Te c h n o l o g y o r u p s k i l l i n g ?
Tr e n d s i n t h e t a s k
composition of jobs
in Central and Eastern Europe
Piotr Lewandowski
R o m a K e i s t e r, Wo j c i e c h H a r d y
Are the well-known task evolution patterns also present
in countries at a lower development level?
35%
ICT Stock per worker, relative to the US
30%
25%
20%
15%
10%
5%
0%
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Czech Republic
Hungary
Poland
Romania
Slovakia
Slovenia
Source: own calculations on Eden & Gaggl (2015) data on ICT capital stock and Eurostat data on employment
Post-transition labour markets were changing rapidly
• Macroeconomic convergence from middle- to high-income status
• Structural shifts
• Agriculture ↘↘
• Manufacturing ↘ but still quite large
• Services ↗
• Educational boom
• Primary and vocational ↘
• Tertiary ↗
SKILLS
„worker’s abilities for performing
various tasks”
How do we measure task contents of jobs?
EU-LFS data for 10 CEE
countries, 1998-2013,
3-digit ISCO occupations
How do we measure task contents of jobs?
EU-LFS data for 10 CEE
countries, 1998-2013,
3-digit ISCO occupations
O*NET data – editions
2003 and 2014
How do we measure task contents of jobs?
EU-LFS data for 10 CEE
countries, 1998-2013,
3-digit ISCO occupations
O*NET data – editions
2003 and 2014
5 annual country-level
task content measures
Autor & Acemoglu (2011)
Familiar picture except for increasing intensity of routine cognitive tasks
20
Average for CEE10
15
Non-routine cognitive
analytical
10
Non-routine cognitive
personal
5
0
-5
Routine cognitive
Routine manual
-10
-15
-20
Non-routine manual
physical
Non-routine cognitive tasks ↗ and manual tasks ↘ across the CEE
Routine manual
Non-routine cognitive analytical
Slovenia
Slovakia
Czech Rep.
30
15
0
-15
-30
Romania
Czech Rep.
Croatia
Estonia
Hungary
Poland
Latvia
Lithuania
30
Slovenia
15
0
Slovakia
-15
-30
Romania
Croatia
Estonia
Hungary
Poland
Latvia
Lithuania
But changes in routine cognitive tasks were heterogenous
Czech Rep.
30
Slovenia
15
Croatia
0
Slovakia
Estonia
-15
-30
Romania
Hungary
Poland
Latvia
Lithuania
What was the correlation between education, technology,
and aggregate task content changes in CEE
• Regression of tasks contents on educational structure and R&D spending
• Panel fixed-effects, 1998-2012, 10 CEE countries
• R&D as a proxy for ICT
log ICT stock per worker
R&D spending / GDP
Constant
Observations
Coefficient
Std. err.
118.9**
39.55
-0.63
0.36
84 (6 countries)
R2 within
0.17
R2 between Correlation
0.59
0.73
Workforce upskilling associated with growing non-routine cognitive
and falling manual tasks
Non-routine
cognitive
analytical
Non-routine
cognitive
personal
Routine
cognitive
Routine
manual
Non-routine
manual
physical
1.49***
0.82***
0.74
-1.18***
-1.74***
0.71***
0.01
0.60
-0.33
-1.17*
R&D spending / GDP
3.73*
3.04*
-4.71
-3.01*
-1.81
R2 (between/within)
0.01/0.84
0.01/0.75
0.03/0.20
0.00/0.80
0.04/0.82
Share of persons with
tertiary education
attained
Share of persons with
secondary education
attained
Shift-share decomposition of tasks content changes
• Structural change
98
𝑡𝑖,𝑗,03
ℎ𝑗13 − ℎ𝑗98 ,
∀𝑖∈𝑇 𝐵𝑆𝑖 =
𝑗∈𝑆
• Educational change
98
𝑡𝑖,𝑗,𝑘,03
∀𝑖∈𝑇 𝐵𝐸𝑖 =
𝑗∈𝑆 𝑘∈𝐸
13
ℎ𝑗,𝑘
98
ℎ𝑗,𝑘
13 − 98
ℎ𝑗
ℎ𝑗
ℎ𝑗98 ,
• Occupational change
13
98
98
(𝑡𝑖,𝑗,𝑘,14
− 𝑡𝑖,𝑗,𝑘,03
) ℎ𝑗,𝑘
∀𝑖∈𝑇 𝑂𝐶𝑖 =
𝑗∈𝑆 𝑘∈𝐸
• Interaction (equation in the paper)
Structural and occupational changes drove the growth of routine
cognitive tasks
Estonia
Poland
Slovenia
20
15
10
5
0
Interaction
Occupational effect
Educational effect
-5
-10
-15
-20
-25
Structural effect
Total observed change
Agriculture contributed most of routine cognitive tasks’ growth
in countries where RC grew most
Contribution of sectors to RC change
30
25
20
15
10
5
0
-5
-10
-15
-20
-25
Slovenia
Hungary
A
Czechia
C+D+E
Slovakia
F+G+H
Croatia
I+J+K
L+O
Estonia
M+N
Poland
Latvia
Total change
Romania
Educational boom fuelled the growth of non-routine analytical tasks
Estonia
Poland
Slovenia
30
25
20
Interaction
15
Occupational effect
10
Educational effect
5
Structural effect
0
Total observed change
-5
-10
-15
Educational boom fuelled the growth of non-routine analytical tasks
and the fall of routine manual tasks
Estonia
Poland
Slovenia
15
10
Interaction
5
Occupational effect
0
Educational effect
-5
-10
-15
-20
-25
-30
Structural effect
Total observed change
So far no de-routinisation in CEE
• Growth of non-routine cognitive tasks and decline of manual tasks
• Prevailing increase in routine cognitive tasks
• Changes largely attributed to workforce upskilling
• Structural change most important for the rise of routine cognitive
Thank you!
Piotr Lewandowski
[email protected]
www.ibs.org.pl
@ibs_warsaw