Public Disclosure Authorized

I
June 2021

Technical Report

Pathways to
Public Disclosure Authorized

Middle-Class
Jobs in Indonesia
Public Disclosure Authorized

Maria Monica Wihardja and Wendy Cunningham

Public Disclosure Authorized

Pathways to Middle-Class Jobs in Indonesia June 2021

Pathways to Middle-Class Jobs in Indonesia
 Disclaimer

This work is a product of the staff of The World Bank, supported by funding from the Australian
Government. The findings, interpretations, and conclusions expressed in this work do not necessarily
reflect the views of the Executive Directors of The World Bank or the governments they represent. The
World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors,
denominations, and other information shown on any map in this work do not imply any judgment on the
part of The World Bank concerning the legal status of any territory or the endorsement or acceptance
of such boundaries.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Disclaimer

Rights and Permissions

© 2021 The World Bank

1818 H Street NW, Washington DC 20433
Telephone: 202-473-1000; Internet: www.worldbank.org

Some rights reserved

The material in this work is subject to copyright. Because The World Bank encourages dissemination of
its knowledge, this work may be reproduced, in whole or in part, for non-commercial purposes as long
as full attribution to this work is given. All queries on rights and licenses, including subsidiary rights,
should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington,
DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org.

Attribution
Please cite the work as follows: “World Bank. 2021. Pathways to Middle-Class Jobs in Indonesia.© World Bank”

Pathways to Middle-Class Jobs in Indonesia June 2021

2 Foreword

Foreword

The World Bank

Satu Kahkonen
Country Director, Indonesia and Timor Leste

Indonesia has made great strides toward becoming a middle-class country. Just before the
Covid-19 pandemic hit in early 2020, it graduated to upper-middle-income country status.
At that point, Indonesia was also well on its way to escaping the “middle-income trap” and
becoming a thriving high-income country, its goal for 2045, one century after it gained
independence. Driven by a strong commitment by the Government of Indonesia to human
capital development and economic transformation, as well as to job creation, jobs have
been a fundamental part of this achievement. They have driven Indonesia’s rapid economic
growth, poverty reduction, and the emergence of a vibrant and growing middle class. Today,
jobs in Indonesia take many forms—40-hours-a-week Jakarta office jobs, food stalls run by
women and family members, on-demand ride-hailing drivers in urban areas, and subsistence
farmers in Papua—weaving a rich tapestry in a diverse economy.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Foreword 3

However, even before the Covid-19 pandemic, Indonesia faced complex challenges in creating
“middle-class jobs,” that is, good–quality jobs that allow an average Indonesian family to
afford a middle-class life. Productivity growth in Indonesia has been insufficient to unlock
the potential of the 47 percent of Indonesians stuck in aspiring middle-class status, partly
due to a structural transformation that has failed to transition enough workers from low-
productivity to high-productivity sectors, firms, and jobs. Most of today’s workforce is not
equipped to hold a middle-class job due to skills shortages or exclusion. Regional partners—
Thailand, the Philippines, China, and Vietnam—have seen much faster middle-class job growth
than Indonesia. The difficulty in creating middle-class jobs is particularly worrisome given
that the tailwind of favorable demography will ease over the coming decade as Indonesia
will begin to age rapidly. These structural challenges have been exacerbated by the Covid-19
pandemic, which has hit jobs in Indonesia hard—in particular in the country’s largely informal
labor market—and interrupted the transition of workers from low-quality to higher-quality,
middle-class jobs.

This report suggests three ways in which Indonesia can build better jobs after the Covid-19
pandemic has passed. First, it follows on the World Bank’s report, Aspiring Indonesia:
Expanding the Middle Class, by asking what reforms could unleash the creation of middle-
class jobs in Indonesia. Second, it is the first attempt to capture the jobs picture in its many
dimensions. While jobs challenges are often perceived in a narrowly sectoral way, this report
brings together all jobs-related dimensions in an integrated and comprehensive manner and
assesses how they intersect and interact. Third, it provides policy reform recommendations
along multiple dimensions and summarizes them in a comprehensive, yet compact, fashion.
This report does not provide short-term guidance but instead asks us to take a step back
and think of the structural changes that need to happen for Indonesia to create a sufficient
number of sustainable middle-class jobs.

The report offers three pathways that together can contribute to the creation of
middle-class jobs. First, Indonesia could accelerate across-the-board productivity growth
by unlocking the entry and growth of new firms to spur competition, innovation, and
productivity in the private sector. Indonesia could improve the productivity of small and
medium enterprises, enabling them to grow and become better creators of middle-class
jobs, and target support to increase productivity in household enterprises where a large
share of jobs exists. Second, Indonesia could transition workers to jobs-friendly sectors and
firms by prioritizing investment promotion strategies in sectors that are amenable to the
creation of middle-class jobs. In tandem with this, Indonesia could close information gaps,
thereby helping workers transition to higher quality sectors, firms, and jobs. Third, Indonesia
could build a middle-class workforce by facilitating learning and providing tailored support
to special groups, in particular women and youth.

It is my hope that this report will stimulate a conversation about the need to create quality
jobs that will help Indonesia build back better in the post-Covid-19 world and realize its dream
of becoming a high-income country by 2045.

Pathways to Middle-Class Jobs in Indonesia June 2021

4 Acknowledgments

Acknowledgments
This report was prepared by a team led by Maria Monica Wihardja (Economist, EEAPV)
and Wendy Cunningham (Lead Economist, HAFH3). The team included (in alphabetical
order) Hamidah Alatas (Consultant, HEASP), Frederico Gil Sander (Lead Economist, EEAM2),
Sarah  Hebous  (Consultant), Sara Johansson de Silva (Consultant), Jonathan Lain (Young
Professional, EA2PV), Abror Tegar Pradana (Consultant, EASP), Devanto Pratomo (Consultant),
Ririn Purnamasari (Senior Economist, EEAPV), Muhamad Raden Purnagunawan (Consultant),
Diana Sari (Consultant), Diahhadi Setyonaluri (Consultant), Victor Steenbergen (Economist,
ETICI), Ekki Syamsulhakim (Senior Social Protection Specialist, HEASP), and Pui Shen Yoong
(Economist, ESAMU).  

The team thanks Camilla Holmemo (former Program Leader, HEADR), Achim Daniel Schmillen

(Program Leader, HEASP), Yongmei Zhou (former Program Leader, EEADR), Habib Rab (Program
Leader, EECDR), Josefina Posadas (Senior Economist, HEASP), Bertine Kamphuis (Senior Private
Sector Specialist, EEAF2), Massimiliano  Calì  (Senior Economist, EEAM2), Francesco Strobbe
(Senior Financial Sector Economist, EEAF2), I Gede Putra  Arsana  (Senior Financial Sector
Specialist, EEAF2), Aufa Doarest (Private Sector Specialist, EEAF2), Laura Ralston (Senior Social
Development Specialist, SEAS2), Michael Weber (Senior Economist, HSPJB), Dino Merotto (Lead
Economist, HSPJB), Reyes Aterido (Consultant, ETIFE), Mauro Testaverde (Senior Economist,
HEASP), and Jaffar Al Rikabi (Economist, EEAM2) for the useful discussions. The team also
thanks Elizabeth Ruppert Bulmer (Lead Economist, HSPJB), Claire Hollweg (Senior Economist,
EEAM1), Kenneth Simler (Senior Economist, EEAPV),  Sjamsu  Rahardja  (Senior Economist,
EECM1), and Truman Packard (Lead Economist, HLCSP) for their in-depth peer review. 

The report team is grateful for the dialogue with the Ministry of National Development
Planning (Kementerian Perencanaan Pembangunan Nasional; Bappenas) and the Coordinating
Ministry of Economic Affairs (Kementerian Koordinator Bidang Perekonomian; CMEA)
throughout the course of the assignment. The team is also grateful for guidance and peer
reviews from Chatib Basri, Vivi Alatas, Christopher Manning, and Titik Anas.

The work was conducted under the general guidance of Satu Kahkonen (Country Director
for Indonesia), Yasser El-Gammal (Practice Manager for Social Protection and Jobs, East Asia
and Pacific Region), Philip O’Keefe (former Practice Manager for Social Protection and Jobs,
East Asia and Pacific Region), and Rinku  Murgai  (Practice Manager for Poverty, East Asia
and Pacific Region). The report has been made possible through a grant from the Australia-
financed Partnership for Knowledge-based Poverty Reduction supported by the Department
of Foreign Affairs and Trade.  

We thank Corinne Bernaldez (Program Assistant, HEASP) and Rizky Fitriany (Team Assistant,
EACIF) for providing excellent administrative support. Thanks are also due to Arsianti
(Consultant, HEASP) for editing the report, to copy editor Katherine Demopoulos, and to
designer Mikael Bima Nainggolan.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Contents 5

Contents
Foreword 2

Acknowledgments 4

Contents 5

Abbreviations, Acronyms, and Indonesian Terms 9

Excecutive Summary 17

1.1 Background and Context 23

Chapter
1.2 Report Objective, Methodology, and Scope 26

01
Annex 1 Data Sources and Definitions 29

Data Sources 29

Type of Job 29

Sector of Employment 30

Introduction Occupation Categories 31

Formality/Informality 33

References 35

2.1 Overview of Macroeconomic Trends 37

Chapter

02
2.2 Why Jobs? 39

2.2.1 Labor income is the main driver of poverty reduction 39

2.2.2 Certain types of jobs help reduce poverty and vulnerability 40

2.2.3 Other types of jobs contribute to the creation of middle-class 41

jobs
2.2.4 Jobs contribute to economic growth 43
Jobs as a Catalyst for
2.3 What Jobs Look Like 43
Economic Growth
2.3.1 Quantity of jobs 43 and a Middle-Class
2.3.2 Quality of jobs 46 Society

2.3.3 Middle-class jobs 56

Pathways to Middle-Class Jobs in Indonesia June 2021

6 Contents

2.4 Factors Underlying Wage Growth and More Middle-Class 61

Jobs
2.5 Conclusions 63

Annex 64

Annex 2.1 Macro-Level Indicators 64

Annex 2.2 Shapley Decomposition – Poverty Reduction 64

Annex 2.3 Shapley Decomposition – Poor-, Vulnerable-, and 64

Middle-Class Jobs
Annex 2.4 Annex Figures and Tables 65

References 79

3.1 Growth Models and the Expansion of the Middle Class 81

Chapter

03
3.1.1 Jobs have been created but the quality of jobs has been 81
insufficiently high
3.1.2 Indonesia’s structural transformation post Asian financial crisis 83
has not brought enough productivity gains
3.2 More and Better Jobs, but Not Yet Good Enough 87

3.3 Incomplete Structural Transformation 91

Structural 3.3.1 Indonesia has not benefited enough from its demographic 91
dividend
Transformation
toward 3.3.2 The contribution of structural transformation to labor 93
productivity growth in Indonesia was smaller than in other
Middle-Class Jobs
regional countries
3.4 Structural Transformation 3.0—More Middle-Class Jobs in 98
the Nonagricultural Sector
3.5 Leveraging FDI for Job Creation 101

3.6 Conclusions 106

References 113

4.1 Overview 117

Chapter

04
4.2 Employment Dynamism in Medium-Sized and Large 122
Manufacturing Plants
4.3 Job Creation and Destruction 125

4.4 How is Demand for Jobs Created in Medium-Sized and Large 130
Manufacturing Firms?
4.4.1 Privately owned firms grow more slowly compared to foreign-owned 131
and state-owned firms
4.4.2 More productive firms are less likely to close down or to become small 132 Enterprise-Level
and micro Dynamics of Job
Creation

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Contents 7

4.5 What Does Manufacturing FDI Mean for Middle-Class Jobs? 134

4.5.1 Foreign firms provide greater benefits to higher-skilled workers 136

4.5.2 Foreign firms provide greater benefits to wage employees 138

4.5.3 Foreign firms provide greater benefits to female workers 139

4.6 Improving Productivity 140

4.7 Improving Wages 143

4.8 Do Household Enterprises Provide a Path to Middle-Class 147

Jobs?
4.8.1 A household enterprise can grow into middle-class status 152

4.9 Conclusions 157

Reform Area 1: Unlocking new firm entry and growth to create 159
competition and innovation to spur productivity in the private
sector
Reform Area 2: Improving the productivity of SMEs so that they 162
grow and become better creators of middle-class jobs
Reform Area 3: Targeting support to increase productivity in 166
HHEs where a large share of jobs exists
Annex 170

Annex 4.1 Additional Figures 170

Annex 4.2 Sector Classification 172

References 173

5.1 Who Are Middle-Class Workers and How Do They Compare to 184
Chapter the Wider Workforce?
5.1.1 Men dominate all jobs and middle-class jobs 184

05
5.1.2 The Indonesian workforce is aging and middle-class workers are 186
also older
5.1.3 The skill level of the Indonesian workforce is lower than 189
education levels might suggest, but a higher level of education is
still correlated with holding a middle-class job
5.1.4 Jobs, especially middle-class jobs, are clustered in urban zones 191
Workers, Jobs, and 5.1.5 While having a moderate disability has little effect on work 192
Barriers to Entry to status, it has a significant effect on middle-class-job status
Middle-Class Jobs 5.2 Barriers to Entry to Middle-Class Jobs 192

5.2.1 Aligning skills with middle-class occupations 193

5.2.2 What can workers do? Skill-up . . . 196

5.2.3 What can workers do? Expand their job-search methods . . . 203

5.2.4 What can workers do? Move to more lucrative jobs . . . 204

Pathways to Middle-Class Jobs in Indonesia June 2021

8 Contents

5.3 Additional Factors Limit Access to Middle-Class Jobs for 206

Vulnerable Populations
5.3.1 Gender roles significantly limit women’s access to middle-class 206
jobs
5.3.2 Youth are at a structural disadvantage 209

5.4 Limited Role of Labor Law in Creating Middle-Class Jobs 212

5.4.1 The limitations of the minimum wage 212

5.4.2 Worker protections are not synonymous with middle-class jobs 214

5.5 Conclusions 217

Reform Area 1: Facilitate Learning 218

Reform Area 2: Support Job Transitions through Information and 224

Financing
Reform Area 3: Provide Tailored Support to Special Groups 229

Annex 231

Annex 5.1 Definition of a Middle-Class Job 231

Annex 5.2 Additional Tables 233

References 235

Chapter
6.1 Summary of Messages 242

06
6.2 The Pathway to Middle-Class Jobs: A Policy Strategy 244

6.2.1 Reform Strategy 1: Accelerate Across-the-Board Productivity 246

Growth
6.2.2 Reform Strategy 2: Structural Transformation 3.0: Transition to 251
Sectors and Firms That Create Good-Quality Jobs
6.2.3 Reform Strategy 3: Build a Middle-Class Workforce 254
6.3 Implementation Challenge 257 Report Summary and
Policy Directions
References 264

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Abbreviations, Acronyms, and Indonesian Terms 9

Abbreviations, Acronyms, and

Indonesian Terms
Term Definition
AI Artificial intelligence

APINDO Indonesian Employers’ Association (Asosiasi Pengusaha Indonesia)

Bappenas Ministry of National Development Planning (Kementerian Perencanaan Pembangunan

Nasional)
BKPM Investment Coordinating Board (Badan Koordinasi Penanaman Modal)

BPJS Badan Penyelenggara Jaminan Sosial (Healthcare and Social Security Agency)

BPS National Statistical Agency (Badan Pusat Statistik)

CAGR Compound annual growth rate

CMEA Coordinating Ministry of Economic Affairs (Kementerian Koordinator Bidang

Perekonomian)
ERIA Economic Research Institute for ASEAN and East Asia

FDI Foreign direct investment

GVC Global value chain

HHE Household enterprise

ICLS International Conference of Labor Statisticians

ICT Information communication technology

ILO International Labour Organization

ISCO International Standard Classification of Occupations

ISIC International Standard Industrial Classification

JKP Job Loss Guarantee (Jaminan Kehilangan Pekerjaan)

KBJI Indonesian Standard Classification of Occupations (Klasifikasi Baku Jenis Indonesia)

KBLI Indonesian Industrial Classification System (Klasifikasi Baku Lapangan Usaha)

KJI Indonesian Classification of Occupations (Klasifikasi Jabatan Indonesia)

KLUI Indonesia Business Classification (Klasifikasi Lapangan Usaha Indonesia)

KPPU Commission for the Supervision of Business Competition (Komisi Pengawas Persaingan
Usaha)
KUR Kredit Usaha Rakyat

Pathways to Middle-Class Jobs in Indonesia June 2021

10 Abbreviations, Acronyms, and Indonesian Terms

Term Definition
Manufacturing Survey of Medium-Sized and Large Manufacturing Firms (Survei Industri Besar/Sedang)
Survey
MIT Massachusetts Institute of Technology

MSMEs Micro, small, and medium enterprises

OECD Organisation for Economic Co-operation and Development

OJK Financial Services Authority (Otoritas Jasa Keuangan)

PIAAC Programme for the International Assessment of Adult Competencies

Sakernas National Labor Force Survey (Survei Angkatan Kerja Nasional)

SMEs Small and medium enterprises

SNI National product standards

STEM Science, technology, engineering, and mathematics

TVET Technical and vocational education and training

UNDP United Nations Development Programme

UNICEF United Nations Children’s Fund

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Figures 11

Figures
No. Title Page
Figure 2.1 Where Have Jobs Been Created? 38
Figure 2.2 Labor Income Starts to Take Over as the Key Contributor to Poverty Reduction at 40
Around US$3.40 a Day Poverty Line
Figure 2.3 Employment in Indonesia Diversifies Away from Agriculture 44
Figure 2.4 Although Making Up a Small Share of Jobs, the High-Value Service Sector Has High 45
Job Prospects
Figure 2.5 Duality of Indonesia’s Labor Market Indicated by High Growth in Wage Employment 46
and Positive Growth in Nonagriculture Casual Employment and Self-Employment
Figure 2.6 Less Than 10 Percent of Jobs Are Stable and Protect Workers 47
Figure 2.7 Occupational Transformation from Skilled Agricultural and Fishery Jobs to 48
Elementary Occupations and Service and Sales Workers
Figure 2.8 Around Half of Informal Workers Were Household Enterprise Owners in 2019 51
Figure 2.9 Indonesian Jobs Are Low Quality by Many Measures 52
Figure 2.10 Manufacturing Had the Highest Compound Annual Earnings Growth Rate in 2008–18 53
Figure 2.11 Composition of Jobs by Income-Class Status Has Barely Changed Since 2001 56
Figure 2.12 The Share of Middle-Class Jobs in Manufacturing among Income Earners Almost 58
Doubled between 2008 and 2018
Figure 2.13 Indonesian Middle-Class Jobs Are Still Subject to Risk 60
Figure 2.14 Wage Employment Is an Important Driver toward More Middle-Income Jobs 61
Annex Figure 2.1 Year-On-Year Real GDP Growth (%) and Employment Elasticity, 1991–2018 64
Annex Figure 2.2 Long-Term View of Employment and Unemployment Rates, 1998–2018 64
Annex Figure 2.3 Share of Contribution to Poverty Reduction, 2000–14 64
Annex Figure 2.4 Average Income and Nonlabor Income in Households below Poverty Line by Different 65
Poverty Line Levels in 2000, 2007, and 2014
Annex Figure 2.5 Shapley Value Estimates by Sector: Contributions to the Decline in Poor- and 66
Vulnerable-Class Jobs, 2011–18
Annex Figure 2.6 Shapley Value Estimates by Sector and Status of Employment: Contributions to the 66
Decline in Poor- and Vulnerable-Class Jobs, 2011–18
Annex Figure 2.7 Shapley Value Estimates by Sector and Formality: Contributions to the Decline in 67
Poor- and Vulnerable-Class Jobs, 2011–18
Annex Figure 2.8 Shapley Value Estimates by Sector: Contributions to the Increase in Middle-Class 67
Jobs, 2011–18
Annex Figure 2.9 Shapley Value Estimates by Sector and Job Status: Contributions to the Increase in 68
Middle-Class Jobs, 2011–18
Annex Figure 2.10 Shapley Value Estimates by Sector and Formality: Contributions to the Increase in 68
Middle-Class Jobs, 2011–18
Annex Figure 2.11 Employment Share by Occupation Level, 2007–15 69
Annex Figure 2.12 Share of Self-Employment and Family Workers, 2018 69

Pathways to Middle-Class Jobs in Indonesia June 2021

12 Figures

No. Title Page

Annex Figure 2.13 Evolution of Earnings and Share of Waged Workers Earning Below the Minimum 70
Wage, 2008–18
Annex Figure 2.14 Underemployment Rate by Location, 2008–18 70
Annex Figure 2.15 Underemployment Rate by Sector, 2018 71
Annex Figure 2.16 Employment Share by Wage Category, 2008–18 71
Annex Figure 2.17 Employment by Employment Status, 2008–18 72
Annex Figure 2.18 Employment Status by Island Region, 2018 72
Annex Figure 2.19 Share of Workers by Educational Attainment, 2008–18 73
Annex Figure 2.20 Share of Employment by Gender, 2008–18 73
Annex Figure 2.21 Female Labor Force Participation Rate, 1991–2019 74
Annex Figure 2.22 Male Labor Force Participation Rate, 1991–2019 74
Annex Figure 2.23 Female Employment Rate, 1991–2019 74
Annex Figure 2.24 Male Employment Rate, 1991–2019 74
Figure 3.1 Respectable Growth and Job-Creation Rates 82
Figure 3.2 But No Convergence with Richer Asian Economies 82
Figure 3.3 Manufacturing Previously Drove Industrial and Overall Growth, but Has Lost 84
Momentum
Figure 3.4 Premature Deindustrialization in Indonesia 84
Figure 3.5 Indonesia Is Not Catching Up in Competitiveness 86
Figure 3.6 Falling Behind in Labor Productivity in Low-Skill Manufacturing 86
Figure 3.7 The Economy Diversified Away from Agriculture into Services . . . 88
Figure 3.8 . . . But Shifts in Labor Have Been More Dramatic and Have Taken Place in Different 88
Sectors from Value Added
Figure 3.9 Slow or Negative Labor Productivity Growth, Especially in Industry and Low-Value 89
Services, after 2010
Figure 3.10 Employment Has Shifted out of Agriculture and Mostly into Lower-Productivity 90
Services
Figure 3.11 Sectors That Are More Productive Tend to Pay Higher Wages, but Employ a Minority 91
of Workers
Figure 3.12 Labor Productivity Growth Has Driven Growth 94
Figure 3.13 But the Contribution of Structural Transformation to Labor Productivity Growth Has 95
Been Limited, Which Has Limited the (Positive) Impacts on Jobs
Figure 3.14 Structural Change Did Not Foster Job Creation in Higher-Productivity Sectors 97
Figure 3.15 More and Even Better Jobs: Structural Transformation 3.0 98
Figure 3.16 Indonesia Is Not Attracting As Much FDI As Comparator Countries 101
Figure 3.17 Foreign-Owned and/or Exporting Plants Pay Higher Wages on Average . . . 103
Figure 3.18 . . . And Create More Jobs 103
Figure 3.19 Foreign Direct Investment Increases the Share of Manufacturing Jobs in Total 104
Employment
Figure 3.20 Foreign Direct Investment in Manufacturing Fosters Labor Productivity Increases in 106
Both Manufacturing and Services
Figure 4.1 A Firm Does Not Increase in Size until it Has Been Operating for 22 Years 118
Figure 4.2 The Hollowing-Out Period of 2000–10 120

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Figures 13

No. Title Page

Figure 4.3 Young Plants (Aged 0–9) Dominated in the 1990s; Older Plants (Aged 10 and above) 123
Dominated after the Asian Financial Crisis
Figure 4.4 Until 2013, Plants Aged 10–19 Absorbed Most Labor, Then Plants Aged 30 Plus Took 123
Over
Figure 4.5 New Firm Growth Slowed during Economic Crises 124
Figure 4.6 Employment Growth Due To Firm Entry Stalled during the Entire Hollowing-Out 124
Decade
Figure 4.7 Most Firms Are Domestically Owned . . . 125
Figure 4.8 . . . Though an Increasing Share of Workers Are Employed in Foreign-Owned Firms 125
Figure 4.9 Most Employment Growth and Destruction Was in Surviving Firms 125
Figure 4.10 Net Job Creation Was Strongest in Textiles, Apparel, and Footwear before the Asian 127
Financial Crisis, Then Joined by Food and Beverages
Figure 4.11 Net Job Creation Greatest in Very Large Firms until 2013; Net Job Creation in Small 128
Firms Resumed in 2013
Figure 4.12 In 2013, Net Job Creation Revived in the Youngest Plants Only 128
Figure 4.13 Net Job Creation in Foreign Firms Was Generally Positive across the Period 129
Figure 4.14 Privately Owned Firms Are Net Job Creators, Publicly Owned Are Net Job Destroyers 129
Figure 4.15 Higher Productivity Growth is Associated with Higher Employment Growth 130
Figure 4.16 Privately Owned Indonesian Firms Grow Slowly Compared to Firms under Public or 131
Foreign Ownership
Figure 4.17 Large, Old, Foreign, and More Productive Plants Do Not Exit 133
Figure 4.18 Productive, State-Owned and Larger Firms Are Less Likely to Significantly Downsize 133
Figure 4.19 Indonesia Has Limited FDI inflows 134
Figure 4.20 Multinational Activity in Indonesia Is Subdued 134
Figure 4.21 Foreign-Owned Firms Pay Better Wages Than Domestically Owned Firms 136
Figure 4.22 Wage Dynamics for Production Workers—Domestic Firms versus Foreign Owned 136
(Average 2010–15)
Figure 4.23 Larger and Foreign-Owned Firms Are More Productive 140
Figure 4.24 Smaller, Older, and Exporting Firms Have Higher Productivity Growth Rates 141
Figure 4.25 The Decomposition of Labor Productivity Growth, Average Annual, 2005–15 141
Figure 4.26 Employment Declines with Productivity Increases 144
Figure 4.27 Employment Declines with Wage Increases 144
Figure 4.28 The Wage-Productivity Gap Widened Significantly in the United States in 1980–2014 145
Figure 4.29 Real Wage and Real Labor Productivity Trends Track Each Other Quite Closely 145
Figure 4.30 Most Sectors Pay below Middle-Class Wages 146
Figure 4.31 Women Work in Lower-Productivity Sectors Relative to Men 147
Figure 4.32 Women Work in Lower-Wage Sectors Relative to Men 147
Figure 4.33 Most Owners of Household Enterprises Wanted to Start a Business 150
Figure 4.34 Monthly Earnings of Owners of Household Enterprises (HHEs) Are Very Similar to 151
Those of Casual Workers

Pathways to Middle-Class Jobs in Indonesia June 2021

14 Figures

No. Title Page

Figure 4.35 Progress of Household Enterprise Owners, 2007–14 153
Figure 4.36 Only 8 Percent of Household Enterprises in a Nationally Representative Survey in 153
2020 Had Profit Per Worker above the Middle-Class Cutoff
Figure 4.37 Most Firms Want to Expand Their Enterprise Premises, but among Nonfarm HHE 155
Who Don’t Want to Expand, Half Are Happy with What They Already Have
Annex Figure 4.1 Job Growth Slowed during the Hollowing-Out Period 170
Annex Figure 4.2 (Log) Productivity Differentials for Apparel 171
Annex Figure 4.3 (Log) Wage Differentials for Apparel 171
Figure 5.1 Labor Force Participation Rates by Gender Have Not Changed in More Than 20 Years 185
Figure 5.2 More Women than Men Drop Out of the Labor Force during Prime Childbearing Years 185
Figure 5.3 Less “Vulnerable” Groups Are Overrepresented in Middle-Class Jobs 186
Figure 5.4 Indonesia’s Labor Force Is of Prime Working Age 187
Figure 5.5 The Number of Workers is Growing in All Age Groups but Declining Among Youth 187
Figure 5.6 The Workforce Is Slowly Acquiring Years of Schooling 190
Figure 5.7 Only 41% of the Labor Force Has Acquired an Upper Secondary Education or More 190
Figure 5.8 Middle-Class Jobs Are Overrepresented among More-Skilled Occupations 193
Figure 5.9 Some of the Fastest-Growing Occupations Offer More Middle-Class Jobs 195
Figure 5.10 Indonesia’s Labor Force Has Education Rates Similar to Those of Asian Middle-Income 196
Countries, but Rates of Return on Par with Less-Advanced Countries
Figure 5.11 Middle-Class Jobs Are Overwhelmingly Associated with a Secondary School Education 197
or Above
Figure 5.12 Middle-Class Occupations Require Analytical, Cognitive, and Interpersonal Skills 200
Figure 5.13 Non-Middle-Class Occupations Require Manual Skills 200
Figure 5.14 Firm Perceptions of Skills Shortage and Skill Training 201
Figure 5.15 Few Indonesian Firms Offer Training 201
Figure 5.16 Women’s Fields of Postsecondary Study Reflect the Distribution for Non-middle-class 202
Jobs, while Men’s Distribution Is More Aligned with Middle-Class Jobs
Figure 5.17 Proactive Job Search Increases with Education Level 203
Figure 5.18 More Than Half of Jobseekers Find a Job within Three Months 204
Figure 5.19 Women Are Overrepresented in Low-Wage Types of Jobs 206
Figure 5.20 Women Have Particularly High Earnings in High Value-Added Service Sector 207
Figure 5.21 Women Enjoy Greater Nonpecuniary Benefits Than Do Men 209
Figure 5.22 Workers Acquire Middle-Class Jobs as They Age, But Women’s Acquisition of Middle- 210
Class Jobs Stalls during Childbearing Years
Figure 5.23 Most Youth Who Leave School Go to Work; High Youth Unemployment Reflects the 211
Large Number of People in Certain Age Cohorts Arriving Simultaneously in the Labor
Market
Figure 5.24 The Distribution of Wages Paid to Those Holding Employment Contracts in West 213
Sumatra and Aceh, and whether Minimum Wages Affect Wages (West Sumatra) or
Do Not Affect Wages (Aceh)
Figure 5.25 Distribution of Employees by Middle-Class Wage, Receipt of Social Benefits, and 216
Contract Status
Figure 6.1 Three-Pronged Strategy Toward Middle-Class Jobs in Indonesia 245

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Tables 15

Tables
No. Title Page
Annex Table 1.1 Industry Codes in Sakernas 30
Annex Table 1.2 Conversion Table 17-Sector to 9-Sector 31
Annex Table 1.3 Occupational Codes Used in the Sakernas 32
Annex Table 1.4 Cross-Tabulation between Job Status and Occupation 33
Table 2.1 Some Jobs Are Successful at Bringing Indonesians out of Poverty and Vulnerability, 42
but Cannot Give Them a Middle-Class Income
Table 2.2 Who are Indonesia’s Informal Workers? 49
Table B2.1.1 Top Ten Five-Digit Sectors for Nonfarm HHEs 49
Table 2.3 Gig Workers Are More Likely Than Average to Be Informal 50
Table 2.4 The Manufacturing Sector Has the Highest Number of Income Earners (in Absolute 57
Terms) Earning Wages above the Middle-Class Cutoff
Table 2.5 If Worker Protection Benefits and Contract Types Are Added to the Criteria for a 59
Middle-Class Job, among Wage Employees, Only the Civil Service Is a Source of Middle-
Class Jobs
Table 2.6 Decomposition of Mean Earnings Growth 2001–18 62
Annex Table 2.1 Number of Employed by 17-Digit Sector, 2008 and 2018 75
Annex Table 2.2 CAGR and Absolute Change in Employment by Wage Category, 2008–18 75
Annex Table 2.3 Wage Employment by Contract and Worker Protection Status, 2019 76
Annex Table 2.4 Share of Informal Employment within Each Economic Sector, 2019 76
Annex Table 2.5 Household Enterprise Owners by One-Digit Level Sector, 2019 77
Annex Table 2.6 Dependent Workers by One-Digit Level Sector, 2019 78
Table 3.1 Foreign Direct Investment Increases Productivity in Manufacturing and Services 105
Table 4.1 Manufacturing Sector as a Source of Middle-Class Jobs in 2018 118
Table 4.2 Descriptive Statistics of Manufacturing Labor Force 121
Table 4.3 Top Net Job Creators and Destroyers 126
Table 4.4 Top Five Subsectors with the Highest Job Churning in 2013 and 2014 127
Table 4.5 Changes in Size of Firms between 2005 and 2015 132
Table 4.6 Within-Sector Effect of FDI By Education Level 137
Table 4.7 Within-Sector Effect of FDI by Occupation 138
Table 4.8 Within-Sector Effect of FDI by Employment Status 138
Table 4.9 Within-Sector Effect of FDI by Gender 139
Table 4.10 Correlation between Labor Productivity and Plant Size 142
Table 4.11 A 10 Percent Increase in Productivity Results in a 2.14 Percent Increase in Wages; 143
Private Domestic Firms Share Gains More with Workers
Table 4.12 Majority of Rural Household Enterprises Operate in the Farm Sector; Majority of 149
Urban Household Enterprises Operate in Wholesale and Retail Trade, and Restaurants
and Hotels
Table 4.13 Distribution of “Self-Employed” Household Enterprises (n=21.92 million) 152
Table 4.14 Very Low Profit per Worker of HHE Reflects the Low Productivity of HHEs in General 154

Pathways to Middle-Class Jobs in Indonesia June 2021

16 Tables & Boxes

No. Title Page

Table 4.15 Key Characteristics of Microenterprises/Household Enterprises, and Small, and 166
Medium-Sized Enterprises
Annex Table 4.1 Manufacturing Subsector Classification 172
Table 5.1 Disability Status and Employment 192
Table 5.2 Occupation for Which Median (Mean) Income is above the Middle-Class Cutoff, Sorted 194
by One-Digit KBJI Occupation
Table 5.3 Percentage of Firms Citing Inadequate Skills When Hiring Workers, by Type of Worker 197
Table 5.4 Share of Workers Who Receive Each Benefit, by Middle-Class Job Status 214
Table B5.7.1 Findings of the First Occupational Employment and Vacancy Survey 227
Annex Table 5.1 Share of workers who receive each benefit, by middle-class job status 232
Annex Table 5.2 Determinant of Middle-Class Jobs, Average Marginal Effects 233
Table 6.1 Summary of Policy Directions 259

Boxes
No. Title Page
Box 1.1 Definition of a Middle-Class Job 25
Box 1.2 Methodology, Definition, and Data 28
Box 2.1 What Are Nonfarm Household Enterprises? 49
Box 2.2 A Transition Interrupted? 53
Box 3.1 What Is Driving Value-Added Growth and Labor Productivity? 92
Box 3.2 Demand-Side Channels: How the COVID-19 Crisis Is Impacting Jobs 95
Box 4.1 Who Are Manufacturing Workers? 120
Box 4.2 Cross-Country Studies on Distributional Impact of Foreign Direct Investment 134
Box 4.3 Definition of Household Enterprises versus Microenterprises and Informal Jobs 147
Box 4.4 The Role of E-commerce in Helping Firms and Household Enterprises Sail through the 155
COVID-19 Crisis
Box 4.5 Case Study: Indonesia’s SME Participation in Global Value Chains: Javara 163
Box 4.6 Banking Agents: Benefits to Household, Small, and Medium-Sized Enterprises 169
Box 5.1 Will COVID Further Hinder Youth from Attaining Middle-Class Jobs? 188
Box 5.2 Estimating the Task Intensity of Middle-Class and Non-Middle-Class Jobs in Indonesia 199
Box 5.3 Transitioning to Adulthood 211
Box 5.4 A Twelve-Step Plan to Boost Learning 219
Box 5.5 Teaching Interpersonal Skills in Schools 222
Box 5.6 Priority Reforms to Build a World-Class Workforce Development System in Indonesia 224
Box 5.7 Workforce Development Tools for Detailed Jobs Information 226
Box 6.1 Creating Better and More Jobs: Recent Reforms to Investment, Access to Skilled 250
Foreign Workers in Shortage, and Manufacturing Inputs
Box 6.2 Finding a Balance between Labor Market Flexibility and Worker Protection 256

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Executive Summary 17

Executive
Summary

Pathways to Middle-Class Jobs in Indonesia June 2021

18 Executive Summary

“Our dream, our ambition is that by 2045, after one century of Indonesian independence,
Indonesia should, Insya Allah [God willing], have escaped the middle-income trap,” President
Joko Widodo declared in the opening lines of his inaugural address in October 2019.
Indonesia is on its way, with single-digit poverty rates and economic growth averaging
5 percent annually. However, half (47 percent) of Indonesians are stuck just below the
middle-class threshold. They belong to an “aspiring middle class” that is neither poor, nor
highly vulnerable to becoming poor, but which has not reached the level and stability of
consumption associated with middle-class status. Indonesia’s jobs situation may be holding
the country back from rapid progress in realizing its dream.

Jobs have contributed to both growth and poverty reduction, and Indonesia creates an
average of 2 million new jobs each year. These take many forms: They may be office jobs
in Jakarta that require 40 hours of work per week and pay a salary. They may be family-
run food stalls that earn a modest profit. They may be on-demand ride-hailing drivers who
spend 50 hours a week on their motorbikes, or subsistence farmers in Papua who feed
their family members from their labors. In 2019, there were more than 120 million working
adults in Indonesia. The employment rate reached a two-decade record high in 2019, with
67.5 percent of adults in the labor force, and unemployment fell to its lowest level in two
decades, at 5.2 percent.

Despite these encouraging signs, Indonesia is not creating the “middle-class jobs” needed to
fuel a middle-class country. Among the 85 million income earners (wage employees, casual
workers, and the self-employed), only 13 million are earning a level of  income sufficient
to allow a family of four to purchase a middle-class way of life.1 And only 3.5 million wage
employees earn above the threshold for middle-class earnings, enjoy full social benefits, and
hold an indefinite-term employment contract.

This report posits that Indonesia is a story of low-productivity job growth. Jobs are being
created, and they fuel the economy, but productivity growth is insufficient to unlock the
backlog of people who aspire to move up to the middle class. The report explores the
factors that are holding back productivity growth and the creation of middle-class jobs, and
it offers policy recommendations that could unlock these barriers.

The quest for better jobs is at the top of the government’s policy agenda. This issue is
perhaps even more crucial in 2021, as Indonesia, and the rest of the world, manage the
economic fallout from the COVID-19 pandemic. This report aims to provide guidance to help
Indonesia “build back better” in its recovery from COVID-19 and to reinvent itself as a middle-
income country.

1 A Propitious Beginning, Cut Short

The structural transformation of the 1980s and 1990s portended a middle-class Indonesia
by 2019. Indonesia was on a path of quality job growth as people moved out of agriculture
and into the rapidly growing manufacturing sector during “Structural Transformation 1.0.”

1 The data do not include the 15 million unpaid workers and the 24 million employers earning profits.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Executive Summary 19

The boom, however, was interrupted by the Asian financial crisis. Foreign direct investment
slowed and manufacturing stalled. Post-crisis, Indonesia entered a “hollowing-out” period,
when manufacturing did not resume and a commodity boom fueled the country. Since
commodities do not create many jobs, “Structural Transformation 2.0” was in fact a
movement out of agriculture and into low-value service jobs. Although Indonesia maintained
5  percent economic growth in this period and continued to reduce poverty, movement
toward middle-class jobs stalled. A decade of potential job improvements was lost.

At the close of the commodity boom in 2012, when low-value services were deeply entrenched
in the economy, manufacturing began to return, in a kind of “Structural Transformation 3.0.”
It was, however, constrained by a range of factors, including a business environment that
was less hospitable to foreign investment and trade than that of the 1980s and 1990s. While
job creation continued, productivity levels were insufficient to create the middle-class jobs
that Indonesians were increasingly aspiring to.

2 Poor Job Quality Exacerbated by the Structure of Firms and

Shortcomings in the Labor Force
On average, Indonesian jobs are of low quality. Two-thirds of jobs are in agriculture or low-
value services, including (in decreasing order of size) informal retail trade of food and
beverages, food stalls, and motorcycle ride-hailing transportation. In 2019, 38.2 percent of
all workers were firm owners, while 61.8 percent were employees. Among firm owners, more
than half were self-employed. Among employees, almost two-thirds were working without
any contract. About three-quarters of Indonesia’s labor market is uncovered or informal.
Labor productivity is low, and this is partly reflected in high unit labor costs relative to the
East Asian region.

The structure of Indonesia’s firms sector is not conducive to the creation of middle-class
jobs. Two-thirds of jobs are in household enterprises, with 45 million owners and 38 million
workers, nearly all of whom are informal. Formal firms are slow to create middle-class jobs.
The manufacturing sector—the source of incredible potential in the 1980s and 1990s—
creates some jobs, but most job creation is in old (more than 30 years) and very large firms.
New firms have difficulty growing and bringing in new staff, thus limiting competition and
increases in productivity. Job churning is significant although it reflects job instability rather
than job upgrading and occurs among existing firms rather than being due to firm entry and
exit. Further, foreign firms, who hire more on average than domestic firms and pay higher
wages, are less prevalent in Indonesia than in those of its regional competitors that have
leveraged foreign direct investment for rapid economic and job growth.

Most of today’s workforce is not equipped to hold a middle-class job. Middle-class jobs tend
to be in more-skilled occupations that require strong cognitive, interpersonal,2 and digital
skills, as well as knowledge in science, technology, engineering, mathematics, or business

2 Interpersonal skills are the behaviors people employ in their relationships. Good interpersonal skills are necessary for
effective collaboration with others.

Pathways to Middle-Class Jobs in Indonesia June 2021

20 Executive Summary

administration. However, 57 percent of the labor force has a lower-secondary education (nine
years of school) or lower. Standardized tests of adult skills (under the Programme for the
International Assessment of Adult Competencies) conducted in 2014–15 ranked Indonesia
last among members of the Organization for Economic Co-operation and Development
(OECD) and middle-income countries;3 Indonesia’s best performers scored worse than the
average performer in the OECD workforce.4 Nearly 14 percent of firms interviewed in the
Doing Business 2015 survey in Indonesia reported that “workforce education” was a major
or severe obstacle to their operations. This compared to 8.4 percent of employers in the
Philippines and 2.5 percent in Thailand.  The skills shortage is particularly severe among
more-skilled types of workers, including managers. The average worker today is aged 40 or
above, is a household head, male, and will not return to school. An absence of systematized
information about quality training programs hinders upskilling and reskilling. Women
and youth face additional barriers to entering and succeeding in the labor market. Job
matches are likely not the best—both for the economy as a whole and for workers—due to
the limited availability of information on job opportunities, and labor legislation that
discourages job turnover.

There are, however, some successes that can be leveraged. A case study of the manufacturing
sector finds that more productive firms, exporting firms, and technology-intensive firms pay
better than those that are not productive, do not export, and use only simple technologies.
More productive firms and foreign firms hire more labor. Importantly, there is no evidence
of a trade-off of labor versus productivity. Instead, more productive firms tend to hire
more labor. On the worker side, virtual learning is opening up adult-learning programs. The
KarirHub database is being further developed to enable it to consolidate information on
job vacancies, learning opportunities, and career options. The goal is to facilitate worker
mobility and better job matches.

3 Productivity for All and Structural Transformation 3.0

for Some
Indonesia could implement a three-pronged strategy if it is to prevent workers from being
trapped in low-productivity jobs.

First, Indonesia could promote productivity growth across the board, not least in sectors
with low value added. The most promising route to increasing productivity is enacting
competition-enhancing policies that will help firms enter the marketplace and grow, and that
will spur innovation. Competition-enhancing policies include those that lower the high cost
of trade, increase access to foreign talent currently in shortage, and attract export-oriented,
efficiency-seeking foreign direct investment with links to global value chains. Competition-
enhancing policies also include those that improve domestic regulations and create a more
predictable and level playing field for firms, and those that strengthen institutions that
3 The middle-income countries surveyed were Czech Republic, Estonia, Slovak Republic, Chile, Indonesia (Jakarta only),
Lithuania, Slovenia, and Turkey.
4 In Indonesia, data was only collected in Jakarta. If data were to be collected for the whole country, the score may have
been even lower.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Executive Summary 21

promote competition and sound business regulation. Indonesia will also need to make a
concerted effort to support productivity growth in household, small, and medium-sized
enterprises; they provide a large share of Indonesia’s jobs but many operate at a low level
of productivity.

Second, Indonesia could facilitate a more decisive shift in economic activity and workers
toward more productive and higher-paying firms, sectors, and jobs. Even if reforms to
increase productivity across the board are sector blind, Indonesia could prioritize certain
policy reforms in sectors in which middle-class jobs are likely to be created. For example, it
could prioritize investment promotion strategies for foreign direct investment attraction
and generation in sectors and projects (including infrastructure projects) likely to be a
source of middle-class jobs. Tax incentives are often considered in this context, but they
are an inferior instrument, especially in Indonesia, which already collects less taxes than
most of its peers and as a result has a revenue base that is insufficient to sustainably fund
the public services such as infrastructure that could attract high-quality investments.
Moreover, if workers are to transition across sectors, firms, and occupations, they also
need robust information systems, a strong job-matching mechanism, legislation that does
not disincentivize job changes, and support for job movements—including unemployment
insurance—so that they may appropriately prepare for, and advance to, these better jobs.

Third, Indonesia could build a workforce that has the necessary skills to take on the new
jobs in higher value added and internationally competitive sectors. This will require changes
to the education system to better prepare today’s youth for modern jobs. Perhaps more
challenging, however, it will also require innovations to upskill the current adult labor force
and benefit more from potential technology transfers and spillovers of foreign direct
investment from multinationals. Additional support to attract women and youth into the
workforce is increasingly necessary as the share of the working-age population begins to
shrink.

4 A Propitious Moment
Indonesia has shown that it can take on the challenge of implementing a multifaceted reform
agenda for jobs. It is one of only a few countries that has approached the jobs agenda from
a holistic perspective, engaged stakeholders across the spectrum, and placed job creation
at the center of its legislative program. The accountability structure and implementation
monitoring will be critical and, perhaps, even more of a challenge than the preparation of
the legislation.

The recent economic disruption caused by the COVID-19 crisis threatens Indonesia’s progress
toward more productive, middle-class jobs. The Asian financial crisis had such an effect.
However, with careful planning to build back better, the disruption could be an opportunity
to systematically and strategically map out a recovery agenda that focuses on productivity
and good jobs.

Pathways to Middle-Class Jobs in Indonesia June 2021

 22 Introduction

1
Introduction

Chapter 1
1 The report has been made possible by a grant from the Australia-financed Partnership for Knowledge-based Poverty Reduction.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Introduction 23

INTRODUCTION
1.1 Background and Context
1

Indonesia has been successful in sustaining growth and

poverty reduction, but it still has a way to go to realize
its middle-class aspirations. Over the past half century,
Indonesia has seen strong economic growth and a substantial
decrease in extreme poverty, enabling it to achieve upper-
middle-income status in 2020.2 In 1967, Indonesia’s GDP was
just US$657 per person, making it one of the poorest countries
in the world.3 Over the next 50 years to 2016, with growth
averaging 5.6 percent annually, GDP per capita grew sixfold
to nearly US$4,000 and extreme poverty4 fell to 7 percent.
However, despite the massive reduction in poverty, 35 percent
of Indonesians remained poor or vulnerable to poverty and a
further 45 percent of the population5—115 million people—
while free from poverty and vulnerability, had not yet achieved
the economic security and lifestyle of the middle class.

Jobs have contributed to both growth and poverty

reduction. Indonesia has made significant progress in job
creation, formalization, and diversification (see Box 1.1
for the definition of “jobs” used in this report). In 2018, the
employment rate reached a two-decade record high, and the
unemployment rate continued to fall. Approximately half of
Indonesia’s workers had a job that paid a wage or salary in 2018,
compared to 2001 when only 36 percent did. The other half of
workers are self-employed, employers, or unpaid workers. The
diversification of jobs out of agriculture and into other sectors
has increased productivity and wages.

2 Indonesia became an upper-middle-income country in July 2020 (see Office

of Assistant to Deputy Cabinet Secretary for State Documents and Translation
2020). However, due to the pandemic and a subsequent decline in GDP and GNI
per capita, in February 2021 the government projected that Indonesia would fall
into lower-middle-income status (see Xinhua 2021).
3 World Bank 2019
4 Based on an international poverty line of US$1.90 per day.
5 World Bank 2019

Pathways to Middle-Class Jobs in Indonesia June 2021

24 Introduction

Despite these encouraging signs, Indonesia is not creating the “middle-class

jobs” needed to fuel a middle-class country. For Indonesia to realize its middle-class
aspirations, it needs the jobs to get it there. Fundamentally, a middle-class job is one that
furnishes a worker with sufficient earnings for them to provide a middle-class standard
of living for their family. The job can take many different forms: for example, a full-time
employee who works a standard 40-hour workweek, collects social benefits, and earns
wages commensurate with middle-class consumption patterns; a self-employed individual
assisted by family members in running a successful microbusiness and who earns at least a
middle-class income, or a successful gig economy worker who earns at or above the middle-
class income cutoff. A middle-class job may be in the formal sector (jobs and firms enrolled
in government registries) or the informal sector; it may be located in an urban or rural area;
it may be in a large business or a microenterprise; and it may be situated in any industrial
sector. The main criterion is that it provides a middle-class standard of living.

However, the quality of jobs in Indonesia is worryingly low. Half of all jobs created in
the past two decades are in the lowest wage and productivity sectors: wholesale and retail
trade, accommodation, food and beverages, and other services. Indonesia’s prospects of
creating jobs with high levels of productivity are dim. In 2016, 85 percent of nonagricultural
enterprises were microenterprises with no legal status and with little prospect of growing.6
Without significant policy reform, this is unlikely to change.

The highest level of government stresses the importance of an agenda for creating
middle-class jobs in Indonesia and, to this end, has paid unprecedented attention
to policy reform. In his 2019 inaugural speech, President Widodo identified jobs as a key
input to the achievement of Indonesia’s middle-class ambitions. Specifically, the president
spoke of the need to develop a “workforce that is hardworking, dynamic, skilled . . .” and to
transform the country economically to create a “a competitive and modern manufacturing
and service-based economy that has high added value for the prosperity of the nation .
. . for all Indonesian people.” He defined specific actions, committing the government to
passing laws to unleash job creation across a wide range of sectors. This culminated in the
Omnibus Law for Job Creation (see Box 6.1 and Box 6.2).

While policy making is rapidly advancing, the empirical foundations for policy
development remain incomplete. Indonesia is data rich and is home to a large intellectual
community that is exploring the many facets of job creation. However, jobs are, by nature,
a multisectoral challenge and levers in one sector may have complementary or negative
effects on other sectors. Thus, effective and efficient policy to shift Indonesia away from
a low-quality jobs equilibrium and toward a middle-class jobs equilibrium must be built on
a multisectoral foundation that identifies the commonalities and makes conscious choices
about trade-offs.

6 Alatas and Wihardja. Forthcoming

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Introduction 25

Box 1.1 Definition of a Middle-Class Job

For the purposes of this report, a job is defined as any income-earning activity that is not illegal. Thus, a job may
be as simple as the work a full-time employee carries out in a factory for a salary, or as complex as a portfolio of
income-earning activities carried out by a rural household in a variety of sectors at different times of the year. 

A middle-class job is defined as a job that pays a level of income and gives a certain level of job satisfaction,
benefits, and security commensurate with the expectations of a middle-class population.a A middle-class job may
be a wage job where a firm remunerates a worker in cash (wages or salary of at least the value of the middle-
class consumption basket) and/or in kind (for example, payments for publicly provided social benefits, social
insurance, paid leave) in line with a legally binding employment contract. A job may also be self-employment,
with the worker earning sufficient profit to enjoy an income and purchase social benefits commensurate with a
middle-class life. 

To facilitate the empirical analysis, this report uses a simple definition of a middle-class job that can be quantified
using data from the National Labor Force Survey (Survei Angkatan Kerja Nasional; Sakernas). Namely, the report
defines  a middle-class  job as  one paying at or above a  middle-class income, equivalent to Rp3,752,000
monthly (at 2018 prices).b

This definition necessarily limits our analysis of middle-class jobs to 68.6 percent of all jobs. The Sakernas does
not include information on profits earned by employers with workers and it does not impute wages of unpaid
workers (whose remuneration likely comes from pooled household income). Indonesia’s Economic Census 2016
does not include job-specific information. Thus, while our analysis of jobs as a whole covers the universe of
jobs in Indonesia—approximately 124 million in 2018—our analysis of middle-class jobs covers the 85 million jobs
worked by paid employees and the self-employed.c 

a Following Brummund, Mann, and Rodriguez-Castelan (2016).

b Following World Bank (2019), a member of the middle class is defined as having less than 10 percent probability of transitioning into poverty (measured
by the poverty line) and vulnerability (measured by the value of household consumption above the poverty line and where the household has a 10
percent chance or more of falling below the poverty line). The threshold for middle-class consumption is approximately 3.5 times the poverty line per
capita. Since some household members do not earn income, a middle-class wage needs to be sufficient to lift the whole household above the middle-class
consumption line. Thus, assuming one full-time adult worker and one part-time adult worker (since only half of Indonesian women work) in the household
and two dependent nonworkers, the middle-class earnings would be equivalent to (3.5*poverty line*4)/1.5, or Rp3,752,000 per month in 2018. This
is a flat amount applied to all income earners. This assumption does not allow for variation based on household size or composition since there is no
information about household characteristics of workers in Sakernas. This report uses the national poverty line with a share of middle-class jobs out of
total income-earning employment of 15.4 percent in 2019. If the provincial, urban/rural poverty line is used, this share is 14 percent. From 2001 to 2020,
the shares of middle-class jobs calculated with the national poverty line and the provincial, urban/rural poverty line track each other closely.

c The Sakernas does not include information on social insurance and benefits, or contract status enjoyed by the self-employed (23.6 million), casual
workers (12.2 million), employers (23.8 million), or the unpaid (15 million). In total, these are equivalent to two-thirds of all jobs. If variables for wages,
social insurance and benefits, and contract status were included in the definition of middle class, all these workers would be excluded from the sample
because the data are missing. Instead, the detail on social insurance, social benefits, and contract status was sacrificed in order to retain in the sample the
self-employed and casual workers.

