While Europe continues to see the ramifications of the crisis and is threatened by the exit of Greece from its fold, a bigger and more long-standing crisis has been brewing in the background since well before the crisis.
The situation of inequality from the youth perspective has received relatively little academic research. However, our research shows that in light of technological change and with the advent of increased automation, not only is the definition of work undergoing a change, but is the employment opportunity landscape for young people in Europe.
Using Agglomerative Hierarchical Clustering Techniques, we compare the situation youth employability and inequality for 28 EU countries, before and after the crisis. Our analysis shows that with technology , especially ICT, making a bigger impact on the definition of tasks and skills, the youth of Europe can no longer simply rely on education as a way of escaping inequality. The youth today require more a more entrepreneurially nourishing landscape coupled with an infrastructure that allows for information to grow in order to have a fighting chance to overcome inequality and define a new sense of work in today's digital age.
Inequality, Technology & Job Polarization of the Youth Labor Market in Europe.
1. Inequality, Technology
and Job Polarization of
the Youth Labor Market
in Europe.
Kariappa Bheemaiah, Mark Smith
Grenoble Ecole de Management.
2. Policies and Literature
Markets have improved since the crisis, but the pace of economic
growth and employment for young people has stagnated in Europe
(Cedefop, 2010) .
One of the challenges of the Europe 2020 Strategy is to solve the
problem of social exclusion
Other factors Globalization + Offshoring + Technology
USA: Autor Levy Murnane (2003), Autor & Acemoglu (2010)
EU/OECD: Goos et all (2011(IMF study)) found that ICT was
having the biggest effect.
YOUTH INEQUALITY ICT
▲Opportunities
▲Income Levels
▲Social Security Systems
▲Economic Mobility
▼Poverty Levels
▼Inequality- education,
jobs, earnings, policies
and other factors.
3. LIT. Review
Technological unemployment - Tinbergen’s canonical education-race model (‘75)
SBTC: Berman, Bound and Machin (‘98)
Limitation of Tinbergen’s model’s lack of a concrete definition for ‘tasks’
ALM Routinization Hypothesis (2003) & the Polarized work environment (2010)
Cognitive non-routine tasks
Cognitive routine tasks
Manual routine tasks
Manual non-routine tasks
Low-Skill
High-Skill
Medium-Skill
Low-Skill
High-Skill
Medium-Skill
SKILL-BIASED
TECHNOLOGICAL CHANGE
SKILL-SUBSTITUTING
TECHNOLOGICAL CHANGE
4. Indicators: 9 ICT related indicators- (Source: Eurostats):
Method: Agglomerative Hierarchical Clustering (Ward’s Method).
Methodology
Centroid method to trace the behavior of the clusters with respect to each
variable.
T 1 Gross value added by Information and Communication Industry.
T 2 Employment in technology sectors at the national level.
T 3 Employment in Information and communication (ICT) industry.
T 4 Sales of Goods and Services via Internet
T 5 Purchase of Goods and Services via internet
T 6 Employment in knowledge-intensive activities.
T 7 Total High-tech Imports as % of total trade.
T 8 Total high-tech Exports as % of total trade.
T 9 GERD in Business Enterprise, Government and Higher Education Sectors.
Low Tech Cluster BG EE EL ES HR IT CY LV LT PL PT RO SI SK
Medium Tech Cluster BE CZ IE FR LU HU MT AT
High Tech Cluster DK DE NL FI SE UK
2007 - 2013
5. Methodology
T 1 T 2 T 3 T 4 T 5 T 6 T 7 T 8 T 9
4 5 3
7
28
43
18
21
2
11
31
8
6
5 5 4
17
53
49
15
15
2
0
10
20
30
40
50
60
70
Profile plot of Technology variables by Cluster (2007)
Cluster 1 Cluster 2 Cluster3
T 1 T 2 T 3 T 4 T 5 T 6 T 7 T 8 T 9
6 5 3
18
44
37
14 15
27
24
31
9 8
5 5 4
26
70
42
14 15
2
0
10
20
30
40
50
60
70
80
90 Profile plot of Technology variables by Cluster (2013)
Cluster 1 Cluster 2 Cluster 3
6. -4.1 -5.2
-3
1.5
-1.5
4.7
13.5
10.6
16.2
-10
-5
0
5
10
15
20
Total
Male
Female
Total
Male
Female
Total
Male
Female
15-24 25-54 55- 64
EU 28: 2014-2002
Overall Employment levels in EU28 have declined =+3.9 PP (2007-2013)
Youth Inequality & Employment.
