KEY INDICATORS OF THE LABOUR MARKET Ninth edition

KEY INDICATORS

OF THE LABOUR MARKET

Ninth edition

Key Indicators of the Labour Market (KILM), Ninth edition

KILM 1. Labour force participation rate

KILM 2. Employment-to-population ratio

KILM 3. Status in employment

KILM 4. Employment by sector

KILM 5. Employment by occupation

KILM 6. Part-time workers

KILM 7. Hours of work

KILM 8. Employment in the informal economy

KILM 9. Unemployment

KILM 10. Youth unemployment

KILM 11. Long-term unemployment

KILM 12. Time-related underemployment

KILM 13. Persons outside the labour force

KILM 14. Educational attainment and illiteracy

KILM 15. Wages and compensation costs

KILM 16. Labour productivity

KILM 17. Poverty, income distribution, employment by economic class

and working poverty

Key

IndIcators

of the Labour

MarKet

Ninth edition

INTERNATIONAL LABOUR OFFICE · GENEVA

First published 2016

Publications of the International Labour Office enjoy copyright under Protocol 2 of the Universal Copyright Convention. Nevertheless, short excerpts

from them may be reproduced without authorization, on condition that the source is indicated. For rights of reproduction or translation, application

should be made to ILO Publications (Rights and Licensing), International Labour Office, CH-1211 Geneva 22, Switzerland, or by email: r[email protected].

The International Labour Office welcomes such applications.

Libraries, institutions and other users registered with a reproduction rights organization may make copies in accordance with the licences issued to

them for this purpose. Visit www.ifrro.org to find the reproduction rights organization in your country.

ILO

Key Indicators of the Labour Market, Ninth edition

Geneva, International Labour Office, 2016

Book and interactive software

Statistical table, labour market, employment, unemployment, labour force, labour cost,

wages, labour mobility, labour productivity, poverty, statistical analysis, comparison,

trend, developed country, developing country. 13.01.2

ISBN: 978-92-2-130121-9 (print)

ISBN: 978-92-2-030122-7 (PDF, Multilingual online edition)

ILO Cataloguing in Publication Data

The designations employed in ILO publications, which are in conformity with United Nations practice, and the presentation of material therein do not

imply the expression of any opinion whatsoever on the part of the International Labour Office concerning the legal status of any country, area or

territory or of its authorities, or concerning the delimitation of its frontiers.

The responsibility for opinions expressed in signed articles, studies and other contributions rests solely with their authors, and publication does not

constitute an endorsement by the International Labour Office of the opinions expressed in them.

Reference to names of firms and commercial products and processes does not imply their endorsement by the International Labour Office, and any

failure to mention a particular firm, commercial product or process is not a sign of disapproval.

ILO publications and digital products can be obtained through major booksellers and digital distribution platforms, or ordered directly from ilo@turpin-

distribution.com. For more information, visit our website: www.ilo.org/publns or contact [email protected].

This publication was produced by the Document and Publications

Production, Printing and Distribution Branch (PRODOC) of the ILO.

Graphic and typographic design, layout and composition, proofreading,

printing, electronic publishing and distribution.

PRODOC endeavours to use paper sourced from forests managed

in an environmentally sustainable and socially responsible manner.

Code: DTP-JMB-CORR-REPRO

This 9th edition of the Key Indicators of the

Labour Market (KILM) is being issued by the ILO

Department of Statistics for the first time. The

series dates back to 1999 and was previously

published by the ILO’s Employment Sector. As

part of the ILO reorganization in 2012, the

Department of Statistics was mandated to consol-

idate all existing ILO statistical databases into

ILOSTAT, the successor to LABORSTA and other

databases published in the past by the ILO.

ILOSTAT is the largest repository of labour statis-

tics in the world, covering all dimensions of

decent work.

The KILM builds on the data reported by

countries to ILOSTAT and is enriched by external

statistical sources from other organizations

including Eurostat, OECD, UNESCO and the World

Bank. It relies on internationally comparable data

derived from statistical standards agreed by the

International Conference of Labour Statisticians.

Indicators in the database are disaggregated

wherever possible, with the goal of identifying

key trends and priority groups for labour market

interventions.

The KILM is a user-friendly and easy-to-

understand database, containing 17 indicators

that capture the most important aspects of the

world’s labour markets. In a joint collaboration

with the ILO Research Department, it also

includes global, regional and national estimates

for selected indicators. These estimates are

explicitly identified in the database and follow

methodologies that are reviewed and improved

with the release of each edition.

The publication of this edition occurs at an

opportune and important time, as the global

community has just adopted the 2030 Agenda for

Sustainable Development. The Agenda calls for a

“data revolution” to strengthen the production

and dissemination of statistics in all domains in

order to better understand developments at the

national, regional and global levels and thereby

enable better informed policy-making.

The indicator framework to track progress on

the Sustainable Development Goals (SDGs) is

currently being discussed by countries and the

international organizations with a goal to

complete the work by the end of 2016. The frame-

work will build on existing indicators and will

also embrace the new ones.

Inclusive and sustainable economic growth

and full and productive employment and decent

work for all are the overarching objective of

Goal 8 of the Agenda. Many of the indicators on

key aspects of decent work will build on existing

ones and databases like KILM will help to set

benchmarks and to monitor labour markets

around the world.

This edition of the KILM also provides

thematic discussions of the educational level of

the labour force and unemployment. It analyses

trends in the share of youth who are not in

employment, education or training, the so-called

NEETs. Reducing the size of this group is a

specific target within Goal 8 of the SDGs.

There are many organizations and colleagues

who are acknowledged below but I want to espe-

cially thank the producers of the data reported

here: national statistical offices, ministries of

labour and other national institutions which dili-

gently, carefully, and often with scarce resources,

produce each survey, registry, census and other

statistical sources to shed light on the world of

work, carefully following international definitions

and standards and the Fundamental Principles of

Official Statistics.

Preface

Rafael Diez De MeDina

Chief Statistician and

Director,

Department of Statistics

International Labour Office

Table of contents

Preface ............................................................................................ v

Summary of KILM 9th Edition ...................................................................... 1

Acknowledgements ................................................................................ 5

Guide to understanding the KILM .................................................................. 7

The history of the KILM .................................................................... 7

The role of the KILM in labour market analysis .................................................. 7

Labour market analyses using multiple KILM indicators ........................................... 10

KILM organization and coverage ............................................................. 11

Information repositories and methodological information ......................................... 12

Summary of the 17 ILO Key Indicators of the Labour Market ....................................... 14

KILM electronic versions ................................................................... 20

References .............................................................................. 20

Education and labour markets: Analysing global patterns with the KILM .............................. 23

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

2. Global trends by indicator ............................................................... 24

2.1. Labour force distribution by level of educational attainment................................ 24

2.2. Unemployment distribution by level of educational attainment ............................. 27

2.3. Unemployment rate by level of educational attainment.................................... 28

2.4. Share of youth not in education, employment or training (NEET) ............................ 31

3. Impact of education on labour market outcomes ............................................. 32

3.1. Unemployment and education ....................................................... 32

3.2. Labour productivity and education.................................................... 35

3.3. Employment-to-population ratio and education .......................................... 37

3.4. Share of employees and education .................................................... 37

4. The current situation in 12 selected countries................................................ 37

4.1 Data for latest year available on the four selected indicators ................................ 37

4.2. Educational attainment compared to other key labour market indicators ...................... 41

4.3. Remaining gaps in education ........................................................ 44

4.3.1. Persistent low levels of educational attainment .................................... 44

4.3.2. Disparities between population groups .......................................... 45

4.3.3. Attention to qualitative factors and field of study................................... 47

5. Conclusion ........................................................................... 47

References .............................................................................. 47

Annex .................................................................................. 48

VIII

Contents

KILM 1. Labour force participation rate............................................................... 51

KILM 2. Employment-to-population ratio.............................................................. 55

KILM 3. Status in employment...................................................................... 61

KILM 4. Employment by sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

KILM 5. Employment by occupation ................................................................. 69

KILM 6. Part-time workers ......................................................................... 73

KILM 7. Hours of work ........................................................................... 77

KILM 8. Employment in the informal economy......................................................... 83

KILM 9. Unemployment........................................................................... 89

KILM 10. Youth unemployment...................................................................... 97

KILM 11. Long-term unemployment .................................................................. 101

KILM 12. Time-related underemployment .............................................................. 105

KILM 13. Persons outside the labour force ............................................................. 111

KILM 14. Educational attainment and illiteracy .......................................................... 113

KILM 15. Wages and compensation costs .............................................................. 119

KILM 16. Labour productivity ....................................................................... 131

KILM 17. Poverty, income distribution, employment by economic class and working poverty ..................... 135

List of figures

2.1 Share of the labour force with primary or less than primary level educational attainment (%) ............. 25

2.2 Share of the labour force with tertiary level educational attainment (%) .............................. 26

2.3 Share of unemployed with tertiary level educational attainment (%) ................................. 27

2.4. Share of young unemployed with tertiary level educational attainment (%) ............................ 28

2.5. Unemployment rate of persons with tertiary level educational attainment (%) ......................... 29

2.6. Unemployment rate of persons with primary or less than primary level educational attainment (%) ........ 30

2.7. Share of youth not in education, employment or training (%) ....................................... 31

3.1. Share of labour force and unemployed with tertiary level of educational attainment (%) ................. 32

3.2. Share of labour force and unemployment rate for persons with tertiary level of educational attainment (%) .. 33

3.3. Tertiary level of educational attainment and labour productivity (PPP US$) ........................... 34

3.4 Employment-to-population ratio and labour force with tertiary level educational attainment .............. 35

3.5. Share of employees in total employment, and share of labour force with tertiary level

of educational attainment................................................................... 36

4.1. Labour force by level of educational attainment ................................................. 38

4.2. Unemployment distribution by level of educational attainment ..................................... 39

4.3. Unemployment rate by level of educational attainment ........................................... 40

4.4. Share of youth (aged 15−24) not in education, employment or training, by sex ........................ 40

4.5. Tertiary level educational attainment and unemployment ......................................... 41

4.6. Tertiary level educational attainment and unemployment rate ...................................... 42

4.7. Tertiary level educational attainment and labour productivity ...................................... 43

4.8. Employment-to-population ratio and educational attainment ....................................... 43

4.9. Share of employees and educational attainment ................................................. 44

4.10. Male and female unemployment rates by educational attainment.................................... 45

4.11. Youth and adult unemployment rates by educational attainment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

List of boxes

1a. Labour market statistics at the ILO............................................................ 8

1b. ILO methodology for producing global and regional estimates of labour market indicators ............... 13

1c. Resolution concerning statistics of work, employment and labour underutilization...................... 19

1.1. Data on the labour market and education: Statistical issues ........................................ 24

2. Resolution concerning statistics of work, employment and labour underutilization,

adopted by the 19th International Conference of Labour Statisticians, October 2013 [relevant paragraphs]... 58

4. International Standard Industrial Classification of All Economic Activities ............................. 66

5a. International Standard Classifications of Occupations: major groups ................................. 71

5b. International Standard Classification of Occupations, 2008......................................... 72

7. Resolution concerning the measurement of working time, adopted by the 18th International

Conference of Labour Statisticians, November–December 2008 ..................................... 80

Table of contents

8a. Avoiding confusion in terminologies relating to the informal economy ............................... 86

8b. Timeline of informality as a statistical concept .................................................. 87

9. Resolution concerning statistics of work, employment and labour underutilization,

adopted by the 19th International Conference of Labour Statisticians, October 2013 [relevant paragraphs]... 95

12a. Resolution concerning the measurement of underemployment and inadequate employment

situations, adopted by the 16th International Conference of Labour Statisticians,

October 1998 [relevant paragraphs] .......................................................... 108

12b. Resolution concerning statistics of work, employment and labour underutilization,

adopted by the 19th International Conference of Labour Statisticians,

October 2013 [relevant paragraphs] .......................................................... 109

15a. The ILO’s Global Wage Report ............................................................... 126

15b. Resolution concerning statistics of labour cost, adopted by the 11th International Conference

of Labour Statisticians, October 1966 [relevant paragraphs] ........................................ 127

15c. Resolution concerning an integrated system of wages statistics, adopted by the 12th International

Conference of Labour Statisticians, October 1973 [relevant paragraphs] .............................. 128

Summary of KILM 9th Edition

The KILM has become

a cornerstone of

information for those

concerned with

the world of work.

The first edition of the Key Indicators of the Labour Market (KILM)

was released in 1999. It has since become a leading product of the

International Labour Office (ILO) and is used daily by researchers and

policy-makers throughout the world.

At the national level, statistical information is generally gathered and

analysed by statistical services and ministries. At the global level, the ILO

plays a vital role in assembling and disseminating labour market informa-

tion and analysis to the world community. ILOSTAT, the ILO consolidated

database, is the biggest repository of labour statistics in the world. Due

to its complexity and wide range of indicators, ILOSTAT includes subsets

of databases which provide more in-depth analysis for key indicators.

This is the case for the KILM, which is based in large part on data from

ILOSTAT, augmented by data from other international repositories and

with estimates and projections carried out by the ILO Research

Department and Department of Statistics. A key aim of the KILM is to

present a core set of labour market indicators in a user-friendly manner.

This ninth edition

of the KILM strengthens

the ILO’s efforts to

support measurement

of national progress

towards the new SDG

of “promoting

sustained, inclusive

and sustainable

economic growth,

full and productive

employment and decent

work for all”.

The KILM also serves as a source of national data for measuring

progress towards Sustainable Development Goal 8 (SDG 8), to “promote

sustained, inclusive and sustainable economic growth, full and produc-

tive employment and decent work for all”. For example, GDP per capita

and GDP growth (KILM A), the share of informal employment in non-

agricultural employment (KILM 8) and the share of youth not in educa-

tion, employment or training (NEET, KILM 10c), when analysed together,

can offer a rich assessment of trends and levels of decent and product-

ive employment in a country. While the Indicator Framework to moni-

tor the 2030 Agenda for Sustainable Development is still under develop-

ment, this set of key indicators will undoubtedly be instrumental for

this purpose.

The KILM also provides valuable information on indicators relating

to other SDGs linked to employment and the labour market. For example,

the statistics on poverty and income distribution contained in KILM table

18a can be a very useful tool for measuring progress towards the SDG 1

of “ending poverty in all its forms everywhere” and SDG 10 of “reducing

inequality within and among countries”.

The KILM offers timely

data and tools for those

seeking to run their

own analysis.

The KILM programme has met the primary objectives set for it in 1999,

namely: (1) to present a core set of labour market indicators; and (2) to

improve the availability of the indicators to monitor new employment

trends. But that is not all that the KILM has to offer. It has evolved into a

primary research tool that provides not only the means for analysis, i.e. the

data, but also guidance on interpretation of indicators and data trends.

These contributions – including those in this KILM 9th edition, described

below – serve the ILO’s agenda of identifying employment challenges in

order to inform policy action that can create more decent work opportu-

nities around the world, especially where the need is greatest.

Summary of KILM 9th Edition

Smart policy-making

requires up-to-date and

reliable labour market

information …

Defining effective labour market strategies at the country level requires

first and foremost the collection, dissemination and assessment of up-to-

date and reliable labour market information.

Once policies and strategy

are decided, continued gathering and analysis of information are essen-

tial to monitor progress towards goals and to adjust policies where

needed. Labour market information and analysis is an essential founda-

tion for the development of integrated strategies to promote fundamen-

tal principles and rights at work, productive employment, social protec-

tion and dialogue between the social partners, as well as to address the

cross-cutting themes of gender and development. This is where the KILM

comes in.

… such as that

provided in the KILM.

The KILM is a collection of 17 “key” indicators of the labour market,

covering employment and other variables relating to employment

(status, economic activity, occupation, hours of work, etc.), employment

in the informal economy, unemployment and the characteristics of the

unemployed, underemployment, education, wages and compensation

costs, labour productivity and working poverty. Taken together, the

KILM indicators provide a strong basis for assessing and addressing key

questions related to productive employment and decent work.

This edition highlights

current labour market

trends:

This ninth edition of the KILM offers a series of noteworthy findings,

some excerpts of which are presented here:

The world’s labour

force is becoming more

and more highly

educated, which could

support increased

productivity.

Education is not always

effective in protecting

against unemployment.

Unemployment

remains elevated in

most countries.

•

The educational level of the world’s labour force is improving. In 62

out of 64 countries with available data spanning the past 15 years, the

share of the labour force with a tertiary education increased. In all

but three of these countries, the share of the labour force that had

attained only primary or less than primary educational level declined

– with significant improvements seen in many low income, lower

middle income and upper middle income economies.

•

An increasing proportion of the labour force with tertiary level

education is associated with higher levels of labour productivity, and

so these favourable education trends could facilitate an expansion in

production of higher value added goods and services and faster

productivity growth, thereby supporting economic growth and

development.

•

In 67 of the 93 countries for which data are available, people with

tertiary education are less likely to be unemployed than people with

lower levels of education. Yet, while higher levels of education are

found to protect workers from unemployment in most high income

countries, among upper middle income economies the situation is

more mixed, and in low income and lower middle income econ-

omies, people with high levels of education tend to be more likely to

be unemployed. In these developing economies, there is a clear

mismatch between the numbers of skilled people and of available

jobs matching their competencies and expectations.

•

Out of 112 countries with comparable KILM unemployment rate

data, 71 (63 per cent) had higher unemployment rates in 2014 (or the

closest available year) than in 2007. The median unemployment rate

across these 112 countries increased from 6.4 per cent in 2007 to

7.2 per cent in 2014.

For examples of how the KILM can be used when formulating policies, see the “Guide to understanding the KILM”.

Summary of KILM 9th Edition

Wide productivity

gaps remain, with

differences in levels

of industrialization

playing a key role

in determining

productivity levels.

Productivity

is growing fastest

in middle income

economies.

•

High income economies saw the number of unemployed increase by

16.2 million between 2007 and 2009, accounting for 56 per cent of

the total global increase in unemployment during the global financial

and economic crisis. However, since 2009, the number of un-

employed in high income economies has declined by 5.7 million,

whereas the number of unemployed in each of the other income

groups continues to grow.

•

The average worker in a high income economy currently produces

62 times the annual output of the average worker in a low income

economy and ten times that of the average worker in a middle income

economy (based on productivity figures in constant 2005 US$).

•

Economic structure is closely related to these productivity differ-

ences. In low income countries, more than two-thirds of all workers

are employed in the agricultural sector – often in low productivity,

subsistence activities – and only 9 per cent are employed in industry.

In middle income economies, less than one-third of workers are

employed in agriculture, while 23 per cent of workers are employed

in the industrial sector.

•

Middle income economies have accounted for nearly all (97 per cent)

of the global growth in industrial employment since 2000. Manufac-

turing employment in high income economies has declined by

5.2 million since 2000, while in middle income economies it has

grown by 195 million.

•

In line with this rapid industrialization, middle income economies

have registered the fastest productivity growth over the past 15 years

(measured as output per worker) and also the fastest growth over the

more recent period following the global economic crisis. Since 2009,

upper middle income economies have seen productivity rise by

4.6 per cent per year on average, with productivity in lower middle

income economies growing by 3.8 per cent per year. Productivity in

low income economies rose by 3.2 per cent per year over the same

period, while high income economies registered an annual increase

of only 1.2 per cent.

•

As of 2015, the vast majority (72 per cent) of the world’s workers are

employed in middle income economies (with per capita gross

national income, GNI, of between US$1,045 and US$12,736). Twenty

per cent of the world’s workers are employed in high income econ-

omies (GNI per capita above US$12,736), while 8 per cent are

employed in low income economies (GNI per capita below US$1,045).

Thus, labour market trends in middle income economies shape, in

large part, overall global labour market trends.

Favourable labour

market trends

in middle income

economies have

helped reduce global

poverty.

•

On the back of rapid industrialization and robust productivity growth,

the number of working poor (workers in households where each

person lives on less than US$2 per day at purchasing power parity,

PPP) declined by 479 million between 2000 and 2015 – with the

share of the working poor in total employment dropping from 57 per

cent of the workforce in middle income economies in 2000 to

25 per cent in 2015. Middle income economies accounted for all of

the world’s reduction in working poverty over this period.

Summary of KILM 9th Edition

The KILM 9th edition

has many features

enabling easy access

and manipulation

of the data, including

a friendly user-driven

projection tool.

The interactive KILM software and Excel add-in (downloadable from

the ILO Department of Statistics website: www.ilo.org/kilm) make

searching for relevant labour market information and analysis quick and

simple. For those who wish to work from the Internet, the KILM indica-

tors can be directly downloaded for individual countries from the KILM

webpage. Each version offers a simple user interface for running queries

on the most up-to-date KILM indicators. Users can also access ILO world

and regional aggregates of selected key indicators directly from the KILM

software, Excel add-in and Internet database.

The KILM 9th Edition is a product of the Data

Production and Analysis Unit within the ILO

Department of Statistics. Marie-Claire Sodergren

coordinated the production of the KILM, with

invaluable contributions in the areas of data

production, analysis and manuscript drafting

from David Bescond, Evangelia Bourmpoula,

Harvey Clavien, Rosina Gammarano, Messaoud

Hammouya, Lê Anh Hua, Arnaud Kunzi, Devora

Levakova, Akiko Minowa, Yves Perardel, Alan

Wittrup and Yanwen Zhu. Alan Wittrup devel-

oped the KILM interactive software. The KILM

was supervised by Steven Kapsos, Chief of the

Data Production and Analysis Unit.

The Team benefited from the strong support

of Rafael Diez de Medina, Director of the ILO

Department of Statistics and Chief Statistician,

Ritash Sarna, Chief of the Department’s

Management Support Unit, and the Microdata

and Knowledge Management Unit through its

Chief, Edgardo Greising and its Database

Administrator, Christophe Vittorelli.

The continuing support from the Deputy

Director-General for Policy, Sandra Polaski, and

Sangheon Lee from her office is greatly appreci-

ated. The ILO Research Department has provided

ILO estimates and for that we thank Raymond

Torres, its Director, as well as Moazam Mahmood,

Ekkehard Ernst, Stefan Kühn, Santo Milasi, Steven

Tobin and Christian Viegelahn. The Employment

Department has also provided valuable comments

and we thank Azita Berar-Awad, its Director, as

well as all her team.

The KILM 9th Edition builds on the efforts of

many ILO colleagues involved in the production

of the KILM databases and manuscripts over the

past 16 years. In particular, the contributions of

Sara Elder, Lawrence Jeff Johnson, Julia Lee,

Dorothea Schmidt, Theo Sparreboom and

Christina Wieser are gratefully acknowledged.

Production of the KILM is dependent on the

continuing efforts of national statistical offices

throughout the world to collect, publish and

disseminate information pertaining to their

labour markets. The KILM also benefits from

collaboration with ILO regional and other field

offices, specifically in the sharing of information

from national and regional statistical sources. A

good deal of the information published is made

available through the continued cooperation and

support of additional organizations, including

The Conference Board, the statistical office of

the European Communities (Eurostat), the

Organisation for Economic Co-operation and

Development (OECD), the UNESCO Institute of

Statistics, and the World Bank. It is with deep

gratitude that we thank our contacts in the vari-

ous organizations for taking the time to share

their data with us.

We would like to express our thanks to the

staff of the ILO Department of Statistics who

have provided technical and administrative

support throughout the whole process, in par-

ticular Catherine Jensen, Agnes Kalinga and

Virginie Woest. We would also like to thank the

ILO Department of Communication and Public

Information for their continued collaboration

and support. Finally, members of the KILM team

wish to express their deep appreciation to any

organization or individual not listed here who

assisted or provided guidance during the devel-

opment and implementation of the project.

Acknowledgements

The history of the KILM

Any organization, institution or government

that advocates labour-related strategies needs rele-

vant data in order to monitor and assess the current

realities of the world of work. In recognition of

this, the International Labour Office (ILO) launched

the Key Indicators of the Labour Market (KILM)

programme in 1999 to complement its regular data

collection programmes and to improve dissemina-

tion of data on the key elements of the world’s

labour markets (for the various statistical activities

carried out by the ILO, see box 1a).

The KILM was originally designed with two

primary objectives in mind: (1) to present a core

set of labour market indicators; and (2) to improve

the availability of the indicators to monitor new

employment trends. The selection of the indicators

was based on the following criteria: (a) conceptual

relevance; (b) data availability; and (c) relative

comparability across countries and regions. Since

the first edition, the design and presentation of the

core indicators have gradually evolved.

The role of the KILM

in labour market analysis

Sound evidence-based policy-making relies on

identifying and quantifying not only best prac-

tices in the labour market but also inefficiencies

– such as labour underutilization and decent work

deficits. This is the first step in designing employ-

ment policies aimed at enhancing the well-being

of workers while also promoting economic

growth. This broad view of the world of work calls

for comprehensive collection, organization and

analysis of labour market information. In this

context, the KILM can serve as a tool in monitor-

ing and assessing many of the pertinent issues

related to the functioning of labour markets. The

following are some examples of how the KILM

can be used to inform policy in key areas of ILO

research.

Promoting the ILO’s Decent Work

Agenda

The ILO’s Decent Work Agenda aims to

promote opportunities for women and men to

obtain productive work, in conditions of free-

dom, equity, security and human dignity.

As a

growing number of governments, employers and

workers investigate options for designing pol-

icies that adhere to the principles of decent work,

it falls to policy-makers to interpret the term

“decent”. Perceptions of what constitutes a

decent job or a decent wage are likely to differ,

depending on national circumstances, the polit-

ical views of policy-makers and each individual’s

position in relation to the labour market. There

are, however, certain conditions relating to the

world of work that are almost universally

accepted as “bad” – for example, working but

earning an income that does not lift one above

the poverty line, or working under conditions

where the fundamental principles and rights at

work

are not respected.

Given that policy formulation should always

be preceded by careful empirical research and

quantitative assessments of the realities of the

world of work, the KILM, as a collection of a broad

range of labour market indicators, can serve as a

tool in addressing many of the pertinent ques-

tions relating to the ILO’s Decent Work Agenda.

The KILM helps to identify where labour is

underutilized and decent work is lacking, not only

in terms of people who are working yet still

unable to lift themselves and their families above

the poverty threshold (KILM 17) but also in terms

of poor quality of work or the lack of any work at

all. The lack of any work can be identified using

unemployment (KILMs 9 and 10) but also more

broadly using inactivity (KILM 13). Poor quality of

work can be assessed using a combination of indi-

cators: for example, by identifying which individu-

als are in vulnerable employment (using status

and sector – KILMs 3 and 4), working excessive

hours (KILM 7), working in the informal economy

(KILM 8), underemployed (KILM 12) or working

in low-productivity jobs (KILM 16).

Since the publication of the Director-General’s report

at the 1999 International Labour Conference (ILO, 1999), the

goal of “decent work” has come to represent the central

mandate of the ILO, bringing together labour standards, funda-

mental principles and rights at work, employment, social

protection and social dialogue in the formulation of policies

and programmes aimed at “securing decent work for women

and men everywhere”.

The ILO Declaration on Fundamental Principles and

Rights at Work aims to ensure that social progress goes hand

in hand with economic progress and development. See http://

www.ilo.org/declaration for more information.

Guide to understanding the KILM

Box 1a. Labour market statistics at the ILO

Statistical activities have always formed an integral part of the work of the International Labour

Organization, as witnessed by the setting up in 1919 of a Statistical Section for “the collection

and distribution of information on all subjects relating to the international adjustment of

conditions of industrial life and labour” (Article 396 of the Versailles Treaty of Peace and article

10(1) of the Constitution of the ILO). Since its inception, the ILO has endeavoured to carry out

its mandate in an ever-changing world. Key statistical functions are performed by the ILO’s

Department of Statistics, the focal point for labour statistics in the United Nations (UN) system.

Formerly a Bureau, the Department of Statistics – established in 2009 – is responsible for

enhancing data compilation, increasing support to countries and constituents to produce,

collect and use more timely and accurate labour and decent work statistics, coordinating and

assessing the quality of ILO statistical activities, setting international statistical standards (by

hosting the International Conference of Labour Statisticians and providing guidelines and

support) and enhancing capacity building in labour and decent work statistics.

For a very long time, a key publication in disseminating labour market statistics was the ILO

Yearbook of Labour Statistics, first issued in 1935. It contained time series data on a wide range

of topics related to the labour market, which changed over time to reflect current interests and

developments. The topics covered have included employment, unemployment, hours of work,

wages, cost of living and retail prices, workers’ family budgets, emigration and immigration,

occupational injuries and industrial relations. Monthly or quarterly updates of the series

published in the Yearbook were first issued in the International Labour Review and its statistical

supplement, and from 1965 in the quarterly Bulletin of Labour Statistics and its supplement.

The Bulletin also contained short articles on statistical practices and methods, and presentations

of the results of special projects carried out by the Department of Statistics.

In 2010 the Department of Statistics embarked on a comprehensive revision of the procedures

used to compile, store and disseminate data, with a view to satisfying the needs of all types of

users of labour market statistics to a fuller extent and in a timelier manner. As a result of this

exercise, the printed publications of the Yearbook of Labour Statistics and the Bulletin were

discontinued and replaced with ILOSTAT, a continuously updated online database containing

annual and short-term statistics. ILOSTAT, available at www.ilo.org/ilostat, also includes data

sets on specific labour-related topics (such as labour migration and social security) and all the

relevant methodological information, including concepts and definitions, classifications and

metadata on the national statistical sources used. The active identification of gaps in the

information helps to inform the technical support the ILO offers to countries. The main focus is

on establishing ILOSTAT as a coordinated and closely monitored database that presents timely

and accurate official figures. The inclusion of short-term data from 2010 has enabled the ILO

to better monitor the employment situation across countries without having to wait for annual

data, improving its capacity to report to important bodies and events such as the G20 and

regional meetings.

The Key Indicators of the Labour Market (KILM) complements this effort by providing consistent

and comparable labour market information. The KILM differs from ILOSTAT’s yearly indicators

in terms of scope and content. Whereas the yearly indicators are the best source of nationally

reported labour statistics, and notwithstanding intensified efforts to obtain comparable data

following the ILO’s preferred concepts and definitions, the KILM has more freedom to enhance

the comparability of series across time and countries, given that it is not restricted to using the

national data as reported. In the case of indicators that cannot be streamlined and remain not

strictly comparable, efforts have been made to select sources and methodologies that provide

series that are as “clean” and comparable as possible; and where anomalies exist in terms of

definitions and methodologies, these are clearly specified in the table notes. Finally, some

indicators are provided in both the yearly indicators and the KILM; however, the full lists of

indicators in each are not identical. For example, labour productivity is included in the KILM,

but not in the yearly indicators, whereas the yearly indicators report data on strikes and lockouts

and occupational injuries, while the KILM does not.

Guide to understanding the KILM

sustainable economic growth, employment and

decent work for all”.

The KILM presents statistics

for several of the indicators currently proposed

for measuring progress towards the eighth SDG,

namely GDP per capita and GDP growth, the

share of informal employment in non-agricultural

employment, the employment-to-population

ratio, the unemployment rate, the youth un-

employment rate, and the share of youth not in

education, employment or training: these corres-

pond to KILM tables A1, 8, 2b, 9b, 10b and 10c,

respectively.

Furthermore, the KILM also

provides valuable information on indicators rele-

vant to monitoring other SDGs linked to employ-

ment and the labour market, such as the statistics

on poverty and income distribution contained in

KILM table 18a, which can be used to measure

progress towards the first SDG, to “end poverty in

all its forms everywhere”.

Monitoring equity in the labour market

Women face specific challenges in attaining

decent work. The majority of KILM indicators are

disaggregated by sex, allowing for comparison of

male and female labour market opportunities.

Many of the “trends” analyses associated with

individual indicators focus on the progress (or

lack thereof) towards the goal of equal opportu-

nity and equal treatment in the labour market.

Assessing employment in a globalizing

world

Globalization has the potential of being bene-

ficial to all, but to date the benefits are not reach-

ing enough people. The goal, therefore, is to

embrace globalization but in a way that shapes it

to encourage creation of decent work opportuni-

ties for all (WCSDG, 2004). One means of doing

so is to make employment a central objective of

macroeconomic and social policies. The KILM

indicators can be useful in this regard by enabling

the employment dynamics associated with

globalization to be monitored. For example, there

are studies indicating that globalization has

impacts on job loss and creation and on changes

in wages and productivity (and thus in interna-

tional competitiveness). If the indicators reflect

negative consequences of globalization, ways can

be sought of altering macroeconomic policies so

as to minimize the costs of adjustment and to

The latest list of indicator proposals available at this

time (first disseminated 11 Aug. 2015) can be found at: http://

unstats.un.org/sdgs/files/List%20of%20Indicator%20Propos-

als%2011-8-2015.pdf.

For a guide on using KILM indicators to assess gender

equality, see ILO, 2010.

Monitoring progress towards

the UN’s Millennium Development Goals

and Sustainable Development Goals

The UN resolved to make the goals of full and

productive employment and decent work for all

a central objective of both its national and inter-

national policies and its national development

strategies as part of its efforts to achieve the

Millennium Development Goals (MDGs).

Recognizing that decent and productive work for

all is central to addressing poverty and hunger,

MDG 1 includes a target 1b (agreed upon in

2008) to “achieve full and decent employment for

all, including women and young people”. The

four indicators selected at the time for monitor-

ing progress towards MDG target 1b are available

within the KILM: (1) employment-to-population

ratio (KILM 2); (2) the proportion of employed

people living below the poverty line (working

poverty rate: KILM 17); (3) the proportion of

own-account and contributing family workers in

total employment (vulnerable employment rate:

KILM 3); and (4) the growth rate of labour

productivity (KILM 16).

With the MDGs concluding in 2015, a series

of 17 Sustainable Development Goals (SDGs) has

been agreed upon to succeed them.

Within the

context of the SDGs, the quest for full and decent

employment for all has been given new promi-

nence, Goal 8 being to “promote inclusive and

See UN, 2005, para. 47. As part of the Millennium Decla-

ration of the United Nations “to create an environment – at

the national and global levels alike – which is conducive to

development and the elimination of poverty”, the interna-

tional community adopted a set of international goals for

reducing income poverty and improving human develop-

ment. A framework of eight goals, 18 targets and 48 indicators

to measure progress was adopted by a group of experts from

the UN Secretariat, ILO, International Monetary Fund (IMF),

Organisation for Economic Co-operation and Development

(OECD) and World Bank. The indicators are interrelated and

represent a partnership between developed and developing

economies. For further information on the MDGs, see http://

www.un.org/millenniumgoals.

KILM, 6th edn (ILO, 2009), Ch. 1, section C offered a

demonstration of how to put all four MDG employment indi-

cators together to arrive at a basic analysis of progress at the

country level. KILM, 7th edn (ILO, 2011), Ch. 1, section A

presented insights into working poverty in the world and

introduced new estimates on working poverty. See also

Sparreboom and Albee, 2011.

During the UN Sustainable Development Summit held

on 25−27 September 2015 in New York and convened as a

high-level plenary meeting of the General Assembly, world

leaders, businesses and civil society groups gathered to

discuss issues relevant to the new development agenda, such

as poverty, hunger, inequality and climate change. The summit

resulted in the adoption of an ambitious new sustainable

development agenda, and a set of 17 Sustainable Develop-

ment Goals. The full list of SDGs and their corresponding

targets is available at: http://www.un.org/sustainabledevelop-

ment/sustainable-development-goals/.

Guide to understanding the KILM

(not working and seeking work: KILMs 9 and 10).

A large share of the population either in un-

employment or outside the labour force, or both,

indicates substantial underutilization of the

potential labour force and thus of the economic

potential of a country. Governments facing this

situation should, if possible, seek to analyse the

reasons for inactivity, which in turn could indi-

cate the policy choices necessary to redress the

situation.

For example, if the majority of the population

outside the labour force is made up of women

who are not working because they have house-

hold responsibilities, the State might wish to

encourage an environment that facilitates female

economic participation through such measures

as the establishment of day-care centres for chil-

dren or flexible working hours. Alternatively, if

disability is a common reason for staying outside

the labour force, programmes to promote the

employment of the disabled could help to lower

the inactivity rate. It is more difficult to recapture

persons who have left the labour market because

they are “discouraged”, that is, because they feel

that no suitable work is available or that they do

not have the proper qualifications, or because

they do not know where to look for work;

however, it may be possible to boost their confi-

dence through participation in training

programmes and jobsearch assistance. In any

particular national context, the correct mix of

policies can only be designed by looking in detail

at the reasons for inactivity.

Unemployment itself should be analysed

according to sex (KILM 9), age (KILM 10), length

(KILM 11) and educational attainment (KILM 14)

in order to gain a better understanding of the

composition of the jobless population and there-

fore to target unemployment policies appropri-

ately. Other characteristics of the unemployed

not shown in the KILM, such as socio-economic

background, work experience, etc., could also be

significant, and should be analysed, if available, in

order to determine which groups face particular

hardships. Paradoxically, low unemployment

rates may well disguise substantial poverty in a

country (see KILM 17), whereas high unemploy-

ment rates can occur in countries with significant

economic development and low incidence of

poverty. In countries without a safety net of

unemployment insurance and welfare benefits,

many individuals, despite strong family solidarity,

simply cannot afford to be unemployed. Instead,

they must eke out a living as best they can, often

in the informal economy or in informal work

arrangements within the formal economy. In

countries with well-developed social protection

schemes or when savings or other means of

distribute the gains of globalization in a more

equitable fashion.

Identifying “best practices”

The KILM can help to identify best-practice

country examples on a number of issues: where

the occupational gender wage gap is non-existent

or minimal; where young people do not face

disadvantages in access to jobs; where labour

productivity and labour compensation are

balanced in such a way as to encourage interna-

tional competitiveness; where economic growth

has gone hand in hand with an expansion of

employment opportunities; where a country

reduces high unemployment; and many others.

The key in each case is to identify policies that

have led to the positive labour market outcome

and to highlight these as possible best practices

that could be implemented elsewhere.

Labour market analyses

using multiple KILM

indicators

More and more countries are producing

national unemployment and aggregate employ-

ment data. Nevertheless, caution is required in

the interpretation of such statistics, given their

limitations if used in isolation, and users are

urged to take a broader view of labour market

developments, combining a range of statistics.

The advantage of using aggregate unemploy-

ment rates, for example, is their relative ease of

collection and comparability for a significant

number of countries. But unemployment is only

one aspect of labour market status, and to look

at this (or any other labour market indicator)

alone is to ignore other elements of the labour

market that are no less significant for being more

difficult to quantify.

The first step in labour market analysis, there-

fore, is to determine the breakdown of labour

force status within the population.

According to

the definitions established in the resolution

concerning statistics of work, employment and

labour underutilization adopted by the 19th

International Conference of Labour Statisticians

in 2013 (ILO, 2013), the working-age population

can be broken down into persons outside the

labour force (formerly known as inactive:

KILM 13), employed (KILM 2) or unemployed

For a specific country example of how to analyse

labour markets using the KILM indicators, see ILO, 2011, Ch. 1,

section C; ILO, 2007,

Appendix F.

Guide to understanding the KILM

programmes is desperately needed in many

developing economies.

KILM organization

and coverage

The Statistics Division of the UN compiles

statistics for approximately 230 countries, areas

and territories.

For each edition of the KILM, the

ILO has made an intensive effort to assemble data

on the indicators for as many countries, areas and

territories as possible. Where there is no informa-

tion for a country, it is usually because that coun-

try was not in a position to provide information

for that indicator, or because such information as

was available was not sufficiently current or did

not meet other criteria established for inclusion

in the KILM.

The KILM groups countries in two different

ways: geographically, distinguishing countries by

region and subregion (broad and detailed); and

according to per capita income, on the basis of

the World Bank’s classification by income group.

There are five main geographical groupings:

(1) Africa; (2) Americas; (3) Arab States; (4) Asia

and the Pacific; and (5) Europe and Central Asia.

These are further divided into 11 corresponding

broad subregions – (1.1) Northern Africa;

(1.2) Sub-Saharan Africa; (2.1) Latin America and

the Caribbean; (2.2) Northern America; (3.1) Arab

States; (4.1) Eastern Asia; (4.2) South-Eastern Asia

and the Pacific; (4.3) Southern Asia; (5.1) Northern,

Southern and Western Europe; (5.2) Eastern

Europe; and (5.3) Central and Western Asia – and

20 corresponding detailed subregions: (1.1.1)

Northern Africa; (1.2.1) Central Africa; (1.2.2)

Eastern Africa; (1.2.3) Southern Africa; (1.2.4)

Western Africa; (2.1.1) Caribbean; (2.1.2) Central

America; (2.1.3) South America; (2.2.1) Northern

America; (3.1.1) Arab States; (4.1.1) Eastern Asia;

(4.2.1) South-Eastern Asia; (4.2.2) Pacific Islands;

(4.3.1) Southern Asia; (5.1.1) Northern Europe;

(5.1.2) Southern Europe; (5.1.3) Western Europe;

(5.2.1) Eastern Europe; (5.3.1) Central Asia; and

(5.3.2) Western Asia. There are four income group-

ings: (1) high income countries; (2) upper middle

income countries; (3) lower middle income coun-

tries; and (4) low income countries.

In the KILM database, indicators are available

for all years since 1980 and data are updated

UN Statistics Division, “Countries or areas, codes and

abbreviations”, available at: http://unstats.un.org/unsd/meth-

ods/m49/m49alpha.htm.

support are available, workers can better afford

to take the time to find more desirable jobs.

Therefore, the problem in many developing econ-

omies is not so much unemployment as rather

the lack of decent and productive work opportu-

nities for those who are employed.

This brings us to the need to dissect the total

employment number as well in order to assess the

well-being of the working population, on the prem-

ise that not all work is “decent work”. If the work-

ing population consists largely of own-account

workers or contributing (unpaid) family workers

(see KILM 3), then the indicator on the total

employed population (KILM 2) loses its value as a

normative measure. Are these people em-

ployed? Yes, according to the international defin-

ition. Are they in decent employment? Possibly not.

Although technically employed, some self-

employed workers or contributing family workers

only have a tenuous hold on employment, and the

line between employment and unemployment is

often thin. If and when salaried jobs open up in the

formal economy, this contingent workforce will

rush to apply for them. Further assessment should

also be undertaken to determine whether such

workers are generally poor (KILM 17b), engaged in

traditional agricultural activities (KILM 4), selling

goods in the informal market with no job security

(KILM 8), working excessive hours (KILM 7a) or

wanting to work more hours (KILM 12).

In an ideal world, the analysis of labour

markets using a broad range of indicators such as

those available in the KILM would be an easy

matter because the data for each indicator would

exist for each country. The reality, of course, is

quite different. Despite recent improvements in

national statistics programmes and in the effi-

ciency of collection on the part of the KILM, a

closer look at the availability of KILM data for

each country shows that many holes still exist

where data are not available.

The coverage of KILM indicators is particu-

larly low in African countries, which is under-

standable given the low priority that is likely to

be placed on conducting labour force surveys in

countries beset by poverty and political unrest.

The paradox is that this is precisely the region

where greater labour market information is

needed to inform both the allocation of scarce

funds and the creation of appropriately targeted

national policies to help people “work out of

poverty”.

Development of national statistical

The ILO strongly advocates placing employment at the

heart of poverty reduction strategies, noting, in particular, that

“it is precisely the world of work that holds the key for solid,

progressive and long-lasting eradication of poverty” (ILO, 2003).

Guide to understanding the KILM

market information through household and

establishment surveys, population censuses and

administrative records, so that the main problem

is not so much the lack of information as its

communication to the global community. In this

and previous editions of the KILM, an extensive

effort was made to tap into the existing datasets

that are increasingly being made public by

national statistical offices through the Internet.

This “data mining” process is ongoing and assists

the KILM, ILOSTAT and other ILO publications

and research programmes in expanding the

coverage of the indicators.

Notes and “breaks”

The collection of labour market indicators

requires the desire for the broadest possible

geographical coverage for a specified time period to

be weighed against the need to ensure the greatest

possible level of comparability or harmonization.

Achieving a harmonious balance between coverage

and comparability is a difficult task; the only realistic

way of reconciling the two is to provide as much

methodological information as possible, and at the

same time to “flag” the issues likely to challenge

users wishing to make valid comparisons between

countries whose statistical methodology and defin-

itions may not match in every respect. Each indica-

tor has a section on “limitations to comparability”,

and notes on methodology and sources are as

explicit as possible in each table.

Historical continuity is important for many

users of labour market information. Without over-

burdening the indicator tables, it is necessary to

alert users to significant changes in the source,

definition or coverage of the information from

year to year.

A “b” placed at the point of a chrono-

logical “break” denotes a change in the methodol-

ogy, scope of coverage and/or type of source used

within the country.

Whether the information has been obtained

from other international repositories, from

regional labour market indicator sets or directly

from official sources, a substantial effort has been

made to provide the links to the source and the

information provider wherever possible.

International comparability

To ensure international comparability, it is

necessary that international standards on labour

statistics exist. Two forms of these are recognized

by the international community: (1) Conventions

and Recommendations adopted by the

International Labour Conference; and (2) resolu-

tions and guidelines adopted by the International

Conference of Labour Statisticians (ICLS). Even

annually. The ILO makes every effort to provide

the KILM in French and Spanish as well as the

original English. These other languages are

provided in the KILM electronic versions only.

Users of the software are able to select their

language – English, French or Spanish – from the

file menu, and can switch between languages at

any time.

Information repositories

and methodological

information

In compiling the KILM, the ILO concentrates

on bringing together information from interna-

tional repositories whenever possible; for coun-

tries not included in these repositories, the infor-

mation is gathered directly from national sources.

The KILM includes compilations made by inter-

national organizations such as the following:

• ILODepartmentofStatistics(ILOSTAT)

• OrganisationforEconomicCo-operationand

Development (OECD)

• StatisticalOfficeoftheEuropeanUnion(Euro-

stat)

• WorldBank

• TheConferenceBoard

• UNESCOInstituteofStatistics

Information maintained by these organizations

has generally been obtained from national sources

or is based on official national publications.

Whenever information was available from

more than one repository, the information and

background documentation from each repository

were reviewed in order to select the data most

suitable for inclusion, based on an assessment of

the general reliability of the sources, the availabil-

ity of methodological information and explana-

tory notes regarding the scope of coverage, the

availability of information by sex and age, and the

degree of historical coverage. Occasionally, two

data repositories have been chosen and presented

for a single country; any resulting breaks in the

historical series are duly noted.

For countries with less developed labour

market information systems, such as those in the

developing economies, information may not be

easily available to national policy-makers and

social partners, let alone to international organi-

zations seeking to compile global datasets. Many

of these countries, however, do collect labour

Guide to understanding the KILM

be affected by the precision of the measurements

made for each country and year, and by system-

atic differences between sources in respect of

the methodology of collection, definitions, scope

of coverage and reference period.

In order to minimize misinterpretation,

detailed notes are provided that identify the

repository, type of source (household/labour

force survey, census, administrative records, etc.),

and changes or deviations in coverage, such as

age groups, geographical coverage (national,

urban, capital city), etc. When analysing or

making reference to a particular indicator, users

are advised to examine closely the section

“limitations to com-parability” and the notes to

the data tables.

though these resolutions are non-binding, they

provide detailed guidelines on conceptual frame-

works, operational definitions and measurement

methodologies for the production and dissemin-

ation of the various labour statistics.

As noted above, there will always be import-

ant caveats relating to the methodologies of

measurement; these require time and effort to

sort out before reasonable international compar-

isons can be made. Limitations to comparability

are often indicator-specific; however, there are

standard issues that require attention with every

indicator. For example, comparisons will certainly

For the most recent relevant ICLS resolution, see box 1c

below.

Box 1b. ILO methodology for producing global and regional

estimates of labour market indicators

The biggest challenge in the production of aggregate estimates is that of missing data. In an ideal

world, producing global and regional estimates of labour market indicators, for example employment,

would simply require summing up the total number of employed persons across all countries in the

world or within a given region. However, because not all countries report data in every year and,

indeed, some countries do not report data for any year at all, it is not possible to derive aggregate

estimates of labour market indicators by merely summing across countries.

To address the problem of missing data, the former ILO Employment Trends Team designed several

econometric models which are actively maintained and used to produce estimates of labour market

indicators in the countries and years for which real data are not available. The Global Employment

Trends Model (GET Model) is used to produce estimates – disaggregated by age and sex – of

employment-to-population ratio, status in employment, employment by sector, unemployment, youth

unemployment and labour productivity (KILMs 2, 3, 4, 9, 10 and 16). The econometric model

described in KILM 17 is used to produce estimates on employment by economic class. The global

and regional labour force estimates found in KILM 1 and KILM 13 are estimated using the Trends

Labour Force model (TLF model).

Each of these models uses multivariate regression techniques to impute missing values at the

country level. The first step in each model is to assemble every known piece of real information (i.e.

every real data point) for each indicator in question. Only data that are national in coverage and

comparable across countries and over time are used as inputs. This is an important selection

criterion when the models are run, because they are designed to use the relationship between the

various labour market indicators and their macroeconomic correlates – such as GDP per capita,

GDP growth rates, demographic trends, country membership in the Heavily Indebted Poor Countries

initiative (HIPC), geographical indicators, and country and time dummy variables – in order to

produce estimates of the labour market indicators where no data exist. Thus, the comparability of

the labour market data that are used as inputs in the imputation models is essential to ensure that

the models accurately capture the relationship between the labour market indicators and the

macroeconomic variables.

The last step of the estimation procedure occurs once the data sets containing both real and im-

puted labour market data have been assembled. In this step, the data are aggregated across

countries to produce the final world and regional estimates. For further information on the Trends

Econometric Models (including the GET and TLF models), readers can consult the technical

background papers available at the following website: http://www.ilo.org/empelm/projects/

WCMS_114246/lang--en/index.htm.

Guide to understanding the KILM

factors such as military service requirements. The

series includes both nationally reported and

imputed data and only estimates that are national,

meaning there are no geographical limitations on

coverage. Table 1b contains labour force partici-

pation rates as nationally reported by sex and age

group: total (15+), youth (15−24) and adult (25+),

where available.

KILM 2. Employment-to-population ratio

The employment-to-population ratio is

defined as the proportion of a country’s working-

age population that is employed (the youth

employment-to-population ratio is the propor-

tion of the youth population – typically defined

as persons aged 15−24 – that is employed).

high ratio means that a large proportion of a

country’s population is employed, while a low

ratio means that a large share of the population is

not involved directly in labour market related

activities, either because they are unemployed or

(more likely) because they are out of the labour

force altogether. Table 2a provides a harmonized

series of employment-to-population ratios as

estim ated and projected by the ILO (like table 1a)

by sex and age group: total (15+), youth (15−24)

and adult (25+). Table 2b contains national estim-

ates of employment-to-population ratios, also by

sex and age group, where available.