Pathways to Middle-Class Jobs in Indonesia June 2021

26 Introduction

1.2 Report Objective, Methodology, and Scope

Pathways to Middle-Class Jobs in Indonesia consolidates the existing, multisectoral
empirical research to provide an evidence-based narrative about the current jobs
situation and the related policy framework that is necessary to spur the creation
of middle-class jobs and a middle-class Indonesia. This report builds on, and contributes
to, several ongoing debates. It was in preparation when the government was reviewing and
debating a range of laws related to job creation and job quality, and followed the Indonesia
Development Forum 2019, where the ideas underlying this report were first tested. It was
first being prepared as the Jokowi administration had just begun its second term.

This report presents the challenge of middle-class jobs from three dimensions: structural
transformation, firms, and workers. The authors intend to give a broad overview of the
challenge, identify key policy interventions, and map out a longer-term agenda for analysis
that can inform future policy. The majority of the work is drawn from existing research
carried out by Indonesia’s intellectual community, the World Bank, and other development
partners. The report is organized into six chapters, including this Introduction, with each
chapter focusing on a specific message.

Chapter 2 posits that Indonesia’s challenge has been the disproportionate creation of low-
quality jobs. The chapter has two parts. The first demonstrates how jobs, and particular
kinds of jobs, are a key to middle-class status since they are the primary mechanism for
poverty reduction and middle-class expansion. The second provides a static picture of the
current jobs situation to demonstrate that it is job quality, not quantity, that is Indonesia’s
challenge. It also provides a snapshot of middle-class jobs and identifies factors that may
lead to growth in middle-class jobs.

Chapter 3 explores jobs from the macro perspective. It argues that structural transformation
is not, under current policies, resulting in the increases in labor productivity that are
necessary to increase the number of middle-class jobs. The authors discuss three ideas:
first, how structural transformation has moved Indonesia toward middle-class jobs and
where it has fallen short; second, the evolution of labor productivity and its sources; third,
how manufacturing foreign direct investment (FDI) has affected job creation, productivity,
and structural transformation in Indonesia. The authors note that it is critical to attract
FDI to sectors likely to be a source of middle-class jobs, to sectors that are critical for the
competitiveness of the overall investment climate, for example to utilities or transport
services, and to low-productivity sectors, particularly agriculture, in which many Indonesians
currently work.

Chapter 4 presents the first comprehensive review of the role that manufacturing firms play
in job creation in Indonesia. The chapter is limited to medium-sized and large manufacturing
firms since there are insufficient firm-level panel data to analyze smaller manufacturing
firms or firms in other sectors, especially services. The chapter starts by identifying the
characteristics of the medium-sized and large manufacturing firms that create jobs. It
then explores whether there are employment-productivity trade-offs, whereby the more

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Introduction 27

productive that workers are, the fewer workers are needed. Finally, the chapter turns to the
other end of the spectrum of businesses and explores the extent to which microenterprises
owned by self-employed individuals or households can be a path to middle-class jobs.

Chapter 5 considers labor supply and what workers can do to align their skills with those
required to carry out middle-class jobs. The chapter begins by describing the overall labor
force and compares it to the middle-class labor force. It then explores two ways that workers
could better prepare for middle-class jobs: acquiring the skills needed in middle-class jobs and
improving the efficiency of their job searches. The chapter closes by identifying additional
factors, such as gender wage gaps, that impede the access of vulnerable groups to middle-
class jobs.

Chapter 6 proposes policy directions, which are organized into three groups that cut across
the previous chapters: policies designed to increase productivity across the board; policies
designed to transition firms and workers toward middle-class jobs; and policies designed
to build a middle-class workforce. The authors draw from the policies discussed in the
analytical chapters and reference sectoral studies. These studies are the sources of the
policy recommendations and have been adapted to the Indonesia context. The authors also
draw from sector-specific dialogue with academics, policy experts, and policy makers.

The report is accompanied by an Overview, which gives an overall picture of the job situation
in Indonesia and summarizes key policy recommendations. The Overview explains why better
jobs are necessary to build a middle-class Indonesia, summarizes the evidence surrounding
barriers to the creation of middle-class jobs, and presents in detail a multisectoral and
coherent agenda for reforms to jobs policies.

The report omits several issues. First, agricultural jobs are not explored in detail; they are
presented in a 2020 study.7 Second, the report is a detailed analysis of the current labor
market, but does not explore how megatrends may affect future jobs. Third, the report
does not go into detail on labor legislation, but instead focuses on issues and policies that
are more immediate to the challenge of creating middle-class jobs. Finally, although the
report was completed during the COVID-19 crisis, it does not present a full treatment of
the impact of the pandemic on middle-class jobs in Indonesia. Instead, it references the
pandemic where appropriate.

7 World Bank 2020

Pathways to Middle-Class Jobs in Indonesia June 2021

28 Introduction

Box 1.2 Methodology, Definition, and Data

The report is a combination of empirical evidence and advice from policy experts. The technical analysis draws
from existing published and unpublished papers and complements them with new data work, which is based on
the following data sets (see Annex 1.1 for descriptions of the first three instruments):

• National Labor Force Survey (Survei Angkatan Kerja Nasional; Sakernas), 2001–August 2019
• Survey of Medium-Sized and Large Manufacturing Firms (Survei Industri Besar/Sedang; Manufacturing
Survey), 1990–2015
• Economic Census, 2006 and 2016
• Indonesia Family Life Survey (Survei Aspek Kehidupan Rumah Tangga Indonesia), 2000, 2007, and 2014
• National Socioeconomic Survey (Survei Sosial Ekonomi Nasional; Susenas), 2001–March 2018

The analysis and discussion examine the jobs picture from various perspectives. In short:

• A middle-class job is any job that provided an income equal to or greater than Rp3,752,000 per month in
2018 (adjusted to the corresponding poverty line for other years). The concept is fully defined in Box 1.1.
• Job type defines the employer-employee relationship. Five job types are defined: self-employed, employer,
casual worker, employee, and unpaid. The term “income earner” refers to paid employees and the self-em-
ployed.
• Sector is a categorization of industrial activity. It may apply to the firm or the worker. The authors use 9-
and 17-sector classifications, as defined in Annex Table 1.2.
• Occupation is a categorization of job titles, based on the Indonesian standard classifications found in the
Indonesia Standard Job Title Classificationa (Klasifikasi Baku Jenis Indonesia; KBJI). The highest level of cat-
egorization (containing the fewest categories, categorized by one-digit KBJI) includes nine occupations, as
defined in Annex I: Annex Table 1.3. Occupational classification is based on the KBJI 2002 (ISCO 88)b
• Institution concerns the type of institution in which a person works. There are seven types of institution:
government, international, nonprofit, profit (including private, state-owned, and local government-owned
firms), cooperative, individual/household enterprises (HHEs), and household. Data on institution type has
been collected since Sakernas 2016.
• Formality concerns whether a firm is legally registered with the state or a worker is covered by labor legis-
lation. The definitions follow the International Conference of Labour Statisticians (ICLS)-17 and are given in
full in Annex 1. In some analysis, the authors use a simpler matrix of occupation and job status as a measure
of formality. This latter definition of formality is the approach taken by the National Statistical Office (Badan
Pusat Statistik; BPS) , and the authors highlight instances in the report when they too adopt it.
The policy analysis was developed by a team of Indonesian experts, and the work was also informed by discussions
with sectoral colleagues and global experts. The analysis relies heavily on policy directions detailed in reports
that deeply explore sector-specific constraints on the issues relevant to Indonesia’s challenge of creating middle-
class jobs.

a BPS 2002

b Eurofound 2010

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Introduction 29

Annex 1 Data Sources and Definitions

Data Sources
The original data work in the study is based around three principal surveys.

The National Labor Force Survey (Survei Angkatan Kerja Nasional; Sakernas). Sakernas
is conducted annually by the National Statistical Agency (Badan Pusat Statistik; BPS). Data
is collected from individuals on a limited set of demographic, education, and economic
indicators. This is the primary tool the government uses to monitor labor market activity.

The survey has changed over time, limiting cross-year comparisons. While the most basic
indicators are available and comparable for the period 2001–18, a fuller set of variables is
available and comparable for the period 2007–18. There are periodic modifications to the
survey, which have enriched the data used in this report.

The Economic Census. The Economic Census is conducted every 10 years by the BPS. The
aim of this census is to identify the Indonesian population and update the directory of
firms, mapping the distribution. The BPS collects information about firms, such as profile,
production costs, capital, and income structure, their prospects and the obstacles they face,
and captures nonfarm micro, small, medium, and large enterprises. In 2016, 26,711,001 firms
falling into 15 sectors were interviewed; of those, 98.3 percent were classified as micro or
small enterprises.8

The Survey of Medium-Sized and Large Manufacturing Firms (Survei Industri Besar/
Sedang; Manufacturing Survey). Each year, the BPS conducts the Manufacturing Survey,
which covers manufacturing establishments or plants with at least 20 workers. The survey
is conducted at the establishment or plant level, rather than at firm level. Estimates of
the incidence of multiplant firms indicate that only around 5–7 percent of plants are part
of a larger firm.9 Hence, treating a plant as a firm in the analysis seems reasonable. This
report uses data from 1990 to 2015 (the latest available). The number of plants in the
survey increased from 16,494 to 26,322 between 1990 and 2015. During the firm census
year (2006), the number of plants covered in the Manufacturing Survey went up.

Type of Job
Unless otherwise specified, labor force participation categories in this report are based on
the primary economic activity performed by the respondent in the seven days prior to the
survey, as defined in the Sakernas Enumerator’s Manual.

Employee: An employee works permanently for another person or institution, office, or

company and receives wages or a salary in the form of money or goods.

8 BPS 2016
9 Asian Development Bank 2019

Pathways to Middle-Class Jobs in Indonesia June 2021

30 Introduction

Casual worker: A casual worker works for another person, employer, or institution in a
nonpermanent capacity (and may have multiple employers in any month). A casual worker
receives remuneration in the form of money or goods, paid daily or by piece rate. A casual
worker may be a farm worker or a nonfarm worker.

Unpaid family worker: An unpaid family worker works for a person, institution, office, or
company (they may or may not be a member of the same household) and does not receive
a wage or salary, whether in cash or in kind.

Self-employed: An individual who is self-employed bears all the economic risk of their job,
including not returning the cost of production that has been incurred, and is not assisted
by paid or unpaid workers.

Employer: An employer is an individual who works at their own risk and is assisted by
permanent, nonpermanent, family, or unpaid worker(s).

Sectors of Employment
The Sakernas provides industry codes derived from the Indonesian Industrial Classification
System (Klasifikasi Baku Lapangan Usaha; KBLI) twice a year. The KBLI builds on the International
Standard Industrial Classification system established by the United Nations.10 The industry
codes reported in the Sakernas are not consistent across years. Some years of Sakernas
present very aggregated sector codes (one-digit codes for 9 or even 17 sectors), while
others present great detail (five-digit codes for each occupation). Annex Table 1.1 provides
the levels of detail and source code book for each corresponding year of the Sakernas.
Annex Table 1.2 provides the classification of one-digit sectors based on the 9- or 17-sector
groupings

Annex Table 1.1 Industry Codes in Sakernas

Number of
Period KBLI/KLUI Number of digits Based on
sectors
2001–06 KBLI 2000 3 9 ISIC Revision 3
2007–10 KBLI 2005 5 9 KBLI 2000
2011–15 KBLI 2009 2 9 ISIC Revision 4
2016–17 KBLI 2015 1 17 KBLI 2009
2018 KBLI 2015 5 17 KBLI 2009
2019 KBLI 2015 1 17 KBLI 2009

Sources: Sakernas 2001–19; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional). The KBLI is the Indonesian
Industrial Classification System (Klasifikasi Baku Lapangan Usaha). The KLUI is the Indonesia Business Classification
(Klasifikasi Lapangan Usaha Indonesia). The ISIC is the International Standard Industrial Classification. The higher the
number of digits, the greater the level of disaggregation.

10 United Nations 2008

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Introduction 31

Annex Table 1.2 Conversion Table 17-Sector to 9-Sector

17-sector 9-sector

Agriculture, forestry, livestock, and fisheries Agriculture, forestry, livestock, and fishing
Mining and quarrying Mining and quarrying
Manufacturing Manufacturing
Electricity and gas supply Electricity, gas, and water supply
Water supply, sewerage, and waste management and
remediation
Construction Construction
Wholesale and retail trade; repair of motor vehicles and Wholesale and retail trade, restaurants and
motorcycles hotels
Accommodation and services activities
Transportation and storage Transportation, storage and communications
Information and communication
Financial and insurance activities Finance, insurance, real estate and business
Real estate services

Business activities
Public administration and defense; compulsory social security Community, social and personal services
Education
Human health services and social work activities
Other service activities

Sources: World Bank staff calculations

Occupation Categories
The Sakernas provides the occupational classification codes used in each survey. The
classifications were developed by the BPS and are found in the Standard Classification of
Occupational Types (Klasifikasi Baku Jenis Pekerjaan; KBJI), formerly known as the Indonesian
Classification of Occupations (Klasifikasi Jabatan Indonesia; KJI). The occupational codes build
on the International Standard Classification for Occupations. The one-digit occupations are
as follows:11

Agriculture, forestry, livestock, and fisheries: This category includes subsistence

farmers with crops and/or livestock, as well as market gardeners, and other crop growers.

Elementary occupations: These are usually unskilled occupations in agriculture,

manufacturing, and the service sector. In Indonesia, these occupations typically include
manual labor in agriculture, construction, and transport.

Machine operators: This category includes plant and machine operators and assemblers,

11 The more digits in the occupational code, the more disaggregated the data. Hence, one-digit codes are the most aggregat-
ed form of occupational classification.

Pathways to Middle-Class Jobs in Indonesia June 2021

32 Introduction

and drivers. Typical occupations are drivers of motorcycle taxis and other vehicles including
heavy trucks, and operators of mobile plants and machines.

Craft and related trades: This category includes workers who produce foodstuffs,
textiles, and other articles by applying their specific skills and knowledge. They typically
work with their hands or with hand tools. Common occupations in this category are workers
constructing the frames of buildings, food processing workers, wood treaters, cabinet
workers, and those in related trades.

Services and sales: This category includes personal services workers, sales workers, and
personal care and protective services workers. Common occupations in this category are
shop salespersons, street and market salespersons, waiters and bartenders, protective
service workers, street food sellers, police officers, hairdressers, and beauticians.

Clerical: This category includes general and keyboard clerks, customer services clerks,
and numerical and material recording clerks, among others. Common clerical occupations
include general office clerks, tellers, and client information workers.

Managers, professionals, and technicians: Common occupations in this category include

schoolteachers, legislators, finance professionals, and managers in various sectors.

The KJI/KBJI classifications change over time, as presented in Annex Table 1.3.

Annex Table 1.3 Occupational Codes Used in the Sakernas

Sakernas years KJI/KBJI Number of digits Based on

2001–06 KJI 1982 3 ISCO 1968

2007–15 KBJI 2002 4 ISCO 1988
2016–19 KBJI 2014 1 ISCO 2008

Sources: Sakernas 2001–19; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional). The KJI is the Indonesian
Classification of Occupations (Klasifikasi Jabatan Indonesia). The KJI became the KBJI, the Standard Classification
of Occupational Types (Klasifikasi Baku Jenis Pekerjaan). The ISCO is the International Standard Classification of
Occupations.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Introduction 33

Formality/Informality
Whether a job or worker is formal or informal is established by the BPS in a cross-tabulation
between job status and occupation (see Annex Table 1.4).

Annex Table 1.4 Cross-Tabulation between Job Status and Occupation

Occupation (based on KJI 1982/ISCO-68)

Agriculture
Production,
Professional, Administrative Clerical and forestry
transport
Status technical, and and Sales Services workers,
equipment Other
and related managerial related workers workers tenders of
operators,
workers workers workers livestock,
and laborers
fishers

Self-employed Formal Formal Formal Informal Informal Informal Informal Informal

Employer assisted Formal Formal Formal Formal Formal Informal Formal Informal
temporary/unpaid
worker(s)
Employer assisted by Formal Formal Formal Formal Formal Formal Formal Formal
permanent worker(s)
Employee Formal Formal Formal Formal Formal Formal Formal Formal
Casual worker in Formal Formal Formal Informal Informal Informal Informal Informal
agriculture sector
Casual worker in Formal Formal Formal Informal Informal Informal Informal Informal
nonagriculture
sector
Unpaid family worker Informal Informal Informal Informal Informal Informal Informal Informal

Sources: Sakernas 2018 Enumerator’s Manual

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional). The KJI is the Indonesian
Classification of Occupations (Klasifikasi Jabatan Indonesia). The ISCO is the International Standard Classification
of Occupations.

Some sections of this report use the new definitions of formal and informal proposed by
the International Labour Organization (ILO) based on the International Conference of Labor
Statisticians-17,12 which is constructed from variables presented in the Sakernas as of
February 2016 or later. The new definitions are based on a combination of job status, type
of accounts kept, social benefits, and type of institution.

A formal unit is defined as:

• government, an international organization, a nonprofit organization, a profit

organization, or a cooperative
OR
• an individual or household enterprise (HHE), any other unit type not mentioned
previously, or the survey respondent does not know, and has complete accounts.

12 This approach was presented by the BPS at a workshop on September 25, 2019.

Pathways to Middle-Class Jobs in Indonesia June 2021

34 Introduction

An informal unit is defined as:

• an individual or HHE, any other unit type, or the survey respondent does not know, and
does not have complete accounts
OR
• a household unit.
A formal job or worker is defined as:
• a self-employed individual, or an employer assisted by temporary, unpaid, or permanent
workers in a formal production unit
OR
• an employee or casual worker who receives at least one social benefit and paid leave
OR
• an employee or casual worker who receives neither social benefits nor paid leave but has
a permanent contract.

An informal job or worker is defined as:

• a self-employed individual, or an employer assisted by temporary, unpaid, or permanent
workers in an informal unit or household production unit
OR
• an unpaid/family worker
OR
• an employee or casual worker who does not receive social benefits and does not have a
permanent contract
OR
• an employee or casual worker who receives at least one social benefit but does not
receive paid leave and does not have a permanent contract.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 References 35

References
Alatas, Hamidah, and Maria Monica Wihardja. Forthcoming. “The Demand Side of Jobs in
Indonesia: Plant-Level Analysis in the Medium and Large Manufacturing Industry.”
Background paper, Pathways to Middle-Class Jobs in Indonesia, World Bank, Jakarta.

Asian Development Bank. 2019. Policies to Support the Development of Indonesia’s Manufacturing
Sector during 2020–2024: A Joint ADB–BAPPENAS Report. Asian Development Bank
Publications.

BPS. 2002. KBJI 2002: Klasifikasi Baku Jenis Pekerjaan Indonesia. https://sirusa.bps.go.id/
webadmin/doc/KBJI2002.pdf.

BPS. 2016. https://se2016.bps.go.id/umkumb/index.php/metadata/index.

Brummund, Peter, Christopher Mann, and Carlos Rodriguez-Castelan. 2016. “Job Quality
and Poverty in Latin America.” Policy Research Working Paper 7927, World Bank,
Washington, DC.

Eurofound. 2010. “Coding and Classification Standards.” December 23. https://www.

eurofound.europa.eu/surveys/ewcs/2005/classification.

Office of Assistant to Deputy Cabinet Secretary for State Documents and Translation. 2020.
“World Bank: Indonesia Is Now Upper Middle Country.” July 20. Cabinet Secretariat
of the Republic of Indonesia. https://setkab.go.id/en/world-bank-indonesia-is-now-
upper-middle-country/.

United Nations. 2008. International Standard Industrial Classification of All Economic Activities
(ISIC), Rev. 4. Statistical Papers, Series M, No. 4, Rev. 4. https://unstats.un.org/unsd/
publication/seriesm/seriesm_4rev4e.pdf.

World Bank. 2019. Aspiring Indonesia: Expanding the Middle Class. Washington, DC: World Bank.

World Bank. 2020. No One Left Behind: Rural Poverty in Indonesia. Washington, DC: World Bank.

Xinhua. 2021. “Indonesia to be Lower-Middle-Income Country Due to Pandemic: Minister.”

February 9. http://www.xinhuanet.com/english/2021-02/09/c_139733430.htm.

Pathways to Middle-Class Jobs in Indonesia June 2021

36 Jobs as a Catalyst for Economic Growth and a Middle-Class Society

Chapter 2

Jobs as a Catalyst for

Economic Growth and a
Middle-Class Society 
June 2021 Pathways to Middle-Class Jobs in Indonesia
 Jobs as a Catalyst for Economic Growth and a Middle-Class Society 37

2.1 Overview of Macroeconomic Trends

Indonesia is not a story of jobless growth. From 1966 to 1996, Indonesia’s economic
growth averaged 7.3 percent annually. It then contracted by 13.4 percent in 1998 and, from
2000 onward, plateaued at a “new normal” of an average of 5.1 percent annually.1 Since
close to the end of the commodity boom, growth has followed a downward trend, from
around 6.2 percent in 2011 to around 5.2 percent by 2018 (Annex Figure 2.1).2 Despite the
growth slowdown, Indonesia’s job creation remained robust and significant progress has
been made. Between 1999 and 2018, an average of 1.8 million additional jobs were created
every year. The employment rate reached a two-decade record high in 2018, while the
unemployment rate continued to fall (Annex Figure 2.2).

However, the growth slowdown in the manufacturing sector since the Asian
financial crisis in 1997 and 1998 also slowed the creation of “good” jobs. Growth in
the manufacturing sector slowed from an average 10.9 percent during the 1990–96 period
to around 4.6 percent during 2000–10 and 4.7 percent during 2010–18. Job opportunities in
the sector were cut just as a flood of people from rural areas were migrating to towns and
cities to find better-paying jobs. Between 2000 and 2010, the compound annual growth rate
(CAGR) of employment in manufacturing “hollowed out” to an annual 1.73 percent, compared
to 5.96 percent between 1990 and 1996, and 3.30 percent between 2011 and 2017.

1 Hill and Negara 2019, Chapter 1

2 Asian Development Bank 2019, Chapter 2

Pathways to Middle-Class Jobs in Indonesia June 2021

38 Jobs as a Catalyst for Economic Growth and a Middle-Class Society

Instead, job growth in the post–Asian financial crisis period came through the
tertiarization of the economy and from low-value services, slowing the creation
of middle-class jobs. By 2018, only one in five (the figure is, however, growing) Indonesians
had the economic security to join the middle class, while the share of aspiring middle class—
that is, Indonesians who have escaped poverty and vulnerability but lack the economic
security to join the middle class—has hovered at around half of the population for the past
two decades.3, 4 While there were around 59 million middle-class Indonesians in 2018, there
were still 124 million who had escaped poverty, but did not have the economic security to
join the middle class. Contentiously, the lack of new middle-class jobs creates a backlog of
Indonesians aspiring to move up to the middle class but who are unable to do so.

Low wages reflect the poor quality of jobs in Indonesia. Most jobs created between
2008 and 2018 have been in low-wage sectors, where the median nominal wage for income
earners5 is far below the minimum wage necessary to join the middle class (around Rp3.7
million per month) (Figure 2.1). Almost half of jobs (45.8 percent) created since 2008 have
accrued to low-value, low-wage services, namely wholesale and retail trade, accommodation,
food and beverages, and other services. The low wages reflect the poor quality of jobs in
Indonesia. That said, low-paid agricultural jobs have contracted substantially over the same
period.

Figure 2.1 Where Have Jobs Been Created?

Median nominal wage 2018, job creation 2008–18, share of employment 2018 (bubble size)

Sources: Sakernas 2008–18; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

3 World Bank 2019

4 The poor are defined as those who live below the poverty line. The vulnerable are defined as those who have more than
10 percent probability of falling into poverty (and a monthly household consumption around 1–1.5 times the poverty line).
Members of the aspiring middle class are defined as those who have more than 10 percent probability of falling into poverty and
vulnerability (and a monthly household consumption around 1.5–3.5 times the poverty line). Members of the middle class are
defined as those who have less than 10 percent probability of falling into poverty and vulnerability (with a monthly household
consumption around 3.5–17 times the poverty line). Members of the upper class are defined as those whose consumption is
more than 17 times the poverty line.
5 Income earners include the self-employed, wage employees, and casual workers. Income earners excludes employers and
unpaid family workers.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Jobs as a Catalyst for Economic Growth and a Middle-Class Society 39

In this chapter, we argue that the shift to low-value jobs has resulted in less economic
growth, greater poverty, and greater inequality than would have occurred had the
structural transformation toward good jobs not been interrupted. First, the report
discusses how jobs are a key to middle-class status since they are the primary mechanism
for poverty reduction and for expansion of the middle class. Second, the report explores
the overall job situation to demonstrate that it is job quality, not quantity, that is Indonesia’s
challenge. The report also looks at the structure of middle-class jobs overall. Third, the
report analyzes drivers of earnings increases in Indonesia over the past two decades.

2.2 Why Jobs?

Indonesia now has a poverty rate in the single digits and is aspiring to become a
high-income country by 2045; jobs, especially middle-class jobs, are becoming ever
more important. In order to better understand the contribution of labor income (vis-à-vis
nonlabor income) to poverty reduction, the report explores the contribution of different
economic sectors and types of employment to the reduction of the share of jobs carried out
by the poor and those vulnerable to poverty (poor- and vulnerable-class jobs), as well as the
expansion of the share of middle-class jobs in Indonesia.

2.2.1 Labor income is the main driver of poverty reduction

Labor income is becoming a more important contributor to poverty reduction

as Indonesia transitions to higher income levels.6 Between 2000 and 2014, while
nonlabor income (including income from cash-transfer programs and other forms of social
assistance) was the key contributor to poverty reduction for the extreme poor (extreme
poverty means an income below US$1.90 a day, which is also the poverty line for low-income
countries), labor income began to take over as the key contributor to poverty reduction at a
poverty line of around US$3.40 a day (slightly above the poverty line for low-middle-income
countries).7 This highlights that the creation of good-quality jobs, alongside the provision
of social assistance, should be a principal policy tool as Indonesia transitions out of the low-
income classification (US$1.90 a day poverty line) and away from its status as a lower-middle-
income country (US$3.20 a day poverty line) and moves toward becoming an upper-middle-
income country (US$5.50 a day poverty line) and a high-income country (US$21.70 a day
poverty line). Indonesia hopes to achieve this latter goal by 2045.

6 Following Azevedo et al. (2013), the authors use the Shapley decomposition analysis. A series of path-dependent
counterfactual simulations— changing one component at a time and keeping the others constant—are generated to estimate
the average contribution of each component to poverty reduction. The components considered are labor income, nonlabor
income, the share of adults in the household, and the consumption-to-income ratio.
7 The time period 2000–14 was selected based on the availability of data in the Indonesian Family Life Survey. The authors
also ran the analysis for the two subperiods 2000–07 and 2007–14. See Annex 2.2 on the Shapley decomposition results for
the subperiod analysis.

Pathways to Middle-Class Jobs in Indonesia June 2021

40 Jobs as a Catalyst for Economic Growth and a Middle-Class Society

Figure 2.2 Labor Income Starts to Take Over as the Key Contributor to
Poverty Reduction at Around US$3.40 a Day Poverty Line
Labor and nonlabor income contribution to poverty reduction, 2000–14

Sources: Indonesian Family Life Survey 2000 and 2014; World Bank staff calculations

2.2.2 Certain types of jobs help reduce poverty and vulnerability

If Indonesia is to expand its middle-class population, better-quality jobs—not

just any jobs—are needed. This report examines which economic sectors, types of
employment, and level of formality help people out of poverty and vulnerability. The report
uses the Shapley decomposition methodology8 and income data for income earners from
the National Labor Force Survey (Survei Angkatan Kerja Nasional; Sakernas). Table 2.1 shows
how different sectors, types of employment, and formality each contributed to helping
Indonesians escape poverty and vulnerability and enter middle-class jobs. The last column
shows the difference in the contributions: negative signs show jobs that can lift Indonesians
out of poverty and vulnerability, but are insufficient to give them a middle-class level of
income, while positive signs show the jobs that can bring Indonesians into the middle class.

Certain sectors and job types are particularly effective at moving workers out of
poverty. Being a wage employee in industry or services or being a formal sector employee
contributed to almost half of the decline in poverty and vulnerability (Annex Figure 2.5–
Annex Figure 2.7). The sectors that were most conducive to lifting people out of poverty were
low-value services (wholesale and retail trade, restaurants and hotels), which contributed
about one-quarter of the decline in the share of poor- and vulnerable-class jobs, followed
by manufacturing (19 percent), other industries (15 percent from construction), high-value
services (15 percent), and agriculture, forestry, livestock, and fisheries (13 percent). Half
of the decline in the share of poor- and vulnerable-class jobs between 2011 and 2018 was

8 A decline in the share of poor- and vulnerable-class jobs is used as a proxy for the decline in poor-quality jobs. The authors
use the period immediately following the commodity boom for the analysis.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Jobs as a Catalyst for Economic Growth and a Middle-Class Society 41

due to wage employment (Annex Figure 2.6). Self-employment and casual employees each
contributed an additional quarter. Formal employment contributed 57 percent to the
decline in the share of poor- and vulnerable-class jobs. This was predominantly driven by
industry and the service sector (Annex Figure 2.7).

Taking these together, the biggest contribution came from wage employment in
the service sector, followed by wage employment in industry. Self-employment in the
service sector and casual jobs in industry also contributed significantly to lifting Indonesians
out of poverty and vulnerability. The agriculture sector, whether wage employment, self-
employment, or casual work is contributing less to the decline in poor- and vulnerable-class
jobs than are self-employment in industry and casual jobs in services.

2.2.3 Other types of jobs contribute to the creation of middle-class jobs

Increases in middle-class jobs are driven by different economic sectors, types of

employment, and levels of formality than those responsible for the decline in poor-
and vulnerable-class jobs. Using a similar methodology as that used in Section 2.2.2, the
authors decomposed the increase in the share of middle-class jobs which, between 2011 and
2018, was 5.7 percentage points.

The manufacturing sector, wage employment, and formal sector employment

were the primary contributors to the increase in the share of middle-class jobs
between 2011 and 2018 (Annex Figure 2.8–Annex Figure 2.10). The manufacturing
sector contributed 27 percent of the increase in the share of middle-class jobs between
2011 and 2018, closely followed by wholesale and retail trade, restaurants and hotels (25
percent). Low-productivity sectors, namely agriculture and community, social, and personal
services, contributed almost nothing to the increase in the share of middle-class jobs.

Wage employment contributed two-thirds of the increase in the share of middle-

class jobs for income earners. Wage employment in industry (32 percent), services (27
percent), and agriculture (6 percent) made a bigger contribution to the growth of middle-
class jobs than to the decline in the share of poor- and vulnerable-class jobs (Annex Figure
2.9). Self-employment in services contributed 11 percent, while self-employment in other
economic sectors did not contribute significantly.

Formal employment contributed more significantly to the increase in the share of

middle-class jobs than to the decline in the share of poor- and vulnerable-class jobs
(Annex Figure 2.10). Formal employment contributed almost three-quarters of the increase
in the share of middle-class jobs, with formal employment in industry and services alone
contributing almost two-thirds. This is higher than the contribution of formal employment
to the decrease in the share of poor- and vulnerable-class jobs (only 57 percent).

Pathways to Middle-Class Jobs in Indonesia June 2021

42 Jobs as a Catalyst for Economic Growth and a Middle-Class Society

Wage employment in industry drives expansion of the middle class, but may not be
widely accessible to Indonesians trying to escape poverty or economic vulnerability.
Wage employment in industry (especially manufacturing) and services (especially wholesale
and retail trade, restaurants and hotels) contributed the most to the increase in the
share of middle-class jobs in Indonesia between 2011 and 2018. Jobs in certain sectors—
the agriculture, forestry, livestock, and fisheries sector, especially self-employment in this
sector; the community, social and personal services sector; self-employment in the service
sector; and self-employment and casual jobs in industry—do not move workers to the middle
class (although they do move workers out of poverty and vulnerability).

Table 2.1 Some Jobs Are Successful at Bringing Indonesians out of Poverty and Vulnerability,
but Cannot Give Them a Middle-Class Income
Contribution to the reduction of poor- and vulnerable-class jobs and the expansion of middle-class jobs

Contribution to
Contribution to
the reduction
the expansion in
in poor- and Difference (%)
middle-class jobs
vulnerable-class
(%)
jobs (%)

Income: agriculture, forestry, livestock, and fishing 12.7 0.3 –12.3

Income: mining and quarrying 3.5 2.7 –0.8
Income: manufacturing 18.7 27.1 8.3
Income: electricity, gas, and water supply 4.9 6.8 1.9
Income: construction 15.2 15.6 0.4
Income: wholesale and retail trade, restaurants and hotels 23.9 24.6 0.7
Income: transportation, storage and communications 8.0 10.9 2.9
Income: finance, insurance, real estate, and business services 6.8 10.9 4.1
Income: community, social, and personal services 6.3 1.2 –5.1

Income: agriculture, self-employment 6.7 –1.2 –7.9

Income: agriculture, employee 6.5 5.5 –0.9
Income: agriculture, casual worker 6.6 5.6 –1.1
Income: industry, self-employment 6.6 4.9 –1.7
Income: industry, employee 20.9 31.8 10.9
Income: industry, casual worker 11.2 10.6 –0.6
Income: service, self-employment 13.2 10.9 –2.3
Income: service, employee 23.8 26.7 2.9
Income: service, casual worker 4.5 5.1 0.6

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Jobs as a Catalyst for Economic Growth and a Middle-Class Society 43

Table 2.1 continued

Income: agriculture, formal 8.3 7.9 –0.3

Income: agriculture, informal 11.5 2.0 –9.4
Income: industry, formal 22.8 34.3 11.5
Income: industry, informal 15.9 13.0 –2.9
Income: service, formal 26.3 30.0 3.7
Income: service, informal 15.2 12.6 –2.6

Sources: Sakernas 2011 and 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

2.2.4 Jobs contribute to economic growth

The creation of middle-class jobs for all workers is critical for economic growth.
Experiences from other countries show that labor productivity drives economic growth,
but economic growth does not always bring good-quality jobs.9 Hence, policies that are
designed to boost economic growth are insufficient to bring good-quality jobs, especially
to the bottom 80 percent of the population who are still not economically secure. Instead,
labor productivity growth has to be the main driver of Indonesia’s higher economic growth
as the population ages. Jobs that command higher salaries and are compatible with the
economic security of the middle-class population are also higher productivity jobs and
commensurate with economic growth. Policies also need to crowd in Indonesians regardless
of age, sex, location, or other factors if middle-class jobs are to go hand in hand with faster
growth and a reduction in inequality. Chapter 3 will discuss in detail how jobs contribute to
economic growth in Indonesia.

2.3 What Jobs Look Like

2.3.1 Quantity of jobs

There were 124 million jobs in Indonesia in 2018. Around 28.8 percent were in the
agricultural sector, 23.2 percent were in industry, and 48.0 percent were in services. This
represents a significant shift over three decades, as employment in Indonesia diversified
away from agriculture (Figure 2.3).

9 Merotto and Weber 2018

Pathways to Middle-Class Jobs in Indonesia June 2021

44 Jobs as a Catalyst for Economic Growth and a Middle-Class Society

Figure 2.3 Employment in Indonesia Diversifies Away from Agriculture

Share of economic sector by age group, 1990 versus 2018

Sources: Sakernas 1990 and 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

The sectors that created the most jobs over the past 10 years differ depending on
the measure used. Approximately 2.1 million jobs were created annually during 2008–18.
The largest number of new jobs were in manufacturing, wholesale and retail trade, and
vehicle repair. However, the highest percentage-point increases in the share of employment
were in manufacturing and low-value services (accommodation, food and beverages, and
education activities). The high-value service sector and utilities experienced the highest
CAGR in the last decade (see Figure 2.4). This sector includes finance and insurance activities;
real estate and business activities; human health and social work activities; and electricity
and gas, water supply, sewage, and waste management and remediation. The job-creation
prospects of the high-value services are higher compared to those of other sectors, but the
sector starts from a very low base (see Annex Table 2.1). If sectors are divided into low-wage,
mid-wage and high-wage paying,10 the high-wage sector experienced the highest CAGR,
although the mid-wage sector created the most jobs (see Annex Table 2.2). Meanwhile, the
shares of total employment attributable to high-skill and mid-skill white-collar workers have
increased (Annex Figure 2.11).

10 The authors divided the sectors into groups based on median earnings. The low-wage sector has a median monthly wage
below Rp2,000,000. It consists of agriculture, forestry, livestock, and fisheries; water supply, sewerage, and waste management
and remediation; accommodation and food service activities; wholesale and retail trade; vehicle repair; and other service
activities. The mid-wage sector has a median monthly wage below Rp2,500,000 and above Rp2,000,000. It consists of
construction; transportation and storage; education activities; and manufacturing. The high-wage sector has a median monthly
wage above Rp2,500,000. It consists of financial and insurance activities; public administration, defense, and compulsory social
security; real estate activities; information and communications; electricity and gas; business activities; human health and social
work activities; and mining and quarrying.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Jobs as a Catalyst for Economic Growth and a Middle-Class Society 45

Figure 2.4 Although Making Up a Small Share of Jobs, the High-Value Service Sector Has High Job Prospects 
CAGR and percentage-point change in employment share by economic sector, 2008–18

Sources: Sakernas 2008 and 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

The growth in different type of jobs from 2008 to 2018 reveals a dual nature to the
labor market in Indonesia. On one hand, Indonesia has experienced positive growth for
nonagriculture casual employment (possibly including gig workers) and self-employment,
both of which are associated with vulnerable jobs (less likely to have a formal arrangement,
social security, or a voice through a union, and more likely to be low productivity). On the
other hand, there has also been positive growth both for employers with permanent workers
and for wage employment, both of which are associated with more formal jobs. The wage
employee sector had by far the highest CAGR and percentage-point increase in both share
of employment and absolute change in job creation (Figure 2.5).

Pathways to Middle-Class Jobs in Indonesia June 2021

46 Jobs as a Catalyst for Economic Growth and a Middle-Class Society

Figure 2.5 Duality of Indonesia’s Labor Market Indicated by High Growth in Wage Employment and Positive
Growth in Nonagriculture Casual Employment and Self-Employment
CAGR and percentage-point change in employment share by employment status, 2008–18

Sources: Sakernas 2008 and 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

2.3.2 Quality of jobs

Job quality is an elusive concept, since a quality element viewed by one worker as
important may be unimportant for another. That said, Indonesia’s jobs are low quality
across different elements. This section reviews job quality using a seven-variable job-quality
index, occupation type, sector, and formality status.

Most jobs in Indonesia are unstable and do not protect workers (Figure 2.6).11 In 2019,
38.2 percent of all workers were firm owners, while 61.8 percent were employees. Among
firm owners, more than half were self-employed. Among employees, almost two-thirds were
working without any work contract (wage employee without a written contract, casual
worker, or unpaid family worker), while another one-fifth were working under fixed-term (or
temporary) contracts. Uncontracted jobs tend to be lower paid and riskier, and do not carry
social benefits. However, even contracted wage employees face vulnerability due to low
compliance with Indonesia’s rigid worker protection policies. In 2019, only 43.5 percent of
wage employees with permanent contracts (10 percent of total wage employees) received
full worker protection benefits and payment above the minimum wage (see Annex Table 2.3).
Compared to other countries in the region, Indonesia’s share of vulnerable workers (based
on the definition of “vulnerable” workers issued by the International Labour Organization as
being those in self-employment or unpaid family workers) is relatively high—higher in 2018
than the Philippines, Malaysia, and Singapore, lower than Vietnam, and on par with Thailand
(see Annex Figure 2.12).

11 See the Annex to Chapter 1 for the definition for each employment type.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Jobs as a Catalyst for Economic Growth and a Middle-Class Society 47

Figure 2.6 Less Than 10 Percent of Jobs Are Stable and Protect Workers 
Distribution of jobs by employment status

Sources: Sakernas 2019; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Between 2007 and 2015,12 one-third of jobs created were high skill and 22 percent
were low skill, while mid-skill jobs (blue- and white-collar combined) amounted to
less than 50 percent.13 Triangulating the evidence for job creation by occupation and sector
shows that low-value services absorbed both high-skilled workers (for example, managers)
and the low skilled (for example, those working in elementary occupations). Figure 2.7 shows
that in terms of occupation, the biggest transformation comes from skilled agricultural
and fishery workers (−18 percent) (mid-skill blue-collar occupation) moving into elementary
occupations (22 percent) (low-skill occupation) or becoming service and sales workers (27
percent) (mid-skill white-collar occupation). The changes in occupation also affect demand
for different types of human skill (for example, analytical, cognitive, interpersonal, manual
routine and nonroutine), which will be discussed more in Chapter 5.

12 Data on occupation by skill level is consistent only for this period of time, that is, 2007–15.
13 See the Annex to Chapter 1 for definitions of each occupation category. Reclassification of one-digit-level occupational
employment into four skill categories: high skill: managers, professionals, technicians and associate professionals; mid skill
white-collar: clerical support, service and sales workers; mid skill blue-collar: skilled agricultural, forestry and fishery workers,
craft and related trade workers, plant and machine operators, and assemblers; low skill: elementary occupations.

Pathways to Middle-Class Jobs in Indonesia June 2021

48 Jobs as a Catalyst for Economic Growth and a Middle-Class Society

Figure 2.7 Occupational Transformation from Skilled Agricultural and Fishery Jobs to Elementary Occupations
and Service and Sales Workers
Median of nominal wage 2015, job creation by occupation, 2007–15, share of employment 2015 (bubble size)

Sources: Sakernas 2007 and 2015; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

About three-quarters of Indonesia’s labor market is informal. Indonesia’s formality

rate dropped to 25.1 percent under a definition based on the International Conference of
Labor Statisticians (ICLS)-17 from 44.3 percent under a simplified definition or 51.5 percent
under the official definition.14 Almost all jobs in agriculture are informal (96.6 percent),
while jobs in education, human health services, and social work (which include teachers and
doctors working as civil servants), as well as finance and insurance are largely formal (Annex
Table 2.4). These informal jobs are largely uncovered by any worker protection regulation
or labor policy. In contrast to formal workers, individual-level data (by name and address) on
where Indonesia’s informal workers work and who they are is largely absent.15

Who are Indonesia’s informal workers? Informal workers are owners of household
enterprises (HHEs), employees, or casual workers without any written work contract
working in HHEs, and unpaid family workers working in HHEs (Table 2.2). Three types of
workers account for approximately 90 percent of Indonesia’s informal workers. Almost half
of informal workers (48.5 percent) are owners of individual enterprises or HHEs (see also
Box 2.1). They are self-employed or employers working in individual enterprises or HHEs and
do not separate their business accounts from the family financial accounting. Another 26
percent are employees and casual workers without a written contract working in HHEs. The
final 15.3 percent are unpaid family workers working in HHEs.

14 See the Annex to Chapter 1 for the revised definition of formality/informality.

15 Consequently, in times of economic crisis, such as that triggered by the COVID-19 pandemic, it is extremely difficult for
the government to assist informal workers.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Jobs as a Catalyst for Economic Growth and a Middle-Class Society 49

Table 2.2 Who are Indonesia’s Informal Workers? 

Share of total
Number of
Type of workers number of informal
workers
workers (%)

Enterprise owners Household enterprise (HHE) owners (self employed) 24,076,130 25

HHE owners with temporary workers or unpaid family workers 18,137,596 19
HHE owners with permanent workers 3,738,569 4
Employees In HHEs
Wage employees 14,159,638 15
Casual workers 10,464,192 11
Unpaid family workers 14,540,420 15
Others
Wage employees 7,200,144 8
Casual workers 1,428,401 2
Unpaid family workers 49,849 0
Others 977,107 1
Total informal workers 94,772,046 100

Sources: Sakernas 2019; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Box 2.1 What Are Nonfarm Household Enterprises?

To give a flavor of nonfarm household enterprises (HHEs), the authors used the National Labor Force Survey (Survei
Angkatan Kerja Nasional; Sakernas) 2018 to generate the top 10 five-digit economic sectors in which nonfarm
HHEs operate (see Table 2.3 below). Most nonfarm HHEs are low-value food and beverages retailers or operate
as motorcycle ride-hailing drivers. Some also operate in tailoring and in motorcycle repair and maintenance.

Table B2.1.1 Top Ten Five-Digit Sectors for Nonfarm HHEs

Five-digit KBLI code Sector Number of HHEs

47112 Retail trade of various items mainly food, beverage or tobacco 3,285,594
not in supermarkets or minimarket (traditional)
56103 Food stalls (kedai makanan) 1,167,102
49424 Motorcycle ride-hailing transportation 1,112,180
56102 Low-end food stalls (warung makanan) 1,094,288
56104 Food traders with mobile/nonpermanent premises 839,678
96999 Other individual service activities 672,702
14120 Tailoring and manufacture of clothes to order 513,358

Pathways to Middle-Class Jobs in Indonesia June 2021

50 Jobs as a Catalyst for Economic Growth and a Middle-Class Society

Box 2.1 continued

47192 Retail trade of various kinds of goods, mainly not food, beverage 488,283
or tobacco (grocery items) not in department stores
47992 Retail trade of food, beverage or tobacco commodities from 432,801
manufacturing industry
45407 Motorcycle repair and maintenance 411,656

Sources: Sakernas 2018; World Bank staff calculations

Note: The KBLI is the Indonesian Industrial Classification System (Klasifikasi Baku Lapangan Usaha Indonesia).
Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

The number of gig workers has increased, driven by the rise of technological
companies and digital platforms. On average, these workers are more uncovered or
informal than other workers (Table 2.3). With the rise of gig workers, the National Statistical
Agency (Badan Pusat Statistik; BPS) added questions to Sakernas 201916 to identify
“dependent contractors” (equivalent to gig workers). The BPS recorded 2.2 million in 2019.17
Gig workers are more informal than non-gig workers more generally; 82.5 percent of gig
workers are informal versus 74.8 percent of non-gig workers. Around 41 percent of gig
workers work in the transportation and storage sectors. The remaining sectors account for
much smaller shares. Around 17 percent of gig workers work in wholesale or retail trade,
15 percent in manufacturing, and 12 percent in agriculture, forestry, livestock, or fisheries.
Around 71.3 percent of all dependent contractors reside in urban areas. Around 97.2 percent
of gig workers identified in Sakernas were also HHE owners (see Figure 2.8).

Table 2.3 Gig Workers Are More Likely Than Average to Be Informal
Number of workers and share of workers in formal versus informal employment

Informal Formal
Type Informal Formal Total
(%) (%)

Dependant 1,838,924 389,043 2,227,967 82.54 17.46

contractor
Other 92,933,122 31,354,030 124,287,152 74.77 25.23
Total 94,772,046 31,743,073 126,515,119 74.91 25.09

Sources: Sakernas 2019; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

16 The BPS started to collect information about dependent contractors in Sakernas of August 2018. However, the number
of dependent contractors was not published as it was the first time that this information had been collected. The BPS is more
confident in the accuracy of the calculation following the 2018 pilot and an increase in the sample size to 300,000 households
in Sakernas of August 2019 from 200,0000 previously. Consequently, the number of dependent contractors was released in
Sakernas of August 2019. Hence, the authors only use information on dependent contractors from Sakernas August 2019.
17 The BPS calls gig workers “dependent contractors” and defines them as follows: A dependent contractor is someone who is
bound to an economic unit (whether a business or company, institution, government, or other unit) to produce goods or services
that can provide benefits to the economic unit, but they are not laborers or employees of the economic unit. A dependent
contractor is also defined as a person who tries to be self-employed or tries to be assisted by a nonpermanent worker, a family
worker, or unpaid labor, but does not have full control over their business in areas such as raw materials, capital goods used,
price or tariffs for goods, services produced, or market or customer access, these being governed by the binding economic unit.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Jobs as a Catalyst for Economic Growth and a Middle-Class Society 51

Figure 2.8 Around Half of Informal Workers Were Household Enterprise Owners in 2019 
Workers in informal employment as dependent contractors and as owners of household enterprises (HHEs)

Sources: Sakernas 2019; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Most employees, who should enjoy a full range of benefits, hold low-quality jobs.
On a scale of 1 to 7 (1 being the fewest positive characteristics), the average job-quality
score is 1.19. This means that Indonesian jobs have slightly more than one positive quality
characteristic. Considering the seven dimensions of the job-quality index developed by
Brummund, Mann, and Rodriguez-Castelan (2016),18 the authors found that the most
prevalent job-quality characteristic is that the worker only has one single job. Nearly 60
percent of workers have held their job for at least three years and 56 percent have a
contract (though only 22.7 percent have a permanent contract). Nearly twice as many jobs
offer health insurance (44.7 percent) as retirement benefits (25 percent). However, the
most elusive job-quality variable is “middle-class remuneration.”

18 The index devised by Brummund, Mann, and Rodriguez-Castelan (2016) uses the variable “above-poverty-line remuner-
ation.” To align the index and the job quality discussion with the focus of this report, middle-class jobs, the authors substituted
the Brummund, Mann, and Rodriguez-Castelan (2016) variable with “middle-class remuneration.”

Pathways to Middle-Class Jobs in Indonesia June 2021

52 Jobs as a Catalyst for Economic Growth and a Middle-Class Society

Figure 2.9 Indonesian Jobs Are Low Quality by Many Measures

Share of jobs with each characteristic

Sources: Sakernas 2018; World Bank staff calculations

Note: By definition, information on social benefits and contract status is only available for wage employees. This
analysis is therefore based on wage employee samples only. Sakernas is the National Labor Force Survey (Survei
Angkatan Kerja Nasional).

Earnings increased from 2008 to 2018,19 particularly in the manufacturing sector.

The median real earnings index for income earners grew 5.42 percent annually (Annex Figure
2.13) from 2008 to 2018. The increase in the median real earnings index is 0.67 percentage
points higher than annual GDP growth during the same period. The manufacturing sector
has the highest compound annual earnings growth (6.4 percent) of all sectors, followed by
mining and quarrying (6.1 percent). The utilities sector has the lowest compound annual
earnings growth (1.3 percent) (Figure 2.10).

19 Wage increases among medium-sized and large manufacturing firms were aligned with labor productivity increases for
the period 1990–2015 (see Chapter 4). However, this may not apply for other sectors of the economy.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Jobs as a Catalyst for Economic Growth and a Middle-Class Society 53

Figure 2.10 Manufacturing Had the Highest Compound Annual Earnings Growth Rate in 2007–18 
Evolution of earnings by sector; index 2007=1

Sources: Sakernas 2011–18; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Box 2.2 A Transition Interrupted?

The COVID-19 pandemic is expected to continue to have deep impacts on the economy. Indeed, after several
years of strong growth, Indonesia’s GDP fell 2.1 percent in 2020, although this was a smaller downturn than in
other countries in the East Asia Pacific region, except China and Vietnam. This partly reflects good policy choices,
less restrictive containment measures, and the diversified structure of the economy.a Although Indonesia’s GDP
decline was much smaller than the 13-percent decline of the Asian financial crisis of 1997 and 1998, the COVID-19
crisis is expected to continue to manifest itself through several channels and thus to continue to have deep and
widespread negative effects.b

First, the crisis triggered by COVID-19 is a broad-based economic crisis. Some industries, such as mining,
plantations, tourism, and some labor-intensive manufacturing industries, have been affected by the external
shock, with demand drying up. Other industries, such as trade, transport, and hotels and restaurants have been
affected by domestic restrictions on movement. Most sectors have been somewhat or severely affected by
global and domestic measures to contain the virus. The handful that have not been so affected include those
providing certain medical goods and services, e-commerce, and less contact-intensive activities such as finance,
education, communications, and telecommunications. Firms across the board are feeling the impacts, and this
includes firms operating in sectors in which middle-class jobs are clustered.

Second, sectors in which mostly urban microbusinesses trade, including offline low-end food and beverages and
motorcycle ride-hailing transportation, have been severely hit. These enterprises have little inventory or savings,

a For example, Indonesia is less dependent on the sectors that were hit hardest by the pandemic, including tourism. Indonesia is also less integrated into the
global production network compared to its regional peers. See, for example, Ing and Kimura (2017).
b Much of the comparison in this report with the Asian financial crisis is based on an understanding of transmission channels. Quantitative analysis has
not yet been carried out.

Pathways to Middle-Class Jobs in Indonesia June 2021

54 Jobs as a Catalyst for Economic Growth and a Middle-Class Society

Box 2.2 continued

and limited access to financing. With social distancing measures, many have failed. Providing financial support
to these urban businesses is difficult since a large share of the self-employed, of daily-wage workers, digital gig
workers, and microenterprise owners operate under the radar.c The COVID-19 government assistance programs
that are aimed at formal workers and firmsd have limited effect on these informal workers and enterprises.
Consequently, the government introduced COVID-19 assistance programs aimed at informal workers and firms;
programs include direct cash transfers for ultra-microenterprises and microenterprises (Banpres Produktif
Usaha Mikro) and credit restructuring programs, including for drivers working via ride-hailing apps. While
these sectors—self-employment, jobs paying daily wages, digital gig workers, and microenterprises—are not a
significant source of middle-class jobs, the COVID-19 crisis may slow the progress of workers holding these roles
in moving up to middle-class status. The pandemic has also reduced demand for waged jobs in the formal sector
and increased the number of unpaid family and casual jobs.