-4.7 -5.8
-3.6
-2 -3.5
-0.3
6.5 4.2
8.6
-10
-5
0
5
10
15
20
Total
Male
Female
Total
Male
Female
Total
Male
Female
15-24 25-54 55- 64
EU 28: 2014-2008
All Clusters: Young adults Level 3-6 Edu. HIGHER NEETs than Level 0-2 peers
ICT replacing analytical tasks of Medium/High Skilled Workers.
15-24 Employment rate NEETS
Low Tech Cluster
High Tech Cluster
7. ICT replacing Medium/High Skilled Workers
Youth more adversely effected than Older Population
27%
-21%
-31%
-7%
-21%
-33% -33%-40%
-30%
-20%
-10%
0%
10%
20%
30%
40%
Professionals Technician & asso.
professional
Clerical support Service and sales Skilled agricultural,
forestry & fishery
Craft and related
trade
Plant & machine
operator
High Skill Jobs Mid- Level Skill Jobs Low-Skill Jobs
15- 24 yr. olds : % change in sectorial employment - EU 27
31%
-4%
-8%
27%
-8%
-15% -15%
-40%
-30%
-20%
-10%
0%
10%
20%
30%
40%
Professionals Technician & asso.
professional
Clerical support Service and sales Skilled agricultural,
forestry & fishery
Craft and related
trade
Plant & machine
operator
Mid- Level Skill Jobs Low-Skill Jobs
25 - 64 yr. olds: % change in sectorial employment - EU 27
Growth in Know. Based Services:
Optimistic Figures for High Tech Cluster (especially FI & NL)
But not so in Low Tech Cluster (especially EL, ES, RO)
8. Cuts and freezes of minimum wage in 21 EU Countries (mostly Low Tech Cluster)
Real hourly minimum wages have reduced
Minimum wages closely linked to economic development
4.6
2.9
0.0
0.8
-1
0
1
2
3
4
5
18 to 24 years 25 to 54 years 55 to 64 years 65 years and above
EU 27 - Average Change in risk of poverty rate according to age (2007-2013)
Youth Inequality & Earnings.
End 2013- Average In-Work-At-risk-of-poverty-rate
18-24 years= 11.4%
25-54 years = 8.8%
55-64 years = 8.5%
9. 17.7
10.1
30.2
9.3
5.6
16.1
11.1
8.5
27.8
16.3 15.5
17.6
15.3
18.1
27.8
22.6 23.6
10.6
7.1
13.8
27.7
14.8
44.9
12.8
16.5
6
13 12.8
-10
0
10
20
30
40
50 EU27
BE
BG
CZ
DK
DE
EE
IE
EL
ES
FR
IT
CY
LV
LT
LU
HU
MT
NL
AT
PL
PT
RO
SI
SK
FI
SE
UK
PP change from 2007 to 2013 for In-Work-at-risk-of-poverty, based on
level of education
(0-2) (3 & 4) (5 & 6)
(0-2) PP change since 2008 (3 & 4) PP change since 2008 (5 & 6) PP change since 2008
Youth Inequality & Education+Skills.
Lower educated face greater inequality risks, especially in Low & Medium Tech
Countries.