The employment-to-population ratio provides

information on the ability of an economy to create

employment; for many countries the indicator

offers more insight than the unemployment rate.

Although a high overall ratio is typically consid-

ered positive, this indicator alone is not sufficient

for assessing the level of decent work or of decent

work deficit: additional indicators are required to

assess such issues as earnings, hours of work,

informal employment, under employment and

working conditions. Employment-to-population

ratios are of particular interest when broken

down by sex, as the ratios for men and women

can provide information on gender differences in

labour market activity in a given country.

KILM 3. Status in employment

Indicators of status in employment distin-

guish between the two main categories of the

employed: (1) employees (also known as wage

and salaried workers) and (2) the self-employed.

The self-employed are further disaggregated

into (a) employers, (b) own-account workers,

(d) contributing family workers. Each of these

categories is expressed as a proportion of the

total number of employed persons. Categorization

by employment status can help in understanding

Global and regional estimates

The ninth edition of the KILM offers users

direct access to ILO global and regional estimates

from 1991 to the present. Tables are presented for

the following indicators: labour force participa-

tion (table R1), employment-to-population ratio

(R2), status in employment (R3), employment by

sector (R4), unemployment rate (R5), youth

unemployment rate (R6), ratio of youth un-

employment rate (R7), labour productivity (R8)

and employment by economic class (R9).

Like other KILM tables based on country-level

data, several of these data sets (R1, R2, R7 and R9)

can be filtered according to year, sex and age

group; users will have access to both raw numbers

and rates. The estimates are derived using one of

three models which apply multivariate regression

techniques to impute missing values at the coun-

try level. The processes used in the ILO global

and regional estimation models are described in

detail in box 1b.

Summary of the 17 ILO Key

Indicators of the Labour

Market

The ninth edition of the KILM provides indi-

cators related to labour force, employment,

unemployment, underemployment, educational

attainment, wages and compensation costs,

productivity and poverty. Each of the 17 indica-

tors is briefly described below.

KILM 1. Labour force participation rate

The labour force participation rate is a

measure of the proportion of a country’s work-

ing-age population that engages actively in the

labour market, either by working or by looking

for work; it provides an indication of the relative

size of the supply of labour available to engage in

the production of goods and services. The break-

down of the labour force (formerly known as

economically active population) by sex and age

group gives a profile of the distribution of the

labour force within a country.

Table 1a contains labour force participation

rate estimates and projections by sex, for the

following standardized age groups: 15+, 15−24,

15−64, 25−34, 25−54, 35−54, 55−64 and 65+, and

for the years 1990 to 2030. The participation rates

are harmonized to account for differences in

national data collection and tabulation method-

ologies as well as for other country-specific

Guide to understanding the KILM

There is widespread interest in this indicator.

Economists use occupation in the analysis of

differences in the distribution of earnings and

incomes over time and between groups – men

and women, for example – as well as in the ana-

lysis of imbalances of supply and demand in

different labour markets. Policy-makers use occu-

pational statistics in support of the formulation

and implementation of economic and social pol-

icies and to monitor progress with respect to

their application, for example in respect of labour

planning and the planning of educational and

vocational training. Managers need occupational

statistics for planning and deciding on personnel

policies and monitoring working conditions,

both at the enterprise level and in the context of

their industry and relevant labour markets.

KILM 6. Part-time workers

There has been rapid growth in part-time

work in the past few decades in the developed

economies. This trend is related to the increase in

the number of women in the labour market, but

also to attempts to introduce labour market flexi-

bility in response to changes in work organiza-

tion within industry, and to the growth of the

services sector.

The indicator on part-time workers focuses

on individuals whose working hours total less

than “full time”, as a proportion of total employ-

ment. Because there is no agreed international

definition as to the minimum number of hours in

a week that constitute full-time work, the dividing

line is determined either on a country-by-country

basis or through the use of special estimations.

Two measures are calculated for this indicator:

total part-time employment as a proportion of

total employment, sometimes referred to as the

“part-time employment rate” or the “incidence of

part-time employment”; and the percentage of

the part-time workforce composed of women.

KILM 7. Hours of work

The number of hours worked has an impact

on the health and well-being of workers as well

as on levels of productivity and labour costs of

establishments. Measuring levels of and trends in

hours worked in a society, for different groups of

workers and for workers individually, is therefore

important when monitoring working and living

conditions as well as when analysing economic

developments.

Two measurements related to working time

are included in KILM 7 in order to give an overall

picture of the time that the employed throughout

the world devote to work activities. The first

both the dynamics of the labour market and the

level of development in any particular country.

Over the years, and with economic growth, one

would typically expect to see a shift in employ-

ment from agriculture to the industrial and

services sectors, with a corresponding increase

in wage and salaried workers and concomitant

decreases in self-employed and contributing

family workers, many of whom will have previ-

ously been employed in the agricultural sector.

The method of classifying employment by

status is based on the 1993 International

Classification by Status in Employment (ICSE),

which classifies the job held by a person at a

point in time with respect to the type of explicit

or implicit employment contract that person has

with other persons or organizations. Such status

classifications reflect the degree of economic risk

entailed in these various types of arrangements,

an element of which is the strength of the attach-

ment between the person and the job, and the

type of authority over establishments and other

workers that the person has or will have.

KILM 4. Employment by sector

This indicator disaggregates employment into

three broad sectors – agriculture, industry and

services – and expresses each as a percentage of

total employment. The indicator shows employ-

ment growth and decline on a broad sectoral

scale, while also highlighting differences in trends

and levels between developed and developing

economies. Sectoral employment flows are an

important factor in the analysis of productivity

trends, because within-sector productivity

growth needs to be distinguished from growth

resulting from shifts from lower to higher produc-

tivity sectors. The addition of further sectoral

detail in tables 4b, 4c and 4d is useful for demon-

strating trends of employment within individual

sectors of the economy.

The sectors of economic activity are defined

according to the International Standard Industrial

Classification of All Economic Activities (ISIC),

Revision 2 (1968), Revision 3 (1990) and Revision 4

(2008).

KILM 5. Employment by occupation

Employment by occupation is presented

according to major classification groups in three

tables: table 5a according to the International

Standard Classification of Occupation, 2008

(ISCO-08); table 5b according to ISCO-88; and

table 5c according to ISCO-68.

All three tables are

disaggregated by sex.

Guide to understanding the KILM

terms of labour markets for countries that regu-

larly collect information on the labour force. The

unemployment rate tells us the proportion of the

labour force that does not have a job, is available

to work and is actively looking for work. It should

not be misinterpreted as a measurement of

economic hardship, although a correlation often

exists. Table 9a provides a harmonized series of

unemployment rates as estimated by the ILO

(like tables 1a and 2a) by sex; table 9b contains

national estimates on total unemployment by sex,

where possible; and table 9c shows flows in and

out of unemployment, measured by the probabil-

ity (hazard rate) of losing a job once employed or

finding a job once unemployed.

The resolution concerning statistics of work,

employment and labour underutilization adopted

by the 19th ICLS, which updates and replaces the

resolution concerning statistics of the econom-

ically active population, employment, unemploy-

ment and underemployment adopted by the 13th

ICLS, defines the unemployed as all persons of

working age who, during the reference period,

were without work, currently available for work

and seeking work. However, it should be recog-

nized that national definitions and coverage of

unemployment can vary with regard to factors

such as age limits, criteria for seeking work, and

treatment of, for example, persons temporarily

laid off, discouraged about job prospects or seek-

ing work for the first time.

KILM 10. Youth unemployment

Youth unemployment is an important policy

issue for many countries at all stages of develop-

ment. For the purpose of this indicator, the term

“youth” covers persons aged 15−24, while “adults”

are defined as persons aged 25 and over, although

national variations in age definitions do occur.

The indicator presents youth unemployment in

the following four ways: (a) the youth unemploy-

ment rate; (b) the ratio of the youth unemploy-

ment rate to the adult unemployment rate; (c) the

youth share in total unemployment; and (d) youth

unemployment as a proportion of the youth

population.

The KILM 10 measures should be analysed

together; any of the four, analysed in isolation,

could present a distorted image. For example, a

country might have a high ratio of youth-to-adult

unemployment but a low youth share in total

unemployment. The presentation of youth un-

employment as a proportion of the youth popula-

tion recognizes the fact that a large proportion of

young people enter unemployment from outside

the labour force. Taken together, the four indica-

tors provide a fairly comprehensive indication of

measure relates to the hours an employed person

works per week (table 7a). This table shows

numbers of employed classified according to

their weekly hours of work, using the following

bands: less than 15 hours worked per week,

15−29 hours, 30−34 hours, 35−39 hours, 40−48

hours, and 49 hours and over, as available. The

data are broken down by sex, age group (total,

youth and adult) and employment status (total,

and employees or wage and salaried workers),

wherever possible. The second measure is the

average annual actual hours worked per person

(table 7b).

KILM 8. Employment in the informal

economy

The informal economy plays a major role in

employment creation, income generation and

production in many countries. In countries with

high rates of population growth or urbanization,

the informal economy tends to absorb most of

the growth in the labour force. Work in the infor-

mal economy is generally recognized as entailing

absence of legal identity, poor working condi-

tions, lack of membership in social protection

systems, higher incidence of work-related acci-

dents and ailments, and limited freedom of asso-

ciation. Knowing how many people are in the

informal economy is a starting point for consider-

ing the extent and content of policy responses

required.

KILM 8 includes national estimates of infor-

mal employment. Table 8 combines two measures

of the informal economy: employment in the

informal sector, the enterprise-based measure

defined by the 15th ICLS; and informal employ-

ment, the broader job-based measure recom-

mended in the 17th ICLS. The latter includes both

persons employed in informal sector enterprises

and persons in informal employment outside the

informal sector (employees holding informal

jobs), as well as contributing family workers in

formal or informal sector enterprises and own-

account workers engaged in the production of

goods for own end-use by their household.

Informal employment and its subcategories are

presented as a share of total non-agricultural

employment.

KILM 9. Unemployment

The unemployment rate is probably the best-

known labour market measure and certainly one

of the most widely quoted by the media in many

countries. Together with the labour force partici-

pation rate (KILM 1) and employment-to-popula-

tion ratio (KILM 2), it provides the broadest avail-

able indicator of economic activity and status in

Guide to understanding the KILM

ICLS, amended in 1998 by the 16th ICLS and

further clarified by the 19th ICLS in 2013. It

includes all persons in employment who “wanted

to work additional hours, whose working time in

all jobs was less than a specified hours threshold,

and who were available to work additional hours

given an opportunity for more work”.

The indicator is important for improving the

description of employment-related problems, as

well as for assessing the extent to which available

human resources are being used in the produc-

tion process of the country concerned. It also

provides useful insights for the design and evalu-

ation of employment, income and social

programmes. The indicator is calculated as time-

related underemployment as a percentage of

total employment.

KILM 13. Persons outside the labour

force

The inactivity rate is defined as the percent-

age of the population that is neither working nor

seeking work (that is, not in the labour force).

Inactivity rates for the age groups 15+, 15−24,

15−64, 25−34, 25−54, 35−54, 55−64 and 65+ are

shown in table 13. The 25−54 age group can be

of particular interest since it is considered to be

the “prime” age band, representing individuals

who are generally expected to be in the labour

force, having normally completed their education

and not yet reached retirement age; it is therefore

worth investigating why these potential labour

force participants are inactive. The inactivity rate

of women, in particular, tells us a lot about the

social customs of a country, attitudes towards

women in the labour force, and family structures

in general.

When the inactivity rate is added to the labour

force participation rate (KILM table 1a) for the

corresponding group, the total will equal 100 per

cent. Data in table 13 have been harmonized to

account for differences in national data collection

and tabulation methodologies as well as for other

country-specific factors such as military service

requirements. The series includes both nationally

reported and imputed data and only estimates

that are national, meaning there are no geograph-

ical limitations in coverage.

KILM 14. Educational attainment

and illiteracy

An increasingly important aspect of labour

market performance and national competitive-

ness is the skill level of the workforce. Information

on levels of educational attainment is currently

the best available indicator of labour force skill

the problems that young people face in finding

jobs. Table 10a provides a harmonized series of

youth unemployment rates as estimated by the

ILO (like tables 1a, 2a and 9a) by sex; table 10b

contains national estimates on total youth un-

employment by sex, where possible. Table 10c

complements the labour market situation of

youth by showing the number of young people

who are not in employment, education or train-

ing (NEET) as a percentage of the youth popula-

tion. The NEET rate is presented for youth aged

15−24 unless otherwise indicated in the notes.

KILM 11. Long-term unemployment

Unemployment tends to have more severe

effects the longer it lasts. Short periods of jobless-

ness can normally be dealt with through un-

employment compensation, savings and, perhaps,

assistance from family members. Unemployment

lasting a year or longer, however, can cause

substantial financial hardship, especially when

unemployment benefits either do not exist or

have been exhausted. Long-term unemployment

is not generally viewed as an important indicator

for developing economies, where the duration of

unemployment often tends to be short, given the

lack of unemployment compensation and the

fact that most people therefore cannot afford to

be without work for long periods. Accordingly,

most of the information available for this indica-

tor comes from the more developed economies.

The data are presented by sex and age group

(total, youth and adult), wherever possible.

Table 11a includes two separate measures of

long-term unemployment: (a) those unemployed

for one year or more as a percentage of the labour

force; and (b) those unemployed for one year or

more as a percentage of the total unemployed

(the incidence of long-term unemployment).

Table 11b includes the number of unemployed

(as well as their share of total unemployed) at

different durations: (a) less than one month;

(b) one month to less than three months; (c) three

months to less than six months; (d) six months to

less than 12 months; (e) 12 months or more. Data

are disaggregated by sex and age group (total,

youth and adult).

KILM 12. Time-related

underemployment

Underemployment reflects underutilization of

the productive capacity of the labour force. Time-

related underemployment is the first component

of underemployment to have been agreed upon

and properly defined within the international

community of labour statisticians. The interna-

tional definition was adopted in 1982 by the 13th

Guide to understanding the KILM

Table 15a presents trends in average monthly

wages, in both nominal and real terms (i.e. adjusted

for changes in consumer prices). Both the nominal

and real average wage series are presented in

national currency. This enables data users to calcu-

late nominal and real wage growth rates without

the distortion caused by exchange rate fluctu-

ations, and to link wage data to other data

expressed in national currency. Table 15b is

concerned with the levels, trends and structures

of employers’ hourly compensation costs for the

employment of workers in the manufacturing

sector. Total compensation is also broken down

into “hourly direct pay” with subcategories “pay for

time worked”, “directly paid benefits” and “social

insurance expenditure and labour-related taxes”;

here all variables are expressed in US dollars.

KILM 16. Labour productivity

Productivity, in combination with hourly

compensation costs, can be used to assess the

international competitiveness of a labour market.

Economic growth in a country or sector can be

ascribed either to increased employment or to

more effective work by those who are employed.

The latter can be described through data on

labour productivity. Labour productivity, there-

fore, is a key measure of economic performance.

An understanding of the driving forces behind it,

in particular the accumulation of machinery and

equipment, improvements in organization and in

physical and institutional infrastructures,

improved health and skills of workers (“human

capital”) and the generation of new technology,

is important in formulating policies to support

economic growth.

Labour productivity is defined as output per

unit of labour input. Two measures are presented

in table 16a: GDP per person engaged and GDP

per hour worked, both in 1990 US dollars and

indexed to 1990 = 100 with information taken

from The Conference Board. Table 16b presents

ILO estimates of labour productivity expressed as

GDP per person engaged in 2005 international

dollars at PPP as well as in 2005 constant US

dollars at market exchange rates.

KILM 17. Poverty, income distribution

and the working poor

Poverty can result when individuals are

unable to generate sufficient income from their

labour to maintain a minimum standard of living.

The extent of poverty, therefore, can be viewed

as an outcome of the functioning of labour

markets. Because labour is often the most signifi-

cant, if not the only, asset of individuals in poverty,

the most effective way to improve the level of

levels. These are important determinants of a

country’s capacity to compete successfully in

world markets and to make efficient use of rapid

technological advances; they are also among the

factors determining the employability of workers.

Table 14a presents information on the educa-

tional attainment of the labour force, with data

broken down by sex and age group (total, youth

and adult) wherever possible. Table 14b presents

the distribution of the unemployed population

by level of educational attainment, with data

broken down by sex and age group (total, youth

and adult) wherever possible. Table 14c presents

the unemployment rates of persons who attained

education at, respectively, primary level or less,

secondary level or tertiary level. The categories

used in the three indicators are conceptually

based on the levels of the International Standard

Classification of Education (ISCED). ISCED was

for assembling, compiling and presenting com-

parable indicators and statistics of education,

both within countries and internationally. Finally,

table 14d is a measure of illiteracy in the popu -

l ation (total, youth and adult).

KILM 15. Wages and compensation

costs

Wages represent a measure of the level and

trend of workers’ purchasing power and an

approximation of their standard of living.

Compensation costs provide an estimate of

employers’ expenditure on the employment of

their workforce. The indicators are, in this sense,

complementary in that they reflect the two main

facets of existing wage measures; one aiming to

track the income of employees, the other show-

ing the costs incurred by employers for employ-

ing them. Information on average wages repre-

sents one of the most important elements of

labour market information. Because wages are a

substantial form of income, accruing to a high

proportion of the economically active popula-

tion, information on wage levels is essential to

evaluate the living standards and conditions of

work and life of this group of workers in both

developed and developing economies.

Average hourly compensation cost is a measure

intended to represent employers’ expenditure on

the benefits granted to their employees as compen-

sation for an hour of labour. These benefits accrue

to employees either directly – in the form of total

gross earnings, or indirectly – in terms of employ-

ers’ contributions to compulsory, contractual and

private social security schemes, pension plans,

casualty or life insurance schemes and benefit

plans in respect of their employees.

Guide to understanding the KILM

Box 1c. Resolution concerning statistics of work, employment

and labour underutilization

In October 2013, the 19th ICLS adopted a “resolution concerning statistics of work, employment and

labour underutilization” in which several concepts in the world of work are redefined and new ones

are introduced (ILO, 2013). The progressive implementation of this resolution will bring about several

changes in how statistics are compiled.

Even though there are no immediate changes to the data in the KILM (statistics such as employment

and unemployment are based on concepts that remain unchanged at their core, despite the expansion

of the overall labour market framework and the introduction of new measures of labour underutilization),

the new resolution will affect the future compilation of labour market statistics, particularly in terms

of indicators related to the concept of work, and forms of work other than employment.

A substantial change to the statistics on employment is the introduction of “five mutually exclusive

forms of work [that are] identified for separate measurement. These forms of work are distinguished

on the basis of the intended destination of the production (for own final use; or for use by others, i.e.

other economic units) and the nature of the transaction (i.e. monetary or non-monetary transactions,

and transfers), as follows:

(a) own-use production work comprising production of goods and services for own final use;

(b) employment work comprising work performed for others in exchange for pay or profit;

experience or skills;

(d) volunteer work comprising non-compulsory work performed for others without pay;

(e) other work activities not defined in this resolution” (para. 7).

Furthermore: “Persons in employment are defined as all those of working age who, during a short

reference period, were engaged in any activity to produce goods or provide services for pay or profit.

They comprise:

(a) employed persons ‘at work’, i.e. who worked in a job for at least one hour;

(b) employed persons ‘not at work’ due to temporary absence from a job, or to working-time arrange-

ments (such as shift work, flexitime and compensatory leave for overtime)” (para. 27).

The resolution extends the definition of unemployment to include examples of “activities to seek

employment” and three specifically defined groups of jobseekers:

(a) future starters defined as persons ‘not in employment’ and ‘currently available’ who did not ‘seek

employment’ … because they had already made arrangements to start a job within a short subse-

quent period, set according to the general length of waiting time for starting a new job in the

national context but generally not greater than three months;

(b) participants in skills training or retraining schemes within employment promotion programmes,

who on that basis, were ‘not in employment’, not ‘currently available’ and did not ‘seek employ-

ment’ because they had a job offer to start within a short subsequent period generally not greater

than three months;

or profit but who were still waiting for the opportunity to leave” (para. 48).

The definition for persons in time-related underemployment was also extended to define this group

of people as “all persons in employment who, during a short reference period, wanted to work

additional hours, whose working time in all jobs was less than a specified hours threshold, and who

were available to work additional hours given an opportunity for more work, where:

(a) the ‘working time’ concept is hours actually worked or hours usually worked, dependent on the

measurement objective (long- or short-term situations) and in accordance with the international

statistical standards on the topic;

(b) ‘additional hours’ may be hours in the same job, in an additional job(s) or in a replacement job(s);

the median or modal values of the hours usually worked of all persons in employment, or on

working time norms as specified in relevant legislation or national practice, and set for specific

worker groups;

Guide to understanding the KILM

have Internet access will be notified by email of

the availability of updates, once they have filled

in the registration material. Users can download

the KILM programme from www.ilo.org/kilm.

The KILM database can also be directly

accessed through the KILM web page, making

access to country-level data for the 17 key labour

market indicators, as well as the descriptive text

explaining their use, definitions and basic trends,

easier than ever. Users can run quick and easy

searches of KILM indicators, and display and

export data in spreadsheet format, directly from

the Internet. As with the software, direct access

to the KILM indicators is available through www.

ilo.org/kilm.

References

International Labour Office (ILO). 1999. Decent

Work, Report of the Director-General,

International Labour Conference, 87th

Session, Geneva, 1999 (Geneva).

—.

2003. Working out of poverty, Report of the

Director-General, International Labour Con-

ference, 91st Session, Geneva, 2003 (Geneva).

—. 2007. KILM, 4th edn (Geneva).

—. 2009. KILM, 6th edn (Geneva).

—. 2010. Women in labour markets: Measuring

progress and identifying challenges (Geneva).

Available at: http://www.ilo.org/empelm/

pubs/WCMS_123835/lang--en/index.htm.

—. 2011. KILM, 7th edn (Geneva).

—. 2013. Resolution concerning statistics of

work, employment and labour underutiliza-

tion, adopted by the 19th International

Conference of Labour Statisticians, Geneva,

Oct. Available at: http://www.ilo.org/global/

statistics-and-databases/standards-and-guide-

lines/resolutions-adopted-by-international-

conferences-of-labour-statisticians/

WCMS_230304/lang--en/index.htm.

well-being is to increase employment opportuni-

ties and labour productivity through education

and training.

Any estimate of the number of people in

poverty in a country depends on the choice of

the poverty threshold. What constitutes such a

threshold of minimum basic needs is a subjective

judgement, varying with culture and national

priorities. Definitional variations create difficul-

ties in making international comparisons.

Therefore, in addition to national poverty

measurements and the Gini index shown in table

17a, this indicator presents data on employment

by economic class, showing individuals who are

employed and who fall within the per capita

consumption thresholds of a given economic

class group. By combining labour market charac-

teristics with household consumption group

data, estimates of employment by economic

class give a clearer picture of the relationship

between economic status and employment.

Because of the important linkages between

employment and material well-being, evaluating

these two components side by side also provides

a more detailed perspective on the dynamics of

productive employment generation, poverty

reduction and growth in the middle class

throughout the world.

KILM electronic versions

The ILO hopes to reach a wider audience by

presenting KILM data in electronic form. As in

previous editions, the electronic version of this

ninth edition of the KILM contains all the data

sets for the indicators, together with an Excel

add-in and interactive software through which

users can select and query the indicators by

country, year, type of source and other user-

defined functions according to specific needs.

Data updates will be automatically downloaded

each time a user opens the programme (if

connected to the Internet). Users who do not

(d) ‘available’ for additional hours should be established in reference to a set short reference period

that reflects the typical length of time required in the national context between leaving one job

and starting another” (para. 43).

For further information and details on the resolution, see http://www.ilo.org/global/statistics-and-

databases/meetings-and-events/international-conference-of-labour-statisticians/19/lang--en/index.htm.

(Box 1c, continued)

Guide to understanding the KILM

plenary meeting of the 60th Session of the

General Assembly, 20 Sep.,

A/60/L.1 (New

York).

World Commission on the Social Dimension of

Globalization (WCSDG). 2004. A fair globaliza-

tion: Creating opportunities for all (Geneva).

Available at: http://www.ilo.org/public/

english/fairglobalization/index.htm.

Sparreboom, T.; Albee, A. (eds). 2011. Towards

decent work in sub-Saharan Africa: Monitoring

MDG employment indicators (Geneva, ILO).

Available at: http://www.ilo.org/global/publica-

tions/ilo-bookstore/order-online/books/

WCMS_157989/lang--en/index.htm2.

United Nations (UN). 2005. World Summit

outcome, resolution adopted by the high-level

environments, workers with greater skills than

those required for their position are more likely to

be found in permanent jobs than in temporary

ones (Ortiz, 2010). Education therefore can provide

protection, to a certain extent, from vulnerable

employment. One study found that the proportion

of youth in vulnerable employment educated at

only primary level or below is greater than that of

similarly educated youth in non-vulnerable employ-

ment (Sparreboom and Staneva, 2014).

At the national level, there is a positive correla-

tion between the proportion of highly educated

adults in the labour force in a given country and

that country’s income per capita (OECD and Stat-

istics Canada, 2000; Holland et al., 2013). A study

involving 18 developing countries found that in

most of the countries analysed, an increase

in national literacy rates was accompanied by a

higher rate of national economic growth. That is,

human capital has a statistically significant positive

impact on economic growth (Vinod and Kaushik,

2007). In addition, higher levels of educational

attainment are associated with lower income

inequalities, and national expenditure (per student)

in education strongly influences a country’s income

distribution (Keller, 2010).

Studies on these critical linkages between

education and labour markets tend to focus on

developed economies. Less is known about the

corresponding dynamics in the developing world;

however, Keller’s finding cited above is particularly

marked for less developed countries. Given that

levels of educational attainment remain compara-

tively low in many developing countries, further

exploration of such linkages is vital (ILO, 2015).

To this end, the ninth edition of the KILM

includes four indicators that directly examine the

link between education and the labour market,

presenting time series for a large number of coun-

tries at all stages of development. These four indi-

cators correspond to table 14a, which looks at the

labour force by level of educational attainment,

disaggregated by sex and age group (total, youth

and adult); table 14b, covering unemployment by

level of educational attainment, disaggregated

by sex and age group; table 14c, which shows

unemployment rates by level of educational

1. Introduction

Education and training are at the core of any

effort to increase a country’s productivity and to

improve people’s likelihood not only of accessing

employment at all, but of accessing good quality

employment. The educational attainment and skills

base of the workforce have a clear impact at both

the individual and the national level. Accordingly,

effective policy-making depends on understanding

the ways in which educational trends and labour

market trends are related, and how these shape

individual and national well-being.

In general terms, higher levels of education are

associated with greater labour market success,

enhancing the opportunities for individuals to

enter the labour market in a better position and

protect them from unemployment. Where highly

educated individuals are unemployed, this may in

some cases reflect their unwillingness to settle for

jobs of lesser quality than they deem appropriate

based on their skill level. The effect of educational

attainment on individuals’ labour market outcomes

refers not only to improved access to employment,

but also to the quality of their employment in terms

of working conditions. Higher levels of educational

attainment are associated with higher salaries. Even

overeducated individuals (those who have a higher

skill level than is requiered for their job) earn more,

overall, than those doing the same job with no

more than the skills needed (Rubb, 2003).

Educational attainment also influences other

crucial aspects of working conditions, such as the

type of contract and working time arrangements.

A higher educational level can put workers in a

better position to negotiate more satisfactory

terms of employment. However, in highly

segmented labour markets, where casual work and

temporary contracts are common and formal

permanent contracts are not widespread, human

capital might be traded off for job security. In such

The analysis of this section was prepared by Rosina

Gammarano and Yves Perardel with the support of colleagues in

the ILO Department of Statistics. Helpful comments on this

section were provided by Laura Brewer, Sara Elder, Lawrence Jeff

Johnson, Sangheon Lee, Sandra Polaski and Theo Sparreboom.

Education and labour

markets: Analysing global

patterns with the KILM

Education and labour markets: Analysing global patterns with the KILM

tries. These countries have been chosen to repre-

sent all income groups, as defined in the World

Bank classification of countries by income (based

on GNI per capita), that is, low income econo-

mies, lower middle income economies, upper

middle income economies and high income

economies. Data for all 12 countries are available

in the KILM, and the group also covers all regions

of the world. Finally, section 5 briefly concludes.

2. Global trends

by indicator

In this section, we present the four KILM indi-

cators at the centre of our analysis. We start by

comparing, for all countries with available data,

the situation in 2000 (or the closest year avail-

able) to that in the latest year available.

The aim

is to understand what has changed during the

past 15 years, and to reveal any trends for these

key indicators.

2.1. Labour force distribution

by level of educational attainment

Table 14a of the KILM provides data on the

distribution of the labour force by level of educa-

In cases where data were not available for 2000, we

selected the closest year for which data were available

between 1997 and 2007, unless otherwise stated. The latest

year available is always after 2009.

attainment, also disaggregated by sex and age

group; and table 10c, containing rates for those

not in education, employment or training (NEET),

disaggregated by sex.

Throughout this chapter, we use the wealth of

data included in the ninth edition of the KILM to

explore the linkages between education and the

labour market, and, more specifically, to examine

whether the expected relationships between

educational attainment and labour market

outcomes are confirmed by the available data.

The KILM offers the advantage of allowing us to

conduct this study at the global level, revealing

patterns for both developed and developing

economies, and across regions.

Section 2 of the chapter begins this explora-

tion by analysing the trends observed during the

past 10−15 years for the four indicators identified

above for all countries for which data are avail-

able. This provides a general picture of the recent

evolution of the educational attainment of the

labour force worldwide. Section 3 investigates

more closely the linkages between education,

labour market outcomes and economic perfor-

mance, by comparing the four education-related

indicators to other KILM indicators, namely,

labour productivity (table 16a), the employment-

to-population ratio (table 2b) and status in

employment (table 3).

Section 4 looks at the same indicators, focus-

ing in more detail on a set of 12 selected coun-

Box 1.1. Data on the labour market and education: Statistical issues

There are a number of challenges associated with the use of data on education, and in particular, of

labour market data in relation to education. The first pertains to data availability. The preferred source

for this type of data is a labour force survey, since it provides reliable information on both the

educational attainment and the labour market status of individuals. Other types of household surveys

and population censuses can also be used to derive these data. This means that, in general terms,

it can be hard to obtain reliable and frequent statistics on the labour force by educational attainment

for those countries that do not have a regular labour force or household survey in place.

Other key challenges relate to the international comparability of education statistics. The configuration

of the national educational system, the levels of education required in the workplace and even

traditions in terms of education are all heavily dependent on the national context. Even though there

is an internationally agreed standard classification of educational levels (the International Standard

Classification of Education, of which the latest version dates from 2011), it cannot be assumed that

educational categories used at the national level always accurately match the categories in this

standard classification.

There is also the potential for confusion as to how a person’s educational level is to be defined.

Ideally, when making cross-country comparisons, all data should refer to the highest level of education

completed, rather than the level in which the person is currently enrolled, or the level begun, but not

successfully completed. However, because data are usually derived from household surveys, the

actual definition ultimately used will inevitably depend on each respondent’s own interpretation.

Education and labour markets: Analysing global patterns with the KILM

upper middle income economies and low and

lower middle income economies have tended to

experience more significant improvements,

though starting from a lower educational base.

The decrease in the share of the labour force

with primary or less than primary level educa-

tional attainment is particularly striking for Macau

(China) and the Occupied Palestinian Territory,

where it fell by around 30 percentage points. The

corollary of this, of course, is that the proportions

of the population attaining some higher level of

education have increased.

It is also crucial to study the changes in the

share of the labour force with a higher level of

educational attainment, in order to establish what

tional attainment. Figure 2.1 focuses specifically

on the share of the labour force with primary or

less than primary level educational attainment.

The key finding is that, worldwide, the educa-

tional level of the labour force is improving. Of

the 64 countries for which data are available, only

three registered an increase in the share of the

labour force that has attained no higher than

primary educational level. Among the developed

economies, the situation does not seem to have

changed appreciably. In most of these countries,

the share of the labour force with no higher than

primary level educational attainment was already

quite low in 2000, and has declined only moder-

ately over the following 15 years. Conversely,

Figure 2.1 Share of the labour force with primary or less than primary level educational

attainment (%)

Note: In all the scatter plots included in this chapter, countries are referred to by the corresponding ISO 3166 alpha-3 country code. The full list

of codes is presented in the annex at the end of this chapter, and is also accessible at https://www.iso.org/obp/ui/#search.

Source: KILM, 9th edn, table 14a, ages 15+, 1997−2004 and latest year available after 2009.

010 20 30 40 50 60 70 80 90 100

100

GEO

RUS

CZE

USA

SVK

2000

High income Upper-middle income Low and lower-middle income

Latest year

GBR

LVA

CAN

SVN

EST

HUN

NOR

LTU

POL

DEU

SWE

HRV

FIN

FRA

BHS

CHE

AUT

DNK

IRL

CYP

BEL

SGP

HKG

NLD

MDA

BGR

LUX

ROU

PAN

MEX

REU

GLP

MTQ

PSE

GUF

MLT

DOM

MAC

BRA

PRT

BLZ

YEM

MDG

CRI

ITA

MNG

URY

ALB

SMR

ESP

ISL

GRC

IND

TUR

NAM

LKA

IDN

MAR

ETH

Education and labour markets: Analysing global patterns with the KILM

for which data are available, 62 registered an

increase: most impressively, Canada, Luxembourg

and the Russian Federation, where the increase

exceeded 20 percentage points, taking these

three countries to the top of the rankings in

terms of share of the labour force with tertiary

education in the latest year available. In contrast,

only two countries, Mexico and Yemen, experi-

enced a (modest) decrease in the share of the

labour force with tertiary education.

One special consideration to keep in mind

when looking at data referring to tertiary educa-

tion is the difference between vocational

e ducation and university degrees, since voca-

tional education plays a significant role in

productivity and sustainable growth for many

countries. However, whether vocational educa-

levels of further educational attainment have

been achieved where the share of workers with

primary education or less has fallen – in particu-

lar, how many people are achieving tertiary level

educational attainment. An increase in the share

of a country’s labour force with tertiary level

education could facilitate an expansion in

production of higher value added goods and

services and a speeding up of productivity

growth, thereby supporting economic growth

and development. Accordingly, the changes in

tertiary level educational attainment are shown

in figure 2.2.

In line with the trends observed in figure 2.1,

figure 2.2 also reflects a general improvement,

this time in the share of the labour force having

completed tertiary education. Of the 64 countries

Figure 2.2 Share of the labour force with tertiary level educational attainment (%)

Source: KILM, 9th edn, table 14a, ages 15+, 1997−2003 and latest year available after 2009.

01020 30 40 50 60 70

GEO

NDL

DNK

SWE

CHE

FRA

RUS

LUX

IRL

CYP

GBR

LTU

HKG

MEX

GLP

PAN

POL

MAC

AUT

SVN

GRC

ISL

LVA

BGR

SGP

MNG

MDA

MLT

PRT

CZE

DOM

TUR

ROU

ALB

ETH

BRA

SMR

YEM

IND

MAR

IDN

NAM

BLZ

MDG

LKA

ITA

SVK

MTQ

HRV

HUN

REU

GUF

URY

BHS

NOR

FIN

BEL

EST

RUS

CZE

USA

CAN

SVK

2000

High income Upper-middle income Low and lower-middle income

Latest year

GBR

LVA

CAN

SVN

EST

HUN

NOR

LTU

POL

DEU

SWE

HRV

FIN

FRA

BHS

CHE

AUT

DNK

IRL

CYP

BEL

SGP

HKG

NLD

MDA

BGR

LUX

ROU

PAN

MEX

REU

GLP

MTQ

PSE

GUF

MLT

DOM

MAC

BRA

PRT

BLZ

YEM

MDG

CRI

ITA

MNG

URY

ALB

SMR

ESP

ISL

GRC

IND

TUR

NAM

LKA

IDN

MAR

ETH

Education and labour markets: Analysing global patterns with the KILM

Table 14b of the KILM presents the distribu-

tion of the unemployed by level of educational

attainment. Figure 2.3 displays the share of un-

employed persons educated to tertiary level in

2000 (or the closest year available) compared to

that in the latest year available for 76 countries

for which data are available.

The share of unemployed persons with tertiary

level educational attainment declined in only ten

of these countries between 2000 and the latest

year available.

A trend increase in the share of the

unemployed with tertiary education is in line with

the rising level of educational attainment of the

labour force experienced by most countries.

However, the findings also indicate that a higher

level of education may be less and less effective in

preventing unemployment. In Saudi Arabia and

tion is considered secondary or tertiary level

varies from country to country, which renders its

analysis more difficult.

2.2. Unemployment distribution by level

of educational attainment

While the level of educational attainment in

the overall labour supply is an important indicator

for assessing changes in an economy’s productive

potential, if sufficient and suitable employment

opportunities are not available, the macroeco-

nomic benefits of a more highly educated labour

force are unlikely to materialize. Assessing the

educational profile of the unemployed alongside

that of the labour force provides important

insights into the extent of the mismatch between

supply and demand in the labour market.

Figure 2.3 Share of unemployed with tertiary level educational attainment (%)

Source: KILM, 9th edn, table 14b, ages 15+, 1997−2003 and latest year available after 2009.

01020 30 40 50 60

GEO

NDL

DNK

SWE

CHE

FRA

RUS

LUX

IRL

CYP

GBR

LTU

HKG

MEX

GLP

PAN

POL

MAC

AUT

SVN

GRC

ISL

LVA

BGR

SGP

MNG

MDA

MLT

PRT

CZE

DOM

TUR

ROU

ALB

RUS

GEO

PHL

UKR

CAN

QAT

IND

PSE

VEN

CYP

LUX

TUN

CHE

EST

IRL

DZA

MAC

AUT

HKG

TUR

GBR

SWE

BHR

CHL

SMR

MEX

DNK

FIN

MYS

BEL

MDA

MNG

PAN

NOR

THA

MAR

LTU

URY

NIC

NLD

HRV

BGR

MLT

SVK

BHS

ITA

NAM

ZWE

BLZ

BRA

CUB

HUN

CZE

ZAF

RWA

IDN

SLV

CRI

DEU

DOM

KGZ

ROU

ALB

POL

PRT

AZE

LVA

SVN

ISL

SGP

GRC

FRA

BRA

SMR

YEM

IND

MAR

IDN

NAM

BLZ

MDG

LKA

ITA

SVK

MTQ

HRV

HUN

REU

GUF

URY

BHS

NOR

FIN

BEL

EST

RUS

CZE

USA

CAN

SVK

2000

High income Upper-middle income Low and lower-middle income

Latest year

GBR

LVA

CAN

SVN

EST

HUN

NOR

LTU

POL

DEU

SWE

HRV

FIN

FRA

BHS

CHE

AUT

DNK

IRL

CYP

BEL

SGP

HKG

NLD

MDA

BGR

LUX

ROU

PAN

MEX

REU

GLP

MTQ

PSE

GUF

MLT

DOM

MAC

BRA

PRT

BLZ

YEM

MDG

CRI

ITA

MNG

URY

ALB

SMR

ESP

ISL

GRC

IND

TUR

NAM

LKA

IDN

MAR

ETH

HND

ETH

Education and labour markets: Analysing global patterns with the KILM

by contrast, such as Cyprus, Republic of Moldova

and Mongolia, highly educated young people

seem to be facing an employment bottleneck. This

could imply a lack of sufficient professional and

high-level technical jobs to absorb the number of

skilled individuals in the labour force. However, it

is important to interpret these results with caution

since tertiary education is usually completed only

towards the end of the youth age band (15−24).

2.3. Unemployment rate by level

of educational attainment

Table 14c of the KILM presents data on un-

employment rates by level of educational attain-

ment. In doing so, it provides insights into changes

in demand for workers with different education

and skill levels. Figure 2.5 focuses on the unemploy-

ment rate of persons with tertiary level education.

Canada, the share of unemployed persons with

tertiary level educational attainment doubled over

these years. In Tunisia, the share of unemployed

persons with a tertiary education increased dramat-

ically, from only 6.6 per cent to 30.9 per cent.

Figure 2.4 focuses on the situation of youth,

specifically on changes in the past 10−15 years in

the share of unemployed youth with tertiary level

educational attainment.

Among the 33 countries for which data are

available, only five experienced a decrease in the

share of unemployed youth with tertiary level

educational attainment. Trends in Spain are worth

highlighting, since high educational levels here

seem to have protected the younger generation

against the considerable increase in unemploy-

ment overall during the period. In other countries,

Figure 2.4. Share of young unemployed with tertiary level educational attainment (%)

Source: KILM, 9th edn, table 14b, ages 15−24, 1997−2003 and latest year available after 2009.

010155 2520 30 35 40

GEO

NDL

DNK

SWE

CHE

FRA

RUS

LUX

IRL

CYP

GBR

LTU

HKG

MEX

GLP

PAN

POL

MAC

AUT

SVN

GRC

ISL

LVA

BGR

SGP

MNG

MDA

MLT

PRT

CZE

DOM

TUR

ROU

ALB

BRA

SMR

YEM

IND

MAR

IDN

NAM

MNG

CYP

MDA

LTU

ESP

NOR

BEL

FRA

LUX

GBR

IRL

GRC

EST

ROU

AUT

POL

HUN

POL

SVK

DNK

CZE

SWE

BGR

MLT

HRV

CHE

NDL

ITA

DEU

FIN

ISL

LVA

MDG

LKA

ITA

SVK

MTQ

HRV

HUN

REU

GUF

URY

BHS

NOR

FIN

BEL

EST

RUS

CZE

USA

CAN

SVK

2000

High income Upper-middle income Low and lower-middle income

Latest year

GBR

LVA

CAN

SVN

EST

HUN

NOR

LTU

POL

DEU

SWE

HRV

FIN

FRA

BHS

CHE

AUT

DNK

IRL

CYP

BEL

SGP

HKG

NLD

MDA

BGR

LUX

ROU

PAN

MEX

REU

GLP

MTQ

PSE

GUF

MLT

DOM

MAC

BRA

PRT

BLZ

YEM

MDG

CRI

ITA

MNG

URY

ALB

SMR

ESP

ISL

GRC

IND

TUR

NAM

LKA

IDN

MAR

ETH

Education and labour markets: Analysing global patterns with the KILM

The largest decreases (of around 10 percentage

points in each case) occurred in Uruguay, Panama

and the Russian Federation.

In order to assess the relationship between

educational level and unemployment, it is import-

ant to compare the respective situations of those

who have achieved tertiary level education and

those with primary level education or less. The

results for the latter group are displayed in

figure 2.6.

Once again, the results are rather scattered,

and generate no visible pattern. In 19 out of the

53 countries included, the unemployment rate fell

for persons with no higher than primary level

education. In some Latin American countries, such

as Uruguay and Panama, the unemployment rate

of persons with primary level education or less

In contrast to the first two indicators studied in

this chapter, where data for the past 10−15 years

showed a clear trend, the results for the unemploy-

ment rate of persons with tertiary level educa-

tional attainment are more scattered. Of the

53 countries for which data are available, 35 experi-

enced an increase in the unemployment rate of the

most highly educated in the labour force over the

period. The situation is particularly critical in

Tunisia, where the unemployment rate of tertiary

graduates increased by more than 21 percentage

points. Increases exceeding 10 percentage points

were also observed in the Occupied Palestinian

Territory, Greece and Cyprus. In Egypt and Georgia,

the unemployment rate among tertiary graduates

was already at very high levels in 2000 and con-

tinued to increase during the period studied.

Conversely, in 18 countries the unemployment rate

of persons with tertiary level education declined.

Figure 2.5. Unemployment rate of persons with tertiary level educational attainment (%)

Source: KILM, 9th edn, table 14c, ages 15+, 1997−2003 and latest year available after 2009.

010155 2520 30

GEO

NDL

DNK

SWE

CHE

FRA

RUS

LUX

IRL

CYP

GBR

LTU

HKG

MEX

GLP

PAN

POL

MAC

AUT

SVN

GRC

ISL

LVA

BGR

SGP

MNG

MDA

MLT

PRT

CZE

DOM

TUR

ROU

ALB

BRA

SMR

YEM

IND

MAR

IDN

NAM

PSE

TUN

GRC

EGY

GEO

JOR

ESP

CYP

PRT

HRV

TUR

URY

RUS

PAN

IRL

MEX

LUX

ISL

NOR

KOR

LKA

ITA

SVK

MTQ

HRV

HUN

REU

GUF

URY

BHS

NOR

FIN

BEL

EST

RUS

CZE

USA

CAN

SVK

2000

High income Upper-middle income Low and lower-middle income

Latest Year

GBR

LVA

CAN

SVN

EST

HUN

NOR

LTU

POL

DEU

SWE

HRV

FIN

FRA

BHS

CHE

AUT

DNK

IRL

CYP

BEL

SGP

HKG

NLD

MDA

BGR

LUX

ROU

PAN

MEX

REU

GLP

MTQ

PSE

GUF

MLT

DOM

MAC

BRA

PRT

BLZ

YEM

MDG

CRI

ITA

MNG

URY

ALB

SMR

ESP

ISL

GRC

IND

TUR

NAM

LKA

IDN

MAR

ETH

SGP

MAC

DEU

HKG

AUS

BEL

NZL

SVN

MNG

SVK

ITA

LVA

FRA

POL

EST

MDA

LTU

IDN

FIN

ISR

CHE

AUT

GBR

CAN

SWE

Education and labour markets: Analysing global patterns with the KILM

than for those with a tertiary degree (16.0 per

cent). Finally, in Slovakia, the unemployment rate

among persons with no higher than primary

education remains very high and, as in Spain, the

highest rate of unemployment is among persons

with the lowest educational level.

In terms of the overall trends in unemploy-

ment rates across different levels of education,

persons in the labour force with a tertiary educa-

tion have the lowest likelihood of being un-

employed in 41 out of 53 countries with available

data. Tertiary graduates are the least likely to be

unemployed in 34 out of 37 high-income econ-

omies, but in only 7 out of 16 middle income

economies. Looking at changes in unemployment

rates over the past 15 years, unemployment rate

decreased significantly, as it did for persons with

tertiary education. In the Occupied Palestinian

Territory, however, where the unemployment rate

increased for the most highly educated, it fell by

7.3 percentage points for those with primary

level education or less. Here, as in many develop-

ing economies, it appears that the country’s less

educated labour force cannot afford to remain

unemployed. Conversely, in Spain and Greece, the

unemployment rate of persons with primary level

education or less increased by an astonishing 20

percentage points in each case, reflecting the

severity of the economic crisis that affected these

two countries after 2008. It is also worth pointing

out that in Spain in 2013, the unemployment rate

was more than twice as high among persons with

primary level education or less (35.1 per cent)

Figure 2.6. Unemployment rate of persons with primary or less than primary level

educational attainment (%)

Source: KILM, 9th edn, table 14c, ages 15+, 1997−2003 and latest year available after 2009.

010155 2520 4530 35 40

GEO

NDL

DNK

SWE

CHE

FRA

RUS

LUX

IRL

CYP

GBR

LTU

HKG

MEX

GLP

PAN

POL

MAC

AUT

SVN

GRC

ISL

LVA

BGR

SGP

MNG

MDA

MLT

PRT

CZE

DOM

TUR

ROU

ALB

BRA

SMR

YEM

IND

SVK

ESP

NAM

LTU

BGR

CZE

GRC

SWE

GBR

SVN

BEL

ITA

CAN

FRA

FIN

EST

RUS

TUN

URY

JOR

DEU

ISR

NZL

AUT

DNK

NDL

GEO

EGY

LUX

CRI

MNG

MDA

HKG

SGP

PAN

MAC

KOR

ISL

MEX

CHE

ROU

AUS

CYP

PRT

HUN

HRV

IRL

LVA

POL

PSE

IDN

NAM

LKA

ITA

SVK

MTQ

HRV

HUN

REU

GUF

URY

BHS

NOR

FIN

BEL

EST

RUS

CZE

USA

CAN

SVK

2000

High income Upper-middle income Low and lower-middle income

Latest year

GBR

LVA

CAN

SVN

EST

HUN

NOR

LTU

POL

DEU

SWE

HRV

FIN

FRA

BHS

CHE

AUT

DNK

IRL

CYP

BEL

SGP

HKG

NLD

MDA

BGR

LUX

ROU

PAN

MEX

REU

GLP

MTQ

PSE

GUF

MLT

DOM

MAC

BRA

PRT

BLZ

YEM

MDG

CRI

ITA

MNG

URY

ALB

SMR

ESP

ISL

GRC

IND

TUR

NAM

LKA

IDN

MAR

ETH

NOR

IDN

Education and labour markets: Analysing global patterns with the KILM

Among the 38 countries for which data are

available, there does not seem to be a clear under-

lying pattern. Nonetheless, it is still noteworthy

that the countries in which the NEET share of

youth has increased the most in the past decade

(Cyprus, Greece, Ireland, Italy, Spain and the

United Kingdom) are all high-income economies

that were badly hit by the global financial crisis.

The crisis in these countries disproportionately

affected young people, leaving them more vulner-

able to unemployment and without the means to

take their education or training further. Conversely,

the countries in which the NEET share of youth

decreased the most are upper-middle-income

economies (Turkey, the former Yugoslav Republic

of Macedonia and Bulgaria). However, for the

majority of the countries included in figure 2.7,

there were no significant changes during the

dynamics were most favourable (either decreas-

ing the most or increasing the least) among those

with a tertiary education in 19 out of the 53 coun-

tries, among those with secondary education in

24 out of 53 countries, and among those with a

primary education in only ten countries.

2.4. Share of youth not in education,

employment or training (NEET)

Table 10c of the KILM presents data on the

share of youth not in education, employment or

training (NEET). By its nature, this indicator

represents a broader measure of potential youth

labour market entrants than either youth un-

employment or youth inactivity. In figure 2.7,

this indicator is displayed for 2003 or the closest

year available and for the latest year available.

Figure 2.7. Share of youth not in education, employment or training (%)

Source: KILM, 9th edn, table 10c, ages 15−24, 1998−2007 and latest year available after 2011.