Third, small and medium enterprises (SMEs), especially young SMEs, continue to be at risk. These firms have
lower levels of capital and little access to the financing that could help them through a period of shutdown.
Indeed, among medium-sized and large manufacturing firms, it was the smaller firms that disappeared during
the Asian financial crisis and never recovered at scale, limiting the dynamism and job-creation potential of young
innovators over the past few decades. If SMEs fail in large numbers again, progress to middle-class jobs may also
stall. Based on a representative firm-level survey conducted in June and October 2020, smaller firms have indeed
been hit harder in terms of average year-on-year change in monthly sales.e

Fourth, net job destruction and declining family incomes leave potentially long-lasting impacts, particularly for
youth (aged 15–24). Many have had to mix school with work, which has disrupted their learning. Between August
2019 and August 2020, 300,000 jobs were destroyed (in net terms), compared to 2.47 million jobs created (in
net terms) between August 2018 and August 2019, and about 2 million jobs created every year on average in
the past decade (2008-2018). The pandemic has impacted the 7 million new graduates (including university and
high-school graduates)—part of the so-called “Generation COVID-19”f—who are ready to enter the labor force.g
Youth were delayed in entering the labor force; in 2020, about 300,000 fewer youth entered the labor force
compared to 2019. Not only has the pandemic caused increased numbers of youth to take on work alongside
their schooling, there has also been an increase in youth idleness (those not in employment, education, or
training).h The increase in employment rates among youth was especially concentrated among those of high-
school age (15–18 years old), who are likely to be working as unpaid family workers. This pattern holds even
for youth working full-time (at least 20 hours a week): the increase was especially concentrated among those
of high-school age. The effects are similar for boys and girls. This may mean interrupted learning (combining
schooling with full-time work) for these children of high-school age. Moreover, graduates starting work in 2020
earned less than their older cohort. A somewhat extensive literature finds that entering the labor force during
economic downturn comes with several issues: it is harder to find a job and this becomes a vicious circle, partly
because people lose skills when they are not working; consequently, it leads to many years of lower wages.

c They are not in government systems such as tax or social security.

d For example, tax measures or exemptions from making social security payments.
e World Bank 2020a
f World Bank 2020b
g Pradana, Hadiwidjaja, and Posadas 2021
h Halim, Hambali, and Purnamasari. Forthcoming

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Jobs as a Catalyst for Economic Growth and a Middle-Class Society 55

Box 2.2 continued

Job losses and falling household incomes could also affect human capital investment in children.i

Fifth, women have been entering the labor force for the first time as “added workers,” meaning that they are
taking on (low-quality) jobs to help compensate for the reduced income or job loss experienced by another
member of the household.j By August 2020, around 5.1 million workers had lost their jobs (both temporarily
and permanently) due to COVID-19, while another 24 million (one in five employed workers) had experienced a
reduction in working hours (and presumably, income).k Previous economic crises have shown that other family
members begin to work to compensate for lost income. By August 2020, six months after the pandemic hit
Indonesia, the crisis had led to negative net employment impacts for men of all ages, but positive net employment
impacts for women, except those in the early childbearing years (19–29 years old). This indicates that women
with caring responsibilities may have some constraints on their ability to engage in labor market work. Both
men and women with children aged under five were less likely to be employed compared to their counterparts
without children aged under five, but the gap is much larger for women.l The increase in the employment of
women is proportionate to the decrease in the share of women outside of the labor force and to the increase in
the share of women working who had no prior job experience. In other words, women entered employment for
the first time as added workers. Evidence also suggests that these female added workers took up low-quality
jobs. The question now is whether these low-quality jobs could be a stepping-stone that will allow these women
to permanently enter the labor force, a longer-term change.

Sixth, with a large number of firms in distress, many laid off workers. Workers and firms will need to go through
a massive rematching process. However, Indonesia does not have the labor market information or matching
tools to facilitate this. This may cause labor market frictions, delay the process of reemployment, and prolong
the recovery in jobs.

By August 2020, the share of middle-class jobs had declined by about 5.2 percentage points compared to August
2019. This trend, if not reversed, threatens the gains Indonesia has made in the last 30 years of structural
transformation toward better-quality jobs. However, this crisis may open new opportunities for accelerated
transformation in some areas, especially when combined with digital adoption, innovation, and training
(upskilling and reskilling) in digitally enabled sectors such as e-commerce.m

Looking forward, crisis means both danger and opportunity. The Asian financial crisis stalled Indonesia’s
progress toward more-productive, middle-class jobs. The recent economic disruption caused by the COVID-19
crisis could have the same results. But with careful planning to “build back better,” the disruption could provide
an opportunity for Indonesia to systematically and strategically map out a recovery agenda that focuses on
productivity and good jobs.

i World Bank 2020b

j Halim, Hambali, and Purnamasari. Forthcoming. See Posadas and Sinha (2010) for “added workers” effects during the Asian financial crisis.
k Sakernas August 2020
l A nationally representative survey shows that mothers are three times more likely to care for children than fathers, with half being in paid work and
struggling to find a balance between work and additional caring responsibilities. See SMERU Research Institute et al. (2021).
m See also World Bank (Forthcoming 2021)

Pathways to Middle-Class Jobs in Indonesia June 2021

56 Jobs as a Catalyst for Economic Growth and a Middle-Class Society

2.3.3 Middle-class jobs

About 15 percent of jobs paid a middle-class income in 2018; this has remained
fairly constant over the preceding two decades. Using dynamic poverty lines,20 among
income earners the composition of jobs by income-class status has barely changed since
2001, with middle-class jobs (using only remunerative criteria) barely moving from 17 to 15
percent (Figure 2.11). The share of poor- and vulnerable-class jobs has increased from 42 to
46 percent. The share of income-earning jobs that are commensurate with aspiring middle-
class status decreased from 42 percent in 2001 to 39 percent in 2018. This is in line with the
fact that the growth in the proportion of the economically secure class in Indonesia was
slower than in other middle-income countries in East Asia.21 Moreover, much of the change in
the composition of consumption-based class status in Indonesia might be more significantly
driven by nonlabor income (including profits, capital return, and social assistance).22

Figure 2.11 Composition of Jobs by Income-Class Status Has Barely Changed Since 2001
Income earners by income-class classification, 2001–18

Sources: Sakernas 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

In 2018, the manufacturing sector had the highest number of income earners (in
absolute terms) earning wages above the middle-class cutoff. Income earners in
the high-value service sectors and in public administration, defense, and compulsory social
security are the most likely23 to earn wages above the middle-class cutoff (Table 2.4 and
Annex Table 2.1). In absolute terms, manufacturing (2.21 million), public administration,
defense and compulsory social security (1.98 million), and education activities
(1.73 million) have the highest number of income earners earning above the middle-class

20 In the Shapley decomposition, the authors use real wages with a March 2018 base year and poverty line. In this analysis,
the authors use wages in nominal terms and a poverty line for the corresponding year.
21 World Bank 2019
22 World Bank 2019
23 Not controlling for any other factor.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Jobs as a Catalyst for Economic Growth and a Middle-Class Society 57

cutoff. Meanwhile, finance and other business activities (including insurance, real estate,
and business activities) (36.5 percent), public administration, defense, and compulsory social
security (42.18 percent), and information and communications (32.69 percent) have the
highest shares of income earners earning wages above the middle-class cutoff.

Table 2.4 The Manufacturing Sector Has the Highest Number of Income Earners (in Absolute Terms) Earning
Wages above the Middle-Class Cutoff
Number of workers as income earners, as income earners with income above middle-class cutoff, and their shares out of total income earners

Total income earners

Total income earners Share of total income
(wage employee,
Sector above the middle-class earners above the
self-employed, casual
cutoff middle-class cutoff (%)
worker)

Agriculture, forestry, livestock, and fisheries 14,371,642 509,462 3.54

Mining and quarrying 1,287,197 375,382 29.16
Manufacturing 14,539,543 2,211,864 15.21
Electricity and gas 322,754 81,625 25.29
Water supply, sewerage, and waste management, 410,302 53,920 13.14
and remediation
Construction 7,783,340 574,070 7.38
Wholesale and retail trade, vehicle repair 14,424,190 1,671,462 11.59
Transportation and storage 5,194,341 791,363 15.24
Accommodation and food services 4,403,044 441,671 10.03
Information and communications 765,874 250,353 32.69
Finance and insurance 1,781,338 770,794 43.27
Real estate 366,782 111,516 30.40
Business activities 1,438,853 426,818 29.66
Public administration, defense, and compulsory 4,681,280 1,974,729 42.18
social security
Education activities 5,956,585 1,730,386 29.05
Human health and social work activities 1,809,562 484,845 26.79
Other service activities 5,497,128 334,399 6.08
Total 85,033,755 12,794,659 15.05

Sources: Sakernas 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Pathways to Middle-Class Jobs in Indonesia June 2021

58 Jobs as a Catalyst for Economic Growth and a Middle-Class Society

Middle-class jobs in the manufacturing sector as a proportion of total middle-class

jobs held by income earners almost doubled between 2008 and 2018. In 2008, two
sectors accounted for the largest shares of middle-class jobs: public administration, defense,
and compulsory social security accounted for 17 percent of middle-class jobs and education
activities 18 percent. In 2008, the share of income earners with middle-class jobs in the
manufacturing sector was 9 percent, only half the share of those with middle-class jobs in
public administration, defense, and compulsory social security, and education activities. By
2018, the share of middle-class jobs in public administration, defense, and compulsory social
security had fallen to 15 percent of total middle-class jobs and in education activities to 14
percent. Over the same period, the manufacturing sector share of middle-class jobs almost
doubled to 17 percent. (Figure 2.12)

Figure 2.12 The Share of Middle-Class Jobs in Manufacturing among Income Earners Almost
Doubled between 2008 and 2018
Distribution of middle-class jobs by sector

Sources: Sakernas 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Using a more restrictive definition of middle-class jobs to include worker protection

benefits and contract status for wage employees only,24 only civil servants are
more than 10 percent25 likely to earn above middle-class wages (Table 2.5). A more
restrictive definition of a middle-class wage employee is one who earns at least a middle-
class income and enjoys worker protection benefits and an indefinite contract. Under this
definition, in absolute terms, public administration, defense, and compulsory social security
contributed 1.4 million middle-class jobs. Education activities contributed 875,882 middle-
class jobs, while the manufacturing sector contributed only 475,111 middle-class jobs.
Meanwhile, construction, wholesale and retail trade, vehicle repair, accommodation, and

24 There is no data on worker protection benefits or contract status for casual workers and the self-employed.
25 Not controlling for any other factor.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Jobs as a Catalyst for Economic Growth and a Middle-Class Society 59

food and beverages have almost no middle-class jobs under the more restrictive criteria,
and even fewer than that of the agriculture, forestry, livestock, and fisheries sector.

Table 2.5 If Worker Protection Benefits and Contract Types Are Added to the Criteria for a Middle-Class Job,
among Wage Employees, Only the Civil Service Is a Source of Middle-Class Jobs
Distribution of middle-class jobs with benefits for workers and indefinite contracts, wage employees, 2018

Total middle-class jobs providing Share of middle-class jobs providing

Employment sector employee benefits and indefinite benefits and indefinite contract to
contract total employees by sector (%)

Agriculture, forestry, livestock, and fisheries 33,354 1.00

Mining and quarrying 47,500 5.59
Manufacturing 475,111 4.31
Electricity and gas 27,690 9.23
Water supply, sewerage, and waste management 15,584 6.53
and remediation
Construction 32,969 0.99
Wholesale and retail trade, vehicle repair 57,584 0.88
Transportation and storage 62,980 2.81
Accommodation and food service activities 17,645 0.84
Information and communication 35,065 5.80
Financial and insurance activities 160,744 9.19
Real estate activities 10,337 3.65
Business activities 50,649 4.33
Public administration, defense, and compulsory 1,399,453 29.89
social security
Education activities 875,882 14.81
Human health and social work activities 187,167 11.30
Other service activities 9,909 0.31

Sources: Sakernas 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Having a middle-class income does not necessarily mean that a worker will enjoy
all dimensions of a quality job. Indonesia’s job-quality index measures 4.8 out of six
for middle-class jobs (as compared to 1.19 out of seven for all jobs).26 The holders of more
than 80 percent of middle-class jobs only hold that one job, which is governed by a signed
contract, have at least three years of tenure, and provides health insurance. However, only

26 Only six dimensions are measured for middle-class jobs since, by definition, all already meet the seventh indicator of
earning a middle-class income. Thus, the seventh indicator for the Job Quality Index for middle-class jobs is met by the entire
middle-class sample.

Pathways to Middle-Class Jobs in Indonesia June 2021

60 Jobs as a Catalyst for Economic Growth and a Middle-Class Society

66 percent pay into retirement schemes, and only half have permanent contracts. In other
words, though holders of middle-class jobs have sufficient income to purchase a middle-class
standard of living and protect themselves against most job losses, many still face risks—of
job loss, of old-age poverty, and of health-induced poverty.

Figure 2.13 Indonesian Middle-Class Jobs Are Still Subject to Risk

Share of middle-class jobs with each characteristic

Sources: Sakernas 2018; World Bank staff calculations

Note: By definition, information on social benefits and contract status is only available for wage employees.
This analysis is based on samples of wage employees with middle-class jobs. Sakernas is the National Labor Force
Survey (Survei Angkatan Kerja Nasional).

Consistent with the findings from the Shapley decomposition presented above,
wage employment is an important driver toward middle-class jobs (Figure 2.14).
Households with higher levels of income are more likely to have wage employment, to be an
employer with permanent or paid workers, or, to a lesser extent, to have nonlabor income
(for example, return of capital or dividends). A household classed as having a level of income
commensurate with poor and vulnerable status relies increasingly on (may be forced to rely
on) self-employment, being an employer with temporary workers or unpaid family workers
(very common among farm HHEs), casual employment, or being an unpaid family worker.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Jobs as a Catalyst for Economic Growth and a Middle-Class Society 61

Figure 2.14 Wage Employment Is an Important Driver toward More Middle-Income Jobs
Share of household by employment type and income class, 2018

Sources: Susenas 2018; World Bank staff calculations

Note: Susenas is the National Socioeconomic Survey (Survei Sosial Ekonomi Nasional).

2.4 Factors Underlying Wage Growth and More Middle-Class

Jobs27
Log earnings increased from 13.96 to 14.34 between 2001 and 2018, mostly due
to the increase in productivity-related characteristics. Using Sakernas 2001, 2011,
and 2018, the authors investigated the main drivers behind changes in mean earnings
and earnings inequality in Indonesia, in order to determine the pace of poverty reduction,
the expansion of the middle class, and the quality of economic growth. The authors
disaggregated the analysis into three subperiods: 2001–11, 2011–18, and 2001–18.28 The
earnings increase was mostly accounted for by the positive changes in the distribution of
the variables of interests, rather than the premium to those characteristics. About 81.6
percent of the increase in average earnings in this period was explained by the increase that
happened during the second half of the period, that is, 2011–18.

27 This section is based on Pradana and Wihardja (Forthcoming).

28 The methodology follows Ferreira, Firpo, and Messina (2017), who use Oaxaca decomposition to decompose changes
in mean earnings and earnings inequality into distribution/endowment and structure/premium effects that each consists of
explanatory variables based on an income earner’s characteristics, namely education, age, job type, gender, location, economic
sector of employment, and occupation.

Pathways to Middle-Class Jobs in Indonesia June 2021

62 Jobs as a Catalyst for Economic Growth and a Middle-Class Society

The primary drivers of change were distribution of education, age, job type,
location, sector, and occupation (Table 2.6). The distribution of education contributed
positively to earnings as Indonesians became more educated and earned more. The authors
found that workers in 2018 were older and closer to the age at which earnings tend to peak
(47 years old).29 The distribution effect from location was positive to earnings, partly due
to workers moving to urban areas where average earnings are higher. Similar was observed
for job type contributed, as there were more wage employees relative to self-employed in
the later period, though the effect is very small. As workers moved out of the low-wage
agricultural sector, earnings also increased on average. As the share of workers with high-skill
jobs (for example, professionals and technicians, administrative and managerial) increased,
occupation also contributed positively to earnings on average.

Greater female labor force participation, return to education, and the job type
premium hindered wage growth. Since more women were participating in the labor force,
and women earn less than men on average (see Chapter 5), the distribution of workers by
gender brought earnings down on average, although this effect was small. Meanwhile, the
decline in educational returns at all levels of education contributed negatively to earnings.
Similarly, the higher penalty (widening wage gap) of being self-employed relative to being a
wage employee contributed to the negative premium effect of job type to earnings.

Table 2.6 Decomposition of Mean Earnings Growth 2001–18

Endowment Structure

2018–2001 2011–2001 2018–2001 2018–2001 2011–2001 2018 - 2011

Education 0.07967 *** 0.04526 *** 0.04299 *** −0.04458 ** -0.00067 −0.0525 ***
Age 0.01898 *** 0.01002 *** 0.01045 *** −0.00072 0.06565 −0.06786 ***
Job type 0.00409 *** −0.00695 *** 0.00464 *** −0.04479 *** 0.01837 * −0.05676 ***
Gender −0.011 *** −0.00127 −0.00883 *** 0.00362 0.01854 *** −0.01582 ***
Location 0.01759 *** 0.02625 *** −0.00411 *** 0.22223 0.13069 0.08698 **
Economic sector 0.00433 *** 0.00164 *** 0.00386 *** 0.04684 0.04112 0.00455
Occupation 0.02031 *** 0.01496 *** 0.01141 *** 0.01606 −0.05465 0.06466 **

Sources: Pradana and Wihardja. Forthcoming

29 A similar exercise that explores the factors underlying the increase of 1.47 percentage points in the earning Gini (from
37.52 to 38.99) find that the largest contributor to the inequality-increasing distribution effect between 2001 and 2018 was
education. This effect can be attributed to the “paradox of progress,” namely the marked convexity of the earning-education
profile at higher levels of education that increases inequality as the level of education improves. (Ferreira, Firpo, and Messina
2017; Bourguignon, Ferreira, and Lustig 2005). For the inequality impacts of other variables, see Pradana and Wihardja
(Forthcoming).

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Jobs as a Catalyst for Economic Growth and a Middle-Class Society 63

The decomposition analysis of mean earnings shows that education contributed

the most to the increase in mean earnings growth between 2001 and 2018.
However, it has also contributed to the increase in earning inequality.30

2.5 Conclusions
This chapter highlights that if Indonesia is to expand its middle-class population,
better-quality jobs—not just any jobs—are needed. Indonesia is creating a limited
number of middle-class jobs and this is hampering its middle-class expansion. Most jobs
created from 2008 to 2018 have been in low-wage sectors, with a median nominal wage far
below the minimum wage necessary to join the middle class. Almost half of jobs (45.8 percent)
created between 2008 and 2018 were in low-value, low-wage services, namely wholesale and
retail trade, accommodation, food and beverages, and other services. In 2018, among wage
employees, private-sector jobs did not account for a significant amount of middle-class
jobs if the definition of middle-class jobs is adjusted to include worker protection benefits
and permanent contract status. Indonesia’s employment-type composition reveals high
vulnerability for most jobs. About three-quarters of Indonesia’s labor market is uncovered
or informal and, of this, half is made up of HHE owners in low value-added sectors. The
decomposition analysis shows the important role of education in increasing mean earnings
and boosting middle-class job prospects, but also highlights that it brings with it increased
risks of earnings inequality.

The following chapters will explore the demand and supply sides of jobs in Indonesia
in more detail. Jobs are created when there is demand and matched supply, and the right
mechanism to link and match them. Indonesia’s low-quality job creation as discussed in this
chapter demands more in-depth empirical evidence from both the demand and supply sides.
Chapters 3 and 4 look at job creation from the demand side, while Chapter 5 looks at the
supply side. Chapter 3 presents analyses and evidence from the macro perspective: how
within- and between-sector labor productivity gains (the latter will be called “structural
transformation”) over the past two decades have moved Indonesia toward middle-class jobs
and where they have fallen short. Chapter 4 looks in detail at firm dynamics in an industry
with among the highest potential to create middle-class jobs, as this chapter suggests,
namely medium-sized and large manufacturing firms. Chapter 4 will also look at whether
HHEs in Indonesia, in which a majority of Indonesians work (either as owner or employee),
can be a pathway toward middle-class jobs. Chapter 5 presents the characteristics of
workers who hold middle-class jobs versus the characteristics of the labor force as a whole
and identifies ways to better prepare Indonesia’s workforce for middle-class jobs.

30 Pradana and Wihardja. Forthcoming

Pathways to Middle-Class Jobs in Indonesia June 2021

64 Jobs as a Catalyst for Economic Growth and a Middle-Class Society

Annex Chapter 2
Annex 2.1 Macro-Level Indicators

Annex Figure 2.1 Year-On-Year Real GDP Growth (%) Annex Figure 2.2 Long-Term View of Employment
and Employment Elasticity, 1991–2018 and Unemployment Rates, 1998–2018

Sources: BPS; CEIC; World Bank staff calculations Sources: Sakernas 1998–2018
Note: The BPS is the National Statistical Agency Note: Sakernas is the National Labor Force Survey (Survei Angkatan
(Badan Pusat Statistik). Kerja Nasional).

Annex 2.2 Shapley Decomposition – Poverty Reduction

Annex Figure 2.3 Share of Contribution to Poverty Reduction, 2000–14

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Jobs as a Catalyst for Economic Growth and a Middle-Class Society 65

Sources: Indonesian Family Life Survey (Survei Aspek Kehidupan Rumah

Tangga Indonesia) 2000, 2007, and 2014; World Bank staff calculations

Annex Figure 2.4 Average Income and Nonlabor Income in Households below Poverty Line by Different Poverty
Line Levels in 2000, 2007, and 2014

Sources: Indonesian Family Life Survey (Survei Aspek Sources: Indonesian Family Life Survey (Survei Aspek
Kehidupan Rumah Tangga Indonesia) 2000, 2007, and 2014; Kehidupan Rumah Tangga Indonesia) 2000, 2007, and 2014;
World Bank staff calculations World Bank staff calculations

Pathways to Middle-Class Jobs in Indonesia June 2021

66 Jobs as a Catalyst for Economic Growth and a Middle-Class Society

Annex 2.3 Shapley Decomposition – Poor-, Vulnerable-, and Middle-Class Jobs

Annex Figure 2.5 Shapley Value Estimates by Sector: Contributions to the Decline in Poor- and Vulnerable-Class
Jobs, 2011–18

Sources: Sakernas 2011 and 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Annex Figure 2.6 Shapley Value Estimates by Sector and Status of Employment: Contributions to the Decline in
Poor- and Vulnerable-Class Jobs, 2011–18

Sources: Sakernas 2011 and 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Jobs as a Catalyst for Economic Growth and a Middle-Class Society 67

Annex Figure 2.7 Shapley Value Estimates by Sector and Formality: Contributions to the Decline in Poor- and
Vulnerable-Class Jobs, 2011–18

Sources: Sakernas 2011 and 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Annex Figure 2.8 Shapley Value Estimates by Sector: Contributions to the Increase in Middle-Class Jobs, 2011–18

Sources: Sakernas 2011 and 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Pathways to Middle-Class Jobs in Indonesia June 2021

68 Jobs as a Catalyst for Economic Growth and a Middle-Class Society

Annex Figure 2.9 Shapley Value Estimates by Sector and Job Status: Contributions to the Increase in Middle-
Class Jobs, 2011–18

Sources: Sakernas 2011 and 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Annex Figure 2.10 Shapley Value Estimates by Sector and Formality: Contributions to the Increase in Middle-
Class Jobs, 2011–18

Sources: Sakernas 2011 and 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Jobs as a Catalyst for Economic Growth and a Middle-Class Society 69

Annex 2.4: Annex Figures and Tables

Annex Figure 2.11 Employment Share by Occupation Level, 2007–15

Sources: Sakernas 2007–15; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Annex Figure 2.12 Share of Self-Employment and Family Workers, 2018

Sources: World Development Indicators

Note: The authors use self-employment and family worker as indicators of a lack of worker protection. (Data derived from ILOSTAT, the
database of the International Labour Organization, at https://ilostat.ilo.org/) . The self-employed and family workers are vulnerable
because they are less likely to have a formal work arrangement, decent working conditions, social security, or effective representation by
a union or similar organization. They are often characterized by inadequate earnings, low productivity, and difficult working conditions,
undermining their fundamental rights. See UN Women. N.d.

Pathways to Middle-Class Jobs in Indonesia June 2021

70 Jobs as a Catalyst for Economic Growth and a Middle-Class Society

Annex Figure 2.13 Evolution of Earnings and Share of Waged Workers Earning Below the Minimum Wage,
2008–18

Sources: Sakernas 2008-18; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Annex Figure 2.14 Underemployment Rate by Location, 2008–18

Sources: Sakernas 2008–18; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Jobs as a Catalyst for Economic Growth and a Middle-Class Society 71

Annex Figure 2.15 Underemployment Rate by Sector, 2018

Sources: Sakernas 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Annex Figure 2.16 Employment Share by Wage Category, 2008–18

Employment share by economic sector, 2008–18

Sources: Sakernas 2008–18; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Pathways to Middle-Class Jobs in Indonesia June 2021

72 Jobs as a Catalyst for Economic Growth and a Middle-Class Society

Annex Figure 2.17 Employment by Employment Status, 2008–18

Sources: Sakernas 2008-18; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Annex Figure 2.18 Employment Status by Island Region, 2018

Sources: Sakernas 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Jobs as a Catalyst for Economic Growth and a Middle-Class Society 73

Annex Figure 2.19 Share of Workers by Educational Attainment, 2008–18

Sources: Sakernas 2008–18; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Annex Figure 2.20 Share of Employment by Gender, 2008–18

Sources: Sakernas 2008–18; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Pathways to Middle-Class Jobs in Indonesia June 2021

74 Jobs as a Catalyst for Economic Growth and a Middle-Class Society

Annex Figure 2.21 Female Labor Force Participation Annex Figure 2.22 Male Labor Force Participation
Rate, 1991–2019 Rate, 1991–2019
Female Male

Source: Lain, Alatas, and Setyonaluri. Forthcoming Source: Lain, Alatas, and Setyonaluri. Forthcoming
Note: Sakernas is the National Labor Force Survey (Survei Angkatan Note: Sakernas is the National Labor Force Survey (Survei Angkatan
Kerja Nasional). Kerja Nasional).

Annex Figure 2.23 Female Employment Rate, Annex Figure 2.24 Male Employment Rate,
1991–2019 1991–2019
Female Male

Source: Lain, Alatas, and Setyonaluri. Forthcoming Source: Lain, Alatas, and Setyonaluri. Forthcoming
Note: Sakernas is the National Labor Force Survey (Survei Angkatan Note: Sakernas is the National Labor Force Survey (Survei Angkatan
Kerja Nasional). Kerja Nasional).

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Jobs as a Catalyst for Economic Growth and a Middle-Class Society 75

Annex Table 2.1 Number of Employed by 17-Digit Sector, 2008 and 2018

Sector 2008 2018

Agriculture, forestry, livestock, and fisheries 41,331,704 35,703,072

Mining and quarrying 1,070,106 1,454,256
Manufacturing 13,047,800 18,251,456
Electricity and gas 145,538 338,447
Water supply, sewerage, and waste management
230,467 471,067
and remediation
Construction 5,438,965 8,300,297
Wholesale and retail trade, vehicle repair 17,981,176 23,073,516
Transportation and storage 5,705,225 5,398,582
Accommodation and food service activities 4,098,585 7,662,236

Information and communication 563,446 894,673

Financial and insurance activities 663,741 1,796,913

Real estate activities 183,818 389,925
Business activities 700,201 1,664,791
Public administration, defense, and compulsory
2,521,401 4,681,280
social security
Education activities 3,286,308 6,066,878
Human health and social work activities 741,015 1,848,460
Other service activities 4,843,251 6,009,100
Total 102,552,747 124,004,949

Sources: Sakernas 2008 and 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Annex Table 2.2 CAGR and Absolute Change in Employment by Wage Category, 2008–18

Change in share (of

CAGR 2008–18 (%) Absolute change total jobs) (percentage
points)

Low wage 1 4,433,808 –7.98

Mid wage 3 10,538,915 3.86
High wage 7 6,479,479 4.11

Sources: Sakernas 2008 and 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Pathways to Middle-Class Jobs in Indonesia June 2021

76 Jobs as a Catalyst for Economic Growth and a Middle-Class Society

Annex Table 2.3 Wage Employment by Contract and Worker Protection Status, 2019

Absolute Share (%)

Employee 51,654,993 100

Employee with permanent contract 11,856,113 22.95
Employee with permanent contract and full benefits (health, work 5,424,279 10.50
accident, death, nongovernment old age, government pension, paid
leave)
Employee with permanent contract, full benefits (health, work 5,157,493 9.99
accident, death, nongovernment old age, government pension, paid
leave), and earnings above the provincial minimum wage

Sources: Sakernas 2019; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Annex Table 2.4 Share of Informal Employment within Each Economic Sector, 2019

Sector Informal employment (%)

Agriculture, forestry, livestock, and fisheries 96.6

Mining and quarrying 64.5
Manufacturing 61.2
Electricity and gas supply 24.9
Water supply, sewerage, and waste management and remediation 76.1
Construction 88.6
Wholesale and retail trade; repair of motor vehicles and motorcycles 86.1
Transportation and storage 73.2
Accommodation and services activities 87.9
Information and communication 45.6
Financial and insurance activities 14.0
Real estate 43.6
Business activities 49.7
Public administration and defense; and compulsory social security 4.2
Education 13.3
Human health and social work activities 18.2
Other service activities 89.9
Total 74.9

Sources: Sakernas 2019; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Jobs as a Catalyst for Economic Growth and a Middle-Class Society 77

Annex Table 2.5 Household Enterprise Owners by One-Digit Sector, 2019

Sector Total Share (%)

Agriculture, forestry, livestock, and fisheries 16,693,786 37.01

Mining and quarrying 271,248 0.6
Manufacturing 4,830,853 10.71
Electricity and gas supply 27,763 0.06
Water supply, sewerage, and waste 195,459 0.43
management and remediation activities
Construction 533,531 1.18
Wholesale and retail trade; repair of motor 12,775,963 28.32
vehicles and motorcycles
Transportation and storage 2,480,220 5.5
Accommodation and food and beverages 4,535,471 10.05
activities
Information and communication 234,502 0.52
Financial and insurance activities 30,150 0.07
Real estate 80,095 0.18
Business activities 325,408 0.72
Education 155,278 0.34
Human health services and social work 144,320 0.32
activities
Other service activities 1,794,452 3.98

Sources: Sakernas 2019; World Bank staff calculations

Note: The total comes to 99.99 percent due to rounding.

Pathways to Middle-Class Jobs in Indonesia June 2021

78 Jobs as a Catalyst for Economic Growth and a Middle-Class Society

Annex Table 2.6 Dependent Workers by One-Digit Level Sector, 2019

Sector Total Share (%)

Agriculture, forestry, livestock, and fisheries 275,932 12.38

Mining and quarrying 14,194 0.64
Manufacturing 341,556 15.33
Electricity and gas supply 4,065 0.18
Water supply, sewerage, waste management, 12,508 0.56
and remediation activities
Construction 39,806 1.79
Wholesale and retail trade; repair of motor 393,399 17.66
vehicles and motorcycles
Transportation and storage 910,108 40.85
Accommodation and services activities 63,077 2.83
Information and communication 45,955 2.06
Financial and insurance activities 8,303 0.37
Real estate 2,978 0.13
Business activities 27,784 1.25
Education 3,027 0.14
Human health services and social work activities 3,120 0.14
Other service activities 82,155 3.69
Total 2,227,967 100

Source: Sakernas 2019

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

June 2021 Pathways to Middle-Class Jobs in Indonesia

 References 79

References
Asian Development Bank. 2019. Policies to Support the Development of Indonesia’s Manufacturing Sector during
2020–2024: A Joint ADB–BAPPENAS Report. Asian Development Bank Publications.

Azevedo, Joao Pedro, Gabriela Inchauste, Sergio Olivieri, Jaime Saavedra, and Hernan Winkler. 2013. “Is Labor
Income Responsible for Poverty Reduction? A Decomposition Approach.” Policy Research Working
Paper 6414, World Bank, Washington, DC.

Bourguignon, Francois, Francisco Ferreira, and Nora Lustig. 2005. The Microeconomics of Income Distribution
Dynamics in East Asia and Latin America. Washington, DC: Oxford University Press and World Bank.

Brummund, Peter, Christopher Mann, and Carlos Rodriguez-Castelan. 2016. “Job Quality and Poverty in Latin
America.” Policy Research Working Paper 7927, World Bank, Washington, DC.

Ferreira, Francisco H. G., Sergio P. Firpo, and Julian Messina. 2017. “Ageing Poorly? Accounting for the Decline in
Earnings Inequality in Brazil, 1995–2012.” Washington, DC: World Bank Group.

Halim, Daniel, Sean Hambali, and Ririn Purnamasari. Forthcoming. “Assessing the Gendered Impacts of the
COVID-19 Pandemic on Indonesia’s Labor Market.”

Hill, Hal, and Siwage Dharma Negara, eds. 2019. The Indonesian Economy in Transition: Policy Challenges in the
Jokowi Era and Beyond. Singapore: ISEAS.

Ing, Lily Yan, and Fukumari Kimura. 2017. Production Networks in Southeast Asia. London: Routledge.

Lain, Jonathan, Hamidah Alatas, and Diahhadi Setyonaluri. Forthcoming. “Women at Work in Indonesia.” World
Bank, Washington, DC.

Merotto, Dino, and Michael Weber. 2018. “Jobs and Economic Transformation in CEMs.” Presentation to
Macroeconomics, Trade, and Investment economists, World Bank, Washington, DC, January 24.

Posadas, Josefina, and Nitish Sinha. 2010. Persistence of the Added Worker Effect: Evidence Using Panel Data from
Indonesia. Internal document, World Bank, Washington, DC.

Pradana, Abror Tegar, Grace Hadiwidjaja, and Josefina Posadas. 2021. “Labor Market Dynamics during the
COVID-19 Pandemic.” World Bank internal document.

Pradana, Abror Tegar, and Maria Monica Wihardja. Forthcoming. “Examining the Drivers of Changes in Mean
Earning and Earning Inequality in Indonesia.” Background paper, Pathways to Middle-Class Jobs in
Indonesia, World Bank, Jakarta.

SMERU Research Institute, Prospera, UNDP (United Nations Development Programme), and the United Nations
Children’s Fund (UNICEF). 2021. Executive Summary Report: The Social and Economic Impacts of Covid-19
on Households and Strategic Policy Recommendations for Indonesia.

UN Women. N.d. “Indicator: Proportion of Migrants Employed in Vulnerable Employment, by Sex.” https://esa.
un.org/unmigration/documents/retreat/UN%20WOMEN_Indicator_vulnerable_employment.pdf

World Bank. 2019. Aspiring Indonesia: Expanding the Middle Class. Washington, DC: World Bank.

World Bank. 2020a. “Covid-19 Impact on Firms in Indonesia: Results from the Covid-19 Business Pulse Survey,
Round 2 (October 2020).” World Bank.

World Bank. 2020b. “The Jobs Response to Covid-19 in East Asia Pacific.” Note prepared by Social Protection and
Jobs Global Practice, World Bank. Version May 27, 2020.

World Bank. Forthcoming 2021. In Bits and Bytes: Foundation of an Inclusive Digital Future in Indonesia. Washington,
DC: World Bank.

Pathways to Middle-Class Jobs in Indonesia June 2021

80 Structural Transformation toward Middle-Class Jobs

Structural
Transformation toward
Middle-Class Jobs1

Chapter 3
1 This chapter is largely based on Gil Sander and Yoong (Forthcoming) and Steenbergen and Hebous (2020).

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Structural Transformation toward Middle-Class Jobs 81

3.1 Growth Models and the Expansion of the Middle Class

3.1.1 Jobs have been created but the quality of jobs has been insufficiently high

From 2008 to 2018, sound macroeconomic policies and a favorable external

environment helped the Indonesian economy to grow, create jobs, and
reduce poverty. Stable economic growth of about 5 percent a year over this
period was associated with job-creation rates of nearly 2 percent a year (Figure 3.1).
On a net basis, the Indonesian economy created 2.1 million jobs annually on average,
labor force participation hovered around 67 percent, and the pool of unemployed
shrank by 240,000 individuals each year. High growth helped bring poverty rates to
below 10 percent of the population. However, Indonesia’s growth was insufficient to
lift average income levels to those of richer economies such as Malaysia or Thailand.
Moreover, Indonesia was overtaken by China in terms of per capita GDP between
2000 and 2010, and the income gap with Vietnam has been steadily reduced—
although Vietnam remains poorer than Indonesia (Figure 3.2).

Pathways to Middle-Class Jobs in Indonesia June 2021

82 Structural Transformation toward Middle-Class Jobs

Figure 3.1 Respectable Growth and Job-Creation Rates

Percent, per year

Figure 3.2 But No Convergence with Richer Asian Economies

Per capita GDP (Indonesia=100)

Sources: Estimates based on Sakernas, National Accounts, and World Development Indicators.
GDP per capita is based on constant 2010 US dollar.
Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Economic growth has, however, not delivered enough good-quality job opportunities
for Indonesians. As explained in Chapter 2, earnings from labor is the main vehicle for
achieving a middle-class standard of living, but too many Indonesians still lack the level
of economic security associated with middle-income status. Most workers remain in low
value-added activities, with low and insecure earnings and little protection. In agriculture
and low-value services—sectors that together account for the majority of employment—
virtually no workers earn a middle-class wage. Those aspiring to middle-class status remain
at risk of slipping back into vulnerability or poverty even in times of relatively high economic
growth.2 Although Indonesia’s recent growth model has created both value added and jobs,
the quality of jobs has been insufficiently high. Hence, it is not only the level of growth that
matters, but also the composition of growth and the transmission channels from growth
to jobs.

2 World Bank 2020a

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Structural Transformation toward Middle-Class Jobs 83

3.1.2 Indonesia’s structural transformation post Asian financial crisis has not brought
enough productivity gains

In the 1990s, Indonesia industrialized rapidly through growth in manufacturing

exports. The rise of global value chains (GVCs) engendered a global expansion of international
trade after 1990. Countries that managed to integrate their economies with GVCs became
more productive, their incomes grew, and poverty fell.3 Indonesia carried out broad reforms
to create a business-friendly climate, including reforms that increased financial openness.
These, combined with a large labor surplus, helped Indonesia attract investment in labor-
intensive industries such as textiles and food and beverages. On average, value added grew
by over 8 percent between 1990 and 1997; manufacturing, other industries, and services
contributed equally (Figure 3.3, left). By 1996, manufacturing accounted for one-quarter
of GDP and manufacturing exports accounted for half of total export value.4 These trends
came to an abrupt halt with the Asian financial crisis in 1997–98. In 1998, Indonesia’s GDP fell
by 13 percent, and it did not recover to 1997 levels until 2003.

After the crisis, Indonesia’s economy shifted into a natural resource–based growth
model. This shift back to a dependence on natural resources was the result of several factors:
the negative impact of the Asian financial crisis, which hurt manufacturing industries; the
growing dominance of China in low-cost and labor-intensive manufacturing industries; the
global increase in demand for (and prices of) commodities in abundance in Indonesia such
as coal and palm oil; a resulting sharp appreciation in the real effective exchange rate,
which hurt manufacturing competitiveness; and weaknesses in the supporting business
environment including insufficient infrastructure, low levels of skills within the workforce,
an unfavorable regulatory framework for foreign direct investment (FDI), and “big-bang”
decentralization that caused regulatory uncertainties at the local level.

Although economic growth remained relatively high, the shift into natural
resources triggered a significant change in the composition of growth, hurting the
quality of jobs created. Compared to the 1990s, economic growth rates in all sectors fell
back, but total value-added growth remained at a respectable 5 percent on average between
2000 and 2012. The sources of growth changed, however, reflecting the shift to commodity
exports and the increasing role of services (especially nontradable services) in the economy
(Figure 3.3, right). The services sector became the single largest source of growth, whereas
manufacturing growth fell, resulting in premature deindustrialization (Figure 3.4). As the
commodity boom came to a halt post 2012, Indonesia did not return to a growth model
driven by manufacturing exports. Growth rates in nonmanufacturing industries and services
fell, but manufacturing growth did not increase sufficiently to replace this lost growth. The
move away from labor-intensive (or inclusive) growth after 2000 also meant there were few
opportunities to improve the quality of jobs for Indonesia’s labor force.

3 World Bank 2020b

4 Wihardja 2016

Pathways to Middle-Class Jobs in Indonesia June 2021

84 Structural Transformation toward Middle-Class Jobs

Figure 3.3 Manufacturing Previously Drove Industrial and Overall Growth, but Has Lost Momentum
Sector contribution to value-added growth, Sector value added, growth rates per annum
percentage points per annum

Sources: World Development Indicators

Note: The crisis years of 1998, 1999, and 2009 are not included.

Figure 3.4 Premature Deindustrialization in Indonesia

Source: Diop 2016

A new set of challenges is emerging as the global economic environment worsens—

Indonesia faces intense competition. Even before the onset of COVID-19, global conditions
were looking less favorable. Signs of slowdown in growth had emerged in major markets
such as the United States, Europe, and China, and a trend toward economic nationalism
(that is, protectionism), was accelerating, especially after the global financial crisis of 2008.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Structural Transformation toward Middle-Class Jobs 85

These global risks have emerged against a backdrop of longer-term trends, such as the rise
in the importance of knowledge economies and the increasing automation of tasks, which
affect the sources of competitive advantage, job creation, and growth.

Indonesia’s economy has become increasingly inward looking and is

underperforming in exports and FDI. Indonesia’s international openness (measured as
merchandise exports and imports as a percentage of GDP) has continued to increase, but at
a slow pace. Indonesia has been surpassed by other regional countries, such as China, India,
and Vietnam.5 Indonesia’s share of global manufacturing exports is falling and FDI is lower
than in comparator economies. Countries such as Thailand and Vietnam have higher levels
of exports and imports, are more integrated into GVCs, have more sophisticated and more
diversified exports, and have a higher share of services exports. Meanwhile, FDI in Indonesia
is not improving the country’s access to international markets as it has increasingly focused
on the extractive industries or access to local markets, rather than on exports. The post–
Asian financial crisis shift in the type of FDI that Indonesia attracts is problematic as it
does not create the quality or quantity of jobs that Indonesia needs. Export-oriented FDI
is typically associated with higher rates of innovation in products and processes and has
been associated with better job creation in the past in Indonesia.6 Estimates by the World
Bank suggest that investment in exports of both manufacturing and commercial services,
including tourism, can have strong effects on job creation.7

Lower integration with GVCs reflects lower competitiveness, which in turn results
from limited worker productivity. Although exports are not becoming more sophisticated
and wages remain low, total unit labor costs have continued to increase in Indonesia—
and more rapidly than in comparator economies (Figure 3.5). The loss in competitiveness
is not due to high growth in labor costs, but to slow growth in labor productivity. Labor
productivity remains low in Indonesia compared to comparators—and competitors—in
Asia, such as Malaysia, Vietnam, Thailand, and the Philippines (Figure 3.6, left). Indonesia is
also falling further behind as growth in labor productivity has been slower than in these
countries.8 Hence, although Indonesia’s workers in several industries are paid less than in
Vietnam or China, low productivity means that Indonesia cannot compete with low-wage
labor, as its unit labor costs are high (Figure 3.6, right).

5 Patunru 2019
6 World Bank 2020a
7 World Bank 2019
8 World Bank 2019

Pathways to Middle-Class Jobs in Indonesia June 2021

86 Structural Transformation toward Middle-Class Jobs

Figure 3.5 Indonesia Is Not Catching Up in Competitiveness

Average annual growth, 2010–19, percent

Sources: World Bank staff calculations based on unit labor cost data from the Economist Intelligence Unit.
Note: The unit labor costs here are not comparable with those in Figure 3.4.

Figure 3.6 Falling Behind in Labor Productivity in Low-Skill Manufacturing

Mean value added per worker, constant 2010 US dollar Average wages and unit labor costs, China=100

Source: Chang et al. 2019

Note: Productivity, wages, and unit labor costs refer only to the finishing, spinning, and weaving of textiles (to control for industry effects).

To achieve jobs that provide economic security and secure incomes, Indonesia
could move toward a productivity-driven growth model that creates value added
and jobs. This growth model must also facilitate the movement of workers into more
productive jobs, both within and between sectors, and must include the skilling up of
workers in low-productivity agriculture and in service sectors with low levels of value added.
Indonesia’s commodity-driven growth model, which has few backward or forward links to
local economies, has weakened the connection between sectors, firms, and occupations
where value is created, and those where most people make a living.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Structural Transformation toward Middle-Class Jobs 87

3.2 More and Better Jobs, but Not Yet Good Enough
Job creation and changes in the structure of employment have played a central role
in Indonesia’s ascension from low- to middle-income status and from high poverty
to low poverty. Between 2000 and 2018, some 5 million workers left jobs in agriculture (on
a net basis), and nearly 40 million jobs were created in nonagricultural sectors. The flow out
of agriculture coincided with rapid urbanization, including of jobs: in 1990, 31 percent of
Indonesia’s population lived in urban areas; in 2018, 55 percent; and in 2030, a majority—63
percent.9

Underemployment in agriculture fell and the increase in agricultural productivity

helped reduce poverty. In 2000, agriculture accounted for only 16 percent of value added,
but employed some 45 percent of the working population. By 2018, the contribution of
agriculture to value added had fallen to 13 percent. The share of the agricultural workforce
fell much more, however, to 29 percent of the total workforce in 2018, and most of this
reduction took place in the period 2010–18. The reduced importance of agricultural jobs
by far exceeded the fall in value added, however, suggesting that many of the workers
that left agriculture had been underemployed. Though the resulting increase in agricultural
productivity contributed to raising rural incomes and lowering poverty, poverty remains
highest in the agricultural sector because of the low earnings available.10

Nonagricultural sectors have provided better jobs with higher incomes and
productivity than agriculture, but not sufficient to shift people decidedly out of
vulnerability and economic insecurity. At an aggregate level, employment shifts have
been much more dramatic than the change in value added—as the agricultural sector shows
(Figure 3.7 and Figure 3.8). Indonesia’s experience of dynamic shifts in employment but
sluggish value-added change stands in contrast to countries like Vietnam or Cambodia where
value added transformed more rapidly than employment.11 In Indonesia, the dramatic change
in employment structure largely reflects one dominant trend: jobs were lost in agriculture,
especially after 2010, and jobs were created in low value-added services. The share of value
added accounted for by these services—retail and wholesale trade, accommodation and
restaurants, and community, personal, and other services—did not increase after 2000,
but they accounted for 70 percent of all new jobs in 2010–18. Hence, low-value services
accounted for 21 percent of value added in 2010 and in 2018, but they accounted for 32
percent of employment in 2010 and 36 percent in 2018. Although the rate of job creation
was higher in higher-value services than in lower-value services (3.8 versus 3.1 percent per
year between 2010 and 2018), high value-added sectors began from a small base, meaning
that they accounted for 16 percent of all new jobs in 2010–18, and 7 percent of total
employment in 2010, rising to 9 percent in 2018.

9 World Bank staff calculations based on United Nations (2019)

10 World Bank 2020c
11 Cunningham and Pimhidzai 2018; Cunningham et al. 2019

Pathways to Middle-Class Jobs in Indonesia June 2021

88 Structural Transformation toward Middle-Class Jobs

Figure 3.7 The Economy Diversified Away from Figure 3.8 . . . But Shifts in Labor Have Been More
Agriculture into Services . . . Dramatic and Have Taken Place in Different Sectors
Share of total value added, percent from Value Added
Share of total employment, percent

Sources: National Statistical Agency via CEIC; World Bank staff Sources: Sakernas; World Bank staff calculations
calculations Note: Sakernas is the National Labor Force Survey
(Survei Angkatan Kerja Nasional).
Note: High value-added services are financial, real estate, business, and insurance services (henceforth abbreviated as “financial services”),
transport and storage, information and communications (henceforth abbreviated as “transport and communications”). Low value-added
services are wholesale and retail trade, motor vehicle repairs, accommodation and food services (“trade, hotels and restaurants”), and
community, personal, and other services. “Other industry” comprises mining, utilities, and construction.

The crowding of workers from farming into lower value-added services—where

there was little growth in value added—limited any gains in earnings. Agricultural jobs
were replaced by jobs for which labor productivity was not sufficiently high to conclusively
shift workers and their families into middle-class income levels. Low entry requirements
(as employees or entrepreneurs) in low-value services12 make these sectors a common
destination for rural and urban low-skilled workers who lack the requisite skills for entering
formal jobs in higher value-added activities. However, since low-value services experienced
little improvement in labor productivity, these jobs did not provide any basis for growth in
earnings either (Figure 3.9). In fact, in the period 2010–18, the agricultural sector—where
employment was shrinking fast—saw the highest levels of productivity growth.

12 By aggregating nine-digit sectors into five economic groups, some high-end services (medical services provided by
doctors, or other professionals providing personal services) are included in the low-value services economic group.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Structural Transformation toward Middle-Class Jobs 89

Figure 3.9 Slow or Negative Labor Productivity Growth, Especially in Industry and Low-Value Services, after 2010 
Growth in value added per worker, average percent a year

Source: BPS via CEIC; Sakernas; World Bank staff calculations

Note: The BPS is the National Statistical Agency (Badan Pusat Statistik). Sakernas is the National Labor Force Survey (Survei Angkatan
Kerja Nasional).

The commodity boom hurt manufacturing jobs and employment in the sector has
been slow to recover. The share of manufacturing jobs in total employment stagnated
from the mid-1990s onward and throughout the Asian financial crisis and the commodity
boom; it hovered around 12–13 percent during 2000–10. The stagnation in manufacturing
jobs was caused by a loss in competitiveness and a lack of investment and growth in the
sector: labor productivity of manufacturing increased only marginally in spite of the
limited job creation, and relative productivity (the ratio of manufacturing sector share in
value added to manufacturing sector share in employment) fell between 2000 and 2018
(Figure 3.10).

Pathways to Middle-Class Jobs in Indonesia June 2021

90 Structural Transformation toward Middle-Class Jobs

Figure 3.10 Employment Has Shifted out of Agriculture and Mostly into Lower-Productivity Services
Relative productivity* and share of employment 2000, 2018

Sources: BPS via CEIC; Sakernas; World Bank staff calculations. *Relative productivity = sector share of value added/sector share of employment.
Note: The BPS is the National Statistical Agency (Badan Pusat Statistik). Sakernas is the National Labor Force Survey (Survei Angkatan Kerja
Nasional).

Nearly two out of three workers in Indonesia are trapped in sectors with
comparatively low levels of labor productivity, which prevents them from earning
higher wages. As shown in Figure 3.11, the relationship between wage levels and labor
productivity is strong and positive, and the vast majority of the workforce is in jobs that
have comparatively low wages and productivity. In spite of the significant flows of workers
out of agriculture, farming still accounts for nearly three in ten jobs. Nearly two out of
three jobs (64 percent) are in agriculture and low value-added services, where productivity
is only marginally higher than in agriculture.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Structural Transformation toward Middle-Class Jobs 91

Figure 3.11 Sectors That Are More Productive Tend to Pay Higher Wages, but Employ a Minority of Workers
Real median monthly wage in 2018 in Rp, millions; log of sector productivity as a share of total productivity, 2018; sector’s share of total
employment, 2018 (bubble size)

Source: Sakernas, August 2018; World Bank staff calculations

Note: The public sector is included in community, personal, and other services. The size of the bubble denotes the sector’s share of total
employment in 2018. Labor productivity is calculated as real output (value added) per worker. The national median monthly wage is Rp1.2 million.
Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

3.3 Incomplete Structural Transformation

3.3.1 Indonesia has not benefited enough from its demographic dividend

Labor productivity growth connected with job creation is needed if improvements

in living standards are to be sustained. Indonesia needs many more productive jobs in
its more productive sectors, while productivity within each sector also needs to increase.
Wages and living standards, as well as the competitiveness of firms and workers, depend
on labor productivity and a stronger link between job creation and value creation. These
links are forged as labor productivity increases in sectors that harbor many workers, and as
workers move from lower-productivity to higher-productivity activities. Productivity growth
happens through many channels: workers become better at the jobs they do (for example,
because of better machines, superior resource management, experience, or skills training);
less productive firms exit a sector and more productive firms enter; and resources—including
workers—move to firms, sectors, and jobs that have higher levels of productivity

The growth-employment trajectory of Indonesia shows that value-added growth

does not necessarily result in better jobs. Per capita output growth can be sustained by
several factors, including more employed workers relative to dependents (due to favorable
demographic trends, higher labor force participation, or lower unemployment), or by higher
labor productivity (Box 3.1). Hence, high growth is not necessarily associated with either
more or more productive jobs. As outlined in Chapter 2, Indonesia is not a story of jobless
growth, but the quality of jobs has not increased in tandem with their number. The current
growth model has resulted in insufficient productivity gains, and, perhaps even more
importantly, a disconnect between sectors creating value and sectors creating jobs.

Pathways to Middle-Class Jobs in Indonesia June 2021

92 Structural Transformation toward Middle-Class Jobs

Box 3.1 What Is Driving Value-Added Growth and Labor Productivity? 

The effect of growth on jobs and welfare depends on its main drivers. These drivers can be identified using the simple
relationship between per capita output, dependency rates, labor force participation, employment, and labor productivity.
The relationship can be expressed as:

where Y= value added, N = population, W= working-age population, L=the labor force, and E=total employment.

It follows that value-added per capita growth is driven by:

• Demographic effects: changes in the working-age population (W) relative to total population (N)

• Participation effects: changes in the labor force participation rate (L/W)

• Employment effects: changes in the share of the employed in the labor force (E/L)

• Productivity effects: changes in output per worker (labor productivity, Y/E).

Labor productivity growth is in turn driven by within-sector productivity effects or a structural transformation of value
added and employment.

• Within-sector productivity improvements occur as workers and firms in a given sector become more efficient. This can
occur through capital deepening or upgrading, technological change, increased skills in the workforce, or improved resource
allocation as less profitable firms or plants exit and are replaced by more efficient ones.

• Structural transformation occurs as workers (and other inputs of production) move from lower- to higher-productivity
sectors of the economy. The shift of Indonesia’s workforce into nonagricultural employment is an example of a structural
transformation.

Structural transformation, finally, is a composition of two effects: the volume of workers changing sectors, and the
differences in productivity levels of sectors. The larger either of these factors, the more significant the structural
transformation taking place.

Indonesia risks losing the tailwind of demographic dividend that has supported its
growth for the past five years, putting additional pressures on labor productivity
growth. Indonesia’s demographic window of opportunity is now closing—and at a lower
per capita level than in many other Asian countries. Demographic drivers—a higher share
of working-age persons providing for fewer dependents—cannot be counted on to provide
impetus to growth in the future. Today, about 50 percent of the population is below 30 years
old. Projections suggest that the share of the working-age population in the total population
will soon reach its peak (in 2025–30) and that the share of elderly in the population will
shortly begin to increase relative to the share of dependents. The period 2015–30 has been
described as the period in which Indonesia could reap a demographic dividend, as the ratio
of dependents relative to the economically active is expected to fall to about 47 percent.
A country’s economic growth usually accelerates during such a period. However, Indonesia
has not benefited as much from this demographic dividend as other Asian countries that

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Structural Transformation toward Middle-Class Jobs 93

have undergone similar. At the end of the demographic-dividend period, Indonesia’s GDP
per capita is predicted to reach somewhere between US$2,583 and US$3,709, which is less
than 25 percent of Korea’s GDP per capita at the end of its demographic-dividend period in
2010 (US$16,219).13 The closing demographic window means that Indonesia cannot count on
factor accumulation, especially not labor surplus, to drive growth in the future.