Young People & Young Adults - 1.1 PP increase of NEET rates for low skilled
7 PP for the medium/high skilled.
10. Youth Inequality & Education+Skills.
10
15
20
25
30
35
40
45
EU27
EUMales
EUFemales
BE
BG
CZ
DK
DE
EE
IE
ES
FR
HR
IT
CY
LV
LT
HU
MT
NL
AT
PL
PT
RO
SI
SK
FI
SE
UK
Level 5-6 Students enrolled in STEM as % of all students
2007 2012 2 per. Mov. Avg. (2007) 2 per. Mov. Avg. (2012)
More young adults ( especially in Low Tech Cluster ) participating in Edu. Levels 5 to 8.
But this is not reflected in employment figures.
….Impact of Education in reducing inequality for the youth?
11. NRI : Exploring ICT +(employment, entrepreneurship & education).
2013 NRI Clusters = 71% Commonality with 2013 Tech Cluster Grouping
Youth Inequality, ICT & NRI
Politicaland
Regulatory
Businessand
Innovation
9 9
Infrastructure
4 Affordability3
Skills
4
Individual
7
Business
6
Government
3
Economic
Impacts
4
Environment
Social
Impacts
3
Economic
Impacts
UsageReadiness
Social
Impacts
Impact
Low Tech Cluster
Low NRI Scores.
Higher youth unemployment, lower earnings (stagnant min.wages),
reduced spending on ALMPs (except PL, SI & SK), qualified NEETs
Med. Tech Cluster
Competitive NRI scores.
Adversely impacted by policies - ALMPs (entrepreneurship) vs PLMPs.
High Tech Cluster
High NRI scores.
Growth of Emp. in Know. Services, Low unemployment levels
Apprenticeships Skills development, flow of knowhow and tacit
knowledge.
12. Conclusions:
Better ICT Better Information Growth
Greater Flow of Information Increases employment, reactivity of
policies, adapt to new tech, more entrepreneurship, greater mobility, spread
of education…gives youth a better chance to reduce the inequality.
Growth of Economies Growth of Flow of Information & ICT
Next Steps:
Youth Focused Entrepreneurship & Innovation policies that allows
Information to Grow (leveraging ICT).
Development of diverse skills in employment to complement formal
schooling (apprenticeship programs).
More granular research in measuring the impact of ICT at micro-levels
ICT and Social Infrastructure development as a priority.
Conclusions, Next steps
Hinweis der Redaktion
Slide 2: POLICIES and LIT
Since crisis most EU countries have begun to recover, but optimism can’t be shared for youth of EU.
And this is especially important as one of the goals of the EU 2020 strategy focuses on inequality or social exclusion and how to reduce it.
Inequality today is largely been focused on education, jobs, earnings, policies and other factors. Spend some time on other factors: We have found that the 3 main avenues: Globalization + Offshoring + Technology.
More recent work has established that of these 3 factors, Technology and especially ICT is the main contributor to changes in employment, earnings and hence inequality.
Work by Autor & Acemoglu and EU work Goos and book by MIT professors Erik Brynjolfsson & Andrew McAfee.
Our paper is in effect looking at the current situation of youth inequality, especially in the period following the crisis, and what role ICT has and is playing in this regard.
Slide 3: LIT REVIEW
The concept of technological unemployment and even understanding ICT’s impact on employment is not a novel new subject
Previous work- Tinbergen and Skill-Based Technological change. Good stuff, but no emphasis on Tasks.
Tasks important because modern technology today replaces the tasks performed by workers.
So to analyse in 2003, Autor, Levy and Murnane came up with the routinization hypothesis in which they classified jobs in different sectors of the economy into 4 skills groups.
Cognitive non-routine tasks (High Skills)
Cognitive routine tasks (Middle Skills)
Manual routine tasks (Middle Skills)
Manual non-routine tasks (Low Skills)
They looked at US data over a 30 year time period, right up to 2010, and found that over time the employment in jobs requiring high skills & low skills was going up, but the jobs which required medium-skill set was dropping, creating this U-shaped polarized work environment.
