Patrick Daru (ILO) and Eduarda Castel-Branco (ETF)

Geneva, 11/05/2017

THE IM(PERFECT) MATCH – ILO INTERNATIONAL CONFERENCE

REGIONAL VIEW: ARAB STATES AND CENTRAL ASIA

DO SKILLS MATTER IN THE MENA REGION?

2

THE SKILLS MISMATCH STORY IN THE ARAB STATES

USUAL STORYLINE

Unfilled vacancies in context of

unemployment

Education and skills programmes

not aligned with the market

Short term training programme to

compensate for the failures of

education system

IN FACT

Lack of datasets to analyze skills

mismatch

Sticky wages that do not allow market to reach equilibrium

Segmented markets: migrants as a

cheaper option

SKILLS MISMATCH NOT ALWAYS A PRIORITY FOR EMPLOYERS

24.5

18.0

36.8

50.1

34.2

9.5

15.3

30.9

24.4

0.0 20.0 40.0 60.0

World

Average

MENA

Average

Algeria

Egypt

Iraq

Jordan

Lebanon

Morocco

Yemen

Based on: Enterprise Surveys

(http://www.enterprisesurveys.org), The World

Bank

Latest surveys available, 2015

Percentage of Firms

Identifying Inadequately

Educated Workforce

as a Major Constraint

in selected MENA Countries (%)

ON THE EMPLOYERS’ SIDE

Employers complain about skills mismatch (not always), and do not train

• 16% Arab Firms train new hires against 36% globally (WB Enterprise Survey)

Skills are not adequately valued

• Wage differentials between most and least educated are the lowest in the world

Short term business vision

• Benefit from laboursurplus in a context of low skilled labourintensive production;

• Longer term investment in business and skills difficult in the context of fragility

Lack of organization of employers

• Impact capacity to structure voice on skills required

• does not prevent the possible poaching by competitors

QUALIFICATION MISMATCH IS HIGH

Country

Latest

Year

Available

Source% Over-

qualified

% Under-

qualified

Total %

qualificati

on

mismatch

Bahrain 2004 Labour Force Survey 13.15 40.03 53.18

Jordan 2013Employment and Unemployment

Survey10.6 12.5

23.1

Morocco 2012 National Employment Survey 7.7 40.9 48.6

oPt 2012 School to Work Transition Survey 13.5 46.4 59.9

Qatar 2012 Labour Force Survey 14.1 38.09 52.19

Saudi

Arabia2013 Labour Force Survey 24.29 23.84

48.13

Yemen 2013-2014 Labour Force Survey 3.35 76.1283

YOUNG WORKERS PERCEPTION OF SKILLS MISMATCH

52.2%

34.2%

1.2%

12.4%

Egypt

Adequate Education and

Skills

Over qualified

Under qualified

Don't Know87.6%

8.2%4.1%

Jordan

Adequate Education and Skills

Over qualified

Under qualified

ILO: School to Work Transition Survey, 2012

FROM WORKERS / JOB SEEKERS PERSPECTIVE

WASTA – HIGHER ON LIST OF JOB

SEEKERS ISSUES (NOT OF

WORKERS)

WHAT SIGNALS? IN A CONTEXT OF

LACK OF TRUSTED CERTIFICATES

INFORMATION ASYMMETRIES – AND

CAREER GUIDANCE

LACK OF CHOICE > INADEQUATE

BEHAVIOR / SOFT SKILLS

“We take on education

we did not choose, that

do not match the market

demand, and for jobs

we will not get because

of Wasta”.

UNICEF Youth Consultation in

Jordan, April 2017

JORDAN: REFUGEE CRISIS RESPONSESKILLS AS ONE ELEMENT ONLY OF JOB MISMATCH

“Replacement” of migrants by

Syrian refugees requires a new

business model.

From “Refugees take

jobs” to “Refugees do not want to

work”

Feb. 2016: Access of

Syrian Refugee to Jordan

Labour Market

10

EASTERN EUROPE AND CENTRAL ASIA

ETF Position Paper (2012) adopted the following

definition of skill mismatch:

“…a broad term that encompasses various types of skill gaps and

imbalances such as over-education, under-education, over-qualification,

under-qualification, over-skilling, skill shortages and surpluses, skills

obsolescence and so forth. Hence skill mismatch can be both qualitative

and quantitative, thus referring to both situations where a person does not

meet the job requirements and where there is a shortage or surplus of

persons with a specific skill. Skills mismatch can be identified at the various

levels: of the individual, the enterprise, the sector or the economy. Several

different types of skill mismatch can coincide”.

11

1.SKILL MISMATCH

12

1.2 SKILL MISMATCH MEASUREMENT IN ETF WORK

Methodology Measures what Strengths/Weaknesse

s

Explored in…

Variance relative rates

(ER, UR)

Dispersion skills.