3.3.2 The contribution of structural transformation to labor productivity growth in

Indonesia was smaller than in other regional countries
As in other East Asian countries, labor productivity has been the main driver of
economic growth in Indonesia. A decomposition of the sources of growth between
2000 and 2017 shows that increases in value added per worker contributed as much as 88
percent of Indonesia’s GDP per capita growth of 3.9 percent, or 3.5 percentage points per
year. Demographic factors also contributed positively (about 1.0 percentage point), as the
deceleration in population growth in the 1980s and 1990s led to an increase in the share of
the working-age population relative to dependents—the demographic dividend. An increase
of the employment share in the labor force (in parallel with a reduction in unemployment
rates) contributed the remainder. Labor force participation rates, by contrast, fell and
therefore contributed negatively. However, as seen in Chapter 5, the falling labor force
participation rates mostly reflected the fact that more young adults remained in school and
so represent an investment in future skills, productivity, and growth.

Labor productivity growth resulted mostly from productivity growth within

sectors, especially in the services sector. Efficiency gains within the main sectors—
agriculture, industry, and services—accounted for 2.5 percentage points of the labor
productivity growth rate, or about two-thirds of total growth. In terms of absolute growth
levels, the contribution of within-sector productivity growth to growth in value added per
worker aligns Indonesia with Asian comparators (except China). The main driver of these
efficiency gains were the services sectors, which accounted for 37 percent of total labor
productivity growth between 2000 and 2017. As shown in Figure 3.13, increased productivity
in agriculture also contributed to growth, whereas labor productivity growth within the
industrial sector was very limited, and was also limited in relation to comparators.

13 World Bank staff calculations based on United Nations population projections and World Development Indicators. Figures
in US dollars at 2000 prices.

Pathways to Middle-Class Jobs in Indonesia June 2021

94 Structural Transformation toward Middle-Class Jobs

The expansion of middle-class jobs has not materialized because structural

transformation has not brought enough productivity gains. The move out of
agriculture and into industry and services contributed positively to labor productivity
growth. However, the changes in the structure of employment have contributed much
less to total productivity growth. The total contribution of structural change to growth,
at 0.9 percentage points, was smaller than in Vietnam (2.2 percentage points), Thailand
(1.3 percentage points), and China (1.1 percentage points), which have all experienced a
faster increase in the middle class in the recent past than Indonesia. A comparison with
Vietnam is telling: Vietnam moved a larger proportion of workers out of agriculture into
other sectors, and into more productive sectors. Average productivity levels are higher in
Vietnam’s industry sector (and in Thailand’s services sector) compared to the corresponding
sectors in Indonesia.

In recent years, within-sector productivity growth has fallen back further. Over
2010–17, the contributions of within-sector changes in industry to labor productivity growth
were minimal, reaching only 0–1 percentage points per year.

Figure 3.12 Labor Productivity Growth Has Driven Growth

Contribution to value–added per capita growth in 2000–17, percentage points, Indonesia and comparators

Sources: World Bank staff calculations using BPS data and Sakernas; World Development Indicators
Note: The BPS is the National Statistical Agency (Badan Pusat Statistik). Sakernas is the National Labor Force Survey (Survei Angkatan
Kerja Nasional).

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Structural Transformation toward Middle-Class Jobs 95

Figure 3.13 But the Contribution of Structural Transformation to Labor Productivity Growth Has Been Limited,
Which Has Limited the (Positive) Impacts on Jobs 
Contribution to annual labor productivity growth over 2000–17, percentage points, Indonesia and comparators

Sources: World Bank staff calculations using data from the National Statistical Agency and Sakernas; WDI
Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional). “Within” refers to within-sector transformation and
“structural” refers to between-sector transformation.of textiles (to control for industry effects).

Box 3.2 Demand-Side Channels: How the COVID-19 Crisis Is Impacting Jobs
The COVID-19 pandemic is expected to continue to have deep impacts on the demand side of jobs; the world has
yet to see the end of the pandemic and uncertainties are looming. Indonesia’s GDP fell 2.1 percent in 2020 with
private consumption falling 2.7 percent and gross fixed investment falling 4.9 percent. Firms across the board
were hit hard, including those firms in sectors where middle-class jobs are clustered. Indonesian firms were hit
harder than those in Vietnam and Malaysia.a Micro, small, and medium-sized enterprises (MSMEs) particularly
struggled to adapt to the shocks. According to a firm-level representative survey, six out of ten firms had to
close temporarily or permanently by June 2020.b Around two-thirds of firms surveyed responded to the crisis
by reducing labor costs: 29 percent reduced the working hours of workers, 28 percent granted unpaid leave
or reduced wages, and 5 percent fired workers. Although by March 2021 almost all firms surveyed had already
reopened and the share of firms reducing labor costs had been halved, as long as the global health crisis persists
and uncertainties about the pandemic drive consumers and investors to be risk averse, the demand side of jobs
may remain weak.

The pandemic has reduced the demand for formal-sector waged jobs and increased the number of unpaid family
and casual jobs, reversing some of the gains made in the creation of wage employment. These are the gains that
have contributed the most to the expansion of Indonesia’s middle-class population over the past three decades.
Indonesia currently needs to create more jobs in manufacturing, as this chapter argues, but the sector has

a World Bank 2020d

b World Bank 2020d

Pathways to Middle-Class Jobs in Indonesia June 2021

96 Structural Transformation toward Middle-Class Jobs

Box 3.2 continued

been hit hard by the pandemic. The Prompt Manufacturing Index is released by Bank Indonesia, the central
bank, to provide a general view of the performance of the manufacturing sector. It plunged from 45.64 in Q1
2020 to 28.55 in Q2 2020. Various labor-intensive manufacturing industries had to temporarily stop operating;
this includes the automotive industry, which was experiencing obstacles in the supply of raw materials and
components, especially from countries that had imposed a lockdown.c Large automotive multinationals in
Indonesia such as Toyota, Daihatsu, and Honda decided to temporarily halt some production in response to the
crisis. Honda stopped production for two weeks because of a shortage of components from India and Malaysia,
where lockdowns interrupted supply. Yamaha and Suzuki also stopped activities for similar reasons. Meanwhile,
in the electronics sector, Gabel suffered from stalled imports from China and appreciation in the US dollar. All
these multinationals employ a vast number of workers and provide middle-class jobs. Although the condition of
the private sector had significantly improved by March 2021,d many firms have made structural adjustments to
cut costs, including labor costs, and have adopted digital technologies.e

Correct mixing and sequencing of short-term versus long-term policies as a response to transitory versus
permanent shocks is critical. In the short term, strategies to “keep the lights on” in firms in financial distress
are critical.f Without such measures, not only would the crisis lead to short-term job losses, but also to greater
frictions when the economy rebounds as businesses will have to devote time and financial resources to
restarting, and will face job-matching difficulties when they begin rehiring workers again.g Once COVID-19 has
been contained, it will be important to help firms restart so they can employ workers again. A policy package
that mixes and sequences job relief measures and job recovery measures provides the best of both worlds.h Job
relief measures would be aimed at helping businesses survive and retain workers, as well as providing income
protection to vulnerable workers. Job recovery measures would be implemented once the outbreak has been
contained and the green shoots of recovery have appeared on the horizon. In addition, during the recovery
period, many firms and workers remain vulnerable, necessitating the mixing of job recovery and relief measures.
Job interventions for recovery may include support for job creation (such as support to formal, informal, and
household enterprises) and linking labor demand with supply (such as employment services and training), as well
as job-creating public works and the provision of unemployment insurance. Firms should also make remote work
more effective by changing norms, and should anticipate changes in consumer and business spending, as well as
bankruptcies that may put pressure on banking, financial, and credit market systems. They should sharpen their
productivity and IT-digital operations, and strengthen their capital efficiency.i

The government has provided fiscal, credit, and cash assistance to firms since the pandemic started. However,
implementation has been constrained by several issues. First is the lack of (consolidated) databases of intended
program beneficiaries (especially for programs targeting firms in the informal sector) and the lack of awareness
among firms about government assistance. The lack of databases is partly because most Indonesian enterprises
are not registered. If more firms were registered, more firms would have received government assistance.

c Sukamdani 2020
d A World Bank firm-level survey conducted in March 2021 shows that the share of firms experiencing a large drop in sales (>20 percent) had dropped
from 82 percent in June 2020 to 38 percent in March 2021. See World Bank 2021
e A World Bank firm-level survey conducted in October 2020 shows that 66 percent of firms had turned to the internet, social media, and specialized
apps or digital platforms in response to the pandemic. See World Bank 2020d
f World Bank 2020e
g World Bank 2020f
h World Bank 2020f; IMF (International Monetary Fund). 2021.
i Azis
June 20212020 Pathways to Middle-Class Jobs in Indonesia
 Structural Transformation toward Middle-Class Jobs 97

Box 3.2 continued

The lack of databases is partly because most Indonesian enterprises are not registered. If more firms were
registered, more firms would have received government assistance. Program awareness is an issue. For example,
by March 2021, only 16 percent of firms surveyed were aware of the government’s fiscal incentives and only 43
percent were aware of its wage subsidy program.j

Maintaining a positive macroeconomic outlook could help prevent firms from firing workers. A recent firm-level
representative survey shows that improved year-on-year sales of firms were significantly associated with higher
hiring probability, while expectation of future sales significantly correlates with a lower probability of firing.k
Stronger demand and hence higher sales via improved economic conditions could help increase the rehiring
of workers in the future, while maintaining a positive macroeconomic outlook could prevent further firing
of workers. Moreover, the same survey shows that exporting firms recovered more quickly. In the longer run,
increased competitiveness in the global market, including through a significant reduction of production costs,
could help induce a rise in exporting firms—which recover faster.

j World Bank 2020d

k World Bank 2020d

The limited impact of structural transformation on labor productivity hence

reflects two constraints: (i) Indonesian workers did not move into more productive
sectors in sufficient numbers; and (ii) for those who did, the level of productivity
of the nonagricultural sectors where they took up jobs was insufficiently high
to represent a substantial transformation. The level of labor productivity in sectors
receiving workers was not sufficiently higher compared to agriculture as to make a
significant difference in terms of productivity, and the contribution of higher-productivity
sectors to job growth was small (Figure 3.14).

Figure 3.14 Structural Change Did Not Foster Job Creation in Higher-Productivity Sectors
Contribution to total employment growth, 2000-18; average labor productivity relative to agriculture, 2000-2018

Sources: World Bank staff calculations using BPS data and Sakernas 2000–18; WDI
Note: The BPS is the National Statistical Agency (Badan Pusat Statistik). Sakernas is the National Labor Force Survey
(Survei Angkatan Kerja Nasional).

Pathways to Middle-Class Jobs in Indonesia June 2021

98 Structural Transformation toward Middle-Class Jobs

3.4 Structural Transformation 3.0—More Middle-Class Jobs in

the Nonagricultural Sector
Indonesia could engender “Structural Transformation 3.0”—the next upward shift
of workers to higher wages and welfare levels. As seen, Indonesian labor markets
have undergone two stages of structural transformation: a shift out of agricultural jobs
and high poverty into higher-productivity jobs, especially in manufacturing, and, after the
Asian financial crisis, a continued shift out of agricultural employment but increasingly into
low value-added services. Indonesia faces three important challenges as it fosters middle-
class jobs (Figure 3.15). First, Indonesia could promote productivity growth across the
board, not least in sectors with low value added, which is where most workers are currently
situated. Second, Indonesia could facilitate a more decisive shift of workers out of low-
paying and low-value activities and into better-paying sectors, firms, and jobs—this is the
essence of a third wave of structural transformation. And third, Indonesia could build a
workforce that has the necessary skills and knowledge to find and take on the new jobs
in higher value-added and internationally competitive sectors (discussed in Chapter 5). These
three elements are linked. Without sufficient productivity growth in agriculture, workers
will not be freed up for opportunities elsewhere. Without upgraded skills and information
on opportunities, workers will not be able to take advantage of increasing labor demand in
sectors that create high-quality jobs. Not all sectors have the same aggregate job-creation
potential and may create jobs for different groups (for example, rural/urban, male/female,
low educated/high educated). Targeted approaches are needed.

Figure 3.15 More and Even Better Jobs: Structural Transformation 3.0
Relative productivity and share of employment 2018 (in percent) and directions for the future

Sources: World Bank staff calculations using BPS data and Sakernas
Note: Dark green arrows represent employment shift. Light green arrows represent productivity shift. The BPS is
the National Statistical Agency (Badan Pusat Statistik). Sakernas is the National Labor Force Survey (Survei Angkatan
Kerja Nasional).

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Structural Transformation toward Middle-Class Jobs 99

Raising productivity in low-value services is necessary if poverty, vulnerability,

and economic insecurity are to be reduced. Low value-added farming and activities in
trade, accommodation and food services, and personal services account for the majority of
Indonesia’s employment; many of the activities in services cater to local markets, although
tourism services are largely externally oriented and employ more women than men. Not
all of these jobs can be replaced by high value-added jobs, especially not in the short term.
Increasing the productivity of these jobs—including those in agriculture—from the current
low base is essential to reduce economic insecurity. Sectors such as agriculture have
significant potential to increase productivity: for example, by capturing a longer supply chain
for farmers and increasing value added, or by diversifying into higher value-added crops
such as horticultural products. Tourism, similarly, can flourish if higher-end quality services
are developed. This will require investment, as well as skills training. However, increasing
productivity within sectors is not enough: the transition of workers from low-productivity
to higher-productivity sectors must also accelerate.

Nonmanufacturing industries have limited potential for creating sustainable

middle-class jobs. Although mining and utilities contribute significantly to value added (21
percent of the total), they are very capital intensive and have limited capacity to absorb labor.
The construction sector may be able to create jobs as infrastructure investments increase
but, as shown in Chapter 2, provides very few jobs with middle-class income levels or with
other aspects of middle-class jobs, such as worker protection. Moreover, the workforces in
these sectors are male dominated and cannot, over the short-to-medium term and in the
Indonesian context, be expected to create jobs for women.

Higher value-added services sectors have potential to grow and to create jobs, but
the overall effect will be limited over the medium term. Jobs in higher value-added
services—such as information and communications technology (ICT) or finance—are modern,
often highly paid, and employ a large share of highly educated workers. As discussed in more
detail in Chapter 5, when women acquire jobs in high-value services, they earn more, on
average, than men. Encouragingly, employment in high-end services has, since 2010, grown
faster than employment in low-value services. Investment in high-end services may well
increase demand for skills and higher education, thus boosting aggregate human capital, and
may lower the out-migration of highly skilled Indonesians. However, these services cannot
be counted on to provide the main bulk of jobs in the short and medium term. Currently,
high-value services employ 8 percent of the employed workforce and, even with very high
growth rates (and supply-side constraints in terms of workforce skills aside), these services
cannot be counted on to develop into a major source of labor demand. They are starting
from a very low base, but, in fact, financial services, real estate, business services, and ICT do
not account for a major share of employment even in high-income countries.

Pathways to Middle-Class Jobs in Indonesia June 2021

100 Structural Transformation toward Middle-Class Jobs

The manufacturing sector may have a better chance of providing higher-quality

jobs, absorbing workers with limited levels of education, and providing gender-
balanced job opportunities. Industrialization, driven by the manufacturing sector, helped
Indonesia recover from commodity booms in the past, for example after the oil booms of
the 1970s and early 1980s.14 Manufacturing firms, especially the medium sized and large,
have provided comparatively “good” jobs in Indonesia as they provide wage employment and
offer higher wages than agriculture or lower value-added services (Chapter 4). Unlike higher
value-added services, however, manufacturing has offered easy entry as it has required
lower skills (43 percent of workers in manufacturing have only primary levels of education—
see Chapter 4—and this is almost identical to the educational profile of workers in lower
value-added services) and has created jobs for women, who currently make up 45 percent
of the workforce in the sector.15 Although productivity growth over time critically depends
on higher skill levels—and strengthening the human capital of Indonesia’s population is one
of the government’s main goals—a strategy to help create jobs for the current stock of
workers will need also to foster jobs requiring low- and mid-level skills. From this perspective
and at least over the short and medium term, manufacturing has significant potential for
creating good jobs for those currently in more insecure jobs—as indeed it has already done
in many other Asian countries.16

14 Wihardja 2016
15 World Bank staff calculations using data from Sakernas
16 Cunningham and Pimhidzai 2018; Cunningham et al. 2019

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Structural Transformation toward Middle-Class Jobs 101

3.5 Leveraging FDI for Job Creation

Indonesia needs FDI to finance its investment-savings gap. At present, Indonesia relies
on portfolio investment to cover its current account deficit. These short-term capital inflows,
while accessible to Indonesia, are inherently volatile. Foreign direct investment provides an
alternative and more stable source of foreign capital that, under the right circumstances,
could be tapped for job creation.

However, so far, Indonesia has not succeeded in attracting sufficient FDI, especially
export-oriented and job-creating FDI. Indonesia has the lowest levels of FDI as measured
by GDP per capita among Asian comparators (Figure 3.16). Indonesia’s underperformance
reflects an overall poor investment climate, as well as constraints specific to FDI. General
reforms have helped Indonesia improve its overall business climate and move up significantly
in the World Bank’s Ease of Doing Business rankings, from 120th in 2014 to 73rd in 2019.
However, Indonesia is ranked below Vietnam (70) and far below China (31), Thailand (21),
and Malaysia (12). The improvement does not capture the fact that the investment climate
remains unfavorable for foreign companies, reflecting the inward-looking nature of
Indonesia’s current policy approach and continued struggles to reform behind-the-border
issues. Until recently (see Box 6.1), Indonesia had some of the tightest restrictions on FDI,
including in local content requirements and limits on foreign ownership. Trade barriers
remain high, hindering investment in export-driven sectors built on GVCs—these export-
driven sectors face barriers in importing some types of capital goods and raw materials.
Beginning with the commodity boom that occurred after the Asian financial crisis, FDI to
Indonesia has moved away from efficiency-seeking investments in export-oriented sectors
in favor of tariff-jumping and rent-seeking investments in natural resources and production
for the domestic market, including in utilities.

Figure 3.16 Indonesia Is Not Attracting As Much FDI As Comparator Countries

(Net inflows, percent of GDP, 2000–18)

Sources: World Development Indicators

Pathways to Middle-Class Jobs in Indonesia June 2021

102 Structural Transformation toward Middle-Class Jobs

Foreign direct investment can accelerate structural transformation and should

be prioritized from a jobs perspective. By bringing in new technology and raising labor
productivity, FDI can generate wage increases.17 Middle-income countries in Latin America
and Asia have seen increasing aggregate labor demand and employment as a result of FDI
inflows into manufacturing.18 Some, although not all, FDI results in improved organizational
and technical improvements among local providers or generates higher labor productivity
overall through technology spillovers to domestic firms.19 By offering both higher
employment growth and higher labor productivity, FDI can increase the rate of structural
transformation. Estimates suggest that were Indonesia to further relax its constraints on
FDI, both foreign and domestic investment would increase in industry, as well as in tradable
services.20

However, FDI may also increase the wage gap between skilled and unskilled workers
and contribute to wage inequality. In fact, FDI flows are likely to have different effects
on job creation and wage inequality, depending on the destination sector. This holds even
more for a country like Indonesia with limited human capital, a factor that blocks many
workers from accessing higher-skill jobs. In general, the evidence suggests that FDI in low-
skill manufacturing has the greatest effect on labor demand, creating the most jobs, and
does not introduce large skill premiums. Higher-skill manufacturing may have a greater
effect on skill premiums, but still increases aggregate demand. Foreign direct investment
in the extractives industry, a capital-intensive sector, is not likely to create many new jobs.

The relationship between FDI in services and employment is more complex. Estimates
suggest that if restrictions were relaxed, domestic investment in nontradable services would
fall as FDI inflows increased. The direct job-creation effects may be mixed. Some investment
in low-end services (for example, retail) may be more inward looking, focusing on accessing
Indonesia’s large domestic market and as such would replace incumbents, rather than
increase aggregate labor demand. However, investment restrictions, for example, in the
mini-marts sector, may raise prices and constrain productivity growth in retail. Meanwhile
investments in services exports—for example, focusing on tourism—could potentially create
more middle-class jobs by increasing productivity in the services exports sector. On the
other hand, FDI in services are known to improve productivity in other sectors and can
therefore help indirect job creation. Evidence from the Czech Republic, India, and Indonesia
shows that opening up services to foreign investors improves the productivity of domestic
firms in manufacturing, which helps to create jobs and raise average wages in this sector.21

17 Steenbergen and Trang 2020

18 Hale and Xu 2016
19 See, for example, Calì, Doarest, and Presidente (Forthcoming); Tong, Kokko, and Seric (2019); and World Bank (2020a).
20 Calì et al. Forthcoming
21 Arnold, Javorcik, and Mattoo 2011; Arnold et al. 2016; Duggan, Rahardja, and Varela 2013

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Structural Transformation toward Middle-Class Jobs 103

Boosting export-oriented FDI in Indonesia is an effective pathway to creating

better jobs and can have strong gender effects. In Indonesia, as in many other
countries, medium-sized and large manufacturing plants that are foreign owned and/or
export oriented have higher labor productivity, tend to generate more jobs, and pay higher
wages on average (Figure 3.17 and Figure 3.18). Foreign-owned manufacturing plants—mostly
food processing firms and textiles factories—are also more export intensive. Apart from
the direct potential contribution to employment, foreign-owned plants can also generate
indirect effects. Vertical technology (or productivity) spillovers can happen through links
with domestic suppliers and customers. Horizontal spillovers (to competitors) happen
through demonstration or imitation effects and through increased competition that can
spur innovation. These technology spillovers are far from automatic, however, and depend
on the strength of links, the types of firms, and the domestic absorptive capacity, among
others.22 District-level analysis of the local effects of manufacturing FDI in the period 2007—
15 shows that when an additional foreign-owned manufacturing firm establishes itself in a
district, total employment increases by 0.4 percent. Extrapolations to national level suggest
that manufacturing FDI created some 4.6 million additional jobs in total in 2007–15, of which
two-thirds were in the manufacturing sector itself.

Figure 3.17 Foreign-Owned and/or Exporting Plants Figure 3.18 . . . And Create More Jobs
Pay Higher Wages on Average . . . Average employment by manufacturing plant type,
Average annual wages by manufacturing plant type, Rp millions number of workers
Average annual wages in Rp, million

Average employment

Source: World Bank staff calculations based on the Manufacturing Source: World Bank staff calculations based on the Manufacturing
Survey 2015. Survey 2015.
Note: The Manufacturing Survey is the Survey of Medium-Sized and Note: The Manufacturing Survey is the Survey of Medium-Sized and
Large Manufacturing Firms (Survei Industri Besar/Sedang). Large Manufacturing Firms (Survei Industri Besar/Sedang).

22 Cunningham and Pimhidzai 2018; Le Quoc and Pomfret 2011

Pathways to Middle-Class Jobs in Indonesia June 2021

104 Structural Transformation toward Middle-Class Jobs

Manufacturing FDI can accelerate structural transformation in Indonesia by

changing the employment and output structure, and also affects employment and
productivity broadly in the economy. The entry of a foreign-owned firm into a district
improves resource allocation and shifts the structure of employment toward better jobs,
by providing a first stepping-stone out of low-productivity sectors. District-level analysis
shows that entry of foreign-owned firms brings a shift out of low-skill services and into
manufacturing jobs, but does not have an effect on agricultural employment or high-
skill services. In total, FDI provided an additional 2.9 percentage points to the share of
manufacturing employment in the period 2007–15, and the contribution of FDI increased
over time (even as the contribution of local manufacturing to total employment fell after
2012) (Figure 3.19).

Figure 3.19 Foreign Direct Investment Increases the

Share of Manufacturing Jobs in Total Employment
Share of manufacturing employment in total, without foreign direct
investment (FDI) and FDI effect, percent

Sources: World Bank staff estimates based on the Manufacturing Survey, Sakernas, and National Accounts.
Note: The Manufacturing Survey is the Survey of Medium-Sized and Large Manufacturing Firms (Survei Industri
Besar/Sedang). Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Manufacturing FDI has generated productivity spillovers in Indonesia. Multinationals

also have a strong impact on total output in the districts they enter, both directly and
through links to other sectors. In fact, the entry of a multinational firm increases real
manufacturing and services output equally and has a small negative effect on agricultural
output. As a result, increasing manufacturing FDI is associated with a significant increase in
output per worker in manufacturing, as well as in services. In the manufacturing sector, FDI
has provided more value-added jobs, because it brings with it greater technology intensity,
management and organization skills, and integration in global markets. In the services sector,
output per worker likely increased as a result of the destruction of low-productivity jobs as

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Structural Transformation toward Middle-Class Jobs 105

workers took up manufacturing employment instead. As seen in Table 3.1, the effects on
productivity were substantial in both manufacturing and services sectors.

Table 3.1 Foreign Direct Investment Increases Productivity in Manufacturing and Services

Primary sector
Manufacturing sector output per Service sector output per worker
(agriculture and extractives)
Year worker (Rp, millions) (Rp, millions)
output per worker, (Rp, millions)

Current No FDI Effect Current No FDI Effect Current No FDI Effect

2007 121 80 40 33 37 –4 86 69 17
2008 128 88 40 37 41 –4 88 70 17
2009 129 89 40 37 41 –4 91 74 18
2010 126 90 37 41 45 –4 96 79 18
2011 122 83 39 50 55 –5 103 83 20
2012 118 81 37 53 58 –6 105 85 20
2013 129 91 39 57 63 –6 113 93 20
2014 136 97 39 60 66 –6 116 96 20
2015 143 102 41 60 66 –6 117 98 20
Average 128 89 39 48 52 -5 102 83 19

Sources: Sakernas 2019; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

In all, these shifts point to the strong potential impact that FDI can have on
structural transformation. Workers left low-end services and the primary sector to take
up jobs in manufacturing, benefiting immediately from higher productivity. Workers who
remained in low-value services likely benefited from the reduction of crowding in services
catering to local demand and saw increases in productivity. The increase in productivity due
to FDI was in fact substantial (Figure 3.20).

Pathways to Middle-Class Jobs in Indonesia June 2021

106 Structural Transformation toward Middle-Class Jobs

Figure 3.20 Foreign Direct Investment in Manufacturing Fosters Labor

Productivity Increases in Both Manufacturing and Services
Employment share and output per worker in 2015, with and without manufacturing
foreign direct investment (FDI)

Sources: World Bank staff estimates based on the Manufacturing Survey, Sakernas, and National Accounts.
Note: The Manufacturing Survey is the Survey of Medium-Sized and Large Manufacturing Firms
(Survei Industri Besar/Sedang). Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

3.6 Conclusions
This chapter has addressed the role of Indonesia’s growth model and labor
productivity growth for accelerating the move toward middle-class jobs. In
particular, the chapter has focused on structural transformation of employment, from
lower-productivity to higher-productivity activities, as a key mechanism for increasing the
quality of jobs. The chapter has shown that a growth model focused on more openness
to international trade and investment can help accelerate such a transformation. The
implication is that Indonesia could improve the investment climate for firms overall, adapt
the regulatory framework to make the country an attractive destination for FDI that can
generate good jobs and spillovers, and pay specific attention to increasing labor demand in
sectors that can offer better jobs.

Stable and relatively high economic growth has delivered better jobs, but these
have not been good enough to shift a substantial part of the labor force into the
middle class. The incomplete transformation of labor markets from low-income activities
to middle-class income levels is due to both insufficient labor productivity growth and a
disconnect between sectors that create value added and those that create jobs. Labor
productivity—a key correlate of wages and other quality indicators of jobs—has not been
growing in the sectors into which most workers moved.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Structural Transformation toward Middle-Class Jobs 107

The increasing dependence on commodity and nontradable sectors for growth

has hurt labor productivity and job creation. In the past forty years, Indonesia has
experienced two key stages of structural transformation. The first stage, from the 1980s
until the Asian financial crisis, was characterized by increased international openness
and rapid industrialization, especially through growth of the manufacturing sector and
merchandise exports, and job creation in nonagricultural sectors. After the crisis, by contrast,
Indonesia’s growth was driven by commodity exports, which are not labor intensive, and a
more inward-looking economic policy, characterized by lower job creation in manufacturing
and by exports. Workers have continued to leave the agricultural sector, but have largely
been confined to jobs in the low value-added services sector, while both output and job
creation in manufacturing stagnated. As a result, labor productivity growth resulted from
productivity growth within certain sectors (especially in agriculture), but not necessarily in
sectors that created good jobs. Overall labor productivity growth was also held back by the
fact that workers did not move into more productive sectors.

“Structural Transformation 3.0” will therefore need a growth model that helps
reconnect labor productivity with jobs. Productivity-driven structural transformation
needs to accelerate if more and better jobs are to be created. This will require Indonesia
to: (i) promote labor productivity growth across the economy, including in low-productivity
sectors such as agriculture and trade where most jobs are currently found; (ii) foster labor
demand in nonagricultural sectors with higher labor productivity, especially manufacturing
and higher value-added services; and (iii) assist workers in accessing those jobs, by closing
gaps related to skills, information, and geographical mobility.

Foreign direct investment can provide a significant impetus in fostering both

higher productivity and better jobs. Inward-looking policies have not delivered sufficient
growth or jobs. More productive jobs will require policies that help Indonesia to open up
further to the opportunities offered by international trade and investment. Foreign
direct investment, especially that which is export oriented, can accelerate structural
transformation as it offers both higher employment growth and higher labor productivity.
In addition, foreign firms can afford to pay higher wages, and often generate productivity
spillovers to other sectors.23 In the manufacturing sector, FDI has delivered these benefits
in the past. In other sectors, especially services, FDI can generate direct and indirect job
creation, but is also critical for increasing the overall effectiveness and competitiveness of
the economy by raising the quality of services.

23 Chang et al. 2019

Pathways to Middle-Class Jobs in Indonesia June 2021

108 Structural Transformation toward Middle-Class Jobs

These findings point to the need for Indonesia to leverage international markets
for investment, inputs, talent, and innovation in order to foster labor demand in
sectors that can create jobs. Increasing the overall competitiveness and resilience of
the Indonesian economy is essential for creating jobs in productive sectors and requires
a stronger investment climate, a competitive, fair and level playing field, investments in
infrastructure and human capital, financial-sector deepening and reform, and responsible
management of natural resources.24 The policy recommendations below related to
structural transformation focus on specific areas that could catalyze labor demand. The
recommendations relate to increasing international openness by encouraging FDI and trade,
which will benefit both foreign and domestic job-creating firms.25 Subsequent chapters
provide a comprehensive take on the overall enabling environment that firms need (Chapter
4), and on preparing Indonesia’s firms (Chapter 4) and workforce (Chapter 5) to access
productive opportunities and middle-class jobs.

Increasing labor productivity in low-productivity sectors such as agriculture will

be essential for fostering structural transformation and improving the quality of
workers at the lower end of the earnings scale. Productivity growth in agriculture,
coupled with outflows of workers to nonagricultural sectors, has been a critical factor in
reducing rural poverty in Indonesia. The higher agricultural productivity has freed up labor
that was previously underemployed in subsistence farming for more productive jobs in
industry and services. Agriculture remains affected by inadequate resource management
(land expansion, deforestation, forest degradation), crops that are poorly matched to land
quality, lack of extension services to facilitate transition to crops with higher yields, trade
restrictions that distort incentives (overproduction of basic crops), and lack of links to
manufacturing industries in, for example, food processing.26 Addressing these distortions
could unlock significant jobs benefits in both agricultural and nonagricultural sectors.

Diversifying and opening up Indonesia’s economy will require broad and significant
policy reform to reverse ground lost during and after the commodity boom and
to meet the challenges related to rapid changes in technology and globalization.
Both the level of FDI and the job creation by FDI are below potential in Indonesia. The
export orientation of FDI has fallen: in 2014, only 35 percent of new foreign manufacturing
plants were export-oriented, compared to 58 percent in 1996.27 Indonesia’s overall export
performance has been worsening, especially compared to competitors (Vietnam, Thailand),
and especially in goods manufacturing. Openness to the world economy comes with risks,
but has historically proven to provide long-term gains that outweigh the costs. However,
the routes to international markets look very different from those that existed during
the rapid industrialization of the 1990s, when low-skill manufacturing offered significant
opportunities for countries with large labor surpluses. Globally, the manufacturing sector
is particularly affected by rapidly changing technologies and intensified knowledge

24 Chang et al. 2019

25 World Bank 2020a; Chang et al. 2019; World Bank 2019
26 World Bank 2020a
27 World Bank 2020a

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Structural Transformation toward Middle-Class Jobs 109

content of products and services, increasing global protectionism, and shifting global
trade patterns. Indonesia’s manufacturing sector will also be affected by regional trade
patterns and the proliferation of bilateral, plurilateral, and regional trade agreements, an
area in which Indonesia lags behind many other countries in the region. Its manufacturing
sector will also be affected by a changing regional context in terms of markets, investment,
and competition (including China’s rise followed by its more recent growth slowdown, and
Vietnam’s catch-up with Indonesia in skills and competitiveness). Low productivity growth
has also meant that Indonesia has lost competitiveness compared to regional competitors.
These shifts pose challenges to policy reforms designed to boost exports. The alternative—
continuing on the inward-looking path—has not proved successful, however. To address the
challenges related to international integration, Indonesia will need to foster innovation and
growth in knowledge-intensive industries and activities. This will require reforms to increase
the competitiveness, capabilities, and connectedness of its economy.28

Indonesia could move decisively and benchmark itself firmly within its regional—
and highly competitive—context. Indonesia competes primarily with countries with
more favorable investment climates (for example, Thailand) or lower labor costs (Vietnam).
Indonesia has taken steps to improve its investment climate, but these reforms have not
been sufficient to induce large-scale inflows of FDI. Indonesia could focus on improving the
access of foreign investors to a greater number of sectors in the Indonesian economy,
attracting job creators from abroad, and strengthening the behind-the-border investment
climate by increasing the overall efficiency and competitiveness of the economy. In parallel,
it will be important to improve the access of local firms and workers to opportunities
offered by increased openness—by fostering technology spillovers, raising skill levels, and
reducing information gaps. Rather than providing a plethora of specific tax and other costly
targeted FDI incentives, Indonesia could strive to create a level playing field by removing
unnecessary and distortive policies and developing a credible and predictable framework.
This will favor competitive domestic firms and attract more efficiency-seeking and export-
oriented FDI with the potential for job creation, higher productivity, and higher wages,
rather than rent-seeking and inward-looking FDI with few effects on labor markets. Analyses
by the World Bank focusing on fostering sustainable development in Indonesia29 show that
this will require reform of regulatory frameworks for investment and trade, of systems for
skills development, migration law, and more.

28 Hallward-Driemeier and Nayyar 2018

29 World Bank 2020a; Chang et al. 2019; World Bank 2019

Pathways to Middle-Class Jobs in Indonesia June 2021

110 Structural Transformation toward Middle-Class Jobs

Strengthening Indonesia’s attractiveness to FDI will serve to create labor demand

both in foreign and domestic firms. Policies to draw in more, more effective, and more
job-creating FDI will require Indonesia to: (i) reduce the restrictions on FDI and thereby
increase investment by job creators and improve the investment climate for both domestic
and foreign firms; (ii) reduce trade barriers, both tariff and nontariff measures, to provide
domestic and foreign firms with access to high-quality inputs; and (iii) remove restrictions
on hiring foreign talent to fill roles for which there is a shortage of domestic workers.

3.7 Recommendations
First, relaxing restrictions on FDI in the Negative Investment List would increase
foreign investment and crowd in domestic investment, especially in tradable
sectors. There is evidence that removing FDI restrictions would substantially increase
foreign investment, crowd in domestic investment,30 and increase technological spillovers
and productivity growth in domestic firms31 in Indonesia. Until recently, by international
standards Indonesia had one of the most restrictive FDI regimes (see Box 6.1). The Negative
Investment List includes limits on foreign equity, minimum local content requirements,
outright prohibition of investment in specific sectors including utilities, and a range of
sectors that are reserved for medium-sized and small enterprises, which thereby indirectly
prohibits foreign (large) multinational companies from entering. These restrictions,
rather than fostering a competitive domestic firm sector and job creation, are limiting
the sophistication and competitiveness of Indonesian goods and services and distorting
incentives for expansion. Foreign equity restrictions are particularly severe in the services
sectors; many of these sectors are important for increasing the competitiveness of the
economy—by efficiently providing high-quality business services, the broader economy
benefits. These restrictions reduce the quality and/or increase the costs of domestic
services, with negative implications for the productivity of firms in downstream sectors.
Combined with reforms to build a middle-class workforce (see Chapter 5), openness to more
FDI should slowly tilt the scale of net job creation toward sectors that are likely to create
more middle-class jobs.

In addition to across-the-board, (almost) sector-blind reforms to attract foreign and

domestic investment, some aspects of investment reform can be prioritized for sectors
that are either potential job creators, especially creators of middle-class jobs (for example,
tourism, ICT, finance, health, education, and specific manufacturing industries), or sectors
that are critical for the competitiveness of the overall investment climate (for example,
electricity and gas supply), or that are potentially transformative to low-productivity
sectors, especially agriculture. This sector-specific investment targeting could, for example,
prioritize investment promotion for infrastructure development projects in sectors that
are more likely to be sources of middle-class jobs (and in other sectors that support these
sectors). The government could dedicate more resources toward meeting the needs of
these prioritized sectors.

30 Calì et al. Forthcoming

31 Calì, Doarest, and Presidente. Forthcoming

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Structural Transformation toward Middle-Class Jobs 111

Second, lowering the high cost of trade would increase firms’ access to quality
inputs. Export-oriented multinationals that are integrated into GVCs also source inputs (and
talent) on international markets. Although the purpose of import restrictions (tariff and
nontariff measures) is to nurture local production, the high cost of trade translates into
restricted access to quality inputs. This in turn lowers Indonesia’s attractiveness, especially
for outward-oriented firms, and prevents domestic firms from competing successfully
in export markets. Between 2000 and 2017 Indonesia increased its average tariffs on
imports, especially for capital and intermediate goods which affect firm productivity, and
also intensified nontariff measures.32 Cumbersome product certification processes and
preshipment inspections raise the indirect costs of importing. The system is also open
to discretionary (and often inconsistent) treatment by different ministerial agencies and
subnational governments. Indonesia could consider lowering the high cost of trade by both
streamlining and lowering tariffs and reducing nontariff measures. Recent reforms have,
however, helped to ease these constraints (see Box 6.1).

Trade agreements, while not a substitute for unilateral trade reforms, can help
Indonesia access other markets for both imports and exports and increase the
credibility of its economic reforms. Indonesia has signed many preferential trade
agreements to obtain market access for its exports. However, Indonesia lags behind
regional peers such as Vietnam or Malaysia in terms of the scope and ambition of these
agreements. Several bilateral trade agreements have expired and Indonesia still has to
finalize negotiations for others, including the EU Comprehensive Economic Partnership
Agreement. Trade agreements afford opportunities beyond export market access. For
example, they could include ambitious investment measures, which could help foster FDI.
As the agreements encompass a range of areas relating to the competitiveness and level
playing field of Indonesia’s economy (procurement, competition policy, and public sector
reform, among others) they may also serve to lock in domestic reforms in areas relating to
the competitiveness of Indonesia’s economy, including on foreign equity.

32 World Bank 2019

Pathways to Middle-Class Jobs in Indonesia June 2021

112 Structural Transformation toward Middle-Class Jobs

Third, firms active in Indonesia could be allowed to fill skills gaps by hiring foreign
talent for roles for which there is a shortage of domestic workers. Indonesia does
not currently have the workforce necessary to fill all middle-class jobs (Chapter 5), and skills
gaps are particularly serious in manufacturing and higher value-added services. Managerial
capacity is a critical factor for business survival and growth,33 but almost 80 percent of
Indonesian firms report that they are unable to fill managerial vacancies. Although investment
in human capital is a top priority for Indonesia, reforms will take time to bear fruit and, at
any rate, such reforms are complementary to foreign-born talent, not a substitute for it. In
fact, foreign talent can bring skill spillovers to domestic workers and firms and may foster
domestic job creation as businesses identify new opportunities through imported skills.34, 35
In spite of the unsatisfied demand for skilled labor, Indonesia’s share of foreign labor in its
workforce is low (0.06 percent in 2016), compared to Asian competitors (more than twice
as high in Vietnam at 0.15 percent, and nearly 5 percent in Thailand). Restrictions on hiring
foreign workers make it difficult to fill such posts.36 Indonesia has, however, moved toward
increased hiring of foreign workers to fill roles for which there is a shortage of skilled
domestic candidates through recent reforms (Presidential Regulation No.  28/2018 and
Ministry of Manpower Decree No. 229/2019). These reforms are deepened by more recent
reforms under Government Regulation No. 34/2021 on Foreign Workers. There remains,
however, room to improve access to foreign talent to fill roles for which there is a shortage
of domestic workers (see Box 6.1).

33 Bloom et al. 2018; Bloom et al. 2010

34 Manning and Aswicahyono 2012
35 Foreign talent can bring skill spillovers to domestic workers following the same logic of spillovers from foreign firms to
domestic firms. A nascent literature on the economywide impacts of high-skilled migration in receiving countries suggests
small but positive impacts on productivity, innovation, trade, and entrepreneurship (Nathan 2014).
36 World Bank 2019

June 2021 Pathways to Middle-Class Jobs in Indonesia

 References 113

References
Arnold, Jens M., Beata Javorcik, Molly Lipscomb, and Aaditya Mattoo. 2016. “Services Reform and
Manufacturing Performance: Evidence from India.” The Economic Journal 126 (590): 1–39.

Arnold, Jens M., Beata Javorcik, and Aaditya Mattoo. 2011. “Does Services Liberalization Benefit
Manufacturing Firms? Evidence from the Czech Republic.” Journal of International Economics
85 (1): 136–46.

Azis, Iwan. 2020. “Coronavirus-Driven Crisis: This Time Is Different.” Slide presentation at an Australian
National University webinar, held on June 3.

Bloom, Nicholas, Aprajit Mahajan, David McKenzie, and John Roberts. 2010. “Why Do Firms in
Developing Countries Have Low Productivity?” American Economic Review 100 (2) 619–23.
https://www.aeaweb.org/articles?id=10.1257/aer.100.2.619.

Bloom, Nicholas, Aprajit Mahajan, David McKenzie, and John Roberts. 2018. “Do Management
Interventions Last? Evidence from India.” Policy Research Working Paper 8339, World Bank,
Washington, DC.

Calì, Massimiliano, Martín Cicowiez, Aufa Doarest, Taufik Hidayat, and Dhruv Sharma. Forthcoming.
“The Economic Impact of Investment Provision: Evidence from Indonesia.” Working paper,
World Bank.

Calì, Massimiliano, Aufa Doarest, and Giorgio Presidente. Forthcoming. “Foreign Entry and Domestic
Performance: Evidence from Indonesia.” Working paper, World Bank.

Chang, Philip Han, Rachita Daga, Gautam Vishram Mehta, and Massimiliano Calì. 2019. Creating Markets
in Indonesia. Unlocking the Dynamism of the Indonesian Private Sector: A Country Private Sector
Diagnostic. Washington, DC: World Bank Group.

Cunningham, Wendy, Claire H. Hollweg, Johanne Buba, Marong Chea, Dilaka Lathapipat, Une Lee, Anne
Lopez, Sodeth Ly, Lan Nguyen, Miguel Sanchez, and Maheshwor Shrestha. 2019. Cambodia’s
Future Jobs: Linking to the Economy of Tomorrow : Summary. Washington, DC: World Bank.

Cunningham, Wendy, and Obert Pimhidzai. 2018. Vietnam’s Future Jobs : Leveraging Mega-Trends for
Greater Prosperity: Main Report (Vietnamese). Washington, DC: World Bank.

Diop, Ndiame. 2016. “Making Indonesian Manufacturing an Engine of Growth Again: Now or Never?”
Internal PowerPoint presentation, World Bank, Jakarta.

Duggan, Victor, Sjamsu Rahardja, and Gonzalo Varela. 2013. “Service Sector Reform and
Manufacturing Productivity: Evidence from Indonesia.” Policy Research Working Paper 6349,
World Bank, Washington, DC.

Gil Sander, Frederico, and Pui Shen Yoong. Forthcoming. “Structural Transformation and Labor
Productivity in Indonesia: Where Are All the Good Jobs?” Background paper, Pathways to
Middle-Class Jobs in Indonesia, World Bank, Jakarta.

Hale, Galina, and Mingzhi Xu. 2016. “FDI Effects on the Labor Market of Host Countries.” Working Paper
2016-25. Federal Reserve Bank of San Francisco, San Francisco, CA.

Hallward-Driemeier, Mary, and Gaurav Nayyar. 2018. Trouble in the Making? The Future of Manufacturing-
Led Development. Washington, DC: World Bank. doi:10.1596/978-1-4648-1174-6. License:
Creative Commons Attribution CC BY 3.0 IGO.

Pathways to Middle-Class Jobs in Indonesia June 2021

114 References

IMF (International Monetary Fund). 2021. “Recessions and Recoveries in Labor Markets: Patterns,
Policies, and Responses to the Covid-19 Shock.” In World Economic Outlook: Managing
Divergent Recoveries. Washington, DC: IMF.

Le Quoc, Hoi and Richard Pomfret. 2011. “Technology Spillovers from Foreign Direct Investment in
Vietnam: Horizontal or Vertical Spillovers?” Journal of the Asia Pacific Economy 16 (2): 183–
201.

Manning, Chris, and Haryo Aswicahyono. 2012. “Trade and Employment in Services: The Case of
Indonesia.” Employment Working Paper No. 132. International Labour Office (ILO), Geneva.

Nathan, Max. 2014. “The Wider Economic Impacts of High-Skilled Migrants: A Survey of the Literature
for Receiving Countries.” IZA Journal of Migration 3, 4. https://doi.org/10.1186/2193-9039-3-
4.

Patunru, Arianto. 2019. “Rising Economic Nationalism in Indonesia.” In The Indonesian Economy in
Transition: Policy Challenges in the Jokowi Era and Beyond, edited by Hal Hill and Siwage Dharma
Negara. Singapore: ISEAS.

Steenbergen, Victor, and Hebous, Sarah. 2020. “The Effect of FDI on Indonesia’s Jobs, Wages and
Structural Transformation.” Background paper, Pathways to Middle-Class Jobs in Indonesia,
World Bank, Jakarta.

Steenbergen, Victor, and Trang, Thu Tran, 2020. “The Distributional Effects of FDI: Evidence from
Ethiopia, Vietnam, and Turkey.” In Global Investment Competitiveness Report 2019/2020:
Rebuilding Investor Confidence in Times of Uncertainty. Washington, DC: World Bank
Publications.

Sukamdani, Hariyadi. 2020. “Changes of the Tourism Paradigm in the Era of New Normal.” Slide
presentation to the Indonesia Tourism Forum. May 15. Asosiasi Pengusaha Indonesia
(APINDO), Jakarta.

Tong, Yee-Siong, Ari Kokko, and Adnan Seric. 2019. “Linking FDI and Local Firms for Global Value Chain
Upgrading: Policy Lessons from Samsung Mobile Phone Production in Vietnam.” Inclusive and
Sustainable Industrial Development Working Paper Series WP5, 2019. UNIDO (United Nations
Industrial Development Organization), Vienna.

United Nations. 2019. World Urbanization Prospects 2018: Highlights. Department of Economic and
Social Affairs, Population Division. New York: United Nations. https://population.un.org/
wup/Publications/Files/WUP2018-Highlights.pdf.

Wihardja, Maria Monica. 2016. “Natural Resources Management: The Effect of the Commodity Boom
on Indonesia’s Macroeconomic Fundamentals and Industrial Development.” International
Organizations Research Journal 11 (1): 48–70.

World Bank. 2019. “Indonesia Development Policy Review.” Internal document, World Bank, Jakarta.

World Bank. 2020a. Indonesia Systematic Country Diagnostic. Indonesia: Eliminating Poverty, Bringing
Economic Security to All. World Bank.

World Bank. 2020b. World Development Report 2020: Trading for Development in the Age of Global Value
Chains. Washington, Washington, DC: World Bank. doi:10.1596/978-1-4648-1457-0. License:
Creative Commons Attribution CC BY 3.0 IGO.

World Bank. 2020c. No One Left Behind. Rural Poverty in Indonesia. Washington, DC: World Bank.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 References 115

World Bank. 2020d. “Covid-19 Impact on Firms in Indonesia: Results from the Covid-19 Business Pulse
Survey, Round 1: May–June 2020” and “Covid-19 Impact on Firms in Indonesia: Results from
the Covid-19 Business Pulse Survey, Round 2: October 2020.” World Bank.

World Bank. 2020e. “COVID-19 in Indonesia: Impact and Responses.” Internal World Bank note prepared
by Frederico Gil Sander, Bertine Kamphuis, and team. April 1.

World Bank. 2020f. “The Jobs Response to Covid-19 in East Asia Pacific.” Internal document prepared
by the Social Protection and Jobs Global Practice. May 27.

World Bank. 2021. “Covid-19 Impact on Firms in Indonesia: Results from the Covid-19 Business Pulse
Survey, Round 3: March 2021.” World Bank.

Pathways to Middle-Class Jobs in Indonesia June 2021

116 Enterprise-Level Dynamics of Job Creation

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 117

Chapter 4

Enterprise-Level
Dynamics of
Job Creation 1

4.1 Overview
Most enterprises in Indonesia do not create jobs, let alone middle-class jobs. In 2016,
an average nonagricultural enterprise in Indonesia had less than three employees by the
fifth year of operation and needed another 15 years to grow to become an enterprise of
five employees (Figure 4.1). In fact, more than 98.75 percent of all enterprises in Indonesia
in 2014 were microenterprises,2, 3 defined as having fewer than five employees. While
the number of enterprises is growing and jobs are being created, employment growth
within these firms is limited. In the manufacturing sector, microenterprises tend to be low
productivity, accounting for 91.6 percent of the total number of establishments in 2015,
but only 6.2 percent of total value added (at market prices)—with limited profits to boost
firm owners or their employees into the middle class.

1 This chapter is based on Alatas and Wihardja (Forthcoming); Steenbergen and Hebous (2020); and Cunningham and
Syamsulhakim (Forthcoming).
2 OECD (Organisation for Economic Co-operation and Development) 2018
3 Data from the Ministry of Cooperatives and SMEs refer to establishments, rather than enterprises, and include the
agricultural sector.

Pathways to Middle-Class Jobs in Indonesia June 2021

118 Enterprise-Level Dynamics of Job Creation

Figure 4.1 A Firm Does Not Significantly Increase in Size until it Has Been
Operating for 22 Years
Employment life cycle of nonagricultural enterprises, 2016; number of workers (y-axis); age (x-axis)

Sources: Economic Census 2016; World Bank staff calculations

Note: The authors regressed number of employees on age dummies to produce an employment life cycle.

In contrast, Indonesia’s medium-sized and large manufacturing firms have higher

productivity, pay higher wages and benefits, and grow more relative to micro and
small enterprises. Medium and large firms, in particular foreign firms, are also more formal
(a larger share have legal status) and hence more are subject to labor regulations—including
those that protect workers—compared to micro and small enterprises. In other words, wage
jobs in medium and large manufacturing firms provide important pathways to the middle
class. The Shapley decomposition analysis laid out in Chapter 2 confirms this.

Relative to other sectors, manufacturing firms, in particular medium-sized and

large manufacturing firms, are key creators of middle-class jobs in Indonesia. In
absolute terms, the manufacturing industry employs the largest number of workers with
wages above the consumption threshold of a middle-class population ( Table 4.1).

Table 4.1 Manufacturing Sector as a Source of Middle-Class Jobs in 2018 

Number of workers

Income earners (wage employees, self- Income earners with wages above middle-
Employment sector
employed, casual workers) class consumption cutoff
Agriculture, forestry, livestock, and
14,371,642 509,462
fisheries
Mining and quarrying 1,287,197 375,382
Manufacturing 14,539,543 2,211,864
Electricity and gas 322,754 81,625
Water supply, sewerage, waste
410,302 53,920
management and remediation
Construction 7,783,340 574,070

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 119

Table 4.1 continued

Wholesale and retail trade, vehicle
14,424,190 1,671,462
repair
Transportation and storage 5,194,341 791,363
Accommodation and food service
4,403,044 441,671
activities
Information and communication 765,874 250,353
Financial and insurance activities 1,781,338 770,794
Real estate activities 366,782 111,516
Business activities 1,438,853 426,818
Public administration, defense, and
4,681,280 1,974,729
compulsory social security
Education activities 5,956,585 1,730,386
Human health and social work activities 1,809,562 484,845
Other service activities 5,497,128 334,399

Sources: Sakernas 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Although job creation in medium-sized and large manufacturing firms was

historically high during the industrialization of the 1980s and early 1990s, it was
largely destroyed during the “hollowing-out” decade that followed the Asian
financial crisis (Figure 4.2 and Annex Figure 4.1).4 Starting in the late 1980s, Indonesia
was able to industrialize by rapidly growing its manufactured exports. At that point,
manufactured exports were not dominated by China. Indonesia also had a large labor surplus,
which attracted labor-intensive industries. As a result, between 1991 and 1996 medium-sized
and large manufacturing firms created a net average of 258,681 jobs per year. However,
between 2000 and 2010, Indonesia entered its hollowing-out decade in which, in seven
of the ten years, medium-sized and large manufacturing firms reported negative net job
creation. This hollowing-out decade partly coincided with the commodity boom of 2001–12,
the complex “big-bang” decentralization and local democratization in Indonesia following the
fall of former president Soeharto, and the global financial crisis of 2008. Prior to the Asian
financial crisis, Indonesia carried out comprehensive and across-the-board deregulation that
led to rapid expansion of the manufacturing sector. This was not replicated after the Asian
financial crisis.5 Over the decade, medium-sized and large manufacturing firms only created
an average annual 12,798 new jobs per year. The following period saw a resumption in job
growth in medium and large manufacturing firms; they created an average of 150,355 jobs
per year during the period 2010–15.