010155 2520 4030 35

GEO

NDL

DNK

SWE

CHE

FRA

RUS

LUX

IRL

CYP

GBR

LTU

HKG

MEX

GLP

PAN

POL

MAC

AUT

SVN

GRC

ISL

LVA

BGR

SGP

MNG

MDA

MLT

PRT

CZE

DOM

TUR

ROU

ALB

BRA

SMR

YEM

IND

TUR

MKD

BGR

MEX

ROU

HRV

ITA

GRC

ESP

USA

CYP

IRL

ISL

LUX

SWE

CHE

SVN

LTU

FIN

NZL

MLT

BEL

EST

TVA

POL

HUN

SVK

CZU

FRA

PRT

CAN

AUS

AUT

DEU

NOR

DNK

NLD

NAM

LKA

ITA

SVK

MTQ

HRV

HUN

REU

GUF

URY

BHS

NOR

FIN

BEL

EST

RUS

CZE

USA

CAN

SVK

2003

High income Upper-middle income

Latest year

GBR

LVA

CAN

SVN

EST

HUN

NOR

LTU

POL

DEU

SWE

HRV

FIN

FRA

BHS

CHE

AUT

DNK

IRL

CYP

BEL

SGP

HKG

NLD

MDA

BGR

LUX

ROU

PAN

MEX

REU

GLP

MTQ

PSE

GUF

MLT

DOM

MAC

BRA

PRT

BLZ

YEM

MDG

CRI

ITA

MNG

URY

ALB

SMR

ESP

ISL

GRC

IND

TUR

NAM

LKA

IDN

MAR

ETH

Education and labour markets: Analysing global patterns with the KILM

to investigate whether there are links between

these indicators and other key labour market

indicators, specifically unemployment rates,

labour productivity, the employment-to-popula-

tion ratio and the share of employees.

3.1. Unemployment and education

Figure 3.1 depicts values for two labour

market indicators among persons with a tertiary

education, comparing the respective shares of

persons educated to this level in the labour force

and in the unemployed.

In 67 of the 93 countries for which data are

available, education seems to be an effective tool

for protecting people from unemployment: that

is, the share of unemployed persons with tertiary

period under review. More specifically, in 23 of

the 38 countries included here, the NEET share of

youth rose or fell by less than 2.5 percentage

points between 2003 (or the closest year avail-

able) and the latest year available.

3. Impact of education

on labour market outcomes

Having analysed trends over the past 10−15

years for the four KILM indicators pertaining to

education, with particular attention to the rela-

tionship between educational attainment and

labour market outcomes, we turn in this section

Figure 3.1. Share of labour force and unemployed with tertiary level of educational

attainment (%)

Source: KILM, 9th edn, tables 14a, 14b, ages 15+, latest year available after 2009.

0103020 7040 50 60

GEO

NDL

DNK

SWE

CHE

FRA

RUS

LUX

IRL

CYP

GBR

LTU

HKG

MEX

GLP

PAN

POL

MAC

AUT

SVN

GRC

ISL

LVA

BGR

SGP

MNG

MDA

MLT

PRT

CZE

DOM

TUR

ROU

ALB

BRA

SMR

YEM

IND

NAM

BLZ

TLS

RWA

HND

IDN

BRA

IND

BHR

MAR

NIC

SMR

DZA

ALB

KGZ

ROU

THA

EGY

TUN

PER

PAN

MNG

COL

MDA

HKG

PRY

URY

TUR

PRT

DOM

SRB

AZE

LKA PHL

ITA

MLT

DEU

BGR

POL

LVA

SVN

KAZ

AUT

GRC

DNK

ISL

MAC

CHE

CYP

SGP

RUS

CAN

ARM

LUX

EST

NOR

IRL

FIN

BEL

GBR

ESP

SWE

FRA

BMU

NDL

USA

HUN

ARG

CZE

KOS

SVK

REU

BHS

MNE

CRI

ETH

BIH

CUB

ZAF

NAM

LKA

ITA

SVK

MTQ

HRV

HUN

REU

GUF

URY

BHS

NOR

FIN

BEL

EST

RUS

CZE

USA

CAN

SVK

Share of labour force with tertiary level of education (%)

High income Upper-middle income Low and lower-middle income

Share of persons with tertiary level of education among the unemployed (%)

GBR

LVA

CAN

SVN

EST

HUN

NOR

LTU

POL

DEU

SWE

HRV

FIN

FRA

BHS

CHE

AUT

DNK

IRL

CYP

BEL

SGP

HKG

NLD

MDA

BGR

LUX

ROU

PAN

MEX

REU

GLP

MTQ

PSE

GUF

MLT

DOM

MAC

BRA

PRT

BLZ

YEM

MDG

CRI

ITA

MNG

URY

ALB

SMR

ESP

ISL

GRC

IND

TUR

NAM

LKA

IDN

MAR

ETH

Education does

not protect from

unemployment

Education

protects from

unemployment

SLV

GTM

TUR

CHI

MYS

ECU

MEX

Education and labour markets: Analysing global patterns with the KILM

differences (over 10 percentage points) in

Bahrain, Egypt, India and Tunisia.

An overview of the situation across countries

in different income groups suggests that higher

levels of education tend to protect workers from

unemployment in high income economies.

Among upper middle income economies the situ-

ation is more mixed, and in low and lower middle

income economies, people with high levels of

education tend to be more likely to be unem-

ployed. In these developing economies, there is a

clear bottleneck, with skilled persons far outnum-

bering the available jobs matching their compe-

tencies and expectations. When studying these

trends in unemployment, it is crucial to take into

account the national context in terms of un-

employment insurance policies. In contexts

level education is lower than the share of the

labour force with the same educational level. The

difference is particularly marked in Lithuania,

where it stood at 24 percentage points (39.5 per

cent of the labour force, but only 15.5 per cent of

the unemployed, have a tertiary degree). In

Belgium, the Cayman Islands, Ireland and the

Russian Federation, the difference is also close to

20 percentage points, indicating that high levels

of education play a major role in preventing

unemployment. On the other hand, there are 26

countries where the opposite is observed, that is,

where persons in the labour force with a tertiary

degree are more likely than those with a lower

level of education to be unemployed. This is espe-

cially the case in the Philippines, Sri Lanka and

Thailand, where the difference exceeds 15

percentage points. We also find considerable

Figure 3.2. Share of labour force and unemployment rate for persons with tertiary level

of educational attainment (%)

Source: KILM, 9th edn, tables 14a, 14c, ages 15+, latest year available after 2009.

01020 4030 7050 60

GEO

NDL

DNK

SWE

CHE

FRA

RUS

LUX

IRL

CYP

GBR

LTU

HKG

MEX

GLP

PAN

POL

MAC

AUT

SVN

GRC

ISL

LVA

BGR

SGP

MNG

MDA

MLT

PRT

CZE

DOM

TUR

ROU

ALB

TUN

PSE

GRC

MKD

EGY

MAR

TLS

SLV

SRB

BIH

DZA

ALB

KOS

REU

PRT

HRV

BHS

VEN

MNG

BGR

POL

ROU

CHI

ECU

URY

PRY

MYS

ARG

CZE

MLT

HKG

DEU

MAC

PAN

KAZ

AUT

ISL

DNK

NLD

MEX

PER

LVA

SVN

TUR

BWA

KGZ

ITA

SVK

ZAF

LKA

DOM

COL

BLZ

NAM

BRA

CUB

THA

MNE

SMR

YEM

IND

NAM

LKA

ITA

SVK

MTQ

HRV

HUN

REU

GUF

URY

BHS

NOR

FIN

BEL

EST

RUS

CZE

USA

CAN

SVK

Share of labour force with tertiary level of education

Unemployment rate of persons with tertiary level education (%)

GBR

LVA

CAN

SVN

EST

HUN

NOR

LTU

POL

DEU

SWE

HRV

FIN

FRA

BHS

CHE

AUT

DNK

IRL

CYP

BEL

SGP

HKG

NLD

MDA

BGR

LUX

ROU

PAN

MEX

REU

GLP

MTQ

PSE

GUF

MLT

DOM

MAC

BRA

PRT

BLZ

YEM

MDG

CRI

ITA

MNG

URY

ALB

SMR

ESP

ISL

GRC

IND

TUR

NAM

LKA

IDN

MAR

ETH

High income Upper-middle income Low and lower-middle income

RWA

GHA

HND

IDN

GTM

BHR

NIC

BMU

SWE

USA

FRA

EST

LTU

BEL

IRL

LUX

CYP

ESP

ARM

GEO

SGP

RUS

CAN

FIN

GBR

CYM

CHE

NOR

Education and labour markets: Analysing global patterns with the KILM

Countries in this group with a high share of the

labour force educated to tertiary level and a low

unemployment rate are all high income econ-

omies. In these cases, education clearly appears

to act as a barrier against unemployment. The

relationship is most marked in Canada,

Luxembourg, Norway, the Russian Federation

and Singapore.

Conversely, the countries in this group with a

relatively low share of the labour force with

tertiary level education but a high unemploy-

ment rate for this category tend to be mostly

upper middle, lower middle and low income

countries. These include Egypt, the former

Yugoslav Republic of Macedonia, Greece, Tunisia

and the Occupied Palestinian Territory. This may

where these are limited or do not exist, un-

employment might not be seen as an option (ILO,

2016). Other possible explanations for the pattern

observed in low and lower middle income econ-

omies include family income being significant

enough for those that have higher education to

remain unemployed while they look for a job that

fully meets their expectations.

Figure 3.2 compares data in tables 14a and

14c for the tertiary level of educational attain-

ment. The coefficient of determination (R2) is

close to zero, reflecting the scattered results.

However, even though there is no clear indica-

tion that a tertiary degree plays a role in protect-

ing people from high unemployment rates, the

figure still presents some interesting patterns.

Figure 3.3. Tertiary level of educational attainment and labour productivity

(PPP US$)

Note: The trend line included in the graph shows to what extent there is a linear relationship between labour productivity and the share of the

labour force with a tertiary education. The coefficient of determination (R2) conveys how well this linear regression fits given the existing data.

Source: KILM, 9th edn, tables 14a, 16a, ages 15+, latest year available after 2009.

ETH

MDG

KHM

KGZ

GHA

IND

MDA

YEM

PHL

GEO

MAR

ARM

IDN

GTM

LKA

COL

PER

THA

BRA

ECU

CRI

DOM

EGY

ALB

AZE

TUN

BGR

MKD

ARG

ZAF

MEX

URY

VEN

ROU

KAZ

RUS

DZA

BIH

LVA

CHL

MYS

HRV

TUR

EST

HUN

LTU

POL

CZE

PRT

SVN

SVK

GRC

MLT

CYP

ISL

GBR

ESP

CAN

DEU

ITA

FIN

DNK

AUT

NLD

SWE

FRA

CHE

BEL

HKG

USA

IRL

NOR

LUX

SGP

140

120

y = 0.3805x +12.353

= 0.43646

100

GEO

NDL

DNK

SWE

CHE

FRA

RUS

LUX

IRL

CYP

GBR

LTU

HKG

MEX

GLP

PAN

POL

MAC

AUT

SVN

GRC

ISL

LVA

BGR

SGP

MNG

MDA

MLT

PRT

CZE

DOM

TUR

ROU

ALB

SMR

YEM

IND

NAM

LKA

ITA

SVK

MTQ

HRV

HUN

REU

GUF

URY

BHS

NOR

FIN

BEL

EST

RUS

CZE

USA

CAN

SVK

Labour productivity per person employed in 2014 (Thousand US$)

GBR

LVA

CAN

SVN

EST

HUN

NOR

LTU

POL

DEU

SWE

HRV

FIN

FRA

BHS

CHE

AUT

DNK

IRL

CYP

BEL

SGP

HKG

NLD

MDA

BGR

LUX

ROU

PAN

MEX

REU

GLP

MTQ

PSE

GUF

MLT

DOM

MAC

BRA

PRT

BLZ

YEM

MDG

CRI

ITA

MNG

URY

ALB

SMR

ESP

ISL

GRC

IND

TUR

NAM

LKA

IDN

MAR

ETH

Labour productivity per person employed in 2014 (thousand US$) Share of labour force with tertiary level of education (%)

Education and labour markets: Analysing global patterns with the KILM

tion. Labour productivity, which we define here

as output per person employed, measures the

efficiency with which inputs are used in an econ-

omy to produce goods and services; it offers an

indication of both competitiveness and living

standards within a country. In figure 3.3, we look

at the relationship between tertiary level educa-

tional attainment and labour productivity.

It is clear from the figure that a link exists

between these two indicators.

A greater propor-

tion of the labour force with a tertiary educa-

tion is associated with higher levels of labour

productivity. When the 74 countries included

are sorted according to their level of labour

productivity, the trend line of the share of the

labour force with tertiary level education is

seem surprising, since in these countries the

labour force educated to tertiary level is not very

large and it might therefore be expected that

these highly educated people would easily find

skilled jobs. However, in these countries there are

still too few employment opportunities for them,

either because the labour market is in a crisis (the

former Yugoslav Republic of Macedonia, Greece)

or because skilled jobs are lacking, revealing a

skills mismatch situation (Egypt, Occupied

Palestinian Territory, Tunisia).

3.2. Labour productivity and education

This section presents information on the rela-

tionship between labour productivity (table 16a)

for the aggregate economy and tertiary educa-

Figure 3.4 Employment-to-population ratio and labour force with tertiary level educational

attainment

Source: KILM, 9th edn, tables 2b, 14a, ages 15+, latest year available after 2009.

010 20 30 40 50 60 70

100

Share of labour force with tertiary-level education (%)

High income Upper-middle income Low and lower-middle income

Employment-to-population ratio (%)

KOS

YEM

BIH

TLS

PSE

GRC

MDA

SRB

MKD

TUN

ZAF

IND

NAM

HND

SLV

GTM

BRA

ROU

KGZ

ECU

CHI

URY

CZE

CRI

PER

PRY

PAN

DEU

VEN

AUT

GEO

LVA

POL

SVN

SVK

PRT

MLT

DOM

BWA

TUR

HRV

BGR

HUN

ESP

FRA

BEL

NLD

DNK

USA

GBR

EST

LTU

IRL

CYP

FIN

LUX

NOR

CHE

SWE

ISL

CYM

RUS

CAN

MYS

CUB

THA

IDN

NIC

RWA

GHA

KHM

MDG

ETH

KWT

ITA

Education and labour markets: Analysing global patterns with the KILM

Figure 3.5. Share of employees in total employment, and share of labour force with

tertiary level of educational attainment

Note: The trend line included in the graph shows to what extent there is a linear relationship between the share of wage and salaried workers

(employees) amongst all persons employed and the share of the labour force with a tertiary level of education.

Source: KILM, 9th edn, tables 3, 14a, ages 15+, latest year available after 2009.

ETH

MDG

RWA

IND

GHA

TLS

AZE

KHM

IDN

GEO

ALB

THA

MAR

HND

GTM

NIC

COL

MNG

PER

KGZ

LKA

DOM

PRY

ECU

ARM

PHL

VEN

EGY

NAM

GRC

YEM

TUR

MEX

MDA

PAN

SRB

ROU

BWA

PSE

CHL

BRA

DZA

KAZ

TUN

URY

MKD

MYS

KOS

BIH

ARG

ITA

CRI

CUB

POL

GLP

PRT

SVN

CZE

CYP

BMU

MNE

ESP

GUF

IRL

NLD

MTQ

REU

HRV

GBR

SVK

CAN

CHE

SGP

BHS

BEL

ZAF

FIN

MLT

AUT

ISL

LTU

BGR

LVA

FRA

HUN

DEU

SWE

HKG

EST

DNK

LUX

CYM

RUS

NOR

MAC

USA

0203010 40 50

70 9060 80 100

Share of wage and salaried workers (employees) (%) Share of labour force with tertiary level of education (%)

y = 0.2884x +11.35

= 0.44324

Education and labour markets: Analysing global patterns with the KILM

The figure shows a clear, positive relationship

between the two indicators. The higher the share

of employees in a country, the higher the propor-

tion of persons with a tertiary education degree.

Educational attainment is clearly related to the

probability of being in the labour market as an

employee.

As in section 3.2,

Armenia, Canada and

the Russian Federation stand out, with a much

higher share of tertiary educational attainment in

the labour force than would be expected given

the share of employees in their employed popula-

tion. Conversely, Namibia, South Africa and, to a

lesser extent, Botswana have very low shares of

employees in relation to the educational attain-

ment of their labour forces. These results may

suggest a regional pattern, reflecting an environ-

ment in which, while the educational level of the

labour force is improving, the configuration of

the economy and labour market is fairly stagnant,

with self-employment remaining prevalent.

This section of the chapter has explored the

links between education and several key labour

indicators. The findings comparing educational

attainment with labour productivity and share of

employees suggest a clear link between the

educational level achieved within a labour force

and labour market outcomes. However, the link

cannot be established with equal confidence for

all the labour market indicators studied. In par-

ticular, the employment-to-population ratio ap-

pears to be completely independent of variations

in educational attainment.

4. The current situation

in 12 selected countries

4.1 Data for latest year available

on the four selected indicators

In the previous section the analysis incorp-

orated all countries for which recent data are

available. In this section, we look in greater detail

at the current situation in a selection of 12 coun-

tries covering all levels of development. Table 4.1

lists the 12 countries, along with selected labour

market data and income group for each.

In figure 4.1, labour force by educational attain-

ment (KILM table 14a) is displayed for the latest

year available across the 12 selected countries.

The

highest share of the labour force with primary or

less than primary educational attainment is found

in El Salvador, where it reaches 85.9 per cent. The

next highest shares are found in Ethiopia and

Cambodia, the two low income economies among

our selection. More than three-quarters of the

clearly positive, with a coefficient of determin-

ation (R2) of 0.44. However, notwithstanding a

clear global underlying trend, there are some

noticeable exceptions, such as Armenia, Canada

and the Russian Federation, where the share of

the labour force with tertiary level education

appears to be much higher than would be

expected, given the corresponding levels of

labour productivity.

3.3. Employment-to-population ratio

and education

Table 2b of the KILM presents data on

employment-to-population ratios based on

national estimates. The employment-to-popula-

tion ratio is defined as the proportion of a coun-

try’s working-age population that is employed.

A high ratio means that a large proportion of a

country’s working-age population is employed,

while a low ratio means that a large share of the

working-age population is not involved directly

in labour market related activities, either because

they are unemployed or (more likely) because

they are not in the labour force. In figure 3.4, this

indicator is shown together with the share of

the labour force educated to tertiary level.

No clear relationship between these two indi-

cators can be deduced from this figure, and the

coefficient of determination (R2) is almost

exactly 0. However, the variability in the employ-

ment-to-population ratio is much higher for coun-

tries where the share of labour force with tertiary

education is low. In Bosnia and Herzegovina and

Ethiopia, for example, a similar share of the labour

force has tertiary level education (14.5 per cent

and 16.4 per cent, respectively), but the two

countries’ employment-to-population ratios differ

by nearly 50 percentage points (31.6 per cent

and 79.4 per cent, respectively). It would seem

that the higher the share of the labour force with

a tertiary degree, the smaller the variability: when-

ever this share exceeds 45 per cent, the employ-

ment-to-population ratio falls within the range

45−65 per cent.

3.4. Share of employees and education

Table 3 of the KILM presents data on employ-

ment by status in employment, according to the

categories set out in the 1993 International

Classification by Status in Employment (ICSE). We

focus here on “employees”, the category of status

in employment that typically benefits from the

highest levels of income and job security in the

labour market. Figure 3.5 shows data on the share

of employees in total employment alongside data

on the share of the labour force with tertiary

level education.

Education and labour markets: Analysing global patterns with the KILM

These two countries are classified as upper middle

income economies. Therefore, in these cases, a rela-

tively high GNI per capita is not accompanied by

a relatively high level of educational attainment. In

contrast, Kyrgyzstan, classified as a lower middle

income economy, has the lowest share of the

labour force in these two countries did not

complete education beyond primary schooling,

indicating a large share of workers with little

education. In Thailand and Algeria, more than half

of the labour force (67.5 and 63.2 per cent, respect-

ively) did not attain a secondary level education.

Table 4.1. Key information for selected countries

Country Working-age population

(000s, aged 15+)

Employment-to-

population ratio (%)

Unemployment

rate (%)

World Bank income group

Canada

29 952 61.4 6.9 High income

Germany

71 875 57.4 5.0 High income

Algeria

29 100 36.2 9.8 Upper middle income

Brazil

157 000 64.0 4.8 Upper middle income

Mexico

90 875 56.9 4.8 Upper middle income

Thailand

55 636 69.4 0.8 Upper middle income

Egypt

58 572 42.1 13.2 Lower middle income

El Salvador

4 572 59.9 5.9 Lower middle income

Kyrgyzstan

3 942 57.2 8.3 Lower middle income

Philippines

67 814 60.0 6.8 Lower middle income

Cambodia

10 811 82.8 0.3 Low income

Ethiopia

57 948 79.4 4.5 Low income

Sources: World Bank, ILOSTAT, KILM, 9th edn, latest year available.

Figure 4.1. Labour force by level of educational attainment

Source: KILM, 9th edn, table 14a, ages 15+, latest year available.

El Salvador, 2013

Ethiopia, 2012

Cambodia, 2012

Thailand, 2013

Algeria, 2011

Brazil, 2013

Egypt, 2013

Mexico, 2011

Philippines, 2008

Germany, 2014

Canada, 2014

Kyrgyzstan, 2013

100

Primary or less Secondary Tertiary

Education and labour markets: Analysing global patterns with the KILM

Figure 4.2. Unemployment distribution by level of educational attainment

Source: KILM, 9th edn, table 14b, ages 15+, latest year available.

El Salvador, 2013

Ethiopia, 2012

Algeria, 2011

Cambodia, 2012

Brazil, 2013

Thailand, 2013

Germany, 2014

Mexico, 2011

Egypt, 2013

Canada, 2014

Philippines, 2008

Kyrgyzstan, 2013

100

Primary or less Secondary Tertiary

labour force with primary education or less

(7.9 per cent). Germany and Canada, the two high

income economies in our sample, have the highest

shares of the labour force with tertiary level educa-

tion (27.1 per cent and 63.4 per cent, respectively).

In figure 4.2, the unemployment distribution

by level of educational attainment (KILM table

14b) is displayed for the latest year available in

our selected countries. Owing to the large share

of the labour force with primary education or

less in many low income economies, the share of

the unemployed with primary education or less

tends to be significant in these countries.

Conversely, high income economies have a higher

proportion of people with tertiary level educa-

tion, which might lead one to expect a larger

share of the unemployed with higher education

in these countries. However, in Germany, unem-

ployment seems to be strongly related to level of

education. While only 13.2 per cent of the labour

force have no more than primary level education,

31.1 per cent of the unemployed have this low

level of education. Less educated workers in

Germany therefore have a higher probability of

being unemployed. In Egypt, on the other hand,

the opposite relationship is observed. Only

18.7 per cent of the labour force have attained a

tertiary level education, but 31.1 per cent of the

unemployed are educated to this level.

In figure 4.3, unemployment rates by level of

educational attainment (KILM table 14c) are

displayed for the latest year available in our

selected countries.

In Kyrgyzstan, Canada,

Germany and Brazil, unemployment rates are

lower among workers with higher levels of

educational attainment. In Germany, individuals

with only primary education or less are more

than four times as likely to be unemployed as

those with tertiary education. In four countries

(Egypt, Philippines, Cambodia and Thailand) the

situation is exactly the opposite: here the rate of

unemployment increases in line with the level of

education. In the Philippines, individuals in the

labour force with a tertiary education are three

times as likely to be unemployed as those with

only a primary (or less) education. The other

three countries do not show any consistent trend.

Figure 4.4 displays the proportion of young

people aged 15−24 not in education, employment

or training (NEET; KILM table 10c) in the selected

Owing to a lack of data in KILM table 14b, Ethiopia is

not included in this section.

Education and labour markets: Analysing global patterns with the KILM

Figure 4.3. Unemployment rate by level of educational attainment

Figure 4.4. Share of youth (aged 15−24) not in education, employment or training, by sex

Source: KILM, 9th edn, table 14c, ages 15+, latest year available.

Source: KILM, 9th edn, table 10c, latest year available.

El Salvador, 2013

Algeria, 2011

Cambodia, 2012

Brazil, 2013

Thailand, 2012

Germany, 2013

Mexico, 2013

Philippines, 2008

Egypt, 2013

Canada, 2013

Philippines, 2008

Kyrgyzstan, 2013

Primary or less Secondary Tertiary

El Salvador, 2013

Algeria, 2013

Cambodia, 2012

Brazil, 2013

Thailand, 2013

Germany, 2014

Mexico, 2012

Cambodia, 2012

Philippines, 2012

Egypt, 2013

Canada, 2013

Philippines, 2008

Ethiopia, 2012

Kyrgyzstan, 2013

Male Female

Education and labour markets: Analysing global patterns with the KILM

their share in the labour force with their share

amongst the unemployed.

In only five of our sample of countries is the

share of unemployed with a tertiary degree actu-

ally lower than the share of persons in the labour

force with a tertiary degree. Moreover, only in

Canada and Germany is the difference significant

(greater than 10 percentage points). In Canada,

63.4 per cent of the labour force but only 48.9

per cent of the unemployed have a tertiary

degree. Thus in Canada (and in Germany) invest-

ing in one’s education can be seen as a means of

reducing the probability of becoming un-

employed. On the other hand, seven countries

show the opposite result, with tertiary graduates

comprising a disproportionately large share of

the unemployed. The largest relative disadvantage

among tertiary graduates is observed in Egypt,

the Philippines and Thailand. In Thailand, only

12.8 per cent of the labour force but 31 per cent

of the unemployed have a tertiary degree. This

indicates a bottleneck, with too many skilled

persons for the number of available jobs match-

ing their competencies and expectations. In

Thailand, the overall unemployment rate remains

very low, but in Egypt and the Philippines, those

countries for the latest year available. While NEETs

are almost non-existent in Thailand and Ethiopia,

they make up significant shares of young people,

and particularly large shares of young women, in

several of the other countries. In Egypt, 40.7 per

cent of young women are NEET (as compared

with 17.3 per cent of young men). In Algeria,

young women are four times as likely to be NEET

than young men (respectively, 34.6 and 8.8 per

cent). The gender gap is also very significant in

the Philippines, Brazil, Kyrgyzstan and Mexico (in

each case more than 10 percentage points). Only

Canada and El Salvador have a higher NEET rate

for males than for females (with gaps of less than

3 percentage points).

4.2. Educational attainment compared

to other key labour market indicators

This section aims to establish whether links

exist in these countries between the four indica-

tors described in the previous section and the

other labour market indicators examined in

section 3.

Figure 4.5 focuses on persons who have

attained a tertiary level education and compares

Figure 4.5. Tertiary level educational attainment and unemployment

Source: KILM, 9th edn, table 14b, ages 15+, latest year available.

El Salvador, 2013

Algeria, 2011

Cambodia, 2012

Brazil, 2013

Germany, 2014

Thailand, 2013

Germany, 2014

Mexico, 2011

El Salvador, 2013

Philippines, 2008

Egypt, 2013

Cambodia, 2012

Canada, 2014

Philippines, 2008

Ethiopia, 2012

Kyrgyzstan, 2013

Ethiopia, 2012

Share of unemployment with tertiary level of education

Share of labour force with tertiary level of education

Education does

not protect from

unemployment

Education

protects from

unemployment

Education and labour markets: Analysing global patterns with the KILM

Figure 4.6. Tertiary level educational attainment and unemployment rate

Note: Owing to lack of data, unemployment rates are for different years in the cases of Germany (2013), Canada (2013), Mexico (2013) and

Thailand (2012).

Source: KILM, 9th edn, tables 14a, 14c, ages 15+, latest year available.

El Salvador, 2013

Algeria, 2011

Cambodia, 2012

Brazil, 2013

Thailand, 2013

Germany, 2014

Mexico, 2011

Philippines, 2008

Egypt, 2013

Canada, 2014

Philippines, 2008

Kyrgyzstan, 2013

Share of labour force with tertiary level of education Unemployment rate of persons with tertiary level of education

11.9

7.7

1.9

3.9

15.1

7.5

22.09

5.6

10.5

2.4

5.0

with a tertiary degree often have difficulty find-

ing jobs matching their level of education.

Figure 4.6 compares data in KILM tables 14a

and 14c

for the tertiary level of educational

attainment. As in section 3, there is no clear link

between these two indicators. In Cambodia,

Thailand and Brazil, the unemployment rate

among those who have attained a tertiary educa-

tion is low and the share of persons with high

education is limited. In Algeria and Egypt, the

share of persons with a tertiary education is

somewhat greater; however, the labour market is

not providing sufficient opportunities for them,

and therefore the unemployment rate of persons

educated to this level is relatively high (22.0 per

cent in Egypt). Finally, in high income economies

(Germany and Canada), the share of the labour

force with a tertiary education is high and the

unemployment rate for tertiary graduates is low.

Education clearly acts as a barrier against un-

employment in these countries.

Owing to a lack of data in KILM table 14c, Ethiopia is

not included in this section.

Figure 4.7 examines the link between attain-

ment of tertiary education and labour productiv-

ity (per person employed in PPP US$).

As in

section 3, a greater proportion of the labour force

with a tertiary education is associated with higher

levels of labour productivity. Yet some countries

show contrasting results. Labour productivity is

particularly low in Ethiopia, Kyrgyzstan and the

Philippines compared to the proportion of the

labour force educated to tertiary level. On the

other hand, Algeria, despite a relatively low share

of the labour force with a tertiary education

(15.2 per cent), has a labour productivity level

above US$50,000 per person employed – the

third highest among our selected countries. This

is probably attributable to the impact on the

results of the national production of oil and gas.

In figure 4.8, the employment-to-population

ratio is displayed together with a breakdown of

educational attainment in the 12 selected coun-

tries.

As in section 3, the data do not show a clear

relationship between educational attainment and

Owing to a lack of data in KILM table 16a, El Salvador is

not included in this section.

Education and labour markets: Analysing global patterns with the KILM

Figure 4.7. Tertiary level educational attainment and labour productivity

Source: KILM, 9th edn, tables 14a, 16a, latest year available.

Thousand US$ / %

Ethiopia, 2012

Algeria, 2011

Cambodia, 2012

Brazil, 2013

Thailand, 2013

Germany, 2014

Mexico, 2011

Philippines, 2008

Egypt, 2013

Canada, 2014

Philippines, 2008

Kyrgyzstan, 2013

Labour productivity per person employed (thousand US$) Share of labour force with tertiary level of education (%)

16.4

2.8

18.2

25.0

12.8

13.4

18.7

23.3

15.2

63.4

27.1

Figure 4.8. Employment-to-population ratio and educational attainment

Source: KILM, 9th edn, tables 2b, 14a, ages 15+, latest year available.

El Salvador, 2013

Ethiopia, 2012

Algeria, 2011

Cambodia, 2012

Brazil, 2013

Thailand, 2013

Germany, 2014

Mexico, 2011

Egypt, 2013

Canada, 2014

Philippines, 2008

Kyrgyzstan, 2013

100

Primary or less Secondary Tertiary

Employment-to-population ratio

57.3

61.4

57.4

58.6

56.5

42.1

54.0

36.2

69.4

82.8

79.4

59.9

Education and labour markets: Analysing global patterns with the KILM

tion of those with a low level of education is simi-

lar to that in the high income countries but the

share of employees stands at only 53.5 per cent.

This section of the chapter has highlighted

once again the clear association that exists

between higher levels of education and positive

outcomes on some key labour indicators, such as

the share of employees in total employment and

labour productivity. However, no consistent link

is apparent between unemployment rates and

education levels. In developing countries, un-

employment may increase with educational

attainment, while in the developed economies in

this sample it tends to do the opposite.

4.3. Remaining gaps in education

The study of the educational patterns of the

labour force in these 12 selected countries

reveals that there are still some gaps that remain

to be addressed, particularly in developing econ-

omies. Here we will present the main areas for

improvement.

4.3.1. Persistent low levels

of educational attainment

There are still a considerable number of coun-

tries where a significant share of the labour force

employment-to-population ratio. Countries with

the lowest shares of primary or less educa-

tional attainment (Kyrgyzstan, Canada, Germany,

Philippines and Mexico) all have employment-to-

population ratios of between 50 and 60 per cent.

Among countries with lower average educational

attainment, there is a wide range of employment-

to-population ratios, from Egypt (42.1 per cent)

and Algeria (36.2 per cent) to Thailand (69.4 per

cent), Cambodia (82.8 per cent) and Ethiopia

(79.4 per cent). El Salvador has an employment-

to-population ratio of 59.9 per cent, similar to

those of Canada and the Philippines, despite

having a very different structure of educational

attainment. These data show no obvious pattern

that could establish a link between levels of

education and employment-to-population ratios.

As in section 3.4, in figure 4.9 we have used

data from KILM table 3 on employment by status

in employment, focusing particularly on the

cat egory “employees” and setting this share of

total employment alongside the share of the

labour force educated to no higher than primary

level. In Ethiopia, the likelihood of having no

more than a primary level of education is high

(almost 80 per cent) while the share of employ-

ees is low (10 per cent). In Germany and Canada,

the results are exactly the opposite. One interest-

ing exception is Kyrgyzstan, where the propor-

Figure 4.9. Share of employees and educational attainment

Source: KILM, 9th edn, tables 3, 14a, ages 15+, latest year available.

El Salvador, 2013

Algeria, 2011

Cambodia, 2012

Brazil, 2013

Thailand, 2013

Germany, 2014

Mexico, 2011

Cambodia, 2012

Philippines, 2008

Egypt, 2013

Canada, 2014

Philippines, 2008

Ethiopia, 2012

Kyrgyzstan, 2013

Share of employees Educational attainment (primary or less)

Education and labour markets: Analysing global patterns with the KILM

Great progress does appear to have been

made in reducing inequalities in access to

education and in educational attainment

between women and men. Indeed, the KILM

data for our selected countries show that in the

great majority of cases the share of the female

labour force having attained tertiary education

is higher than that of the male labour force (see

KILM table 14a). Also, as shown in figure 4.10,

while female unemployment rates remain

higher than male unemployment rates for all

levels of education in some countries, this

disparity is not widespread.

However, even though progress has been

made in reducing gender disparities in educa-

tional attainment, girls still face major obstacles

in accessing school in some parts of the world. It

is important to address these barriers, to ensure

that girls around the world have the opportunity

to complete secondary and, where appropriate,

tertiary education (UNESCO, 2012).

When we turn to examine the differences by

age group, it becomes evident that young people

is educated to no higher than primary level.

The

six countries in our sample with the highest share

of the labour force having only primary education

or less (El Salvador, Ethiopia, Cambodia, Thailand,

Algeria and Brazil) are all low or middle income

economies. More specifically, in El Salvador,

Ethiopia, Cambodia, Thailand and Algeria, the

percentage of the labour force educated to primary

level or below is well over 60 per cent. This shows

that there is still much to be done to increase

general levels of educational attainment in low and

middle income economies, including to improve

access to higher quality employment.

4.3.2. Disparities between population

groups

Research shows (UNESCO, 2015a, b) that there

are still strong disparities in educational attainment

and in returns to education not only between

countries with different levels of income and

development, but also between different popula-

tion groups within countries. Of particular concern

is the persistence of vulnerable groups for whom

access to quality education is very difficult.

Figure 4.10. Male and female unemployment rates by educational attainment

Note: Owing to lack of data, Cambodia, Ethiopia and the Philippines are not included in this figure.

Source: KILM, 9th edn, table 14c, ages 15+, latest year available.

Male

Female

Male

Female

Male

Female

Male

Female

Male

Female

Male

Female

Male

Female

Male

Female

Male

Female

El Salvador AlgeriaBrazil GermanyMexicoEgypt CanadaKyrgyzstanThailand

Primary education or Less Secondary education Tertiary education

Education and labour markets: Analysing global patterns with the KILM

nation of gender and age disparities in access to

and achievement in education is significantly

hindering the entry of young women into the

labour market.

There are still, too, marked inequalities in

educational attainment by household. Notable

differences in youth literacy persist between

people in the richest households and those in the

poorest. Bridging the wealth gap in opportunity

for education is fundamental to promoting inclu-

sive and sustained growth and development for all

countries (UNESCO, 2015a, b).

Finally, we need to consider the situation of

migrants in respect of their access to education and

the labour market in host countries. Extraordinary

efforts are needed to ensure that migrant youth

have equitable access to the acquisition of the skills

they need to enter the labour market (UNESCO,

2015a, b). The increasing flows of labour migration

are also creating an urgent need to consider the

“internationality” of educational qualifications and

other educational arrangements.

are particularly vulnerable in the labour market. In

general, unemployment rates for youth tend to be

higher than corresponding rates for adults, for all

levels of educational attainment. Figure 4.11

provides some examples of this comparison.

Research shows that, for the younger gener-

ation, completion of secondary level education is

no longer enough to secure a satisfactory situ-

ation

within the labour market (Sparreboom and

Staneva, 2014).

It is also important to highlight that, as shown

in figure 4.4, the share of youth not in education,

employment or training is considerably higher

among women than among men in seven of the

selected countries. Thus it seems that the combi-

The four countries were chosen for reasons concerning

the availability and reliability of data.

Based on the ILO School-to-Work Transition Survey defi-

nition, youth are considered “transited” if they have a stable

job, if they are in a satisfactory temporary job or in satisfactory

self-employment.

Figure 4.11. Youth and adult unemployment rates by educational attainment

Source: KILM, 9th edn, table 14b, latest year available for each country.

Youth (15-24) Adult (25+) Youth (15-24) Adult (25+) Youth (15-24) Adult (25+) Youth (15-24) Adult (25+)

El Salvador BrazilGermanyMexico

Primary education or less Secondary education Tertiary education

Education and labour markets: Analysing global patterns with the KILM

where a large share of the labour force has

received only primary education or less.

In some countries there seems to be a

mismatch between supply of and demand for

skilled labour. Where the demand is lower than

the supply, high levels of education are unlikely

to protect against unemployment. However, in

some national contexts highly educated indi-

viduals may have higher expectations in terms

of potential jobs, and be less willing to compro-

mise. In other contexts, a high level of educa-

tional attainment can provide individuals with

easier access to jobs of better quality, offering

higher salaries, improved working conditions,

permanent contracts, full-time employment and

other benefits.

In addition to its positive effects at the indi-

vidual level, increased educational attainment,

coupled with sufficient productive employment

opportunities, can also have a positive impact at

the national level, promoting inclusive economic

growth and helping to reduce income inequalities.

References

Elder, S. 2015. What does NEETs mean and why

is the concept so easily misinterpreted?,

Work4Youth Technical Brief No. 1 (Geneva,

ILO).

Hanushek, E.A.; Woessmann, L. 2008. “The role of

cognitive skills in economic development”, in

Journal of Economic Literature, Vol. 46, No. 3,

pp. 607−668.

Holland, D.; Liadze, I.; Rienzo, C.; Wilkinson, D.

2013. The relationship between graduates

and economic growth across countries, BIS

Research paper No. 110.

ILO. 2008. Conclusions on skills for improved

productivity, employment growth and devel-

opment, International Labour Conference

(Geneva).

—.

2015. Global Employment Trends for Youth

2015: Scaling up investments in decent jobs

for youth (Employment Policy Department,

Geneva).

—.

2016. Women at Work – Trends 2016 (Geneva).

Keller, K.R.I. 2010. “How can education policy

improve income distribution? An empirical

analysis of education stages and measures on

income inequality”, in Journal of Developing

Areas, Vol. 43, No. 2, pp. 51−77.

Organisation for Economic Co-operation and

Development (OECD); Statistics Canada. 2000.

Literacy in the information age: Final report

4.3.3. Attention to qualitative factors

and field of study

This chapter is based primarily on quantitative

indicators. Nevertheless, it is crucial to bear in

mind the qualitative factors that also influence

the role of education in labour market outcomes.

For example, a study of 11 African countries found

that in all these countries “learning deficits”, that

is, lack of ability to provide all students with the

necessary knowledge, skills, cultural and social

understandings,

etc., are considerably greater

than simple “access deficits”, that is, the lack of

universal enrolment in the schooling system

(Spaull and Taylor, 2015). Another study revealed

that the cognitive skills of the population are

much more closely linked to individual earnings,

income distribution and economic growth than

they are to level of educational attainment alone.

The authors found that international comparisons

incorporating cognitive skills show much larger

skills deficits in developing countries than those

generally derived solely from educational enrol-

ment and attainment measures (Hanushek and

Woessmann, 2008). The key point arising from

both these examples is that an exclusive focus on

enrolment and educational attainment may gener-

ate misleading inferences.

The quality and relevance of the national

schooling system (in terms of compulsory educa-

tion, for instance) will also strongly influence the

impact of educational attainment on the individ-

ual’s labour market status, and on returns to

education in general.

With respect to tertiary education, the choice

of the field of study and its relevance in the labour

market will greatly influence returns to educa-

tion. Ideally, all individual choices of field of study

should be made in a way that their sum would

align skills and educational supply closely with

skills and educational demand, thus minimizing

mismatch. However, given the difficulty in

predicting future demand for skills, aiming at this

ideal is likely to remain a challenge.

5. Conclusion

This overview of educational patterns among

the labour force has revealed the importance of

educational attainment in achieving satisfactory

labour market outcomes. Education has a positive

effect not only in facilitating access to employ-

ment, but also in improving the chances of gain-

ing quality employment. Thus, it is clear that

promoting higher levels of educational attain-

ment should remain a priority for countries

Education and labour markets: Analysing global patterns with the KILM

of the international adult literacy survey

(Paris).

Ortiz, L. 2010. “Not the right job, but a secure one:

Over-education and temporary employment in

France, Italy and Spain”, in Work, Employment

and Society, Vol. 24, No. 1, pp. 47−64.

Rubb, S. 2003. “Overeducation in the labour

market: A comment and re-analysis of a meta-

analysis”, in Economics of Education Review,

Vol. 22, No. 6, pp. 621−629.

Sparreboom,

T.; Staneva,

A. 2014. Is education

the solution to decent work for youth in

developing economies? Identifying qualif-

ications mismatch from 28 school-to-work

transition surveys, Work4Youth Publication

Series No. 23 (Geneva, ILO).

Spaull, N.;

Taylor, S. 2015. “Access to what?

Creating a composite measure of educational

quantity and educational quality for 11 African

countries”, in Comparative Education

Review, Vol. 59, No. 1, pp. 133−165.

United Nations Educational, Scientific and

Cultural Organization (UNESCO). 2012. Youth

and skills: Putting education to work, EFA

Global Monitoring Report (Paris).

—.

2015a. Education for All 2000−2015:

Achievements and challenges, EFA Global

Monitoring Report (Paris).

—. 2015b. Education 2030: Equity and quality

with a lifelong learning perspective. Insights

from the EFA Global Monitoring Report’s

World Inequality Database on Education

(WIDE), Policy Paper 20 (Paris).

Vinod, H.D.; Kaushik, S.K. 2007. “Human capital

and economic growth: Evidence from devel-

oping countries”, in American Economist,

Vol. 51, No. 1, pp. 29−39.

Annex

International Organization

for Standardization country codes

ISO 3166 – alpha 3

Code Country / territory

ABW Aruba

AFG Afghanistan

AGO Angola

AIA Anguilla

ALB Albania

AND Andorra

ANT Netherlands Antilles

ARE United Arab Emirates

ARG Argentina

ARM Armenia

ASM American Samoa

ATG Antigua and Barbuda

AUS Australia

AUT Austria

AZE Azerbaijan

BDI Burundi

BEL Belgium

BEN Benin

BFA Burkina Faso

BGD Bangladesh

BGR Bulgaria

BHR Bahrain

BHS Bahamas

BIH Bosnia and Herzegovina

BLR Belarus

BLZ Belize

BMU Bermuda

BOL Bolivia, Plurinational State of

BRA Brazil

BRB Barbados

BRN Brunei Darussalam

BTN Bhutan

BWA Botswana

CAF Central African Republic

CAN Canada

CHA Channel Islands

CHE Switzerland

CHL Chile

CHN China

CIV Côte d'Ivoire

CMR Cameroon

COD Congo, Democratic Republic of the

COG Congo

COK Cook Islands

COL Colombia

COM Comoros

CPV Cabo Verde

CRI Costa Rica

CUB Cuba

CUW Curaçao

CYM Cayman Islands

CYP Cyprus

CZE Czech Republic

DEU Germany

DJI Djibouti

DMA Dominica

DNK Denmark

Code Country / territory

Education and labour markets: Analysing global patterns with the KILM

DOM Dominican Republic

DZA Algeria

ECU Ecuador

EGY Egypt

ERI Eritrea

ESH Western Sahara

ESP Spain

EST Estonia

ETH Ethiopia

FIN Finland

FJI Fiji

FLK Falkland Islands (Malvinas)

FRA France

FRO Faroe Islands

FSM Micronesia, Federated States of

GAB Gabon

GBR United Kingdom

GEO Georgia

GGY Guernsey

GHA Ghana

GIB Gibraltar

GIN Guinea

GLP Guadeloupe

GMB Gambia

GNB Guinea-Bissau

GNQ Equatorial Guinea

GRC Greece

GRD Grenada

GRL Greenland

GTM Guatemala

GUF French Guiana

GUM Guam

GUY Guyana

HKG Hong Kong, China

HND Honduras

HRV Croatia

HTI Haiti

HUN Hungary

IDN Indonesia

IMN Isle of Man

IND India

IRL Ireland

IRN Iran, Islamic Republic of

IRQ Iraq

ISL Iceland

ISR Israel

ITA Italy

JAM Jamaica

JEY Jersey

JOR Jordan

JPN Japan

KAZ Kazakhstan

KEN Kenya

KGZ Kyrgyzstan

KHM Cambodia

KIR Kiribati

KNA Saint Kitts and Nevis

KOR Korea, Republic of

KOS Kosovo

KWT Kuwait

LAO Lao People’s Democratic Republic

LBN Lebanon

LBR Liberia

LBY Libya

LCA Saint Lucia

LIE Liechtenstein

LKA Sri Lanka

LSO Lesotho

LTU Lithuania

LUX Luxembourg

LVA Latvia

MAC Macau, China

MAF Saint Martin (French part)

MAR Morocco

MCO Monaco

MDA Moldova, Republic of

MDG Madagascar

MDV Maldives

MEX Mexico

MHL Marshall Islands

MKD Macedonia, the former Yugoslav

Republic of

MLI Mali

MLT Malta

MMR Myanmar

MNE Montenegro

MNG Mongolia

MNP Northern Mariana Islands

MOZ Mozambique

MRT Mauritania

MSR Montserrat

MTQ Martinique

MUS Mauritius

MWI Malawi

MYS Malaysia

MYT Mayotte

NAM Namibia

NCL New Caledonia

Code Country / territory Code Country / territory

Education and labour markets: Analysing global patterns with the KILM

SVN Slovenia

SWE Sweden

SWZ Swaziland

SXM Sint Maarten (Dutch part)

SYC Seychelles

SYR Syrian Arab Republic

Code Country

TCA Turks and Caicos Islands

TCD Chad

TGO Togo

THA Thailand

TJK Tajikistan

TKL Tokelau

TKM Turkmenistan

TLS Timor-Leste

TON Tonga

TTO Trinidad and Tobago

TUN Tunisia

TUR Turkey

TUV Tuvalu

TWN Taiwan, China

TZA Tanzania, United Republic of

UGA Uganda

UKR Ukraine

URY Uruguay

USA United States

UZB Uzbekistan

VCT Saint Vincent and the Grenadines

VEN Venezuela, Bolivarian Republic of

VGB British Virgin Islands

VIR United States Virgin Islands

VNM Viet Nam

VUT Vanuatu

WLF Wallis and Fortuna Islands

WSM Samoa

YEM Yemen

ZAF South Africa

ZMB Zambia

ZWE Zimbabwe

NER Niger

NFK Norfolk Island

NGA Nigeria

NIC Nicaragua

NIU Niue

NLD Netherlands

NOR Norway

NPL Nepal

NRU Nauru

NZL New Zealand

OMN Oman

PAK Pakistan

PAN Panama

PER Peru

PHL Philippines

PLW Palau

PNG Papua New Guinea

POL Poland

PRI Puerto Rico

PRK Korea, Democratic People’s

Republic of

PRT Portugal

PRY Paraguay

PSE Occupied Palestinian Territory

PYF French Polynesia

QAT Qatar

REU Réunion

ROU Romania

RUS Russian Federation

RWA Rwanda

SAU Saudi Arabia

SDN Sudan

SEN Senegal

SGP Singapore

SHN Saint Helena

SLB Solomon Islands

SLE Sierra Leone

SLV El Salvador

SMR San Marino

SOM Somalia

SPM Saint Pierre and Miquelon

SRB Serbia

SSD South Sudan

STP Sao Tome and Principe

SUR Suriname

SVK Slovakia

Code Country / territory Code Country / territory

estimates that are national, meaning there are

no geographical limitations in coverage. This

series of harmonized estimates serves as the

basis of the ILO’s global and regional aggregates

of the labour force participation rate as reported

in the Global Employment Trends series and

made available in the KILM 9th edition software

as table R1. Table 1b on the software is based on

available national estimates.

Use of the indicator

The labour force participation rate indicator

plays a central role in the study of the factors

that determine the size and composition of a

country’s human resources and in making

projections of the future supply of labour. The

information is also used to formulate employ-

ment policies, to determine training needs and

to calculate the expected working lives of the

male and female populations and the rates of

accession to, and retirement from, economic

activity – crucial information for the financial

planning of social security systems.

The indicator is also used for understand-

ing the labour market behaviour of different

categories of the population. The level and

pattern of labour force participation depend

on employment opportunities and the demand

for income, which may differ from one cat-

egory of persons to another. For example,

studies have shown that the labour force

participation rates of women vary systemat-

ically, at any given age, with their marital status

and level of education. There are also import-

ant differences in the participation rates of the

urban and rural populations, and among differ-

ent socio-economic groups.

Malnutrition, disability and chronic sick-

ness can affect the capacity to work and are

therefore also considered as major determin-

ants of labour force participation, particularly

Introduction

The labour force participation rate is a

measure of the proportion of a country’s work-

ing-age population that engages actively in the

labour market, either by working or looking for

work; it provides an indication of the size of the

supply of labour available to engage in the

production of goods and services, relative to

the population at working age. The breakdown

of the labour force (formerly known as the

economically active population) by sex and age

group gives a profile of the distribution of the

labour force within a country.

The labour force participation rate is calcu-

lated by expressing the number of persons in

the labour force as a percentage of the working-

age population. The labour force is the sum of

the number of persons employed and the

number of unemployed. The working-age

population is the population above the legal

working age – often aged 15 and older, but with

variation from country to country based on

national laws and practices.