Magnitude.

Macro. Data avail. MOLD, KAZ, KYR,

Coefficient of variation Dispersion skills.

Magnitude

Macro. Data avail.

Proportion of

unemployed vs employed

Direction mismatch:

which educ levels in

shortage / excess

Macro. Data avail GEORGIA.

MOLD, KAZ, KYR,

Mismatch by occupation Ratio employed

occup/educ: over-,

under-qualificatio

Unemployed pop – not

considered. Data avail

MOLD

Other measures used in ETF analysis: Beveridge curve, relative

wages by educational levels

13

EASTERN EUROPE

ARMENIAAZERBAIJAN

BELARUSGEORGIAMOLDOVAUKRAINE

SOME FIGURES INCLUDE RUSSIAN FEDERATION

14

2. EDUCATIONAL ATTAINMENT POPULATION (2015)

high23%

low8%mediu

m69%

Armenia (15-75)

high22%

low13%mediu

m65%

Azerbaijan (15-64)-2013

high44%

low7%

medium49%

Ukraine (15-70)

high35%

low4%

medium

61%

Georgia (25-64)

Sources: DB Torino process 2016

15

EE: YOUTH UNEMPLOYMENT RATE AND PARTICIPATION IN VET (UPPER-SECONDARY LEVEL)

AM

AZ

GE

MD

RU

UA

1020

3040

You

th u

nem

ploy

me

nt r

ate

(%

)

10 20 30 40 50% of VET students in upper secondary education

0

10

20

30

40

50

60

Armenia Azerbaijan Georgia Republic ofMoldova

RussianFederation

Ukraine

Youth unemployment rate (15-24) and % VET students in upper sec education - 2014

VET stud % upper sec Youth UR (15-24)

16

EE: A) UNEMPLOYMENT RATE (+15; 15-24) – 2010, 2015 B) NEET RATE (15-24) – 2013, 2015

38.9

32.5

14.913.4

36.4

30.8

14.912.8

17.4

22.4

0

5

10

15

20

25

30

35

40

45

2010 2015 2010 2015 2010 2015 2010 2015 2011 2015 2010 2015

Armenia Azerbaijan Belarus Georgia Moldova Ukraine

Unemployment rate by sex (age group +15) and

youth unemployment rates (15-24), %

Total Male Female Youth UR

0

5

10

15

20

25

30

35

40

45

Tota

l

Ma

le

Fem

ale

Tota

l

Ma

le

Fem

ale

Tota

l

Ma

le

Fem

ale

Tota

l

Ma

le

Fem

ale

Armenia Georgia Republic ofMoldova

Ukraine

NEETs Rates (15-24) by sex (%) - 2013 and 2015

2015

2013

17

EE: SKILL GAPS (2013)

AM AZ BY GE MD RU UA

2013 6.4 0.5 17.9 9.9 31.2 7.5

0

5

10

15

20

25

30

35

Skill gap (2013)

Based: World Bank Enterprise Surveys

% firms identifying

and inadequately

educated Workforce

as a major constraint

18

EE SKILL MISMATCH: OVER-QUALIFICATION YOUTH

Source: ILO SWTS 2012-2013

21.527.5

23.2

11.66.6

8.9

66.9 65.9 67.9

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Armenia Moldova Ukraine

Overqualification Underqualification Matched qualification

19

EE SKILL MISMATCH: VARIANCE UR AND ER - MOLDOVA

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

2010 2011 2012 2013 2014 2015

Variance relative employment rates - Mold

Total Men Women

0.00

0.02

0.04

0.06

0.08

0.10

2010 2011 2012 2013 2014 2015

Variance relative unemployment rates -Mold

Total Men Women

0.00

0.10

0.20

0.30

2010 2011 2012 2013 2014 2015

Variance relative employment and unemployment rates (F+M) - Moldova

E/Ei (empl) U/Ui (unem)

20

MOLDOVA: PROPORTION OF UNEMPLOYED VSEMPLOYED BY EDUCATIONAL LEVEL

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1.60

2010 2011 2012 2013 2014 2015

Proportional mismatch - Moldova

Low Medium High

Levels education - LOW: ISCED 0-2; MED: ISCED 3-4; HIGH: ISCED 5-8

Excess supply of low

skilled labour

Persisting shortage

highly educated but

matched in last 2

years

Medium level

qualifications (VET):

matched; trend

towards shortage

21

MOLDOVA: OCCUPATIONAL MISMATCH (ISCO)

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00

Overqualific(HE)

Overqualificat(second level)

Matched qualif(HE)

Matched qualif(second lev)

Underqualif

Mismatch by occupation of employed population - trend (Moldova)

2010

2011

2012

2013

2014

2015

Levels education - LOW: ISCED 0-2; MED: ISCED 3-4; HIGH: ISCED 5-8

22

GEORGIA: PROPORTION OF UNEMPLOYED VSEMPLOYED BY EDUCATIONAL LEVEL

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2009 2010 2011 2012 2013 2014 2015