Goos, Manning and Salomons, looked at data from 1980 to 2004, & for 16 EU states and was focused on the populace as a whole and saw the same polarized curve in the European context. With technological change, skill adaptability and cognitive flexibility is playing an increasingly influential role in inequality.
As a result, DESKILLING--------------no longer SBTC but SSTC.
So having looked at this, we noticed that there were 2 gaps that needed to be filled:
What was the inequality situation for young people in Europe,
What is the impact of ICT in this respect?
Slide 4: METHODOLOGY
With these goals as objectives, we also wanted to see how we could compare EU countries. So we considered 9 Technology variables (selection based on emp. Research).
Ward’s method Country clusters.
Euclidean Distances
Maximum homogeneity within Cluster
Maximum heterogeneity between clusters.
Low tech, Medium and High Tech. Migration 2007 to 2013.
Having identified the grouping, which variable is responsible for the change? CENTROID METHOD.
GERD: Gross domestic expenditure on R&D
SLIDE 5: METHODOLOGY
THESE VARIABLES were related to
Trade,
Sectors of manufacturing industries
Employment variables
And so having identified these arenas, we focused on these subjects from the perspective of the youth and traced the changes occurring over this 2007-2013 time period for each cluster.
Slide 6: EMPLOYMENT
The inequality was what caused the crisis. (Goda, 2013) (OECD, 2011).
Employment in the EU has actually gone down by 3.9PP over this period overall.
But the youth are the worst affected and this is even from before the crisis.
We focused on the 15-24 age group bracket, we found that
Countries in the High Tech Cluster had highest employment rates
Countires in the Low Tech Cluster had the lowest emp. rates as well, namely Greece, Hungary, Bulgaria and Spain.
For Spain this extends even into the 25-54 age group and in terms of employment, it has been the most adversely effected country in this period.
NEETS :
Countries in Low Tech Cluster, also had the highest NEET rates
High Tech Cluster had lower NEET rates.
SUPRISINGLY, the population of young people & young adults with-
At least a Levels 3-6 education level had higher NEET rates than their peers with lower (Levels 0-2) educational attainments.
This result draws strong parallels with the polarization findings of Goos and Manning (2007) and the ALM (2003) hypothesis which found that technology led inequality, had little effect was seen on low-skill jobs, as ICT began replacing the analytical tasks of medium/high-skill workers.
Slide 7: SECTORS
Looking at Skills replacement begets the need to look at how employment in sectors is changing.
We selected 7 sectors and on the basis of the level of Skill normally required to perform jobs in these sectors.
First we see that once again the younger population has been more effected by the older
We can also notice the polarization effect although we don’t see the same curves but this can be because of the fact that we only found consistent data for these 7 sectors.
In the Middle Skills sector, we did find growth, but:
It is in the Tech intensive Services industries
It is been much more beneficial to the older population
3.High Tech Clusters were found to have absorbed a larger share of this growth in the services sector.
Slide 8: EARNINGS
This dichotomy on the basis of Employment has also had an effect on EARNINGS
Youth has been seen a 4.6PP increase in poverty rate b/w 2007-2013.
When we compared the median equivalised net income in terms of Purchasing Power Standards and found that
Countries in Low Tech Cluster had the lowest increases in earnings with Ireland and Greece as the worst performers
Also min wage had either been reduced or stagnated for 21 EU countries and this was concerning as empirical research has shown that in most low income countries, min wage is closely linked to economic development.
Youth labour market in Europe has been in a state of decline since the turn of the century, with market forces more favourable towards the prime age and older worker segments rather than the youth.