4 Indonesia’s hollowing-out decade is similar to that which occurred in the United States, although the causes and impact
were different. Between 2000 and 2010, US manufacturing employment fell by 5.8 million—one-third—from 17.3 million to
11.5 million. The sector has only recovered 12 percent of those jobs lost. The decade was associated with social disruption,
because manufacturing was an important pathway to becoming middle class, to winning a secure job and benefits. See
Guilford 2018.
5 Kuncoro 2019

Pathways to Middle-Class Jobs in Indonesia June 2021

120 Enterprise-Level Dynamics of Job Creation

Figure 4.2 The Hollowing-Out Period of 2000–10

Net job creation (people)

Sources: Manufacturing Survey; World Bank staff calculations

Note: The year 2006 was a census year and the survey of that year contains more firms than in other years. It has been
excluded for consistency reasons. The Manufacturing Survey is the Survey of Medium-Sized and Large Manufacturing
Firms (Survei Industri Besar/Sedang).

The hollowing-out period in the manufacturing sector between 2000 and 2010
amounts to a decade of lost middle-class jobs. Not only did manufacturing cease to
be a source of growth of good-quality jobs, a significant portion of those who might have
worked in manufacturing without the hollowing out instead moved into low-paying jobs in
the service sector. During this period, inequality skyrocketed: Indonesia’s Gini coefficient
rose from 30 points in 2000 to 37.8 points in 2010, and 41.4 points in 2014, the highest
level ever recorded in Indonesia. Indonesia’s Gini, once relatively moderate by international
standards, had become high and climbed faster than the Ginis of most of its East Asian
neighbors.6

Box 4.1 Who Are Manufacturing Workers?

Manufacturing sector workers tend to be gender balanced, young, have a low- to mid-level of education, and are
blue-collar. Between 2008 and 2015, on average 57 percent of the labor force in manufacturing was male with
an average age of 33.7. In 2015, 43 percent of workers had only a primary education and another 43 percent had
a secondary education, a significant improvement from 2008 when 54 percent of manufacturing workers had
only a primary education. Eighty-three percent of workers in manufacturing in 2015 were employees; the sector
has become increasingly employer-employee based and increasingly formalized, as evidenced by the decline in
the share of casual workers in the sector. The majority of jobs in manufacturing in 2015 were blue-collar (86
percent of the labor force), while only a small share (14 percent) of the workforce held white-collar positions.

6 World Bank 2014

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 121

Box 4.1 continued

Table 4.2 Descriptive Statistics of Manufacturing Labor Force

General Level of education Employment type Occupation type

Year Self- Casual Blue- White-

Male Average None Primary Secondary Tertiary Employee
employed worker collar collar
(%) Age (%) (%) (%) (%) (%)
(%) (%) (%) (%)

2008 55 33 9 54 32 4 16 75 9 88 12
2009 54 33 12 49 35 5 11 78 11 87 13
2010 55 33 10 48 38 5 21 73 6 86 14
2011 57 33 10 49 38 3 13 81 6 87 13
2012 57 34 9 47 39 5 10 84 6 86 14
2013 58 34 9 46 41 5 11 85 4 84 16
2014 58 34 8 45 42 5 11 83 6 86 14
2015 58 35 8 43 43 6 11 83 6 84 16
Average 57 34 9 48 39 5 13 80 7 86 14

Sources: Steenbergen and Hebous 2020, using Sakernas

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

The services sector is also key to creating middle-class jobs. However, firm-level
panel data are unavailable, preventing deep analysis of the demand side of jobs.
Using data from the Economic Census 2016, the authors found that 81 percent of total
nonagricultural enterprises were in the services sector (21.7 million enterprises, 89 percent
of which were microenterprises), while 17 percent were in manufacturing, 1 percent in
other industries, and another 1 percent in construction. As Table 4.1 shows, there were 9
million middle-class jobs in total in all services subsectors combined in 2018, compared to
2.2 million middle-class jobs in manufacturing, 1.1 million in other industries, and 510,000
in agriculture. The services subsectors that created a significant number of middle-class
jobs (relative to manufacturing) include: (i) wholesale and retail trade, and vehicle repair;
(ii) public administration, defense, and compulsory social security; and (iii) education
activities. However, the lack of firm-level panel data for services7 makes it very challenging
to conduct any deep analysis on the demand side of jobs in that sector. Literature on the
topic is also extremely scarce.8

7 The 2006 and 2016 editions of the Economic Census did not collect detailed firm-level information and they are not panel-
data series. The 2009 and 2015 editions of the World Bank Enterprise Survey did not include all services subsectors and were
only representative for some subsets of Indonesian firms in two groups of services, namely “other services” and retail. For the
World Bank Enterprise Survey, 1,444 interviews were conducted in 2009 and 1,320 in 2015. Of the interviewees, 491 were
included in both surveys. Moreover, the Survey did not collect the detailed financial information needed to calculate value
added and labor productivity for the services firms. See Gomez-Mera and Hollweg (2018) for descriptive analysis of firm-level
employment growth using the 2009 and 2015 World Bank Enterprise Survey.
8 See Manning and Aswicahyono (2012); 2005 input-output tables were used in examining the creation of jobs in the services
sector from export activities.

Pathways to Middle-Class Jobs in Indonesia June 2021

122 Enterprise-Level Dynamics of Job Creation

This chapter begins to unpack the firms side of the jobs issue by exploring the dynamics of
medium-sized and large manufacturing firms. The choice of sector and firm size was driven
by the sample frame of the available data. Data from the Manufacturing Survey is available
for the period 1990–2015; it is an annual survey of all medium and large manufacturing
firms in Indonesia with 20 employees or more. The Manufacturing Survey provides rich panel
data about the firms, including age, ownership, number of workers by type, and production
inputs and outputs. This chapter also contains deeper work on the role of manufacturing
foreign direct investment (FDI) in the creation of middle-class jobs.

4.2 Employment Dynamism in Medium-Sized and Large

Manufacturing Plants
Indonesia’s flagship manufacturing industries have changed over the period 1990–
2015, with implications on the demand for and quality of jobs. The main shift in key
industries within the manufacturing sector between 1990 and 2015 in terms of (real) value
added has been the increasing share of automotive and other transport equipment from
4 percent to 13 percent and the increased share of computers, electronics, optics, and
electrical equipment from 2 percent to 8 percent. The food and beverages sector has seen
a more modest increase of 2 percentage points in its share of (real) value added. Meanwhile,
the tobacco share of value added has declined from 15 percent to 5 percent. Textiles and
textile products have seen a modest 1-percentage-point decline in share in (real) value added.

In terms of employment, changes in the shares of selected sectors are more

subdued, due to the increasing dominance of more capital-intensive or high-
technology industries. The share of automotive and other transport equipment increased
from 3 percent to 5 percent, and of computers, electronics, optics, and electrical equipment
from 2 percent to 5 percent. The less rapid changes in the size of shares of employment
compared to (real) value added and output in these two sectors indicate the more capital-
intensive/high-technology and/or more productive nature of the new industries. Meanwhile,
tobacco, and textiles and textile products have both seen a 1-percentage-point decline in
share of employment. The less rapid decline in the employment share in tobacco relative
to (real) value added and output suggest that this sector is more labor-intensive (and less
productive) than other industries with increasing shares in (real) value added and output.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 123

Figure 4.3 Young Plants (Aged 0–9) Dominated in the Figure 4.4 Until 2013, Plants Aged 10–19 Absorbed
1990s; Older Plants (Aged 10 and above) Dominated Most Labor, Then Plants Aged 30 Plus Took Over
after the Asian Financial Crisis Employment share by age group
Firm share by age group

Sources: Manufacturing Survey 1990–2015; World Bank staff calculations

Note: The Manufacturing Survey is the Survey of Medium-Sized and Large Manufacturing Firms (Survei Industri Besar/Sedang).

Between 1990 and 2015, employment was increasingly absorbed by very old and
very large plants. Young plants aged 0–9 dominate in terms of number of plants from
1992 to 1997. However, since 1998, plants aged 10 and above have taken over and accounted
for the highest number of plants until 2015 (Figure 4.3). In terms of net employment, young
plants cut many jobs during the Asian financial crisis of 1997–98 (see Figure 4.4). This was
partly due to firm exit, since small firms (below 50 employees) and firms aged younger
than 10 were more likely to exit during this period than older and larger firms.9 Firms with
1,000–2,999 employees and those that had been in operation for 30 years or more were
the least likely to exit during this crisis period, while ownership status was not significant in
determining how likely a firm was to exit. In short, the Asian financial crisis seemed to deter
entrance of new firms while killing off young and small firms, thus leaving only old and large
firms once the crisis had passed. Between 1990 and 2015, plants aged 10–19 absorbed the
most jobs, but, in 2013, plants aged 30 plus took over (Figure 4.4).

Indonesian manufacturing plants are aging . . . Plants were already aging from 1995,
even before the Asian financial crisis. The share of old plants (aged 20 and over) out of
the total number of plants increased significantly from 1995 (Figure 4.3). Similarly, the
employment share of old plants (as a share of total manufacturing employment) increased
even faster from about the same year (Figure 4.4). As a result, older plants had higher-than-
average numbers of employees per plant relative to younger plants.

9 Based on an estimated probit model on firm exit during 1997–98.

Pathways to Middle-Class Jobs in Indonesia June 2021

124 Enterprise-Level Dynamics of Job Creation

. . . because few plants entered during the Asian financial crisis and few young
plants survived it. During the Asian financial crisis and the global financial crisis of 2008,
the entrance of start-ups into the market slowed significantly. Hiring in new firms was
already low in 1993–96, even before the Asian financial crisis. Although there has been a
surge in young firms, and somewhat in hiring, since 2013, average levels of both remain
below the pre-1993 level (Figure 4.5 and Figure 4.6).

Figure 4.5 New Firm Growth Slowed during Economic Figure 4.6 Employment Growth Due To Firm Entry
Crises Stalled during the Entire Hollowing-Out Decade
New firms and existing firms, as share of total Employment share: new versus surviving

Sources: Manufacturing Survey 1990–2015; World Bank staff calculations

Note: The Manufacturing Survey is the Survey of Medium-Sized and Large Manufacturing Firms (Survei Industri Besar/Sedang).

Most Indonesian businesses are privately owned domestic firms, and these are
the firms that create most employment. More than 90 percent of Indonesian firms
are domestically owned, and nearly all are privately owned. In 1990, more than 90 percent
of workers were employed in domestic firms, while slightly more than 80 percent were in
privately held firms and more than 10 percent worked in the small share of publicly owned
firms. By 2015, employment in privately and publicly held firms had fallen; 70 percent of
workers were employed in privately held domestic firms and 3 percent in publicly owned
firms.

While domestic firms remain important for jobs, the presence of foreign-owned
plants increased in Indonesia; their share of total firms and employment expanded
significantly between 1990 and 2015.10 The employment share of foreign-owned plants
increased from 7 percent in 1990 to 26 percent in 2015 (Figure 4.8), while the share of
foreign-owned plants in the total number of plants increased from 2 percent in 1990 to 9
percent in 2015 (Figure 4.7). In 2011, about half of employment in foreign-owned plants was
in firms that exported 50 percent or more of their production. In absolute terms, in 2015,
foreign-owned plants employed more than seven times as many people as they did in 1990.

10 A plant is identified as “foreign owned” if a foreign shareholder owns the largest proportion of the investment.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 125

Figure 4.7 Most Firms Are Domestically Owned . . . Figure 4.8 Though an Increasing Share of
Firm share by ownership Workers Are Employed in Foreign-Owned Firms
Employment share by ownership

Sources: Manufacturing Survey 1990–2015; World Bank staff calculations

Note: In 2000 and 2001, immediately after the Asian financial crisis, the share of state-owned enterprises, in terms of number of plants and
employment, spiked because of the nationwide government restructuring of the collapsing private sector. The Manufacturing Survey is the Survey
of Medium-Sized and Large Manufacturing Firms (Survei Industri Besar/Sedang).

4.3 Job Creation and Destruction

Job creation was offset by job destruction, and this dynamism mostly came from
surviving plants. From 1990 to 2015, weak net job creation in the manufacturing sector,
especially between 2000 and 2010, was not due to a lack of job creation, but because
job destruction was equally high (Figure 4.9).11 Moreover, most jobs created or destroyed
came from surviving plants, rather than from new plants (job creation) or exiting plants
(job destruction).12 This high job churning mostly came from production workers vis-à-vis
nonproduction workers.

Figure 4.9 Most Employment Growth and Destruction Was in Surviving Firms
Job creation and job destruction, 1991–2015

Sources: Manufacturing Survey 1991–2015; World Bank staff calculations based on Sakernas 2017
Note: The year 2006 was a census year and the survey of that year contains more firms than in other years. It has been excluded for
consistency reasons. The Manufacturing Survey is the Survey of Medium-Sized and Large Manufacturing Firms (Survei Industri Besar/Sedang).

11 Plants that were in the survey, and exited and reentered are treated as new plants.
12 Also, see Ferro and Kuriakose 2018

Pathways to Middle-Class Jobs in Indonesia June 2021

126 Enterprise-Level Dynamics of Job Creation

Labor-intensive, low-technology sectors were responsible for the most (net) job
creation and job destruction in the manufacturing sector during 2013–14. In
2013 and 2014, garments produced from textiles, special paper, knitted and embroidered
products, edible palm oil, and smoked rubber were the top five subsectors in terms of net
job creation (Table 4.3).13 Meanwhile, the most job destruction in net terms occurred in
stripping, cleaning, and sorting coffee, and in coffee and tea processing.

Table 4.3 Top Net Job Creators and Destroyers

Harmonized
Sector Net job creation
system code
Top five net job creators
14111 Garments produced from textiles 60,326
17014 Special paper 40,216
13911 Knitted and embroidered products 21,645
10431 Edible palm oil 20,021
22122 Smoked rubber 17,593
Top five net job destroyers
10612 Stripping, cleaning and sorting coffee (40,404)
10761 Coffee and tea processing (15,310)
20112 Basic chemical and inorganic gas (14,239)
12091 Tobacco drying (5,261)
22123 Crumb Rubber (4,870)
Sources: Manufacturing Survey 2013–14; World Bank staff calculations
Note: The Manufacturing Survey is the Survey of Medium-Sized and Large Manufacturing Firms (Survei
Industri Besar/Sedang).

Net job creation hides the degree of job churning. Subsectors with the highest rates of
job churning (the sum of the gross values of job creation and job destruction) between 2013
and 2014 include some of the subsectors that have the highest net job creation, namely
garments produced from textiles, edible palm oil, and special paper (Table 4.4). Subsectors
with the highest job churning were generally labor-intensive industries (for example, textiles
and textile products, footwear, food and beverages), but were also those that typically have
short-term purchase orders from buyers, that have employees on short-term contracts (for
example, textiles and textile products), or low-technology or light industries.

13 The authors used 2013 and 2014 data because the “reasons for exit” variable was not available for 2015 at the time
of writing. Without the variable, the authors could not distinguish between firms exiting and firms downsizing.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 127

Table 4.4 Top Five Subsectors with the Highest Job Churning in 2013 and 2014
Number of jobs
Job churning Job churning (lower
Subsector Description
(upper bound)a bound)b
14111 Garments produced from textiles 126,000 123,821
10431 Edible palm oil 61,407 61,270
10612 Stripping, cleaning, and sorting coffee 44,656 44,656
17014 Special paper 40,230 40,230
10761 Coffee and tea processing 34,898 34,898
13121 Weaving (excluding sacks) 34,502 34,206
13133 Fabric printing 31,646 31,646
15201 Everyday footwear 29,274 29,192
13112 Yarn spinning 28,699 28,099
a Exit includes firms downsizing and becoming micro or small.
b Exit does not include firms that downsize to micro or small.

Sources: Manufacturing Survey 2013–14; World Bank staff calculations

Figure 4.10 Net Job Creation Was Strongest in Textiles, Apparel, and Footwear before the Asian
Financial Crisis, Then Joined by Food and Beverages
Net job creation by manufacturing subsector, 1991–2015

Sources: Manufacturing Survey 1991–2015; World Bank staff calculations.

Note: The year 2006 was a census year and the survey of that year contains more firms than in other years. It has been excluded for consistency
reasons. The Manufacturing Survey is the Survey of Medium-Sized and Large Manufacturing Firms (Survei Industri Besar/Sedang)

All plants experienced a broad-based weakening in net job creation between 2000
and 2010, except those that were foreign owned. Net job creation was generally strong
between 1991 and 1996 for almost all subsectors, but was generally very weak between
1997 and 2009 for almost all subsectors, followed by signs of strengthening from 2010. Pre–
Asian financial crisis, strong net job creation came from the textiles, apparel, and footwear
subsectors. Net job creation slumped both overall and in these three subsectors during
the Asian financial crisis and the hollowing-out decade, then rebounded in 2014–15. Net job
creation in the food and beverages subsector grew strongly in 2011–12 (Figure 4.10). Net job
creation was positive for almost all employment-size groups between 1991 and 1995, but

Pathways to Middle-Class Jobs in Indonesia June 2021

128 Enterprise-Level Dynamics of Job Creation

was generally very weak for all employment-size groups between 1996 and 2009, except in
1999, 2000, and 2004 when net job creation was positive for some employment-size groups
(large companies with 1,000+ employees). Net job creation resumed between 2010 and 2015,
although it was not as consistently strong as during the 1991–95 period (Figure 4.11). Net
job creation was generally strong for all firm-age groups in 1991–95 (Figure 4.12); it was very
weak for younger plants (aged under 9) between 1996 and 2009, but strengthened again
for most age groups between 2010 and 2015, except those aged 10–19. Only in 2013–15 did
net job creation revive among the youngest plants (aged 5 or younger).

Figure 4.11 Net Job Creation Greatest in Very Large Firms until 2013; Net Job Creation in Small Firms
Resumed in 2013
Net job creation by manufacturing subsector, 1991–2015

Sources: Manufacturing Survey 1991–2015; World Bank staff calculations

Note: The year 2006 was a census year and the survey of that year contains more firms than in other years. It has been excluded for consistency
reasons. The Manufacturing Survey is the Survey of Medium-Sized and Large Manufacturing Firms (Survei Industri Besar/Sedang)

Figure 4.12 Net Job Creation Was Very Weak for Younger Plants (Aged under 9) between 1996 and 2009
Net job creation by manufacturing subsector, 1991–2015

Sources: Manufacturing Survey 1991–2015; World Bank staff calculations.

Note: The year 2006 was a census year and the survey of that year contains more firms than in other years. It has been excluded for consistency
reasons. The Manufacturing Survey is the Survey of Medium-Sized and Large Manufacturing Firms (Survei Industri Besar/Sedang)

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 129

Net job creation was generally positive for foreign plants from 1991 to 2015.
Excluding 2000–02, when the government’s restructuring of the private sector took place,
net job creation was generally consistently positive, albeit small, for foreign plants except
in 2004, 2009, and 2013. Meanwhile, net job creation by private domestically owned plants
was strong in 1991–95, weak in 1996–2009 (except in 1999 and 2004) and started to pick up
again in 2010–14 (Figure 4.13).

Figure 4.13 Net Job Creation in Foreign Firms

Was Generally Positive across the Period
Net job creation by ownership, excluding the private sector
restructuring phase (2000–02)

Sources: Manufacturing Survey 1991–2015; World Bank

staff calculations
Note: The Manufacturing Survey is the Survey of Medium-
Sized and Large Manufacturing Firms (Survei Industri Besar/
Sedang).

Comparing employment in 2005 with the 2015 situation, privately owned domestic
and foreign-owned plants were net job creators, while publicly owned plants were
net job destroyers. Between 2005 and 2015, private domestically owned plants created
a net 583,640 jobs, and foreign plants created a net 564,555.14 Meanwhile, state-owned
plants destroyed a net 127,466 jobs. Net job creation by foreign-owned plants mostly came
from relatively old (aged 10 and above) plants, while net job creation by privately owned
plants was more mixed and young firms (aged 5 and under) also contributed significantly
(Figure 4.14).

Figure 4.14 Privately

Owned Firms Are Net
Job Creators, Publicly
Owned Are Net Job
Destroyers
Net job creation by age,
size, and ownership,
2005–15

Sources: Manufacturing Survey

2000 and 2015; World Bank staff
calculations
Note: The Manufacturing Survey
is the Survey of Medium-Sized
and Large Manufacturing Firms
(Survei Industri Besar/Sedang).

14 Net job creation was calculated based on the differences in the number of jobs in the age, size, and ownership categories in
2005 and 2015.

Pathways to Middle-Class Jobs in Indonesia June 2021

130 Enterprise-Level Dynamics of Job Creation

4.4 How is Demand for Jobs Created in Medium-Sized and Large

Manufacturing Firms?
High growth in productivity and export share were associated with higher employ-
ment growth from 1990 to 2015.15 Higher productivity expands the market and creates
demand for more labor, outweighing the substitution effect on labor of higher produc-
tivity. Although Figure 4.13 and Figure 4.14 suggest foreign-owned plants were strong net
job creators from 2005 to 2015, after controlling for employment size, age, productivity
growth (growth in real value added per worker), and growth of exported share of produc-
tion, foreign ownership was not the key factor that determined employment growth. In
fact, foreign ownership marginally reduced employment growth, after controlling for the
other variables. Instead, large and old plants with high productivity growth and high growth
of exported share of their production were found to have higher employment growth on
average (Figure 4.15).

Figure 4.15 Higher Productivity

Growth is Associated with Higher
Employment Growth
Employment growth determinants in
elasticity terms

Sources: Manufacturing Survey 1990–2015; World Bank

staff calculations.
Note: Elasticity with respect to private, domestic plants
with 20-49 employees, aged 0-5. Controlled for province,
subsector, and year. All coefficients above are significant
at 1 percent confidence interval, except “productivity
growth (public)” and “age 10-19.” The Manufacturing
Survey is the Survey of Medium-Sized and Large
Manufacturing Firms (Survei Industri Besar/Sedang).

The positive association between growth in labor productivity and employment

growth is due to the lower production costs (especially unit labor costs), greater
competitiveness, and market expansion. In essence, higher labor productivity expands
the market and the output effect outweighs the substitution effect. For example, Dutz (2018)
uses data from Brazil and finds that “a 10 percent decrease in the average manufacturing
price-cost margin, as would likely occur with greater competition, is associated with an
increase in labor productivity growth of over 3 percent per year—which corresponds to an
expansion in employment of about 1.4 million jobs per year.”16

15 The authors follow the definitions of employment growth at the plant level in Davis and Haltiwanger (1992) and Davis,
Haltiwanger, and Schuh (1996).
16 Greater competition can also lead to a higher average price-cost margin, as higher competition leads to a higher exit
probability from the less efficient firms. The average price-cost margin could increase due to technical efficiency.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 131

4.4.1 Privately owned firms grow more slowly compared to foreign-owned and
state-owned firms

Indonesia’s medium-sized and large manufacturing plants did not grow significant-
ly over their 40-year life cycles, relative to trends in other countries. Plants that
were held privately under Indonesian ownership grew more slowly on average over their life
cycles than plants held under foreign ownership and state-owned plants (Figure 4.16). Five
years after establishment, foreign-owned and state-owned plants added, on average, 10 em-
ployees per year, but a privately owned plant started adding only 5 employees per year and
only from its 9th year of establishment. For privately owned domestic plants, employment
doubled only by the 32nd year following establishment, with growth occurring very slowly
across the life cycle. In other words, privately owned domestic Indonesian plants on average
barely grew over their 40-year life cycle. The fact that privately owned domestic plants
grew even more slowly than state-owned plants, which are not subject to competitive pres-
sures, is a concerning sign; growth of firms shows that the market is vibrant and contains
competition that drives productivity, competitiveness, and growth. Even compared to the
life cycle of manufacturing plants (with 20 employees or more) in Vietnam, which is less
developed than Indonesia in terms of GDP per capita, Indonesian manufacturing plants look
stunted.17

Figure 4.16 Privately Owned Indonesian

Firms Grow Slowly Compared to Firms
under Public or Foreign Ownership
Employment life cycle, by ownership status; number
of employees (y-axis); age (x-axis)

Sources: Manufacturing Survey 1990–2015; World

Bank staff calculations
Note: The authors regressed number of employees
on age dummies to produce an employment life cycle.
The Manufacturing Survey is the Survey of Medium-
Sized and Large Manufacturing Firms (Survei
Industri Besar/Sedang).

The age-growth profiles suggest that the first four years of a plant’s life cycle
are most critical for growth spurts in real value added and employment. The age-
growth profiles suggest that the growth of real value added and employment among very

17 Cunningham, Ly, and Chea 2018

Pathways to Middle-Class Jobs in Indonesia June 2021

132 Enterprise-Level Dynamics of Job Creation

young plants play an important role in the growth of most manufacturing subsectors.18
Findings from one study show that, regardless of the definition used, high-growth firms
tend to be younger than average.19 Further, a study on high-growth firms in Indonesia20
shows that high-growth firms tend to have spurts of high growth in the early years. This
may be due to several factors. First, younger firms face “up or out” dynamics—they either
need to grow quickly and stabilize, or they will fail to grow and exit. Second, younger firms
tend to invest in the most advanced technology, while older firms continue to use older
capital. Third, younger firms revise their beliefs more frequently than older firms regarding
the demand for their products, as they are uncertain of their product appeal.21 This enables
surviving young firms to grow more rapidly than older firms.22 Overall, the growth rate for
value added is flat after four years of operation, while employment growth continues to
slow for at least the next 20 years.

4.4.2 More productive firms are less likely to close down or to become small and
micro
Between 2005 and 2015, out of the total number of plants that existed in 2005, 46
percent downsized, 24 percent grew in size, and 27 percent exited.23 Younger plants
are more likely to exit and are more likely to reduce in size to below 20 employees (Table
4.5). Among plants aged 0–5 in 2005, 32 percent had exited by 2015, while 19 percent had
reduced in size to micro or small. Among plants aged 6–10 in 2005, 28 percent had exited by
2015, while 17 percent had reduced in size to micro or small. Among plants aged above 10 in
2005, 25 percent had exited by 2015, while 15 percent had reduced in size to micro or small.
Taking the whole sample together, 46 percent of firms downsized in the ten-year period and
27 percent exited; only 24 percent grew and 3 percent maintained their size.

Table 4.5 Changes in Size of Firms between 2005 and 2015

Number of firms by size and years after establishment

Firm size in 2015

Years after
Firm size in 2005  Medium Large Small or micro (<20) Exited
establishment
0–5 Medium 1,091 225 725 1,041
  Large 159 465 27 274
6–10 Medium 1,091 216 646 780
  Large 226 667 17 338
>10 Medium 3,861 448 1,841 2,164
  Large 815 2,552 75 984
Sources: Manufacturing Survey 2005 and 2015; World Bank staff calculations
Note: The Manufacturing Survey is the Survey of Medium-Sized and Large Manufacturing Firms (Survei Industri Besar/Sedang).

18 Alatas and Wihardja. Forthcoming

19 Grover Goswami, Medvedev, and Olafsen 2019
20 Ferro and Kuriakose 2018
21 A firm that experiences higher demand than expected will revise its forecasts upward and expand production, while a firm
that experiences lower demand than expected will revised its forecasts downward and reduce production or exit the market.
22 Grover Goswami, Medvedev, and Olafsen 2019
23 The Manufacturing Survey data shows whether a firm exited the market or whether it exited the survey. The data shows
why a firm exited the survey, including whether it exited because it became smaller (less than 20 employees).

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 133

Plants with higher productivity, foreign ownership, and that are larger and older
are less likely to exit. Controlling for other variables, the probability of a plant exiting
decreases with size, age, and productivity (Figure 4.17). More productive plants have higher
survival rates, while less productive plants are more likely to exit. Similar to many other
countries, smaller (20–49 employees) and younger (0–5 years old) plants are most likely to
exit. Given that (surviving) young plants (0–5 years old) have the highest growth spurts in
value added and employment, the high likelihood that young plants will exit has negative
implications for job creation.

More productive, state-owned, and larger firms are less likely to become small and
micro (and are less likely to close down), and the probability of this occurring is
not significantly affected by age. Similar to the probability of such a plant exiting, the
probability of the plant downsizing to micro or small decreases as size and productivity
increase, but is not affected by age (Figure 4.18).24 On the contrary, while the probability of
exiting is higher for state-owned plants relative to privately owned plants, the probability of
becoming micro or small is lower for state-owned plants relative to privately owned plants.
Also, while the probability of exit is lower for older plants, the probability of becoming micro
or small is not significantly affected by age.

Figure 4.17 Large, Old, Foreign, and More Figure 4.18 Productive, State-Owned and Larger
Productive Plants Do Not Exit Firms Are Less Likely to Significantly Downsize
Probit regression coefficient estimates Probit regression coefficient estimates

Sources: Manufacturing Survey 1990–2015; World Bank staff Sources: Manufacturing Survey 1990–2015; World Bank staff
calculations calculations
Note: The probability of exit (other than getting smaller), with respect Note: The probability of becoming micro or small (with respect to
to private, domestic plants with 20–49 employees, 0–5 years from the private, domestic plants with 20–49 employees, 0–5 years from the
date of establishment. Controlled for province, subsector, and year; all date of establishment). Controlled for province, subsector, and year;
coefficients are significant at 10 percent confidence interval, except all coefficients are significant at 10 percent confidence interval,
foreign and 6–9 years after establishment. The Manufacturing Survey except foreign, 6–9 years after establishment, 20–29 years after
is the Survey of Medium-Sized and Large Manufacturing Firms (Survei establishment, and 30+ years after establishment. The Manufacturing
Industri Besar/Sedang) Survey is the Survey of Medium-Sized and Large Manufacturing Firms
(Survei Industri Besar/Sedang)
24 Using a probit model, the authors also calculated the probability of downsizing to micro or small from medium or large.

Pathways to Middle-Class Jobs in Indonesia June 2021

134 Enterprise-Level Dynamics of Job Creation

4.5 What Does Manufacturing FDI Mean for Middle-Class Jobs?

Attracting FDI in manufacturing can provide opportunities for growth in middle-
class employment. Foreign-owned manufacturing firms can be important for stimulating
employment growth and they also pay wage premiums, thereby contributing to middle-class
jobs. However, Indonesia is not as welcoming as it could be to foreign investors. According to
the OECD FDI Restrictiveness Index 2017, Indonesia is the third-most restrictive out of 68 rich
and middle-income countries. Restrictions have limited foreign investment and hampered
job creation, including in the labor-intensive manufacturing sector. Indonesia’s FDI inflows as
a share of GDP have averaged less than 2 percent over the past decade, among the lowest
in the region (Figure 4.19). The shares of value added, output, and employment contributed
by multinationals25 were subdued between 2007 and 2015 (see Figure 4.20).

Figure 4.19 Indonesia Has Limited FDI Inflows Figure 4.20 Multinational Activity in Indonesia Is Subdued
FDI inflows in GDP in percent, Indonesia versus comparators, Multinational activity shares (out of 1): firm average, 2007–15
2009–18

Sources: Steenbergen and Hebous 2020; Source: Manufacturing Survey 2007–15

World Development Indicators Note: The Manufacturing Survey is the Survey of Medium-Sized and Large
Manufacturing Firms (Survei Industri Besar/Sedang).

Box 4.2 Cross-Country Studies on Distributional Impact of Foreign Direct

Investment
There is considerable evidence from a range of countries that foreign direct investment (FDI) is
beneficial across the entire sector in which it is deployed, raising overall employment and average
wages. Much of the existing evidence points to FDI raising wages, a result driven primarily by new technology
and increased labor productivity.a In many cases, the literature also finds FDI has a positive effect on aggregate
employment. For example, FDI was found to lift the employment rate in China,b the Czech Republic,c and Uruguay.d
In addition, FDI also had a strong positive effect on employment in Mexico’s manufacturing sector, with stronger
effects in export-oriented industries.e
a Hale and Xu 2016
b Karlsson et al. 2009
c Dinga and Münich 2010
d Peluffo 2015
e Waldkirch, Nunnenkamp, and Alatorre Bremont 2009
25 In this section, a firm is classified as a multinational if its foreign capital share is at least 10 percent.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 135

Box 4.2 continued

In Indonesia, FDI has also raised manufacturing employment.f Sjöholm, Lipsey, and Sun (2010) find that
foreign establishments played a role in Indonesia’s growing manufacturing employment between 1975 and
2005. While plants with some foreign ownership made up less than 10 percent of manufacturing employment
in 1975, they employed around 20 percent in 2005. The authors also find that employment in foreign-owned
plants grew about 5 percent faster than in domestic-owned firms, while plants that were acquired by foreigners
grew about 10 percent faster. Considering that foreign plants are on average considerably larger than domestic
plants, the difference in the number of jobs created is large.g

The increased employment associated with FDI inflows has also helped pull workers out of low-
productivity sectors and into higher-productivity sectors. While the number of studies is limited, there
is a growing body of evidence that finds important across-sector employment effects from FDI. Escobar and
Mühlen (2018) cover the period 2006–16 and find that FDI flows to Mexico led to considerable reallocation of
labor between sectors. Using Mexican states as the unit of observation, they find that direct investments in the
industrial sector helped draw workers out of (low-productivity) agriculture and increase the size of the share of
employment in manufacturing. Interestingly, as an illustration of wider economic spillovers, they also found a
positive effect on the employment share in relatively productive service sectors. As a result, labor reallocation
occurred for both low- and medium-skilled workers. The effect of FDI on Indonesia’s intersectoral employment
is discussed in Chapter 3 of this report.

However, FDI may also increase the wage gap between skilled and unskilled workers. Foreign investment
often introduces new technologies that raise the demand for higher-skilled workers. There is considerable
empirical evidence confirming that FDI contributes to rising wage inequality in host countries. In developing
countries, wage inequality increases as stocks of inward FDI increase.h A rise in Japanese FDI in developing
countries is associated with an increase in nonproduction wages (for more-skilled workers) relative to production
wages (for less-skilled workers).i Similar effects of foreign investments have been found for firms in Indonesia
and Mexico.j However, technological change is not necessarily biased in favor of skilled workers; the skill level
required depends largely on the type of FDI attracted.k An FDI injection into some types of low-skill sectors
(such as textiles or food processing) could disproportionally benefit unskilled workers.l In Ethiopia, Ghana, and
Mozambique, FDI in labor-intensive manufacturing and infrastructure was associated with declining inequality.m
f For additional studies on FDI in Indonesia see, for example, Blomström and Sjöholm (1999); Lipsey and Sjöholm (2004); Takii (2005); Blalock and Gertler
(2008); and Arnold and Javorcik (2009).
g Lipsey and Sjöholm 2004
h Figini and Görg 2011
i Head and Ries 2002
j Lee and Wie 2015; Feenstra and Hanson 1997
k Luo 2017
l Leamer 1998; Cruz et al. 2018
m Cornia 2016

Pathways to Middle-Class Jobs in Indonesia June 2021

136 Enterprise-Level Dynamics of Job Creation

While both domestic and foreign manufacturing firms are a source of middle-class
jobs, average wages are considerably higher in foreign-owned firms. Using the
income bracket classification in Aspiring Indonesia: Expanding the Middle Class,26 it is evident
that the manufacturing sector could provide an important contribution for higher-paying
employment. At domestic firms, an average of 42 percent of all jobs have wages at a level
commensurate with aspiring middle class, while the same is true for 40 percent of all jobs at
foreign-owned firms (Figure 4.21). However, wages are higher on average in foreign-owned
firms: Rp3.2 million per month on average, as compared to Rp1.8 million (Figure 4.22). A larger
number of foreign-owned firms pay wages classified as “middle class” than do domestic firms
(30 percent versus 16 percent).

Figure 4.21 Foreign-Owned Firms Pay Better Wages Figure 4.22 Wage Dynamics for Production
Than Domestically Owned Firms Workers—Domestic Firms versus Foreign Owned
Share of wages in each class category, by firm ownership (Average 2010–15)
Average monthly wages, by firm type, in Rp, millions

Sources: Steenbergen and Hebous 2020; Manufacturing Survey Sources: Steenbergen and Hebous 2020; Manufacturing Survey
2010–15 2010–15
Note: Where percentages do not total 100 percent, this is due to Note: The Manufacturing Survey is the Survey of Medium-Sized and
rounding. The Manufacturing Survey is the Survey of Medium-Sized Large Manufacturing Firms (Survei Industri Besar/Sedang).
and Large Manufacturing Firms (Survei Industri Besar/Sedang).

In Indonesia’s manufacturing sector, there is a positive relationship between the

activity of foreign-owned companies and wages across districts. Holding all other
factors constant, when a sector in a given region sees a 10-percentage-point increase in
output share going to foreign-owned companies, the average wages in that sector of that
region are likely to increase by 2.25 percent. The magnitude of this effect in Indonesia is in
line with that of other countries.27

4.5.1 Foreign firms provide greater benefits to higher-skilled workers

Foreign direct investment provides greater benefits to higher-skilled workers
in Indonesia’s manufacturing sector. The positive effect of the activity of foreign

26 World Bank 2019a

27 See Steenbergen and Trang (2020) for similar estimates for Turkey, Vietnam, and Ethiopia.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 137

multinationals on wages increases with an individual’s educational outcome.28 Interacting

the output share of multinationals with educational level shows a negative but statistically
insignificant effect for those with no education, and positive effects on wages for all other
education levels. Relatedly, the effect is larger for white-collar jobs than it is for blue-collar
jobs.

Breaking down the firms sampled in the Manufacturing Survey into low-skill and
high-skill manufacturing,29 the benefits of FDI are more widespread for the low-
skill manufacturing sectors than the high-skill manufacturing sectors. Increased
multinational activity in low-skill manufacturing is associated with benefits for workers with
primary and secondary education but the benefits are especially pronounced for those with
tertiary education. For high-skill manufacturing, the benefits are concentrated in workers
with secondary education only (and are statistically insignificant for other groups). A similar
finding holds for occupation type, where FDI in low-skill sectors benefits both blue- and white-
collar workers, but the benefits are restricted to the latter for high-skill sectors. This result
could be explained by the difference in overall skill requirement for these types of sectors.
Another reason may be that the results are more widespread for low-skill manufacturing,
as they make up most of Indonesia’s manufacturing sector, and multinationals are also
concentrated in low-skill manufacturing (Table 4.6 and Table 4.7).

Table 4.6 Within-Sector Effect of FDI By Education Level

Coefficient regression estimates

  (1) (2) (3)

Outcome variable Wage (log)
Low-skill High-skill
Manufacturing sector All
manufacturing manufacturing
Output share of foreign-owned multinationals
–0.0180 −0.00561 −0.377
(lagged one period) * No education
  (0.171) (0.181) (0.302)
Output share of foreign-owned multinationals
0.256*** 0.330*** −0.0167
(lagged one period) * Primary education
  (0.0771) (0.0843) (0.0424)
Output share of foreign-owned multinationals
0.229*** 0.310*** 0.107**
(lagged one period) * Secondary education
  (0.0705) (0.0750) (0.0420)
Output share of foreign-owned multinationals
0.359*** 0.607*** 0.0762
(lagged one period) * Tertiary education
  (0.120) (0.158) (0.0920)
Controls Education, gender,
  age, sectoral tariffs
Fixed effects Region-year, sector
Observations 154,292 138,153 15,928

Source: Steenbergen and Hebous 2020

Note: Second stage IV results

28 Steenbergen and Hebous 2020

29 See Annex 4.2 for a list of low- and high-skill manufacturing sectors.

Pathways to Middle-Class Jobs in Indonesia June 2021

138 Enterprise-Level Dynamics of Job Creation

Table 4.7 Within-   (1) (2) (3)

Sector Effect of FDI Outcome variable Wage (log)
by Occupation Low-skill High-skill
Manufacturing sector All
Coefficient regression manufacturing manufacturing
estimates Output share of foreign-owned multinationals 0.0656*** 0.0820*** −0.00314
(lagged one period) * Blue-collar
  (0.0192) (0.0213) (0.00917)
Output share of foreign-owned multinationals 0.0823*** 0.0985*** 0.0359***
(lagged one period) * White-collar
  (0.0216) (0.0288) (0.0110)
Controls Education, gender,
Source: Steenbergen and   age, sectoral tariffs
Hebous 2020 Fixed effects Region-year, sector
Note: Second stage IV
results
Observations 209,086 191,425 17,471

4.5.2 Foreign firms provide greater benefits to wage employees

Most of the wage benefits of multinational activity in Indonesia’s manufacturing
sector accrue to employees. Table 4.8 shows results that interact the share of output
attributed to multinationals with different job types. For both casual workers and the
self-employed, multinational activity has a negative effect on average wages. There are
several possible explanations, but the most likely is that FDI increases the demand for formal
employees. As the most skilled workers are absorbed by multinationals, average wages for
casual workers and the self-employed may go down. An alternative, less likely scenario is that
the self-employed and small, informal firms, which employ casual labor, face increased price
competition for their products from manufacturing multinationals and thus see downward
pressure on their earnings. Another possibility is that wages go up in foreign firms because
they have to follow the minimum wage policy, while domestic firms tend to pay market
rates. Foreign (minimum-wage paying) multinational penetration is sufficiently small that it
does not distort local labor markets. Interestingly, multinational activity seems to benefit
the self-employed in high-skill manufacturing; this could indicate increased demand for
inputs for high-skill manufacturing.
Table 4.8 Within-   (1) (2) (3)
Sector Effect of Outcome Variable Wage (log)
FDI by Employment Low-skill High-skill
Manufacturing Sector All
Status manufacturing manufacturing

Coefficient regression Output share of foreign-owned

multinationals (lagged one period) * −0.246* −0.308*** 0.220**
estimates Self-employed
  (0.126) (0.111) (0.0864)
Output share of foreign-owned
multinationals (lagged one period) 0.379*** 0.523*** 0.0753*
* Employee
  (0.0831) (0.0877) (0.0433)
Output share of foreign-owned
multinationals (lagged one period) * −1.048*** −0.956*** −1.339***
Source: Steenbergen and Casual worker
Hebous 2020   (0.191) (0.192) (0.330)
Note: Second stage IV
results
Controls Education, gender,
  age, sectoral tariffs
June 2021 Fixed effects
Pathways to Middle-Class Jobs in Indonesia Region-year, sector
Observations 154,292 138,153 15,928
 Enterprise-Level Dynamics of Job Creation 139

4.5.3 Foreign firms provide greater benefits to female workers

Female workers in low-skill manufacturing are among the biggest beneficiaries
of increased multinational activity in Indonesian manufacturing. Table 4.9 shows
the effect of the interaction between gender and the output share of foreign-owned
multinationals on wages. The overall effect is positive but insignificant for males. For females,
the overall effect is positive and highly significant, especially in the low-skill manufacturing
subset. This may be explained by the gender bias in the types of industries in which foreign-
owned multinationals tend to operate (for example, food products, textiles)—such industries
employ a large number of females. Alternatively, the results may be explained by a rise in the
total demand for labor, which creates job opportunities for otherwise informally working
females.30 It might also point to wage discrimination against females by domestic firms or a
preference of (higher-paying) multinationals for hiring women.31

Table 4.9 Within-Sector   (1) (2) (3)

Effect of FDI by Gender Outcome variable Wage (log)
Coefficient regression Low-skill High-skill
Manufacturing sector All
estimates manufacturing manufacturing
Output share of foreign-owned
multinationals (lagged one period) * 0.622*** 0.685*** 0.161**
Female
  (0.0956) (0.115) (0.0736)
Output share of foreign-owned
multinationals (lagged one period) * 0.00295 −0.00802 0.0120
Male
  (0.0750) (0.0955) (0.0350)
Source: Steenbergen and
Hebous 2020 Controls Education, gender,
Note: Second stage IV   age, sectoral tariffs
results
Fixed effects Region-year, sector
Observations 154,292 138,153 15,928

In summary, manufacturing multinationals have important, but heterogenous

effects on wages within a sector. An injection of FDI increases average wages and
benefits most workers who have at least a primary education. However, some workers
benefit more than others. There is a skills premium, so workers with higher levels of
education benefit more, and white-collar workers see greater benefits than blue-collar
workers. Interestingly, women appear to benefit more than men (likely as a result of the
type of sectors multinationals engage in). Overall across the entire manufacturing sector,
the effects are more widespread for low-skill than high-skill manufacturing sectors.

30 See, for example, Ver Beek (2001)

31 Tang and Zhang 2017

Pathways to Middle-Class Jobs in Indonesia June 2021

140 Enterprise-Level Dynamics of Job Creation

4.6 Improving Productivity

Labor productivity in large plants (100+ employees) and foreign-owned plants is
skewed toward higher levels of productivity. Large plants (100+ employees) are more
productive on average than the medium sized (50–99 employees) or small (20–49 employees).
Given economies of scale, it is not surprising to find that larger plants are more productive.
This is evident from the labor productivity distribution function graph, which is skewed
more toward higher levels of productivity (Figure 4.23). Labor productivity differences are
also pronounced by ownership type, with foreign plants being more productive on average.
The labor productivity distribution function graph is most skewed toward higher levels of
productivity for foreign-owned plants (Figure 4.23). The labor productivity distribution
function graph for state-owned plants has a fat right-hand tail (almost on par with foreign
plants) that indicates relatively high shares of state-owned plants in the higher levels of
productivity. The labor productivity distribution function graph for privately owned
domestic plants is more concerning because these plants are supposed to be subject to
higher competition (hence, plants with lower productivity should have been competed out).

Figure 4.23 Larger and Foreign-Owned Firms Are More Productive

Distribution function of labor productivity by ownership and size

Sources: Manufacturing Survey 2015; World Bank staff calculations

Note: The Manufacturing Survey is the Survey of Medium-Sized and Large Manufacturing Firms (Survei Industri Besar/Sedang).

Higher growth in exports is associated with higher growth in labor productivity.32

Productivity growth is critical for generating better jobs, but it also improves living standards
by lowering prices and improving the quality of goods people consume. A multivariate
analysis was used to see factors that are associated with productivity growth. Being small
(<50 employees) and old (>5 years of operation), as well as having higher growth in exports,
is associated with higher growth in labor productivity. Foreign-owned firms have statistically
similar productivity growth rates to domestically owned firms (Figure 4.24).

32 The authors use definitions of productivity growth at the plant level found in Davis and Haltiwanger (1992) and Davis,
Haltiwanger, and Schuh (1996).

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 141

(Real) Labor productivity growth was mostly driven by the “within” effect
between 2000 and 2015, although the entry effect is still indispensable (Figure
4.25). Overall, growth in labor productivity was low, less than 1 percent (0.54 percent),
during 2000–15. During the Asian financial crisis, the contribution of the within effect was
relatively high, probably because firms were pushed hard to increase efficiency by cutting
jobs while maintaining production. Determinants of labor productivity growth also vary
by industry.33 Because of the exclusion of micro and small enterprises, decomposition of
aggregate productivity gains based on truncated data (based only on medium-sized and
large enterprises) tend to overestimate the between effect; the gains (losses) from firm
entry tend to be overestimated (underestimated), and the gains (losses) from firm exit tend
to be underestimated (overestimated).34

Table 4.24 Smaller, Older, and Exporting Firms

Have Higher Productivity Growth Rates
Productivity growth determinants in elasticity terms

Sources: Manufacturing Survey 1990–2015; World Bank staff

calculations.
Note: Elasticity with respect to private, domestic plants
with 20–49 employees, 0–5 years following establishment.
Controlled for province, subsector, and year. All coefficients are
significant at 10 percent confidence interval, except foreign
and 6–9 years following establishment. The Manufacturing
Survey is the Survey of Medium-Sized and Large Manufacturing
Firms (Survei Industri Besar/Sedang).

Figure 4.25 The Decomposition of Labor Productivity Growth, Average Annual, 2005–15
Percentage point contribution to labor productivity growth

Sources: Manufacturing Survey 2005–2015; World Bank staff calculations

Note: The Manufacturing Survey is the Survey of Medium-Sized and Large Manufacturing Firms (Survei Industri Besar/Sedang).

33 Alatas and Wihardja. Forthcoming

34 Li and Rama 2015

Pathways to Middle-Class Jobs in Indonesia June 2021

142 Enterprise-Level Dynamics of Job Creation

Evidence suggests declining allocative efficiency in terms of labor productivity,

but increasing allocative efficiency in terms of total factor productivity and
unchanged allocative efficiency in terms of capital productivity. Some studies imply
that over time firms that are more productive expand at the expense of those that are less
productive.35 As a result, a positive correlation between productivity and firm size can be
expected, indicating allocative efficiency.36 Table 4.10 shows that the correlation between
labor productivity (real value added per worker) and plant size is positive but very small in
magnitude. Moreover, it is significant for early years in the survey, that is 1990–2005, but not
for later years. This may indicate that allocative efficiency decreased in more recent years
in terms of labor productivity. However, Table 4.10 also shows that the correlation between
total factor productivity and plant size is generally increasing and remains significant over
the years, indicating increasing allocative efficiency in terms of total factor productivity.
When broken down into different subsectors, the results vary, with the strongest positive
correlation found in food and beverages. Moreover, the correlation between capital
productivity and plant size declined relative to that of the year 1990 but remained stable
from 2000 to 2015.

Table 4.10 Correlation between Labor Productivity and Plant Size

1990 2000 2005 2010 2015
Correlation between labor productivity and plant
0.0357*** 0.0194*** 0.0337*** 0.0052 0.0035
size
Correlation between total factor productivity
and plant size 
All sectors 0.0349*** 0.0861*** 0.0563*** 0.0566*** 0.0737***
Textiles and textile products −0.1768*** -0.0528*** 0.0724*** 0.1375*** 0.1521***
Food and beverages 0.3399*** 0.3977*** 0.4028*** 0.3685*** 0.3687***
Computers, electronics, optics, and electricity 0.0937 0.206*** 0.1745*** 0.1529** 0.0839
Automotive and other transport equipment 0.257*** 0.1568** 0.154** 0.2774*** 0.0999*
Correlation between capital productivity and
0.0671*** 0.0245*** 0.0214** 0.0298*** 0.0207**
plant size

Sources: Manufacturing Survey 1990, 2000, 2005, 2010, and 2015; World Bank staff calculations
Note: The Manufacturing Survey is the Survey of Medium-Sized and Large Manufacturing Firms (Survei Industri Besar/Sedang).

35 Li and Rama 2015

36 Bartelsman, Haltiwanger, and Scarpetta 2009a and 2009b

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 143

4.7 Improving Wages

Plants that are more productive pay higher real wages. Across plants, those that
are 10 percent more productive (real value added per worker) on average pay real wages
that are 2.14 percent higher (Table 4.11). In Vietnam, in formal waged jobs in industry and
services, a 10 percent increase in productivity is associated with a 1.3 percent increase in
wages.37 In Indonesia, a higher share of female workers is associated with marginally lower
real wages. The positive effect on real wages of being a foreign plant diminishes once
productivity is controlled for. Workers in larger plants with up to 999 employees earn higher
wages compared to those in plants with 20–49 employees, controlling for everything else.
Age of plant does not seem to be a significant factor in determining wages.

Over time, plants that increase their productivity (real value added per worker) also
increase real wages: this applies particularly to private, domestic plants; the effect
is lower in foreign plants. A 10 percent increase in productivity growth is associated with
a 2.1 percent increase in real wage growth in privately owned, domestic plants, a 1.5 percent
increase in real wage growth in state-owned plants, and a 1.3 percent increase in real wage
growth in foreign-owned plants (Table 4.11). Similar results emerge from Vietnam, where
foreign firms share less of their productivity gains with workers, as compared to privately
owned, domestic firms and state-owned firms.38 The gap is even larger in Vietnam than
in Indonesia, where a 10 percent increase in productivity is associated with a 1.3 percent
increase in domestic firm wages, a 3.3 percent increase in state-owned firm wages, and a 0.6
percent increase in foreign firm wages.

Table 4.11 A 10 Percent Increase in Productivity Results in a 2.14 Percent Increase

in Wages; Private Domestic Firms Share Gains More with Workers
Regression coefficient estimates in elasticity terms

Log (real) wages (across firms) Log (real) wages (within firms)
Coefficient Private Public Foreign
Log (real value added per worker) 0.214 *** 0.21*** 0.15*** 0.13***
Public 0.058 *** n.a. n.a. n.a.
Foreign −0.0005 n.a. n.a. n.a.
Share of female workers −0.002*** n.a. n.a. n.a.
Control: province, year, subsector; ***, **, *=significant at 1 percent, 5 percent, and 10 percent confidence intervals

Sources: Manufacturing Survey 1990–2015; World Bank staff calculations

Note: The Manufacturing Survey is the Survey of Medium-Sized and Large Manufacturing Firms (Survei Industri Besar/Sedang).

37 Aterido and Hallward-Driemeier 2017

38 Aterido and Hallward-Driemeier 2017

Pathways to Middle-Class Jobs in Indonesia June 2021

144 Enterprise-Level Dynamics of Job Creation

Labor productivity and real wages tend to be higher in subsectors that intensively
use high technology, but employment is lower (Figure 4.26 and Figure 4.27).39
Mapping the share of employment against the conditional probability of productivity and
wages (with respect to apparel) shows a negative relationship: low-productivity and low-
wage subsectors, such as apparel, textiles, and food and beverages, have high shares of
employment, while high-productivity and high-paying wage subsectors, such as motor
vehicles, chemicals, and other transport equipment, have low shares of employment.

Figure 4.26 Employment Declines with Figure 4.27 Employment Declines with Wage
Productivity Increases Increases
Share of employment versus productivity (relative to the Share of employment versus real wage (relative to the apparel
apparel subsector) subsector)

Sources: Manufacturing Survey 2015; World Bank staff calculations Sources: Manufacturing Survey 2015; World Bank staff calculations
Note: The Manufacturing Survey is the Survey of Medium-Sized and Note: The Manufacturing Survey is the Survey of Medium-Sized and
Large Manufacturing Firms (Survei Industri Besar/Sedang). Large Manufacturing Firms (Survei Industri Besar/Sedang).