Table 1 contains national estimates of labour

force participation rates by sex and age group

(total, youth and adult, referring to ages 15+,

15-24 and 25+ years, respectively, unless excep-

tions are indicated in the table). This series

covers 219 economies over the years 1980 to

2014. The KILM contains an additional table of

ILO estimates of labour force participation rates

according to the following standardized age

groups: 15+, 15-24, 15-64, 25-34, 25-54, 35-54,

55-64 and 65+ years. The participation rates in

table 1a of the software version are harmonized

to account for differences in national data and

scope of coverage, collection and tabulation

methodologies as well as for other country-

specific factors such as military service require-

ments.

The series includes both nationally

reported and imputed data and includes only

These labour force estimates, along with projections

of labour force participation rates are also published in the

ILO’s online database ILOSTAT. For further information on

the methodology used to produce harmonized estimates,

see Bourmpoula, V.; Kapsos, S.; Pasteels, J.M.: ILO esti-

mates and projections of the economically active popula-

tion: 1990-2030 (2013 edition) (Geneva, ILO, 2013),

KILM 1. Labour force participation rate

available at: http://www.ilo.org/ilostat/content/conn/ILOSTAT-

ContentServer/path/Contribution%20Folders/statistics/web_

pages/static_pages/EAPEP/EAPEP%20Methodological%20

paper%202013.pdf.

Note 1 continued

KILM 1 Labour force participation rate

workers may be underestimated – particularly

the number of employed persons who (a) work

for only a few hours in the reference period,

especially if they do not do so regularly, (b) are

in unpaid employment, or (c) work near or in

their home, thus mixing work and personal

activities during the day. Since women, more so

than men, are found in these situations, it is to

be expected that the number of women in

employment (and thus the female labour force)

will tend to be underestimated to a larger

extent than the number of men.

Definitions and sources

The labour force participation rate is defined

as the ratio of the labour force to the working-

age population, expressed as a percentage. The

labour force is the sum of the number of

persons employed and the number of persons

unemployed.

Thus, the measurement of the

labour force participation rate requires the

measurement of both employment and un-

employment. Employment should, in principle,

include members of the armed forces, both the

regular army staff and temporary conscripts.

The labour force participation rate is related

by definition to other indicators of the labour

market. The inactivity rate is equal to 100 minus

the labour force participation rate, when the

participation rate is expressed as a number

between 0 and 100. KILM 13 shows the harmo-

nized inactivity rates of persons according to

the standardized age bands used in table 1a of

the KILM software. The employment-to-popu-

lation ratio (KILM 2) is equal to the labour force

participation rate after the deduction of un-

employment from the numerator of the rate.

The unemployment rate (KILM 9) is related to

the labour force participation rate and employ-

ment-to-population ratio in such a way that two

of them determine the value of the third.

Labour force surveys are typically the

preferred source of information for determin-

ing the labour force participation rate and

related indicators. Such surveys can be designed

Resolution concerning statistics of work, employ-

ment and labour underutilization, adopted by the 19th

International Conference of Labour Statisticians, Geneva,

October 2013; available at:

http://www.ilo.org/global/statis-

tics-and-databases/meetings-and-events/international-

conference-of-labour-statisticians/19/WCMS_230304/lang-

-en/index.htm

(see box 2 in KILM 2 for excerpts relating to

employment and box 9 in KILM 9 for excerpts relating to

unemployment, the sum total of which equal the “labour

force” (currently active population)).

in low-income environments. Another aspect

closely studied by demographers is the rela-

tionship between fertility and female labour

force participation. This relationship is used to

predict the evolution of fertility rates from the

current pattern of female participation in

economic activity.

Comparison of the overall labour force

participation rates of countries at different

stages of development reveals a U-shaped rela-

tionship. In less developed economies, labour

force participation rates can be seen to decline

with economic growth. Economic growth is

associated with expanding educational facilities

and longer time spent studying, a shift from

labour-intensive agricultural activities to urban

economic activities, and a rise in earning oppor-

tunities, particularly for the “prime” working

age (25–54 years) of the head of household so

that other household members with lower

earning potential may choose not to work.

These factors together tend to lower the overall

labour force participation rate for both men

and women, although the effect is weaker for

the latter and shows a wider variation.

It is also instructive to look at labour force

participation rates for males and females by age

group. Labour force activity among the young

(15–24 years) reflects the availability of educa-

tional opportunities, while labour force activity

among older workers (55–64 years or

65+ years) gives an indication of the attitude

towards retirement and the existence of social

safety nets for the retired. Labour force partici-

pation is generally lower for females than for

males in each age category. Among persons of

prime working age, the female rates are not

only lower than the corresponding male rates,

but they also typically exhibit a somewhat

different pattern. During this period of their life

cycle, women tend to leave the labour force to

give birth to and raise children, returning – but

at a lower rate – to economically active life

when the children are older. In developed

economies, the profile of female participation

is, however, increasingly becoming similar to

that of men.

To some degree, the way in which the labour

force is measured can have an effect on the

extent to which men and women are included

in labour force estimates. Unless specific prob-

ing questions are built into the survey question-

naire, participation among certain groups of

See, for example, ILO: “Female labour force partici-

pation rate and fertility”, in Key Indicators of the Labour

Market, third edition, Chapter 1 (Geneva, 2003).

KILM 1 Labour force participation rate

mal economy who fall outside the scope of the

survey or census.

For international comparisons of labour

force data, the most comprehensive source is

undoubtedly labour force surveys. Nevertheless,

despite their strength, labour force survey data

may contain non-comparable elements in terms

of scope and coverage, mainly because of differ-

ences in the inclusion or exclusion of certain

geographic areas, and the incorporation or

non-incorporation of military conscripts. Also,

there are variations in national definitions of

the labour force concept, particularly with

respect to the statistical treatment of “contrib-

uting family workers” and “unemployed and

not looking for work”.

Non-comparability may also arise from

differences in the age limits used in measuring

the labour force (formerly known as the

economically active population). Some coun-

tries have adopted non-standard upper-age

limits for inclusion in the labour force, with a

cut-off point of 65 or 70 years, which will affect

broad comparisons, and especially compari-

sons of those at the higher age levels. Finally,

differences in the dates to which the data refer,

as well as the method of averaging over the

year, may contribute to the non-comparability

of the resulting statistics.

To a large extent, these comparability issues

have been addressed in the construction of the

ILO estimates of labour force participation

rates shown in table 1a. Only household labour

force survey and population census data that

are representative of the whole country (with

no geographical limitation) were used in the

construction of the estimates. In countries with

more than one survey source, only one type of

source was used. If a labour force survey was

available for the country, labour force participa-

tion rates derived from this source were chosen

in favour of those derived from population

censuses.

to cover virtually the entire non-institutional

population of a given country, all branches of

economic activity, all sectors of the economy

and all categories of workers, including the self-

employed, contributing (unpaid) family work-

ers, casual workers and multiple jobholders. In

addition, such surveys generally provide an

opportunity for the simultaneous measure-

ment of the employed, the unemployed and

persons outside the labour force in a coherent

framework.

Population censuses are another major

source of data on the labour force and its

components. The labour force participation

rates obtained from population censuses,

however, tend to be lower, as the vastness of

the census operation inhibits the recruitment

of trained interviewers and does not typically

allow for detailed probing on the labour market

activities of the respondents.

Limitations to comparability

National data on labour force participation

rates may not be comparable owing to differ-

ences in concepts and methodologies. The

single most important factor affecting data

comparability is the data source. Labour force

data obtained from population censuses are

often based on a restricted number of ques-

tions on the economic characteristics of indi-

viduals, with little possibility of probing. The

resulting data, therefore, are generally not

consistent with corresponding labour force

survey data and may vary considerably from

one country to another, depending on the

number and type of questions included in

the census. Establishment censuses and surveys

can – by their nature – only provide data on the

employed population, leaving out the un-

employed and, in many countries, workers

engaged in small establishments or in the infor-

Employment Trends series and made available

in the KILM 9th Edition software as table R2.

Table 2b is based on available national esti-

mates of employment-to-population ratios.

Use of the indicator

The employment-to-population ratio

provides information on the ability of an econ-

omy to create employment; for many countries

the indicator is often more insightful than the

unemployment rate. Although a high overall

ratio is typically considered as positive, the

indicator alone is not sufficient for assessing

the level of decent work or decent work defi-

cits.

Additional indicators are required to

assess such issues as earnings, hours of work,

informal sector employment, underemploy-

ment and working conditions. In fact, the ratio

could be high for reasons that are not necessar-

ily positive – for example, where education

options are limited, young people tend to take

up any work available rather than staying in

school to build their human capital. For these

reasons, it is strongly advised that indicators

should be reviewed collectively in any evalua-

tion of country-specific labour market

policies.

The concept that employment – specifically,

access to decent work – is central to poverty

reduction was firmly acknowledged in the

framework of the Millennium Development

Goals (MDG) with the adoption of an employ-

ment-based target under the goal of halving the

share of the world’s population living in

extreme poverty. The employment-to-popula-

tion ratio was adopted as one of four indicators

to measure progress towards target 1b on

“achieving full and productive employment and

decent work for all, including women and

Since the publication of ILO: Decent Work, Report of

the Director-General, International Labour Conference,

87th Session, 1999 (Geneva, 1999), the goal of “decent

work” has come to represent the central mandate of the

ILO, bringing together standards and fundamental prin-

ciples and rights at work, employment, social protection

and social dialogue in the formulation of policies and

programmes aimed at “securing decent work for women

and men everywhere”. For more information, see: http://

www.ilo.org/decentwork.

Introduction

The employment-to-population ratio

defined as the proportion of a country’s work-

ing-age population that is employed. A high

ratio means that a large proportion of a coun-

try’s population is employed, while a low ratio

means that a large share of the population is not

involved directly in market-related activities,

because they are either unemployed or (more

likely) out of the labour force altogether.

Virtually every country in the world that

collects information on labour market status

should, theoretically, have the requisite infor-

mation to calculate employment-to-population

ratios, specifically, data on the working-age

population and total employment. Both

components, however, are not always

published, nor is it always possible to obtain

the age breakdown of a population, in which

case data are provided for employment only

with no accompanying ratio. Table 2 in the

KILM shows employment-to-population ratios

for 215 economies, disaggregated by sex and

age group (total, youth and adult), where

possible.

KILM 2 also contains ILO estimates of

employment-to-population ratios, which can

help complement missing observations. The

series (table 2a) is harmonized to account for

differences in national data and scope of cover-

age, collection and tabulation methodologies

as well as for other country-specific factors such

as military service requirements.

It includes

both nationally reported and imputed data and

includes only estimates that are national, mean-

ing there are no geographic limitations in

coverage. This series of harmonized estimates

serves as the basis of the ILO’s global and

regional aggregates of the employment-to-

population ratio reported in the Global

In this text, we sometimes shorten the term to

“employment ratio”.

For further information on the methodology used to

harmonize estimates, see Annex 4, “Note on global and

regional estimates”, in ILO: Global Employment Trends

2011 (Geneva, 2011); http://www.ilo.org/global/publica-

tions/books/WCMS_150440/lang--en/index.htm.

KILM 2. Employment-to-population ratio

KILM 2 Employment-to-population ratio

and programmes that allow them to balance

work and family responsibilities.

Definitions and sources

The employment-to-population ratio is the

proportion of a country’s working-age popula-

tion that is employed. The youth and adult

employment-to-population ratios are the propor-

tions of the youth and adult populations – typ-

ically persons aged 15 to 24 years and 25 years

and over, respectively – that are employed.

Employment is defined in the resolution

adopted by the 19th International Conference

of Labour Statisticians (ICLS) as persons of

working age who, during a short reference

period, were engaged in any activity to produce

goods or provide services for pay or profit,

whether at work during the reference period

(i.e. who worked in a job for at least one hour)

or not at work due to temporary absence from

a job, or to working-time arrangements

(see

box 2).

For most countries, the working-age popu-

lation is defined as persons aged 15 years and

older, although this may vary from country to

country. For many countries, this age corres-

ponds directly to societal standards for educa-

tion and work eligibility. However, in some

countries, particularly developing ones, it is

often appropriate to include younger workers

because “working age” can, and often does,

begin earlier. Some countries in these circum-

stances use a lower official bound and include

younger workers in their measurements.

Similarly, some countries have an upper limit

for eligibility, such as 65 or 70 years, although

this requirement is imposed rather infrequently.

The variations on age limits also affect the

youth and adult cohorts.

Apart from issues related to age, the popula-

tion base for employment ratios can vary across

countries. In most cases, the resident non-insti-

tutional population of working age living in

private households is used, excluding members

of the armed forces and individuals residing in

mental, penal or other types of institution.

Resolution concerning statistics of work, employ-

ment and labour underutilization, adopted by the 19th

International Conference of Labour Statisticians, Geneva,

2013; available at: http://www.ilo.org/global/statistics-and-

databases/standards-and-guidelines/resolutions-adopted-

by-international-conferences-of-labour-statisticians/

WCMS_230304/lang--en/index.htm.

young people”.

With the MDGs coming to an

end in 2015, the crucial role of decent work in

poverty reduction was reinforced in the succes-

sors to the MDGs, the Sustainable Development

Goals (SDGs). In fact, the eighth SDG const-

itutes the goal of “promoting inclusive and

sustainable economic growth, employment and

decent work for all”.

Employment-to-population ratios are

becoming increasingly common as a basis for

labour market comparisons across countries or

groups of countries. Employment numbers

alone are inadequate for purposes of compari-

son unless expressed as a share of the popula-

tion who could be working. One might assume

that a country employing 30 million persons is

better off than a country employing 3 million

persons, whereas the addition of the working-

age population component would show

another picture; if there are 3 million persons

employed in Country A out of a possible

5 million persons (60 per cent employment-to-

population ratio) and 30 million persons

employed in Country B out of a possible

70 million (43 per cent employment-to-popula-

tion ratio), then the employment-generating

capacity of Country A is superior to that of

Country B. The use of a ratio helps determine

how much of the population of a country – or

group of countries – is contributing to the

production of goods and services.

Employment-to-population ratios are of

particular interest when broken down by sex,

as the ratios for men and women can provide

information on gender differences in labour

market activity in a given country. However, it

should also be emphasized that this indicator

has a gender bias in so far as there is a tendency

to undercount women who do not consider

their work as “employment” or are not

perceived by others as “working”. Women are

often the primary child caretakers and respon-

sible for various tasks at home, which can

prohibit them from seeking paid employment,

particularly if they are not supported by socio-

cultural attitudes and/or family-friendly policies

The first MDG included three targets and nine indica-

tors; see the official list at: http://mdgs.un.org/unsd/mdg/

Host.aspx?Content=Indicators/OfficialList.htm. The

remaining indicators under the target on decent work were

the growth rate of GDP per person engaged (i.e. labour

productivity growth; KILM 17), working poverty (KILM 18)

and the vulnerable employment rate (KILM 3).

The official list of SDGs and their corresponding

targets (including for Goal 8) can be found at: http://www.

un.org/sustainabledevelopment/sustainable-development-

goals/.

KILM 2 Employment-to-population ratio

ment statuses. For example, some countries

measure persons employed in paid employ-

ment only and some countries measure only

“all persons engaged”, meaning paid employ-

ees plus working proprietors who receive some

remuneration based on corporate shares.

Additional variations that apply to the “norms”

pertaining to measurement of total employ-

ment include hours limits (beyond one hour)

placed on contributing family members before

inclusion.

For most cases, household labour force

surveys are used, and they provide estimates

that are consistent with ILO definitional and

collection standards. A small number of coun-

tries use other sources, such as population

censuses, official estimates or specialized living

standards surveys, which can cause problems of

comparability at the international level.

Comparisons can also be problematic when

the frequency of data collection varies widely.

The range of information collection can run

from one month to 12 months in a year. Given

the fact that seasonality of various kinds is

undoubtedly present in all countries, employ-

ment ratios can vary for this reason alone. Also,

changes in the level of employment can occur

throughout the year, but this can be obscured

when fewer observations are available.

Countries with employment-to-population

ratios based on less than full-year survey peri-

ods can be expected to have ratios that are not

directly comparable with those from full-year,

month-by-month collections. For example, an

annual average based on 12 months of observa-

tions, all other things being equal, is likely to be

different from an annual average based on four

(quarterly) observations.

Such exceptions are noted in the “Coverage limita-

tion” field of all KILM tables relating to employment. The

higher minimum hours used for contributing family work-

ers is in keeping with an older international standard

adopted by the ICLS in 1954. According to the 1954 ICLS,

contributing family workers were required to have worked

at least one-third of normal working hours to be classified

as employed. The special treatment was abandoned at the

1982 ICLS.

Many countries, however, include the armed

forces in the population base for their employ-

ment ratios even when they do not include

them in the employment figures. In general,

information for this indicator is derived from

household surveys, including labour force

surveys. Some countries, however, use “official

estimates” or population censuses as the source

of their employment figures.

Limitations to comparability

Comparability of employment ratios across

countries is affected most significantly by vari-

ations in the definitions used for the employ-

ment and population figures. Perhaps the

biggest differences result from age coverage,

such as the lower and upper bounds for labour

force activity. Estimates of both employment

and population are also likely to vary according

to whether members of the armed forces are

included.

Another area with scope for measurement

differences has to do with the national treat-

ment of particular groups of workers. The

international definition calls for inclusion of all

persons who worked for at least one hour

during the reference period.

The worker could

be in paid employment or in self-employment,

including in less obvious forms of work, some

of which are dealt with in detail in the resolu-

tion, such as unpaid family work, apprentice-

ship or non-market production. The majority of

exceptions to coverage of all persons employed

in a labour force survey have to do with slight

national variations in the international recom-

mendation applicable to the alternate employ-

The application of the one-hour limit for classification

of employment in the international labour force framework

is not without its detractors. The main argument is that clas-

sifying persons who engaged in economic activity for only

one hour a week as employed, alongside persons working

50 hours per week, leads to a gross overestimation of labour

utility. Readers who are interested to find out more on the

topic of measuring labour underutilization may refer to ILO:

Beyond unemployment: Measurement of other forms of

labour underutilization, Room Document 13, 18th Inter-

national Conference of Labour Statisticians, Working Group

on Labour Underutilization, Geneva, 24 November –

5 December 2008; http://www.ilo.org/global/statistics-and-

databases/meetings-and-events/international-conference-of-

labour-statisticians/WCMS_100652/lang--en/index.htm.

KILM 2 Employment-to-population ratio

Box 2. Resolution concerning statistics of work, employment

and labour underutilization, adopted by the 19th International

Conference of Labour Statisticians, October 2013

[relevant paragraphs]

Concepts and definitions

Employment (paras 27 to 31)

27. Persons in employment are defined as all those of working age who, during a short reference

period, were engaged in any activity to produce goods or provide services for pay or profit. They

comprise:

a. employed persons “at work”, i.e. who worked in a job for at least one hour;

b. employed persons “not at work” due to temporary absence from a job, or to working-time

arrangements (such as shift work, flexitime and compensatory leave for overtime).

28. “For pay or profit” refers to work done as part of a transaction in exchange for remuneration

payable in the form of wages or salaries for time worked or work done, or in the form of profits

derived from the goods and services produced through market transactions, specified in the most

recent international statistical standards concerning employment-related income.

a. It includes remuneration in cash or in kind, whether actually received or not, and may also

comprise additional components of cash or in-kind income.

b. The remuneration may be payable directly to the person performing the work or indirectly to

a household or family member.

29. Employed persons on “temporary absence” during the short reference period refers to those who,

having already worked in their present job, were “not at work” for a short duration but maintained

a job attachment during their absence. In such cases:

a. “job attachment” is established on the basis of the reason for the absence and in the case of

certain reasons, the continued receipt of remuneration, and/or the total duration of the

absence as self-declared or reported, depending on the statistical source;

b. the reasons for absence that are by their nature usually of short duration, and where “job

attachment” is maintained, include those such as sick leave due to own illness or injury

(including occupational); public holidays, vacation or annual leave; and periods of maternity

or paternity leave as specified by legislation;

c. reasons for absence where the “job attachment” requires further testing, include among

others: parental leave, educational leave, care for others, other personal absences, strikes or

lockouts, reduction in economic activity (e.g. temporary lay-off, slack work), disorganization

or suspension of work (e.g. due to bad weather, mechanical, electrical or communication

breakdown, problems with information and communication technology, shortage of raw ma-

terials or fuels):

i. for these reasons, a further test of receipt of remuneration and/or a duration threshold

should be used. The threshold should be, in general, not greater than three months taking

into account periods of statutory leave entitlement specified by legislation or commonly

practiced, and/or the length of the employment season so as to permit the monitoring of

seasonal patterns. Where the return to employment in the same economic unit is guaran-

teed this threshold may be greater than three months;

ii. for operational purposes, where the total duration of the absence is not known, the elapsed

duration may be used.

30. Included in employment are:

a. persons who work for pay or profit while on training or skills-enhancement activities required

by the job or for another job in the same economic unit, such persons are considered as

employed “at work” in accordance with the international statistical standards on working time;

b. apprentices, interns or trainees who work for pay in cash or in kind;

c. persons who work for pay or profit through employment promotion programmes;

d. persons who work in their own economic units to produce goods intended mainly for sale or

barter, even if part of the output is consumed by the household or family;

KILM 2 Employment-to-population ratio

(box 2 continued)

e. persons with seasonal jobs during the off season, if they continue to perform some tasks and

duties of the job, excluding, however, fulfilment of legal or administrative obligations (e.g. pay

taxes), irrespective of receipt of remuneration;

f. persons who work for pay or profit payable to the household or family,

in market units operated by a family member living in the same or in another household; or

ii. performing tasks or duties of an employee job held by a family member living in the same

or in another household;

g. regular members of the armed forces and persons on military or alternative civilian service

who perform this work for pay in cash or in kind.

31. Excluded from employment are:

a. apprentices, interns and trainees who work without pay in cash or in kind;

b. participants in skills training or retraining schemes within employment promotion programmes,

when not engaged in the production process of an economic unit;

c. persons who are required to perform work as a condition of continued receipt of a government

social benefit such as unemployment insurance;

d. persons receiving transfers, in cash or in kind, not related to employment;

e. persons with seasonal jobs during the off season, if they cease to perform the tasks and duties

of the job;

f. persons who retain a right to return to the same economic unit but who were absent for

reasons specified in paragraph 29(c), when the total duration of the absence exceeds the

specified threshold and/or if the test of receipt of remuneration is not fulfilled. For analytical

purposes, it may be useful to collect information on total duration of absence, reason for

absence, benefits received, etc.;

g. persons on indefinite lay-off who do not have an assurance of return to employment with the

same economic unit.

Introduction

The indicator of status in employment

distinguishes between two categories of the

total employed. These are: (a) wage and sala-

ried workers (also known as employees); and

(b) self-employed workers. These two groups

of workers are presented as percentages of the

total employed for both sexes and for males

and females separately. Information on the

subcategories of the self-employed group – self-

employed workers with employees (employ-

ers), self-employed workers without employees

(own-account workers), members of produc-

ers’ cooperatives and contributing family work-

ers (formerly known as unpaid family workers)

– is not available for all countries but is

presented wherever possible. Table 3 currently

covers 194 countries.

Use of the indicator

This indicator provides information on the

distribution of the workforce by status in

employment and can be used to answer ques-

tions such as what proportion of employed

persons in a country (a) work for wages or sala-

ries; (b) run their own enterprises, with or

without hired labour; or (c) work without pay

within the family unit? According to the

International Classification of Status in

Employment (ICSE), the basic criteria used to

define the status groups are the types of

economic risk that they face in their work, an

element of which is the strength of institutional

attachment between the person and the job,

and the type of authority over establishments

and other workers that the job-holder has or

will have as an explicit or implicit result of the

employment contract.

Breaking down employment information by

status in employment provides a statistical basis

Resolution concerning the international classifica-

tion of status in employment, adopted by the 15th Interna-

tional Conference of Labour Statisticians, Geneva, 1993;

available at: http://www.ilo.org/public/english/bureau/stat/

download/res/icse.pdf.

for describing workers’ behaviour and condi-

tions of work, and for defining an individual’s

socio-economic group.

A high proportion of

wage and salaried workers in a country can

signify advanced economic development. If, on

the other hand, the proportion of own-account

workers (self-employed without hired employ-

ees) is sizeable, it may be an indication of a

large agricultural sector and low growth in the

formal economy. Contributing family work is a

form of labour – generally unpaid, although

compensation might come indirectly in the

form of family income – that supports produc-

tion for the market. It is particularly common

among women, especially women in house-

holds where other members engage in self-

employment, specifically in running a family

business or in farming. Where large shares of

workers are contributing family workers, there

is likely to be poor development, little job

growth, widespread poverty and often a large

rural economy.

Own-account workers and contributing

family workers are less likely to have formal

work arrangements, and are therefore more

likely to lack elements associated with decent

employment, such as adequate social security

and a voice at work. Therefore, the two statuses

are summed to create a classification of “vulner-

able employment”, while wage and salaried

workers together with employers constitute

“non-vulnerable employment”. The vulnerable

employment rate, which is the share of vulner-

able employment in total employment, was an

indicator of the (no longer applicable) MDG

employment target on decent work.

Globally,

just below half of the employed are in vulner-

able employment, but in many low-income

countries this share is much higher.

The indicator of status in employment is

strongly linked to the employment by sector

indicator (KILM 4). With economic growth, one

would expect to see a shift in employment from

United Nations: Handbook for producing national

statistical reports on women and men, Social Statistics and

Indicators, Series K, No. 14 (New York, 1997), p. 217.

See ILO: Key Indicators of the Labour Market, sixth

edition (Geneva, 2009), Chapter 1, section C; ILO: Key Indi-

cators of the Labour Market, fifth edition (Geneva, 2007),

Chapter 1, section A; and United Nations: The Millennium

Development Goals report 2013 (New York, 2013).

KILM 3.

Status in employment

KILM 3 Status in employment

the agricultural to the industrial and services

sectors, which, in turn, would be reflected in an

increase in the number of wage and salaried

workers. Also, a shrinking share of employment

in agriculture would result in a lower propor-

tion of contributing family workers, who are

often widespread in the rural sector in develop-

ing economies. Countries that show falling

proportions of either own-account workers or

contributing family workers, and a complemen-

tary rise in the share of employees, accompany

the move from a low-income situation with a

large informal or rural sector to a higher-

income situation with high job growth. The

Republic of Korea and Thailand are examples,

where large shifts in status in employment have

accompanied economic growth.

Shifts in proportions of status in employ-

ment are generally not as sharp or as clear as

shifts in sectoral employment. Countries with a

large informal economy, in both the industrial

and services sectors, may have larger propor-

tions of both self-employed and contributing

family workers (and thus higher rates of vulner-

able employment) than a country with a smaller

informal economy. It may be more relevant to

view status in employment within the various

sectors in order to determine whether there

has been a change in their relative shares. Such

a degree of detail is likely to be available in

recently conducted labour force surveys or

population censuses.

Definitions and sources

International recommendations for the

status in employment classification have existed

since before 1950.

In 1958, the United Nations

Statistical Commission approved the

International Classification by Status in

Employment (ICSE). At the 15th International

Conference of Labour Statisticians (ICLS) in

1993, the definitions of categories were revised.

The 1993 revisions retained the existing major

categories, but attempted to improve the

conceptual basis for the distinctions made and

the basic difference between wage employment

and self-employment.

See ILO: Key Indicators of the Labour Market, fifth

edition (Geneva, 2007), Chapter 1, section B.

The Sixth International Conference of Labour Statisti-

cians (1947) and the 1950 Session of the United Nations

Population Commission both made relevant recommenda-

tions for statistics on employment and unemployment and

on population censuses respectively.

The 1993 ICSE categories and extracts from

their definitions follow:

i. Employees are all those workers who hold

the type of jobs defined as “paid employment

jobs”, where the incumbents hold explicit

(written or oral) or implicit employment

contracts that give them a basic remuneration

that is not directly dependent upon the rev-

enue of the unit for which they work.

ii. Employers are those workers who, working

on their own account or with one or a few

partners, hold the type of jobs defined as

“self-employment jobs” (i.e. jobs where the

remuneration is directly dependent upon the

profits derived from the goods and services

produced), and, in this capacity, have engaged,

on a continuous basis, one or more persons to

work for them as employee(s).

iii. Own-account workers are those workers

who, working on their own account or with

one or more partners, hold the type of jobs

defined as “self-employment jobs” [see ii

above], and have not engaged on a continuous

basis any employees to work for them.

iv. Members of producers’ cooperatives are

workers who hold “self-employment jobs”

[see ii or iii above] in a cooperative producing

goods and services.

v. Contributing family workers are those

workers who hold “self-employment jobs” as

own-account workers [see iii above] in a

market-oriented establishment operated by a

related person living in the same household.

vi. Workers not classifiable by status include

those for whom insufficient relevant informa-

tion is available, and/or who cannot be

included in any of the preceding categories.

The status-in-employment indicator pre-

sents all six groups used in the ICSE definitions.

The two major groups – self-employed and

employees – cover the two broad types of status

in employment. The remaining four – employ-

ers (group ii); own-account workers (group iii);

members of producers’ cooperatives (group iv);

and contributing family workers (group v) – are

sub-categories of total self-employed. The

number in each status category is divided by

total employment to arrive at the percentages

shown in table 3. As was mentioned before, the

vulnerable employment rate is calculated as the

sum of contributing family workers and own-

account workers as a percentage of total

employment.

KILM 3 Status in employment

Most of the information for this indicator was

gathered from international repositories of

labour market data, including the ILO

Department of Statistics online database

(ILOSTAT), the Statistical Office of the European

Union (EUROSTAT), the Organisation for

Economic Co-operation and Development

(OECD) Labour Force Statistics Database, and

the Latin America and Caribbean Labour

Information System (QUIPUSTAT) with additions

from the websites of national statistical offices.

Limitations to comparability

The indicator on status in employment can

be used to study how the distribution of the

workforce by status in employment has changed

over time for a particular country; how this

distribution differs across countries; and how it

has developed over the years for different coun-

tries. However, there are often differences in

definitions, as well as in coverage, across coun-

tries and for different years, resulting from

variations in information sources and method-

ologies that make comparisons difficult.

Some definitional changes or differences in

coverage can be overlooked. For example, it is

not likely to be significant that status-in-employ-

ment comparisons are made between countries

using information from labour force surveys

with differing age coverage. (The generally

used age coverage is 15 years and over, but

some countries use a different lower limit or

impose an upper age limit.) In addition, in a

limited number of cases one category of self-

employed – the members of producers’ co-

operatives – are included with wage and salaried

workers. The effects of this non-standard

grouping are likely to be small. More detailed

comparisons within the group of self-employed

are difficult if only combinations of subcat-

egories are available; for example, in a number

of countries own-account workers include

employers, members of producers’ cooper-

atives or contributing family workers for certain

periods.

It is also important to note that information

from labour force surveys is not necessarily

consistent in terms of what is included in

employment. For example, reporting civilian

employment can result in an underestimation

of “employees” and “workers not classifiable by

status”, especially in countries that have large

armed forces. The other two categories, self-

employed and contributing family workers,

would not be affected, although their relative

shares would be.

With respect to geographic coverage, infor-

mation from a source that covers only urban

areas or only particular cities cannot be

compared fairly with information from sources

that cover both rural and urban areas, that is,

the entire country.

For “wage and salaried workers” one needs

to be careful about the coverage, noting

whether, as mentioned above, it refers only to

the civilian population or to the total popula-

tion. Moreover, the status-in-employment

distinctions used in this chapter do not allow

for finer distinctions in working status – in

other words, whether workers have casual or

regular contracts and the kind of protection the

contracts provide against dismissals, as all wage

and salaried workers are grouped together.

When performing queries on this table and tables

4a-d on employment by sector, we strongly recommend

removing countries that do not have national coverage

from the selection when making comparisons across coun-

tries. On the software, this can be done by performing the

query for all data and then refining the parameters to select

“National only” under “Geographic coverage”.

Introduction

The indicator for employment by sector

divides employment into three broad groupings

of economic activity: agriculture, industry and

services. Table 4a presents data for 193 countries

for the three sectors as a percentage of total

employment. Although data are limited to very

few years in the majority of countries in some

regions (such as sub-Saharan Africa and the

Middle East and North Africa), every region is

covered. Because users may be interested in

analysing trends in employment in greater

sectoral detail, the KILM also includes three

tables showing detailed breakdowns of employ-

ment by sector as defined by the International

Standard Industrial Classif ication of All Economic

Activities (ISIC). Table 4b presents employment

by the latest revision, ISIC Revision 4 (2008)

tabulation category as a per centage of total

employment, table 4c presents the same accord-

ing to ISIC Revision 3 (1990) and table 4d pre-

sents the disaggregation according to ISIC

Revision 2 (1968) major divisions (see box 4 for

the list of 1-digit sector levels for each ISIC revi-

sion). Sectoral breakdowns are shown by sex for

virtually all countries covered.

Use of the indicator

Sectoral information is particularly useful in

identifying broad shifts in employment and

stages of development. In the textbook case of

economic development, jobs are reallocated

from agriculture and other labour-intensive

primary activities to industry and finally to the

services sector; in the process, workers migrate

from rural to urban areas. In a large majority of

countries, services are currently the largest

sector in terms of employment. In most of the

remaining countries employment is predomi-

nantly agricultural.

Classification into broad groupings may

obscure fundamental shifts within industrial

patterns. An analysis of tables 4b to 4d, there-

fore, allows identification of individual indus-

tries and services where employment is growing

or stagnating. Teamed with information on job

vacancies by sector, the more detailed data,

viewed over time, should provide a picture of

where demand for labour is focused and, as

such, could serve as a guide for policy-makers

designing skills and training programmes that

aim to improve the match between labour

supply and demand. Of particular interest to

many researchers is employment in the manu-

facturing sector (ISIC 4, tabulation category C,

ISIC 3, tabulation category D and ISIC 2, major

division 3). One could also investigate, for exam-

ple, how employment in the accommodations

and food services sector (ISIC 4, tabulation cat-

egory I and ISIC 3 tabulation category H) has

evolved in countries where tourism comprises a

major portion of gross national product.

It is also interesting to study sectoral employ-

ment flows in connection with productivity

trends (see KILM 16) in order to separate within-

sector productivity growth (e.g. resulting perhaps

from changes in capital or technology) from

productivity growth resulting from shifts of work-

ers from lower- to higher-productivity sectors.

Finally, the breakdown of the indicator by

sex allows for analysis of gender segregation of

employment by sector. Are men and women

equally distributed across sectors, or is there a

concentration of females among the services

sector? Women may be drawn into lower-paying

service activities that allow for more flexible

work schedules thus making it easier to balance

family responsibilities with work life.

Segregation of women in certain sectors may

also result from cultural attitudes that prevent

them from entering industrial employment.

Definitions and sources

For the purposes of the aggregate sectors

shown in table 4a, the agriculture, industry and

services sectors are defined by the ISIC system.

The agriculture sector comprises activities in

United Nations: International Standard Industrial

Classification of All Economic Activities, Series M, No. 4,

Rev. 3 (New York, 1989; Sales No. E.90.XVII.11). Also avail-

able in Arabic, Chinese, French, Russian and Spanish. All

ISIC versions may be found at: http://unstats.un.org/unsd/

cr/registry/.

KILM 4. Employment by sector

KILM 4 Employment by sector

Box 4. International Standard Industrial Classification of All

Economic Activities

Revision 2, 1968 – Major divisions

0 Activities not adequately defined

1 Agriculture, hunting, forestry and fishing

2 Mining and quarrying

3 Manufacturing

4 Electricity, gas and water

5 Construction

6 Wholesale and retail trade and restaurants and hotels

7 Transport, storage and communication

8 Financing, insurance, real estate and business services

9 Community, social and personal services

Revision 3, 1990 – Tabulation categories

A Agriculture, hunting and forestry

B Fishing

C Mining and quarrying

D Manufacturing

E Electricity, gas and water supply

F Construction

G Wholesale and retail trade; repair of motor vehicles, motorcycles and personal and household

goods

H Hotels and restaurants

I Transport, storage and communications

J Financial intermediation

K Real estate, renting and business activities

L Public administration and defence; compulsory social security

M Education

N Health and social work

O Other community, social and personal services activities

P Private households with employed persons

Q Extra-territorial organizations and bodies

X Not classifiable by economic activity

Revision 4, 2008

Revision 4 of ISIC was adopted in August 2008 by the United Nations Statistical Commission and

countries were expected to begin reporting data accordingly in 2009. The revision’s objectives are

to enhance its relevance and comparability with other standard classifications used around the world,

while ensuring its continuity. ISIC Revision 4 incorporates new economic production structures and

activities. Moreover, the structure differs significantly from ISIC Revision 3 in order to better reflect

current economic organization throughout the world. Meanwhile, the proposed classification structure

allows for improved comparison with other standards, such as the Classification of Economic

Activities in the European Community (NACE), the North American Industry Classification System

(NAICS) and the Australian and New Zealand Standard Industrial Classification (ANZSIC). Specifically,

a comprehensive alignment has been retained with NACE at all levels of the classification, while clear

links with NAICS and ANZSIC have been developed at the two-digit level.

In May 2002, ISIC Revision 3.1 superseded Revision 3.0. Because the changes pertain to the more detailed level of the

classification hierarchy only, that is, the 2- to 4-digit level, the 1-digit level data presented in table 4c remain unaltered under

Revision 3.1.

KILM 4 Employment by sector

Information for this indicator has been

assembled from a number of international

repositories and is derived from a variety of

sources, including household or labour force

surveys, official estimates and censuses. In a

very few cases and only where other types of

sources are not available, information is

derived from administrative records and estab-

lishment surveys. The primary repositories

used for the indicator are the ILO’s ILOSTAT

database, and EUROSTAT data, which are based

on the European Labour Force Survey. These

sources are augmented by various regional

repositories, such as QUIPUSTAT, the ILO’s

Latin America and Caribbean Labour

Information System, and by data gathered

directly from publications or websites of

national statistical offices.

Limitations to comparability

Information on a country provided by the

employment-by-sector indicator can differ

according to whether the armed forces, the self-

employed and contributing family members

are included in the estimate. These differences

agriculture, hunting, forestry and fishing, in

accordance with major division 1 of ISIC 2, cat-

egories A and B of ISIC 3 and category A of

ISIC 4. The industry sector comprises mining

and quarrying, manufacturing, construction

and public utilities (electricity, gas and water),

in accordance with major divisions 2 to 5 of

ISIC 2, categories C to F of ISIC 3 or categories

B to F of ISIC 4. The services sector consists of

wholesale and retail trade, restaurants and

hotels, transport, storage and communications,

finance, insurance, real estate and business

services, and community, social and personal

services. This sector corresponds to major divi-

sions 6 to 9 of ISIC 2 or categories G to Q of

ISIC 3 or categories G to U of ISIC 4. See

the table below for a representation of how the

aggregate sectors are calculated according to

the different ISIC revisions:

Aggregate sector ISIC 2

major

divisions

ISIC 3

categories

Agriculture 1 A+B A

Industry 2-5 C-F B-F

Services 6-9 G-Q G-U

Sector not adequately

defined

0 X n/a

(box 4 continued)

Tabulation categories:

A Agriculture, forestry and fishing

B Mining and quarrying

C Manufacturing

D Electricity, gas, steam and air conditioning supply

E Water supply; sewerage, waste management and remediation activities

F Construction

G Wholesale and retail trade; repair of motor vehicles and motorcycles

H Transportation and storage

I Accommodation and food service activities

J Information and communication

K Financial and insurance activities

L Real estate activities

M Professional, scientific and technical activities

N Administrative and support service activities

O Public administration and defence; compulsory social security

P Education

Q Human health and social work activities

R Arts, entertainment and recreation

S Other service activities

T Activities of households as employers; undifferentiated goods- and services-producing activities

of households for own use

U Activities of extraterritorial organizations and bodies

Full details on the latest revision and links to crosswalks between previous revisions are available at: http://unstats.un.org/unsd/

cr/registry/isic-4.asp.

KILM 4 Employment by sector

For some years in certain countries, the sectoral

information relates only to urban areas, so that

little or no agricultural work is recorded. This

is the case for some Latin American countries.

Caution should be used in the analysis of such

data.

Since 1980, different ISIC systems have

sometimes been used concurrently. A slight

majority of countries use Revision 3 as opposed

to Revision 2 or the new Revision 4. The notes

to table 4a show the version of the ISIC used for

each country and year. On occasion, a country

may have continued to use ISIC 2 even after

starting a new data series according to ISIC 3.

In such cases, where two series based on differ-

ent classification systems exist for the same

year, the most recent classification is shown in

table 4a. Although these different classification

systems can have large effects at detailed levels

of industrial classification, changes from one

ISIC to another should not have a significant

impact on the information for the three broad

sectors presented in table 4a.

When performing queries on the employment-by-

sector tables (4a-4d) and table 3 on status in employment,

we strongly recommend removing countries that do not

provide national coverage from the selection when making

comparisons across countries. On the software, this can be

done by performing the query for all data and then refining

the parameters to select “National only” under “Geographic

coverage”.

introduce elements of non-comparability across

countries. When the armed forces are included

in the measure of employment they are usually

allocated to the services sector; the services

sector, therefore, in countries that do not include

armed forces tends to be understated in compar-

ison with countries where they are included.

Information obtained from establishment

surveys covers only employees (wage and salary

earners); thus, the self-employed and contribut-

ing family members are excluded. In such cases,

the employment share of the agriculture sector

in particular is severely under-represented in

comparison with countries that report total

employment without exclusion of status groups.

In table 4a, the only records from an establish-

ment survey or an establishment are found for

Ethiopia (1994) and Belarus (1987–94).

Where information is reported for total

employment or civilian employment for the

entire country, comparability across countries

is reasonable for the employment-by-sector

indicator, because of the similarity in coverage.

Introduction

The indicator for employment by occupa-

tion comprises statistics on jobs classified

according to major groups as defined in one or

more versions of the International Standard

Classification of Occupations (ISCO). The most

recent version of the ISCO, ISCO-08, distin-

guishes ten major groups: (1) Managers; (2)

Professionals; (3) Technicians and associate

professionals; (4) Clerical support workers; (5)

Service and sales workers; (6) Skilled agricul-

tural, forestry and fishery workers; (7) Craft and

related trade workers; (8) Plant and machine

operators and assemblers; (9) Elementary

oc cupations; and (10) Armed forces occupa-

tions. Since 2008 countries have progressively

adapted their national systems to permit them

to report data according to ISCO-08. Data for

earlier years, and for countries that have not yet

adapted their national systems, are classified

according to earlier versions of the classifica-

tion: ISCO-88 and ISCO-68 (see box 5a for the

occup ational groups covered by these two

class ification standards).

Table 5a presents data for the major groups

according to ISCO-08, which are available for

98 countries, and table 5b presents data for the

major groups according to ISCO-88 for

149 countries. Although at least some observa-

tions are available for every region, data are

lacking for numerous countries in sub-Saharan

Africa, and are sparse for the Middle East and

North Africa. Table 5c presents data according

to ISCO-68. This table mostly covers earlier

years, but some countries continue to report

major groups from ISCO-68 alongside

those from ISCO-88. Table 5c contains data on

seven countries.

All tables include both the number of work-

ers by occupation and the share of workers in

an occupational group as a percentage of the

total number of persons employed, and for

men and women separately.

Use of the indicator

Occupational statistics are used for research

on labour market topics ranging from occupa-

tional safety and health to labour market

segmentation. Occupational analyses also

inform economic and labour policies in areas

such as educational planning, migration and

employment services. Occupational informa-

tion is particularly important for the identifica-

tion of changes in skill levels in the labour force.

In many advanced economies, but also in devel-

oping economies, occupational employment

projection models are used to inform policies

aiming to meet future skills needs, as well as to

advise students and jobseekers on expected job

prospects. Ideally, these are conducted on a

more detailed level than the ISCO major groups

and go beyond the information contained in

tables 5a through 5c of the KILM.

Changes in the occupational distribution of

an economy can be used to identify and analyse

stages of development. In the textbook case of

economic development, when labour flows

from agriculture to the industrial and services

sectors, these flows will be visible in the occu-

pational distribution as well. The share of

skilled agricultural and fishery workers will

typically decrease, while rising skill require-

ments are likely to be reflected in a decreasing

share of elementary occupations, rising shares

of high-skilled occupational groups such as

professionals and technicians, and the need for

rising educational attainment levels.

In developed economies, which already

have relatively well-educated labour forces,

increases in the shares of high-skilled occupa-

tional groups (see box 5a) are associated with

the advance of the knowledge economy and

additional changes in the structure of econ-

omies. Furthermore, shifts within occupational

groups may be equally important. For example,

the growing importance of information and

communication technology (ICT) has resulted

in a proliferation of ICT-related jobs.

The breakdown of the indicator by sex allows

for an analysis of gender segregation of employ-

ment. Division of labour markets on the basis of

KILM 5. Employment by occupation

KILM 5 Employment by occupation

skill levels is summarized in box 5a.

The use of

ISCED categories to assist in defining the four

skill levels does not imply that the skills neces-

sary to perform the tasks and duties of a given

job can be acquired only through formal educa-

tion. The skills may be, and often are, acquired

through (informal) training and experience. In

addition, it should be emphasized that the

focus in both ISCO-88 and ISCO-08 is on the

skills required to carry out the tasks and duties

of an occupation, and not on whether a worker

employed in a particular occupation is more or

less skilled, or more or less qualified, than

another worker in the same occupation.

Although the ten major groups defined in

ISCO-88 and ISCO-08 are similar in content

and in name, some occupations are classified in

different major groups according to each of

these two versions. These changes reflect

changes in skill requirements arising from tech-

nological change as well as changes in the way

the concept of skill level was applied to the

design of the classification, to give less empha-

sis to formal educational requirements. Data

classified at major group level according to the

two versions are therefore not strictly compar-

able, and represent a break in series.

Information for this indicator has primarily

been assembled from international reposito-

ries, which have been augmented by some data

gathered directly from the publications or

websites of national statistical offices. The main

repositories for this indicator are ILOSTAT and

EUROSTAT. Additional information is obtained

from national statistical offices. Most of the

information derives from labour force surveys,

but in a limited number of countries, the infor-

mation is gathered from other household

surveys, population censuses, official estimates

and, in particular for table 5c, establishment

surveys.

Limitations to comparability

Information on a country provided by the

employment by occupation indicator can differ

according to whether the armed forces are

included in the estimate. Armed forces constitute

a separate major group, but in some countries

are included in the most closely matching civilian

occupation, depending on the type of work

performed by the individual armed forces

member concerned, or are included in non-clas-

The concept of skills level was introduced in ISCO-88,

and was not used explicitly or systematically in ISCO-68.

sex is one of the most pervasive characteristics

of labour markets around the world, which is

reflected in differentials in occupational distri-

butions between men and women (as well as in

sectoral distributions).

Such differentials can be

analysed at detailed levels of the occupational

classification,

but even at the most aggregated

level, large differences by sex are evident.

Definitions and sources

Tables 5a to 5c classify jobs by occupation.

A job is defined, according to ISCO-08, as a set

of tasks and duties performed, or meant to be

performed, by one person, including for an

employer or in self-employment. An occupa-

tion is defined as a set of jobs whose main tasks

and duties are characterised by a high degree of

similarity.

Occupational classifications categor-

ize all jobs into groups, which are hierarchically

structured in a number of levels. ISCO-08 has a

four-level hierarchy and breaks down its ten

major groups into sub-major groups, minor

groups and unit groups of occupations at its

most detailed level. At the most aggregate level,

there are ten major groups (see box 5a). The

box also lists the major groups defined in

ISCO-88 and ISCO-68. For more details on

ISCO-08, please refer to box 5b.

The ten major groups in ISCO-08 (and in

the previous ISCO-88), are associated with four

broad skill levels. These levels are defined in

relation to the levels of education specified in

the International Standard Classification of

Education (ISCED).

In ISCO-08, the nature of

the work performed in relation to characteristic

tasks, defined for each skill level, takes prece-

dence over formal educational requirements.

The relationship between major groups and

See for example, ILO: Global employment trends for

women 2012 (Geneva, 2012); available at: http://www.ilo.

org/global/research/global-reports/global-employment-

trends/WCMS_195447/lang--nl/index.htm.

See for example, Anker, R.: Gender and jobs. Sex

segregation of occupations around the world (Geneva, ILO,

1998).

Resolution concerning updating the International

Standard Classification of Occupations, adopted by the

Tripartite Meeting of Experts on Labour Statistics on Updat-

ing the International Classification of Occupations (ISCO),

3–6 December 2007; available at: http://www.ilo.org/public/

english/bureau/stat/isco/docs/resol08.pdf.

For further details about ISCED, see the chapter on

KILM 14. The relevant documents related to the latest

version of the ISCED (2011) are available at: http://www.

uis.unesco.org/Education/Pages/international-standard-

classification-of-education.aspx.

KILM 5 Employment by occupation

Box 5a. International Standard Classifications

of Occupations: major groups

Occupational classification ISCO skill level

(see Key below)

ISCO-2008 – Major groups

1 Managers 3+4

2 Professionals 4

3 Technicians and associate professionals 3

4 Clerical support workers 2

5 Service and sales workers 2

6 Skilled agricultural, forestry and fishery workers 2

7 Craft and related trade workers 2

8 Plant and machine operators and assemblers 2

9 Elementary occupations 1

0 Armed forces occupations 1+2+4

ISCO-1988 – Major groups

1 Legislators, senior officials and managers --

2 Professionals 4

3 Technicians and associate professionals 3

4 Clerks 2

5 Service workers and shop and market sales workers 2

6 Skilled agricultural and fishery workers 2

7 Craft and related trades workers 2

8 Plant and machine operators and assemblers 2

9 Elementary occupations 1

0 Armed forces --

ISCO-1968 – Major groups

0/1 Professional, technical and related workers n.a.

2 Administrative and managerial workers n.a.

3 Clerical and related workers n.a.

4 Sales workers n.a.

5 Service workers n.a.

6 Agricultural, animal husbandry and forestry workers, fishermen and hunters n.a.

7/8/9 Production and related workers, transport equipment operators and labourers n.a.

X Workers not classifiable by occupation n.a.

Y Members of the armed forces n.a.

Key: ISCO skill levels

(1) The first ISCO skill level was defined with reference to ISCED category 1, comprising primary

education, which generally begins at the age of 5, 6 or 7 years and lasts about five years.

(2) The second ISCO skill level was defined with reference to ISCED categories 2 and 3, comprising

first and second stages of secondary education. The first stage begins at the age of 11 or 12 years

and lasts about three years, while the second stage begins at the age of 14 or 15 years and also

lasts about three years. A period of on-the-job training and experience may be necessary, some-

times formalized in apprenticeships or traineeships. This period may supplement the formal train-

ing or replace it partly or, in some cases, in full.