Furthermore high gender biased NEET rates towards young women (Degryse, 2013) especially in UK the UK which had the second highest figures for women between the ages of 16 to 24. These findings also coincide with the high NEET rates found for women in the UK belonging to the same age group, when compared to other countries of Cluster 3.
restrictive wage policies as part of the austerity measures implemented in 21 EU countries where national minimum wage exists (Schulten, 2014) has resulted in cuts and freezes of minimum wages in countries belonging to Cluster 2, such as Greece, Ireland, Portugal, Romania , Latvia and Spain. As a result of these policies, real hourly minimum wages have reduced in a range of countries where as a rule; minimum wages was closely linked to economic development, affecting even advanced economies like the UK (Cluster 1)
SLIDE 9+ 10: The Education slides
The last part of our institutional analysis looked at the effect of Education on Earnings as we wanted to see the impact that it was having.
We found that firstly, the lower educated did face the greatest risk of poverty especially in the Low the Tech Cluster.
BUT! We also saw that the NEET rates for low skilled workers was lower than that of middle skilled workers in Low tech Clusters in spite of the fact that there were more students enrolled in STEM education at the higher levels in Low Tech Country Clusters
SO THIS QUESTIONS THE ROLE OF EDUCATION
Education by itself does not serve as a tool for breaking away from unemployment and income inequality. With forecasts predicting the increased impact of ICT technologies in work, education and skill development, changes which will help the youth adapt to technological change, whilst aiding in reducing inequality, need to be addressed from an institutional level across all EU member states.
SLIDE 11: NRI
Having analyzed the way these Clusters were performing in terms of these macro socio-economic parameters from the perspective of a polarized market, we were still facing a quandary as we needed to link the level of a country’s ICT capability with its socio-economic performance. This was after all one of the primary objectives.
To complete this last blank we needed to have some kind of a composite reference indicator that was a measure of ICT as well as the effect of ICT on employment, earnings, social welfare and innovation.
Luckily we did not have to construct this kind of an indicator as it already exists.
The WEF along with Cornell and INSEAD have created an indicator which measures exactly these aspects and is known as the NETWORK READINESS INDEX, which essential giving a score to a country based on its ICT readiness and competitiveness.
The NRI measures 4 sub-aspects of every country’s ICT standing:
The technological Environment
The state of readiness
The level of usage
The Impact of ICT
To do so, each aspect is measured using selected indicators. So the Environment is measured in terms of its Political and Regulatory environment and its Business environment which is in turn measured by 9 indicators, each, and so on….
The data for these indicators is collected from Eurostat, OECD, ITU and the WEF’s own survey (1-7 Likert Scale). So this is the reason why we wanted to select it since a lot of our data is from the same sources and also because of the gamut of the indicators that the NRI considers.
Finally by aggregating the individual indicators by using the arithmetic mean in a pyramid like continuum, each country is given a NRI score on a 1 to 7 scale, which aligns them with the Survey’s results.
Taking these scores for each country fir the final year of our analysis (2013), we performed another AHC cluster and found that for we got almost the same country groupings all over again. There were a few discrepancies, but overall 71% of the countries found themselves in the same grouping as before.
Slide 12: CONCLUSION
We started our research with the objective of finding a relationship between the situation of youth inequality and ICT in Europe. And to do so we wanted to compare different countries simultaneously. Our cluster analysis helped us achieve these goals and shows that countries which are more advanced in terms of ICT preparedness are also those which have lower levels of inequality. Increased Social Capital - the ability of people to connect.
This is primarily because greater ICT capabilities leads to greater ease of information flow which leads to the creation of a flux within the socio-economic context that facilitates the spread of ideas, creates new opportunities more easily, allows for greater transparency, improved access to educational resources and makes the population more easily adaptable to technological changes.
Policy makers thus need to encourage development of diverse skills for workers whilst in employment to complement formal schooling.
ICTs can be used at the community level in order to help localised or segregated communities with job retraining programs for unemployed workers.
But most importantly, EU member states need to understand that the growth of economies is a consequence of the growth of information and that prosperity is directly related to making information grow.