39 Using the definition of the United Nations Industrial Development Organization for technology grouping:
https://stat.unido.org/content/learning-center/classification-of-manufacturing-sectors-by-technological-intensity-%28isic-
revision-4%29;jsessionid=B99E902A3918AB9F3DF9859923DFC4F4.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 145

Monitoring the wage-productivity gap is critical. Figure 4.28 The Wage-Productivity Gap Widened
A breakdown of the social contract in the workplace, for Significantly in the United States in 1980–2014
example in the United States starting in the early 1980s, Wage-productivity gap in the United States, 1945–2014,

resulted in a divergence between real average wage indexed in 1945

growth and real labor productivity growth. When wages

and productivity moved upward together from 1945 to
about 1980, “the American middle class expanded and
ushered in a sustained era of broadly shared prosperity.”40
“Beginning in the late 1970s, the postwar social contract
came under relentless pressure from the confluence of four
major economic developments: globalization, deregulation,
financialization, and automation.”41 In response to this
pressure, there were four dominant responses by
employers: the containment of wages, salaries, and benefits;
domestic outsourcing; a combination of global outsourcing
and offshoring; and the automation of work processes.42
In the United States, from the early 1980s until at least
2014, wage growth has been lagging behind productivity
Sources: Khochan 2019, “Breakdown of the Old Social Contract
growth due to the breakdown of the social contract and and Starting Assumptions for a New Social Contract”, original
source of data: Economic Policy Institute.
worker-management relations (Figure 4.28). Divergence of
productivity growth from wage growth is not sustainable
from economic, social, or political points of view. It is Figure 4.29 Real Wage and Real Labor Productivity
therefore critical to monitor the wage-productivity gap. Trends Track Each Other Quite Closely
Wage and productivity per capita (real terms, mean, in Rp ’000)
Unlike the United States, within Indonesian medium-
sized and large manufacturing industries, real wage
and real labor productivity trends tracked each other
quite closely from 1990 to 2015 (Figure 4.29). The
correlation between the three-year moving average of real
wage growth and real labor productivity growth was 0.68
(significant at the 1-percent confidence interval). Unlike
the United States, Indonesia was not so much impacted by
globalization, financialization, deregulation, or automation
in the period 1990–2015. Indonesia was not particularly
integrated to the regional or global value chain.43 Indonesia
also did not experience the financialization and deregulation
of the market to the same degree as the United States, and
automation came at a much slower rate. Post Asian financial
crisis, labor unions also became stronger in Indonesia. Sources: Manufacturing Survey 1990–2015; World Bank staff
calculations
Note: Keeping only the 2nd–99th percentiles. The
Manufacturing Survey is the Survey of Medium-Sized and
40 Kochan and Dyer 2019b Large Manufacturing Firms (Survei Industri Besar/Sedang)
41 Kochan and Dyer 2019a
42 Kochan and Dyer 2019a
43 Shepherd and Soejachmoen 2017

Pathways to Middle-Class Jobs in Indonesia June 2021

146 Enterprise-Level Dynamics of Job Creation

High-technology, capital-intensive sectors pay higher wages than do labor-intensive

sectors. In only one industry subsector did medium-sized and large manufacturing
companies provide average monthly wages (excluding overtime and other benefits) that
were commensurate with the expectations of the middle class in 2015 (that is, more than
Rp3.2 million per month) (Figure 4.30). Large plants in high-technology/capital-intensive
subsectors, namely other transport equipment, computers, electronics, optical, base metals,
electrical, motor vehicles, and chemicals pay the largest average nominal wages. Labor-
intensive sectors that absorb most employment in industry—including food and beverages,
tobacco, textiles and textile products, and footwear—pay wages that are much lower than
the middle-class population’s expenditure cutoff. Large firms pay better than medium firms
for each subsector.

Figure 4.30 Most Sectors Pay Below Middle-Class Wages

Nominal average monthly wages in rupiah in medium-sized and large manufacturing firms, 2015
(vertical line is the cutoff for a middle-class wage in 2015)

Sources: Manufacturing Survey 2015; World Bank staff calculations

Note: The Manufacturing Survey is the Survey of Medium-Sized and Large Manufacturing Firms (Survei Industri Besar/Sedang).

Women are more likely to work in low-wage and low-productivity sectors. The
wage premium of women working in manufacturing is about 20 to 25 percent lower than
that of men on average.44 Among medium-sized and large manufacturers, although almost
equal numbers of men and women are employed, women are more likely to be hired in low-
productivity and low-wage sectors (Figure 4.31 and Figure 4.32).

44 Based on the authors’ recent estimate using Sakernas and an augmented Mincerian regression. See also Chapter 5 of this
report.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 147

Figure 4.31 Women Work in Lower-Productivity Figure 4.32 Women Work in Lower-Wage Sectors
Sectors Relative to Men Relative to Men
Share of female workers versus productivity (relative to the Share of female workers versus real wages (relative to the apparel
apparel subsector) subsector)

Sources: Manufacturing Survey 2015; World Bank staff calculations Sources: Manufacturing Survey 2015; World Bank staff calculations
Note: The Manufacturing Survey is the Survey of Medium-Sized and Note: The Manufacturing Survey is the Survey of Medium-Sized and
Large Manufacturing Firms (Survei Industri Besar/Sedang). Large Manufacturing Firms (Survei Industri Besar/Sedang).

4.8 Do Household Enterprises Provide a Path to Middle-Class

Jobs?45
Household enterprises (HHEs) are ubiquitous in Indonesia, and are an important
source of jobs and a vital part of the Indonesian economy. These small businesses
that employ the owner and no more than four individuals46 take many forms. Some are
run as an extension of the home; such businesses might include a neighborhood bakery, a
local laundry, a motorcycle mechanic, or even a physician’s private practice. Others operate
outside the home and go to the clients; such businesses include food carts selling fried rice,
ride-hailing motorcycle taxis, or traveling welders. See Box 4.3 for the definition of an HHE
used in this report.

Box 4.3 Definition of Household Enterprises versus Microenterprises and

Informal Jobs
While the term “microenterprise” usually refers to the size of the enterprise (either by employment or revenue
size), the term “household enterprise” (HHE) refers to the institution in which someone works. Since 2016, this
“institutions” variable has been made available in the National Labor Force Survey (Survei Angkatan Kerja Nasi-
onal; Sakernas). There is a huge overlap between microenterprises and HHEs, since the latter tend to be micro.
However, the term HHE is used to differentiate between small firms that are standalone entities and those that
are the extension of a household.
45 This section is based on Cunningham and Syamsulhakim (Forthcoming).
46 Sakernas does not provide information on the number of paid or unpaid employees in family businesses. However, data
from other countries (Vietnam, Cambodia, and Mexico, for example), find that most HHEs have one to two employees, and
many are unpaid family members.

Pathways to Middle-Class Jobs in Indonesia June 2021

148 Enterprise-Level Dynamics of Job Creation

Box 4.3 continued

Using the Sakernas, an HHE owner is defined as the job status of a worker who is either: (i) self-employed; (ii) an
employer assisted by one to four paid or permanent worker(s); or (iii) an employer assisted by unpaid or tem-
porary worker(s); AND who either: (a) does not have any written accounts; or (b) keeps simple accounts; AND
the type of work institution is identified as an individual or household business. The inclusion of bookkeeping
standards in the definition is intended to capture the casual nature of these firms where the line between the
household and the business is undefined. The restriction to one to four paid or permanent worker(s) is intended
to capture only the smallest enterprises (data on the number of workers is not available for employers assisted
by unpaid or temporary workers). In this report, this definition is referred to as the most restrictive definition
of an HHE; it encompasses 98.8 percent of HHEs. The remaining 1.2 percent of HHEs are considered formal
firms. The less restrictive definition does not control for the financial account variable and/or the number of
employees hired. The report uses Sakernas, in particular Sakernas August 2018, for all data analysis.

Due to data availability, the income-related analysis for HHE owners in this section will be limited to only self-em-
ployed HHE owners because the income data in Sakernas is only available for self-employed, wage employees,
and casual workers. Though HHE owners are either self-employed or employers, there is no income data for the
latter. There are about 22 million self-employed HHE owners in Indonesia, and 75.5 percent are in the nonfarm
sector.

Household enterprises employ approximately 83 million workers, which comprises

about 69 percent of total employment in Indonesia. Of these, 44 million are HHE owners,
meaning that there are 44 million HHEs in Indonesia. Half are self-employed, meaning that
no additional people work in the HHE, not even unpaid family workers. The remaining HHEs
employ 39 million people, an average of two unpaid, temporary, or permanent workers in
the enterprise, in addition to the owner.47 Most HHE owners regard their business as their
main job, but about one-quarter have an additional job.

A slight majority of HHEs operate in rural areas though a minority operate in the
farm sector. Nearly 55 percent of HHEs are located in rural areas, and the majority (54
percent) of HHEs are located in the island of Java. This is not surprising, not only because more
than half of the population of Indonesia lives in Java, but also because strong infrastructure,
such as the transportation and telecommunications found in Java, make it easier to conduct
and develop business. About 58 percent of rural HHEs operate in the farm sector, as do 15.4
percent of urban HHEs (Table 4.12). Of urban HHEs, 52.5 percent are engaged in wholesale
and retail trade, as is a sizable share (25 percent) of rural HHEs. Approximately 12 percent
of HHEs in urban areas engage in manufacturing, as compared to 9 percent of rural HHEs.

47 The data do not allow us to identify if an unpaid worker is employed in a firm with only other nonpermanent, unpaid, or
family members, or if they work in a firm that has paid employees.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 149

Table 4.12 Majority of Rural Household Enterprises Operate in the Farm Sector; Majority of Urban
Household Enterprises Operate in Wholesale and Retail Trade, and Restaurants and Hotels
Share of household enterprises (HHEs) by sector: urban and rural

Rural and
Sector Rural (%) Urban (%) urban (%)
Electricity, gas, and water supply 0.2 0.8 0.4
Mining and quarrying 0.9 0.4 0.7
Finance, insurance, real estate, and business services 0.3 1.6 0.9
Construction 0.9 1.3 1.1
Community, social, and personal services 2.4 6.9 4.4 Sources: Sakernas August
2018; World Bank staff
Transportation, storage, and communications 3.3 9.3 6.0 calculations
Manufacturing 9.5 11.9 10.6 Note: Sakernas is the National
Labor Force Survey (Survei
Wholesale and retail trade, restaurants and hotels 24.8 52.5 37.5
Angkatan Kerja Nasional).
Agriculture, forestry, livestock, and fisheries 57.8 15.4 38.4 Where percentages do not
total 100 percent, this is due
Total 100 100 100 to rounding.

Household enterprises are generally run by middle-aged men with a low level
of education. More than 28 million men own an HHE, as compared to 16 million women.
However, the share of men and women owning an HHE is proportional to their overall share
of the workforce. Among the 44 million HHE owners, about 80 percent are aged 35 or older.
Both HHE owners and casual workers have a similar average age (46 versus 41 years old),
while employees are younger (36 years old). Slightly more than half of HHE owners have
a primary education or less. Further, HHE owners and casual workers have similar average
years of schooling (7 years versus 6.6 years), while employees have an average 12 years of
schooling. After controlling for other relevant variables, the probability of being an HHE
owner is 3 percentage points less for high school graduates, as compared to those with no
formal education. Those with a university education are even less likely to be HHE owners;
the probability is 13 percentage points lower.

The majority of HHE owners started out because they wanted to run their own
businesses (62 percent), although around one-fifth started a business because
they could not get a salaried job. Using a nationally representative household-level
survey, the Digital Economy Household Survey, conducted in March 2020 with 1,542 sampled
HHEs,48 it can be seen that 62 percent of nonfarm and farm HHE owners started a business
because they wanted to run their own business (Figure 4.33). Independence was cited by
44 percent of the nonfarm HHE sample and 49 percent of the farm HHE sample. A higher
share of nonfarm HHE owners than farm HHE owners cited a desire to balance personal
and professional lives (24 versus 11 percent), supplement household income (16 versus 8
percent), and earn more than in a waged job (18 versus 14 percent). Nearly 30 percent of
farm HHEs cited family tradition. Around one-fifth of HHE owners started their enterprises
because they could not find a job as a salaried employee.49

48An HHE is defined in the survey as an enterprise owned by a household member that: (i) sells goods and/or services; AND
(ii) has no legal status; AND (iii) has no separate financial account from that of the family financial account; AND (iv) has at
most four paid workers.
49 Similarly, a study of HHEs in Mexico found that 16 percent of owners started a small firm because they could not find work
elsewhere (Cunningham and Maloney 2001). Similar results were found for Vietnam (Pasquier-Doumer, Oudin, and Thang
2017)

Pathways to Middle-Class Jobs in Indonesia June 2021

150 Enterprise-Level Dynamics of Job Creation

Figure 4.33 Most Owners of Household Enterprises Wanted to Start a Business

Rationale for starting a household enterprise (HHE), as share of HHE owners, by farm versus nonfarm HHE, 2020

Sources: Authors’ calculations using the Digital Economy Household Survey 2020.
Note: More than one response is possible.

Self-employed HHE owners earn lower average income per month than employees,
but are on par with casual workers. Due to data availability, as explained in Box 4.3, the
income-related analysis for HHE owners in this section is limited to HHE owners who are
self-employed (firm owners with any type of worker are not included in the income-related
analysis). This reduces the HHE-owner sample from 44 million HHE owners to 22 million self-
employed HHE owners. The mean and median of income per month (also per hour) for self-
employed HHE owners is about half that of an employee. The earnings gap is partly due to
the lower level of education of self-employed HHE owners as compared to employees, but
may also indicate that self-employed HHE owners pay a price when they choose to operate a
firm rather than work elsewhere. In contrast, the monthly earnings of HHE owners are very
similar to those of casual workers: the mean is slightly higher than that of casual workers,
but the median is slightly lower. Notably education levels are also similar, so casual work may
be the viable alternative form of employment for an HHE owner, indicating that owning an
HHE is not an inferior option. That said, the wage distributions of different types of workers
partly overlap, so there are HHE owners who earn more than employees of any type, and
there are also some that earn less (Figure 4.34, left panel). Comparing the three density
functions, the average, as well as the median, of the monthly earnings of HHE owners and
casual workers is lower than the minimum wage.

Earnings are higher in HHEs owned by educated men and in nonfarm sectors. For
example, owners who have tertiary education earn about 32 percent more than owners
who have only a primary education or less. Male HHE owners who have the same personal
and firm characteristics as female HHE owners earn about 42 percent more than the female
owners. Owners of nonfarm HHEs earn more than owners in the farm sector. For example,

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 151

HHE owners in construction earn about 57 percent more than HHE owners in the farm sector,
whereas HHE owners in mining and quarrying earn 33 percent more than HHE owners in the
farm sector.

Household enterprises use rudimentary business management tools. Only 21 percent

of HHEs use simple accounting practices and just 1.2 percent have complete accounts.
This means that most HHEs have neither sufficient awareness of the importance of basic
financial management, nor sufficient knowledge of it.50 Nearly 90 percent of HHEs do not
use the internet, potentially limiting enterprise productivity, market share, and business
management. Further, HHEs tend to be older, with an average age of 12 years. About 15
percent are start-ups (two years old or under), 38 percent are aged between 2 and 10 years
old, and 47 percent are old, with between 10 and 20 years of operation under the current
owner.

Figure 4.34 Monthly Earnings of Owners of Household Enterprises (HHEs)

Are Very Similar to Those of Casual Workers
Distribution of monthly earnings of HHE owners

Sources: Sakernas August 2018; World Bank staff calculations Sources: Sakernas August 2018; World Bank staff calculations
Note: Sakernas is the National Labor Force Survey (Survei Note: Sakernas is the National Labor Force Survey (Survei
Angkatan Kerja Nasional). Angkatan Kerja Nasional).

50 In calculating account ownership, to be able to show “complete account” and “don’t know” categories, the authors tempo-
rarily used the more restrictive definition of HHE: self-employed, employer assisted by one to four permanent worker(s), or
employer assisted by temporary worker(s).

Pathways to Middle-Class Jobs in Indonesia June 2021

152 Enterprise-Level Dynamics of Job Creation

4.8.1 A household enterprise can grow into middle-class status

Household enterprises can be a path to middle-class jobs and would be even more
so if some of the productivity constraints were addressed. As mentioned earlier,
income-related analysis linking HHEs to middle-class jobs will be limited to self-employed
HHEs. Approximately 9 percent of the self-employed HHEs produce earnings sufficient to
be classed as middle class, and most (8 percent of the total self-employed HHEs) are in the
nonfarm sector, leaving about 1 percent (about 250,000) operating in the farm sector (Ta-
ble 4.13).51

Table 4.13 Distribution of “Self-Employed” Household Enterprises (n=21.92 million)

Share of farm and nonfarm, by income class

Above middle Below middle

Total (%)
class (%) class (%)
Farm 1.2 23.4 24.5
Nonfarm 8.0 67.4 75.5
Total (%) 9.2 90.8 100.0

Sources: Sakernas August 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Some HHE owners can achieve middle-class earnings, but not at the rate that
employees do. Middle-class HHE owners, on average, earn more than casual workers but
less than employees (Figure 4.34, right panel). The median of the earnings, however, is the
same for middle-class HHE owners and employees.

Household enterprises can grow into middle-class status. From 2017 to 2018, the
number of HHEs achieving middle-class income in all sectors except agriculture, construction,
and community or social sectors, increased by 125,823 or about 9 percent. But these data do
not tell us if middle-class HHEs entered the labor market or if below-middle-class HHE grew
into middle-class HHE. To unravel this, the authors use the panel data from the Indonesian
Family Life Survey (Survei Aspek Kehidupan Rumah Tangga Indonesia) to see how HHEs with
different income-class status evolved over a seven-year period (2007–14). The authors find
that 8.5 percent of HHE owners without paid employees that were below middle-class status
in 2007 were able to move up to middle-class status by 2014. In addition, 0.9 percent of the
8.5 percent were able to hire paid workers (Figure 4.35). Another 0.7 percent were able to
obtain middle-class status by leaving their HHE and becoming an employee. The rest either
stayed as non-middle-class HHE owners (71 percent) or moved to non-middle-class employee
jobs (17.5 percent). A similar study (2020)52 looks at longer-term dynamics of HHE owners,
using 2000–14 panel observation from the Indonesian Family Life Survey. The study finds
that in 2014, after 14 years, about 69 percent of the panel sample from 2000 were still self-
employed.

51 As only 1.2 percent of farm HHEs have achieved middle-class status, the analysis focuses only on the nonfarm sector,
totaling 16.5 million HHEs.
52 World Bank 2020a

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 153

Figure 4.35 Progress of Household

Enterprise Owners, 2007–14
Number of workers, percentage share

Sources: Data from the Indonesian Family Life Survey

2007–14; World Bank staff calculations
Note: The Indonesian Family Life Survey is the Survei
Aspek Kehidupan Rumah Tangga Indonesia.

Household enterprise owners classified as middle class tend to have a minimum

of a secondary school education, and are usually male and without a disability.
Education is positively related with the probability of owing a middle-class HHE. More
than half of above-middle-class HHE owners had attained a secondary education or higher,
compared to only about one-third of below-middle-class HHE owners. An HHE owner with
a secondary education is 30 percentage points more likely to become middle class than
an owner with only a primary school education. Meanwhile, an HHE owner who is also a
university graduate is 62 percentage points more likely than a primary school graduate to
become middle class. In addition to education, 10 percent of the above-middle-class HHE
owners were receiving training in 2018, compared to 6.6 percent of below-middle-class HHE
owners, though this is still above the 2 percent for the whole workforce. Looking at other
characteristics, male-owned HHEs are 30 percentage points more likely to achieve middle-
class income than female-owned HHEs. A person with a disability is more likely than a person
without a disability to become an HHE owner, but only 6.5 percent of the 1.7 million above-
middle-class HHE owners have a disability, compared to 9.3 percent of the 14.1 million of
below-middle-class HHE owners.

Figure 4.36 Only 8% of Household Enterprises

in a Nationally Representative Survey in 2020
Had Profit Per Worker above the Middle-Class
Cutoff
Distribution of profit per worker

Sources: Author’s calculations using the Digital Economy

Household Survey, 2020
Note: Vertical line shows the middle-class cutoff.

Pathways to Middle-Class Jobs in Indonesia June 2021

154 Enterprise-Level Dynamics of Job Creation

Using more sophisticated management tools, including the internet, at work also
contributes to the middle-class status of HHEs. The probability of a nonfarm HHE
reaching middle-class status increases by about 38 percentage points if the HHE owner uses
the internet in their business, and by 37 percentage points if they use simple or complex
accounting methods. In addition, the likelihood of attaining middle-class status is higher for
businesses that are located in urban areas, or in Java, with more access to infrastructure
and a market. Lastly, the longer an HHE runs, the higher the chance that it will achieve
middle-class status.

Data suggest that labor productivity is too low in HHEs for owners to easily reach
middle-class status. Available data do not permit good estimates of labor productivity in
HHE, but “earnings” of self-employed HHEs and “profit per worker” in (any type of) HHEs from
the Digital Economy Household Survey can be used as rough proxies for labor productivity.
As discussed above, Figure 4.35 shows that less than 10 percent of self-employed HHE
owners have monthly earnings above the middle-class income cutoff. Data from the Digital
Economy Household Survey paint a worse picture. Using the Survey data, the upper bound
of the mean53 profit per worker in the sampled HHE was below the middle-class threshold.
Table 4.14 shows an upper bound of profit per worker of below Rp2 million regardless of
whether the owner plans to expand. The share of HHE with profit per worker above the
middle-class cutoff is only 8  percent (see Figure 4.36). Owners identify different factors
as impediments to doing better. Nonfarm HHE owners identify their top constraints as the
high cost of rent, lack of access to loans or credit, and excessive offline competition. The
top three problems of farm HHE owners are lack of access to loans or credit, weather, and
lack of skills or knowledge.

Table 4.14 Very Low Profit per worker

Profit per Worker Do you have any plan to
of HHE Reflects the develop
Lower Upper
Low Productivity of expand this Mean
bound bound
enterprise in the
HHEs in General
future?
Profit per worker; mean,
Yes 1,466,973 997,302 1,936,645
upper, and lower bound;
in Rp No 1,091,164 669,662 1,512,666

Source: Authors’ calculations using the Digital Economy Household Survey 2020
Note: Lower and upper bound of the 95 percent confidence interval

Around 80 percent of nonfarm HHE and two-thirds of farm HHE want to expand
their businesses. Most want to expand by expanding the premises of the enterprise,
followed by upgrading skills and knowledge. A desire to upgrade skills and knowledge is
particularly prevalent among farm HHE owners. A smaller share hope to borrow or attract
more funding in order to diversify their products. Only 11 percent of nonfarm HHE and

53 Using the 95-percent confidence interval.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 155

6 percent of farm HHE wanted to expand their business by selling online, or by joining a
digital platform (the survey was carried out before COVID-19; see Box 4.4). Among the 25–35
percent of the sample who did not want to expand their business, more than half said they
did not have enough money, but half of the nonfarm HHE said they were happy with what
they already had (Figure 4.37).

Figure 4.37 Most Firms Want to Expand Their Enterprise Premises, but among Nonfarm HHE
Who Don’t Want to Expand, Half Are Happy with What They Already Have
Reasons for not wanting to expand business, share who selected each option

Source: Author’s calculations using the Digital Economy Household Survey, 2020
Note: More than one response is possible.

Box 4.4 The Role of E-commerce in Helping Firms and Household Enterprises
Sail through the COVID-19 Crisis
The COVID-19-driven crisis and the containment measures put in place by government to limit the spread of the
extremely contagious infection—such measures include home isolation, large-scale social distancing, and a ban
on mudik (the exodus of workers for Eid al Fitr) in 2020 and 2021—have damaged the performance of many
firms, including those that were previously healthy. However, this crisis has also created a new opportunity for a
group of merchants across the world, namely those that are digitally connected and sufficiently digitally skilled.
In China, businesses relying on physical space and shops suffered significant losses, while online retailers with
apps built into social media became popular.a In fact, firms in emerging markets accelerated their e-commerce
implementation and broader diffusion of technologies such as online payment and (advanced) data analytics.b

a Bouey 2020
b World Bank 2021b

Pathways to Middle-Class Jobs in Indonesia June 2021

156 Enterprise-Level Dynamics of Job Creation

Box 4.4 continued

Online buying has helped keep demand afloat. The number of e-buyers has more than doubled during the pan-
demic, and 80 percent of these new e-buyers plan to continue buying online.c Retailers that had stayed away
from e-commerce were forced by the pandemic to adapt in order to survive (or to capture new business op-
portunities). Not only did e-commerce grow at the extensive margin (new users and buyers), but it also grew at
the intensive margin (intensity of selling and buying). A representative survey conducted in December 2020 in
collaboration with a regional e-marketplace, Shopee, and with more than 15,000 digital merchants participating
shows that e-commerce has been a valuable alternative source of income during the pandemic.d Among digital
merchants who joined Shopee in the last two years, 50 percent joined during the pandemic. On average, digital
merchants surveyed reported that although both online and offline sales decreased in April 2020, total sales
had, by December 2020, recovered to pre-pandemic levels, while online sales were stronger than pre-pandemic
levels. Close to 80 percent have so far kept their business open throughout the pandemic compared to 36 per-
cent of general firms,e reflecting their ability to sell online. Shopee has also reported an increase in the intensity
of buying and selling by existing buyers and sellers in terms of frequency and time spent.f

A representative firm-level survey conducted in October 2020 shows that two-thirds of firms were either al-
ready using, had started using, or were increasingly using the internet, social media, specialized apps, or digital
platforms. Their reasons for doing so were serving sales (82 percent) and marketing (75 percent). Controlling
for firm size, having digitalized a business by June 2020 was also found to help mitigate a year-on-year monthly
sales drop in October 2020. However, going online is not a panacea per se. Among enterprises that were already
online, such as e-marketplace merchants or food and beverages micro merchants in on-demand food delivery
digital platforms, surveys show that not all were successful. Analog complements appeared also to be factors
determining a firm’s performance: for example, an e-marketplace merchant changing their product offering to
respond to changing consumer demand, or the level of entrepreneurial experience of an online food merchant. g

In addition to e-commerce, the crisis could trigger the broader adoption of digital financial services. A nationally
representative survey, the Financial Inclusion Insights Survey, conducted by the National Council for Financial
Inclusion (Sekretariat Nasional Keuangan Inklusif) and Kantar between March and May 2019 showed that the
vast majority of Indonesians had no experience with e-money.h Within every educational level, the share of those
owning a smartphone was much higher than those who use e-money, indicating a gap and room for change.

With still a small share of the population using the internet at their workplace, or to buy and sell online, this crisis
has opened an opportunity for widespread digital adoption.i

c Google, Temasek, and Bain 2020

d World Bank 2021c
e The term “general firms” refers to both online and offline firms in selected sectors. The firms range in size from micro to large. See World Bank
2020d.
f Sea Limited 2020
g World Bank 2021c; World Bank and Grab 2021
h Moorena et al. N.d.
i Only 27.2 percent of workers in Indonesia use the internet at work (Sakernas 2019), while 12.8 percent of internet users (aged above 5 years old) buy
online and 5.1 percent sell online (Susenas 2019).
Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional). Susenas is the National Socioeconomic Survey (Survei Sosial
Ekonomi Nasional).

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 157

4.9 Conclusions
This chapter aims to better understand the key dynamics of job creation and job quality
among medium-sized and large manufacturing firms from 1990 to 2015 in their role as
creators of middle-class jobs in Indonesia. Three key findings have emerged from the analysis
of job dynamics across firms in this industry; these could motivate priority areas for policy
reform.

First, Indonesia’s hollowing-out decade of 2000–10 impaired job creation in industry

and firm dynamism remained muted after 2010, with fewer firms entering and
fewer new entrants surviving; more needs to be done to ensure greater dynamism
in industry. After the Asian financial crisis, jobs in industry were increasingly absorbed
by very large and very old firms, but even firms that entered and survived did not grow
much over their 40-year life cycle relative to other countries. In other words, Indonesian
manufacturing firms were stunted. This demands a closer examination of barriers to
entry and growth, among others, for Indonesian small and medium enterprises (SMEs) in
manufacturing.

Second, foreign-owned firms played an increasingly important role in the growth

of labor productivity and job creation in industry; improving competitiveness to
attract foreign investment is key.54 The share of foreign-owned firms out of total firms
in industry rose from 7 percent in 1990 to 26 percent in 2015. Foreign-dominated firms have
characteristics (for example, higher growth in export share) that support higher growth
of labor productivity, higher employment growth, and higher wages. Despite the growth,
however, the shares of value added, output, and employment attributed to multinationals
in the manufacturing sector were subdued between 2007 and 2015. There is ample room
for more FDI in the industry, especially because it is associated with a skills premium and a
wage premium for employees generally, and for female workers in low-skill manufacturing.

Third, firms with higher growth of labor productivity were associated with higher
employment growth; policies to increase labor productivity and support the
movement of workers to firms that are more productive and to higher-paying
manufacturing subsectors are necessary if middle-class jobs are to be created in
industry. There is no evidence that growth in labor productivity cuts jobs. On the contrary,
it creates jobs. The more productive firms are using their productivity gains to expand the
workforce; they are not cutting jobs. In other words, the expansion effects outweigh the
substitution effects. As labor productivity increases, wages also rise, albeit not proportionally
and not equally across the different ownership types. Increases in labor productivity are
still driven primarily by within-firm productivity growth, with very limited between-firm
productivity growth (involving workers moving to more productive firms). Moreover, only
high-technology, capital-intensive sectors paid wages that were commensurate with middle-
class status.

54 During 2008–15, using a different definition of foreign-owned firms (share of foreign ownership>50 percent) and exporters
(share of exports in sales>50 percent), foreign-owned firms and exporters were less likely to be new entrants; World Bank.
Forthcoming.

Pathways to Middle-Class Jobs in Indonesia June 2021

158 Enterprise-Level Dynamics of Job Creation

Some services sectors are also key to the creation of middle-class jobs, although
firm-level data in services is lacking. The services sector could become an even more
important source of new middle-class jobs in the face of Indonesia’s premature deindustrialization,
concerns of rising automation and Industry 4.0, shifting patterns of globalization, concerns over
labor-intensive manufacturing-led or trade-led growth,55 development in the region and in the
developing world (for example, China’s rise and subsequent slowdown and the emergence of
low-cost regional competitors, among others). However, though there is potential in the services
sector for creating middle-class jobs, the manufacturing sector is no less relevant (also discussed
in Chapter 3). Indonesia could avoid becoming a country that is too expensive to compete in
low-cost manufacturing, but whose workforce is not sufficiently skilled to compete in high
value-added manufacturing. A panel survey of services firms that collects more detailed data on
the trade in services would help develop evidence-based policy recommendations in the future.
Studies could also be carried out on the demand side of (middle-class) jobs in the services sector,
should this level of data on services firms become available. In the short term, since evidence on
the demand side of jobs in the manufacturing sector is available and analysis has already been
carried out, policies to create middle-class jobs can begin here. In the medium term, further
studies are needed to provide more evidence on which to base policies designed to create
middle-class jobs in the services sector.

The chapter also presents initial analyses on HHEs in Indonesia.

Household enterprises can be a path to middle-class jobs and would be even more
so if some of the productivity constraints were addressed. Given that the overall
share of income earners who hold middle-class jobs is only 15 percent, the 10-percent share
of self-employed HHE owners who hold middle-class jobs is not so far from the overall income
earners, albeit slightly lower. Having a middle-class job is more likely for those with certain
characteristics and ensuring household owners obtain these characteristics will increase the
opportunities for creating more middle-class jobs in Indonesia. The owner of an HHE who
has a minimum of senior secondary school education, uses the internet, carries out some
bookkeeping, is male, does not have a disability, is older, and whose enterprise is located
in an urban area or in Java is more likely to achieve middle-class status. This chapter also
highlights the low productivity of average HHEs as reflected in their low earnings.

The following policy recommendations apply to both manufacturing and services firms,
although most evidence is drawn from firm-level analysis of medium-sized and large manufacturing
firms. Based on earlier studies cited in Indonesia’s Country Private Sector Diagnostic56
and the Indonesia Systematic Country Diagnostic,57 reform priorities are: strengtheningthe
dynamism of private domestic firms; improving competitiveness to attract more foreign
investment; increasing labor productivity; and supporting the movement of workers into
more productive firms and into manufacturing subsectors that pay higher wages. Reforms
to the following policy areas are recommended to support the creation of middle-class jobs.

55 See World Bank 2020c; Hallward-Driemeier and Nayyar 2018.

56 Chang et al. 2019
57 World Bank 2020b

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 159

Reform Area 1: Unlocking new firm entry and growth to create

competition and innovation to spur productivity in the private
sector
The private sector accounted for 86.8 percent of all jobs in 2019 in Indonesia58 and reducing
the obstacles that hinder private firms from entering the market or growing is key for
generating robust growth in employment. However, to create quality jobs and expand
Indonesia’s middle class, Indonesia could do more—firms in Indonesia need to be better
integrated into global value chains (GVCs). This chapter has highlighted that Indonesian firms
that export more of their production are associated with higher growth in employment
and productivity (with implications for higher wages, albeit disproportionally). Leveling the
playing field between domestic and foreign firms, and small and large firms, among others,
is key if Indonesia is to build dynamism among firms and robust job creation in the private
sector. Export-oriented FDI is typically associated with higher rates of product and process
innovation. In developing countries, greater openness to FDI is associated with better
infrastructure and labor quality, and higher participation in GVCs for firms in the electronics
and automotive sectors.59 Along with reforms to close gaps in infrastructure, human capital,
and financial services, Indonesia needs a more predictable regulatory framework that will
enable firms to enter and grow. As firm dynamism grows and efficiency-seeking FDI arrives,
Indonesia will need to support workers to transition to higher-quality jobs and will also
need to increase the capacity of domestic firms and workers to benefit from technology
spillovers.

Five areas of intervention are critical to unlock new firm entry and growth to create the
competition and innovation that spur productivity and create middle-class jobs in the
private sector.

First, Indonesia could integrate more domestically owned firms into GVCs.60 One of
the most serious constraints to the GVC integration of Indonesian firms is insufficient access
to inputs, markets, and skills. Four policy reforms could be prioritized (these are discussed
in more detail in Chapter 3):

i. Indonesia could reduce the high cost of trade in order to increase firms’ access to quality
inputs;
ii. Indonesia could complete key free trade agreements in order to increase access to
international markets;
iii. Indonesia could reduce restrictions on work permits in high-skill occupations for which
there is a shortage of domestic workers and provide incentives to help firms nurture in-
house talent; this will increase the country’s access to necessary skills;
iv. Indonesia could relax restrictions on investment in order to increase FDI.

58 Sakernas, August 2019. The private sector includes profit-making institutions, cooperatives, individual enterprises, and
HHEs.
59 Soejachmoen 2012
60 Also see World Bank 2020c

Pathways to Middle-Class Jobs in Indonesia June 2021

160 Enterprise-Level Dynamics of Job Creation

Second, Indonesia could strengthen its competition regulatory framework. A

regulatory framework that is conducive to competition is crucial to ensure that allocative
efficiency is improved—that is, finite labor, capital, and knowledge or technology should be
able to go to more productive firms. The 1999 Indonesian Competition Law established the
Business Competition Supervisory Commission (Komisi Pengawas Persaingan Usaha; KPPU),
which is tasked with enforcing competition policy. However, both the competition framework
and the KPPU still suffer from limitations that make Indonesia’s competition regime one
of the least effective of 49 countries surveyed by the OECD.61 Effective implementation
of the Competition Law requires both a stronger KPPU and a more robust government
commitment. There are three priority areas for policy reform that would make the current
competition regulatory framework more effective: (i) strengthening the KPPU’s technical
capacity to enforce competition laws and advocate pro-competition policies; (ii) revisiting
the existing Competition Law (No. 5/1999) to improve the ability of Indonesia’s competition
framework to identify and sanction anticompetitive behavior; and (iii) mainstreaming
competition considerations in the policy-making process to ensure that regulations do not
unduly restrict competition.62

Third, Indonesia could support firms with fewer, but more targeted instruments
and a predictable regulatory framework. The first reform effort should focus on
improving the policy-making process and the introduction of a regulatory oversight
body to reduce regulatory uncertainty.63 Compulsory impact assessments of regulations
should be complemented by a simplified process for the incorporation and operation of
firms and the establishment of a regulatory oversight body designed to promote good
regulatory practices across the vertical and horizontal government levels. Over the
longer term, introducing compulsory public consultation—that follows international best
practice—into the regulatory process and providing incentives for local governments to
streamline business licensing will be critical. The second reform effort should focus on
consolidating and targeting innovation policy instruments to better assist firms to perform
innovative activities that will spur productivity growth and new product development.64
The low productivity of firms in Indonesia partly reflects the low levels of innovation. Public
spending on innovation is highly fragmented and ineffective at addressing the needs of the
private sector.65 Specific policy interventions may include combining disparate innovation
instruments and better targeting new instruments to address the needs of firms, including
in terms of technology adoption.

61 Chang et al. 2019

62 Chang et al. 2019
63 World Bank 2020b
64 World Bank 2020b
65 World Bank. Forthcoming

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 161

Fourth, Indonesia could increase the capacity of domestic firms and workers to
benefit from the technology spillovers of FDI. To enhance the economywide benefits of
FDI, policies need to help domestic firms benefit more from interaction with multinationals
through links or imitation. Experiences from countries such as South Korea (with huge levels
of Japanese FDI that started in the 1970s)66 and Vietnam (with Samsung’s massive investment
in mobile phone production since 2009)67 provide lessons learned on technology spillovers
from FDI. Policies to increase the capacity of domestic firms and workers to benefit from
the technology spillovers of FDI include increasing the quality of human capital and domestic
workers (detailed policies on improving education and skills are discussed in Chapter 5),
promoting firm links, providing incentives and advocacy to promote firm-based training
programs, building business networks, establishing institutional partnerships, and providing
information services—all of which can increase both absorptive and adaptive capacity among
local firms. It is difficult for local firms to establish direct links to multinationals, but it is
possible for them to supply the tier-three or higher suppliers of multinationals, since many
large multinationals coming to Indonesia often bring their own tier-one, tier-two, and even
tier-three suppliers to the country. However, in the medium to long term, national policy
could address the question of raising the capability of domestic firms to supply relatively
high value-added products to higher-tier suppliers of multinationals.

Fifth, Indonesia could close information gaps to help workers transition to higher-
quality jobs. Higher labor productivity growth that comes from a systematic movement of
workers from firms with low levels of productivity to more productive firms is necessary, as
highlighted in this chapter. This requires not only new firms to enter and firms that are more
productive to outcompete less productive firms, but also for people to be able to move out
of their current jobs and into more productive sectors, firms, or jobs. In Indonesia, labor
mobility transition costs for workers in the manufacturing sector are high by international
standards.68 While skills gaps are an important constraint to labor mobility, transitions are
also low between sectors with similar (and low-to-medium) skill requirements. Policies to help
workers move to better jobs are needed. The first reform effort should focus on developing
a robust labor market information system and job-matching mechanism. (Details of policies
to close information gaps and help workers transition to higher-quality jobs are discussed
in Chapter 5.) The second reform effort should focus on reskilling and upskilling current
workers so that they can move to better jobs. (Details of policies on upskilling adult workers
are discussed in Chapter 5.)

66 Castley 1996
67 Tong, Kokko, and Seric 2019
68 Calì, Hidayat, and Hollweg 2018

Pathways to Middle-Class Jobs in Indonesia June 2021

162 Enterprise-Level Dynamics of Job Creation

Reform Area 2: Improving the productivity of SMEs so that they

grow and become better creators of middle-class jobs
In Indonesia, SMEs (excluding microenterprises) accounted for only 10 percent of
employment in 2014 and contributed about 27 percent of GDP in 2013. Average employee
productivity in medium-sized enterprises is about one-third that of large enterprises, and
in small enterprises it drops to about one-fifth.69 Indonesia’s SMEs are its “missing middle”—
Indonesia has many microenterprises and large firms, but not many medium-sized firms.
Although the shares of SME employment and GDP as a proportion of the total are small
compared to microenterprises (in terms of employment) and large enterprises (in terms
of GDP), SMEs have higher growth prospects and are more likely than microenterprises
to create middle-class jobs. They are also an important part of the economic growth and
social inclusion story of Indonesia. However, compared to other countries in ASEAN, missing-
middle SMEs in Indonesia remain less competitive due to their low productivity and levels of
innovation.70 Ineffective government policies toward SMEs have contributed to this outcome.
The government has issued a range of policies to improve productivity growth among SMEs,
including policies designed to make business registration cheaper and faster, policies on
technology transfer and e-commerce, on access to market and to export markets, and on
financial assistance. However, these policies continue to take a welfare approach, instead
of one that emphasizes competition, efficiency, and productivity. The policies also contain
excessive protectionism: for example, some business sectors are reserved for SMEs and
other programs shield SMEs from competition.71

There are five priority policy areas of intervention that would improve the productivity of
SMEs, helping them to grow and become better creators of middle-class jobs in Indonesia.

First, Indonesia could develop specific measures and strategies to assimilate its SMEs
into GVCs.72 As global trade becomes more integrated, SME integration into GVCs is essential
for a large economy like Indonesia. It has become increasingly important for Indonesia
to design strategies specifically aimed at improving SME competitiveness and links with
larger businesses, both domestic and multinational, to integrate them into regional and
international value chains. For many developing countries, including Indonesia, integrating
smallholders into agriculture value chains and home-based workers into manufacturing
and services GVCs will ensure that supply chains benefit the poorest.73 Moreover, it will
also be critical to address information gaps between SMEs and multinationals about the
quality standards SMEs must meet to supply to multinationals or their tier-three or higher
suppliers, and to support SMEs to meet these quality standards and deliver goods in a timely
manner. In order to develop links between SMEs and multinationals, the government could
also provide tax incentives for multinationals that invest in the upgrading of local suppliers

69 Anas, Mangunsong, and Panjaitan 2017

70 Anas, Mangunsong, and Panjaitan 2017
71 OECD and ERIA (Economic Research Institute for ASEAN and East Asia) 2018
72 Also see World Bank 2020c
73 World Bank 2020c

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 163

through training, mentoring, or staff secondment.74 Other reforms are also critical if SMEs
are to be integrated into GVCs. Such reforms include: lowering transaction costs for SMEs
when they use free trade agreements, simplifying the Certificate of Origin procedure, and
helping SMEs to arrange Certificates of Origin and reducing the associated costs to zero.75

Box 4.5 Case Study: Indonesia’s SME Participation in Global Value Chains:
Javara
Javara was founded in 2008 by Helianti Hilman. She was inspired by indigenous farmers and food artisans
who were striving to preserve Indonesia’s food biodiversity, indigenous wisdom, and spiritualism. Javara is a
mission-driven enterprise working across supply chains, from production to distribution, in order to preserve
Indonesia’s biodiversity and bring community-produced, organic products to broader markets. Starting small,
to date, Javara has created more than 600 artisanal products and works with tens of thousands of smallholder
farmers across Indonesia. One of Javara’s goals is to create inclusive economies in remote and rural areas and
reverse the “forced” urbanization. Javara helps farmers to remain in their villages and increase their incomes
by improving their skills to increase value added and capture more of the supply chain. Javara intervenes
along supply chains to strengthen the production capacity of suppliers, improve workplace safety, and market
products nationally and internationally, securing premium prices for farmers and processors. Javara currently
holds a portfolio of more than 250 organic products that are certified under the EU, the US National Organic
Program, and Japanese Agricultural Standards, and has exported to more than 22 countries in 4 continents.
In 2017, Ms. Hilman founded Sekolah Pangan, a food-related entrepreneurial program and business ecosystem
dedicated to nurturing businesses among rural youth, women, and indigenous communities. The curriculum
includes innovative farming methods, added-value processing, culinary businesses, and eco-cultural tourism.

Source: Javara website. https://javara.co.id/our-journey/

Second, Indonesia could promote the incremental adoption of innovative activities

to boost productivity. Although a number of programs exist, they appear to be mainly
focused on collaboration between research and development (R&D) institutions and larger
companies. However, for developing countries like Indonesia, where innovation capabilities
are less developed, it is more usual to incrementally adopt innovative activities—for example,
upgrading of processes and quality—and new business models, as well as product imitation.
An incremental approach can yield great payoffs in terms of boosting productivity and
employment.76 Innovation policy in Indonesia could prioritize and facilitate this incremental
process of learning and of accumulating innovation capabilities. In line with this, Indonesia
could focus on employing instruments that support absorptive capacity and foster
collaboration and simple innovation projects. Policy reforms could focus on a mix of policy
instruments that may include technology extension and diffusion programs, early-stage
infrastructure and advisory services, incentives, vouchers for collaboration, and direct
grants for business innovation (with embedded advisory services).77

74 OECD (Organisation for Economic Co-operation and Development) 2018

75 Anas, Mangunsong, and Panjaitan 2017
76 Cirera et al. 2020
77 Cirera et al. 2020

Pathways to Middle-Class Jobs in Indonesia June 2021

164 Enterprise-Level Dynamics of Job Creation

Third, Indonesia could close the digital infrastructure gap and ensure the availability
of affordable, reliable, and ubiquitous high-speed internet. Digital technologies are the
newest tool in the productivity box and have already shown their role in disrupting markets
and leapfrogging some enterprises to productivity levels that would not have been feasible
in the pre-digital economic space. During the Covid-19 crisis, being digitally connected has
become even more pertinent (Box 4.6). However, the digitalization of Indonesian SMEs is still
relatively limited.78 A study by Deloitte (2015) shows that greater use of digital technologies
(including social media, broadband, and e-commerce) could increase SME revenues by up to
80 percent, and makes them 17 times more likely to introduce innovations and one and a
half times more likely to create jobs.79 Considering that Indonesia had 105 million internet
users in 2017, digitalization strategies could help domestic SMEs to scale up and become
more productive. However, there is a significant digital divide in Indonesia and this remains
a challenge. Only 27.2 percent of Indonesia’s workers used the internet in their workplace in
2019, with many factors at play, including a lack of meaningful digital literacy (literacy that
could increase productivity) and a digital infrastructure divide. Indonesia has uneven mobile
broadband penetration and is lagging considerably on fixed broadband penetration. Fixed
broadband, which matters particularly to businesses, is available to less than 5 percent of
the population and about 15 percent of households. Increasing digital skill is critical (details
of potential policies are discussed in Chapter 5). Specific policy reform areas that address
the digital infrastructure divide include: (i) accelerating access to high-speed broadband
internet (fixed and mobile 4G) nationwide through reforms to spectrum management
and infrastructure sharing (this will stimulate further private investment); and (ii) closing
remaining internet access gaps through targeted public-private partnerships.

Fourth, Indonesia could expand access to financial services. At 27.4 percent, the
proportion of firms with a line of credit is low compared to actual need. A vast majority of
firms that do not have access to a line of credit are microenterprises or SMEs. Only one in four
small enterprises in Indonesia have received a bank loan in recent years. Kredit Usaha Rakyat
(KUR) is the main program providing SMEs with access to finance. However, it faces many
challenges.80 More generally, the current financial market infrastructure does not make it
easy for SMEs to conduct their business. Further, the majority of SMEs are not included in
the country’s credit information system: they may be unaware of the system or unable to
meet the requirements to be included in it or to get a loan. However, global evidence shows
that increased access to finance results in higher employment growth, especially among
microenterprises and SMEs.81 These findings should lead to policy interventions on access to
finance, including interventions that are targeted to produce job growth. More recent global
evidence shows that among other factors, developing the finances of SMEs contributes

78 OECD (Organisation for Economic Co-operation and Development) 2018

79 Deloitte 2015
80 OECD (Organisation for Economic Co-operation and Development) 2018
81 Ayyagari et al. (2015) provide evidence on the effect of access to finance on job growth in 50,000 firms across 70
developing countries. See also Gomez-Mera  and Hollweg (2018) for the correlation between access to finance and the
performance of Indonesian firms.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 165

significantly to increasing the probability of a high-growth episode.82 With reforms, SMEs

can obtain access to finance and generate more job opportunities.

The first priority policy step to increase SME access to financial services is to prepare a
roadmap with a view to leveling the playing field for delivery of financial services. The
second priority policy reform is to strengthen financial sector infrastructure (such as
credit registries and secured transactions). Sharing of credit information in Indonesia is
evolving into a new phase of development since the regulation and management of the
Credit Information System was transferred from Bank Indonesia to the Financial Services
Authority (Otoritas Jasa Keuangan; OJK). In parallel, two private credit bureaus (Lembaga
Penyedia Informasi Perkreditan) have been granted licenses and are now formally
operating to facilitate commercial bank and nonbank financial institutions access credit
information. The information stored in these systems can relate to both individuals and
businesses. However, a weak regulatory framework, an underdeveloped oversight function,
an underdeveloped consumer redress function, and limited technical capacity across the
financial sector represent major limitations. The third suggested policy reform is promoting
the use of dedicated liquidity and risk-sharing facilities, as well as other lending instruments,
through the financial sector. This has become particularly urgent given the COVID-19 crisis.

82 Grover Goswami, Medvedev, and Olafsen 2019

Pathways to Middle-Class Jobs in Indonesia June 2021

166 Enterprise-Level Dynamics of Job Creation

Reform Area 3: Targeting support to increase productivity in

HHEs where a large share of jobs exists
Household enterprises, which constitute more than 98 percent of microenterprises, are
significantly different to SMEs in many aspects (Table 4.15). Policies and programs to support
and increase the productivity of HHEs should consider these differences. Unfortunately,
current policies around micro, small, and medium enterprises (MSMEs) tend to favor SMEs.83

Table 4.15 Key Characteristics of Microenterprises/Household Enterprises, and Small,

and Medium-Sized Enterprises

Issue Microenterprise Small enterprise Medium-sized enterprise

Degree of informality is lower
Degree of informality is high All operate in the formal
Formality (many operate in the formal
(most operate informally) sector
sector)
Many are non-primitive units All have formal organizational
Organization and
Primitive/traditional with modern management structure with modern
management
systems management system
Most are family businesses
Many use wage-paid
Workers relying on unpaid family All use wage-paid employees
employees
members as workers/helpers
Degree of automation is much
Production process Traditional/manual Many are highly mechanized
higher
Most are very locally oriented
Market orientation and serve local low-income Local, national and/or export National and/or export
households
Most are well educated and
Economic and social No education, or low level of Many are well educated and
from medium- to high-income
profile of owner education, and poor from nonpoor families
families
In general, out-of-date
Use of modern technology
machinery is used or manual Many use machines and
Technology is much higher and all use
processes, and no use of information technology
information technology
information technology
Gender of owner or Many are owned or managed Fewer women are involved as Very few women are owners
entrepreneur by women owners or entrepreneurs or entrepreneurs
Reason for running own Mostly for profit All for profit
In general, to survive
business
Spirit of entrepreneurship In general, low Mostly high All high

Source: Tambunan 2019

83 For example, Kredit Usaha Rakyat (KUR; a credit scheme with interest subsidies) can be accessed by MSMEs. However,
banks require collateral and most microenterprises cannot supply it. In addition, HHEs usually need small loans of a size that are
too small for banks, as they will be too costly to provide. For example, a small food outlet requires a small amount of working
capital on a daily basis, and will be reluctant to borrow large amounts of money for a longer period of time. In short, HHEs need
a different loan scheme that matches their requirements. None is yet available in the current KUR scheme.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 167

One of the key characteristics of HHEs is that they are not as growth-oriented
as SMEs—this applies particularly to those owned by women—so specific policy
reforms should focus on increasing their productivity and hence their income
incrementally without demanding a significant amount of additional resources
(time, worker, capital, or financing, among others). Although according to the latest
household-level survey, around 80 percent of nonfarm HHEs and two-thirds of all HHEs want
to expand their businesses, among women entrepreneurs in Indonesia, the majority of them
fall into the category of “necessity” enterprises.84 Specific policy reforms around HHEs should
focus less on encouraging growth, and more on small increases in productivity, and hence
in income. If small productivity gains could be realized across more than 44 million HHEs in
Indonesia, the total productivity gain would be enormous. Intervention in two particular
policy areas would improve the productivity of HHEs.

First, the basic business skills of HHEs can be improved with training and education.
Indonesian HHE owners use rudimentary (or no) formal business practices. But, as highlighted
in this chapter, the probability of a nonfarm HHE reaching middle-class status could be
increased by 37 percentage points if the owner uses an accounting method, whether simple
or complex. Improving the basic business management skills of HHE owners or workers
through standardized or tailormade training would raise productivity and increase their
chances of expanding. At least three skill types need to be improved: financial literacy,
management, and entrepreneurial.85 Strategies to improve the business skills of HHEs
should be, in sequence (short term then medium term): (i) creating short training courses
with simplified rule-of-thumb modules in business skills, combined with training in personal
initiative (a soft skill), competition, and innovation; (ii)  providing information about the
benefits of training and ways of acquiring it; and (iii) in the medium term, synchronizing and
integrating government training and support programs across ministries and bodies.

Second, inclusive and supportive ecosystems can be created for HHEs, targeting,
in particular, a reduction in the vulnerability of HHE owners. Chapter 4 discusses how
infrequently a self-employed HHE owner can increase their earnings to middle-class level.
Both HHE owners and their workers have very little economic security. However, greater
access to social benefits may bring them closer to a more secure livelihood than if they
relied solely on labor income. Only 2.4 million nonwage workers were registered in the
social security program (Badan Penyelenggara  Jaminan Sosial Ketenagakerjaan; BPJS
Ketenagakerjaan) in 2018 out of a population of 44 million, partly due to limited knowledge
of the benefits and importance of the scheme. Willingness to pay is also low. The owners of
HHEs and their workers are vulnerable, and policy and encouragement is urgently needed to
increase their membership in the scheme. Specific policy reforms that could create inclusive
and more secure ecosystems for HHEs are: (i) encouraging HHE owners and workers to join
BPJS Ketenagakerjaan by providing tailored information on the benefit of registration in
the program; (ii) providing co-contributions from government to ease entry and sustain

84 Arsana and Alibhai 2016

85 See evidence of the importance of these skills in Drexler, Fischer, and Schoar (2014); Mangla, Agarwal, and Pandey
(2015); Campos et al. (2017); Duening (2010); and Kusumawardhani et al. (2019).