KILM 5 Employment by occupation

areas only. Urban coverage is available for some

Latin American countries, and caution should be

used in the analysis of such data.

When performing queries on the employment by

occupation tables (5a to c), we strongly recommend remov-

ing countries that do not have national coverage from the

selection when making comparisons across countries. On

the software, this can be done by performing the query for

all data and then refining the parameters to select “National

only” under “Geographic coverage”.

(box 5a continued)

(3) The third ISCO skill level was defined with reference to ISCED category 5, comprising education

which begins at the age of 17 or 18 years, lasts about four years, and leads to an award not

equivalent to a first university degree.

(4) The fourth ISCO skill level was defined with reference to ISCED categories 6, 7 and 8, comprising

education which also begins at the age of 17 or 18 years, lasts about three, four or more years,

and leads to a university or postgraduate university degree, or the equivalent.

Box 5b. International Standard Classification of Occupations, 2008

ISCO-1988, which was until recently the most widely used international classification of occupations,

has now been superseded by ISCO-08. The revised classification aims to provide:

• acontemporaryandrelevantbasisfortheinternationalreporting,comparisonandexchangeof

statistical and administrative information about occupations;

• ausefulmodelforthedevelopmentofnationalandregionalclassificationsofoccupations;and

• asystemthatcanbeuseddirectlyincountriesthathavenotdevelopedtheirownnationalclas-

sifications.

It should be emphasized that, while serving as a model, ISCO-08 is not intended to replace any

existing national classification of occupations, as the occupation classifications of individual countries

should fully reflect both the structure of the national labour market and information needs for nationally

relevant purposes. However, countries whose occupational classifications are aligned with ISCO-08

in concept and structure will find it easier to develop the procedures to make their occupational

statistics internationally comparable.

Even though the framework and the concepts underpinning ISCO-08 are essentially unchanged from

those used in ISCO-88, there are significant differences in the treatment of some occupational

groups. Some of the more significant changes include (see source for a comprehensive overview):

The sections of the classification dealing with managerial occupations have been reorganized so as

to overcome problems experienced by users of ISCO-88.

Occupations associated with information and communication technology have been updated and

expanded, allowing for the identification of professional and associate professional occupations in

this field as sub-major groups.

Occupations concerned with the provision of health services have been expanded, in order to provide

sufficient detail to allow ISCO-08 to be used as the basis for the international reporting of data on

the health workforce. These occupations have been grouped together, where possible, to provide

two sub-major groups and a separate minor group devoted to occupations in health services.

Source: ILO: International Standard Classification of Occupations (ISCO-08), Geneva, 2012, available at: http://www.ilo.org/

wcmsp5/groups/public/---dgreports/---dcomm/---publ/documents/publication/wcms_172572.pdf.

sifiable workers. In some countries, members of

the armed forces are excluded from important

data sources, such as the Labour Force Survey.

Furthermore, in several countries, certain major

groups are combined into one more aggregated

group. These differences introduce elements of

non-comparability across countries.

If information is based on establishment

surveys, which is mostly limited to table 5c, only

employees are covered, which results in non-

comparability with sources covering all employ-

ment such as labour force surveys. In terms of

the number of countries affected, an even more

important difference is the non-comparability of

data if occupational information relates to urban

Introduction

The indicator on part-time workers focuses

on individuals whose working hours total less

than “full time”, as a proportion of total employ-

ment. Because there is no internationally

accepted definition as to the minimum number

of hours in a week that constitute full-time

work, the dividing line is determined either on

a country-by-country basis or through the use

of special estimations. Two measures are calcu-

lated for this indicator: total part-time employ-

ment as a proportion of total employment,

sometimes referred to as the “part-time employ-

ment rate”; and the percentage of the part-time

workforce comprised of women. Table 6

contains information for 104 economies.

Use of the indicator

There has been rapid growth in part-time

work in the past few decades in developed

economies. This trend is related to the increase

in female labour force participation, but also

results from policies attempting to raise labour

market flexibility in reaction to changing work

organization within industries and to the

growth of the services sector. Of concern to

policy-makers in the apparent move towards

more flexible working arrangements is the risk

that such working arrangements may be less

economically secure and less stable than full-

time employment.

Part-time employment has been seen as an

instrument to increase labour supply. Indeed,

as part-time work may offer the chance of a

better balance between working life and family

responsibilities, and suits workers who prefer

shorter working hours and more time for their

private life, it may allow more working-age

persons to actually join the labour force. Also,

For a review of recent trends in working- time

arrangements, see Boulin, J.-Y. et al. (eds): Decent working

time: New trends, new issues (Geneva, ILO, 2006) and

Messenger, J.C. (ed.): Working time and workers’ prefer-

ences in industrialized countries: Finding the balance

(Routledge, 2004).

policy-makers have promoted part-time work in

an attempt to redistribute working time in

countries of high unemployment, thus lower-

ing politically sensitive unemployment rates

without requiring an increase in the total

number of hours worked.

Part-time employment, however, is not

always a choice. A review of KILM 12, time-

related underemployment, confirms that a

substantial number of part-timers would prefer

to be working full time. While flexibility may be

one advantage of part-time work, disadvantages

may exist in comparison with colleagues who

work full time. For example, part-time workers

may face lower hourly wages, ineligibility for

certain social benefits and more restricted

career and training prospects.

Since the early

1990s, most OECD countries have introduced

measures to improve the quality of part-time

work, for example with respect to social bene-

fits for part-time workers in line with those of

full-time workers. Nevertheless, occupational

segregation between part-time and full-time

work remains an issue in most countries as it

limits the occupational choices of part-time

workers.

Looking at part-time employment by sex is

useful to see the extent to which the female

labour force is more likely to work part time

than the male labour force.

Age breakdowns

are also significant and often demonstrate that

young workers (aged 15 to 24 years) are more

likely than adults (25 years and over) to work

part time.

A suggested virtue of part-time work

is that it facilitates the gradual entry of young

persons into the labour force and the exit of

older workers from the labour market.

“How good is part-time work?”, OECD Position paper,

July 2010; available at: http://www.oecd.org/employment/

emp/48806797.pdf.

See Sparreboom, T.: “Gender equality, part-time work

and segregation”, paper presented at the 73rd Decent Work

Forum, February 2013, ILO, Geneva.

See ILO: Key Indicators of the Labour Market, seventh

edition (Geneva, 2011), Chapter 1, section B, “Gender

equality, employment and part-time work in developed

economies”.

See ILO: Global Employment Trends for Youth 2013:

A generation at risk (Geneva, 2013);

available at:

http://

www.ilo.org/moscow/information-resources/publications/

WCMS_345429/lang--en/index.htm

KILM 6. Part-time workers

KILM 6 Part-time workers

member countries, using a threshold of 30 usual

hours of work in the main job, which is displayed

in the data for table 6 for countries with the

OECD as repository.

Labour force surveys serve as the source of

information on part-time work for nearly all

countries included in table 6. Establishment-

based surveys, in which information on employ-

ees comes directly from payroll records of

establishments, are unlikely to provide infor-

mation on the number of hours that individuals

work and thus cannot be used as a reliable

source for this indicator.

Another reason that labour force surveys are

the preferred source of information for distin-

guishing between full- and part-time work is

that a certain, varying proportion of workers in

all countries possesses more than one job. In

such cases, accounting for the primary jobs of

survey respondents may result in their classifi-

cation as part-time workers, but adding infor-

mation on the second (and possibly third) jobs

may boost their hours over the full-time mark.

In other words, it is the total number of hours

that a person normally works in a week that

determines full- or part-time status, not that

person’s job per se. Only labour force surveys

(and population censuses with fairly extensive

questionnaires) can provide information on the

total number of hours that individuals work.

Nonetheless, many of the countries with infor-

mation based on labour force surveys still

report the number of hours worked on the

main job only, thus disregarding the fact that a

person may work the equivalent of full-time

hours in multiple jobs.

The table notes include the distinction

between “usual” and “actual” hours worked.

“Usual hours” indicates that it is the number of

hours that people typically work in a short

reference period such as one week, over a long

observation period of a month, quarter, season

or year that comprises the short reference

measurement period used.

Usual hours

comprise normal working hours as well as

overtime or extra time usually worked, whether

paid or not. Usual hours do not take into

consideration unplanned leave. As an example,

Users will find information on jobs covered – all jobs,

main job only, etc. – in the “job coverage” field of the data

table.

Resolution concerning the measurement of working

time, adopted by the 18th International Conference of

Labour Statisticians, Geneva, 2008; available at: http://www.

ilo.org/global/statistics-and-databases/standards-and-guide-

lines/resolutions-adopted-by-international-conferences-of-

labour-statisticians/WCMS_112455/lang--en/index.htm.

Definitions and sources

There is no official ILO definition of full-time

work, largely because it is difficult to arrive at an

internationally agreed demarcation point

between full-time and part-time work given the

national variations of what these terms mean. At

the 81st Session of the International Labour

Conference in 1994, the ILO defined “part-time

worker” as “an employed person whose normal

hours of work are less than those of comparable

full-time workers”.

Thus, the demarcation point

is left to the individual countries to define. Some

countries use worker interpretation of their own

employment situation for distinguishing full-time

versus part-time work; that is, survey respondents

are classified according to how they perceive their

work contribution. (See, for example, results in

table 6 based on the European Labour Force

Survey with EUROSTAT as the source.) Other

countries use a cut-off point based on weekly

hours usually or actually worked. Dividing lines

are typically somewhere between 30 and 40

hours a week. Thus, people who work, say, 35

hours or more per week may be considered “full-

time workers”, and those working less than 35

hours “part-time workers”.

The definition of a standard workweek can,

and often does, provide a legal or cultural basis

for the establishment of starting points for

requirements of employee benefits, such as

health care, and overtime premiums for hours

worked in excess of the standard week. It

should be recognized that what might be

thought of as the “standard” workweek for one

country, could be higher than the official

demarcation point for full-time work in a statis-

tical sense. In other words, while a 35- to

40-hour workweek is the probable cut-off stan-

dard for full-time work for many industries and

workplaces throughout much of the world,

national statistical definitions for full-time work

are often somewhere between 30 and 37 hours.

In 1997, the OECD initiated an analysis of

part-time work definitions and concluded that a

definition of part-time work based on a thresh-

old of 30 hours would better suit the purposes

of international comparisons.

Since then, it has

carried out work to harmonize data for its

The 81st Session of the International Labour Confer-

ence adopted the Part-Time Work Convention (No. 175)

and Recommendation (No. 182); texts available at: http://

www.ilo.org/ilolex/english/ convdisp1.htm.

OECD: “The definition of part-time work for the

purpose of international comparisons”, in Labour Market

and Social Policy, Occasional Paper No. 22 (Paris, 1997).

KILM 6 Part-time workers

Use of the OECD data set, discussed in the

previous section, while largely of benefit to

cross-country comparisons, can also have some

negative effects. These will depend on the indi-

vidual situation for each country included in

the set, as countries vary in terms of each of the

following: the range of full-time/part-time

hours cut-off points; standard work-weeks in

general or in particular industries or occupa-

tions; individual conceptual frameworks for

full- and part-time measurement; and the

extent of information available to the OECD for

the estimation and adjustment process.

Although harmonized to the greatest extent

possible, part-time measurement still varies

according to the usual or actual hours criterion

applied. A criterion based on actual hours will

generally yield a part-time rate higher than one

based on usual hours, particularly if there are

temporary reductions in working time as a

result of holiday, illness, etc. Therefore, seasonal

effects will play an important role in fluctua-

tions in actual hours worked. In addition, the

specification of main job or all jobs may be

important. In some countries, the time cut-off is

based on hours spent on the main job; in others,

on total hours spent on all jobs. Measures may

therefore reflect usual or actual hours worked

on the main job or usual or actual hours worked

on all jobs.

Because of these differences, as well as

others that may be specific to a particular coun-

try, cross-country comparisons must be made

with great care. These caveats notwithstanding,

measures of part-time employment can be

quite useful for understanding labour market

behaviour, more for individual countries but

also across countries.

Users with a keen interest in these comparisons

should examine OECD: “The definition of part-time work

for the purpose of international comparisons”, op. cit.

a person who usually works 40 hours a week,

but who was sick for one day (eight hours) in

the survey period, will nevertheless be classi-

fied as a full-time worker (for a country with a

35-hour break point for full-time work).

Limitations to comparability

Information on part-time work can be

expected to differ markedly across countries,

principally because countries use different defi-

nitions of full-time work and also because they

may have different cultural or workplace norms.

The age inclusions for labour force eligibility

can also be an important source of variation.

Entry ages vary across countries, as do upper

age limits. If one country counts everyone over

the age of 10 in the survey, while another starts

at age 16, the two countries can be expected to

have differences in part-time employment rates

for this reason alone. Similarly, some countries

have no upper age bounds for coverage eligibil-

ity, while others draw the line at some point,

such as 65 years. Any cut-off linked to age will

result in some people being missed among the

“employed” counts; as part-time work is particu-

larly prevalent among the older and younger

cohorts, this will lower the measured incidence

of part-time employment. Yet another basis for

variation stems from the definitions used for

“contributing family workers”. Countries that

have no hourly bound for inclusion (one hour

or more) can be expected to have more part-

time workers than those with higher bounds

such as 15 hours.

Introduction

Two measurements related to working time

are included in KILM 7 in order to give an over-

all picture of the time that the employed

throughout the world devote to work activities.

The first measure relates to the hours that

employed persons work per week (table 7a)

while the second measure is the average annual

hours actually worked per person (table 7b).

The statistics in 7a are presented separately for

male and female; according to age group (total,

youth and adult); and employment status

(total, wage and salaried workers and self-

employed). The following hour bands are

applied in table 7a: less than 15 hours worked

per week, between 15 and 29 hours, between

30 and 34 hours, between 35 and 39 hours,

between 40 and 48 hours, and 49 hours and

over, as available. Currently, statistics for

98 economies are presented in table 7a and for

62 economies in table 7b.

Use of the indicator

Issues related to working time have received

intensive attention following labour market

dynamics triggered by the global economic

crisis. Low and stable unemployment rates

despite large drops in output in some advanced

economies have been claimed to be related to

flexibility in working time.

Beyond the medium

run, the number of hours worked has an impact

on the health and well-being of workers.

Some

persons in developed and developing econ-

omies working full time have expressed

concern about their long working hours and its

effects on their family and community life.

Hijzen, A.; Martin, S.: “The role of short-time work

schemes during the global financial crisis and early recov-

ery: a cross-country analysis”, in IZA Journal of Labor Policy,

Vol. 2, No. 5 (Bonn, Institute for the Study of Labor, 2013).

Spurgeon, A.: Working time: Its impact on safety and

health (Geneva, ILO, 2003); available at:

http://www.ilo.

org/travail/whatwedo/publications/WCMS_TRAVAIL_

PUB_25/lang--en/index.htm.

Messenger, J.C. (ed.): Working time and workers’ pref-

erences in industrialized countries: Finding the balance

(Routledge, 2004).

Additionally, the number of hours worked has

an impact on workers’ productivity and on the

labour costs of establishments. Measuring the

level of, and trends in, working time in a soci-

ety, for different groups of persons and for indi-

viduals, is therefore important when monitoring

working and living conditions as well as for

analysing economic and broader social

developments.

Employers have also shown interest in

enhancing the flexibility of working arrange-

ments. They are increasingly negotiating non-

standard working arrangements with their

workers.

Employees may work only part of the

year or part of the week, work at night or on

weekends, or enter or leave the workplace at

different times of the day. They may have vari-

able daily or weekly schedules, perhaps as part

of a scheme that fixes their total working time

over a longer period, such as one month or one

year. Consequently, employed persons’ daily or

weekly working time may show large vari-

ations, and a simple count of the number of

people in employment or the weekly hours of

work is insufficient to indicate the level of, and

trend in, the volume of work.

“Excessive” working time may be a concern

when individuals work more than a “normal”

workweek due to inadequate wages earned

from the job or jobs they hold. In table 7a,

persons could be considered to work excessive

hours if they fall within the 49 hours and over

band. (Workers within the 40–48 hours per

week band are more debatable, and depen-

dant, to a degree, on national circumstances.

Only those at the upper end of the range could

be safely categorized as working excessive

hours.) Long hours can be voluntary or invol-

untary (when imposed by employers).

“Inadequate employment related to excessive

ILO: Report II: Measurement of working time, 18th

International Conference of Labour Statisticians, Geneva,

November-December 2008; http://www.ilo.org/wcmsp5/

groups/public/---dgreports/---stat/documents/publication/

wcms_099576.pdf.

Policy suggestions that preserve health and safety, are

family friendly, promote gender equality, enhance produc-

tivity and facilitate workers’ choice and influence their

working hours are provided in: Lee, S.; McCann, D.;

Messenger, J.: Working time around the world (Geneva,

ILO, 2007).

KILM 7. Hours of work

KILM 7 Hours of work

resolution) and on other activities that are part

of the tasks and duties of the job concerned

(“related hours”). The latter can include, for

example, cleaning and preparing working

tools, and certain on-call duties. The concept

also includes time spent at the place of work

when the person is inactive for reasons linked

to the production process or work organization

(“down time”), as during these periods paid

workers, for example, still remain at the

disposal of their employer, while self-employed

workers will continue working on other tasks

and duties. “Hours actually worked” also

include short rest periods (“resting time”)

spent at the place of work as they are necessary

for human beings and because they are difficult

to distinguish separately, even if paid workers,

for example, are not “at the disposal” of their

employer during those periods. Explicitly

excluded are lunch breaks if no work is

performed, as they are normally sufficiently

long to be easily distinguished from work

pe riods. The international definition relates to

all types of workers – whether in salaried or

self-employment, paid or unpaid, and carried

out in any location, including the street, field,

home, etc.

For some countries, data are available only

according to “hours usually worked”. This

measure identifies the most common weekly

working schedule of a person in employment

over a selected period. The internationally

agreed statistical definition of “usual hours of

work” refers to the hours worked in any job

during a typical short period such as one

week, over a longer period of time, or more

technically, as the modal value of the “hours

actually worked” per week over a longer obser-

vation period.

Average annual hours actually worked, as

presented in table 7b, is a measure of the total

number of hours actually worked during a year

per employed person. The measure incorp-

orates variations in part-time and part-year

employment, in annual leave, paid sick leave

and other types of leave, as well as in flexible

daily and weekly working schedules.

Conventional measures of employment and

weekly hours worked (as in table 7a) cannot do

so. Household-based surveys, unless continu-

ous, are rarely able to measure accurately the

hours actually worked by the population for a

long reference period, such as a year.

Establishment surveys may use longer reference

periods than household surveys but, unlike the

latter, do not cover the whole working popula-

tion. Consequently, the “average annual hours

hours”, also called “over-employment” has

been referred to as “a situation where persons

in employment wanted or sought to work fewer

hours than they did during the reference

period, either in the same job or in another job,

with a corresponding reduction of income”.

Few countries have actually measured “over-

employment” so the measure of persons in

employment for more than 48 hours a week

could be used as a proxy for persons in employ-

ment who usually work beyond what is consid-

ered “normal hours” in many countries.

However, whether or not this situation is actu-

ally desired cannot be assessed, so nothing can

be assumed about how many hours people

might wish to work. Clearly, the number of

hours worked will vary across countries and

depends on, other than personal choice, such

important aspects as cultural norms, real wages

and levels of development.

Definitions and sources

Statistics on the percentage of persons in

employment and in paid employment by hours

worked per week (table 7a) are mostly calculated

on the basis of information on employment and

employees by actual-hour bands provided

primarily by household-based surveys which

cover all persons in employment (exceptions are

identified in the notes to table 7a). In general,

persons totally absent from work during the

reference week are excluded. Annual hours actu-

ally worked per person (table 7b) are estimated

from the results of both household and establish-

ment surveys. For the most part, coverage

comprises total employment or employees (wage

earners and salaried workers).

The “actual hours of work” per week identi-

fies the time that persons in employment effec-

tively spent directly on, and in relation to,

productive activities; down time; and resting

time during the corresponding reference

period (see box 7).

That is, the “actual hours

of work” includes time spent at the workplace

on productive activities (“direct hours” in the

ILO: Final report, 16th International Conference of

Labour Statisticians, Geneva, October 1998; http://www.ilo.

org/public/english/bureau/stat/download/16thicls/repconf.pdf.

Resolution concerning the measurement of working

time, adopted by the 18th International Conference of

Labour Statisticians, Geneva, November-December 2008;

http://www.ilo.org/global/statistics-and-databases/standards-

and-guidelines/resolutions-adopted-by-international-confer-

ences-of-labour-statisticians/WCMS_112455/lang--en/index.

htm. (See box 7 for a summary and relevant paragraphs.)

KILM 7 Hours of work

average number of employed persons during

the year.

Limitations to comparability

Statistics based on hours actually worked

are not strictly comparable to statistics based

on hours usually worked. A criterion using

hours actually worked will generally yield a

higher weekly average than usual hours,

par ticularly if there are temporary reductions in

working time as a result of holiday, illness, etc.

that will have an impact on the measure of aver-

age weekly hours. Seasonal effects will also play

an important role in fluctuations in hours actu-

ally worked. In addition, the specification of

main job or all jobs may be an important one.

In some countries, the time cut-off is based on

hours spent in the main job; in others on total

hours spent in all jobs. Measures may therefore

reflect hours actually or usually worked in the

main job or in all jobs. Because of these and

other differences that may be specific to a

particular country, cross-country comparison in

table 7a should be undertaken with great care.

The different estimation methods for annual

hours of work depend to a large extent on the

type and quality of the information available

and may lead to estimates that are not compar-

able. All estimates presented are derivations

from numbers gathered from surveys and other

sources, usually produced within the national

statistical agency. It is difficult to evaluate the

impact of estimation differences on their

comparability across countries.

The various data collection methods also

represent an important source of variation in the

working time estimates. Household-based

surveys (including population censuses) that

obtain data from working persons or from other

household members can and often do cover the

whole population, thus including the self-

employed. As they use the information respon-

dents provide, responses may contain substantial

errors. On the other hand, the data obtained

from establishment surveys depend on the type,

range and quality of their records on attendance

and payment. While consistency in reporting

overtime may be higher, the information may

contain undetected biases. Furthermore, their

worker coverage is never complete, as these

surveys tend to cover medium-to-large establish-

ments in the formal sector with regular employ-

ees, and exclude managerial and peripheral staff

as well as self-employed persons.

actually worked” is often estimated on the basis

of statistics from both sources.

The Organisation for Economic Co-operation

and Development (OECD) is the source of most

estimates of annual hours actually worked per

person in table 7b. The OECD estimates for

nine countries

reproduced in table 7b are

based on national accounts questionnaires that

measure hours worked by employed persons

(employees and self-employed) in domestic

production during one year. Hours worked

refer to production within effective and normal

working hours, with addition for overtime while

deducting absences due to sickness, leave of

absence, vacations and any labour conflicts. The

estimated hours generated are the same that are

used by national accountants as input for the

calculation of productivity (output per hour

worked). Additional countries provide data

based on their own series that are consistent

with the national accounts (Canada, Finland,

France, Germany, Norway and Sweden).

OECD estimates for Belgium, Ireland,

Luxembourg, the Netherlands and Portugal

apply a second estimation procedure, taking

information from legislation or collective agree-

ments that concern “normal hours”. This

consists of multiplying the weekly “normal

hours” (measured in the European Labour

Force Survey) by the number of weeks that

workers have been in employment during the

year. Annual leave and public holidays are

subtracted to obtain a net amount of “annual

normal time”. Estimates of overtime obtained

from sources such as household or establish-

ment surveys are added, and estimates of time

taken in substantial forms of absences, obtained

from household surveys or administrative

sources, are then subtracted. In practice, some

additional adjustments may be needed when

the “normal hours” vary over the year.

The remainder of OECD country estimates

are based on statistics for time actually worked

for each week of the year, derived from continu-

ous household surveys. Statistics for a month

or quarter when used need to be adjusted for

the number of working days in that period.

Further adjustments are made for public holi-

days and strike activity, normally on the basis of

information obtained from administrative

sources. The resulting estimates may then be

added up to obtain the total annual “hours

actually worked”, which is then divided by the

Austria, Denmark, Greece, Italy, Republic of Korea,

Slovenia, Spain, Switzerland and Turkey.

KILM 7 Hours of work

Box 7. Resolution concerning the measurement of working time,

adopted by the 18th International Conference of Labour

Statisticians, November-December 2008

Summary

In 2008, the International Conference of Labour Statisticians (ICLS) adopted a resolution concerning

the measurement of working time. The resolution revises the existing standards on statistics of hours

of work (Resolution concerning statistics of hours of work, adopted by the 13th ICLS in 1962) in order

to reflect the working time of persons in all sectors of the economy and in all forms of productive

activity towards the achievement of decent work for all, and to provide measurement methodologies

and guidelines on a larger number of measures than previously defined internationally, thereby

enhancing the standards’ usefulness as technical guidelines to States and the consistency and

international comparability of related statistics.

The resolution provides definitions for seven concepts of working time associated with the productive

activities of a person and performed in a job:

• Hours actually worked, the key concept of working time defined for statistical purposes appli-

cable to all jobs and to all working persons;

• Hours paid for, linked to remuneration of hours that may not all correspond to production;

• Normal hours of work, refers to legally prevailing collective hours;

• Contractual hours of work individuals are expected to work according to contractual relationships

as distinct from normal hours;

• Hours usually worked, most commonly in a job over a long observation period,

• Overtime hours of work, performed beyond contracts or norms; and

• Absence from work hours, when working persons do not work;

It also provides definitions for two concepts of working-time arrangements that describe the

characteristics of working time in a job, namely the organization and scheduling of working time,

regardless of type of job, and formalized working-time arrangements, that are specific combinations

of the characteristics having legal recognition.

Relevant paragraphs

Concepts and definitions

Hours actually worked

11.

(1) Hours actually worked is the time spent in a job for the performance of activities that contribute

to the production of goods and/or services during a specified short or long reference period.

Hours actually worked applies to all types of jobs (within and beyond the SNA production

boundary) and is not linked to administrative or legal concepts.

(2) Hours actually worked measured within the SNA production boundary includes time spent directly

on, and in relation to, productive activities; down time; and resting time.

(a)

“Direct hours” is the time spent carrying out the tasks and duties of a job. This may be performed

in any location (economic territory, establishment, on the street, at home) and during overtime

periods or other periods not dedicated to work (such as lunch breaks or while commuting).

(b) “Related hours” is the time spent maintaining, facilitating or enhancing productive activities

and should comprise activities such as:

(i) cleaning, repairing, preparing, designing, administering or maintaining tools, instruments,

processes, procedures or the work location itself; changing time (to put on work clothes);

decontamination or washing up time;

(ii) purchasing or transporting goods or basic materials to/from the market or source;

(iii) waiting for business, customers or patients, as part of working-time arrangements and/or

that are explicitly paid for;

KILM 7 Hours of work

ity of the national estimates presented, is care-

ful to note that “the data [on average annual

hours worked per person] are intended for

comparisons of trends over time; they are

unsuitable for comparisons of the level of aver-

age annual hours of work for a given year,

because of differences in their sources”.

See Statistical Annex of the annual OECD publication:

Employment Outlook.

Comparability of statistics on working time

is complicated even further by the fact that esti-

mates may be based on more than one source

– results may be taken primarily from a house-

hold survey and supplemented with informa-

tion from an establishment survey (or other

administrative source) or vice versa. In such

cases, more than one survey type is noted in the

corresponding column of the notes. For these

reasons, the OECD, which provided the major-

(iv) on-call duty, whether specified as paid or unpaid, that may occur at the work location

(such as health and other essential services) or away from it (for example from home). In

the latter case, it is included in hours actually worked depending on the degree to which

persons’ activities and movements are restricted. From the moment when called back for

duty, the time spent is considered as direct hours of work;

(v) travelling between work locations, to reach field projects, fishing areas, assignments,

conferences or to meet clients or customers (such as door-to-door vending and itinerant

activities);

(vi) training and skills enhancement required by the job or for another job in the same economic

unit, at or away from the work location. In a paid-employment job this may be given by

the employer or provided by other units.

cannot work due to machinery or process breakdown, accident, lack of supplies or power or

Internet access, etc., but continues to be available for work. This time is unavoidable or inher-

ent to the job and involves temporary interruptions of a technical, material or economic nature.

(d) “Resting time” is time spent in short periods of rest, relief or refreshment, including tea, coffee

or prayer breaks, generally practised by custom or contract according to established norms

and/or national circumstances.

(3) Hours actually worked measured within the SNA production boundary excludes time not worked

during activities such as:

(a) Annual leave, public holidays, sick leave, parental leave or maternity/paternity leave, other

leave for personal or family reasons or civic duty. This time not worked is part of absence from

work hours (defined in paragraph 17);

(b) Commuting time between work and home when no productive activity for the job is performed;

for paid employment, even when paid by the employer;

(vi); for paid employment, even when authorized, paid or provided by the employer;

(d) Longer breaks distinguished from short resting time when no productive activity is performed

(such as meal breaks or natural repose during long trips); for paid employment, even when

paid by the employer.

Hours usually worked

15.

(1) Hours usually worked is the typical value of hours actually worked in a job per short reference

period such as one week, over a long observation period of a month, quarter, season or year that

comprises the short reference measurement period used. Hours usually worked applies to all

types of jobs (within and beyond the SNA production boundary).

(2) The typical value may be the modal value of the distribution of hours actually worked per short

period over the long observation period, where meaningful.

(3) Hours usually worked provides a way to obtain regular hours worked above contractual hours.

(4) The short reference period for measuring hours usually worked should be the same as the refer-

ence period used to measure employment or household service and volunteer work.

Introduction

The KILM 8 indicator is a measure of employ-

ment in the informal economy as a percentage

of total non-agricultural employment. There are

wide variations in the definitions and methodol-

ogy of data collection related to the informal

economy. Some countries now provide data

according to the 2003 guidelines concerning a

statistical definition of informal employment.

The KILM 9th edition contains national esti-

mates on informal employment. Where avail-

able, KILM 8 reports on informal employment,

employment in the informal sector and infor-

mal employment outside the informal sector.

Information on employment in the informal

sector, measured according to the resolution of

the 15th International Conference of Labour

Statisticians (ICLS), is also included. Users are

advised to review the specific definitions of each

record carefully and to use caution when

making country-to-country comparisons.

Table 8 contains national estimates for

62 countries in total. A gender-specific break-

down for the indicator is given where the data

are available. In most cases, information on

persons in informal employment is given as

absolute numbers and as a percentage of total

non-agricultural employment.

Use of the indicator

The informal sector represents an impor-

tant part of the economy, and certainly of the

labour market, in many countries and plays a

major role in employment creation, produc-

tion and income generation. In countries with

high rates of population growth or urbaniza-

tion, the informal sector tends to absorb most

of the expanding labour force in urban areas.

Guidelines concerning a statistical definition of

informal employment, adopted by the 17th International

Conference of Labour Statisticians, Geneva, 2003; http://

www.ilo.org/global/statistics-and-databases/standards-and-

guidelines/guidelines-adopted-by-international-conferences-

of-labour-statisticians/WCMS_087622/lang--en/index.htm.

Informal employment offers a necessary

survival strategy in countries that lack social

safety nets, such as unemployment insurance,

or where wages and pensions are low, espe-

cially in the public sector. In these situations,

indicators such as the unemployment rate

(KILM 9) and time-related underemployment

(KILM 12) are not sufficient to describe the

labour market completely.

Globalization is also likely to have contrib-

uted to raising the share of informal employ-

ment in many countries. Global competition

erodes employment relations by encouraging

formal firms to hire workers at low wages with

few benefits or to subcontract (outsource) the

production of goods and services.

In addition,

the process of industrial restructuring in the

formal economy is seen as leading to greater

decentralization of production through subcon-

tracting to small enterprises, many of which are

in the informal sector.

The informal economy represents a chal-

lenge to policy-makers that pursue the follow-

ing goals: improving the working conditions

and legal and social protection of persons in

informal sector employment and of employees

in informal jobs; increasing the productivity of

informal economic activities; developing train-

ing and skills; organizing informal sector

producers and workers; and implementing

appropriate regulatory frameworks, govern-

mental reforms, urban development, and so

on. Poverty, too, as a policy issue, overlaps with

the informal economy. There is a link – although

not a perfect correlation – between informal

employment and being poor. This stems from

the lack of labour legislation and social protec-

tion covering workers in informal employment,

and from the fact that persons in informal

employment earn, on average, less than work-

ers in formal employment.

Statistics on informal employment are essen-

tial to obtaining a clear idea of the contributions

See evidence presented in Bacchetta, M. et al.:

Globalization and informal jobs in developing countries

(Geneva, ILO and WTO, 2009); available at: http://www.wto.

org/english/res_e/booksp_e/jobs_devel_countries_e.pdf.

KILM 8. Employment

in the informal economy

KILM 8 Employment in the informal economy

the broader measure do not exist, the statistical

definition of the latter is also included.

Employment in the informal sector

and informal sector enterprises

The definition of employment in the infor-

mal sector that was formally adopted by the

15th ICLS is based on the concept of the infor-

mal sector enterprise, with all jobs deemed to

fall under such an enterprise included in the

count. In other words, employment in the

informal sector basically comprises all jobs in

unregistered and/or small-scale private un-

incorporated enterprises that produce goods

or services meant for sale or barter.

There are considerable nuances and

complexities to the definition. The term “enter-

prise” is used in a broad sense, as it covers both

units which employ hired labour and those run

by individuals working on their own account or

as self-employed persons, either alone or with

the help of unpaid family members. Workers of

all employment statuses are included if deemed

to be engaged in an informal enterprise. Thus,

self-employed street vendors, taxi drivers and

home-based workers are all considered enter-

prises. The logic behind establishing the criter-

ion based on employment size was that

enterprises below a certain size are often

exempted, under labour and social security

laws, from employee registration and are

unlikely to be covered in tax collection or

labour law enforcement due to lack of govern-

ment resources to deal with the large number

of small enterprises (many of which have a high

turnover or lack easily recognizable features).

Certain activities, which are sometimes iden-

tified with informal activities, are not included

in the definition of informal enterprises for

practical as well as methodological reasons.

Excluded activities include: agricultural and

related activities, households producing goods

exclusively for their own use, e.g. subsistence

farming, domestic housework, care work, and

employment of paid domestic workers; and

volunteer services rendered to the community.

The definition of informal sector enterprises

was subsequently included in the System of

National Accounts (SNA 1993 and 2008), adopted

by the United Nations Economic and Social

Council on the recommendation of the United

Nations Statistical Commission.

Inclusion in the

Information on the System of National Accounts (SNA

1993) is available from the Statistics Division, United Nations,

New York. See http://unstats.un.org/unsd/nationalaccount/.

of all workers, women in particular, to the econ-

omy. Indeed, the informal economy has been

considered as “the fallback position for women

who are excluded from paid employment. [...]

The dominant aspect of the informal economy

is self-employment. It is an important source of

livelihood for women in the developing world,

especially in those areas where cultural norms

bar them from work outside the home or where,

because of conflict with household responsibil-

ities, they cannot undertake regular employee

working hours”.

Definitions and sources

In 1993, the statistical conception of infor-

mal sector activities was adopted at the 15th

ICLS.

More than 20 years later, the concept of

informality has evolved, broadening in scope

from employment in a specific type of produc-

tion unit (or enterprises) to an economy-wide

phenomenon, with the current focus now on

the development and harmonization of infor-

mal economy indicators.

The conceptual

change from the informal sector to the informal

economy (described further below), while

certainly technically sound and commendable

as a reflection of the evolving realities of the

world of work, has resulted in challenges for the

measurement of a concept that was already

fraught with difficulties. The current statistical

concept of informal employment is also

described below. However, because it takes time

for a “new” statistical concept to take hold, some

countries will continue to report on the concept

of employment in the informal sector for a few

years to come. The national statistics are repro-

duced in table 8 and where data according to

United Nations: Handbook for Producing National

Statistical Reports on Women and Men, Social Statistics and

Indicators, Series K, No. 14 (New York, 1997), p. 232.

Large parts of text in this section, including boxes,

are reproduced from ILO: The informal economy and

decent work: A policy resource guide supporting transitions

to formality, chapter 2 (Geneva, ILO, 2013); http://www.ilo.

org/emppolicy/pubs/WCMS_212688/lang--en/index.htm.

Resolution concerning statistics of employment in the

informal sector, adopted by the 15th International Confer-

ence of Labour Statisticians, Geneva, 1993; http://www.ilo.

org/global/statistics-and-databases/standards-and-guidelines/

resolutions-adopted-by-international-conferences-of-labour-

statisticians/WCMS_087484/lang--en/index.htm.

For more details on a comprehensive measurement

of informality, refer to the ILO manual Measuring informal-

ity: A statistical manual on the informal sector and infor-

mal employment (ILO, 2013);

http://www.ilo.ch/global/publi-

cations/ilo-bookstore/order-online/books/WCMS_222979/

lang--en/index.htm.

KILM 8 Employment in the informal economy

In tracing the evolution of the informal

concept (see box 8b), it is important to bear in

mind that the purpose of the expansion to an

informal economy concept was not to replace

one term with another (see box 8a), but rather

to broaden the concept to take into consider-

ation different aspects of the “informalization

of employment”. It is also worth bearing in

mind that for statistical purposes the 17th ICLS

did not endorse using the term “employment

in the informal economy” to represent the

totality of informal activities. The reasons are

(i) that the different types of observation unit

involved (enterprise vs job) should not be

confused, (ii) that some policy interventions

would have to be targeted to the enterprise and

others to the job, and (iii) that the informal

sector concept from the 15th ICLS needed to

be retained as distinct from informal employ-

ment since it had become a part of the SNA and

a large number of countries were already

collecting statistics based on this definition.

The 17th ICLS defined informal employ-

ment as the total number of informal jobs,

whether carried out in formal sector enter-

prises, informal sector enterprises, or house-

holds, during a given reference period.

Included are:

i. Own-account workers (self-employed with no

employees) in their own informal sector

enterprises;

ii. Employers (self-employed with employees) in

their own informal sector enterprises;

iii. Contributing family workers, irrespective of

type of enterprise;

iv. Members of informal producers’ cooperatives

(not established as legal entities);

v. Employees holding informal jobs as defined

according to the employment relationship (in

law or in practice, jobs not subject to national

labour legislation, income taxation, social

protection or entitlement to certain employ-

ment benefits (paid annual or sick leave, etc.));

vi. Own-account workers engaged in production

of goods exclusively for own final use by their

household.

Only items i, ii and iv would have been

captured in full under the statistical framework

for employment in the informal sector. The

remaining statuses might or might not have

been included, depending on the nature of the

production unit under which the activity took

place (i.e. if deemed an informal enterprise).

The major new element of the framework was

SNA was considered essential, as it was a pre-

requisite for identifying the informal sector as a

separate entity in the national accounts and

hence for quantifying the contribution of the

informal sector to gross domestic product.

Informal employment

The definition of the 15th ICLS relates to the

informal sector and the employment therein. But

it has also been recognized within the statistical

community, that there are aspects of informality

that can exist outside informal sector enterprises

as currently defined. Casual, short-term and

seasonal workers, for example, could be infor-

mally employed – lacking social protection,

health benefits, legal status, rights and freedom

of association, but when they are employed in the

formal sector are not considered within the

measure of employment in the informal sector.

In the early 2000s, there was growing

momentum behind the call for more and better

statistics on the informal economy, statistics that

capture informal employment both within and

outside the formal sector. There was a gradual

move among users of the statistics, spearheaded

by the Expert Group on Informal Sector Statistics

(the Delhi Group) – an international forum of

statisticians and statistics users concerned with

measurement of the informal sector and improv-

ing the quality and comparability of informal

sector statistics – towards promotion of this

broader concept of informality. The idea was to

complement the enterprise-based concept of

employment in the informal sector with a

broader, job-based concept of informal employ-

ment. At its fifth meeting in 2001, the Delhi

Group called for the development of a statistical

definition and measurement framework of infor-

mal employment to complement the existing

standard of employment in the informal sector.

The ILO Department of Statistics and the

17th ICLS took up the challenge of developing

new frameworks which could better capture

the phenomenon of informality. The ILO

conceptualized a framework for defining the

informal economy that was presented and

adopted at the 2002 International Labour

Conference. The informal economy was defined

as “all economic activities by workers or

economic units that are – in law or practice –

not covered or sufficiently covered by formal

arrangements”. In 2003, the 17th ICLS adopted

guidelines endorsing the framework as an

international statistical standard.

Guidelines concerning a statistical definition of

informal employment, op. cit.

KILM 8 Employment in the informal economy

• Unregistered employees who do not have

explicit, written contracts or are not subject

to labour legislation;

• Workers who do not benefit from paid annual

or sick leave or social security and pension

schemes;

• Most paid domestic workers employed by

households;

• Most casual, short-term and seasonal workers.

item v, employees holding informal jobs. This

category captures the bulk of the “informal

employment outside the informal sector” in

many countries and includes workers whose

“[…] employment relationship is, in law or in

practice, not subject to national labour legisla-

tion, income taxation, social protection or enti-

tlement to certain employment benefits

(advance notice of dismissal, severance pay,

paid annual or sick leave, etc.)”. These include:

Box 8a. Avoiding confusion in terminologies relating

to the informal economy

Within the statistical community, application of accurate terminology is important. To the layperson,

the terms “informal sector”, “informal economy”, “employment in the informal sector” and “informal

employment” might all seem to be interchangeable. They are not. The nuances associated with each

term are extremely important from a technical point of view. The following can serve as an easy

reference for the terminology associated with informality and the technical definitions:

(a) Informal economy all economic activities by workers or economic units that

are – in law or practice – not covered or sufficiently covered

by formal arrangements (based on ILC 2002)

(b) Informal sector a group of production units (unincorporated enterprises

owned by households) including “informal own-account

enterprises” and “enterprises of informal employers”

(based on 15th ICLS)

enterprises engaged in non-agricultural activities with at

least some of the goods or services produced for sale or

barter (based on 15th ICLS)

(d) Employment in the informal sector all jobs in informal sector enterprises (c), or all persons who

were employed in at least one informal sector enterprise,

irrespective of their status in employment and whether it

was their main or a secondary job (based on 15th ICLS)

(e) Informal wage employment all employee jobs characterized by an employment

relationship that is not subject to national labour legislation,

income taxation, social protection or entitlement to certain

employment benefits (based on 17th ICLS)

(f) Informal employment total number of informal jobs, whether carried out in formal

sector enterprises, informal sector enterprises, or

households; including employees holding informal jobs (e);

employers and own-account workers employed in their

own informal sector enterprises; members of informal

producers’ cooperatives; contributing family workers in

formal or informal sector enterprises; and own-account

workers engaged in the production of goods for own end

use by their household (based on 17th ICLS)

(g) Employment in the informal

economy

sum of employment in the informal sector (d) and informal

employment (f) outside the informal sector; the term was

not endorsed by the 17th ICLS

Source: Reproduced from ILO: The informal economy and decent work: A policy resource guide supporting transitions to formality,

chapter 2 (Geneva, ILO, 2013).

KILM 8 Employment in the informal economy

date country situations and specific country

needs. In practice, this has led to a collection of

national statistics on employment in the infor-

mal sector, with countries reporting on their

preferred variation of the criteria laid out in the

resolution of the 15th ICLS. Some countries

apply the criterion of non-registered enterprises

but registration requirements can vary from

country to country. Others apply the employ-

ment size criterion only (which may vary from

country to country) and other countries still

apply a combination of the two. As a result of the

national differences in definitions and coverage,

the international comparability of the employ-

ment in the informal sector indicator is limited.

In summary, problems with data compar-

ability for the measure of employment in the

informal sector result in particular from the

following factors:

• differences in data sources;

• differences in geographic coverage;

• differences in the branches of economic activ-

ity covered. At one extreme are countries that

cover all kinds of economic activity, including

agriculture, while at the other are countries

that cover only manufacturing;

The ILO Department of Statistics has played

a leading role in developing methods for the

collection of data on the informal sector, in

compiling and publishing official statistics in

this area, and in providing technical assistance

to national statistical offices to improve their

data collection. In 1998, the department estab-

lished a database on the informal sector, which

was subsequently used as the basis for some of

the more dated statistics in table 8. The data set

was updated in 2001, along with a compen-

dium of available official national statistics and

related methodological information, and again

in 2012. As of 2014, the Department has been

including the topic of informality in its regular

annual data compilation effort in order to

provide regular updates for the ILO’s online

database, ILOSTAT. These data, supplemented

with some additional data from the ILO

Regional Office for Latin America and the

Caribbean, served as the repositories used for

the production of table 8.

Limitations to comparability

The concept of the informal sector was

consciously kept flexible in order to accommo-

Box 8b. Timeline of informality as a statistical concept

1993: Definition of informal sector adopted by the 15th ICLS.

1999: Third meeting of the Expert Group on Informal Sector Statistics (Delhi Group), where it was

concluded that the group should formulate recommendations regarding the identification of

precarious forms of employment (including outwork/home-work) inside and outside the

informal sector.

2001: Fifth meeting of the Delhi Group, where it was concluded that the definition and measurement

of employment in the informal sector needed to be complemented with a definition and

measurement of informal employment and that Group members should test the conceptual

framework developed by the ILO.

2002: 90th Session of International Labour Conference (ILC), where the need for more and better

statistics on the informal economy was emphasized and the ILO was tasked to assist countries

with the collection, analysis and dissemination of statistics. The ILC also proposed a definition

for the informal economy.

2002: Sixth meeting of the Delhi Group, recognized the need for consolidating the country

experiences and recommended further research for developing a statistical definition of

informal employment and methods of compiling informal employment statistics through

labour force surveys.

2003: Guidelines on a definition of informal employment as an international statistical standard

adopted by th 17th ICLS.

2013: A manual on measuring informality on methodological issues for undertaking surveys of the

informal economy at the country level published by the ILO

ILO: Measuring informality: a Statistical Manual on the informal sector and informal employment (ILO, Geneva, 2013), available

at: http://www.ilo.ch/global/publications/ilo-bookstore/order-online/books/WCMS_222979/lang--en/index.htm..

KILM 8 Employment in the informal economy

limit the measurement of informal employment

to employee jobs only. The built-in flexibility of

the statistical concept, while certainly a

commendable and necessary feature for a new

concept, does create limitations when it comes

to the comparability of statistics across coun-

tries. More comparability will only be achieved

in the long run when good practices have

driven out less good ones.

In order to reduce comparability issues and

to improve the availability and quality of data,

the ILO, in collaboration with members of the

Delhi Group, has published the manual

Measuring informality: A statistical manual

on the informal sector and informal employ-

ment.

This manual pursues two objectives:

(1) to assist countries that are planning a

programme to produce statistics on the infor-

mal sector and informal employment, in under-

taking a review and analysis of their options;

and (2) to provide practical guidance on the

technical issues involved with the development

and administration of the surveys used to

collect relevant information, as well as on the

compilation, tabulation and dissemination of

the resulting statistics.

ILO: Measuring informality: A statistical manual on

the informal sector and informal employment (Geneva,

2013), available at: http://www.ilo.ch/global/publications/

ilo-bookstore/order-online/books/WCMS_222979/lang--en/

index.htm.

• differences in the criteria used to define the

informal sector, for example, size of the enter-

prise or establishment versus non-registration

of the enterprise or the worker;

• different cut-offs used for enterprise size;

•

inclusion or exclusion of paid domestic workers;

• inclusion or exclusion of persons who have a

secondary job in the informal sector but

whose main job is outside the informal sector,

e.g. in agriculture or in public service.

As with the concept of the informal sector,

the concept of informal employment was

designed in such a way as to allow countries to

accommodate their own situations and needs.

The 17th ICLS guidelines specifically say that

“the operational criteria for defining informal

jobs of employees are to be determined in

accordance with national circumstances and

data availability”. Some countries (especially

developing countries) may choose to develop a

measure that includes informal jobs of own-

account workers, employers and members of

producers’ cooperatives, while other countries

(especially developed countries) may wish to

They can contribute to the better understand-

ing of variations in unemployment that are the

result of variations in the rate at which workers

move from employment to unemployment and

vice versa. Unemployment flows are available

for 67 economies.

Use of the indicator

The overall unemployment rate for a country

is a widely used measure of its unutilized labour

supply. If employment is taken as the desired

situation for people in the labour force (formerly

known as economically active population),

unemployment becomes the undesirable situ-

ation. Still, some short-term unemployment can

be necessary for ensuring adjustment to

economic fluctuations. Unemployment rates by

specific groups, defined by age, sex, occupation

or industry, are also useful in identifying groups

of workers and sectors most vulnerable to

joblessness.

While it may be considered the most infor-

mative labour market indicator reflecting the

general performance of the labour market and

the economy as a whole, the unemployment rate

should not be interpreted as a measure of

economic hardship or of well-being. When

based on the internationally recommended stan-

dards (outlined in more detail under “defini-

tions and sources” below), it simply reflects the

proportion of the labour force that does not

have a job but is available and actively looking

for work. It says nothing about the economic

resources of unemployed workers or their family

members. Its use should, therefore, be limited

to serving as a measurement of the utilization of

labour and an indication of the failure to find

work. Other measures, including income-related

indicators, would be needed to evaluate

economic hardship.

An additional criticism of the aggregate

unemployment measure is that it masks infor-

mation on the composition of the jobless popu-

lation and therefore misses out on the

particularities of the education level, ethnic

origin, socio-economic background, work

experience, etc. of the unemployed. Moreover,

the unemployment rate says nothing about the

Introduction

The unemployment rate is probably the

best-known labour market measure and

certainly one of the most widely quoted by

media in many countries as it is believed to

reflect the lack of employment at national levels

to the greatest and most meaningful extent.

Together with the employment-to-population

ratio (KILM 2), it provides the broadest indica-

tor of the labour market situation in countries

that collect information on the labour force.

The KILM 9th edition complements national

estimates with ILO estimates of unemployment

rates. To supplement the stock indicator of

unemployment with a more dynamic view of

the labour market, KILM 9 also includes an indi-

cator on unemployment flows, namely inflows

into and outflows from unemployment.

National estimates of unemployment rates

are available for a total of 215 economies in

table 9b. Information on the number of un-

employed persons is available for additional

countries in both tables, but the lack of statistics

on the labour force, the necessary denominator,

prevents the calculation of the unemployment

rate for them. ILO estimates of unemployment

rates (table 9a) are harmonized to account for

differences in national data and scope of cover-

age, collection and tabulation methodologies as

well as for other country-specific factors such as

differing national definitions.