Proportional mismatch (Women) - Georgia

Primary & less Basic Medium High

0

0.2

0.4

0.6

0.8

1

1.2

1.4

2009 2010 2011 2012 2013 2014 2015

Proportional mismatch - (Men) - Georgia

Primary & less Basic Medium High

23

CENTRAL ASIA

KAZAKHSTANKYRGYZSTAN

TAJIKISTANTURKMENISTAN

UZBEKISTAN

Sources: World Bank

24

CENTRAL ASIA: EDUCATIONAL ATTAINMENT (25-64)

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2010 2015 2010 2015 2009

Kazakhstan Kyrgyzstan Tajikistan

Educational attainment adult population (25-64), %

Low Medium High

25

CENTRAL ASIA: A) EMPLOYMENT RATES BY SEX (20-64); B) UNEMPLOYMENT RATES (+15) AND YOUTH UR (15-24)

0

20

40

60

80

100

2010 2015 2010 2015 2009

Kazakhstan Kyrgyzstan Tajikistan

Employment rate by sex (20-64) - 2009 and 2015

Total Male Female 0

2

4

6

8

10

12

14

16

18

2010 2015 2010 2015 2009

Kazakhstan Kyrgyzstan Tajikistan

Unemployment rates by sex (15 +) and youth unemployment rates (15-24), %

Total Male Female Youth

26

CENTRAL ASIA: VET STUDENTS AS % UPPER-SECONDARY BY SEX

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

90.0

100.0

Kazakhstan Kyrgyzstan Tajikistan Uzbekistan

Students in VET as % upper sec students by sex - 2010, 2015

2010 2015

27

KAZAKHSTAN: VARIANCE UR AND ER (+15)

0.00

0.05

0.10

0.15

0.20

0.25

0.30

2011 2012 2013 2014 2015

Variance: relative unemployment and employment rates - KAZ (total)

VAR Ui/U VAR Ei/E

0.00

0.20

0.40

0.60

0.80

2011 2012 2013 2014 2015

Variance of relative unemployment rates by gender - KAZ

Total Men Women

0.00

0.05

0.10

0.15

0.20

0.25

2011 2012 2013 2014 2015

Variance relative employment rates by gender - KAZ

Total Men Women

28

KYRGYZSTAN: VARIANCE UR AND ER (+15)

0.00

0.10

0.20

0.30

0.40

0.50

2011 2012 2013 2014 2015

Variance relative employment and unemployment rates (F+M) - Kyrgyzstan

Ui/U Total Ei/E Total

0.00

0.50

1.00

1.50

2011 2012 2013 2014 2015

Variance relative unemployment rate (Ui/U) -Kyrg

Ui/U Total Men Women

0.00

0.10

0.20

0.30

2011 2012 2013 2014 2015

Variance relative employment rate (Ei/E) -Kyrg

Ei/E Total Men Women

VET graduates: ETF

tracer study 2015 –

¾ agree: skills not

matching

employers’ needs

hamper job search

29

KAZAKHSTAN: PROPORTION OF UNEMPLOYED VSEMPLOYED BY EDUCATIONAL LEVEL

0.00

0.50

1.00

1.50

2.00

2.50

3.00

2011 2012 2013 2014 2015

Proportional mismatch KAZ (total - F+M)

Primary and less Basic Secondary general Initial VET Secondary VET Incomplete higher Higher

30

KYRGYZSTAN: PROPORTION OF UNEMPLOYED VSEMPLOYED BY EDUCATIONAL LEVEL)

0.00

0.50

1.00

1.50

2.00

2.50

3.00

2011 2012 2013 2014 2015

Proportional mismatch (M+F) - Kyr

primary and less basic general secondary (compl)

primary profess secondary profess incompl higher

higher

0.00

0.50

1.00

1.50

2.00

2.50

3.00

2011 2012 2013 2014 2015

Proportional mismatch (M) - Kyrg

0.00

1.00

2.00

3.00

4.00

5.00

2011 2012 2013 2014 2015

Proportional mismatch (F) - Kyrg

31

CONCLUSIONS

Concepts and methodologies for skill mismatch measurement: need for

shared views

Better use of available data (in particular: statistical; special surveys; more

qualitative information) to analyse/ measure skill mismatch. Data

inconsistencies to be addressed (e.g.: education)

A simple indicator-based approach to quantifying on-the-job skills mismatch

across countries is likely to be unreliable. Combined analysis results different

methodologies – complementarity angles.

Instead, more careful country-specific analysis is needed to verify the extent

of "genuine" skills mismatch and its drivers to devise adequate policies.

Difficult solely on the basis of employer survey data, to gauge the extent of

genuine skills shortages

THANK YOU