Pathways to Middle-Class Jobs in Indonesia June 2021

168 Enterprise-Level Dynamics of Job Creation

participation in the program; and (iii) promoting the benefits of becoming a legal entity by
bundling membership with other government support programs designed to increase the
productivity of HHEs, such as access to finance and to skill development.

Third, product competitiveness could be improved. Early evidence shows that a low
level of labor productivity in an HHE will prevent its owner from easily reaching middle-class
status. The low level of productivity leads to low competitiveness. Most HHEs have limited
working assets and limited knowledge of innovating, diversifying, and differentiating their
products from those of their competitors. Consequently, they struggle to compete in fierce
markets. For example, in agriculture, 56 percent of farm HHEs have less than 0.5 hectares of
land, meaning they cannot diversify their crops, let alone take advantage of the economies
of scale that are possible on larger farms. Most nonfarm HHE also face similar problems.
Farmer field schools, where farmers can learn to adapt old methods or train in new ones,
significantly increase their skills.86 In some cases, farmers have come up with better or more
efficient ways of using the new methods they learn. Increasing economies of scale of HHE can
be achieved through use of shared factories. First, basic work mechanisms must be developed
among HHEs, for example, assigning autonomy and responsibilities, setting decision-making
arrangements, coordinating objectives and efforts, networking, and managing problems.
Establishing such cooperation among HHEs requires guidance from local government, which
acts as coordinator. Product standardization is beneficial for every business entity, including
HHEs. Businesses gain credibility, can build a brand, competition increases, profits grow, and
so does the HHE. Most HHEs have no knowledge of standardization and its benefits. Specific
policy reforms recommended to improve product competitiveness are: (i) promoting on-
the-job training and apprenticeships; (ii) providing shared factories or machines that can
be used by HHEs; and (iii) supporting product standardization and cooperation among HHEs.

As highlighted in this chapter, one-fifth of HHE owners started their enterprises

because they could not get a job as a salaried employee. Indonesia needs policy reform
to close information gaps and help HHE owners and workers transition to higher-quality
jobs. The reforms should include the development of a robust labor market information
system and job-matching mechanism, as well as reskilling and upskilling of current workers.
These reforms will help HHE owners and workers to move to salaried jobs.

86 Luther et al. 2018

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 169

Box 4.6 Banking Agents: Benefits to Household, Small, and Medium-Sized

Enterprises
The existence of the networks of banking agents (such as Laku Pandai and Layanan Keuangan Digital) have sev-
eral direct and indirect advantages for household enterprises (HHE), and small and medium-sized enterprises
(SMEs):

1. First, HHEs and SMEs acting as banking agents earn income from every banking transaction they conduct
(especially those conducted electronically) and from every cash-in-cash-out transaction. The fee obtained from
each transaction can contribute quite significantly to the total income of an HHE or SME, especially in areas
where access to banking facilities is limited or nonexistent. A traditional Laku Pandai agent offering banking
services could earn up to Rp0.4–2 million per month, or 10–50 percent of their net income, by offering basic
banking services.a

2. Government social assistance programs consist of benefits distributed either as conditional cash transfers
or in-kind benefits (similar to e-vouchers) that can be liquidated or used to buy food at banking agents under
the HIMBARA network (HIMBARA is the abbreviation for Himpunan Bank Negara, the Association of State-Owned
Banks). These schemes accelerate recruitment of banking agents and expand the scope of services provided
by banking agents in the regions. For example, the number of Laku Pandai agents increased from 3,734 in June
2015 to 1,146,131 in September 2019.b In addition, HHEs and SMEs that own stores may also receive an additional
benefit, namely increased footfall as people visit the store to carry out banking transactions, pay bills, or with-
draw cash.

3. Many of the HHEs and SMEs recruited by state-owned banks have improved their digital and financial knowl-
edge, because they are also expected to provide digital and financial education both directly and indirectly to
their customers. In many cases, children of older HHE and SME owners have better digital and financial knowl-
edge and teach their parents, helping them cater to their customers.

4. Various kinds of banking and other transactions (for example, paying bills) can be carried out closer to home
and potentially more cheaply than at a traditional bank.

5. The existence of banking agents can also enhance the culture of saving or accumulating assets among house-
holds, individuals, and enterprises; services are easy to access and have relatively informal procedures. Custom-
ers can save at any time and any amount through a banking agent without needing to self-conscious about the
amount they are saving. The new financial records of these customers, saved by the agents, enable banks to
assess the credentials of customers should they apply for credit. This could potentially relax banks’ collateral
requirements.

a See, for example, Boston Consulting Group and MicroSave Consulting (2019).
b OJK (Otoritas Jasa Keuangan). N.d.

Pathways to Middle-Class Jobs in Indonesia June 2021

170 Enterprise-Level Dynamics of Job Creation

Annex Chapter 4
Annex 4.1 Additional Figures

Annex Figure 4.1 Job Growth Slowed during the Hollowing-Out Period
Job growth (upper and lower bound)

Sources: Manufacturing Survey; World Bank staff calculation

Note: The upper bound of job growth is calculated by assuming that the number of employees of firms exiting, but not closing down, falls to 19 (20
employees is the threshold of the Manufacturing Survey). The lower bound of job growth is calculated by assuming that the number of employees
of firms exiting, for any reason, falls to 0. The Manufacturing Survey is the Survey of Medium-Sized and Large Manufacturing Firms (Survei Industri
Besar/Sedang).

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Enterprise-Level Dynamics of Job Creation 171

Annex Figure 4.2 (Log) Productivity Differentials for Apparel

Sources: Manufacturing Survey; World Bank staff calculations

Note: All coefficients are significant at 10 percent confidence interval, except nonmetallic mineral for productivity. Dark Green means low-
technology-intensity sectors, green means medium-technology-intensity sectors, and light green means high-technology-intensity sectors. The
apparel subsector is outperformed by most other subsectors in terms of both real wages and productivity. Except for food, textile, tobacco, and
nonmetallic metal, other manufacturing subsectors outperformed the apparel subsector in real wages. All subsectors outperformed apparel in
productivity (real value added per worker). The Manufacturing Survey is the Survey of Medium-Sized and Large Manufacturing Firms (Survei Industri
Besar/Sedang).

Annex Figure 4.3 (Log) Wage Differentials for Apparel

Sources: Manufacturing Survey; World Bank staff calculations

Note: All coefficients are significant at 10 percent confidence interval, except nonmetallic mineral for productivity. Dark Green means low-
technology-intensity sectors, green means medium-technology-intensity sectors, and light green means high-technology-intensity sectors. The
apparel subsector is outperformed by most other subsectors in terms of both real wages and productivity. Except for food, textile, tobacco, and
nonmetallic metal, other manufacturing subsectors outperformed the apparel subsector in real wages. All subsectors outperformed apparel in
productivity (real value added per worker). The Manufacturing Survey is the Survey of Medium-Sized and Large Manufacturing Firms (Survei Industri
Besar/Sedang).

Pathways to Middle-Class Jobs in Indonesia June 2021

172 Enterprise-Level Dynamics of Job Creation

Annex 4.2 Sector Classification

Annex Table 4.1 Manufacturing Subsector Classification

Broad sector Subsectors

Food, beverages, and tobacco products
Wood and wood products
Other nonmetallic mineral products
Fabricated metal
Low-skill
Paper and paper product, printing and publishing
manufacturing
Rubber and plastics products
Base metals
Textiles, clothing, and leather products
Furniture; manufacturing not elsewhere classified
Coke and refined petroleum products
Chemicals and chemical products
Machinery and equipment not elsewhere classified
High-skill
Transport equipment
manufacturing
Electrical machinery and equipment
Computers, electronics, and optical equipment
Pharmaceutical products

Sources: Steenbergen and Hebous 2020

June 2021 Pathways to Middle-Class Jobs in Indonesia

 References 173

References
Alatas, Hamidah, and Maria Monica Wihardja. Forthcoming. “Firm-Level Analysis of Job
Creation Dynamics in the Medium and Large Manufacturing Industry in Indonesia.”
Background paper, Pathways to Middle-Class Jobs in Indonesia, World Bank, Jakarta.

Anas, Titik, Carlos Mangunsong, and Nur Afni Panjaitan. 2017. “Indonesian SME Participation
in ASEAN Economic Integration.” Journal of Southeast Asian Economies 34 (1) 77–117.

Anas, Titik, and Thaliya Wikaspuspita. 2019. “Indonesia’s Services Sector: Performance,
Policies and Challenges.” In The Indonesian Economy in Transition: Policy Challenges in
the Jokowi Era and Beyond, edited by Hal Hill and Siwage Dharma Negara. Singapore:
ISEAS.

Arnold, Jens. M., and Beata S. Javorcik. 2009. “Gifted Kids or Pushy Parents? Foreign
Direct Investment and Plant Productivity in Indonesia.” Journal of International
Economics 79 (1): 42-53.

Arsana, I Gede Putra, and Salman Alibhai. 2016. Women Entrepreneurs in Indonesia: A

Pathway to Increasing Shared Prosperity. Washington, DC. : World Bank Group.
http://documents.worldbank.org/curated/en/738881467782741648/Women-
entrepreneurs-in-Indonesia-a-pathway-to-increasing-shared-prosperity

Aswicahyono, Haryo, Douglas H. Brooks, and Chris Manning. 2011. “Exports and Employment
in Indonesia: The Decline in Labor-Intensive Manufacturing and the Rise of
Services.” ADB Economics: Working Paper Series No. 279, Manila, Asian Development
Bank.

Aterido, Reyes, and Mary Hallward-Driemeier. 2017. “Private Sector Firms in Vietnam.”
(PowerPoint presentation for internal use.) Social Protection and Jobs;
Macroeconomics, Trade, and Investment; Finance, Competitiveness, and
Innovation, World Bank Group.

Ayyagari, Meghana, Pedro Juarros, Maria Soledad Martinez Peria, and Sandeep Singh.
2015. “Access to Finance and Job Growth: Firm-Level Evidence across Developing
Countries.” Policy Research Working Paper 7604, World Bank, Washington, DC.
https://home.gwu.edu/~ayyagari/Finance-growth.pdf.

Bartelsman, Eric, John Haltiwanger, and Stefano Scarpetta. 2009a. “Cross-Country

Differences in Productivity: The Role of Allocation and Selection.” NBER Working
Paper 15,490, National Bureau of Economic Research, Cambridge, MA.

Bartelsman, Eric, John Haltiwanger and Stefano Scarpetta. 2009b. “Measuring and
Analyzing Cross-Country Differences in Firm Dynamics.” In Producer Dynamics:
New Evidence from Micro Data, edited by T. Dunne, J. B. Jensen, and M. J. Roberts.
Cambridge, MA: National Bureau of Economic Research.

Pathways to Middle-Class Jobs in Indonesia June 2021

174 References

Blalock, Garrick, and Paul J. Gertler. 2008. “Welfare Gains from Foreign Direct Investment
Through Technology Transfer to Local Suppliers.” Journal of International Economics
74 (2): 402–21.

Blomström, Magnus, and Fredrik Sjöholm. 1999. “Technology Transfer and Spillovers: Does
Local Participation with Multinationals Matter?” European Economic Review 43 (4–6):
915–23.

Boston Consulting Group and MicroSave Consulting. 2019. “CICO Economics in Indonesia:
Research Findings.” July, slide 11. https://connect.bcg.com/cicoeconomics/6-cico-
economics-in-indonesia_aug-2019-2/.

Bouey, Jennifer. 2020. “Assessment of COVID-19’s Impact on Small and Medium-Sized

Enterprises: Implications from China.” Santa Monica, CA: RAND Corporation.

Calì, Massimiliano, Martín Cicowiez, Aufa Doarest, Taufik Hidayat, and Dhruv Sharma.
Forthcoming a. “The Economic Impact of Investment Provision: Evidence from
Indonesia.” Working paper, World Bank.

Calì, Massimiliano, Aufa Doarest, and Giorgia Presidente. Forthcoming b. “Foreign Entry and
Domestic Performance: Evidence from Indonesia.” Working paper, World Bank.

Calì, Massimiliano, Taufik Hidayat, and Claire H. Hollweg. 2018. “What is behind Labor Mobility
Costs? Evidence from Indonesia.” Background paper, Pathways to Middle-Class Jobs
in Indonesia, World Bank, Jakarta.

Campos, Francisco, Michael Frese, Markus Goldstein, Leonardo Iacovone, Hillary C. Johnson,
David McKenzie, and Mona Mensmann. 2017. “Teaching Personal Initiative Beats
Traditional Training in Boosting Small Business in West Africa. Science 357 (6357):
1287–90. https://doi.org/10.1126/science.aan5329.

Castley, Robert. 1997. Korea’s Economic Miracle: The Crucial Role of Japan. Basingstoke, UK:
Palgrave Macmillan.

Chang, Philip Han, Rachita Daga, Gautam Vishram Mehta, and Massimiliano Calì. 2019.

Creating Markets in Indonesia. Unlocking the Dynamism of the Indonesian Private
Sector: A Country Private Sector Diagnostic. Washington, DC: World Bank Group.

Cirera, Xavier, Jaime Frias, Justin Hill, and Yanchao Li. 2020. A Practitioner’s Guide to
Innovation Policy: Instruments to Build Firm Capabilities and Accelerate Technological
Catch-Up in Developing Countries. Washington, DC: World Bank. https://
openknowledge.worldbank.org/handle/10986/33269 License: CC BY 3.0 IGO.

Cornia, Giovanni A. 2016. “An Econometric Analysis of the Bifurcation of Within-Country

Inequality Trends in Sub-Saharan Africa, 1990–2011.” UNDP Africa Report No.
267781, United Nations Development Programme (UNDP).

June 2021 Pathways to Middle-Class Jobs in Indonesia

 References 175

Cruz, Marcio, Gaurav Nayyar, Gerhard Toews, and Pierre-Louis Vézina. 2018. “FDI and the Skill
Premium: Evidence from Emerging Economies.” Policy Research Working Paper
8613, World Bank, Washington, DC.

Cunningham, Wendy. 2018. Vietnam Future Jobs: Leveraging Mega-Trends for Greater
Prosperity. World Bank Study. Washington, DC: World Bank.

Cunningham, Wendy, Sodeth Ly, and Marong Chea. 2018. Cambodia’s Future Jobs: The
Economist Version. A World Bank Study. World Bank Publications, World Bank Group.

Cunningham, Wendy, and William F. Maloney. 2001. “Heterogeneity among Mexico’s

Microenterprises.” Economic Development and Cultural Change 50 (1): 131–56.

Cunningham, Wendy, and Ekki Syamsulhakim. Forthcoming. “Household Enterprise in

Indonesia.” Background paper, Pathways to Middle-Class Jobs in Indonesia, World
Bank, Jakarta.

Davis, Steven J., and John C. Haltiwanger. 1992. “Gross Job Creation, Gross Job Destruction,
and Employment Reallocation.” The Quarterly Journal of Economics 107 (3): 819–63.

Davis, Steven J., John C. Haltiwanger, and Scott Schuh. 1996. Job Creation and Destruction.
Cambridge, MA: MIT Press.

Deloitte. 2015. SMEs Powering Indonesia’s Success: The Connected Archipelago’s Growth Engine.
Sydney, Australia: Deloitte Access Economics.

Dinga, Marian, and Daniel Münich. 2010. “The Impact of Territorially Concentrated FDI on
Local Labor Markets: Evidence from the Czech Republic.” Labour Economics 17 (2):
354–67.

Drexler, Alejandro, Greg Fischer, and Antoinette Schoar. 2014. “Keeping It Simple: Financial
Literacy and Rules of Thumb.” American Economic Journal: Applied Economics 6 (2):
1–31.

Duening, Thomas N. 2010. “Five Minds for the Entrepreneurial Future: Cognitive
Skills as the Intellectual Foundation for Next Generation Entrepreneurship
Curricula.” The Journal of Entrepreneurship 19 (1): 1–22. https://doi.
org/10.1177/097135570901900101.

Duggan, Victor, Sjamsu Rahardja, and Gonzalo Varela. 2013. “Service Sector Reform and
Manufacturing Productivity: Evidence from Indonesia.” Policy Research Working
Paper 6349, World Bank, Washington, DC.

Dutz, Mark A. 2018. Jobs and Growth: Brazil’s Productivity Agenda. Washington, DC: World
Bank.

Pathways to Middle-Class Jobs in Indonesia June 2021

176 References

Escobar, Octavio, and Henning Mühlen. 2018. “The Role of FDI in Structural Change:
Evidence from Mexico.” Hohenheim Discussion Papers in Business, Economics and
Social Sciences No. 22-2018. Stuttgart: University of Hohenheim.

Feenstra, Robert, and Gordon Hanson. 1997. “Foreign Direct Investment and Relative Wages:
Evidence from Mexico’s Maquiladoras.” Journal of International Economics 42 (3–4):
371–93.

Ferro, Esteban, and Smita Kuriakose. 2018. “Indonesia: High-Growth Firms.” Background
paper, High Growth Firms, World Bank, Washington, DC.

Figini, Paolo, and Holger Görg. 2011. “Does Foreign Direct Investment Affect Wage
Inequality? An Empirical Investigation.” The World Economy 34 (9): 1455–75.

GDP Venture. 2020. “The Impact of the COVID-19 Pandemic.” (PowerPoint presentation.)

Gil Sander, Frederico, and Pui Shen Yoong. Forthcoming. “Structural Transformation and
Labor Productivity in Indonesia: Where Are All the Good Jobs?” Background paper,
Pathways to Middle-Class Jobs in Indonesia, World Bank, Jakarta.

Gomez-Mera, Laura, and Claire H. Hollweg. 2018. “Firm Performance and Constraints in
Indonesia.” Unpublished paper, World Bank.

Google, Temasek, and Bain. 2020. e-Conomy SEA 2020: At Full Velocity: Resilient and Racing
Ahead. https://storage.googleapis.com/gweb-economy-sea.appspot.com/assets/
pdf/e-Conomy_SEA_2020_Report.pdf

Grover Goswami, Arti, Denis Medvedev, and Ellen Olafsen. 2019. High-Growth Firms: Facts,
Fiction, and Policy Options for Emerging Economies. Washington, DC: World Bank.
doi:10.1596/978-1-4648-1368-9. License: Creative Commons Attribution CC BY 3.0
IGO.

Guilford, Gwynn. 2018. “The Epic Mistake about Manufacturing That’s Cost Americans
Millions of Jobs.” Quartz, May 3. https://qz.com/1269172/the-epic-mistake-about-
manufacturing-thats-cost-americans-millions-of-jobs/.

Hale, Galina, and Mingzhi Xu. 2016. “FDI Effects on the Labor Market of Host Countries.”
Working Paper 2016-25. Federal Reserve Bank of San Francisco, San Francisco, CA.

Hallward-Driemeier, Mary, and Gaurav Nayyar. 2018. Trouble in the Making? The Future of
Manufacturing-Led Development. Washington, DC: World Bank. doi:10.1596/978-1-
4648-1174-6. License: Creative Commons Attribution CC BY 3.0 IGO.

Haltiwanger, John C., Ron S. Jamin, and Javier Miranda. 2010. “Who Creates Jobs? Small
vs. Large vs. Young.” NBER Working Paper 16300, National Bureau of Economic
Research, Cambridge, MA.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 References 177

Head, Keith, and John Ries. 2002. “Offshore Production and Skill Upgrading by Japanese
Manufacturing Firms.” Journal of International Economics 58 (1): 81–105.

Javorcik, Beata, Fitria Fitriani, and Leonardo Iacovone. 2012. Productivity Performance in
Indonesia’s Manufacturing Sector. World Bank policy note, World Bank, Jakarta.

Karlsson, Sune, Nannan Lundin, Fredrik Sjöholm, and Ping He. 2009. “Foreign Firms and
Chinese Employment.” The World Economy 32 (1): 178–201.

Khochan, Thomas A. 2019. “Breakdown of the Old Social Contract and Starting Assumptions
for a New Social Contract.” Original source of data: The Economic Policy Institute.

Kochan, Thomas A., and Lee Dyer. 2019a. “Demise of the Old Social Contract and the
Challenges and Opportunities That Lie Ahead.” Reading material from week three
of Shaping Work of the Future, a course of study offered by MITx, MIT, Cambridge,
MA.

Kochan, Thomas A., and Lee Dyer. 2019b. “What Was the Postwar Social Contract, Where
Did It Come From, and What Made It Work for Three Decades?” Reading material
from week two of Shaping Work of the Future, a course of study offered by MITx,
Massachusetts Institute of Technology (MIT), Cambridge, MA.

Kuncoro, Ari. 2019. “Trends in the Manufacturing Sector under the Jokowi Presidency:
Legacies of Past Administrations.” In The Indonesian Economy in Transition: Policy
Challenges in the Jokowi Era and Beyond, edited by Hal Hill and Siwage Dharma
Negara. Singapore: ISEAS.

Kusumawardhani, Niken, Daniel Suryadarma, Luca Tiberti, and Veto Tyas Indrio. 2019.
“What Skills Lead to Entrepreneurial Success? Evidence from Non-Farm-Household
Enterprises in Indonesia.” Working Paper 2019-14, Partnership for Economic Policy.

Leamer, Edward. 1998. “In Search of Stolper-Samuelson Effects on U.S. Wages.” In Exports,
Imports and the American Worker, ed. Susan M. Collins, 141–214. Washington, DC:
Brookings Institution Press.

Lee, Jong-Wha, and Dainn Wie. 2015. “Technological Change, Skill Demand, and Wage
Inequality: Evidence from Indonesia.” World Development 67 (C): 238–50.

Li, Yue, and Martín Rama. 2015. “Firm Dynamics, Productivity Growth, and Job Creation in
Developing Countries: The Role of Micro- and Small Enterprises.” The World Bank
Research Observer 30 (1): 3–38. Published by Oxford University Press on behalf of
the World Bank.

Lipsey, Robert E., and Fredrik Sjöholm. 2004. “Foreign Direct Investment, Education, and
Wages in Indonesian Manufacturing.” Journal of Development Economics 73 (1):
415–22.

Pathways to Middle-Class Jobs in Indonesia June 2021

178 References

Luo, Rui. 2017. “Skill Premium and Technological Change in the Very Long Run: 1300–1914.”
Discussion Papers in Economics 17/09, University of Leicester.

Luther, Gregory C., Joko Mariyono, , Raden M. Purnagunawan, Ben Satriatna, and Martin
Siyaranamual. 2018. “Impacts of Farmer Field Schools on Productivity of Vegetable
Farming in Indonesia.” Natural Resource Forum 42 (2): 71–82. doi:10.1111/1477-
8947.12144.

Mangla, Puneet, Ashish Agarwal, and Pulak M. Pandey. 2015. “Inventory Management
Performance in MSMEs: What Factors do Influence Them?” International Journal of
Advanced Manufacturing Systems 15 (2): 45–58.

Manning, Chris, and Haryo Aswicahyono. 2012. “Trade and Employment in Services: The
Case of Indonesia.” Employment Working Paper No. 132. International Labour Office
(ILO), Geneva.

Moorena, Lolita, Mikaela Rabb, Aliyyah Rusdinar, Simone Schaner, Caroline Tangoren, and
Natalie Theys. N.d. Towards Inclusive Digital Finance in Indonesia: A Literature Review
and Landscape Analysis. https://www.povertyactionlab.org/sites/default/files/
review-paper/ifii-whitepaper.pdf.

OECD (Organisation for Economic Co-operation and Development). 2018. SME

and Entrepreneurship Policy in Indonesia 2018. OECD Studies on SMEs and
Entrepreneurship. Paris: OECD Publishing.

OECD and ERIA (Economic Research Institute for ASEAN and East Asia). 2018. SME Policy
Index: ASEAN 2018: Boosting Competitiveness and Inclusive Growth. Paris: OECD
Publishing; Jakarta: ERIA.

OJK (Otoritas Jasa Keuangan). N.d. “Laku Pandai: Sahabat Masyarakat: Layanan Keuangan
Tanpa Kantor Dalam Rangka Keuangan Inklusif.” Otoritas Jasa Keuangan. https://
www.ojk.go.id/id/Pages/Laku-Pandai.aspx.

Pasquier-Doumer, Laure, Xavier Oudin, and Nguyen Thang. 2017. “The Importance of
Household Business and the Informal Sector for Inclusive Growth in Vietnam.” Thế
Giới Publishers, 9786047733453. hal-01827323.

Patunru, Arianto. 2019. “Rising Economic Nationalism in Indonesia.” In The Indonesian

Economy in Transition: Policy Challenges in the Jokowi Era and Beyond, editd by Hal Hill
and Siwage Dharma Negara. Singapore: ISEAS.

Peluffo, Adriana. 2015. “Foreign Direct Investment, Productivity, Demand for Skilled Labour
and Wage Inequality: An Analysis of Uruguay.” World Economy 38 (6): 962–83.

Sea Limited. 2020. “Sea Limited (SE) Q3 2020 Earnings Call Transcript.” Motley Fool. https://
www.fool.com/earnings/call-transcripts/2020/11/18/sea-limited-se-q3-2020-
earnings-call-transcript/.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 References 179

Shepherd, Ben, and Moekti Soejachmoen. 2017. “Why is Indonesia Left Behind in Regional
Production Networks? In The Indonesian Economy: Trade and Industrial Policies, ed.
Lili Yan Ing, Gordon H. Hanson, and Sri Mulyani Indrawati. Routledge–ERIA Studies in
Development Economics. London: Taylor and Francis.

Sjöholm, Fredrik, Robert E. Lipsey, and Jing Sun. 2010 “Foreign Ownership and Employment
Growth in Indonesian Manufacturing.” IFN Working Paper No. 831, Research
Institute of Industrial Economics, Stockholm.

Soejachmoen, Moekti Prasetiani. 2012. “Why is Indonesia Left Behind in Global Production
Networks?” PhD thesis, The Australian National University.

Steenbergen, Victor, and Hebous, Sarah. 2020. “The Effect of FDI on Indonesia’s Jobs,
Wages and Structural Transformation.” Background paper, Pathways to Middle-Class
Jobs in Indonesia, World Bank, Jakarta.

Steenbergen, Victor, and Trang, Thu Tran, 2020. “The Distributional Effects of FDI: Evidence
from Ethiopia, Vietnam, and Turkey.” In Global Investment Competitiveness Report
2019/2020: Rebuilding Investor Confidence in Times of Uncertainty. Washington, DC:
World Bank Publications.

Takii, Sadayuki, 2005. “Productivity Spillovers and Characteristics of Foreign Multinational

Plants in Indonesian Manufacturing 1990–1995.” Journal of Development Economics
76 (2): 521–42.

Tambunan, Tulus. 2019. “Recent Evidence of the Development of Micro, Small and Medium
Enterprises in Indonesia.” Journal of Global Entrepreneurship Research 9:18. https://
doi.org/10.1186/s40497-018-0140-4.

Tang, Heiwai, and Zhang, Yifan. 2017. “Do Multinationals Transfer Culture? Evidence on
Female Employment in China.” CESifo Working Paper Series 6295, CESifo, Munich.

Ver Beek, Kurt Alan, 2001. “Maquiladoras: Exploitation or emancipation? An overview of the
situation of maquiladora workers in Honduras.” World Development 29 (9): 1553–67.

Waldkirch, Andreas, Peter Nunnenkamp, and Jose Eduardo Alatorre Bremont. 2009.
“Employment Effects of FDI in Mexico’s Non-Maquiladora Manufacturing.” Journal of
Development Studies 45 (7): 1165–83.

World Bank. 2014. Indonesia: Avoiding the Trap. Development Policy Review. Jakarta: World
Bank.

World Bank. 2019a. “Aspiring Indonesia: Expanding the Middle Class.” Background paper,
Development Policy Review 2019, World Bank, Jakarta.

World Bank. 2019b. Closing the Human Capital Gap. Development Policy Review. World Bank,
Jakarta.

Pathways to Middle-Class Jobs in Indonesia June 2021

180 References

World Bank. 2020a. “Growth prospects for the self-employed / micro firms in Indonesia.”
Presentation as part of a series of work on firm productivity, innovation, and
entrepreneurship in Indonesia. Findings presented in Indonesia Public Expenditure
Review 2020: Spending for Better Results.

World Bank. 2020b. Indonesia Systematic Country Diagnostic. Indonesia: Eliminating Poverty,
Bringing Economic Security to All. World Bank.

World Bank. 2020c. World Development Report 2020: Trading for Development in the Age of
Global Value Chains. Washington, Washington, DC: World Bank. doi:10.1596/978-1-
4648-1457-0. License: Creative Commons Attribution CC BY 3.0 IGO.

World Bank. 2020d. “Covid-19 Impact on Firms in Indonesia: Results from the Covid-19
Business Pulse Survey, Round 2 (October 2020).” World Bank.

World Bank. 2021a. Uneven Recovery. East Asia and Pacific Economic Update. April.
Washington, DC: World Bank. doi: 10.1596/978-1-4648-1702-1. License: Creative
Commons Attribution CC BY 3.0 IGO.

World Bank and Grab. 2021. “Does Access to GrabKios & GrabFood Services Improve the
Performance of Micro Firms?” Internal document, World Bank, Jakarta.

World Bank. Forthcoming. Public Expenditure Review for Innovation and Business Support
Program in Indonesia. World Bank internal document.

World Bank, and Shopee. 2021b. “Insights from the Covid-19 Digital Merchant Survey 2021.”
World Bank, Jakarta.

June 2021 Pathways to Middle-Class Jobs in Indonesia

Enterprise-Level Dynamics of Job
References
Creation 181

Pathways to Middle-Class Jobs in Indonesia June 2021

182 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

Chapter 5

Workers, Jobs,
and Barriers
to Entry to
Middle-Class Jobs 

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 183

In 2018, there were 124 million working youth and adults in Indonesia, which has
one of the largest workforces in the world. Two in every three people aged 15–64 were
either working (65.6 percent) or looking for work (3.3 percent), on par with countries of
the same level of development as Indonesia. The labor force participation rate has crept up
slowly from 66.9 percent in 2007 to 67.2 percent in 2018, as unemployment rates have fallen
from 9.1 percent to 5.4 percent over the same period.

However, in 2018,  only 13 million  income earners out of 85 million held “middle-
class” jobs.1, 2 That is, only 15 percent of income earners—comprised of wage employees,
casual workers, and the self-employed—were earning income that would allow a family of
four to purchase a middle-class  way of life.  Using a more restrictive definition of middle-
class job, namely that a worker earns above the threshold for middle-class earnings, enjoys
full social benefits,3  and holds an indefinite-term employment  contract,4  only  3.5 million
wage employees (out of the 85 million income earners) can be considered to hold a middle-
class job. 

1 See Chapter 1 for a discussion of the definitions of “middle-class job” that can be measured with Sakernas data. 
2 While there are 124 million workers, Sakernas provides income data only on the 85 million workers who are self-
employed, wage employees, or casual workers, and therefore only allows middle-class status to be assigned to these types
of workers. This report refers to these workers as “income earners.” There is no earnings information for employers or
unpaid workers. By definition, there is no information on social benefits or contract status for casual workers and the
self-employed.
3 The job-associated benefits that are included in the Sakernas are: health insurance, work accident insurance, death
benefits, old-age benefits (nongovernment), pension benefits (government), and/or paid annual leave / sick leave /
maternity leave.
4 The Sakernas allows for four categories of contract: indefinite term, definite term, verbal, and none.

Pathways to Middle-Class Jobs in Indonesia June 2021

184 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

This chapter will explore: (i) who holds middle-class jobs; (ii) barriers to obtaining a middle-class

job; (iii) additional barriers faced by women and youth; and (iv) policy recommendations.5

5.1 Who Are Middle-Class Workers and How Do They Compare

to the Wider Workforce? 
To understand how to help more workers transition to middle-class jobs, the first step is to
identify the differences between those who are already in those kinds of jobs and those
who are not. This section presents  the profile of  the overall workforce and compares  it
to the profile of workers in middle-class jobs. 

5.1.1 Men dominate all jobs and middle-class jobs 

Indonesia’s workforce is dominated by men; this has changed little over the past 20
years (Figure 5.1). While there are equal numbers of working-age men and women, only 39
percent of workers are women. Approximately half of working-age women (aged 15–64) hold
a job, as compared to 80 percent of men, with significant variation by province.6  While
Indonesia’s female labor force participation rate is on par with other countries at its level of
development, it is significantly below other developing countries in the East Asia and Pacific
region, where more than 58 percent of women work.

More curious, though, is that  female labor force participation rates  have not
moved in twenty years  despite improvements in the educational attainment
of women, falling fertility rates, and rapid urbanization.  This apparent stagnation
reflects an improvement in aspects of women’s lives that are correlated with lower work
tendencies. Women’s educational attainment is rapidly increasing. Most now have at least
some secondary school education, which global evidence shows is the level of education at
which labor force participation dips due to the opportunity cost of working.7  Fertility
rates continue to fall slowly—from 2.5 children in 2000 to 2.3 in 2018—and early marriages
declined from 24 percent in 2003 to 13 percent in 2013. However, women continued to drop
out of the labor force at a higher rate than men during prime childbearing years (Figure
5.2).8, 9 Urbanization makes working harder since women have to go out and find a job rather
than working on the family farm.10 In addition, they have to contend with the high cost of

5 To simplify definitions and comparisons, this section defines a middle-class job as one that meets the  threshold of
middle-class earnings (see Chapter 1).
6 The rates at which women participate in the labor force differ significantly from region to region. In Bali, Yogyakarta,
and Papua, more than 70 percent of working-age women either hold a job or are seeking a job. In contrast, 45 percent of
women in West Java are in the labor force (Lain, Alatas, and Setyonaluri 2020).
7 In other words, the earnings from working are insufficient to cover the costs incurred by working, such as the cost
of childcare or transportation.
8 In 2018, one child under the age of three in the household reduced the likelihood of a woman working by 7 percent and
increased the likelihood of a man working by 2 percent. A second child further reduced the likelihood of a woman
working by 3 percent and increased slightly the likelihood of a man working. The presence of informal childcare,
provided by older inactive men or women in the household, increased women’s participation by nearly 6 percent and
men’s by nearly 2 percent (Lain, Alatas, and Setyonaluri 2020).
9 Approximately one-third of women who were not working and who experienced a job termination in 2018 attributed
their exiting of the labor market to domestic work. A further 13 percent suggested pregnancy was the main reason (Lain,
Alatas, and Setyonaluri 2020). A qualitative study by Artiawati (2007) that focused on teachers confirms that work-
family conflict escalates when children are young. Indeed, the likelihood of working becomes less negative as the age of
children in the household increases (Lain, Alatas, and Setyonaluri 2020).
10 Provinces in which a larger share of the workforce is employed in agriculture tend to have higher female labor force

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 185

transportation11  and  weak laws (and practices) for women’s safety in work spaces.12  The
economic transformation toward the service industry, where women are overrepresented
and  hold more of the  professional and technical jobs than they do in other sectors, and
the rise of the (digital) gig economy and its potential for part-time work and working from
home, may offer greater opportunities to women.13 

Figure 5.1 Labor Force Participation Rates by Gender Figure 5.2 More Women than Men Drop Out of the
Have Not Changed in More Than 20 Years  Labor Force during Prime Childbearing Years 
Labor force participation rate by gender in Indonesia Proportion of all labor market dropouts at each age cohort 

Sources: Sakernas 2007–18; World Bank staff calculations Source: Lain, Alatas, and Setyonaluri 2020 
Note: Sakernas is the National Labor Force Survey (Survei Angkatan Note: The sample is women and men who were out of the labor
Kerja Nasional). force in 2018 but who had worked in the year prior to the survey. 

While women hold 35 percent of jobs that pay an income, they only hold 28.6 percent
of middle-class jobs  (Figure 5.3a). About  12.3  percent of the  30  million  income-
earning  women have jobs that earn at least a  middle-class  wage  (3.6  million  women), as
compared to 16.5 percent of the 55 million income-earning men (9.1 million) (Figure 5.3b). The
underrepresentation of women in middle-class jobs is partly due to the large number of
women who work but are unpaid. That said, women’s employment in middle-class jobs has
grown more quickly than men’s: increasing by 2.8 percentage points over the period 2001–
18 as compared to 0.7 percentage points for men.14 

participation (Lain, Alatas, and Setyonaluri 2020). This aligns with existing evidence at the desa/kelurahan (rural/
urban villages) level from Cameron, Contreras Suarez, and Rowell (2019). 
11 A survey on Transjakarta’s new bus route and female employment found that around 15 percent of women with
young children quit their job because of a long commute (Witoelar et al. 2017).
12 Aeberhard-Hodges 2020
13 Currently, women are underrepresented among gig workers. According to  Sakernas  2019, only 1.1 percent
of  working  women hold gig jobs, as compared to 2.2  percent of  working  men.  Gig jobs are not limited to  young
workers. Only 10 percent of female gig workers are aged 15–24. 
14 Women’s share of income-earning jobs is 36 percent, while their share of all jobs is 39 percent. The shift is due
to the high number of women (22.6 percent) who are unpaid workers—and thus omitted from the “income-earning”
calculation—as compared to only 5.7 percent of working men

Pathways to Middle-Class Jobs in Indonesia June 2021

186 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

Figure 5.3 Less “Vulnerable” Groups Are Overrepresented in Middle-Class Jobs 

a. Distribution of income-earning jobs and distribution of middle-class jobs across defined demographic groups

b. Share of income-earning workers who hold middle-class jobs, by characteristic 

Sources: Sakernas 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

5.1.2 The Indonesian workforce is aging and middle-class workers are also older 

The Indonesian workforce is aging.  As of 2018, the average Indonesian worker was
40.5 years old, two years older than in 2007. Prime-age workers (aged 25–54) comprised
70 percent of the workforce, while youth comprised only 13 percent (Figure 5.4). This is 4
percentage points less than the youth share of the workforce in 2007. In fact, the number
of workers in each age group has increased between 2007 and 2018, with the exception
of youth. In 2018, 200,000  fewer  youth were working as compared to 2007  (Figure 5.5).
The aging tendency of the workforce is due to several factors: delayed entrance to the

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 187

labor force in favor of more schooling,15 later retirement of older workers, and falling birth
rates leading to equalizing cohort sizes. United Nations projections show that the number
of youth and children in Indonesia has stabilized and will begin to decline in 2025, leading to
further aging of the workforce.16

Figure 5.4 Indonesia’s Labor Force Is of Prime Figure 5.5 The Number of Workers is Growing in All Age
Working Age  Groups but Declining Among Youth 
Distribution of the employed in 2018 Employment in 2007 and 2018

Sources: Sakernas 2018; World Bank staff calculations  Sources: Sakernas 2007 and 2018; World Bank staff calculations
Note: Sakernas is the National Labor Force Survey (Survei Note: The declining number of youth (15–24) is not visible in the graph due
Angkatan Kerja Nasional). to the large magnitude of workers. Sakernas is the National Labor Force
Survey (Survei Angkatan Kerja Nasional).

Prime-age  income earners  hold middle-class jobs.  More than 14 percent of  income
earners in the age ranges 25–34, 35–44, and 45–54  earn at least a middle-class income,
as compared to 12.5 percent of the oldest income earners and only 6.7 percent of the
youngest  (Figure 5.3b).  The cohort  groups 35–44 and 45–54 are overrepresented in
middle-class jobs, relative to their  share of total jobs. For example, the 45–54 age group
holds 19.6 percent of income-earning jobs but 26.8 percent of middle-class jobs. The youngest
workers (age 15–24)  only hold 6.4 percent of middle-class jobs, though they account for
14.4 percent of all workers. Older workers are also underrepresented (Figure 5.3a).

15 The share of youth in school increased from 30 to 36 percent over the period 2007–18 (Sakernas). All other activities—
working, searching for work, housekeeping, and discouraged worker—declined by approximately one percentage point over
the period. 
16 United Nations, Department of Economic and Social Affairs, Population Division 2019

Pathways to Middle-Class Jobs in Indonesia June 2021

188 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

Box 5.1 Will COVID-19 Hinder Youth from Attaining Middle-Class Jobs? 
The COVID-19 crisis differs from the Asian financial crisis and the global financial crisis of 2008 in fundamental ways. Most
importantly, it is an economic crisis spurred by a health pandemic and requires policymakers to consider trade-offs between
public health and economic recovery. The highly contagious and potentially deadly disease has not only debilitated the
demand side of jobs and led to firm closures and domestic and external demand shocks, for example, but has also had
significant effects on the supply side of jobs through illnesses, deaths, anxiety, fear, and depression. Illness, mobility
restrictions, home isolation and other containment measures have reduced work productivity and disrupted the education
and learning processes of the future workforce, especially of those students who are not connected digitally. While some
of the demand-side shocks can be reversed once the pandemic is controlled and economic activity resumes, the supply-side
shocks may have longer-term impact or may be permanent.

The COVID-19 period and its aftermath is a bad time to leave school. While the modern world has never experienced a joint
health-economic crisis of the severity of the COVID-19 pandemic, lessons from economic crises of the past several decades
can provide some predictions about how the current crisis may affect the long-term labor market prospects of today’s
youth.

Entering the labor market during an economic downturn has long-lasting “scarring” consequences for the new worker.
Reviews of international literaturea have concluded that youth who enter the labor force during an economic downturn
earn lower wages for many decades as compared to those entering in economically robust years. For example, youth who
graduate from a Canadian university during an economic recession in Canada can expect their wages to be 9 percent lower
than average in their first job and 5 percent lower than average five years later; the gap is only closed ten years after
graduation, mostly through frequent job switching.b Similarly, wages are 10–21 percent lower among UK graduates,c and 2.5
percent lower among US graduatesd who leave school for work during a recession. Similar results have also been found for
Mexican youth.e

The scarring also leads to greater difficulty in finding jobs later in life. Low-skilled workers who enter the labor force during
an economic downturn experience unemployment rates one to two percentage points higher for 16 years after leaving
school in Argentinaf and in Norway.g More rapid entry, perhaps through self-employment, can still have scarring effects. For
example, Indonesian youth who are self-employed are more likely to still be self-employed 7 and 14 years after the initial
observation as compared to youth who began in other types of work.h Unemployment experienced at an early age has also
been found to be associated later in life with illness, stress, depression, and lower life expectancy.i

This pandemic has impacted the seven million new graduates—part of the so-called “Generation COVID-19”j—who are ready
to enter the labor force.k in 2020, about 300,000 fewer youth entered the labor force compared to 2019. A closer look
at the impacts of the pandemic on youth suggests they are more likely to mix school and employment, and that there
has been an increase in youth idleness (those not in employment, education, or training).l An increase in employment
rates among youth was especially concentrated among those of high-school age (15–18), who were likely to be occupied
as unpaid family workers. This pattern holds even when we look at youth working full time (at least 20 hours a week):
the increase was especially concentrated among those of high-school age. The effects are similar for boys and girls.
This may mean interrupted learning (combining schooling with full-time work) for these children of high-school age.
Moreover, new graduates starting work in 2020 earned less than their 2019 cohort. A somewhat extensive literature
finds that entering the labor force during economic downturn comes with several issues: it is harder to find a job and
this then becomes a vicious circle, partly because people lose skills when they are not working; it also leads to many
years of lower wages. Job losses and reduced household incomes could also affect human capital investment in children.m

The potential long-term costs of this scarring of the labor force highlight the need to proactively integrate new jobseekers
into jobs as economic activity resumes. The wounds will be more difficult to heal later.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 189

Box 5.1 continued

a Naidoo, Packard, and Auwalin 2015; Duryea 2012

b Oreopoulos, von Wachter, and Heisz 2006
c Gregg and Tominey 2005; Gregory and Jukes 2001
d Kahn 2010
e Székely and Karver 2015
f Duryea 2012
g Bratsberg, Raaum, and Røed 2006
h Naidoo, Packard, and Auwalin 2015
i Bell and Blanchflower 2010
j World Bank 2020a
k Pradana, Hadiwidjaja, and Posadas 2021
l Halim, Hambali, and Purnamasari. Forthcoming
m World Bank 2020a

5.1.3 The skill level of the Indonesian workforce is lower than education

levels might suggest, but a higher level of education is still correlated with holding
a middle-class job 
The educational level of the workforce—measured by level of education completed
by those aged 15–64 who are not in school—rapidly increased between 2007 and
2018.  The number of workers who had completed senior high school doubled over the
period, reaching 36 million in 2018, and the number who had completed university increased
from 6 million to 15 million (Figure 5.6). That said, Indonesia’s education profile in 2018 was
similar to that of the United States in 1940. Nearly 60 percent of workers have less than
a senior high school education  (Figure 5.7).  This is worrisome for two reasons.  First, the
quality of education is likely much lower than it should be for the attainment rate. The
World Bank’s  Human  Capital  Index 2020 estimates that an Indonesian child entering the
education system today can expect to complete 12.4 years of schooling, but will only
learn the equivalent of 7.8 years;17 in other words, the rapid increase in attainment is not
necessarily leading to equivalently higher skills.18 Second, the stock of workers with low levels
of education is so large that flows of more educated youth into the labor force will only
slowly shift the education profile. Over the past 10 years, the average level of education of
the workforce increased by one year (Figure 5.6). 

17 World Bank. N.d.

18 OECD 2016

Pathways to Middle-Class Jobs in Indonesia June 2021

190 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

Figure 5.6 The Workforce Is Slowly Acquiring Years Figure 5.7 Only 41 percent of total workers has
of Schooling acquired an upper secondary education or more
Share of workers with each level of education completed and Educational distribution of workers, 2018
average years of education of the labor force, 2007–18

Sources: Sakernas 2007–18; World Bank staff calculations Sources: Sakernas 2018; World Bank staff calculations
Note: Sakernas is the National Labor Force Survey Note: Sakernas is the National Labor Force Survey (Survei Angkatan
(Survei Angkatan Kerja Nasional). Kerja Nasional).

Similar trends emerge by gender: the educational attainment of both men and
women has increased over time, though women lag behind men. Girls and boys have
similar school enrolment rates, hovering near 90 percent for primary enrolment, 70 percent
for secondary school enrolment, and  30 percent for tertiary school enrolment. Notably,
women’s enrolment in tertiary passed that of men in 2012, with nearly 40 percent of young
women of tertiary education age enrolled in training or university, compared to 33 percent
of young men. However, the educational attainment rate for adults (aged 25 or older) gives
men the advantage;  rates of attainment of upper secondary or tertiary remain about
7 percentage points lower for women than men. This gap has been constant since 2006.19

More educated individuals hold a disproportionate share of middle-class jobs. For

example, workers with a primary education or less hold 34.4 percent of income-earning jobs
but only 8 percent of middle-class jobs. In contrast, workers with a postsecondary education
hold  15.6  percent of all jobs but 47 percent of middle-class jobs  (Figure 5.3a).  A higher
education does not guarantee a middle-class job, though.  Only  40  percent of those with
a postsecondary education hold a middle-class job (Figure 5.3b). 

19 Lain, Alatas, and Setyonaluri 2020

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 191

5.1.4 Jobs, especially middle-class jobs, are clustered in urban zones 

Although jobs are relatively equally distributed between urban and rural
zones,  better jobs are clustered in urban areas. Approximately 66 million  jobs are
in urban zones compared to 57 million in rural. However, a large share of rural jobs are
unpaid  roles on family farms. If  only income-earning jobs are considered, 62 percent are
in urban areas.  However, 82 percent of middle-class jobs are in urban areas (Figure 5.3a).
Only 20 percent of the urban population earns a middle-class  income, partly due to the
high incidence of low value-added services in urban zones, but this is far higher than the 7
percent of rural dwellers (Figure 5.3b). 

The  clustering of middle-class jobs in urban areas is partly due to  educational
attainment and agglomeration effects. A one-year increase in the average educational
level in Indonesia leads to a 10 percent increase in high-skilled wages and a 2 percent increase
in low-skilled wages in urban areas, but no increase occurs in rural zones. This wage premium
is driven by the direct effect of skills on productivity and by the indirect effects of worker
proximity as workers learn from each other. For example, a doubling of population density
results in a 10 to 13 percent increase in productivity in urban Indonesia, which translates to
higher wages. Increased density does not affect rural productivity. So, even though urban
jobs are mostly in low value-added services, the impacts of education and agglomerations
result in higher productivity and wages than in less densely populated rural zones. There are
few noticeable job spillovers from urban to rural zones.20 

Rural households outside of Java and Bali are at risk of losing ground. Rural households
are increasing incomes by diversifying their sources of income. While 53 percent of rural
households engage in agriculture, only 35 percent of rural household income is derived from
this source. The rest is derived from off-farm jobs and cash transfers. Wage employment
in services and industry has increased rural earnings in Java and Bali by 62 percent since
2017, but the absence of such jobs in other rural zones has resulted in a stagnation in labor
earnings since 2017. For example, 83 percent of new service jobs in Papua were in low-paying
community and social services. This jobs gap may expand for rural populations outside of Java
and Bali due to the increasing incidence of natural disasters and weather variability brought
on by climate change. Given the low incidence of internal migration, these populations will
fall even further from middle-class jobs.21 

20 Roberts, Gil Sander, and Tiwari 2019

21 World Bank 2020b

Pathways to Middle-Class Jobs in Indonesia June 2021

192 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

5.1.5  While having a moderate disability has little effect on work  status, it has a
significant effect on middle-class-job status 
Having a  severe  disability reduces engagement in the labor force.  People with
a moderate disability  represent  a greater share  of the labor force (6.9 percent)  and
workers  (7.1  percent) than expected given that they make up only 6.2  percent of the
working-age population.22  In contrast, those with a severe disability represent  0.9
percent of the working population but only 0.3 percent of labor force participants or
workers. Those with any physical disability are less likely to hold a job as a wage employee
(3.9 percent)  than their work patterns would suggest, instead being overrepresented in
self-employment (10.9 percent), as employer (11.5 percent), casual worker (7.3 percent), and
unpaid worker (7.2 percent) (Table 5.1).

Workers with disabilities have less access to middle-class jobs.  Eleven percent of
workers with a moderate disability and 7 percent of those with a severe disability hold a
middle-class job (Figure 5.3b). However, while 6.1 percent of workers holding income-earning
jobs have a moderate disability—on par with their share of the working-age population—only
4.7 percent of middle-class jobs go to those with a moderate disability (Figure 5.3a). People
with a severe disability  represent only 0.3 percent of income earners and 0.1 percent of
middle-class jobs.

Table 5.1 Disability Status and Employment  

Engagement in the labor force (%) Type of job (%)

Disability Waged
intensity Working Labor worker with Self- Casual Unpaid
Employment Employer Employee
age force a middle- employed worker worker
class job
None 92.9 92.8 92.6 95.2 89.2 88.5 96.1 92.7 92.8
Moderate 6.2 6.9 7.1 4.7 10.4 11.1 3.8 7.0 6.8
Severe 0.9 0.3 0.3 0.1 0.5 0.4 0.1 0.3 0.4

Sources: Sakernas 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional). Due to rounding, percentages in this table may not total
100 percent.

5.2 Barriers to Entry to Middle-Class Jobs 

If the macroeconomy and firms play their part to create middle-class job openings, workers
will need to step up to effectively fill those jobs. To acquire a middle-class job, workers should
prepare for and seek one.  Chapter 2  gave insights into  the types of jobs that are more
associated with middle-class earnings. All sectors except agriculture, mining, and community
and personal services are associated with growth in middle-class jobs. Being an employee
is more associated with middle-class jobs. Formal sector jobs also offered movement into
middle-class jobs.

22 The most common disability is moderate sight loss (4.1 percent of the working-age population). This is followed by
moderate mobility challenges (1.6 percent) and hearing or psychological difficulties (1.3 percent each). Less than 0.3
percent of the working-age population identifies with having a severe physical or psychological disability.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 193

Workers have control over three factors that will enhance their success at reaching middle-
class jobs. First, they can acquire skills that are appropriate for occupations that tend to
pay  middle-class  incomes. Second, they can  search for jobs in a way that may be more
conducive to finding and securing a middle-class job. Third, they can move to jobs that are
more aligned with middle-class earnings. The barriers to each are discussed in this section,
as is the role of labor legislation in creating middle-class jobs.

5.2.1 Aligning skills with middle-class occupations 

The most populated occupations tend to have a disproportionately small share of
middle-class jobs. Service and sales (22.4 percent), elementary occupations (24.1 percent),
and craft and related trades (12.9  percent) are the most prevalent occupations overall.
However, only 6.6 percent of all middle-class jobs are in craft and related trades and 5
percent are in elementary occupations. In contrast, managers, professionals, technicians,
and clerical workers each represent no more than 9 percent of all jobs, yet they each host
8–20 percent of middle-class jobs. Service and sales workers comprise almost similar shares
of all jobs (22.4 percent) and middle-class jobs (17.1 percent) (Figure 5.8). 

Figure 5.8 Middle-Class Jobs Are Overrepresented among More-Skilled Occupations 

Distribution of jobs and middle-class jobs, by occupation type

Sources: Sakernas 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

A closer look at occupations  finds that  23  (30) occupations have  a 2015 median
(mean) wage above the  middle-class  cutoff.23  The Sakernas has a sufficiently large
sample that one can identify which, among 113 occupations,24 earn a median (or average)
wage that is commensurate with a middle-class job (Table 5.2). Twenty-one (28) of the middle-
class occupations were white-collar jobs at the top of the income distribution: managers (and
military), professionals, and technicians and associates. The remaining two  occupations

23 This calculation employs the survey sample weights. If these weights are not employed, the number of occupations
(by mean cutoff) decreases, dropping archivist, librarian, and related information professionals; life science technicians
and related professionals; and administrative associate professionals.
24 The occupations are defined at the three-digit KBJI, which is the most the data can be disaggregated and still have
meaningful cell sizes.

Pathways to Middle-Class Jobs in Indonesia June 2021

194 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

were plant operators in mining and mineral processing, and locomotive engine drivers and
related workers. The occupations vary widely, and include senior government officials, law
professionals, biologists, computing specialists, business professionals, engineers, police
inspectors, finance professionals, and plant operators in mining and mineral processing. 