Table 9a is based

on available national estimates of unemploy-

ment rates and includes these reported rates as

well as imputed data for 177 economies. ILO

estimates of unemployment rates are national

data, meaning there are no geographical limita-

tions in coverage. This series of harmonized

estimates serves as the basis of the ILO’s global

and regional aggregates of the unemployment

rates reported in the World Employment and

Social Outlook series and made available in the

KILM 9th edition as table R5. Estimates of un-

employment flows (table 9c) are calculated on

the basis of data on unemployment by duration.

For further information on the methodology used to

harmonize estimates, see Annex 4, “Note on global and

regional estimates”, in ILO: Global employment trends

2011 (Geneva, 2011); http://www.ilo.org/global/publica-

tions/books/WCMS_150440/lang--en/index.htm.

KILM 9. Unemployment

KILM 9 Unemployment

occurrence of intense inflationary pressures.

Because of this supposed trade-off, the un-

employment rate is closely tracked over time.

The usual policy goal of governments,

employers and trade unions is to have a rate

that is as low as possible yet also consistent

with other economic and social policy object-

ives, such as low inflation and a sustainable

balance-of-payments situation. When using the

unemployment rate as a gauge for tracking

cyclical developments, we are interested in

looking at changes in the measure over time. In

that context, the precise definition of unem-

ployment used (whether a country-specific

definition or one based on the internationally

recommended standards) does not matter

nearly as much – so long as it remains

unchanged – as the fact that the statistics are

collected and disseminated with regularity, so

that measures of change are available for study.

Internationally, the unemployment rate is

frequently used to compare how labour markets

in specific countries differ from one another or

how different regions of the world contrast in

this regard. Unemployment rates may also be

used to address issues of gender differences in

labour force behaviour and outcomes. The

unemployment rate has often been higher for

women than for men. Possible explanations are

numerous but difficult to quantify; women are

more likely than men to exit and re-enter the

labour force for family-related reasons; and

there is a general “crowding” of women into

fewer occupations than men so that women

may find fewer opportunities for employment.

Other gender inequalities outside the labour

market, for example in access to education and

training, also negatively affect how women fare

in finding jobs.

The indicator on unemployment flows

(table 9c) included in the KILM 9th edition

provides estimates of the inflows into and

outflows from unemployment in order to

uncover the adjustment dynamics of unem-

ployment that are underlying net changes in

unemployment rates. The data series intend to

improve the understanding of varying un-

employment rates across time and countries.

Unemployment rates alone often do not

reveal the full picture of the labour market situ-

ation in an economy as they do not say much

about the driving forces behind variations in

unemployment. In particular, changes in un-

employment rates result from the net effect of

flows into unemployment and flows out

of unemployment. Both flow margins can be

type of unemployment – whether it is cyclical

and short-term or structural and long-term –

which is a critical issue for policy-makers in the

development of their policy responses, espe-

cially given that structural unemployment

cannot be addressed by boosting market

demand only.

Paradoxically, low unemployment rates may

well disguise substantial poverty,

as high

unemployment rates can occur in countries

with significant economic development and

low incidence of poverty. In countries without

a safety net of unemployment insurance and

welfare benefits, many individuals, despite

strong family solidarity, simply cannot afford to

be unemployed. Instead, they must eke out a

living as best they can, often in the informal

economy or in informal work arrangements. In

countries with well-developed social protec-

tion schemes or when savings or other means

of support are available, workers can better

afford to take the time to find more desirable

jobs. Therefore, the problem in many develop-

ing countries is not so much unemployment

but rather the lack of decent and productive

work, which results in various forms of labour

underutilization (i.e. underemployment, low

income, and low productivity).

A useful purpose served by the unemploy-

ment rate in a country, when available on at

least an annual basis, is the tracking of business

cycles. When the rate is high, the country may

be in recession (or worse), economic condi-

tions may be bad, or the country somehow

unable to provide jobs for the available work-

ers. The goal, then, is to introduce policies and

measures to bring the incidence of unemploy-

ment down to a more acceptable level. What

that level is, or should be, has often been the

source of considerable discussion, as many

consider that there is a point below which an

unemployment rate cannot fall without the

Information relating to poverty, working poverty and

inequality is provided in the chapter on KILM 17.

Readers interested in the broader topic of labour

underutilization should refer to ILO, Beyond unemploy-

ment: Measurement of other forms of labour underutiliza-

tion, Room Document 13, 18th International Conference of

Labour Statisticians, Working group on Labour underutiliza-

tion, Geneva, 24 November – 5 December 2008; http://www.

ilo.org/global/statistics-and-databases/meetings-and-events/

international-conference-of-labour-statisticians/

WCMS_100652/lang--en/index.htm or ILO, “Report and

proposed resolution of the committee on work statistics”,

19th International Conference of Labour Statisticians,

Committee on Work Statistics, Geneva, 2 November –

11 November 2013; http://www.ilo.org/global/statistics-and-

databases/meetings-and-events/international-conference-of-

labour-statisticians/19/WCMS_223719/lang--en/index.htm.

KILM 9 Unemployment

“labour force” and “employment” are some-

times mistakenly used interchangeably.

According to the Resolution concerning

statistics of work, employment and labour

underutilization adopted in 2013 by the 19th

International Conference of Labour Statisticians

(ICLS), the standard definition of unemploy-

ment refers to all those persons of working age

who are without work, seeking work (carried

out activities to seek employment during a

recent past period), and currently available for

work

(see box 9). Future starters, that is,

persons who did not look for work but have a

future labour market stake (made arrange-

ments for a future job start) are also counted as

unemployed.

In many national contexts there may be

persons not currently in the labour market who

want to work but do not actively “seek” work

because they view job opportunities as limited,

or because they have restricted labour mobility,

or face discrimination or structural, social or

cultural barriers. The exclusion of people who

want to work but are not seeking work (in the

past often called the “hidden unemployed”,

which also included persons formerly known

as “discouraged workers”) is a criterion that

will affect the count of both women and men,

although women may have a higher probability

of being excluded from the count of un-

employed because they suffer more from social

barriers overall that impede them from meeting

this criterion.

Another factor leading to exclusion from the

unemployment count concerns the criterion

that workers be available for work during the

short reference period. A short availability

period tends to exclude those who would need

to make personal arrangements before starting

work, such as for care of children or elderly rela-

tives or other household affairs, even if they are

“available for work” soon after the short refer-

ence period. As women are often responsible for

household affairs and care, they are a significant

part of this group and would therefore not be

included in measured unemployment. Various

countries have acknowledged this coverage

problem and have extended the “availability”

period to the two (or more) weeks following the

reference period. Even then, women – more

Resolution concerning statistics of work, employ-

ment and labour underutilization, 19th International

Conference of Labour Statisticians, Geneva, 2013; available

at: http://www.ilo.org/global/statistics-and-databases/stan-

dards-and-guidelines/resolutions-adopted-by-international-

conferences-of-labour-statisticians/WCMS_230304/lang--en/

index.htm.

affected by different factors that may vary over

the course of the business cycle or follow longer

term trends. In order to allow for a more

detailed analysis of these dynamics, the inflows

into and outflows from unemployment are

constructed in an attempt to shift from a simple

stock approach to an understanding of the vari-

ation in unemployment as the variation in the

rate at which workers move from one labour

market state to another. More specifically, the

flow approach illustrates how quickly workers

transition from employment into unemploy-

ment (inflow) and unemployment into employ-

ment (outflow). The estimated inflow and

outflow rates that are shown in table 9c are

directly related to the probability that an

employed worker becomes unemployed

(inflow) and the probability that an unemployed

worker finds a job (outflow). These measures

provide an essential tool to target labour market

policies more specifically at certain groups of

the labour market or to adjust them according

to which aspect of the unemployment dynamics

dominate in a particular situation.

It can be very insightful to track the behav-

iour of inflows and outflows during economic

up- and downturns, or to use flow measures in

forecasting the unemployment rate. Moreover,

an analysis of unemployment flows together

with other labour market indicators can be

useful to better understand labour market

distress and develop policy recommendations.

For a deeper understanding of fluctuations in

unemployment, it is essential to understand

fluctuations in the transition rate from employ-

ment to unemployment and vice versa.

Definitions and sources

The unemployment rate is defined math-

ematically as the ratio resulting from dividing

the total number of unemployed (for a country

or a specific group of workers) by the corre-

sponding labour force, which itself is the sum

of the total persons employed and unemployed

in the group. It should be emphasized that it is

the labour force (formerly known as the

economically active population) that serves as

the base for this statistic, not the total popula-

tion. This distinction is not necessarily well

understood by the public. Indeed, the terms

For a detailed analysis of factors influencing labour

flows and their consequences for unemployment dynamics

see, e.g., Ernst, E.; Rani, U.: “Understanding unemploy-

ment flows”, in Oxford Review of Economic Policy Making,

Vol. 27, No. 2 (Oxford, 2011), pp. 268–294.

KILM 9 Unemployment

On the other hand, administrative records

can overstate registered unemployment

because of double-counting, failure to remove

people from the registers when they are no

longer looking for a job, or because it allows

inclusion of persons who have done some

work. Due to these measurement limitations,

national unemployment data based on the

registered unemployed should be treated with

care; registered unemployment data can serve

as a useful proxy for the extent of persons with-

out work in countries where data on total

unemployment are not available. Time-series of

registered unemployment data by country can

serve as a good indication of labour market

performance over time, but due to the limita-

tion in comparability with “total unemploy-

ment”, the two measures should not be used

interchangeably. Tables 9a and 9b provide data

on total unemployment only.

The indicator on unemployment flows

(table 9c) is calculated based on data on un-

employment by duration (see KILM 11 on long-

term unemployment) and quarterly unemploy-

ment rates. In cases in which quarterly

unemployment rates are not available, annual

unemployment rates are used (as indicated in

the column “use of annual data for unemploy-

ment rates” of table 9c). Data to compute unem-

ployment flows (unemployment by duration,

unemployment rates and labour force) are taken

from labour force or household surveys.

Unemployment flows in table 9c are

displayed as alternative estimates of monthly

transition rates from employment to un-

employment (inflow) and vice versa (outflow).

The different estimates of inflow and outflow

rates are calculated on the basis of data on

unemployment spells of different durations,

namely less than one month, less than three

months, less than six months, and less than

12 months. The estimates on the basis of un-

employment with a duration of less than one

month follow the methodology applied by

Shimer (2012).

The table also contains weighted flow rate

estimates. Weights are calculated on the basis

of the methodology proposed by Elsby et al.

(2013) who use these weighted flow rate esti-

mates in case of no evidence for duration

dependence and flow rates calculated on the

basis of unemployment with a duration of less

For more details, see Shimer, R.: “Reassessing the ins

and outs of unemployment”, in Review of Economic

Dynamics, Vol. 15, No. 2, pp. 127–148 (2012).

than men – tend to be excluded from unemploy-

ment, probably because this period is still not

sufficiently long to compensate for constraints

that are more likely to affect them.

With a view to overcoming these limitations

of the concept of unemployment, and in order

to acknowledge the two population groups

mentioned above (persons without work but

either not available or not actively seeking

work), the 19th ICLS resolution introduced the

concept of the “potential labour force”. This

potential labour force comprises “unavailable

jobseekers”, defined as persons who sought

employment even though they were not avail-

able, but would become available in the near

future, and “available potential jobseekers”,

defined as persons who did not seek employ-

ment but wanted it and were available. Thus,

persons without work formerly included in the

“relaxed definition” of unemployment are now

included in the potential labour force. The 19th

ICLS resolution also identifies a particular

group within the available potential jobseekers,

the “discouraged jobseekers”, made up of those

persons available for work but who did not

seek employment for labour market-related

reasons (such as past failure to find a suitable

job or lack of experience).

Household labour force surveys are gener-

ally the most comprehensive and comparable

sources for unemployment statistics. Other

possible sources include population censuses

and official estimates. Administrative records

such as employment office records and social

insurance statistics are also sources of unem-

ployment statistics; however, coverage in such

sources is limited to “registered unemployed”

only. A national count of either unemployed

persons or work applicants who are registered

at employment offices is likely to be only a

limited subset of the total number of persons

seeking and available for work, especially in

countries where the system of employment

offices is not extensive. This may be because of

eligibility requirements that exclude those who

have never worked or have not worked recently,

or to other discriminatory impediments that

preclude going to register.

For more details on the potential labour force and

the changes to the definition in unemployment, please

refer to ILO: “Report III - Report of the Conference”, 19th

International Conference of Labour Statisticians, Geneva,

2−11 October 2013; available at: http://www.ilo.org/global/

statistics-and-databases/meetings-and-events/international-

conference-of-labour-statisticians/19/WCMS_234124/lang-

-en/index.htm.

KILM 9 Unemployment

1. Different sources. To the extent that sources

of information differ, so will the results.

Comparability difficulties resulting from the

differences between sources measuring regis-

tered unemployment and total unemployment

have been removed by separating the two and

only including total unemployment. The

remaining sources in KILM 9 – labour force

surveys, official estimates and population

censuses – can still pose issues of comparabil-

ity in cross-country analyses. Official estimates

are generally based on information from differ-

ent sources and can be combined in many

different ways. A population census generally

cannot probe deeply into labour force activity

status. The resulting unemployment estimates

may, therefore, differ substantially (either

upwards or downwards) from those obtained

from household surveys where more ques-

tions are asked to determine respondents’

labour market situation. For more information

regarding sources, users may also refer to the

discussion of the pros and cons of various

sources in the corresponding section of

KILM 1 (labour force participation rates).

2. Measurement differences. Where the in-

formation is based on household surveys or

population censuses, differences in the ques-

tionnaires can lead to different statistics –

even allowing for full adherence to ILO guide-

lines. In other words, differences in the

measurement tool can affect the comparabil-

ity of labour force results across countries.

3. Conceptual variation. National statistical

offices, even when basing themselves on the

ILO conceptual guidelines, may not follow the

strictest measurements of employment and

unemployment. They may differ in their

choices concerning the conceptual basis for

estimating unemployment, as in specific

instances where the guidelines prior to the

19th ICLS resolution used to allow for a

relaxed definition, thereby causing the labour

force estimates (the base for the unemploy-

ment rate) to differ. They may also choose to

derive the unemployment rate from the civil-

ian labour force rather than the total labour

force. To the extent to which such choices

vary across countries, so too will the statistics

displayed in KILM 9.

4. Number of observations per year (refer-

ence period). Statistics for any given year

can differ depending on the number of obser-

vations – monthly, quarterly, once or twice a

year, and so on. Among other things, a consid-

erable degree of seasonality can influence the

results when the full year is not covered.

than one month otherwise.

The column

“result of test for no duration dependence”

indicates the result of the test to determine

which flow rate estimate to choose, following

Elsby et al. (2013).

For countries for which the hypothesis of

“no duration dependence” is rejected, Elsby et

al. (2013) follow the approach of Shimer (2012)

and use flow estimates that are calculated on

the basis of unemployment with a duration of

less than one month. For countries for which

the above hypothesis is not rejected, the

weighted estimate is preferred.

Limitations to comparability

A significant amount of research has been

carried out over the years in producing un-

employment rates that are fully consistent concep-

tually, in order to contrast unemployment rates

of different countries for different hypotheses.

Interested users can compare the series of

“ILO-comparable” unemployment estimates

with the information shown in table 9b. In a few

cases the adjusted rates are the same as those

found in table 9b; elsewhere they are quite differ-

ent as the information in table 9b may be obtained

from multiple sources, while the adjusted

“ILO-comparable” rates are always based on a

household labour force sample survey.

There are a host of reasons why measured

unemployment rates may not be comparable

between countries. A few are provided below,

to give users some indication of the range of

potential issues that are relevant when attempt-

ing to determine the degree of comparability of

unemployment rates between countries. Users

with knowledge of particular countries or

special circumstances should be able to expand

on them:

See Elsby, M.; Hobijn, B.; Sahin, A.: “Unemployment

dymamics in the OECD”, in Review of Economics and

Statistics, Vol. 95, No. 2, pp. 530–548 (2013).

The “ILO-comparable” unemployment rates are

national labour force survey estimates that have been

adjusted to make them conceptually consistent with the

strictest application of the ILO statistical standards. The

unemployment rates obtained are based on the total labour

force including the armed forces. For more information

regarding the methodology, see Lepper, F.: Comparable

annual employment and unemployment estimates, Depart-

ment of Statistics Paper, ILO (Geneva, 2004); available at:

http://www.ilo.org/global/statistics-and-databases/

WCMS_087893/lang--en/index.htm. The estimates (up to

2005 only) are available at: http://laborsta.ilo.org.

KILM 9 Unemployment

Geographical coverage. Survey coverage

that is less than national coverage – urban areas,

city, regional – has obvious limitations in

comparability to the extent that coverage is not

representative of the country as a whole.

Unem

ployment in urban areas may tend to be

higher than total unemployment because of

the exclusion of the rural areas where workers

are likely to work, although they may be under-

employed or unpaid family workers, rather

than seeking work in a non-existent or small

formal sector.

6. Age variation. The generally used age cover-

age is 15 years and over, but some countries

use a different lower limit or impose an upper

age limit.

7. Collection methodology. Sample sizes,

sample selection procedures, sampling frames,

and coverage, as well as many other statistical

issues associated with data collection, may

make a significant difference. The better the

sample size and coverage, the better the

results. Use of well-trained interviewers,

proper collection and processing techniques,

adequate estimation procedures, etc. are

crucial for accurate results. Wide variations in

this regard can clearly affect the comparabil-

ity of the unemployment statistics.

When viewing the unemployment rate as a

gauge for tracking cyclical developments within

a country, one would be interested in looking

at changes in the measure over time. In this

context, the definition of unemployment used

(whether a country-specific definition or one

based on the internationally recommended

standards) does not matter as much – so long

as it remains unchanged – as the fact that the

statistics are collected and disseminated with

regularity, so that measures of change are avail-

able for study. Still, for users making cross-

country comparisons it will be critical to know

the source of the data and the conceptual basis

for the estimates. It is also important to recog-

nize that minor differences in the resulting

statistics may not represent significant real

differences.

When performing queries on this table and others,

users have the option to omit records that are of sub-national

geographic coverage. On the software, this can be done by

performing the query for all data and then refining the

parameters to select “National only” under “Geographical

coverage”.

Two examples of substantial difference in

household surveys may be useful for under-

standing some of the complexities of optimal

comparisons. The first concerns “job search”.

The ILO conceptual framework assumes that

persons looking for work must indicate one or

more “active” methods – such as applying

directly to employers or visiting an employ-

ment exchange office – in order to be counted

as unemployed. Among the potential methods

is the consultation of “newspaper advertise-

ments”. In many parts of the world, this may

not be a common or readily available means. In

others, newspapers are an excellent source of

information about potential jobs, and many

jobseekers do indeed consult them. However,

some countries accept the mere reading or

looking at advertisements as a search method,

whereas others require that persons actually

answer one or more advertisement before the

newspaper search is counted as an acceptable

method. The issue comes down to whether the

“passive” versus the “active” search is allowed,

and countries vary in their approach to this.

The second example relates to “discouraged

jobseekers”: persons who are not currently

looking for work but may have looked in the

past and clearly desire a job “now” (see “defin-

itions and sources” above). Most surveys do not

include them among the unemployed (as indi-

cated by the ILO definition of unemployment),

but some do. Users wishing to account for such

a definitional difference would need to obtain

relevant information (perhaps at the “micro”

level) in order to adjust for differences in

unemployment rates.

The above two examples illustrate aspects of

conceptual variation and measurement differ-

ence. The degree of complexity of these and

other differences in the measurement and esti-

mation of unemployment that can occur

around the world serve as a reminder that great

care should be taken in any attempt to draw

exacting comparisons.

The proposed resolution of the committee on work

statistics extended the “activities to seek employment” and

includes examples of such activities. For more details, see

ILO: Report and proposed resolution of the committee on

work statistics,19th International Conference of Labour

Statisticians, Committee on Work Statistics, Geneva,

2–11 November 2013; available at: http://www.ilo.org/

global/statistics-and-databases/meetings-and-events/inter-

national-conference-of-labour-statisticians/19/

WCMS_223719/lang--en/index.htm.

KILM 9 Unemployment

Box 9. Resolution concerning statistics of work, employment

and labour underutilization, adopted by the 19th International

Conference of Labour Statisticians, October 2013

[relevant paragraphs]

Concepts and definitions

…

Unemployment (paras 47-48)

47. Persons in unemployment are defined as all those of working age who were not in employment,

carried out activities to seek employment during a specified recent period and were currently

available to take up employment given a job opportunity, where:

a. “not in employment” is assessed with respect to the short reference period for the measure-

ment of employment;

b. to “seek employment” refers to any activity when carried out, during a specified recent period

comprising the last four weeks or one month, for the purpose of finding a job or setting up a

business or agricultural undertaking. This includes also part-time, informal, temporary,

seasonal or casual employment, within the national territory or abroad. Examples of such

activities are

i. arranging for financial resources, applying for permits, licences;

ii. looking for land, premises, machinery, supplies, farming inputs;

iii. seeking the assistance of friends, relatives or other types of intermediaries;

iv. registering with or contacting public or private employment services;

v. applying to employers directly, checking at worksites, farms, factory gates, markets or

other assembly places;

vi. placing or answering newspaper or online job advertisements;

vii. placing or updating résumés on professional or social networking sites online;

c. the point when the enterprise starts to exist should be used to distinguish between search

activities aimed at setting up a business and the work activity itself, as evidenced by the

enterprise’s registration to operate or by when financial resources become available, the

necessary infrastructure or materials are in place or the first client or order is received, depend-

ing on the context;

d. “currently available” serves as a test of readiness to start a job in the present, assessed with

respect to a short reference period comprising that used to measure employment:

i. depending on national circumstances, the reference period may be extended to include a

short subsequent period not exceeding two weeks in total, so as to ensure adequate

coverage of unemployment situations among different population groups.

48. Included in unemployment are:

a. future starters defined as persons “not in employment” and “currently available” who did

not ”seek employment”, as specified above, because they had already made arrange-

ments to start a job within a short subsequent period, set according to the general length

of waiting time for starting a new job in the national context but generally not greater than

three months;

b. participants in skills training or retraining schemes within employment promotion

programmes, who on that basis, were “not in employment”, not “currently available” and

did not “seek employment” because they had a job offer to start within a short subsequent

period generally not greater than three months;

c. persons “not in employment” who carried out activities to migrate abroad in order to work

for pay or profit but who were still waiting for the opportunity to leave.

Use of the indicator

Young men and women today face increasing

uncertainty in their hopes of undergoing a satis-

factory entry to the labour market, and this

uncertainty and disillusionment can, in turn,

have damaging effects on individuals, communi-

ties, economies and society at large. Unemployed

or underemployed youth are less able to contrib-

ute effectively to national development and have

fewer opportunities to exercise their rights as

citizens. They have less to spend as consumers,

less to invest as savers and often have no “voice”

to bring about change in their lives and commu-

nities. In certain case

s, this results in social

unrest and a rejection of the existing socio-

economic system by young people. Widespread

youth unemployment and underemployment

also prevents companies and countries from

innovating and developing competitive advan-

tages based on human capital investment, thus

undermining future prospects.

Knowing the costs of non-action, many

governments around the world prioritize the

issue of youth unemployment and attempt to

develop appropriate policies and programmes.

Measuring the impact of such policies requires

age-disaggregated indicators, such as those

provided in KILM 10. The KILM youth indica-

tors also constitute the basis for the ILO’s

Global Employment Trends for Youth, which

serves as a key product for quantifying and

analysing the current labour market trends and

challenges of young people.

While KILM 10 is the only one of the 17 KILM

indicators relating specifically to youth, age-

disaggregation has been included for numerous

other indicators in the KILM. Thus, KILM users

can access and analyse data for youth (in

comparison to the adult and total populations)

for labour force participation rates (tables 1a

See, for example, the inventory of crisis-response

programmes and policies put into place by countries in

ILO: Global Employment Trends for Youth: Special issue on

the impact of the global economic crisis on youth (Geneva,

2010); http://www.ilo.org/empelm/pubs/WCMS_143349/

lang--en/index.htm.

All ILO Global Employment Trends for Youth publica-

tions are available at: http://www.ilo.org/trends.

Introduction

Youth unemployment is widely viewed as

an important policy issue for many countries,

regardless of their stage of development. For

the purpose of this indicator, the term “youth”

covers persons aged 15 to 24 years and “adult”

refers to persons aged 25 years and over.

KILM 10 consists of three tables. Tables 10a

and 10b contain ILO estimates and national

estimates, respectively, of four distinct

measurements of aspects of the youth unem-

ployment problem. The four measurements

are: (a) youth unemployment rate (youth

unemployment as a percentage of the youth

labour force); (b) ratio of the youth un-

employment rate to the adult unemployment

rate; (c) youth unemployment as a proportion

of total unemployment; and (d) youth un-

employment as a proportion of the youth

population. Table 10c presents estimates of

the proportion of young people not in

employment, education or training, the

“NEET” rate. The information in table 10c

follows the standard definition of youth, that

is, it refers to persons aged 15 to 24, unless

otherwise indicated. The information in all

three tables is disaggregated by sex.

ILO estimates of youth unemployment in

table 10a are harmonized to account for

differences in scope of coverage, collection

and tabulation methodologies as well as for

other country-specific factors such as military

service requirements. This table includes

both nationally reported and imputed data

and includes only estimates that are national,

meaning there are no geographic limitations

in coverage. It is this series of harmonized

estimates that serve as the basis of the ILO’s

global and regional aggregates of the labour

force participation rate as reported in the

Global Employment Trends series and made

available in the KILM 9th edition software as

table R6.

The youth unemployment rates and related

measurements are available in table 10a for

178 economies, and in table 10b for 196 econ-

omies. NEET rates in table 10c are available

for 119 economies.

KILM 10. Youth unemployment

KILM 10 Youth unemployment

jobseekers, young people will suffer most from

economically induced reductions or freezes in

hiring by establishments.

Information on the other two aspects of the

youth unemployment problem captured by KILM

10, namely the share of unemployed youth in

total unemployment and the proportion of

unemployed youth in the youth population,

helps to complete a portrait of the depth of the

youth employment challenge. The former

complements the ratio of youth-to-adult un-

employment rate in reflecting to what degree the

unemployment problem is a youth-specific prob-

lem as opposed to a general problem. If, in addi-

tion to a high youth unemployment rate, the

proportion of youth unemployment in total

unemployment is high, this would indicate an

unequal distribution of the problem of un-

employment. In this case, employment policies

might usefully be directed towards easing the

entry of young people into the world of work.

The proportion of youth unemployed in the

youth population places the youth unemploy-

ment challenge in perspective by showing what

share of the youth population unemployment

actually touches. Youth who are looking for work

might have great difficulty finding it but when

this group only represents less than 5 per cent of

the total youth population then policy-makers

may choose to address it with less urgency.

The proportion of youth unemployed in the

youth population is also an element in the total

proportion of youth not in employment, educa-

tion or training. The NEET rate is a broad

measure of the untapped potential of youth

who could contribute to national development

through work. Because the NEET group is

neither improving its future employability

through investment in skills nor gaining experi-

ence through employment, this group is partic-

ularly at risk of both labour market and social

exclusion.

In addition, the NEET group is

already in a disadvantaged position due to

lower levels of education and lower household

incomes.

In view of the fact that the NEET

group includes unemployed youth as well as

economically inactive youth, the NEET rate

provides important complementary informa-

tion to labour force participation rates and

unemployment rates. For example, if youth

participation rates decrease during an economic

downturn due to discouragement, this may be

reflected in an upward movement in the NEET

Note that youth in education and youth in employ-

ment are not mutually exclusive groups.

Eurofound (European Foundation for the Improve-

ment of Living and Working Conditions): Young people and

NEETs in Europe: First findings (résumé) (Dublin, 2011).

and 1b), employment-to-population ratios

(tables 2a and 2b), part-time employment

(table 6), employment by hours worked per

week (table 7a), long-term unemployment

(table 11), time-related underemployment

(table 12), inactivity rates (table 13), labour

force by level of educational attainment

(table 14a), unemployment by level of educa-

tional attainment (table 14b), illiteracy

(table 14d), and working poverty (table 17b).

The KILM information on youth unemploy-

ment illustrates the different dimensions of the

lack of jobs for young people. In general, the

higher the four rates presented in tables 10a and

10b, the worse the employment situation of

young people. These measurements are likely to

move in the same direction, and should be

looked at in tandem, as well as together with

other indicators now available in the KILM for

the youth cohort, in order to assess fully the situ-

ation of young people within the labour market

and guide appropriate policy initiatives.

In a country where the youth unemployment

rate is high and the ratio of the youth unemploy-

ment rate to the adult unemployment rate is

close to one, it may be concluded that the prob-

lem of unemployment is not specific to youth,

but is country-wide. However, unemployment

rates of youth are typically higher than those of

adults, reflected by a ratio of youth-to-adult

unemployment rates that exceeds one. There

are various reasons why youth unemployment

rates are often higher than adult unemployment

rates and not all of them are negative. On the

supply side, young persons might voluntarily

engage in multiple short spells of unemploy-

ment as they gain experience and “shop around”

for an appropriate job. Moreover, because of the

opening and closing of educational institutions

over the course of the year, young students are

far more likely to enter and exit the labour force

as they move between employment, school

enrolment and unemployment.

However, high youth unemployment rates

are also the consequences of a labour market

biased against young people. For example,

employers tend to lay off young workers first

because the cost to establishments of releasing

young people is generally perceived as lower

than for older workers. Also, employment

protection legislation usually requires a mini-

mum period of employment before it applies,

and compensation for redundancy usually

increases with tenure. Young people are likely

to have shorter job tenures than older workers

and will, therefore, tend to be easier and less

expensive to dismiss. Finally, since they

comprise a disproportionate share of new

KILM 10 Youth unemployment

for youth aged 15 to 24 years unless otherwise

indicated.

As in KILM table 9, information on unem-

ployment is commonly obtained from one of

three sources: household surveys of the labour

force, official estimates and population censuses.

In tables 10b and 10c the most commonly used

source is the labour force survey.

Limitations to comparability

There are numerous limitations to the

comparability of KILM 10 data across countries

and over time; some are more significant than

others.

One major limitation to comparability

relates to the source used in deriving unem-

ployment rates. The main difficulty with using

population censuses as the source is that, owing

to their cost, they are not undertaken frequently

and the information on unemployment is

unlikely to be up to date. In addition, sources

other than labour force surveys often do not

include probing questions related to employ-

ment and therefore may not produce a compa-

rable estimate of employment across different

groups of workers. On occasion, unemploy-

ment information is based on official estimates.

Again, these are unlikely to be comparable and

are typically based on a combination of admin-

istrative records and other sources. In any

event, users should be aware of the primary

source and take this into account when compar-

ing data across time or across countries.

An additional point should be made regard-

ing the definition of unemployment. For some

countries, the unemployment figures exclude

those who have not been previously employed

(i.e. excluding first time jobseekers). In those

cases, this is indicated clearly in the notes. This

definition will tend to lower the level of

reported youth unemployment.

Although less important than other factors,

differences in the age groups utilized should

also be mentioned as the age limits applied for

both youth and adults may vary across coun-

tries. In general, where a minimum school-leav-

ing age exists, the lower age limit of youth will

usually correspond to that age. This means that

the lower age limit often varies between 10 and

16 years, according to the institutional arrange-

ments in the country. This should not greatly

For the sake of completeness, users are also advised

to review the corresponding discussion in the chapter on

KILM 9.

rate.

More generally, a high NEET rate and a

low youth unemployment rate may indicate

significant discouragement of young people. A

high NEET rate for young women suggests their

engagement in household chores, and/or the

presence of strong institutional barriers limit-

ing female participation in labour markets.

Definitions and sources

Young people are defined as persons aged 15

to 24; however, countries vary somewhat in their

operational definitions. In particular, the lower

age limit for young people is usually determined

by the minimum age for leaving school, where

this exists. Differences in operational definitions

have implications for comparability, which are

discussed below. The resolution concerning

statistics of work, employment and labour under-

utilization, adopted by the 19th International

Conference of Labour Statisticians (ICLS),

outlines the international standards for (youth)

unemployment. The resolution states that the

unemployed comprise all persons above a speci-

fied age who, during the reference period, were:

(a) without work; (b) currently available for

work; and (c) actively seeking work.

As is the

case for KILM 9, the unemployment rate is

defined as the number of unemployed in an age

group divided by the labour force for that group.

In the case of youth unemployment as a propor-

tion of the young population, the population for

that age group replaces the labour force as the

denominator.

The NEET rate in table 10c is defined as the

number of youth who are not in employment,

education or training as a percentage of the

youth population. The NEET rate is presented

ILO: Global Employment Trends for Youth 2013: A

generation at risk (Geneva, 2013), chapter 2.1; http://www.

ilo.org/moscow/information-resources/publications/

WCMS_345429/lang--en/index.htm.

Resolution concerning statistics of work, employ-

ment and labour underutilization, adopted by the 19th

International Conference of Labour Statisticians, October

2013; http://www.ilo.org/global/statistics-and-databases/

standards-and-guidelines/resolutions-adopted-by-interna-

tional-conferences-of-labour-statisticians/WCMS_230304/

lang--en/index.htm. Readers can find the excerpts pertain-

ing to the definition of unemployment in box 9 in the chap-

ter on KILM 9 and may also wish to review the text in the

“Definitions and sources” section in there.

Youth unemployment as a percentage of the youth

population is sometimes called the youth unemployment

ratio or the youth unemployment-to-population ratio. The

(youth) unemployment-to-population ratio and the (youth)

employment-to-population ratio (KILM 2) add up to the

(youth) labour force participation rate (KILM 1).

100

KILM 10 Youth unemployment

vary significantly over the year as a result of

different school opening and closing dates.

Most of the information reported relates to

annual averages. In other cases, however, the

figures relate to a specific month of the year (as

is the case with census data). The implications

of the particular month chosen will vary across

countries, owing to differences in institutional

arrangements.

As mentioned previously, NEET rates are

available for youth (aged 15 to 24 years), but it

is important to keep in mind, when studying

these rates, that not all persons complete their

education by the age of 24 years. However,

differences in age groups for unemployment

rates and NEET rates may hamper a coherent

analysis of youth employment issues, which is

why information regarding both groups has

been included whenever available.

affect most of the youth unemployment

measures. However, the size of the age group

may influence the measure of the young un-

employed as a percentage of total unemploy-

ment. Other things being equal, the larger the

age group the greater will be this percentage.

In a few cases there is a larger discrepancy in

the lower and upper age limits applied. There

are also differences in the operational definition

of adults. In general, adults are defined as all

individuals aged 25 years and over, but some

countries apply an upper age limit.

Reference periods of the information

reported might also vary across countries.

Because there will be a substantial group of

school-leavers (either permanently or for the

extended holiday break) in the reported figures,

the level of youth unemployment is likely to

The duration of unemployment matters, in

particular in countries where well-developed

social security systems provide alternative

sources of income. In this respect, an increas-

ing proportion of long-term unemployed is

likely to reflect structural problems in the

labour market. During the economic crisis for

example, many economies saw a sharp rise in

the unemployment rate, often as a result of

longer unemployment durations.

Reducing the duration of periods of unem-

ployment is a key element in many strategies to

reduce overall unemployment. Long-duration

unemployment is undesirable, especially in

circumstances where unemployment results

from difficulties in matching supply and

demand because of demand deficiency. The

longer a person is unemployed, the lower his

or her chance of finding a job. Drawing income

support for the period of unemployment

certainly diminishes economic hardship, but

financial support does not last indefinitely. In

any case, unemployment insurance coverage is

often insufficient and not available to every

unemployed person; the most likely non-recip-

ients are persons entering or re-entering the

labour market. Eligibility criteria and the extent

of coverage, as well as the very existence of

insurance, vary widely across countries.

Research has shown that the duration of

unemployment varies with the length of time

that income support can be drawn. This occurs

largely because jobless persons with long-dura-

tion unemployment benefits are able to extend

their periods of joblessness in order to find the

job most consistent with their skills and finan-

cial needs. It might also indicate simply that

unemployment is caused by a long-term defi-

ciency in the supply of jobs. Evidence of the

effect of “generosity” – that is, a high level of

income supplement benefits – on the duration

of unemployment periods is less clear.

Before drawing conclusions about the

effects of features of the benefit system on

unemployment duration, it is necessary to

The International Social Security Association

publishes useful reports that detail social security coverage

by country. See the “Social Security Programs Throughout

the World” series and database at: www.issa.int.

Introduction

The indicators on long-term unemployment

look at duration of unemployment, that is, the

length of time that an unemployed person has

been without work, available for work and look-

ing for a job. KILM 11 consists of two indicators,

one containing long-term unemployment (refer-

ring to people who have been unemployed for

one year or longer); and the other containing

different durations of unemployment.

The first type of indicator, displayed in table

11a, includes two separate measures of long-

term unemployment: (a) the long-term un-

employment rate – persons unemployed for one

year or longer as a percentage of the labour

force; and (b) the incidence of long-term un-

employment – persons unemployed for one

year or longer as a proportion of total unemp-

loyment. Both measures are given for a total of

100 countries, and are disaggregated by sex and

age group (total, youth, adult), where possible.

The second type of indicator, displayed in

table 11b, includes the number of unemployed

(as well as their share in total unemployment) at

different durations: (a) less than one month; (b)

one month to less than three months; (c) three

months to less than six months; (d) six months

to less than twelve months; (e) twelve months or

more. Table 11b is available for 91 economies.

Use of the indicator

While short periods of joblessness are of less

concern, especially when unemployed persons

are covered by unemployment insurance

schemes or similar forms of support, prolonged

periods of unemployment bring with them

many undesirable effects, particularly loss of

income and diminishing employability of the

jobseeker. Moreover, short-term unemploy-

ment may even be viewed as desirable when it

allows time for jobless persons to find optimal

employment; also, when workers can be tempo-

rarily laid off and then called back, short spells

of unemployment allow employers to weather

temporary declines in business activity.

KILM 11. Long-term unemployment

102

KILM 11 Long-term unemployment

over); it is expressed as a percentage of the

overall labour force (long-term unemployment

rate) and of total unemployment (incidence of

long-term unemployment). For more details on

the international definition of unemployment,

users should refer to the corresponding section

in KILM 9.

Data on long-term unemployment are often

collected in household labour force surveys.

Some countries obtain the data from adminis-

trative records, such as those of employment

exchanges or unemployment insurance

schemes. In the latter instances, data are less

likely to be available by sex; moreover, since

many insurance schemes are limited in their

coverage, administrative data are likely to yield

different distributions of unemployment dura-

tion. In addition, the use of administrative data

reduces, and may even totally preclude, the

likelihood that ratios can be calculated using a

statistically consistent labour force base.

Therefore, all the data for this indicator come

from labour force or household surveys, alter-

native sources having been eliminated as likely

to cause inconsistency across the countries for

which data are provided.

Because the data relate to the period of

unemployment experienced by persons who

are still unemployed they necessarily reflect

persons in a “continuing spell of unemploy-

ment”. The duration of unemployment

(table 11b) refers to the duration of the period

during which the person recorded as unem-

ployed was seeking and available for work.

Data on the duration of unemployment are

collected in labour force or household surveys

and the durations consist of a continuous

period of time up to the reference period of the

survey. Table 11b breaks down total unemploy-

ment into different unemployment durations.

For each unemployment duration, data are

expressed in thousands of persons and as a

share of total unemployment.

Statistics on unemployment by duration are

gathered using the databases of the ILO (ILOSTAT);

the Organisation for Economic Co-operation and

Development (OECD); the Statistical Office for the

the European Union (EUROSTAT); and National

Statistical Offices. To facilitate cross-country

comparison, data from OECD and EUROSTAT

were preferred. Unemployment by duration is

broken down by the following durations:

• Unemployment with a duration of less than

one month

• Unemployment with a duration of one month

to less than three months

ana lyse the qualifying and eligibility conditions

as well as the extent of nominal and real income

replacement. Nevertheless, experts and policy-

makers agree that long-term unemployment

merits special attention and even, at times,

political action. There are concerns that unem-

ployment statistics fail to record significant

numbers of people who want to work but are

excluded from the standard definition of un-

employment because of the requirement that an

active job search be undertaken in the reference

period. Alternatively, one may wish to apply a

broader statistical concept known as “long-term

joblessness”, covering working-age persons not

in employment and who have not worked

within the past one or two years. This measure

of the long-term jobless includes “discouraged

jobseekers”, that is, persons who are unem-

ployed but not seeking work due to specific

labour market-related reason, such as the belief

that no work is available to them. If long-term

joblessness is high, then unemployment, as

strictly defined, is less reliable as an indicator to

monitor effective labour supply, and macroeco-

nomic adjustment mechanisms may not bring

unemployment down.

Long-term unemployment is clearly related to

the personal characteristics of the unemployed,

and often affects older or unskilled workers, and

those who have lost their jobs through redun-

dancy. High ratios of long-term unemployment,

therefore, indicate serious unemployment prob-

lems for certain groups in the labour market and

often a poor record of employment creation.

Conversely, a high proportion of short-term

unemployed indicates a high job creation rate

and more turnover and mobility in the labour

market (see further details on the indicator on

labour flows – table 9c). Such generalizations

must be made with great care, however, as there

are many factors, including the issue of unem-

ployment benefit programmes cited above, that

can influence the relationship between long-

term unemployment and the relative economic

health of a given country. Indeed, in the absence

of some sort of compensatory income (or a

limited period of support), unemployed workers

may be obliged to lower their expectations and

take whatever job is available, thereby shorten-

ing their period of unemployment.

Definitions and sources

The standard definition of long-term unem-

ployment (table 11a) is all unemployed persons

with continuous periods of unemployment

extending for one year or longer (52 weeks and

103

KILM 11 Long-term unemployment

It should also be acknowledged that the

length of time that a person has been unem-

ployed is, in general, more difficult to measure

than many other statistics, particularly when the

data are derived from labour force surveys.

When unemployed persons are interviewed,

their ability to recall with any degree of preci-

sion the length of time that they have been

jobless diminishes significantly as the period of

joblessness extends. Thus, as it nears a full year,

it is quite easy to say “one year”, when in reality

the respondent may have been unemployed

between 10 and 14 months. If the household

respondent is a proxy for the unemployed for

person, the specific knowledge and the ability to

recall are reduced even further. Moreover, as the

jobless period lengthens, not only is the likeli-

hood of accurate recall reduced, but the jobless

period is more likely to have been interrupted

by limited periods of work (or of stopping

searching), but either this is forgotten over time

or the unemployed person may not consider

that work period as relevant to his or her “real”

unemployment problem (which is undoubtedly

consistent with society’s view as well).

All things considered, then, it must be

clearly understood that data on the duration of

unemployment are more likely to be unreliable

than most other labour market statistics.

However, this problem ought not to diminish

the importance of this indicator for individual

countries. The fact remains that the indicator

covers a group of individuals with serious diffi-

culties in the labour market. Whether the

period of joblessness is one year and longer or

ten months and longer, the group taken as a

whole is markedly afflicted by an undesirable,

unwanted status.

• Unemployment with a duration of three

months to less than six months

• Unemployment with a duration of six months

to less than 12 months

• Unemployment with a duration of 12 months

or more

The category of the unemployed with a

duration of 12 months or more (long-term

unemployment) is included in both tables

(table 11a and 11b); however, data in one table

may slightly differ from the other as different

sources or coverage may be used.

Limitations to comparability

Because all data presented in tables 11a and

11b come from labour force surveys or house-

hold surveys, fewer caveats need to accompany

cross-country comparisons. Nevertheless, while

data from household labour force surveys make

international comparisons easier, varius issues

can remain. Questionnaire design, survey

timing, differences in the age groups covered

and other issues affecting comparability (see

the discussion under KILM 9) mean that care is

required in interpreting cross-country differ-

ences in levels of unemployment. Also, as

mentioned above, users will want to know

something about the nature of unemployment

insurance coverage in countries of interest to

them, as substantial differences in such cover-

age – especially the lack of it altogether – can

have a profound effect on differences in long-

term unemployment.

Table 11b was constructed as an input for the calcula-

tion of labour flows (table 9c) and hence durations of un-

employment (table 11b) were included in the KILM with the

aim of contructing the longest possible time series, while

long-term unemployment (table 11a) was constructed with

the aim of using repositories that are consistent with other

indicators of the KILM (such as KILM 1, KILM 9, and

KILM 10).

Introduction

This indicator relates to the number of

employed persons whose hours of work in the

reference period are insufficient in relation to

a more desirable employment situation in

which the person is willing and available to

engage. The indicator was previously known as

“visible underemployment”. Two time-related

underemployment rates are presented: one

gives the number of persons in time-related

underemployment as a percentage of the

labour force, and the other as a percentage of

total employment. The information presented

in table 12 covers 78 countries. All information

is based on results from household surveys and

is disaggregated by sex and age group (total,

youth and adult), where possible.

Use of the indicator

Underemployment reflects underutilization

of the productive capacity of the labour force.

The concept of “underutilization” is a complex

one with many facets. In order to draw a more

complete picture of underutilization in relation

to the decent work deficit, one needs to exam-

ine a set of indicators, which includes but is not

limited to labour force, employment-to-popu-

lation ratios, inactivity rates, status in employ-

ment, working poverty and labour productivity.

Utilizing a single indicator to paint a picture of

underutilization will often provide an incom-

plete picture.

Underemployment has been broadly inter-

preted and has come to be used to imply any

sort of employment that is “unsatisfactory” (as

perceived by the worker) in terms of insufficient

hours, insufficient compensation or insufficient

use of one’s skills. The fact that the judgement

about underemployment is based on personal

assessment that could change daily at the whim

of the respondent, makes it a concept that is

difficult to quantify and to interpret. It is better

to deal with the more specific (more quantifi-

able) components of underemployment sepa-

rately; the “visible” underemployment can be

measured in terms of hours of work (time-

related underemployment) whereas “invisible”

underemployment, which is measured in terms

of income earned from the activity, low produc-

tivity, or the extent to which education or skills

are underutilized or mismatched, is much more

difficult to quantify. Time-related under-

employment is the only component of under-

employment to date that has been agreed on

and properly defined within the international

community of labour statisticians.

Statistics on time-related underemployment

are useful as a supplement to information on

employment and unemployment, particularly

the latter, as they enrich an analysis of the effi-

ciency of the labour market in terms of the abil-

ity of the country to provide full employment to

all those who want it.

In fact, the resolution

concerning statistics of work, employment and

labour underutilization, adopted by the 19th

International Conference of Labour Statisticians

(ICLS) in 2013, restated the definition of time-

related underemployment and its central role as

a measure of labour underutilization. A new

indicator meant to account for time-related

underemployment and supplement the unem-

ployment rate was also introduced, the

“combined rate of time-related underemploy-

ment and unemployment” (calculated as the

number of persons in unemployment or time-

related underemployment as a percentage of

the labour force).

Thus, the indicator on time-

related underemployment can provide insights

Time-related underemployment is listed as one vari-

able of “labour slack” in the framework for statistically

capturing the wider concept of “labour underutilization”.

Interested readers can refer to ILO: “Beyond unemploy-

ment: Measurement of other forms of labour underutiliza-

tion”, Room Document 13, 18th International Conference

of Labour Statisticians, Working group on Labour underuti-

lization, Geneva, 24 November – 5 December 2008; http://

www.ilo.org/wcmsp5/groups/public/---dgreports/---stat/

documents/meetingdocument/wcms_100652.pdf.

Resolution concerning the statistics of work, employ-

ment and labour underutilization, adopted by the 19th

International Conference of Labour Statisticians, Geneva,

2013; http://www.ilo.org/global/statistics-and-databases/

standards-and-guidelines/resolutions-adopted-by-interna-

tional-conferences-of-labour-statisticians/WCMS_230304/

lang--en/index.htm.

KILM 12.

Time-related

underemployment

106

KILM 12 Time-related underemployment

Definitions and sources

The international definition of time-related

underemployment was adopted by the 16th ICLS

in 1998 and several revisions to the text were

proposed by the 19th ICLS in 2013 in order to

clarify ambiguities.

The international definition

is based on three criteria: it includes all persons

in employment who, during a short reference

period (a) wanted to work additional hours,

(b) had worked less than a specified hours

threshold (working time in all jobs), and (c) were

available to work additional hours given an

opportunity for more work. Each of these criteria

is defined in further detail in the resolution itself

(see boxes 12a and 12b). Regarding the first

criterion, for example, workers should report

that they (1) want another job or jobs in addition

to their current employment, (2) want to replace

any of their current jobs with another job or jobs

with increased hours of work, (3) want to

increase the hours of work of any of their current

jobs, or (4) want a combination of these three

possibilities.

The current international definition of time-

related underemployment includes all workers

who report a desire to work additional hours.

This contrasts with the definition of unemploy-

ment, which includes non-employed persons

who would like to work only if they report

having actively sought work. There is evidence

that the number of time-related underem-

ployed persons would decrease significantly if

the definition were to include only those who

report having actually sought to work addi-

tional hours. This change would almost

certainly result in a greater decrease for women

than for men and would, therefore, illustrate

the fact that women tend not to look for addi-

tional work even if they actually want it, perhaps

because the time required for job seeking

would compete with the time needed for activi-

ties related to the gender role assigned to them

by society: that of caring for their households

and family members, for example.

Resolution concerning the measurement of underem-

ployment and inadequate employment situations, adopted

by the 16th International Conference of Labour Statisti-

cians, Geneva, 1998; http://www.ilo.org/global/statistics-

and-databases/standards-and-guidelines/resolutions-

adopted-by-international-conferences-of-labour-statisti-

cians/WCMS_087487/lang--en/index.htm. Report II of the

19th International Conference of Labour Statisticians,

Geneva, 2013: http://www.ilo.org/wcmsp5/groups/public/---

dgreports/---stat/documents/publication/wcms_220535.pdf.

for the design, implementation and evaluation

of employment, income and social policies and

programmes. Particularly in developing econo-

mies people only rarely fall under the clear-cut

dichotomy of either “employed” or “un-

employed”. Rather, the vast majority of the

population will be the underemployed who eke

out a living from small-scale agriculture and

other types of informal activities. As noted in a

study on the subject in Namibia,

very few

persons working only a few hours per week on

their small plots or guarding goats considered

themselves to be employed, particularly since

the earnings, in cash or kind from these activi-

ties were minimal. They were, however, classi-

fied as employed by the labour force survey

according to the international definition of

employment. In such situations, where the

majority of the population do not consider

themselves to be gainfully employed, an attempt

should be made to distinguish between the fully

employed and the underemployed.

