher number of observations.
population and in FI the share was close to 4 % class 2: 48.15 % < WI ≤ 60 %; ( ) Third column: * and ** refer to the coefficient if
91
while the highest share of inactive highly skilled class 3: 33.33 % < WI ≤ 48.15 %; it is (in absolute terms) greater than 1.96 and
adults is in CZ, SK, IT, and PL (more than 20 %). class 4: 16.67 % < WI ≤ 33.33 %; 2.576 times the standard error, resp. As the true
85
( ) Various studies had already found a link between class 5: WI ≤ 16.67 %. coefficient then lies in the middle a confidential
national levels of educational attainment among ( ) The ICT experience variable is negatively interval greater than 95 % and 99 %, resp., the
89
EU Member States and the level of workforce expressed, i.e. equal to 2 if there is no estimated sign of the coefficient is significant at
training (Badescu et al., 2011). experience, otherwise it is 1. minimum 5 % and 1 %, resp.
118
health is thus an important precondition for policies, which cannot all be presented The aim of the regression is to find evi-
maintaining the available human capital. and analysed here. We decided to focus dence for the human capital deprecia-
on the importance of work intensity on tion phenomenon, such as the impact of
3.3. Using human capital skills performance, on women and labour skills and work intensity on PIAAC perfor-
segmentation and skill mismatches, as mance, i.e. the scores in literacy, numer-
86
Research suggests that the availabil- potential key explanatory variables ( ). acy, and problem-solving. Tables 3 to 5
90
ity of human capital does not gener- show the results of the regressions ( )
ate benefits, notably economic ones, if Skill proficiency is higher among those with the respective coefficients together
81
it is left idle or under-utilised ( ) with who are active on the labour market — a with the standard error of estimation.
the quality of the national institutional finding which may be part of a vicious The higher the coefficient relative to the
and policy frameworks being important circle: the inactive part of the workforce standard error, the greater its statistical
82
in this respect ( ). Examples of poor suffers from skill depreciation and has significance as indicated by the stars in
human capital usage are reflected in evi - lower participation in education and the respective third column ( ).
91
dence of outcomes such as high rates of training, further worsening their pros-
unemployment and inactivity, especially pects to find a job. Such a result reveals Results for the socio-demographic controls:
of women, migrants and young people; that human capital capacity extends well
high levels of (early) retirement; part- beyond the ‘stock’ accumulated during • Age has a clear, negative impact
time work or skill-mismatch. On the other formal education and develops through across all three disciplines. The older
hand, institutions and policies such as: use at work. people are, the poorer they perform in
active labour market policies; financial literacy, numeracy and problem solv-
incentives to work through tax and social 3.3.1. Work intensity, ing. Though cohort effects may play a
welfare systems; retirement policy of the use of skills and skills role, this is an unsustainable situation
increasing retirement age and extend- performances in view of workforce ageing, which
ing working life, can all improve the uti- calls for stronger investment in their
lisation of human capital and thereby The phenomenon of skill usage and its work-related qualifications and skills.
indirectly support further investment in impact on maintaining human capital
human capital. can be investigated through a series of • Women score less favourably in
regression analyses using PIAAC micro- all disciplines.
Moreover, use of skills in and out- data. We build a new variable, which
side work is the best way to maintain takes into account the full working his- • As expected, higher educational
83
and even increase them ( ), as being tory of individuals. ‘Work intensity’ can be attainment leads to better scores in
employed generally corresponds to bet- proxied as the total number of years a all disciplines.
ter skills (see Chart A.1 in Annex). Recent person is paid to work, relative to his or
87
PIAAC data allow us to shed more light her age ( ). The so-defined ‘work inten- • Being foreign-born strongly reduces
on the importance of using skills at work. sity’ variable has been classified into the chance of achieving high scores
The data not only show that the potential quintiles ( ). We include two variables in all disciplines.
88
of highly skilled adults is not exploited reflecting the use at work of those skills
84
to the same extent in all countries ( ), which may be particularly relevant: the Of particular relevance for our analysis
but also stresses the role of a person’s frequency of ‘solving complex problems’ are the impact of work intensity and
89
employment status on skill usage and (of any nature) and ICT experience ( ). the use of skills. A number of results
maintenance ( ). We then control for a number of core stand out.
85
socio-demographic variables: educa-
Good utilisation of human capital cov- tion (highest educational attainment • Not being exposed to ICT in one’s
ers a broad range of problems and achieved), age, gender, and ‘foreign working environment strongly reduces
born’ — a dummy variable reflecting the scores in all three disciplines. The
( ) Knowledge and skill are workers’ capabilities where the respondent was born. same is true for a low frequency of
81
for performing various tasks, and they can ‘solving complex problems’ at work.
be used differently. Therefore Acemoglu and
Autor (2011) argue for adding additional ( ) European Commission publishes extensively
86
variables in the models to distinguish about activation problems and policies. See • The longer someone has been working
between skills (availability) and their use e.g. European Commission (2014g, 2012c,
in order to better understand the labour 2012d, 2012e). for pay, the higher their relative per-
market trends, and the impact of technology 87 formance in numeracy, literacy and,
on employment and earnings. Also, within ( ) A more accurate definition of ‘work intensity’
the companies, the value of its human would have been the number of years one has to a lesser extent, problem solving.
capital depends on its potential to contribute worked for pay, relative to the time elapsed
to the competitive advantage of the firm. since finishing formal education. However, as
See Lepak and Snell (1999 in Baron (2011). many people start working long before they
( ) See review of human capital policies in the finish formal education, we dropped that idea 90
82
because so-defined work intensity would often
EU in Heckman and Jacobs (2009). have exceeded 100 %. We have no information ( ) The number of valid cases per country
83
( ) Reder (1994). on the work history after finishing education. involved may be too small (ranging from
around 2.500 to around 6.000). The last
84
( ) In CZ, PL, NL and Flanders (BE) highly skilled ( ) Dividing the population into five equal classes column shows the international average for
88
individuals who are out of the labour force with respect to people’s ‘work intensity’: countries where the results are most reliable,
represented more than 2 % of the total adult Class 1: Work intensity (WI) > 60 %; due to a higher number of observations.
population and in FI the share was close to 4 % class 2: 48.15 % < WI ≤ 60 %; ( ) Third column: * and ** refer to the coefficient if
91
while the highest share of inactive highly skilled class 3: 33.33 % < WI ≤ 48.15 %; it is (in absolute terms) greater than 1.96 and
adults is in CZ, SK, IT, and PL (more than 20 %). class 4: 16.67 % < WI ≤ 33.33 %; 2.576 times the standard error, resp. As the true
85
( ) Various studies had already found a link between class 5: WI ≤ 16.67 %. coefficient then lies in the middle a confidential
national levels of educational attainment among ( ) The ICT experience variable is negatively interval greater than 95 % and 99 %, resp., the
89
EU Member States and the level of workforce expressed, i.e. equal to 2 if there is no estimated sign of the coefficient is significant at
training (Badescu et al., 2011). experience, otherwise it is 1. minimum 5 % and 1 %, resp.
118