Whereas unemployment is the most

common indicator used to assess the perfor-

mance of the labour market, in isolation it does

not provide sufficient information for an

understanding of the shortcomings of the

labour market in a country. For example, in the

situation above, employment as measured by

the standard labour force survey would be high

and unemployment low. Low unemployment

rates in these countries, however, do not

necessarily mean that the labour market is

effective. Rather, the low rates mask the fact

that a considerable number of workers work

fewer hours, earn lower incomes, use their

skills less, and, in general, work less produc-

tively than they could do and would like to do.

As a result, many are likely to be competing

with the unemployed in their search for alter-

native jobs and a clearer picture of the under-

utilization of the productive potential of the

country’s labour force can be gained by adding

the number of underemployed to the number

of unemployed as a share of the overall labour

force, as suggested by the resolution mentioned

in the preceding paragraph. Therefore, adding

an indicator of time-related underemployment

can assist in building a better understanding of

the true employment situation.

Scott, W.: “Simplified measurement of underemploy-

ment: Results of a labour force sample survey in Namibia”,

in Bulletin of Labour Statistics (Geneva, ILO), 1994-3;

http://www.ilo.org/wcmsp5/groups/public/---dgreports/---

stat/documents/publication/wcms_087909.pdf.

107

KILM 12 Time-related underemployment

the table is the number of hours of work (actual

or usual) at which a person is no longer counted

in the underemployment estimate.

As mentioned above, statistics for this indi-

cator are based exclusively on household

surveys. They were obtained mainly from inter-

national data repositories such as the OECD’s

labour statistics database, the Statistical Office

of the European Communities (EUROSTAT),

and the ILO’s online database (ILOSTAT).

National publications were also used in some

specific cases.

Limitations to comparability

National definitions of time-related under-

employment vary significantly between coun-

tries. Based on a review of country practices,

most national definitions include workers who

want to work additional hours (definition

code 2). Many other definitions include only

workers who report involuntary reasons either

for not working more hours or for working the

current number of hours (definition code 1).

The specific reasons considered as “involun-

tary”, however, vary significantly across coun-

tries. A certain number of countries obtain this

information in two stages. The first stage identi-

fies workers who usually work less than a thresh-

old for involuntary reasons, while the second

stage identifies workers whose actual hours are

below their usual hours for economic or techni-

cal reasons. The reasons considered as “involun-

tary” are not equivalent for the two groups of

workers identified, however. Some economies

apply the definition requiring workers to seek to

work additional hours (definition code 3).

Most definitions include persons whose

“hours actually worked” during the reference

week were below a certain threshold. Some

definitions include persons whose “hours

usually worked” were below a certain threshold

and other definitions include both groups of

workers. Perhaps because no international defi-

nition of “part time” exists, national determina-

tions of hourly thresholds are not always

consistent. In a few countries the threshold is

defined in terms of the legal hours or the usual

hours worked by full-time workers. Some coun-

tries enquire directly as to whether workers

work part time, or define the threshold in terms

of the worker’s own usual hours of work. As a

consequence, the threshold used varies signifi-

cantly from country to country. The hours cut-

off for Costa Rica, for example, used to be (until

2012) the full-time equivalent of 47 hours,

Despite the improvements in the clarity of

the definition of underemployment over the

last 20 years,

few countries apply the defini-

tion consistently because the criteria on which

it is specified are still not entirely precise. (This

is similar to the imprecise full-time/part-time

cut-off points, as discussed in KILM 6.) This lack

of precision has discouraged the production of

regular statistics on the subject and has made it

difficult to compare the levels of time-related

underemployment between countries. For

example, countries differ according to whether

actual or usual hours are used to identify

persons working less than the normal duration,

an issue also touched upon in KILM 6. The

resolution adopted by the 19th ICLS encour-

ages the separate identification of persons in

time-related underemployment according to

their usual hours of work and their actual hours

of work (and all combinations of these).

The indicator, as shown in table 12, reflects

the variety of interpretations of the standard

definition of time-related underemployment.

The national definitions are grouped according

to the following three common concepts (or

definition codes):

(1) Persons in employment who reported that

they were working part-time or whose hours

of work (actual or usual) were below a certain

cut-off point, and who also reported involun-

tary reasons for working fewer than full-time

hours – these are also known as “involuntary

part-time workers”.

(2) Persons in employment whose hours of work

(actual or usual) were below a certain cut-off

point and who wanted to work additional

hours.

(3) Persons in employment whose hours of work

(actual or usual) were below a certain cut-off

point and who sought to work additional

hours.

It is possible to compare countries that apply

the strictest definition (code 3) with countries

that apply a wider definition (codes 1 or 2) to

see to what extent the definition applied affects

the count of underemployed workers. The

hours cut-off information shown in the notes to

Underemployment was first addressed in resolu-

tion III adopted by the 11th ICLS concerning measurement

and analysis of underemployment and underutilization

of manpower (1966), and in resolution I adopted by the

13th ICLS concerning statistics of the economically active

population, employment, unemployment and underem-

ployment (1982).

KILM users should consult the notes to table 12 to

clarify which definition applies to each country.

108

KILM 12 Time-related underemployment

countries. Despite the fact that all the informa-

tion for this measurement comes from house-

hold surveys, a variety of other potential

limitations to comparability result from differ-

ences in the timing of surveys, sampling proce-

dures, collection questionnaires, and so on. A

succinct description of such limitations is

provided in the section of the chapter on KILM 9

regarding “Limitations to comparability”.

Box 12a. Resolution concerning the measurement

of underemployment and inadequate employment

situations, adopted by the 16th International Conference

of Labour Statisticians, October 1998

[relevant paragraphs]

Objectives

1. The primary objective of measuring underemployment and inadequate employment situations is

to improve the analysis of employment problems and contribute towards formulating and evaluating

short-term and long-term policies and measures designed to promote full, productive and freely

chosen employment as specified in the Employment Policy Convention (No. 122) and Recommendations

(Nos. 122 and 169) adopted by the International Labour Conference in 1964 and 1984. In this context,

statistics on underemployment and indicators of inadequate employment situations should be used

to complement statistics on employment, unemployment and inactivity and the circumstances of the

economically active population in a country.

2. The measurement of underemployment is an integral part of the framework for measuring the

labour force established in current international guidelines regarding statistics of the economically

active population; and the indicators of inadequate employment situations should as far as possible

be consistent with this framework.

Scope and concepts

3. In line with the framework for measuring the labour force, the measurement of underemployment

and indicators of inadequate employment should be based primarily on the current capacities and

work situations as described by those employed. Outside the scope of this resolution is the concept

of underemployment based upon theoretical models about the potential capacities and desires for

work of the working-age population.

4. Underemployment reflects underutilization of the productive capacity of the employed population,

including those which arise from a deficient national or local economic system. It relates to an

alternative employment situation in which persons are willing and available to engage. In this

resolution, recommendations concerning the measurement of underemployment are limited to time-

related underemployment, as defined in subparagraph 8(1) below.

5. Indicators of inadequate employment situations that affect the capacities and well-being of

workers, and which may differ according to national conditions, relate to aspects of the work situation

such as use of occupational skills, degree and type of economic risks, schedule of and travel to work,

occupational safety and health and general working conditions. To a large extent, the statistical

concepts to describe such situations have not been sufficiently developed.

6. Employed persons may be simultaneously in underemployment and inadequate employment

situations.

Measures of time-related underemployment

7. Time-related underemployment exists when the hours of work of an employed person are

insufficient in relation to an alternative employment situation in which the person is willing and

available to engage.

whereas most OECD countries report involun-

tary part-time only, meaning persons working at

or below 30 hours a week.

It should be clear from the foregoing discus-

sion concerning the wide variety of possibilities

for measuring time-related underemployment

that failure to isolate the definitional compo-

nents will greatly limit comparability between

109

KILM 12 Time-related underemployment

Box 12b. Resolution concerning statistics of work, employment and

labour underutilization, adopted by the 19th International

Conference of Labour Statisticians, October 2013

[relevant paragraphs]

155. The resolution incorporates guidelines for the measurement of time-related underemployment

based on the recommendations of the 16th ICLS resolution on this topic. The operational definition

of time-related underemployment has not been changed. However, several revisions to the text are

proposed in order to clarify ambiguities identified by countries in applying the international standards.

These relate particularly to the defining criteria of time-related underemployment, the relevant

working-time concepts used, and the different subgroups that may be identified to shed light on

structural and cyclical situations of time-related underemployment.

Time-related underemployment

156. As set forth in the 16th ICLS resolution, the definition of time-related underemployment

comprised three criteria. It referred to persons in employment who, in the short reference period,

wanted to work additional hours, had worked less than an hours threshold set at national level, and

who were available to work additional hours in a subsequent reference period. A main source of

ambiguity relates to the requirement to establish an hour’s threshold as part of the definition. This

criterion was introduced in order to focus the measure on situations related to insufficient quantity

of employment, as evidenced by the number of hours actually worked at all jobs in the reference

week. Exclusion of the threshold from the definition would result in the inclusion of persons who

wanted to work additional hours because of issues not related to insufficient quantity of work,

particularly due to low income, thus no longer being a measure of time-related underemployment.

157. To establish the hours threshold, countries may use a variety of approaches, including a

distinction based on notions of part-time/full-time employment, or on median or modal values of

hours usually worked. At the time when the standards were adopted by the 16th ICLS, an international

definition of hours usually worked did not exist. As a result, the resolution used the notion of normal

hours. Even then, however, the intention was to recommend the concept of hours usually worked in

order to have a measure in reference to the typical working time associated with specific groups of

8(1) Persons in time-related underemployment comprise all persons in employment, as defined in

current international guidelines regarding employment statistics, who satisfy the following three

criteria during the reference period used to define employment:

(a) “willing to work additional hours”, i.e. wanted another job (or jobs) in addition to their current job

(or jobs) to increase their total hours of work; to replace any of their current jobs with another job

(or jobs) with increased hours of work; to increase the hours of work in any of their current jobs;

or a combination of the above. In order to show how “willingness to work additional hours” is

expressed in terms of action which is meaningful under national circumstances, those who have

actively sought to work additional hours should be distinguished from those who have not.

Actively seeking to work additional hours is to be defined according to the criteria used in the

definition of job search used for the measurement of the economically active population, also

taking into account activities needed to increase the hours of work in the current job;

(b) “available to work additional hours”, i.e. are ready, within a specified subsequent period, to work

additional hours, given opportunities for additional work. The subsequent period to be specified

when determining workers’ availability to work additional hours should be chosen in light of

national circumstances and comprise the period generally required for workers to leave one job

in order to start another;

in all jobs during the reference period, as defined in current international guidelines regarding

working-time statistics, were below a threshold, to be chosen according to national circumstances.

This threshold may be determined by e.g. the boundary between full-time and part-time employ-

ment, median values, averages, or norms for hours of work as specified in relevant legislation,

collective agreements, agreements on working-time arrangements or labour practices in countries.

(Box 12a continued)

110

KILM 12 Time-related underemployment

persons in employment. As different industries may have different typical working-time patterns, for

example in agriculture, the draft resolution allows the setting of different hours thresholds for different

worker groups, depending on national circumstances.

158. A second source of ambiguity concerns the reference period against which to assess the

availability criterion. The 16th ICLS resolution provides detailed guidelines for establishing the reference

period for availability as comprising the “period generally required for workers to leave one job in order

to start another”. In practice however, most countries have used a similar period as that used for

establishing availability as part of the definition of unemployment. Such practice is likely to result in an

underestimation of time-related underemployment by referring to a situation in the past when the

person would not have made arrangement to become available for additional work. This would be, in

particular, the case for persons with responsibilities outside of employment, including those providing

care for dependent members of the household, and those engaged also in other forms of work.

159. A final source of ambiguity is the distinction between the two categories of persons in time-

related underemployment, namely, those who work usually less than the hours threshold and those

who usually work more than the hours but who, during the short reference period, were not at work

or actually worked reduced hours for economic reasons. These two groups are mutually exclusive:

(a) The first group is in a prolonged situation of time-related underemployment (with both hours actu-

ally worked and hours usually worked below the threshold for time-related underemployment).

As such, when separately identified, this group may be useful for examining structural situations

of insufficient quantity of employment among the employed.

(b) The second group is in a temporary situation of time-related underemployment. As such it reflects

situations of insufficient quantity of employment due to cyclical or seasonal factors.

Relevant paragraphs of the resolution adopted by the 19th ICLS

43. Persons in time-related underemployment are defined as all persons in employment who, during

a short reference period, wanted to work additional hours, whose working time in all jobs was less

than a specified hours threshold, and who were available to work additional hours given an opportunity

for more work, where:

(a) the “working time” concept is hours actually worked or hours usually worked, dependent on the

measurement objective (short or long-term situations) and in accordance with the international

statistical standards on the topic;

(b) “additional hours” may be hours in the same job, in an additional job(s) or in a replacement job(s);

the median or modal values of the hours usually worked of all persons in employment, or on

working time norms as specified in relevant legislation or national practice, and set for specific

worker groups;

(d) “available” for additional hours should be established in reference to a set short reference period

that reflects the typical length of time required in the national context between leaving one job

and starting another.

44. Depending on the working time concept applied, among persons in time-related underemployment

(i.e. who wanted and were “available” to work “additional hours”), it is possible to identify the following

groups:

(a) persons whose hours usually and actually worked were below the “hours threshold”;

(b) persons whose hours usually worked were below the “hours threshold” but whose hours actually

worked were above the threshold;

economic reasons (e.g. a reduction in economic activity including temporary lay-off and slack

work or the effect of the low or off season).

45. In order to separately identify the three groups of persons in time-related underemployment,

information is needed on both hours usually worked and hours actually worked. Countries using only

one working time concept will cover, for hours usually worked, the sum of groups (a) and (b); for hours

actually worked, the group (c), so long as the reasons for being “not at work” or for working below

the “hours threshold” are also collected.

46. To assess further the pressure on the labour market exerted by persons in time-related

underemployment, it may be useful to identify separately persons who carried out activities to seek

“additional hours” in a recent period that may comprise the last four weeks or calendar month.

(Box 12a continued)

Introduction

The inactivity rate is the proportion of the

working-age population that is not in the labour

force. Summing up the inactivity rate and the

labour force participation rate (see KILM 1) will

yield 100 per cent. Information on this indicator

is given for 189 economies for the same stan-

dardized age groupings provided in KILM

table 1a: 15+, 15-24, 15-64, 25-54, 25-34, 35-54,

55-64 and 65+ years. The estimates are harmon-

ized to account for differences in countries’ data

collection and tabulation methodologies as well

as for other country-specific factors such as mili-

tary service requirements. The series includes

both country reported and imputed data.

Use of the indicator

Although labour market economists tend to

focus on the activities and characteristics of

people in the labour force, there has been

continued, if less visible, interest in individuals

outside the labour market, especially those

who want to work but are not currently seeking

work.

Much of this growing interest stems

The resolution concerning the statistics of work,

employment and labour underutilization, adopted by the

19th International Conference of Labour Statisticians in

2013 taps into this pool of inactive persons by identifying

situations of inadequate absorption of labour beyond

those captured by unemployment. The resolution intro-

duces a definition of potential labour force and proposes

that the definition cover persons who have indicated

some interest in employment but who are currently

counted as being outside of the labour force. It distin-

guishes three mutually exclusive groups:

a) unavailable jobseekers, referring to persons without

employment who are seeking employment but are not

available;

b) available potential jobseekers, referring to persons

without employment who are not seeking employment

but are available; and

c) willing potential jobseekers,

comprising persons with-

out employment who are neither seeking nor available

for employment but who want employment.

Further information can be sought at: http://www.ilo.

org/wcmsp5/groups/public/---dgreports/---stat/documents/

publication/wcms_220535.pdf.

from concern over improving the availability of

decent and productive employment opportuni-

ties in developing and developed economies

alike. Individuals are considered to be outside

the labour force, if they are neither employed

nor unemployed, that is, not actively seeking

work. There is a variety of reasons why some

individuals do not participate in the labour

force; such persons may be occupied in caring

for family members; they may be retired, sick or

disabled or attending school; they may believe

no jobs are available; or they may simply not

want to work.

In some situations, a high inactivity rate for

certain population groups should not necessar-

ily be viewed as “bad”; for instance, a relatively

high inactivity rate for young people aged 25 to

34 years may be due to their non-participation

in the labour force to receive education.

Furthermore, a high inactivity rate for women

aged 25 to 34 years may be due to their leaving

the labour force to attend to family responsibili-

ties such as childbearing and childcare. Using

the data in KILM 13, users can investigate the

extent to which motherhood relates to the

labour force patterns of women. It has long

been recognized that aspects of household

structure are associated with labour market

activity. For example, female heads of house-

holds tend to have relatively high inactivity

rates. Among married-couple families, husbands

typically have low inactivity rates, especially if

there are children in the family. However, a low

rate of female inactivity could coincide with a

high rate for men, for instance if the male is

completing his education or is physically unable

to work, thus making the wife the primary

wage earner.

A subgroup of persons outside the labour

force comprises those known as discouraged

jobseekers, defined as persons not in the

labour force, who are available for work but no

longer looking for work due to specific labour

market-related reasons, such as the belief that

there are no jobs available. This is typically for

personal reasons associated with their percep-

tion of lack of job availability. Regardless of

their reasons for being discouraged, these

potential workers are generally considered to

KILM 13. Persons outside

the labour force

112

KILM 13 Persons outside the labour force

survey source, only one type of source was

used. If a labour force survey was available for

the country, inactivity rates derived from it were

chosen in favour of those derived from a popu-

lation census. Only inactivity rates that are suffi-

ciently representative of the standardized age

groups (15+, 15-24, 15-64, 25-34, 25-54, 35-54,

55-64 and 65+ years) were used in the construc-

tion of the series.

Table 13 includes both real (country reported)

inactivity rates as well as rates that were imputed

using econometric modelling techniques. GDP

levels and growth rates, population age structure

variables and dummy variables to capture time

trends, region-specific trends and country fixed

effects were among the explanatory variables

used to generate the imputed labour force

participation rates in KILM table 1a, which were

then used in the construction of table 13. These

rates were estimated separately both for each age

group as well as for the sexes.

Limitations to comparability

The usual comparability issues stemming

from differences in concepts and methodolo-

gies according to types of survey, variations in

age groups, geographic coverage, etc., do not

apply in the case of table 13. The table is derived

from the harmonized labour force participation

rates in table 1a, where only data deemed suffi-

ciently comparable across countries were used,

which makes table 13 harmonized (and compa-

rable) by default. The selection criteria for

creating the harmonized data set were

explained in the previous section.

be underutilized. The presence of discouraged

jobseekers is implied if the measured labour

force grows when unemployment is falling

(although demographic pressures should also

be taken into consideration). People who were

not counted as unemployed (because they

were not actively searching for work) when

there were few jobs to be had may change their

mind and look for work when the odds of find-

ing a job improve. Furthermore, when numbers

of discouraged jobseekers are high, policy-

makers may attempt to “recapture” members

of this group by improving job placement

services. (See also the discussion on “discour-

agement” in the chapter on KILM 10.)

Definitions and sources

There are several aspects of the definition to

consider for the indicator on persons outside the

labour force. Foremost is the fact that estimates

must be made for the entire population, either

through labour force surveys, population

censuses, or similar means. Typically, determina-

tions are made as to the labour force status of the

relevant population. The labour force is defined

as the sum of the employed and the unemployed.

The remainder of the population is the number

of persons outside the labour force.

Only labour force participation rates and

population figures deemed sufficiently compa-

rable across countries were used in the

construction of table 13.

To this end, only

labour force survey and population census-

based data were used in the construction of the

estimates. In countries with more than one

See the corresponding section in the chapter on

KILM 1 for details of the construction of the harmonized

table 1a. Since table 13 complements table 1a, the same

methodologies for its construction apply.

Introduction

KILM 14 reflects the levels and distribution

of the knowledge and skills base of the labour

force and the unemployed. Tables 14a and 14b

show the distribution of the educational attain-

ment of the labour force and the unemployed

for 137 and 141 countries, respectively, accord-

ing to five levels of schooling – less than one

year, pre-primary level, primary level, second-

ary level, and tertiary level. Table 14c provides

information on the unemployment rate, that is,

the share of the unemployed in the labour

force, according to three groupings of educa-

tional attainment: primary or less, secondary

and tertiary, for 128 countries. Finally, table 14d

presents information on illiteracy rates as the

percentage of illiterate persons in the popula-

tion for 166 countries.

The data in tables 14a, 14b, 14c, and 14d are

broken down by sex and wherever possible by

the following age cohorts: total (15 years and

over), youth (15 to 24 years), and adult

(25 years and over).

Use of the indicator

In all countries, human resources represent,

directly or indirectly, the most valuable and

productive resource; countries traditionally

depend on the health, strength and basic skills

of their workers to produce goods and services

for consumption and trade. The advance of

complex organizations and knowledge require-

ments, as well as the introduction of sophisti-

cated machinery and technology, means that

economic growth and improvements in welfare

increasingly depend on the degree of literacy

and educational attainment of the total popula-

tion. The population’s predisposition to acquire

such skills can be enhanced by experience,

informal and formal education, and training.

Although the natural endowments of the

labour force remain relevant, continuing

economic and technological change means that

the bulk of human capital is now acquired, not

only through initial education and training, but

also increasingly through adult education and

enterprise or individual worker training, within

the perspective of lifelong learning and career

management. Unfortunately, quantitative data

on lifelong learning, and indicators that moni-

tor developments in the acquisition of knowl-

edge and skills beyond formal education, are

sparse. Statistics on levels of educational attain-

ment, therefore, remain the best available indi-

cators of labour force skill levels to date. These

are important determinants of a country’s

capacity to compete successfully and sustain-

ably in world markets and to make efficient use

of rapid technological advances. They should

also affect the employability of workers.

The ability to examine education levels in

relation to occupation and income is also useful

for policy formulation, as well as for a wide

range of economic, social and labour market

analyses. Statistics on levels of, and trends in,

educational attainment of the labour force can:

(a) provide an indication of the capacity of

countries to achieve important social and

economic goals; (b) give insights into the broad

skill structure of the labour force; (c) highlight

the need to promote investments in education

for different population groups; (d) support

analysis of the influence of skill levels on

economic outcomes and the success of differ-

ent policies in raising the educational level of

the workforce; (e) give an indication of the

degree of inequality in the distribution of

educational resources between groups of the

population, particularly between men and

women, and within and between countries;

and (f) provide an indication of the skills of the

existing labour force, with a view to discover-

ing untapped potential.

By focusing on the educational characteris-

tics of the unemployed, the KILM 14 indicator

can also help to shed light on how significant

long-term events in a country, such as skill-

based technological change, increased trade

openness or shifts in the sectoral structure of

the economy, alter the experience of high- and

low-skilled workers in the labour market. The

information provided can have important

KILM 14. Educational attainment

and illiteracy

114

KILM 14 Educational attainment and illiteracy

programmes to a level of education were clari-

fied and tightened, and the fields of education

were further elaborated.

Many countries

continue to classify education levels according to

the levels of ISCED-76, but more and more coun-

tries have made the change to the nine levels and

ten subcategories of ISCED-97. In 2011, a new

classification ISCED 2011 was introduced;

however, reporting according to ISCED-11 did

not start until 2014.

Tables 14a to 14c clearly

identify which classification system applies for

each record. The main education levels are also

summarized in the table below.

The major attainment levels in KILM 14 are

primary, secondary and tertiary education.

Primary education aims to provide the basic

elements of education (for example, at elemen-

tary or primary school and lower secondary

school) and corresponds to ISCED levels 1

and 2. Curricula are designed to give students

a sound basic education in reading, writing and

arithmetic, along with an elementary under-

standing of other subjects such as history, geog-

raphy, natural science, social science, art, music

and, in some cases, religious instruction. Some

vocational programmes, often associated with

relatively unskilled jobs, as well as apprentice-

ship programmes that require further educa-

tion, are also included. Students generally

begin primary education between the ages of

and 7 years and end at 13 to 15 years. Literacy

programmes for adults, similar in content to

programmes in primary education, are also

classified under primary education.

Secondary education is provided at high

schools, teacher-training schools at this level,

and schools of a vocational or technical nature.

General education continues to be an important

constituent of the curricula, but separate subject

presentation and more specialization are also

found. Secondary education consists of ISCED

levels 3 (designated “upper secondary educa-

tion”) and 4 (designated “post-secondary non-

tertiary education”), and students generally

begin between 13 and 15 years of age and finish

between 17 and 18 years of age. It should be

noted that the KILM classifications of primary

and secondary education differ from the classifi-

cations used in UNESCO publications, in which

level 2 is termed “lower secondary education”.

For further details about the ISCED see UNESCO:

International Standard Classification of Education/ISCED

1997 (Paris, 1998); http://www.uis.unesco.org/Education/

Pages/international-standard-classification-of-education.aspx.

For further details on ISCED 2011, see UNESCO: Inter-

national Standard Classification of Education/ISCED 2011

(Paris, 2012); http://www.uis.unesco.org/Education/Pages/

international-standard-classification-of-education.aspx.

implications for both employment and educa-

tion policy. To the extent that persons with low

education levels are at a higher risk of becom-

ing unemployed, the political reaction may be

either to seek to increase their education level

or to create more low-skilled occupations

within the country.

Alternatively, a higher share of unemploy-

ment among persons with higher education

could indicate a lack of sufficient professional

and high-level technical jobs. In many coun-

tries, qualified jobseekers are being forced to

accept employment below their skill level.

Where the supply of qualified workers outpaces

the increase in the number of professional and

technical employment opportunities, high

levels of skills-related underemployment (see

the chapter on KILM 12 for more information)

are inevitable. A possible consequence of the

presence of highly educated unemployed in a

country is a “brain drain”, whereby educated

professionals migrate in order to find employ-

ment in other areas of the world.

While not a labour market indicator in itself,

the illiteracy rate of the population may be a

useful proxy for basic educational attainment in

the potential labour force. Literacy and numer-

acy are increasingly considered to be the basic

minimal skills necessary for entry into the

labour market.

Definitions and sources

Educational attainment

The six categories of educational attainment

used in KILM 14 are conceptually based on the

ten levels of the International Standard

Classification of Education (ISCED). The ISCED

was designed by the United Nations Educational,

Scientific and Cultural Organization (UNESCO)

in the early 1970s to serve as an instrument suit-

able for assembling, compiling and presenting

comparable indicators and statistics of educa-

tion, both within countries and internationally.

The original version of ISCED (ISCED-76) classi-

fied educational programmes by their content

along two main axes: levels of education and

fields of education. The cross-classification vari-

ables were maintained in the revised ISCED-97;

however, the rules and criteria for allocating

For more information relating to definitions of the

labour force and unemployment, users can consult the

chapters on KILMs 1 and 9, respectively.

115

KILM 14 Educational attainment and illiteracy

Education classifications used in KILM table 14

KILM Level ISCED-11 Level ISCED-97 Level ISCED-76 Level Description

Less than

primary

X: No schooling X: No schooling X: No schooling Less than one year of schooling

0: Early childhood

education

0: Pre-primary

education

0: Education

preceding the first

level

Education delivered in kindergartens,

nursery schools or infant classes

Primary

1: Primary education 1: Primary education

or first stage of basic

education

1: First level Programmes are designed to give students

a sound basic education in reading, writing

and arithmetic. Students are generally

5-7 years old. Might also include adult

literacy programmes.

2: Lower secondary

education

2: Lower secondary

or second stage

of basic education

2: Second level,

first stage

Continuation of basic education, but with

the introduction of more specialized subject

matter. The end of this level often coincides

with the end of compulsory education

where it exists. Also includes vocational

programmes designed to train for specific

occupations as well as apprenticeship

programmes for skilled trades.

Secondary

3: Upper secondary

education

3: Upper secondary

education

3: Second level,

second stage

Completion of basic level education, often

with classes specializing in one subject.

Admission usually restricted

to students who have completed the 8-9

years of basic education or whose basic

education and vocational experience

indicate an ability to handle the subject

matter of that level.

4: Post-secondary

non-tertiary

education

4: Post-secondary

non-tertiary

education

Captures programmes that straddle

the boundary between upper-secondary

and post-secondary education.

Programmes of between six months and

two years typically serve to broaden

the knowledge of participants who have

successfully completed level 3

programmes.

Tertiary

5: Short-cycle

tertiary education

5: First stage of

tertiary education

(not leading directly

to an advanced

research qualification);

subdivided into:

6: Bachelor’s or

equivalent level

5A 6: Third level,

first stage, leading

to a first university

degree

Programmes are largely theoretically based

and are intended to provide sufficient

qualifications for gaining entry into

advanced research programmes. Duration

is generally 3-5 years.

5B 5: Third level,

first stage, leading

to an award

not equivalent

to a first university

degree

Programmes are typically of a “practical”

orientation designed to prepare students

for particular vocational fields (high-level

technicians, teachers, nurses, etc.).

7: Master’s or

equivalent level

6: Second stage

of tertiary education

(leading to an

advanced research

qualification)

7: Third level,

second stage

Programmes are devoted to advanced

study and original research and typically

require the submission of a thesis

or dissertation.

8: Doctoral or

equivalent level

116

KILM 14 Educational attainment and illiteracy

Limitations to comparability

A number of factors can limit the appropri-

ateness of using the indicators for comparisons

of statistics on education between countries or

over time. First, it should be noted that the

same limitations relating to comparability of

other indicators based on labour force apply

here as well. The discussion in the correspond-

ing section of the chapters on KILMs 1 and 9

should be read for additional details on the

caveats relating to comparability.

In addition to the differences associated

with varying information sources, the way in

which individuals in the labour force are

assigned to educational levels can also severely

limit the feasibility of cross-country compari-

sons. Many countries have difficulty establish-

ing links between their national classification

and ISCED, especially with respect to technical

or professional training programmes, short-

term programmes and adult-oriented

programmes (ranging around levels 3 and 5 of

ISCED-76 and levels 3, 4 and 5 of ISCED-97). In

numerous situations, ISCED classifications are

not strictly adhered to: a country may choose

to include level 3 (secondary) with levels 5, 6

and 7 (tertiary), or levels 1 or 2 (primary) may

include level 0 (pre-primary). It should also be

noted that in a few countries ISCED levels are

combined in a different way; for instance,

levels 1 and 2 (taken together as the primary

level) may refer to level 1 only, as in many coun-

tries in Latin America and the Caribbean, or to

level 2 only. It is necessary to pay close atten-

tion to the notes – specifically, the notes given

in the column “Classification note” – in order

to ascertain the actual distribution of education

levels before making comparisons.

An issue that affects several countries in the

European Union subgroup of the Developed

Economies originates from the way in which

those who have received their highest level of

education in apprenticeship systems are classi-

fied. The classification of apprenticeship in the

“secondary” level – despite the fact that this

involves one or more years of study and train-

ing beyond the conventional length of second-

Tertiary education is provided at universi-

ties, teacher-training colleges, higher profes-

sional schools, and sometimes distance-learning

institutions. It requires, as a minimum condi-

tion of admission, the successful completion of

education at the secondary level or evidence of

the attainment of an equivalent level of knowl-

edge. It corresponds to ISCED levels 5, 6, 7

and 8 (levels 5A, 5B and 6 in ISCED-97 and

levels 5, 6 and 7 in ISCED-76).

In addition to primary, secondary and

tertiary education, KILM 14 also covers two

other categories of educational attainment that

correspond to ISCED levels: less than primary

(levels X and 0); and level of education not

definable (level 9).

The statistics on educational attainment of

the labour force, including the unemployed,

were obtained from the ILO online database

(ILOSTAT); the Caribbean Labour Statistics

Dataset; the OECD and EUROSTAT online data-

bases; and information collected from national

statistical offices. Information on educational

attainment is typically collected through house-

hold surveys, official estimates and population

censuses conducted by national statistical

services.

Illiteracy rates

Literacy is defined as the skills to read and

write a simple sentence about everyday life.

Illiteracy is the inverse, that is, the lack of the

skills to read and write a simple sentence about

everyday life. The source of information for the

number of illiterate persons and the illiteracy

rates is UNESCO’s Institute for Statistics (UIS).

The estimates are either national, based on

data collected during national population

censuses and household surveys, or are UIS

estimates. Information about the model estima-

tion methodology is available on the UIS

website.

The UIS literacy and illiteracy estimates are available

at: http://www.uis.unesco.org/Literacy/Pages/default.aspx.

KILM Level ISCED-11 Level ISCED-97 Level ISCED-76 Level Description

Not definable

9: Not elsewhere

classified

9: Education

not definable by level

Programmes for which there are no

entrance requirements.

Not stated

?: Level not stated ?: Level not stated

(Education classifications used in KILM table 14, continued)

117

KILM 14 Educational attainment and illiteracy

different social and cultural contexts, different

definitions and standards of literacy, and differ-

ent methodologies for collecting and compiling

the literacy data, as well as variations in the qual-

ity of data collected, and caution is needed in

comparing the literacy situations among coun-

tries and regions. Some countries define illiter-

acy not by reading and writing aptitude, but by

the years of schooling attained. For example, a

person is categorized as illiterate in Estonia

(2000) if they have not completed primary

education, while in Malaysia (2010), an illiterate

person is someone who has never been to

school. These data points, therefore, should not

be compared against, say, Angola (2012), where

illiterate persons are defined as those who

cannot easily read a letter or a newspaper.

ary schooling in other countries – can lower the

reported proportion of the labour force or

population with a tertiary education, compared

with countries where the vocational training is

organized differently. This classification issue

substantially holds down the levels of tertiary

education reported by Austria and Germany, for

instance, where the participation of young

people in the apprenticeship system is

widespread.

Limitations to comparability of information

on illiteracy rates, as given in table 14d, exist

because of variations in the definition of illiter-

acy. The most common definition is the inability

to read and write a simple statement about

everyday life. However, different countries have

in implementing and assessing employment,

wages and other social policies. Information on

labour cost per unit of labour input (that is, per

time unit) is particularly useful in the analysis

of certain industrial problems, as well as in the

field of international economic cooperation

and international trade.

Introduction

Table 15a presents trends in average

monthly wages, both in nominal and real terms

(i.e. adjusted for changes in consumer prices),

where available. Average wages represent one

of the most important aspects of labour market

information as wages are a substantial form of

income, accruing to a high proportion of the

labour force, namely persons in paid employ-

ment (employees). In most developed econo-

mies, more than 85 per cent of the employed

population are paid employees, and the share

of paid employees has been constantly rising in

many of the newly industrializing countries

(see KILM 3). Information on wage levels is

essential to evaluate the living standards and

conditions of work and life of this group of

workers in both developed and developing

economies. It helps to assess how far economic

growth and rising labour productivity (KILM 16)

translate into better living standards for ordi-

nary workers and to the reduction of working

poverty (KILM 17).

There is also a particular need for informa-

tion on average wages in planning economic and

social development, establishing income and

fiscal policies, fixing social security contributions

and benefits, and in regulating minimum wages

and for collective bargaining. Policy-makers, as

well as employers and trade unions, pay close

attention to wage trends. At the global level, the

ILO’s biennial Global Wage Report analyses wage

trends across different regions and discusses the

This was also the rationale for including average real

wages in the ILO’s list of Decent Work Indicators; see Guide

to the Millennium Development Goals Employment Indi-

cators (Geneva, ILO, 2nd edition, 2013); http://www.ilo.

org/empelm/what/WCMS_208796/lang--en/index.htm.

This chapter presents two distinct and

complementary indicators. The first, table 15a,

shows trends in average monthly wages in the

total economy for 126 countries, while the

second, table 15b, presents the trends and struc-

ture of employers’ average compensation costs

for the employment of workers in the manufac-

turing sector, available for 33 countries.

These two indicators differ in their nature

and primary objectives. Wages are important

from the workers’ point of view and represent a

measure of the level and trend of their purchas-

ing power and an approximation of their stan-

dard of living, while the second indicator

provides an estimate of employers’ expenditure

toward the employment of its workforce. The

indicators are, nevertheless, complementary in

that they reflect the two main facets of existing

wage measures; one aiming to measure the

income of employees, the other showing the

costs incurred by employers for employing them.

For most employees, wages – the income

they receive from paid employment – represent

the main part of their total labour-related

income. Information on workers’ wages is a

valuable economic indicator for planners,

policy-makers, employers and workers them-

selves. The statistical series in table 15a show

nominal average wages and real average wages.

From the employers’ standpoint, wages are

only one component of the cost of employing

labour, which is usually referred to as labour

costs (according to the ILO concept), employ-

ment costs or compensation costs. Other cost

elements include employers’ expenditure on

social security benefits, provided either as

direct payments to the employees or as contri-

butions to funds set up for the purpose, as well

as the cost of various benefits, services and

facilities (such as housing, vocational training

and welfare provisions) which are primarily

intended to benefit workers. Table 15b pre-

sents both the level and structure of compen-

sations costs, with distinction made between

total hourly direct pay and hourly social insur-

ance expenditures and labour-related taxes.

Assessing the change in labour costs over time

can play a central role in wage negotiations and

KILM 15. Wages

and compensation costs

120

KILM 15 Wages and compensation costs

labour costs (see box 15b). In particular, the

costs of recruitment, employee training, and

plant facilities and services, such as cafeterias,

medical clinics and welfare services, are not

included. It is estimated that the labour costs

not included in hourly compensation costs

account for around 1 to 2 per cent of total

labour costs for those countries for which infor-

mation is presented. This measure is also

closely related to the “compensation of employ-

ees” measure used in the system of national

accounts,

which can be considered a proxy for

total labour costs.

Use of the indicator

Real wages in an economic activity are a

major indicator of employees’ purchasing

power and a proxy for their level of income,

independent of the actual work performed in

that activity. Real wage trends are, therefore,

useful indicators, both within countries and

across them. Significant differences in the

purchasing power of wages, over time and

between countries, reflect the modern world

economy, and comparisons of the movement of

real wages can provide a measure of the mate-

rial progress (or regression) of the working

population. Real average wages are therefore

an important indicator for monitoring changes

in working conditions. And they should be

reviewed in conjunction with trends in working

poverty (KILM 17) and low pay incidence.

Trends in nominal wages can be used to

inform adjustments in minimum wages, the

lowest remuneration that employers may

legally pay to workers under national law.

While there is no single, recommended ratio

between minimum wages and average wages,

information on average wages can inform

policy-makers when setting minimum wages

and enable them to monitor whether those at

the bottom of the distribution fall behind

general wage increases.

Social partners – workers’ and employers’

organizations – rely on wage data for collective

See System of National Accounts 2008 (New York,

United Nations, 2009); http://unstats.un.org/unsd/nationa-

laccount/docs/SNA2008.pdf.

Note that minimum wages are set in nominal terms,

so nominal average wages are the primary comparator. For

a review of minimum wage legislation, see ILO: Working

Conditions Laws Report 2010 (Geneva, 2010).

See Chapter 5.2 of ILO: Global Wage Report 2010/11:

Wage policies in times of crisis (Geneva, 2010).

role of wage policies (see box 15a).

In addition

to the relevance of wage data, international stan-

dards were long ago developed, adopted and

implemented for the concepts, scope and meth-

ods of collection, as well as for the compilation

and classification of wage statistics (see “defini-

tions and sources”). This should, in principle,

facilitate international comparisons.

The indicator of table 15b is concerned with

the levels, trends and structures of employers’

hourly compensation costs for the employment

of workers in the manufacturing sector. The

measure is shown for all employees and it

includes the total compensation cost levels

expressed in absolute figures in US dollars and

as an index relative to the costs in the United

States (on the basis of US = 100). Total compen-

sation is also broken down into “hourly direct

pay” with subcategories “pay for time worked”

and “directly paid benefits”, and ”social insur-

ance expenditure and labour-related taxes”

with all variables expressed in US dollars.

Average hourly compensation cost is a

measure intended to represent employers’

expenditure on the benefits granted to their

employees as compensation for an hour of

labour. These benefits accrue to employees,

either directly – in the form of total gross earn-

ings – or indirectly – in terms of employers’

contributions to compulsory, contractual and

private social security schemes, pension plans,

casualty or life insurance schemes and benefit

plans in respect of their employees. This latter

group of benefits is commonly known as “non-

wage benefits” or “non-wage labour costs”

when referring to employers’ expenditure and

in table 15b is captured in “social insurance

expenditures and labour-related taxes”.

Compensation cost is closely related to

labour cost, although it does not entirely corre-

spond to the ILO definition of total labour cost

contained in the 1966 ILO resolution concern-

ing statistics of labour cost, adopted by the 11th

International Conference of Labour Statisticians

(ICLS),

in that it does not include all items of

For the latest report, please refer to ILO: Global Wage

Report 2014/15: Wages and income inequality (Geneva,

2015); http://www.ilo.org/global/research/global-reports/

global-wage-report/2014/lang--en/index.htm. Data associ-

ated with the ILO Global Wage Report are available in

ILOSTAT; www.ilo.org/ilostat.

Resolution concerning statistics of labour cost,

adopted by the 11th International Conference of Labour

Statisticians, Geneva, 1966; http://www.ilo.org/global/statis-

tics-and-databases/standards-and-guidelines/resolutions-

adopted-by-international-conferences-of-labour-statisti-

cians/WCMS_087500/lang--en/index.htm (see box 15b).

121

KILM 15 Wages and compensation costs

units, and information on hours of work –

required to calculate hourly labour costs – is

not always available from the countries covered.

International comparisons are thus hampered

by a lack of harmonization in terms of defini-

tions, methodology and measurement units.

National definitions of earnings differ consider-

ably, earnings do not include all items of labour

compensation and the omitted items of

compensation may represent a large propor-

tion of total compensation.

For these reasons, table 15b is based on

another source of information, namely the esti-

mates of hourly compensation costs of employ-

ees in manufacturing as compiled by The

Conference Board.

The Conference Board

series adjust published earnings data for items

of compensation not included in earnings and

although these estimates do not entirely cor-

respond to the ILO definition of total labour

costs, they are closely related to it and account

for nearly all labour costs in any country

presented within the indicator, resulting in the

most reliable available series in terms of inter-

national comparability.

Information on compensation or labour

costs is not generally available separately for

men and women. Many establishments from

which this information is collected do not

maintain separate data by sex for non-direct

pay. In addition, the distribution of male and

female workers according to occupation, levels

of skill and supervisory responsibilities are

often dissimilar within an industry, between

establishments and among countries.

Therefore, comparisons of compensation cost

information between men and women based

on an allocation of costs proportional to the

respective number of persons or the amount of

earnings could lead to erroneous conclusions.

The same remarks apply to the measurement of

total labour costs, where it is even more diffi-

cult to allocate the cost of certain components,

such as welfare services or vocational training,

between men and women. With these difficul-

ties in mind, the ILO resolution concerning

statistics of labour cost did not recommend the

Capdevielle, P. and Sherwood, M.: “Providing compar-

able international labor statistics”, in Monthly Labor Review

(Washington, DC, BLS, June 2002); http://www.bls.gov/

opub/mlr/2002/06/art1full.pdf.

The United States Bureau of Labor Statistics (BLS)

had an International Labor Comparisons (ILC) program

which compiled this data but it was eliminated in 2013.

Upon its termination, The Conference Board acquired all

the BLS data files and now continues to produce updates

to the series; see https://hcexchange.conference-board.org/

ilcprogram/index.cfm.

bargaining. A fundamental concern of employ-

ees and trade unions is to protect the purchas-

ing power of wages, particularly in periods of

high inflation, by raising nominal wages in line

with changes in consumer prices. Real wage

increases become feasible without putting the

sustainability of enterprises into jeopardy when

labour productivity is growing.

When used together with other economic

variables such as employment, production, and

income and consumption, trends in average real

wages are valuable indicators for the analysis of

overall macroeconomic trends, as well as in

economic planning and forecasting. Importantly,

they can indicate the extent to which economic

growth and rising labour productivity translates

into income gains for workers. These, in turn,

influence aggregate demand, and countries with

external surpluses can utilize wage policies to

re-balance their economies by strengthening

domestic consumption.

Information on hourly compensation costs

(table 15b), like total labour cost, is valuable for

many purposes. The level and structure of the

cost of employing labour and the way costs

change over time can play a central role in

every country, not only for wage negotiations

but also for defining, implementing and assess-

ing employment, wage and other social and

fiscal policies that target the distribution and

redistribution of income. At both the national

and international levels, labour costs are a

crucial factor in the abilities of enterprises and

countries to compete. When specific to the

manufacturing sector, labour costs serve as an

indicator of competitiveness of manufactured

goods in world trade. This is why governments

and social partners, as well as researchers and

national and international institutions, are

interested in labour cost information that can

be compared between countries and indus-

tries. Also, the measurement and analysis of

non-wage labour costs have become an impor-

tant issue in debates on labour market flexibil-

ity, employment policies, analyses of cost

disparities, and comparisons of productivity

levels among countries.

Not all countries compile statistics on total

labour costs as defined in the relevant ILO reso-

lution. This is because special surveys are

required, which tend to be costly and burden-

some, particularly for employers. Although

guidelines are given to ILO constituents with

regard to the type of information to be compiled

and published, ILO information on average

labour costs in manufacturing is derived from

various sources. It is expressed in different time

122

KILM 15 Wages and compensation costs

link wage data to other data expressed in

national currency. It also takes account of the

fact that wage levels may not be strictly compa-

rable across countries due to methodological

differences, while growth rates are less likely to

be affected by statistical effects.

Table 15a shows average monthly wage series

from the ILO’s Global Wage Database that were

compiled for the latest edition of the Global

Wage Report on the basis of official, national

sources. The series referring to real average

monthly wages are generally taken directly from

the national statistical office if available.

Otherwise, nominal values are collected and

deflated by the International Monetary Fund

(IMF) CPI. In cases where neither real values nor

the IMF CPI are available, data on the CPI are

collected directly from national sources. Real

wages are standardized to a common base year,

namely the base year that individual countries

use as the CPI base year.

Nominal average monthly wages are based

on a variety of national sources, as published by

national statistical agencies. In an ideal case,

the indicator refers to monthly average wages

in the sense of “earnings” (as defined by the

12th ICLS; see box 15c)

for the entire econ-

omy and all employees in a given country.

However, countries use different approaches

when collecting wage data. Methodological

differences relate to the type of source used,

the coverage of the source, and how the data

are aggregated to produce monthly average

wages. When data for the target concept were

not available, closely related wage series were

used instead (for details refer to “limitations to

comparability”).

The most common source for wage data – in

particular in advanced economies, in Central

and South-Eastern Europe and the CIS coun-

tries – are labour-related establishment surveys.

They collect data at the source, namely from

establishments that employ workers. Since

establishments usually keep accurate records of

all wages paid for their own book-keeping and

for tax purposes, this approach has the advan-

tage of producing reliable wage data without

having to rely on the memory of individual

employees. However, in countries where enter-

In most cases, data on CPIs from the IMF’s World

Economic Outlook are used. The base year information for

individual countries can be found in the IMF metadata.

Resolution concerning an integrated system of

wages statistics, adopted by the 12th International Confer-

ence of Labour Statisticians, Geneva, 1973; http://www.ilo.

org/public/english/bureau/stat/download/res/wages.pdf.

compilation of labour cost statistics according

to sex.

Definitions and sources

The annual average wages (table 15a) are

presented in both nominal and real terms. The

series of wage statistics are generally available

in nominal terms, expressed in absolute figures

and in national currency. This reflects the way

these data are collected, usually from those

who pay wages (enterprises) or from those who

receive them (paid employees). Wage statistics

in nominal terms (and in national currency) are

required by policy-makers, who set minimum

wages in nominal terms, or by employers and

trade unions, who bargain over nominal wage

rates. Other data users also need nominal wage

data, for example if they want to compare wage

levels to other indicators that are available in

nominal form (such as poverty thresholds or

prices of goods), or if they want to convert

them from one currency to another.

However, changes in nominal average wages

are not necessarily very informative when it

comes to assessing changes in the welfare of

wage earners: they indicate only the earnings

of an average employee in monetary terms, but

not the amount of goods and services that can

be purchased with wages. In other words,

nominal wages do not provide information on

the purchasing power of employees. This

purchasing power is influenced by, among

other factors, increases (or decreases) in prices

of goods and services that employees acquire,

use, or pay for – i.e. by the inflation rate.

Average monthly wages are therefore not only

presented in nominal terms, but also in real

terms by adjusting for changes in consumer

prices. Note, however, that the consumer price

index (CPI) reflects price changes as viewed

from the perspective of the average consumer

and that some wage earners might experience

a different rate of price changes (for example,

when they spend a higher proportion of their

income on food items than the average

consumer).

Both the nominal and real average wage

series are presented in national currency. This

enables data users to calculate nominal and

real wage growth rates without distortion

caused by exchange rate fluctuations, and to

For some purposes, other price measures such as the

producer price index (PPI) or implicit GDP deflator might

be more appropriate.

123

KILM 15 Wages and compensation costs

obtain estimates of earnings based on “hours

actually worked”.

Adjustment factors are obtained from various

sources, such as periodic labour cost surveys

(interpolated on the basis of other information

for non-survey years), annual tabulations of

employers’ social security contribution rates,

and information on contractual and legislated

changes in fringe benefits. The statistics are

further adjusted, where necessary, to take

account of major differences in workers’ cover-

age, industrial classification systems and changes

over time in survey coverage or frequency.

A country’s compensation costs are

computed in national currency units and

converted into US dollars using the average

daily exchange rate as published by either the

US Federal Reserve Board or the IMF. For euro

area countries, data are converted to US dollars

using the euro to dollar exchange rate only for

years in which the euro was officially currency

in those countries. For years prior to adoption

of the euro, the data in the old national currency

for all years are converted into US dollars using

historical US dollar to national currency

exchange rates or fixed exchange rates estab-

lished at the time of the country’s conversion

to the euro.

The hourly compensation measures relate to

manufacturing as defined by the North American

Industry Classification System (NAICS). NAICS

is the common industrial classification used by

the United States, Canada, and Mexico. The

NAICS definition of manufacturing differs some-

what from the definition of manufacturing used

in other countries. In such cases, the Bureau of

Labor Statistics makes adjustments to ensure

comparability across the series.

The following definitions apply to the data

series in table 15b:

Total hourly compensation costs include

(1) total hourly direct pay, (2) employer social

insurance expenditures and (3) labour-related

taxes.

Total hourly direct pay includes all payments

made directly to the worker, before payroll

deductions of any kind, consisting of pay for

time worked and directly paid benefits. This

definition is the equivalent of the ILO concept

of “gross earnings”, which consists of (a) pay for

time worked, including basic time and piece

rates, overtime premiums, shift differentials,

other premiums and bonuses paid regularly

each pay period, and cost-of-living adjustments,

prises routinely pay wages outside their normal

book-keeping (so-called “envelope wages”) in

order to avoid taxes and social security contri-

butions, the establishment-based approach has

limitations.

Household surveys, the second major

source for wage data, have the advantage that

they cover all employees regardless of where

they work.

Wage data from household surveys

usually cover the public and private sectors,

formal and informal enterprises and all indus-

trial sectors. There are, however, a number of

subtle methodological differences that can

affect comparability between countries of wage

levels based on household surveys (for details

refer to “limitations to comparability”).

Finally, a few countries rely on administra-

tive data sources such as social security records

to compile wage data, or combine several

different primary sources to produce a synthetic

wage series. In some countries the national

accounts sections of central statistical offices

produce the wage series that match the desired

concept most closely. However, national

accounts are only a useful source for data on

average wages when compensation of employ-

ees is disaggregated into its two major compo-

nents – wages and salaries and employers’

social contributions – and when matching data

on total wage employment exist.

While most countries report wages with a

calendar month as a reference period, some

report only daily, weekly or annual wages. In

order to ensure comparability, these source

data were standardized into the same monthly

reference period, e.g. annual wages were

divided by 12 months to produce average

monthly wages.

Hourly compensation costs for employees

in manufacturing (table 15b) are estimates

from The Conference Board based on national

statistics from establishment and labour cost

surveys. Earnings statistics are obtained from

country-specific surveys of employment, hours

and earnings, or from manufacturing surveys or

censuses. Total compensation is computed by

adjusting each country’s average earnings

series for items of total hourly direct pay, social

insurance and labour-related taxes and subsi-

dies not included in earnings. Where countries

measure earnings on the basis of “hours paid

for”, the figures are also adjusted in order to

Persons living in institutional households, such as

military barracks, prisons or monasteries, are commonly

excluded.

124

KILM 15 Wages and compensation costs

also limit coverage to the private sector (i.e.

exclude the public sector) or to specific indus-

tries within the private sector (such as manufac-

turing). If small enterprises pay lower wages

than large enterprises or wages differ between

the public and the private sector, these exclu-

sions will affect the level of the collected wage

data – depending on how large differences are,

and how many employees are excluded from

the coverage. However, if wages in the excluded

establishment move roughly in line with those

enterprises for which data are available, these

exclusions will only have a marginal effect on

trends over time. Even data with less than full

coverage can therefore be a useful proxy in

analysing wage growth in an economy.

Establishment surveys usually draw their

samples from an establishment register that is

maintained either by the central statistical office

or another institution, such as the registrar of

companies. In developing countries with a

large informal sector, this is a serious limitation

since many small, unregistered establishments

are missing from the sample frame. Also

excluded are individual households employing

paid domestic workers, which account for a

significant proportion of total paid employ-

ment in some developing regions.

In some

developing countries, establishment surveys

therefore capture only a small proportion of all

wage employees (those in the public sector and

those in large, modern enterprises). Under

these circumstances, collecting information

from the recipients of wages can be the better

alternative.

Household surveys encompass a greater

range of jobs and workers than establishment

surveys. However, they tend to experience

problems associated with self-reporting of earn-

ings. Furthermore, household surveys display

methodological differences that can affect

comparability. For instance, some surveys

collect data on the usual monthly wages while

others ask for the actual wage received in the

past month. At times it is also not clear whether

respondents are asked to report their gross or

net wages (i.e. before or after deduction of

taxes and compulsory social security contribu-

tions). These differences can have a material

effect on the reported level of wages, while they

are less likely to have a major impact on trends

According to ILO estimates, the global share of

domestic workers in paid employment was 3.6 per cent in

2010, but it reached 11.9 per cent in Latin America and the

Caribbean and 8.0 per cent in the Middle East. See ILO:

Global and regional estimates on domestic workers,

Domestic Work Policy Brief No. 4 (Geneva, 2011).

and (b) other direct pay, such as pay for time

not worked (vacations, annual holidays and

other paid leave for personal or family reasons,

civic duties, and so on, except sick leave),

seasonal or irregular bonuses and other special

payments, selected social allowances and the

cost of payments in kind.

Social insurance expenditures refer to the

value of social contributions (legally required

as well as private and contractual) incurred by

employers in order to secure entitlement to

social benefits for their employees; these contri-

butions often provide delayed, future income

and benefits to employees.

Labour-related taxes refer to taxes on pay-

rolls or employment (or reductions to reflect

subsidies), even if they do not finance program-

mes that directly benefit workers.

All employees include production workers

as well as all others employed full or part time

in an establishment during a specified payroll

period. Temporary employees are included.

Persons are considered employed if they receive

pay for any part of the specified pay period. The

self-employed, unpaid family workers and

workers in private households are excluded.

Limitations to comparability

As mentioned in the preceding section,

country-specific practices differ with respect to

the sources and methods used for wage data

collection and compilation, which in turn have

an influence on the results and comparability

across countries. The main sources of informa-

tion (establishment censuses and surveys, and

household surveys) usually differ in terms of

objectives, scope, collection and measurement

methods, survey methodology and so on. The

scope of the information may vary in terms of

geographical coverage, workers’ coverage (for

example, exclusion of part-time workers)

and

establishment and enterprise coverage (based

on establishment size or sector covered).

While most countries include firms regard-

less of size in establishment surveys, some

countries exclude small firms with less than five

or less than ten employees. Some countries

It should be noted here that wage series covering all

persons employed should not be directly compared with

series covering employees only, since a bias may be intro-

duced with the inclusion of working proprietors and

contributing family members.

125

KILM 15 Wages and compensation costs

when used alone, may be misleading. It is also

important to remember that this indicator

measures compensation of employees specific

to manufacturing and is significant only in so far

as countries strive to compete in the manufac-

turing sector. However, when used in conjunc-

tion with other indicators, such as labour

productivity (KILM 16), relative changes can be

helpful in assessing trends in competitiveness.

Care should also be taken not to interpret

hourly compensation costs as the equivalent of

the purchasing power of worker incomes, for

two reasons. The first relates to the compo-

nents and nature of compensation costs. In

addition to the payments made directly to

workers, compensation includes employers’

payments to funds for the benefit of workers.

Such “non-direct pay” can include current

social security benefits such as family or depen-

dants’ allowances, deferred benefits, as in

payments to retirement and pension funds, or

various types of insurance entitlements, such as

unemployment and health benefit funds, which

will represent income to workers only under

certain conditions. In a few countries, non-

wage costs also include some taxes paid by

employers – or deductions for subsidies

received – for the employment of labour, such

as taxes on employment or payroll.

The second reason for differentiating hourly

compensation costs from the concept of work-

ers’ purchasing power lies in the fact that the

prices of goods and services vary greatly among

countries, and the commercial exchange rates

used here to convert national figures into a

single currency do not indicate relative differ-

ences in prices. A more meaningful interna-

tional comparison of the relative purchasing

power of workers’ income would involve the

use of purchasing power parities (PPPs), that is,

rates at which the currency of one country must

be converted into the currency of another

in order to buy an equivalent basket of goods

and services.

In spite of the various adjustments made to

the series of hourly compensation costs of

employees in manufacturing (table 15b) in

order to ensure a high level of comparability

across countries and over time, differences may

still be found in the information presented. The

average earnings series used as a basis for these

estimates may be influenced by changes over

time in the industrial structure, that is, the

growth or decline of establishments, levels of

activity and changes in the structure of the

workforce employed (changes in the relative

proportions of men and women, skilled and

over time as long as the survey instrument

remains unchanged.

Even when using the same concept of wages

(for example, earnings), there are likely to be

differences with regard to the inclusion or

exclusion of various components (such as peri-

odic bonuses and allowances, or payments in

kind). Earnings statistics show fluctuations that

reflect the influence of both changes in wage

rates and supplementary payments. In addi-

tion, daily, weekly and monthly earnings are

dependent on variations in hours of work (in

particular, hours of paid overtime or short-time

working), while hourly earnings are influenced

by the concept of hours of work – hours actu-

ally worked, hours paid for, or normal hours of

work – used in the computation (see KILM 7 for

information on the various concepts pertaining

to hours of work).

When making comparisons of real wage

trends between countries, one should keep in

mind that this indicator is not only based on

country-specific series of wages, but also that

measures of real wages will be affected by the

choice of the price deflator, that is, the CPI. The

scope of CPIs can vary not only in terms of the

types of household or population groups

covered, but also in terms of the geographical

coverage. Country-specific practices also differ

regarding the treatment of certain issues relat-

ing to the computation of CPIs, including the

treatment of seasonal items, new products and

quality changes, durable goods and owner-

occupied housing, the inclusion or exclusion of

financial services and indirect taxes, and so on.

Other factors may influence the comparabil-

ity of real wage trends – and therefore purchas-

ing power – across countries. One is the

reference period of both wages and CPIs. Annual

averages of hourly or monthly wages may be

averages of information based on weekly,

monthly or quarterly reference periods. In some

cases, they are based on the whole calendar or

financial year. On the other hand, the CPI data

are annual averages of an index that is compiled,

in most cases, monthly, or in a few cases quar-

terly or biannually. When nominal wages and

CPI information do not refer to exactly the same

period, this can give rise to problems for coun-

tries experiencing rapid inflation.

When using the information presented in

table 15b on hourly compensation costs to

make comparisons of international competitive-

ness, it should be borne in mind that differ-

ences in hourly compensation costs are only

one factor in competitiveness and therefore,

126

KILM 15 Wages and compensation costs

tials. Furthermore, changes over time in rela-

tive compensation cost levels in US dollars are

also affected by (a) the differences in underly-

ing national wage and benefit trends measured

in national currencies, and (b) frequent and

sometimes sharp changes in relative currency

exchange rates.

Further to limitations to comparability for

each of the indicators, there are also limitations

concerning comparisons between the indica-

tors. Making comparisons of wage rates, earn-

ings or labour costs over time and between

countries is probably one of the most difficult

tasks for the users of the information presented

in this publication. Users should, in particular,

be aware of the following issues:

(1) Within each of the indicators, the infor-

mation may be affected by differences in

sources; that is, there may not be a close corre-

spondence between the concepts and defini-

tions used, the scope and coverage, the

methods used for compilation, and the ways in

which the information is presented. Table 15a

is based on unadjusted national data that reflect

these differences. A number of adjustments

unskilled labour, full-time and part-time work-

ers, and so on). All these factors influence the

levels of earnings and workers’ benefits within

a country.

Hourly compensation costs are partly esti-

mated, and each year the most recent informa-

tion is subject to revision by The Conference

Board. For example, in 2001 the hourly

compensation costs series were revised for the

United States from 1997 onwards to incorpo-

rate results on non-wage costs from an annual

survey of manufacturers. In 2006, data for

Mexico were revised back to 1999 to incorpo-

rate benchmark data from an industrial census

and data for Ireland and Norway were revised

back to 2001 to incorporate non-wage compen-

sation costs from the 2004 labour cost surveys.

The comparative-level figures are averages

for all manufacturing industries and are not

necessarily representative of all component

industries. In some countries, such as the

United States and Japan, differentials in hourly

compensation cost levels by industry group are

quite wide, while other countries, such as

Germany and Sweden, have narrower differen-

Box 15a. The ILO’s Global Wage Report

The biennial Global Wage Report is the ILO’s flagship publication on wage trends and wage policies.

It uses a number of indicators to analyse global wage developments, including the growth of average

real wages, the low-pay incidence (defined as the share of wage workers with earnings below two-thirds

of the median) and the wage share in national income. It is unique in its global scope and builds on

data from 130 countries and territories (in its last edition) that between them account for approximately

95.8 per cent of the world’s wage workers. Based on a standard methodology that corrects for the

remaining response bias, the report documents wage growth for the world and in seven regions.

The report also provides practical illustrations of how collective bargaining, minimum wages and

income policies can be building blocks of effective wage policies that contribute to equitable

outcomes. It is inspired by the objective to promote “policies in regard to wages and earnings, hours

and other conditions of work, designed to ensure a just share of the fruits of progress to all and a

minimum living wage to all employed and in need of such protection”, one of the central elements of

the Decent Work Agenda (see ILO Declaration on Social Justice for a Fair Globalization). The relevance

of this approach has been underscored by the global economic crisis, during which many governments

expanded income support policies and wage subsidies in order to stabilize domestic demand and

to support recovery.

Further information is available from the Inclusive Labour Markets, Labour Relations and Working

Conditions Branch (INWORK), the ILO’s lead programme on wage data analysis and policy advice,

or online at http://www.ilo.org/travail/lang--en/index.htm.The econometric model developed in the

paper utilizes available national household survey-based estimates of the distribution of employment

by economic class, augmented by a larger set of estimates of the total population distribution by

class together with key labour market, macroeconomic and demographic indicators. The output of

the model is a complete panel of national estimates and projections of employment by economic

class for 142 developing countries, which serve as the basis for the production of regional aggregates.

Source: Kapsos, S.; Bourmpoula, E. (2013). Employment and economic class in the developing world, ILO Research Paper No. 6;

available at: http://www.ilo.org/wcmsp5/groups/public/---dgreports/---inst/documents/publication/wcms_216451.pdf.

127

KILM 15 Wages and compensation costs

Box 15b. Resolution concerning statistics of labour cost,

adopted by the 11th International Conference of Labour

Statisticians, October 1966 [relevant paragraphs]

The 11th ICLS (Geneva, 1966) adopted a resolution concerning statistics on labour cost, recommending

the following International Standard Classification of Labour Cost:

I. Direct wages and salaries

1. Straight-time pay of time-rated workers

2. Incentive pay of time-rated workers

3. Earnings of piece-workers (excluding overtime premiums)

4. Premium pay for overtime, late shift and holiday work

II. Remuneration for time not worked

1. Annual vacation, other paid leave, including long-service leave

2. Public holidays and other recognized holidays

3. Other time off granted with pay (e.g. birth or death of family members, marriage of employ-

ees, functions of titular office, union activities)

4. Severance and termination pay where not regarded as social security expenditure

III. Bonuses and gratuities

1. Year-end and seasonal bonuses

2. Profit-sharing bonuses

3. Additional payments in respect of vacation, supplementary to normal vacation pay and

other bonuses and gratuities

IV. Food, drink, fuel and other payments in kind

V. Cost of workers’ housing borne by employers

1. Cost for establishment-owned dwellings

2. Cost for dwellings not establishment-owned (allowances, grants, etc.)

3. Other housing costs

VI. Employers’ social security expenditure

1. Statutory social security contributions (for schemes covering old age, invalidity and survi-

vors; sickness; maternity; employment injury; unemployment; and family allowances)

2. Collectively agreed, contractual and non-obligatory contributions to private social security

schemes and insurances (for schemes covering: old age; invalidity and survivors; sickness;

maternity; employment injury; unemployment and family allowances)

3a. Direct payments to employees in respect of absence from work due to sickness, maternity

or employment injury, to compensate for loss of earnings

3b. Other direct payments to employees regarded as social security benefits

4. Cost of medical care and health services

5. Severance and termination pay where regarded as social security expenditure

VII.

Cost of vocational training, including also fees and other payments for services of outside

instructors, training institutions, teaching material, reimbursements of school fees to workers, etc.

VIII. Cost of welfare services

1. Cost of canteens and other food services

2. Cost of education, cultural, recreational and related facilities and services

3. Grants to credit unions and cost of related services for employees

128

KILM 15 Wages and compensation costs

benefits, the relative contributions of employ-

ers, employees and the State to such schemes,

and so on.

(3) Finally, it should be noted that the

series presented in table 15a show the trends

in real and nominal monthly wages based on

information expressed in national currency,

while table 15b shows the levels and trends of

hourly compensation costs and their structure

in US dollars. In the first indicator, account has

been taken of changes in the CPI in each coun-

try, while in table 15b, to produce a real series

in addition to the nominal series, nominal

national data have been converted into US

dollars and are thus affected by variations, over

have been made in table 15b by The Conference

Board to ensure a high level of comparability

between countries; however, some disparities

may still exist. Users should take account of the

notes to the tables for each indicator.

(2) Care should be taken when comparing

trends in annual average wages and hourly

compensation costs for the same countries. It

should be noted that wages and total compen-

sation costs are not substitutes for each other.

The difference between the two may be

affected by factors such as the rapid growth (or

the freeze) of nominal wages and the develop-

ment of non-wage benefits, changes over time

in the nature of social security schemes and

IX. Labour cost not elsewhere classified, such as costs of transport of workers to and from work

undertaken by employer (including reimbursement of fares, etc.), cost of work clothes, cost of

recruitment and other labour costs

X. Taxes regarded as labour cost, such as taxes on employment or payrolls, included on a net

basis, i.e. after deduction of allowances or rebates made by the State.

(Box 15b continued)

Box 15c. Resolution concerning an integrated system

of wages statistics, adopted by the 12th International

Conference of Labour Statisticians, October 1973

[relevant paragraphs]

8. The concept of earnings, as applied in wages statistics, relates to remuneration in cash and in

kind paid to employees, as a rule at regular intervals, for time worked or work done, together with

remuneration for time not worked, such as for annual vacation, other paid leave or holidays. Earnings

exclude employers’ contributions in respect of their employees paid to social security and pension

schemes and also the benefits received by employees under these schemes. Earnings also exclude

severance and termination pay.

9. Statistics of earnings should relate to employees’ gross remuneration, i.e. the total before any

deductions are made by the employer in respect of taxes, contributions of employees to social security

and pension schemes, life insurance premiums, union dues and other obligations of employees.

10. (i) Earnings should include: direct wages and salaries, remuneration for time not worked (excluding

severance and termination pay), bonuses and gratuities and housing and family allowances paid by

the employer directly to his employees.

(a) Direct wages and salaries for time worked, or work done, cover: (i) straight-time pay of time-rated

workers; (ii) incentive pay of time-rated workers; (iii) earnings of piece-workers (excluding over-

time premiums); (iv) premium pay for overtime, shift, night and holiday work; (v) commissions paid

to sales and other personnel. Included are: premiums for seniority and special skills, geographical

zone differentials, responsibility premiums, dirt, danger and discomfort allowances, payments

under guaranteed wage systems, cost-of-living allowances and other regular allowances.

(b) Remuneration for time not worked comprises direct payments to employees in respect of public

holidays, annual vacations and other time off with pay granted by the employer.

of vacation period (supplementary to normal pay) and profit-sharing bonuses.

(ii) Statistics of earnings should distinguish cash earnings from payments in kind.

129

KILM 15 Wages and compensation costs

time and between countries, in the US dollar

exchange rates.

In spite of these comparability issues, which

are inherent in the underlying statistical series,

every effort has been made to choose informa-

tion that is as close as possible to the target

concept and thus comparable across countries.

As long as users are alert to these issues, the

wage and labour compensation indicators

presented can provide valuable insights for

socio-economic analyses.

Introduction

This chapter presents information on labour

productivity for the aggregate economy with

labour productivity defined as output per unit of

labour input (persons engaged or hours worked).

Labour productivity measures the efficiency with

which inputs are used in an economy to produce

goods and services and it offers a measure of

economic growth, competitiveness, and living

standards within a country.

Use of the indicator

Economic growth in a country can be

ascribed either to increased employment or to

more effective work by those who are employed.

The latter effect can be described through

statistics on labour productivity. Labour produc-

tivity therefore is a key measure of economic

performance. The understanding of the driving

forces behind it, in particular the accumulation

of machinery and equipment, improvements in

organization as well as physical and institu-

tional infrastructures, improved health and

skills of workers (“human capital”) and the

generation of new technology, is important for

formulating policies to support economic

growth. Such policies may focus on regulations

on industries and trade, institutional innova-

tions, government investment programmes in

infrastructure as well as human capital, tech-

nology or any combination of these.

Labour productivity estimates can support

the formulation of labour market policies and

monitor their effects. For example, high labour

productivity is often associated with high levels

or particular types of human capital, indicating

priorities for specific education and training pol-

icies. Likewise, trends in productivity estimates

can be used to understand the effects of wage

settlements on rates of inflation or to ensure that

such settlements will compensate workers for

(part of) realized productivity improvements.

Finally, productivity measures can contribute

to the understanding of how labour market

performance affects living standards. When the

intensity of labour utilization – the average

number of annual working hours per head of the

population – is low, the creation of employment

opportunities is an important means of raising

per capita income in addition to productivity

growth.

In Europe, for example, with productiv-

ity levels relatively close to those of the United

States but with lower per capita income levels,

living standards can be improved by increasing

labour utilization. This can be achieved by

encouraging a higher labour force participation

rate or by encouraging workers to work more

hours, e.g. by creating more decent and produc-

tive employment opportunities for economic

activity. In contrast, when labour intensity is

already high, for example in East Asia, increasing

productivity is essential to improving living stand-

ards. In any case, increasing labour force partici-

pation is at best a transitional source of growth

depending on the rate of population growth and

the age structure of the population. In the long

run, it is the productivity of labour which deter-

mines the rise in per capita income.

Definitions and sources

Productivity represents the amount of

output per unit of input. In KILM 16, output is

measured as gross domestic product (GDP) for

the aggregate economy expressed at purchas-

ing power parities (PPPs) to account for price

differences in countries; as well as at market

exchange rates for table 16a, which reflect the

market value of the output produced.

Labour productivity growth may be due to

either increased efficiency in the use of labour,

without more of other inputs, or because each

worker works with more of the other inputs,

such as physical capital, human capital or inter-

mediate inputs. More sophisticated measures,

such as “total factor productivity”, which is the

It is clear that living standards do not equal per capita

income, but the latter can still be viewed as a reasonably

good proxy of the former, even though the link is not auto-

matic. For example, the United Nations Development

Programme (UNDP) Human Development Report 2014

reveals that, out of 186 economies with information on

both the human development index (HDI) and GDP per

capita in 2012, 107 rank higher in HDI than in GDP, two

rank the same and 77 rank higher in GDP than in HDI.

KILM 16. Labour productivity

132

KILM 16 Labour productivity

In table 16b, GDP estimates for OECD coun-

tries after 1990, are mostly obtained from the

OECD National Accounts, Volumes I and II

(annual issues) and the Eurostat New Cronos

database. The series up to 1990 are mostly

derived from Maddison (1995).

To compute labour productivity per person

engaged in table 16b, GDP is divided by total

employment. These employment estimates are

primarily taken from OECD: Labour Force

Statistics (annual issues); Eurostat’s New

Cronos database; the ILO estimates on employ-

ment; and the Vienna Institute for Comparative

Economic Studies (WIIW). To compute labour

productivity per hour worked, estimates on

annual hours worked are based on a variety of

sources deemed to be most appropriate source

of the preferred concept of “actual hours

worked per person employed” in each individ-

ual country. National sources are used as well as

collections such as that of the OECD Growth

Project, which are updated by Scarpetta et al.

(2000).

In later years, the trend of the OECD

Employment Outlook has been used. Full

details on sources used for each variable – GDP,

employment and hours – are available on the

Total Economy Database website and displayed

in the notes sections of the KILM data tables.

For countries outside of the OECD, the

national accounts and labour statistics which

were assembled from national sources by inter-

national organizations such as the World Bank,

the Asian Development Bank, the Food and

Agriculture Organization (FAO), the ILO and the

United Nations Statistical Office were used as the

point of departure.

These series were comple-

mented by the series from Maddison (1995) in

particular to cover the period 1980–90. Maddison

(1995) also provides benchmark estimates of

annual hours worked for a significant number

A. Maddison, database on “Historical statistics”,

available at Maddison’s homepage: http://www.ggdc.net/

maddison/.

See Scarpetta, S. et al.: Economic growth in the OECD

area: Recent trends at the aggregate and sectoral level,

Economics Department Working Papers, No. 248 (Paris,

OECD, 2003), Table A.13.

World Bank: World Development Indicators (various

issues); Asian Development Bank: Key Indicators of Devel-

oping Asian and Pacific Countries (annual issues); ILO:

Yearbook of Labour Statistics (annual issues); United

Nations: National Account Statistics: Main Aggregates and

Detailed Tables (annual issues).

output per combined unit of all inputs, are not

included in KILM 16.

Estimated labour produc-

tivity may also show an increase if the mix of

activities in the economy or in an industry has

shifted from activities with low levels of produc-

tivity to activities with higher levels, even if

none of the activities have become more

productive by themselves.

For a constant “mix” of activities, the best

measure of labour input to be used in the

productivity equation would be “total number

of annual hours actually worked by all persons

employed”. In many cases, however, this labour

input measure is difficult to obtain or to esti-

mate reliably. For this reason, two series for

labour productivity are shown in table 16b,

GDP per person engaged and GDP per hour

worked; and one series in table 16a, GDP per

worker.

To compare labour productivity levels across

economies, it is necessary to convert output to

US dollars on the basis of purchasing power

parity (PPP). A PPP represents the amount of a

country’s currency that is required to purchase

a standard set of goods and services worth one

US dollar. Through the use of PPPs one takes

account of differences in relative prices between

countries. Had official currency exchange rates

been used instead, the implicit assumption

would be that there are no differences in rela-

tive prices across countries. The labour product-

ivity estimates in table 16b are expressed in

terms of 1990 US dollars converted at PPPs (as

the 1990 PPP made it possible to compare the

largest set of countries – see details below) and

in table 16a in terms of 2005 international

dollars converted at PPPs as well as constant

2005 US dollars.

The labour productivity estimates in table

15b are derived from the Total Economy

Database of The Conference Board and are

available for 123 economies. This database also

includes measures of labour compensation to

obtain unit labour cost. A full documentation

of sources and methods by country and under-

lying documentation on the use of PPPs, etc.

can be downloaded from the database website.

For recent estimates of total factor productivity

growth, please refer to The Conference Board Total Econ-

omy Database™, May 2015, available at: http://www.confer-

ence-board.org/data/economydatabase/. Estimates at the

industry level can be obtained from the EU KLEMS Growth

and Productivity Accounts (http://www.euklems.net).

The Total Economy Database is maintained at: http://

www.conference-board.org/data/productivity.cfm. The

database was previously housed at the Groningen Growth

and Development Centre of the University of Groningen,

(Note 3 continued)

Netherlands. This research centre still undertakes research

on comparative analysis of levels of economic performance

and differences in growth rates. See http://www.ggdc.net/

index.htm for the latest publications.

133

KILM 16 Labour productivity

However, despite common principles that are

mostly based on the United Nations System of

National Accounts, there are still significant

problems in international consistency of national

accounts estimates, in particular for economies

outside the OECD. Such factors include:

(a) different treatment of output in services

sectors. In a considerable number of econ-

omies, especially for non-market services,

output is often estimated on the basis of

inputs, such as total labour compensation, or

on an implicit assumption concerning produc-

tivity growth; in other cases – where output

measures are available – quality changes are

often insufficiently reflected in the measures

of output volume.

(b)

different procedures in correcting output

measures for price changes, in particular

the use of different weighting systems in

obtaining deflators. Traditionally output

trends in constant prices have been weighted

at values that are kept fixed for several years.

Fixed weights usually imply an overestima-

tion of volume growth rates, creating a bias

that increases the further one moves away

from the base year. Most economies therefore

change weights every five or ten years. Over

the past year an increasing number of OECD

countries have been shifting to using annual

chain weights.

Another important source of

methodological difference between coun-

tries is the use of deflators for information

and communication technology (ICT) prod-

ucts. Price declines of these goods are often

insufficiently chosen with traditional price

measurement methods. The United States has

introduced a range of hedonic price deflators

for ICT goods, which measure the price

change of a commodity on the basis of

changes in the major characteristics that have

an impact on the price. Many other countries

are introducing this type of price measure in

their national accounts, but at a much slower

pace than the United States. In the estimates

for the manufacturing sector the latter prob-

lem has been tackled by using harmonized

deflators for ICT industries, based on hedonic

deflators for the United States, for those coun-

tries that have no adequate ICT deflator

themselves.

economic activities in developing econ -

omies and of the underground economy in

The method of using chain weights allows for the

use of different weights at different segments of a time

series which are then “chained together”.

of non-OECD economies.

In some cases, use

has also been made of national accounts statistics

for individual countries.

Whenever data for employment are unavail-

able, The Conference Board supplements em-

ployment data with data on the total labour

force, which happens in about one third of all

cases – primarily in developing countries. Since

labour force is not necessarily a sufficient proxy

for employment, indicators on labour produc-

tivity by The Conference Board (table 16b) are

supplemented with a table on labour productiv-

ity (16a), utilizing employment data from the

ILO Trends Econometric Models (see KILM 2).

Labour productivity in table 16a is calcu-

lated using data on GDP in constant 2005 inter-

national dollars in PPPs, derived from the

World Development Indicators database of the

World Bank.

To compute labour productivity

as GDP per person engaged, ILO estimates for

total employment are used.

Countries for

which no real data on employment exist

(meaning that all data points are estimates

rather than reported data) in and after the year

2000 were excluded. Furthermore, table 16a is

complemented by a series of GDP at market

exchange rates (rather than PPPs) to get a

better idea of labour productivity estimates

when used for the purpose of competitiveness

indicators. GDP figures (at constant 2005 US

dollars) are also derived from the World

Development Indicators database. Table 16a is

available for 140 economies with coverage

extending to all KILM regional groupings.

Limitations to comparability

The limitations to the international and

historical comparability of the estimates are

summarized under the following headings:

Output measures in national currencies,

employment, and working hours.

Output measures in national currencies

Output measures are obtained from national

accounts and represent, as much as possible,

GDP at market prices for the aggregate economy.

Maddison, A.: Monitoring the World Economy 1820-

1992 (Paris, OECD Development Centre, 1995).

For more detail, please refer to the website of the

World Development Indicators database at: http://data.

worldbank.org/data-catalog/world-development-indicators.

For more details, please see KILM 2a.

134

KILM 16 Labour productivity

force estimates include a substantial part of

(part-time and seasonal) family workers.

However, the estimates presented for the econ-

omies in this data set are meant to cover all

economic activity. Furthermore, limitations to

comparability of ILO employment estimates

discussed in KILM 2 apply.

Working hours

Estimates of annual working hours are often

unavailable or are relatively unreliable. Even for

developed economies, annual working hours are

not consistently defined. For example, statistics

on working hours often refer to paid hours

rather than to hours actually worked, implying

that no adjustments are made for paid hours that

are not worked, such as hours for paid vacation

or sickness, or for hours worked that are not

paid for. Moreover, statistics on working hours

often are only available for a single category of

the workforce (in many cases, only employees),

or only for a particular industry (such as manu-

facturing), or for particular types of establish-

ment (for example, those above a certain size or

in the formal sector). As always, these problems

are particularly serious for a substantial number

of low-income economies. Whether and how the

estimates of annual hours worked have been

adjusted for such weaknesses in the primary

statistics are often undocumented.

Readers may wish to review the corresponding

section relating to comparability issue for working hours in

KILM 7.

developed (industrialized) economies in

national accounts. Some economies use data

from special surveys for “unregistered activi-

ties”, or indirect estimates from population

censuses or other sources to estimate these

activities, and large differences in coverage

between economies remain.

In addition to such inconsistencies there are

significant differences in scope and quality of

the primary national statistics and the staff

resources available for the preparation of the

relevant national estimates.

Employment

Estimates of employment are, as much as

possible, for the average number of persons

with one or more paid jobs during the year.

Particularly for low- and middle-income econ-

omies in Asia and Latin America, statistics on the

number of self-employed and family workers in

agricultural and informal manufacturing activ-

ities are probably less reliable than those for

paid employees. As in the case of output estim-

ates, the employment estimates are s ensitive to

under-coverage of informal or underground

activities, which harbour a substantial part of

labour input. In some cases, informal activities

are not included in the production and employ-

ment statistics at all. In agriculture the labour

For an overview of methods, see, for example,

OECD: Measuring the Non-Observed Economy. A Hand-

book (Paris, 2002).

because similar data do not exist for most high-

income economies, where extreme poverty is a

rarer occurrence. The Gini index is shown only

in those countries for which poverty information

is available; however, this statistic is also available

for many high-income economies from the origi-

nal data repository (the World Bank). Table 17b

contains estimates of the “working poor” –

defined as the proportion of employed persons

in a household whose members are living below

the US

1.90 international poverty line as well as

the full distribution of employment across five

economic classes.

Use of the indicator

The value of measures of poverty, the distri-

bution of workers across different economic

class groups and income inequality lies in the

information these indicators provide on the

outcome of economic processes at the national

level, as a reflection of the access of different

groups of people to goods and services. The

information relating to poverty shows the abso-

lute number and the proportion of the popula-

tion that has “unacceptably” low consumption

or income levels, while the employment by

economic class and inequality series show the

disparity between different groups of people

within a country in terms of consumption or

income levels. Measurements of poverty are

extremely important as an indication of the

well-being and living conditions of a popula-

tion. In addition, a poverty line helps focus the

attention of governments and civil society on

the living conditions of the people in poverty

and can be used to gauge the need to devise

public policies and programmes to reduce

poverty and enhance the welfare of individuals

within a society. Analysing information on

poverty over time, when comparable, is crucial

for monitoring any increase or decrease in the

incidence of poverty and can help in assessing

the results of poverty reduction programmes.

Any assessment of poverty can also contribute

Introduction

Tables 17a and 17b present two of the indi-

cators that were used for monitoring progress

towards the first Millennium Development

Goal (MDG), “eradicating extreme poverty and

hunger”, while the MDGs were in force. The

proportion of the population living below the

international poverty line was a selected indica-

tor under the first target (1a) of MDG 1 (on the

eradication of poverty), while the proportion of

employed persons living below the inter-

national poverty line (the “working poor”), was

an indicator selected for monitoring the MDG 1

second target (target 1b) on decent work.

With

the MDGs coming to an end in 2015,

17 Sustainable Development Goals (SDGs)

have been set to succeed them.

The first SDG

being that of “ending poverty in all its forms

everywhere”, an indicator on the population

living below the international poverty line was

kept as a measure of progress towards tar -

get 1.1. Tables 17a and 17b also present other

measures of economic well-being, including

the employed population living in different

economic class groups (denoted by different

per capita household consumption thresh-

olds), estimates of the population living below

nationally defined poverty lines and the Gini

index as a measure of the degree of inequality

in income distribution.

Information on poverty in tables 17a and 17b

relates almost entirely to developing economies

The first MDG included three targets and nine indica-

tors. See the official list at: http://mdgs.un.org/unsd/mdg/

Host.aspx?Content=Indicators/OfficialList.htm. The

remaining indicators under the target on decent work were

the growth rate of GDP per person engaged (i.e. labour

productivity growth; KILM 16), the employment-to-popu-

lation ratio (KILM 2) and the vulnerable employment rate

(KILM 3).

The official list of SDGs and their corresponding

targets (including for the first goal) can be found at: http://

www.un.org/sustainabledevelopment/sustainable-develop-

ment-goals/.

KILM 17.

Poverty, income

distribution, employment

by economic class

and working poverty

136

KILM 17 Poverty, income distribution, employment by economic class and working poverty

fostering an enabling environment in which the

employment opportunities and incomes of the

working poor are improved.

It is important to note that the poverty,

employment by economic class and inequality

measures presented here focus on only one

aspect of absolute and relative deprivation.

They concentrate on personal income or private

consumption and do not directly address depri-

vation related to other spheres, such as access

to health care, education, productive employ-

ment, and social and political participation. A

comprehensive analysis of poverty and inequal-

ity should include a link to these other dimen-

sions, which are captured at least partially in

some of the other KILM indicators.

Definitions and sources

Because of the multiple dimensions of

poverty, there are various theoretical conceptions

of its measurement. Three are described below:

1. One common approach is to analyse informa-

tion on monetary income or personal

consumption as opposed to human develop-

ment. The underlying information relates, in

most cases, to personal consumption expen-

diture and, in only a few cases, to personal

income. This is because obtaining information

on income from surveys can be difficult and

because such information may not fully reflect

the “real” living standard of households. A

drawback of measuring poverty in this

manner is that household surveys often vary

across countries and over time, thus reducing

the comparability of the information (see

“Limitations to comparability” below).

A key feature of using income or personal

consumption as measures of poverty is the

establishment of a poverty line, the predeter-

mined level of income or consumption below

which a person (or household) is considered

to be poor. The incidence of poverty is typic-

ally measured as the fraction of the population

whose consumption expenditure falls below

this predetermined level. Many countries have

adopted national income poverty lines, using

thresholds based on the amount of income

necessary to buy a specified quantity of food.

Measurement of poverty using internationally

comparable poverty lines is also useful

because it allows poverty estimates to be

developed on a global basis. The World Bank

has established two international poverty

to explaining its possible causes, an important

step in finding a solution.

During the 1990s, a decade characterized by

increased globalization and an increase in the

number of market-based economies, poverty was

increasingly recognized as a major challenge for

the international community. The first of the

MDGs

was to “eradicate extreme poverty and

hunger”, with the specific target of halving the

share of people in the world living on less than

1 a day between 1990 and 2015.

The fight

against poverty was kept and reinforced in the

successors to the MDGs, the SDGs, with the first

SDG being that of “ending poverty in all its forms

everywhere”. The corresponding specific target

is to achieve, by 2030, the eradication of extreme

poverty for all people everywhere.

While poverty in the developed world is

often associated with unemployment, the

extreme poverty that exists throughout much of

the developing world is largely a problem associ-

ated with persons who are working, which is

why the second target under MDG 1 was to

“achieve full and productive employment and

decent work for all, including women and young

people”. The majority of working-age people in

poverty must work in order to survive and

support their families in a context where no effi-

cient social protection schemes or social safety

nets exist. For these workers who live in poverty,

the problem is typically one of poor employ-

ment quality, including low wages or incomes

and low levels of labour productivity. Thus,

reducing overall poverty rates in line with the

former MDG and subsequent SDG necessitates

As part of the Millennium Declaration of the United

Nations “to create an environment – at the national and

global levels alike – which is conducive to development and

the elimination of poverty”, the international community

adopted a set of international goals for reducing income

poverty and improving human development. A framework

of eight goals, 21 targets and 60 indicators to measure pro-

gress was adopted by a group of experts from the United

Nations Secretariat, ILO, IMF, OECD and the World Bank.

The indicators are interrelated and represent a partnership

between developed and developing economies. For further

information on the MDGs, see: http://www.un.org/millen-

niumgoals/.

The MDG on poverty was expressed in terms of

shares. That is, the goal was to reduce by half the propor-

tion of people living on less than US

1 a day. Because

populations tend to rise over time, a falling share of the

poor population will not necessarily translate into a decline

in the actual number of poor people. US

1.25 was the

threshold for the international “

1 per day poverty line”.

The poverty line was then raised to US

1.90 in October

2015. It has been updated by the World Bank on the basis of

2005 price levels, and subsequently on the basis of 2011 price

levels, and new price data collected through the International

Comparison Program.

137

KILM 17 Poverty, income distribution, employment by economic class and working poverty

ment indicators.

The data sets included in

tables 17a involve the use of poverty lines, with

poverty rates calculated as the percentage of

the population living below the line. National

poverty lines are based on the World Bank’s

country poverty assessments, while interna-

tional poverty lines are based on tabulations

from nationally representative primary house-

hold surveys and published in the PovcalNet

database. Estimates of the Gini index are based

on national household surveys, supplemented

by the Luxembourg Income Study database for

high-income economies.

Employment by economic class estimates,

which provide the distribution of employment

across five household consumption-based

economic classes (see box 17), are also based

on nationally representative primary house-

hold surveys, but only those surveys that

include questions on employment status. In

order for an estimate of employment across

economic classes to be included in table 17b,

the definition of employment must be found to

be sufficiently in line with the international

definition of employment as provided in the

resolution adopted by the 19th International

Conference of Labour Statisticians (ICLS).

For

countries and years with available distribu-

tional data from the World Bank’s PovcalNet

database but for which no national employ-

ment by economic class estimate is available,

the employment by economic class estimates

are derived from an ILO econometric model

referenced in box 17.

The national, urban and rural poverty

lines are specific to each country. Several

factors may have influenced the choice of

poverty threshold, such as nutritional require-

ments, basic consumption needs or minimum

acceptable consumption levels. The population

National and international poverty data and the Gini

index were extracted from the World Bank, World Develop-

ment Indicators Online. Data on the population distribu-

tion across economic class thresholds were downloaded

from PovcalNet, an interactive web-based computational

tool managed by the World Bank that allows users to repli-

cate the calculations by the World Bank’s researchers in

estimating the extent of absolute poverty in the world.

PovcalNet is available online at: http://iresearch.worldbank.

org/povcalnet/. It is important to note that alternatives to

World Bank estimates of poverty do exist and the issue of

“best” poverty estimation is a topic of debate in the research

community. See, for example, the ILO study on alternative

estimates of poverty, Karshenas, M.: Global Poverty: New

National Accounts Consistent and Internationally Compa-

rable Poverty Estimates, ILO mimeo (Geneva, 2002).

For additional information regarding the Luxem-

bourg Income Study, see: http://www.lisproject.org/.

See the chapter on KILM 2 for further details on the

ICLS definition of employment.

lines, currently at US$1.90 and US$3.10 of

consumption per person a day.

A second perspective relies upon a “basic

needs” approach and reflects deprivation in

terms of material requirements for minimally

acceptable fulfilment of human needs, includ-

ing food and employment. The concept goes

beyond the lack of income because it takes

into account the need for basic health care

and education, as well as essential services

such as access to safe water. In addition to its

Human Development Index, the United

Nations Development Programme in 1997

introduced the concept of the Human Poverty

Index (HPI) for developing economies, and

later replaced it, in 2010, with the Multidimen-

sional Poverty Index (MPI).

The HPI is a

composite index that aims to capture the

extent of deprivation in human life, particu-

larly accounting for overlapping deprivations

suffered.

The third approach, which combines

elements of the two previous perspectives, is

related to the capabilities required for a

person to function in a particular society,

under the assumption that a minimally

acceptable level of such capabilities exists.

This approach covers a wide range of cap-

abilities, and can vary from the capability of

being well nourished in a low-income econ-

omy to more complex social achievements in

a high-income economy, such as the capabil-

ity of gaining computer literacy (on the

assumption that a person lacking computer

literacy is likely to face difficulties in entering

the labour market in a developed economy).

Poverty is defined in terms of being out of the

mainstream of a society, notably being outside

the labour market. Poverty analysis from this

angle has led to development of the concept

of “social exclusion”.

4. Finally, the Gini index is a well-known direct

measure of the degree of distributional

inequality in income or consumption. It looks

at the cumulative distribution of income or

consumption (represented by the Lorenz

curve) and estimates the extent to which it

deviates from perfect equality.

The data presented for national and inter-

national poverty lines and the Gini index were

obtained from the set of World Bank develop-

For more information on the MPI, see: http://hdr.

undp.org/en/content/multidimensional-poverty-index-mpi.

138

KILM 17 Poverty, income distribution, employment by economic class and working poverty

these two components side by side also

provides a more detailed perspective on the

dynamics of productive employment genera-

tion, poverty reduction and growth in the

middle class throughout the world.

Limitations to comparability

Cross-country comparisons should not be

made using national poverty lines, as these do

not reflect any single agreed-upon international

norm on poverty. However, when the focus is

narrowed to one country and the same poverty

line has been used consistently over time,

ana lyses of trends and patterns of poverty may

be informative and in many cases more useful

for national inferences than analysis of interna-

tional poverty lines.

At the country level, comparisons over time

may be affected by such factors as changes in

survey types or data collection procedures.

Both agricultural conditions and the occur-

rence of natural and economic disasters affect

poverty rates, and membership of the poor

group may change from year to year, as some

individuals climb out of poverty while others

fall into it.

In the case of estimates based on an inter-

national poverty line, the use of PPPs rather than

market exchange rates ensures that differences

in price levels across countries are taken into

account. However, it cannot be categorically

asserted that two people in two different coun-

tries, consuming at US$1.90 (or US$3.10) a day

at PPP, face the same degree of deprivation or

have the same degree of need. Apart from the

well-known problems in economics in making

interpersonal comparisons of welfare, there are

other problems, such as rural–urban price

differentials and differences in required calorie

intake due to climatic variations, which may or

may not have been taken into account. One esti-

mate may relate to consumption and the other

to income, and a daily income of US$1.90 (or

US$3.10) may permit less consumption than a

daily consumption expenditure of the same

amount. The adjustments that are often made

to convert income estimates into consumption

estimates can also impart bias to the resulting

consumption distributions. The extent of non-

market activity and the way in which non-market

production and consumption are valued could

substantially hamper comparability.

Even if measurements of poverty and

economic class groups using international

below country-specific poverty lines cannot

readily be compared between countries. Also,

over time, these poverty lines may have been

changed to take account of new developments

or new data, casting doubts on comparability

over time as well.

The international poverty lines use a sum

of money in constant US dollars, converted into a

sum of money for the country concerned using

purchasing power parity (PPP) conversion

factors rather than market exchange rates.

Taking the US

1.90 poverty line as an example,

this amount is converted into an equivalent

amount in the currency of the country in ques-

tion, using the PPP conversion factor. This

measure has the virtue of allowing comparisons

over space and time.

The third data set for the indicator, the Gini

index, is a convenient and widely used measure

of the degree of income inequality. It measures

the extent to which the distribution of income

(or, in some cases, consumption expenditure)

among individuals or households within a

country deviates from a perfectly equal distri-

bution. A Lorenz curve plots the cumulative

percentages of total income received against

the cumulative percentages of recipients, start-

ing with the poorest individual or household.

The Gini index measures the area between the

Lorenz curve and the hypothetical line of abso-

lute equality, expressed as a percentage of the

maximum area under the line.

The Gini index

has a value of zero for perfect equality of

incomes and 100 for perfect inequality. As with

all summary measures, it cannot fully capture

differences between countries and over time in

the cumulative share of different clusters (frac-

tals) of the population in income or consump-

tion, which is represented by the Lorenz curve.

Finally, the employment by economic

class estimates indicate individuals who are

employed and who fall within the per capita

consumption thresholds of a given economic

class group. By combining labour market char-

acteristics with household consumption group

data, employment by economic class estimates

give a clearer picture of the relationship

between economic status and employment.

Because of the important linkages between

employment and material welfare, evaluating

Readers may wish to consult other sources for addi-

tional information and alternative measures of inequality.

See, for example, Tabatabai, H.: Statistics on Poverty and

Income Distribution: An ILO Compendium of Data

(Geneva, ILO, 1996); and the World Income Inequality

Database (WIID) of the United Nations University at: http://

www.wider.unu.edu/research/Database/en_GB/database/.

139

KILM 17 Poverty, income distribution, employment by economic class and working poverty

expected to show greater inequality of income

than of consumption. Whether the index is

based on income or consumption is made clear

in the notes to the tables, and it is important for

users to bear the distinction in mind when

attempting to make comparisons. The cumula-

tive distributions of consumption or income

used in constructing the index relate to per

capita levels, and the percentiles are of popula-

tion, not households. Apart from possible

weaknesses in the quality of the underlying

consumption or income data, the adjustments

made to convert the index into a cumulative

distribution of the population may introduce

additional bias or error into the estimates.

Nevertheless, despite these numerous imper-

fections, the index is very useful for studying

trends in inequality across space and time.

poverty lines were perfect, several unanswered

questions would remain. For example, is a

person with a particular consumption level (say

5 a day) in a poor country better or worse

off than a person with the same consumption

level in a rich country? Or is a person living on

5 a day worse off if he or she lives in a coun-

try that has high inequality?

The Gini index, in principle, makes it pos-

sible to compare inequality levels in different

countries and over time, without defining a

particular poverty line, national or interna-

tional. In practice, however, it involves other

problems of comparability. The index is calcu-

lated from survey data, which may relate to

income or consumption. If both consumption

and income information were available in the

requisite detail, the Gini index would be

Box 17. New ILO estimates of employment across economic classes

Recent ILO research has provided a picture of the workforce in the developing world in terms of the

distribution of workers across five economic classes. Currently these are: (1) the extreme working

poor (less than US$1.90 a day); (2) the moderate working poor (between US$1.90 and US$3.10);

(3) the near poor (between US$3.10 and US$5); (4) developing world middle-class workers, who are

those workers living in households with per capita consumption between US$5 and US$13); and

(5) developed world middle-class and above, which are those workers living in households with per

capita consumption greater than US$13 per person per day).

Building on earlier work by the ILO to produce global and regional estimates of the working poor,

a methodology was developed to produce country-level estimates and projections of employment

across five economic classes (Kapsos and Bourmpoula, 2013). This has facilitated the production

of the first ever global and regional estimates of workers across economic classes, providing new

insights into the evolution of employment in the developing world. The aim of the work is to

enhance the body of evidence on trends in employment quality and income distribution in the

developing world – a desirable outcome given the relative dearth of information on these issues

in comparison with indicators on the quantity of employment, such as labour force participation

and unemployment rates.

The authors define workers living with their families on between US$4 (now US$5) and US$13 at PPP

as the developing world’s middle-class, while workers living above US$13 are considered middle

class and upper middle class based on a developed world definition. Growth in middle-class

employment in the developing world can provide substantial benefits to workers and their families,

with evidence suggesting that the developing world’s middle class is able to invest more in health

and education and therefore live considerably healthier and more productive lives than the poor and

near-poor classes. This, in turn, can benefit societies at large through a virtuous circle of higher

productivity employment and faster development. The rise of a stable middle class also helps to

foster political stability through growing demand for accountability and good governance (see

Ravallion, 2009).

The econometric model developed in the paper utilizes available national household survey-based

estimates of the distribution of employment by economic class, augmented by a larger set of

estimates of the total population distribution by class together with key labour market, macroeconomic

and demographic indicators. The output of the model is a complete panel of national estimates and

projections of employment by economic class for 142 developing countries, which serve as the basis

for the production of regional aggregates.

Source: Kapsos, S. and Bourmpoula, E. (2013). “Employment and economic class in the developing world”, ILO Research Paper

No. 6. http://www.ilo.org/wcmsp5/groups/public/---dgreports/---inst/documents/publication/wcms_216451.pdf.

140

KILM 17 Poverty, income distribution, employment by economic class and working poverty

from which table 17a draws. However, the

employment by economic class estimates in table

17b compiled by the ILO on the basis of national

survey data are disaggregated by age (total, youth

and adults, defined as persons aged 15+, 15-24

and 25+ years, respectively) and by sex, allowing

for comparisons across these groups.

Aside from disaggregation into rural and

urban areas for national poverty lines, the

poverty and inequality data in table 17a are

provided at the aggregate level only, without

disaggregation by age and sex. This is due to the

fact that disaggregated poverty data are not avail-

able in the major international data repositories