Table 5.2 Occupation for Which Median (Mean) Income is above the Middle-Class Cutoff, Sorted by One-Digit KBJI
Occupation

One-digit KBJI Three-digit KBJI

Additional occupations where median

income is below the middle-class cutoff
but mean income is above it
Legislators, senior • Parliament members* • General managers**
officials and managers, • Directors and chief executives*
armed forces
• Senior government officials**
• Other specialist managers*
• Production and operations managers**
• Members of the military (army, navy, air force)
Professionals • Legal professionals** • Archivist, librarians and related information
• Health professionals (not nurses)* professionals

• College/university professors**
• Computing professionals**
• Business professionals**
• Architects/engineers**
• Biologist, botanist, zoologist and related
professionals**
• Mathematicians, actuaries and statisticians;
• Physicists, chemist, and related professionals*
• Social science professionals and other
researchers not mentioned above
Technicians and • Ship and aircraft controllers and • Physical and engineering science technicians;
associates technicians** • Safety, health and quality inspectors;
• Police inspectors and detectives • Life science technicians and related associate
• Customs, tax and related government professionals
associate professionals** • Finance and sales associate professionals
• Administrative associate professionals
Plant and machine • Plant operators in mining and mineral
operators processing
• Locomotive engine drivers and related
workers
* Occupations that fall in the top 25 percentiles of middle-class occupations (that is, occupations whose median income
falls above the middle-class threshold). ** Occupations that fall between the 50th and 75th percentiles of middle-class
occupations.

Sources: Sakernas 2015; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional). The KBJI is the Indonesian Standard Classification of
Occupations (Klasifikasi Baku Jenis Indonesia).

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 195

Some occupations with a high share of middle-class jobs are increasing in size,
while others are decreasing. While none of the occupations in Table 5.2 are among the
20 fastest-growing occupations, in eight of the 20 fastest-growing occupations, at least
15 percent of jobs earn at least a middle-class wage. However, most of the fastest-growing
occupations are low-skill services and trades, which provide few jobs with middle-class
wages. When considering the fastest-shrinking occupations, non-middle-class jobs dominate
(Figure 5.9).

Figure 5.9 Some of the Fastest-Growing Occupations Offer More Middle-Class Jobs
Fastest-growing occupations (striped bar indicates that at least 15 percent of the jobs in the occupation are middle class),
2007–15 

Fastest-shrinking occupations, 2007–15

Sources: Sakernas 2007 and 2015; World Bank staff calculations.

Note: Occupations are at the three-digit level. The statistic is the difference in the occupation as a share of all jobs in 2007 and 2015. A striped
bar indicates that at least 15 percent of the jobs in the occupation earn at least a middle-class wage. Sakernas is the National Labor Force Survey
(Survei Angkatan Kerja Nasional).

Pathways to Middle-Class Jobs in Indonesia June 2021

196 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

5.2.2 What can workers do? Skill-up . . .

As discussed, Indonesia’s labor force is very low skilled. In this section, the available
information on the nature of middle-class jobs is used to inform the discussion on education
level, skills, and the constraints to skilling-up in Indonesia. 

Middle-class jobs require at least an upper secondary education, and preferably a

postsecondary education 

The average rate of return to education is 6.9 percent in Indonesia, on par with low-

income countries in the East Asia region, and far below that of other middle-income
countries  (Figure 5.10).  Skills acquired  in school by Indonesians  are less valued by the
market as compared to the skills acquired in schools in other middle-income countries in the
region. Firms confirm that the skill level of the workforce is a problem. Nearly 14 percent
of firms interviewed in the Doing Business 2015 survey in Indonesia report that “workforce
education” is a major or severe obstacle to doing business, as compared to 8.4 percent of
employers in the Philippines and 2.5 percent in Thailand.25  There is a particularly severe
shortage of the skills needed for more-skilled occupations, including managerial roles (Table
5.3), which are more aligned with middle-class jobs.  While  75 percent of Indonesian firms
that are searching for managers report a shortage of managerial skills in the market, only
30 percent of Malaysian firms seeking managers are faced with skills shortages. In contrast,
only 21.8 percent of Indonesian firms have trouble finding unskilled production workers, as
compared to 38 percent of Malaysian hiring firms. 

Figure 5.10 Indonesia’s Labor Force Has Education Rates Similar to Those of Asian Middle-Income Countries,
but Rates of Return on Par with Less-Advanced Countries
Returns to and average years of education, regional comparators in various years

Source: Montenegro and Patrinos 2014; Sakernas 2018; Demombynes and Testaverde 2018
Note: Montenegro and Patrinos use data circa 2011. The rate of return on an additional year of schooling is calculated using the standard Mincerian
approach, where the log of (daily) wages is regressed against years of school, experience (calculated as age minus years of schooling minus five)
and sex. The data on average years of schooling is from Barro and Lee (2010). The years in brackets on the x-axis is the year of the data for each
particular country.

25 Gomez-Mera and Hollweg 2018

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 197

Table 5.3 Percentage of Firms Citing Inadequate Skills When Hiring Workers, by Type of Worker

TType of worker Indonesia Malaysia Thailand Philippines

(2009–15) (%) (2015–16) (%) (2015–16) (%) (2015–16) (%)

Managers 76.7 30.2 75 34.2

Nonproduction technicians, associate
67.3 50 86.7 55.6
professionals, and sales workers
Skilled production workers 55.1 39.5 46 69.3
Unskilled nonproduction workers 43.4 25.6 57.1 38.9
Unskilled production workers 21.8 38.5 25 48.1

Source: Gomez-Mera and Hollweg 2018

Note: The years in the top row of the table are the years for which the data were collected for each country; two surveys are pooled for
Indonesia.

Middle-class  jobs are overwhelmingly associated with  completion of upper

secondary school or above. More than 80 percent of people holding a middle-class job
have  attended either secondary school (37 percent), technical and vocational education
and training (TVET) (9 percent), or university (38.5 percent), as compared to 41.7 percent
of people with non-middle-class  jobs.  Only 4 percent of people who have only a primary
school education hold  middle-class  jobs.  Although most  middle-class  jobs are held by the
more educated, having a secondary school or postsecondary education does not guarantee
a  middle-class  job. Approximately  23  percent of people with non-middle-class jobs have a
postsecondary degree (Figure 5.11).

Figure 5.11 Middle-Class Jobs Are Overwhelmingly Associated with a Secondary School
Education or Above
Distribution of level, by middle-class job status

Source: Sakernas 2008–18; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Pathways to Middle-Class Jobs in Indonesia June 2021

198 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

Middle-class jobs also require a range of cognitive skills 

Middle-class jobs  are characterized by tasks that require  higher-level analytical,

cognitive, interpersonal, and digital skills, while non-middle-class jobs  are
characterized by tasks requiring  manual skills.  A job can be characterized by the
skills required to carry out the standard  tasks in that job: cognitive (thinking) tasks
versus manual  (physical labor). They can also be characterized by  the  nature  of those
tasks: routine processes or (nonroutine) discretionary processes. These can be mapped to
five categories that have been developed in the literature26 (see Box 5.2 for methodology): 

1. Analytical (nonroutine cognitive) tasks require the person carrying out the task to
analyze and process abstract information and actively use it to solve complex problems; 

2. Interpersonal (nonroutine cognitive interpersonal) tasks  require the person

carrying out the task to manage people and relationships; 

3. Cognitive (routine cognitive) tasks  require the person carrying out the task to
carry out repetitive mental tasks with a high degree of accuracy without using their
discretion; 

4. Adaptable manual (nonroutine manual) tasks require the person carrying out the
task to exercise judgment and physical dexterity; 

5. Routine manual tasks require the person carrying out the task to carry out repetitive
physical motions. 

The authors added a sixth category: digital tasks. These require the person carrying out
the task to work with computers and electronics.

26 Adapted from Acemoglu and Autor (2011) 

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 199

Box 5.2 Estimating the Task Intensity of Middle-Class and Non-Middle-Class

Jobs in Indonesia
The estimation methodology of task intensity trends is based on Acemoglu and Autor (2011). The cited paper
provides scores for the task content of five skill sets per occupation. The skill sets are roughly organized
according to skill type (manual, cognitive, interpersonal) and degree of repetitiveness. This returns five skill
types (interpersonal repetitive is not measured): 

1. Analytical skills (nonroutine cognitive): These require individuals to analyze and process abstract infor-
mation and actively use it to solve complex problems. Examples of tasks: identifying underlying principles,
reasons, or facts by breaking down information or data into separate parts; thinking creatively to develop,
design, or create new applications, ideas, systems, or products. Examples of occupations: engineers, doc-
tors, lawyers, and scientists. 

2. Interpersonal skills (nonroutine interpersonal): These require an ability to manage people and relation-
ships. Examples of tasks: Developing and maintaining constructive and cooperative working relationships
with others; using persuasion and negotiation skills; adapting to and coordinating the actions of others;
coaching others to improve their performance. Examples of occupations: managers, business administra-
tors, schoolteachers, police, salespeople, and those engaged in public relations. 

3. Cognitive skills (routine cognitive):  These require individuals to perform repetitive mental tasks with a
high degree of accuracy. They are very structured and do not allow workers to determine their own tasks
and priorities. Examples of tasks: typing, data entry, bookkeeping, or checking ledgers. Examples of occupa-
tions: clerks, keyboard operators, and tellers.

4. Adaptable manual skills (nonroutine manual): These require individuals to exercise judgment and physical
dexterity to perform certain tasks. Examples of tasks: driving and operating mechanized equipment such
as forklifts or other vehicles; using hands and arms to grasp, manipulate, or assemble objects or control
mechanized equipment. Examples of occupations: drivers, crew on boats or ships, laborers in mining and
construction.

5. Routine manual: These require individuals to spend a significant amount of time making repetitive physical
motions.  Examples of tasks:  Operating machines or controlling processes by using direct physical activi-
ty or mechanical mechanisms, not including the operation of vehicles or computers. Examples of occupa-
tions: textile workers, metal processers, and operators of plants and machines.

The score for each skills type is averaged across occupations that are classified as middle class or non-middle
class, using the middle-class wage as the sorting variable.

A sixth score is created for “digital,” based on Pradana and Wihardja (Forthcoming), which draws from Muro et
al. (2017). The digital task intensity is generated from two task items in O*NET:a (i) “Knowledge: Computer and
Electronics,” which measures the overall knowledge of computers and electronics required to carry out a job;
and (ii) “Work Activity: Interacting with Computers,” which quantifies the centrality of computers to the overall
work activity of the occupation. 
a
O*NET Online. N.d. https://www.onetonline.org.

Pathways to Middle-Class Jobs in Indonesia June 2021

200 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

Middle-class jobs require high levels of cognitive skills. Analytical, routine cognitive, and

interpersonal skills are consistently the most necessary skills for middle-class jobs (Figure
5.12). Analytical skills are clearly used in the tasks associated with many of the occupations
described in Box 5.2: legal professionals, biologists, and engineers. Interpersonal skills are
also critical for middle-class jobs. Managers, for example, may need strong interpersonal
skills. Manual skills  are less necessary for middle-class  jobs. Digital skills are important in
middle-class jobs, but cognitive and interpersonal skills more so.

Figure 5.12 Middle-Class Occupations Require Figure 5.13 Non-Middle-Class Occupations

Analytical, Cognitive, and Interpersonal Skills Require Manual Skills
Task intensity of middle-class occupations, 2007–15 Task intensity of middle-class occupations, 2007–15

Source: Sakernas 2007–15 and O*Net task content of occupation; World Bank staff calculations.
Note: The task intensity score is estimated for each occupation and then aggregated across a sample of middle-class occupations to generate a
score for each task intensity type for each year. The sample of middle-class jobs was based on the 2015 Sakernas and the same set of occupations
is used for each sample year. Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

In contrast, the task intensity of non-middle-class occupations is highest in manual

skills and lowest in digital  skills.  Adaptable manual tasks, such as driving or cleaning,
appeared more frequently than tasks requiring routine manual skills, such as those found
when working on an assembly line. In non-middle-class jobs, routine cognitive tasks were the
most frequently appearing cognitive skills and this increased slightly over time. Digital skills
are the skills least used to perform non-middle-class jobs, though the slight upward trend
may suggest that this is changing. 

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 201

Figure 5.14 Firm Perceptions of Skills Shortage Figure 5.15 Few Indonesian Firms Offer Training
and Skill Training Share of firms offering formal training, by country
Skills shortages and firm training

Source: Gomez-Mera and Hollweg 2018

Interviews with a small group of Indonesian employers uncovered that the most

severe skills shortages  are  of skills that are necessary to perform tasks
commonly a feature of middle-class jobs. The World Bank Enterprise Survey asks firms
that are searching for new employees  to rank  on a scale of 1 (not difficult) to 4 (very
difficult), the degree of difficulty in finding employees with each of seven skills. Indonesian
employers  identified foreign language skills  (analytical), managerial skills  (interpersonal),
work ethic and commitment (interpersonal), and interpersonal and communications skills as
most lacking (Figure 5.14). These skills are less about knowledge in a specific field; rather, they
are behavioral. They are also prominent requirements of middle-class jobs. These skills were
ranked even higher than  writing skills, even though the standardized tests of adult skills
in 2014–1527 found that Indonesian adult literacy is, on average, at a low level. Computer
and IT skills ranked lowest, perhaps not surprising given the low intensity of digital tasks in
Indonesia’s occupations in 2015. 

Turning to knowledge needed for middle-class jobs, postsecondary studies are

roughly aligned with middle-class occupations. The middle-class jobs discussed above
require technical knowledge in business administration, health (more specialized than nurses),
information and communications technology (ICT), engineering, finance, and science. The
distribution of field of postsecondary education study roughly reflects the job market, with
a larger share in engineering, business administration, and law. There are a few exceptions:
a significant share of postsecondary graduates specializes in education—especially women—
which is aligned with formal sector jobs but not with middle-class earnings. Medicine, natural
sciences, math, and statistics form a smaller share of middle-class than non-middle-class
jobs, but this may reflect  the prevalence of nursing and other health professionals, who
tend to earn less and tend to be women. In fact, women’s fields of postsecondary study
reflect the distribution for non-middle-class jobs, while men’s fields of postsecondary study
are more aligned with the distribution for middle-class jobs (Figure 5.16). 

27 Conducted by the Programme for the International Assessment of Adult Competencies (PIAAC)

Pathways to Middle-Class Jobs in Indonesia June 2021

202 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

Figure 5.16 Women’s Fields of Postsecondary Study Reflect the Distribution for Non-middle-class Jobs,
while Men’s Distribution Is More Aligned with Middle-Class Jobs
Distribution of field of postsecondary education study, by gender and middle-class/non-middle-class jobs

Sources: Sakernas 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Middle-class jobholders pursue training courses once leaving school, both in the market and in
firms 

Workforce upskilling is  constrained by  low take-up in the training sector and
limited employer provision of training.  Working adults rarely go back to formal
education, so they depend on short-duration TVET or on employer training. While 10 percent
of workers report having participated in formal training courses, 26 percent of middle-class
workers have done so.28 The low rates of upskilling are perhaps not surprising given the low
supply of training by Indonesian firms.  About 13 percent of Indonesian firms  report that
they provide training to their employees, compared to 32 percent of firms in East Asia and
Pacific countries and more than half of Filipino firms (Figure 5.15). When Indonesian firms do
provide training, they tend to train more-skilled workers. For example, 55 percent of firms
said that they provided training to skilled production workers, as compared to 23 percent
who provided training to unskilled production workers. Training is concentrated in technical
skills specific to the firm, which are not the main skills shortages identified by employers.
Thus, adult workers are acquiring new skills via training from firms, but these skills may not
be the right ones to move them into middle-class jobs. 

28 Derived from Sakernas 2018. A person who has received training is 5 percentage points more likely to be a middle-class
worker, even when controlling for years of education, age, gender, and location (urban or rural). The nature of the data do not
allow us to determine if the training was acquired before or after the worker entered the middle-class job.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 203

5.2.3 What can workers do? Expand their job-search methods . . .

Middle-class jobs are  acquired through a mix of proactive methods, while non-
middle-class jobs are primarily found through informal personal networks.  While
the data will not allow for identification of the search method used to find a middle-class job,
“education level” can be used as a proxy for the types of workers who end up in middle-
class  jobs.  Jobseekers who have an upper secondary or postsecondary education  rely
heavily on personal contacts, but they also use a range of other job-search methods: they
respond to job advertisements, they reach out directly to companies and firms that may
be hiring, and they sign up for job-search services, such as websites and employment
agencies. As the level of education increases, the proactive job-search measures increase. In
contrast, jobseekers with a lower-secondary education or below almost exclusively rely on
family or other acquaintances to find out about jobs (Figure 5.17).

Figure 5.17 Proactive Job Search Increases with Education Level

Method of job search, total and by education level

Sources: Sakernas, 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

A large share of  jobseekers find a job within three months or less.  Sixty percent
of jobseekers find a job within three months and another 20 percent within six months
(Figure 5.18).  This is somewhat fast when considering the  complexity  of the Indonesian
labor market and the absence of consolidated job vacancy information, which brings into
question whether people are taking the time to secure the best job possible. Approximately
42 percent of those who are searching for jobs are actually employed. Most say that they
are looking for a new job to increase their income. The primary reason given for leaving a
job is “insufficient income”(23 percent). This suggests poor job matches.

Pathways to Middle-Class Jobs in Indonesia June 2021

204 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

Figure 5.18 More Than Half of Jobseekers Find a Job within Three Months
Number of people by duration of job search, in months, at time of interview, by employment status at time of interview

Sources: Sakernas, 2018; World Bank staff calculations. The statistics come from the question, “How long has (NAME) been looking for a job/
preparing a business?”.
Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Notably, skills mismatch—meaning that people are not in the right job for their
skill set—is on the rise.  In 2016,  1 percent of  urban workers with at least a Bachelor’s
degree or equivalent who left a job in the past year did so because they felt that the job
was not  commensurate with their skills; by 2018, this had increased to 5 percent.29  The
skill mismatch is more likely to occur for a person who has previously worked as a wage
employee, who resides in an urban area, and has a diploma from TVET training, a university
degree, or a postgraduate degree.

Few middle-class jobholders wish to move to another job. Only 1.1 percent of those

with a middle-class job are engaged in job search, compared to (a still small) 4.4 percent of all
workers. Reasons for job search do not differ much between those who have a middle-class
job to those who have a non-middle-class job; both groups say earning more money is their
primary reason for looking for a job (50 percent and 40 percent, respectively), while being
responsible for providing the household comes second (21 percent and 34 percent).

5.2.4 What can workers do? Move to more lucrative jobs . . .

Workers with a strong skill set and information about lucrative jobs may still not
transition to middle-class jobs, due to impediments to labor mobility. Indonesian
labor is mobile. Labor mobility covers: (i) moving to a different state of employment, such
as from unemployment to employment or from casual work to a permanent contract; (ii)
moving within or across firms (or sectors) but maintaining the same state of employment;
and (iii) moving geographic relocation. In 2018, more than 10 million people were searching
for a job at the time of the Sakernas interview. About 6 million unemployed workers

29 Sakernas, August 2016–18; World Bank staff calculations

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 205

were looking for a job, while another 4.3 million employed workers were seeking to move
between jobs. Chapter 4 reported significant job churning in the manufacturing sector. The
Indonesian Family Life Survey (Survei Aspek Kehidupan Rumah Tangga Indonesia) (2007–14)
reports that approximately 32 percent of internal migrants relocated for work reasons;
the tendency is particularly strong among those who have lost their jobs, with 49 percent
migrating after a job loss.30 That said, relocation is quite unusual.31

Worker mobility might increase if the cost of changing job was lower. The cost of
labor mobility may be monetary (such as the costs incurred during job search or relocation,
or the cost of training to transition to a new industry), indirect (such as the foregone
earnings while not working or scarred labor market prospects while not working—see Box
5.1), or nonmonetary (psychological stress or loss of self-esteem caused by a job search).
Taking these together (since the methodology used does not allow each motivation to be
estimated separately), Indonesia has exceptionally high labor mobility costs relative to other
countries.32 A 47-country study found that only 4 countries had higher labor mobility costs
than Indonesia.33 Labor mobility costs in Indonesia in 2007 were estimated at 2.8 times the
annual mean wage. Women seem to have higher mobility costs than men, while youth have
lower mobility costs than adults. More-skilled workers have lower mobility costs. Workers in
regions further from Jakarta (especially Eastern Indonesia) face particularly high mobility
costs.

Labor mobility costs impede entry into some sectors in which middle-class jobs are
prevalent, but may be less of a barrier to entry in other sectors. In 2007, manufacturing
was one of the lowest-cost sectors for workers to enter, costing about 1.8 times the annual
average wage. High-value services were quite costly to enter, equal to more than four times
the annual average wage. The high mobility cost is likely due to the education requirements
in high-value services relative to other sectors. Trade and accommodation seems to be the
least costly sector for entry, as well as one of the most populous sectors in 2007.34

Weak social protection and labor policy may add to the cost of labor mobility.
Indonesia has generous severance payment laws, which should reduce the cost of mobility.
However, few workers who are eligible for severance pay actually collect it.35 Indonesia passed
its first unemployment insurance program in October 2020. Until it becomes effective, a
worker who loses a job is unlikely to have sufficient income to finance a job search or other
transition costs. The absence of a social protection and labor policy that provides benefits to
ease job transition may lock workers into less-productive jobs, thus restricting their chances
of moving up to higher-paying, perhaps middle-class, jobs.

30 Hollweg et al. 2016

31 World Bank 2020b
32 Hollweg et al. 2016
33 Calì, Hidayat, and Hollweg 2018
34 Hollweg et al. 2016
35 Weber and Kudo 2020

Pathways to Middle-Class Jobs in Indonesia June 2021

206 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

5.3 Additional Factors Limit Access to Middle-Class Jobs for

Vulnerable Populations
Education, skills, job search, and labor mobility costs will be challenges for all  who wish
to acquire a middle-class job. However, some groups may face particular challenges. This
section explores three such groups: women, youth, and those with disabilities. 

5.3.1 Gender roles significantly limit women’s access to middle-class jobs 

Women’s underrepresentation among middle-class jobholders is due to occupational
sorting, job type, and traditional gender wage gaps, and is underpinned by the
continuance of traditional gender roles. Middle-class jobs are  commonly employee
positions in higher-paying occupations. However, women—especially  married women with
children and women who are head of a household—are underrepresented among employees,
holding  35 percent of employee jobs but 39 percent of all jobs  (Figure 5.19).  Their self-
selection into unpaid employment or self-employment36  seems to reflect childcare
duties37 and social norms that undervalue women’s jobs.38

Figure 5.19 Women Are Overrepresented in Low-Wage Types of Jobs

Share of gender, by job type, 2018

Sources: Adapted from Lain, Alatas, and Setyonaluri (2020)

36 Women with children in the household are 1.1–1.9 percentage points more likely to be self-employed than are women
without young children in the household. Men with children in the household are 0.1–1.4 percentage points more likely to be
self-employed than are men without. Women who head a household are 17 percentage points more likely to be self-employed
than women who do not head a household, compared to 8 percentage points for male household heads. 
37 Indraswari (2006) finds that Indonesian women choose to open small-scale businesses (and become own-account workers)
with the aim of contributing to family income, but without needing to enter the formal labor market, which they perceive would
entail sacrificing their position as a mother.
38 Fagertun (2017) reports that men and women in the Bali tourism industry perceive that a woman’s work is supplementary
to that of her husband or male kin. 

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 207

The growth of the service sector provides women with opportunities to find middle-
class jobs, but women tend to cluster in the lower-paying segments of the service
sector, working in more traditionally female jobs. Around 84 percent of middle-class
jobs carried out by women are in the service sector compared with 65 percent of middle-
class jobs carried out by men. However, women dominate the low-value service jobs. More
than 53  percent  of working women are employed in low value-added service sectors, as
compared to 43 percent of men. Another 3.4 percent of working women are employed in
high value-added service sectors, compared to 11.4 percent of men. In the service sector,
women  are more successful in holding  middle-class  occupations (professional,  technical,
clerical, but not managerial) than in other sectors, though they tend to work in the lower-
paying jobs within occupations. For example, 79 percent of female service professionals are
in teaching, which pays very low wages compared to other service professionals. Similarly,
more than half of technical service jobs held by women are in life sciences and health, and
most of these positions are low-paid nursing roles. In other words, women cluster into
“feminized” occupations within the service sector; these occupations pay less than other
occupations in the sector. That said, women also hold the majority of the professional jobs in
services that are non-middle class. Very few women work in high-value services, which offer
more  middle-class jobs than do low-value services. However, those who do earn more, on
average (and at the median), than do men (Figure 5.20). 

Figure 5.20 Women Have Particularly High Earnings in High Value-Added Service Sector
Mean and median earnings in high- and low-value services and other sectors by gender in 2018

Sources: Lain, Alatas, and Setyonaluri (2020) using Sakernas 2018

Note: Earnings data taken from income earners (wage workers, own-account workers, and casual workers) only.
High-value services are: (i) transport, storage, and communications; and (ii) financing, insurance, real estate,
and business services. Low-value services are: (i) wholesale and retail trade, and hotels and restaurants; and (ii)
community, social, and personal services.

Pathways to Middle-Class Jobs in Indonesia June 2021

208 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

Even if men and women did acquire similar occupations and work types, the gender
wage gap persists, which keeps women at a distance from men in acquiring jobs
that pay a middle-class income. Indonesian men earn 25 percent more for an hour of
work than do Indonesian women  (down from 38 percent in 2001)39,40; this is double the
gender wage gap in other Southeast Asian countries.  Gender gaps are becoming increasingly
difficult to explain through different “endowments” of education, experience, and location
over the years. Moreover, the gender wage gap is not explained by different occupations,
sectors, or types of work.41 Qualitative data suggests that social norms may play a role. For
example, the OECD (2019) finds that nearly one-third of respondents in a perceptions survey
disagreed with the statement “it is perfectly acceptable for any women in their family to
have a paid job outside the home if she wants one.” The perception that a woman’s work
is supplementary to that of their male partner or male kin42 may discourage women from
pursuing careers or from staying attached to a labor market in which they can move up the
earnings ladder. 

A women may trade off income in favor of a job with benefits that are perceived
as family friendly. While gender wage gaps favor men, and the gap is smaller for middle-
class jobs, women enjoy greater nonpecuniary benefits. Women holding middle-class jobs
are about 5 percentage points more likely than men to have written contracts,43 to hold
family-friendly benefits (annual, sick, or maternity leave), and  work shorter hours than
men44 (Figure 5.21). This evidence aligns with recent analytical work in Vietnam, which finds
that women trade off higher wages for family-friendly social benefits.45 

39 All available earnings data cover income earners (wage-employed, own account workers, and casual workers).
Blinder-Oaxaca-style decomposition is implemented using technique described by Elder, Goddeeris, and Haider 2010.
Control variables used are: age; educational attainment; urbanicity; and province fixed effects. Data source is Sakernas
2001-2018. See Lain, Alatas and Setyonaluri (2020).
40 The hourly wage gap is 45 percent among the poorest (first decile) of workers and nearly absent among the
tenth decile of workers.
41 Adding fixed effects for job status (wage work versus own account work versus casual work), sector, and
occupation reduces the gender earnings gap by just 1-2 percentage points. Thus, even after accounting for sorting
between different types of jobs, unexplained gender earnings gaps remain large. Source: Lain, Alatas and Setyonaluri,
2020.
42 Fagertun 2017
43 Approximately 92 percent of women holding middle-class jobs have a written contract, versus 88 percent of men
holding middle-class jobs. Further, 50 percent of women with a non-middle-class job have a written contract, versus 45
percent of men with a non-middle-class job.
44 Women work a median of 40 hours weekly, as compared to men who work a median of 45 hours weekly.
45 Chowdhury et al. 2018

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 209

Figure 5.21 Women Enjoy Greater Nonpecuniary Benefits Than Do Men

Share of men and women who are employees in middle-class and non-middle-class jobs who receive each type of social benefit

Sources: Lain, Alatas, and Setyonaluri (2020) using Sakernas 2018.

Note: The sample is wage employees only.

Gender roles may also limit women’s jobs search. Being married or having children in
the household is negatively correlated with women’s job-search activity, while it is positively
correlated for men. Women who are working are less likely than working men to search for
a higher-paying job, especially if they have children in the household.

5.3.2 Youth are at a structural disadvantage 

Middle-class jobs come with experience, especially in certain occupations. The share
of middle-class jobs increases with age, with a sharp drop-off once people begin to retire
and collect a pension (Figure 5.22). Since middle class is measured in this report with an
income cutoff, this means that income increases with age. Youth have more education
than do older adults, so it might be expected that they would also have higher wages. But
instead, experience in the workforce, particularly in certain occupations, drives the wage
gaps between youth and adults. Notably, women’s acquisition of middle-class jobs stalls
during childbearing years, resuming in their late forties.

Pathways to Middle-Class Jobs in Indonesia June 2021

210 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

Figure 5.22 Workers Acquire Middle-Class Jobs as They Age, But Women’s Acquisition
of Middle-Class Jobs Stalls during Childbearing Years
Middle-class jobs as a share of income-earning jobs, by age

Sources: Sakernas 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Youth earn 32 percent less than adults, on average.46 Observable characteristics explain
about 58 percent of the difference in earnings, and most of that is explained by years of
experience in the workforce.47 In other words, if youth had the same level of experience as
adults, they would earn more (on average) than what they earn now relative to those adults.
Each year of experience increases wages by an average of 3.5 percent (at a decreasing rate
of increase), which accumulates with time. However, the annual rate of wage increase is
actually higher for youth than adults, meaning that youth receive big pay raises—about 10
percent on average—in their early years in the labor force, thus closing the wage gap faster
than the average increase would suggest.48

Young people’s underrepresentation in higher-skill occupations also contributes to

the wage gap with adults—and therefore their more limited access to middle-class
jobs. If youth held managerial, professional, and semi-professional white-collar jobs at the
same rates as adults, the wage gap would shrink slightly. But perhaps more importantly,
if they were paid similarly within occupation as older adults (holding all else, including
education level, equal), the wage gap would reduce by 40 percent. The return to experience
for high-skill occupations is particularly strong. For example, a young person who is in a
professional career earns 50 percent less than an older professional; the wage differential
is 25 percent among managers.

46 In 2018, adults earned an average of Rp2.5m per month and youth earned Rp1.8m per month.
47 The authors estimate a Oaxaca-Blinder decomposition to identify which factor endowments, and returns on those
endowments, explained the wage gap between youth (aged 15–24) and adults (aged 25–64). The regressors were:
experience (and its square), education level, gender, location (urban or rural), formal sector employment, occupation,
sector, and regional dummies. About 58 percent of the wage gap is explained by endowments.
48 The authors estimate a standard Mincerian regression for the entire labor force and separately for youth (aged 15–24)
and adults (aged 25–64). The regressions include experience (and its square), education level, gender, location (urban or
rural), formal sector employment, occupation, sector, and regional dummies. The authors estimate the regression again,
with an interaction term for each occupation and a youth dummy.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 211

Box 5.3 Transitioning to Adulthood

Indonesian youth have high unemployment rates relative to adults and youth in the East Asia and
Pacific region. While 15.8 percent of male and female Indonesian youth are unemployed, only 4.5 percent of
adult men and 4.1 percent of adult females are searching for a job (as a share of all those who are working or
searching).

Indonesia’s high youth unemployment rates reflect a turbulent transition from school to work. In
2018, about 2.7 million people aged 17 in Indonesia were in education, as compared to 1.5 million people aged 18,
1 million aged 19, and 300,000 aged 23. In 2018, 3.3 million youth (aged 15–24) left school and transitioned into
another state (work; job search; or not in education, employment, training, or job search [NEET]). No other age
group experiences such big upheaval in a short period.

Figure 5.23 Indonesia’s high youth unemployment rates reflect a turbulent transition from
school to work
(Change in the number of people by each state, by age)

Sources: Sakernas 2018; World Bank staff calculations. The size of sample for each age is standardized to the sample size of the
24-year-olds.
Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

Most youth who leave education go to work. Except for the youngest school-leavers (aged under 18), most
school-leavers transition directly from school to work. Of the year group with the biggest exit from school
(those aged 18), 39 percent went to work, 35 percent moved to job search, and the remaining 26 percent were
NEETs. In other words, secondary school releases 1.2 million people into the labor force at the same time and
nearly 40 percent are working by the next year. A similar release of people from education (university) to work
occurs at age 22, and most are working within a year.

Insufficient job creation and inadequate job-search tools may impede the school-to-work transition
in Indonesia. While one would expect youth in any country to face a turbulent period, it seems to be more of
a challenge for Indonesian youth. Indeed, Indonesian youth are 5 percentage points more likely than Malaysian
youth and youth in all of East Asia and Pacific to be unemployed (10.2 percent of young Malaysian men and 11.9
percent of young Malaysian women are unemployed; in East Asia and Pacific, 10.9 percent of young men and
10.0 percent of young women are unemployed). Youth “idleness” (NEET rate) is also higher compared to East

Pathways to Middle-Class Jobs in Indonesia June 2021

212 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

Box 5.3 continued

Asian and Pacific countries: 15.5 percent of men and 28 percent of women in Indonesia, compared to 8.8 and
17.8 in Malaysia. This may be due to several factors. First, an average of two million jobs are created annually in
Indonesia, while more than three million Indonesians leave school annually. This will necessarily limit job entry,
even if a large share of youth opt out. Second, the preponderance of low-wage jobs may result in youth choosing
to wait for a better job opportunity.

In spite of extensive job search, young people struggle to find jobs. Compared to
adults, youth use increasingly varied job-search methods and search for longer than adults.
Only 42 percent of jobseekers aged 15–24 use only one job-search method, as compared to
49 percent of jobseekers aged 25–34 and 91 percent of those aged 55–64. More than one-
third of youth use three or more job-search methods. The most common forms are personal
connections (87.8 percent), searching or applying to newspaper or digital advertisements
(49.9 percent), contacting companies directly (36 percent), and registering in a jobs database
(25 percent). These alternative methods drop off sharply at age 35. Only 63 percent of
youth acquire a job within three months, despite the fact that they conduct a more intense
job search than that carried out by workers aged 45 and above. In comparison, more than 72
percent of jobseekers aged 45 and above acquire a job with three months. However, within
one year, about 95 percent of jobseekers of all ages have found a job.

5.4 Limited Role of Labor Law in Creating Middle-Class Jobs

Labor legislations—minimum wages, worker protections, social benefits—are well defined in
Indonesia. The Labor Law and associated regulations lay out a rich set of guidelines for the
rules of the game for employers and workers,49 aiming to balance rights and responsibilities
between parties. But can this law and these regulations also serve as tools to increase
middle-class jobs in Indonesia? Two types of labor regulation may influence middle-class jobs:
regulations on minimum wages and on worker protection.

5.4.1 The limitations of the minimum wage

The minimum wage could be the primary policy lever for influencing the wages of
low-skilled workers, but it is not a tool for creating middle-class jobs. In principle, the
minimum wage should increase the wage floor. In practice, it rarely serves as a wage floor
in developing countries—even among employees with labor contracts.50 An estimated 43
percent of workers earn below the province-mandated minimum wage in Indonesia.51 For
example, Figure 5.24 shows that in the province of West Sumatra, nearly half of workers
with a labor contract—who are legally mandated by law to earn at least the minimum wage—
earn significantly below. The distribution of subminimum wages reflects the distribution
of low productivity among a significant number of workers. Only by enacting measures to
increase labor productivity will employers be able to pay closer to the minimum wage and,

49 See Weber and Kudo (2020) for a recent review.

50 Cunningham 2007; Del Carpio and Pabón 2014
51 Minimum wages are set at the provincial level in Indonesia, although district and district-sector minimum wages can
be set at a higher level than the corresponding provincial minimum wages.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 213

eventually, a middle-class wage.

A higher minimum wage that is not accompanied by productivity increases will

have a limited effect on wage increases in Indonesia. Studies exploring the impact
of the minimum wage on employment outcomes in Indonesia, mostly from before 2010,52
generally find that an increase in the minimum wage has a small positive effect on average
wages, with larger impacts for those earning below or near the minimum wage. However,
more recent data suggest that a minimum wage can be too high, and thus lose its ability
to affect wages. For example, the minimum wage is 80 percent above the median wage in
the province of Aceh. Figure 5.24 shows the distribution of wages paid to those holding
employment contracts in both provinces, and a vertical line at the minimum wage. If
minimum wages did not have any effect on wages, one would see a relatively smooth wage
distribution at the point where the minimum wage is, as seen in the graph for Aceh. Nothing
remarkable happens to the wage distribution at Rp2.7 million. However, the graphs for wages
in West Sumatra show a clear spike in the wage distribution at the minimum wage. It might
be interpreted that wages just below and above Rp2,119,067—the minimum wage in that
region—are “pulled” toward that wage but the rest of the wage distribution is unaffected.
Thus, setting a minimum wage close to the middle-class income cutoff is not likely to pull
low-wage earners up to middle-class jobs, though it may help those who are earning just
below it. As the graphs show, a minimum wage that is not commensurate with productivity
levels largely leads to evasion.

Figure 5.24 The Distribution of Wages Paid to Those Holding Employment Contracts in West Sumatra and
Aceh, and whether Minimum Wages Affect Wages (West Sumatra) or Do Not Affect Wages (Aceh)
Distribution of (log) nominal monthly wage; minimum wages

Sources: Sakernas 2018; World Bank staff calculations

Note: Vertical lines show minimum wages. Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

52 Hohberg and Lay 2015; Alatas and Cameron 2008; Harrison and Scorse 2010; Chun and Khor 2010

Pathways to Middle-Class Jobs in Indonesia June 2021

214 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

Since the middle-class cutoff is above the market wage, and noncompliance with
the minimum wage is high, it is very unlikely that the minimum wage can move a
significant share of workers to middle-class status. While these data are not conclusive,
they do suggest that for many workers, productivity falls below the minimum wage. The
province-specific minimum wages account for some of this productivity difference, but a
rich literature shows that the minimum wage does not increase worker productivity. Without
efforts to increase worker productivity, simply setting a minimum wage that is equal to the
middle-class wage and enforcing it is more likely to lead to job losses and bankruptcies than
to a larger middle class.

5.4.2 Worker protections are not synonymous with middle-class jobs

Theoretically, labor legislation could improve the quality of jobs by lowering risk.
The middle-class income cutoff is determined by the lowest income at which a household
has less than a 10 percent likelihood of falling into poverty and vulnerability if faced with an
economic shock (see Box 1.1). If households accumulate risk protection measures through
their jobs in the form of health insurance, severance pay (to guard against income loss),
job protection when pregnant, occupational safety regulations, paid annual leave (to stay
healthy), and pensions (to guard against old-age impoverishment), for example, they may
be able to maintain that middle-class status even when faced with an economic shock. In
other words, the cutoff for a middle-class income would be lower if the value of worker
protections provided under the Labor Law were monetized.

Table 5.4 Share of Workers Who Receive Each Benefit, by Middle-Class Job Status

Middle-class job No middle-class job

Health insurance 86.4 % 28.9 %

Work accident benefit 78.4 % 24.5%
Death benefit 68.5 % 16.6 %
Old-age benefit (nongovernment) 62.1 % 13.1%
Pension (government) 56.3 % 9.7 %
Paid annual leave 75.2% 23.0 %

Sources: Sakernas 2018; World Bank staff calculations

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

The current labor legislation regime is not effective at creating middle-class jobs
due to incomplete coverage and low levels of enforcement.53 While 22.7 percent of
wage employees have an indefinite-term work contract and should be guaranteed a full
regiment of social benefits, only half actually receive them (Figure 5.25). Among all wage
employees, only 7 percent earn a middle-class income, receive full benefits, and have a
permanent contract. Further, not all benefits are equally received by all holders of middle-

53 Health insurance, work accident insurance, death benefits, old-age benefits (non-government), pension
benefits (government), and/ or paid annual leave / sick leave / maternity leave. 

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 215

class (or non-middle-class) jobs (Table 5.4). Health insurance is the most common benefit; it
is provided for 86.4 percent of those in middle-class jobs, as compared to 28.9 percent of
those in non-middle-class jobs. Just under half of all workers with a middle-class job are not
paying into a pension system, meaning their middle-class status may be threatened once
they stop working.

Labor legislation that guarantees worker protections is often not complied with.
For example, workers with a permanent contract are entitled to generous severance pay
when dismissed from their job. However, statistics on job terminations (2008–13) report
that while 11.3 percent of workers whose employment was terminated were legally entitled
to severance pay, only 3.2 percent received it, and only 0.5 percent received the full amount
as per Indonesian legislation.54

In sum, while labor legislation is intended to improve wages, labor protections,

and access to social benefits, its current structure and coverage suggest it only
benefits a small segment of the workforce. There is little evidence that increased
benefits (minimum wages, social benefits) affect employment decisions made by firms or
firm productivity. Some worker protections may even hinder transitions to middle-class
jobs; for example, severance pay that increases with tenure may discourage long-term
employees from leaving a job and searching for a better one. High severance payments may
discourage firms from hiring permanent workers (who are eligible for labor-related social
benefits), instead contributing to the pool of workers on fixed-term contracts, where social
benefits and job security are tenuous. These questions point to a large research agenda that
is needed to understand the role that labor legislation can and cannot play in the efforts to
move Indonesia to middle-class status.

54 Weber and Kudo 2020

Pathways to Middle-Class Jobs in Indonesia June 2021

216 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

Figure 5.25 Distribution of Employees by Middle-Class Wage, Receipt of Social Benefits, and Contract Status

Sources: Sakernas 2018; World Bank staff calculations

Note: The sample excludes unpaid workers and employers, since earnings cannot be imputed for either. An employer, by definition, does not
collect benefits and does not have a contract, since they do not have an employer. The middle-class remuneration cutoff is defined as Rp3,752,000
monthly. Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional).

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 217

5.5 Conclusions55
This chapter has argued that Indonesia’s labor force is not ready to take on mid-
dle-class jobs. Of the 124 million workers, 85 million are income earners. Only 15 percent
of the 85 million earn income sufficient to be classified as middle class. Men, older workers,
the more educated, and urban dwellers without a severe physical disability are more likely to
hold a middle-class job. Women are at a particular disadvantage given their low labor force
participation rates, their crowding into low-wage sectors, their tendency to take on unpaid
family work, and the continued conflict between work and home. Youth can only hope to
acquire middle-class jobs in the future since work experience is key to these jobs.

Workers can undertake three strategies to better prepare for and integrate into
the labor market. First, workers need skills that align with the needs of the market. The
current workforce has secondary school completion rates on par with other middle-income
countries in the East Asia region, but learning seems particularly low. Indonesian employers
are less happy with the skills offered by the labor force as compared to employers from
other countries, with particular dissatisfaction among employers offering more-skilled
(middle-class) jobs. The future workforce is also lagging, with poor learning outcomes. At the
same time, middle-class jobs are becoming more intense in tasks that require interpersonal,
routine cognitive, analytical, and digital skills, all of which are best learned in upper secondary
school. Second, jobseekers (and employers) need to employ a broader range of methods if
they are to find (or offer) middle-class jobs. Personal networks are insufficient, as they do
not provide the information jobseekers need to plan careers, prepare for them, and identify
job opportunities. Third, workers need to be willing to move themselves up the job ladder by
changing job or even location in search of greater opportunities. That said, labor mobility is
costly in Indonesia, especially for women, the less skilled, and those living in remote areas.

Labor legislation may not be a viable tool for creating middle-class jobs. The scarce
information about the minimum wage suggests that it may successfully increase wages,
but it seems to lose impact as it increases, relative to the market wage. Since the middle-
class cutoff is above the market wage, and noncompliance is high, it is very unlikely that,
in the absence of large labor productivity increases, the minimum wage can move workers
to middle-class status. Worker protections can provide nonpecuniary benefits that may
protect workers against falling out of the middle class, but compliance is low and there is
little enforcement.

Public policies and programs can help workers better prepare for, and transition to,
middle-class jobs. First, policies can facilitate learning of the right skills. They can help
youth complete upper secondary school and acquire a broad range of knowledge, and the
analytical and interpersonal skills necessary in occupations that are aligned with middle-
class jobs.56 The 124 million workers who will not return to school will require affordable and

55 This section is partly drawn from Pratomo 2020.

56 Indotask (World Bank 2021a) finds that the skills most in demand include speaking, reading comprehension, and
active listening. These skills are often taught and reinforced in primary and secondary school, as opposed to tertiary
technical or academic programs.

Pathways to Middle-Class Jobs in Indonesia June 2021

218 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

market-relevant training to build these skills. Second, a more comprehensive and service-

oriented labor market information system can shorten the time needed to find a job and
increase the efficiency of job matches. This can lead to more productive job placements
that are more aligned with the skills of workers and the needs of employers. The revamped
system will need to cover four key functions: job matching, career guidance, labor market
intelligence,57 and a virtual one-stop shop for access to government support.58 The labor
market intelligence function, combined with the career guidance, should provide updated
and reliable information about skills in demand, occupational prospects, and individual
assessments of aptitude and preference to enable individuals to make informed decisions
about their future. The current job-search process is too limited for a labor market as rich
as that of Indonesia. Workers may need tailored support through career services that use
the labor market information system to coach jobseekers on how to find, prepare for, and
transition to jobs. Third, certain populations will need additional support to overcome
barriers to middle-class jobs. The policy section below discusses the system that is leading
to the poor workforce results and proposes policies to improve it.

Reform Area 1: Facilitate Learning

The development of human resources is a main priority of President Joko Widodo’s second
term. In his inaugural speech, he stated the government’s intention to build “a workforce
that is hard-working, dynamic, skilled . . . [T]his cannot be achieved with old methods. New
methods must be developed.”59 The analysis in this chapter finds that the skills development
system will need to teach more students a broader range of skills over a longer period of
time if Indonesia is to meet the skills demands of modern jobs. Four areas of intervention
are critical for helping workers reach middle-class jobs.

First, Indonesia could provide support to at-risk students to ensure they complete
secondary school. This chapter highlighted the role of a secondary education as a gateway
to a middle-class job. Although children are required to complete 12 years of basic education,
30 percent of children aged 16–18 are not in school. The gaps are even larger among youth
from the poorest households.60 Youth leave school for various reasons, including poverty,
early marriage, remote location of secondary schools, or having a disability that impedes
learning.61 A high-quality education system that confers clear benefits on youth and their
families will go a long way to keeping youth in education all the way through secondary
school. Such a reform agenda, summarized in Box 5.4, will take many years to realize.
57 “Labor market intelligence” can be defined as information about labor market trends that provides users with a
panorama of the overall labor market and how it is changing. Information may include the distribution of occupations in
specific locations, average earnings per occupation, the skills profiles of different jobs, unemployment rates by sector or
region, and job descriptions to help users who are selecting careers, identifying areas of study, considering relocating, or
considering a career change.
58 World Bank 2020d
59 Tehusijarana 2019
60 Based on the National Socioeconomic Survey (Survei Sosial Ekonomi Nasional; Susenas), almost 30 percent of
children aged 16 to 18 are excluded from school. Among enrolled students, 61 percent from the richest households reach
grade 12, but only 23 percent from the poorest households reach that level.
61 Poverty remains an important determinant in school enrollment, and is cited as a primary reason for school dropout
by about 50 percent of children. Early marriage is cited by 12.3 percent of girl dropouts. See World Bank (2020e).

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 219

Box 5.4 A Twelve-Step Plan to Boost Learning

Boosting learning requires more than a good curriculum or a star teacher. Instead, it requires developing
the building blocks for learning, while ensuring their interconnectedness and coordination among a range
of stakeholders. A recent assessment of the formal education system identified twelve goals for students,
teachers, management, and parents to improve learning.
1. Ensure that students reach at least minimum learning and development standards at each level of the
system
2. Ensure quality early-childhood education is accessible to all
3. Act to guarantee equitable access to good-quality education and learning for the children most excluded
from the system
4. Act to improve the learning outcomes of the lowest performers
5. Ensure that all students, including those with disabilities, succeed
6. Improve teacher recruitment, training, and professional development; calibrate incentives to increase
accountability
7. Improve the quality of preservice institutions and the candidates that enter them
8. Support existing institutions to improve service delivery
9. Strengthen accountability mechanisms
10. Expand access to and improve the quality and relevance of technical and vocational education and training
11. Raise the performance of tertiary education
12. Increase system resilience to future shocks

Source: World Bank 2020e

In the short run, policymakers can provide targeted support to at-risk youth to help them
complete secondary school. This entails developing a systematic protocol for mapping
youth who are at risk of dropping out and the reasons for this. With these data in hand,
national and local governments can identify the needs of specific subpopulations and fine-
tune response interventions.62 To some extent, this strategy is already being implemented
through the Program Keluarga Harapan, which provides conditional cash transfers to parents
whose children are in school,63 thus addressing poverty constraints that may lead to school
dropouts. However, children with learning disabilities, young girls subject to early marriage,
and children living in remote areas, for example, need different kinds of support. It is
increasingly important to expand the reach of the Paket C (a diploma equivalency program
for those aged over 18 who are unable to gain a diploma through traditional schooling) so
that adults can gain an upper secondary school diploma. This will require more proactive
outreach by program staff, as well as a greater use of distance learning techniques to
facilitate participation by the target audience, primarily adults.64

62 The list of potential interventions is long, ranging from providing facilities for children with physical impairments,
community programs to prevent early marriage, distance learning programs for small schools, or even redesigning the
Program Keluarga Harapan to better incentivize secondary school completion (Barrera Osorio et al. 2011).
63 Ministry of Social Affairs 2016
64 Herlyna, Mukhtar, and Sari 2018

Pathways to Middle-Class Jobs in Indonesia June 2021

220 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

Second, Indonesia could support the development of and access to online distance
learning in short TVET courses for adult workers. As discussed in this chapter, more
than 120 million workers will need to upskill and reskill throughout their 50 years of work.
They will not return to school and will receive few benefits from training offered by an
employer. They must rely on the public or private training systems, but these training
programs cannot be easily accessed by adult workers. They are time constrained, given
that they work, and have limited free time for study. They need to support families so
cannot take time off from their job. There is a limited number of providers of short-term
training and they are largely of unknown quality.65 Further, workers may not live or work
near training centers.

A greater use of training courses that are of short duration, time flexible, and can be
accessed online is a feasible way for workers to proactively continue learning. However,
workers often do not have information about online training programs, the quality of a
service, or the types of training that employers will value. The Kartu Prakerja program
launched at the outset of the COVID-19 pandemic is a first step in addressing this gap. The
program provides an e-voucher to be used for short-term online training; it is targeted at
the low-skilled segment of the population not in education. It has been widely advertised
and more than 20 million people have registered for the e-voucher. In addition, the
government can support the market for short-course online training by: (i) using the existing
TVET accreditation system, together with the occupational skills competency framework,
to assess the quality of web-based training courses and award accreditation accordingly;
(ii) providing public financing, such as training vouchers or individual (lifelong) learning
accounts, for adult learners, with greater incentives for training in critical occupations
or skills;66 (iii) incentivizing employers to provide more on-the-job training, for example by
providing (total or partial) reimbursements for investing in critical skills; and (iv) supporting
the development of a database of accredited online training courses.

Third, Indonesia could revise its curriculum and pedagogical methods to teach
nonroutine interpersonal, analytical, and digital skills to students and adult
learners. This chapter showed that both general education and TVET secondary schools
need to expand their instruction in applied subjects such as business, computers, and
English. Middle-class jobs require that these knowledge subjects are complemented by
interpersonal, analytical, routine cognitive, and digital skills.67 Similar to reading or math,
these skills can be taught and they are cumulative, meaning that learners begin with the
basics and build up to more complex areas.68

Classrooms and training institutions can teach a range of nonacademic skills through modest
adjustments to curricula and pedagogical methods. Interpersonal skills have successfully
been taught through direct instruction, indirect instruction (by incorporating interpersonal

65 World Bank 2020c

66 For example, Singapore established personal interest-earning training accounts that workers can access when
needed. Several European countries have legal instruments to promote adult training. They include leave for training
and payback clauses.
67 A growing literature finds that these skills are increasingly important in an automated workplace.
68 Guerra, Cunningham, and Modecki 2014

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 221

skill development into pedagogical methods or by training teachers in those skills so they
can model them for students), or through after-school activities (see Box 5.5). Appropriate
methodologies are sensitive to the age of the student and the context.69 Analytical skills
can also be directly or indirectly taught through school subject matter learned interactively.
Modern pedagogical methods are increasingly moving in this direction.70 The best method for
learning digital skills is still under debate, but there is a growing consensus that digital skills
require hands-on practice and can largely be learned through self-instruction. While ad hoc
efforts to teach nonacademic skills are underway in Indonesia, it is increasingly necessary to
systematize learning in these areas. This will require: (i) a strategy for developing analytical,
routine cognitive, interpersonal, and digital skills from early childhood through adulthood,
which identifies the skills to be taught and the way this will happen for each age group;
(ii) the development of pedagogical methods and the adaptation of learning materials to
implement the expanded curricula; (iii) the training of teachers to adopt these materials;
and (iv) a measurement protocol to track progress.

69 Acosta, Cunningham, and Muller 2016

70 Crawford, Hasan, and Bentaouet Kattan 2018

Pathways to Middle-Class Jobs in Indonesia June 2021

222 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

Box 5.5 Teaching Interpersonal Skills in Schools

The Collaborative for Academic, Social, and Emotional Learninga  keeps an inventory of evidence-based
interventions for developing interpersonal skills. The interventions can be summarized in five categories:

Developing socioemotional skills in teachers and principals so they model the behaviors in the classroom.
For example, the Peruvian Ministry of Education’s Escuela Amiga (Friendly School) program included a year-
long university-level course for teachers and principals to develop interpersonal skills. Nearly 15,000 classroom
teachers and 81 principals participated in the program’s first year (2013). Impact results are pending.

Improving the school environment so children and youth have a safe place to practice and reinforce
socioemotional skills. The internationally applied School-Wide Positive Behavior Support approach helps schools
define and establish a set of school norms, practices for reinforcing positive behavior, total support from all
school staff (not only teachers), and data-driven decision making. Randomized controlled trials show that the
approach positively affects student behavior.

Incorporating the development of socioemotional skills into the teaching of other subjects by altering
pedagogical methods. For example, Facing History and Ourselves is an organization that trains history teachers
in more than 110 countries to use analytical and interpersonal instruction in their teaching. The program has
a statistically significant impact on student capacity to analyze evidence, on prosocial behavior, tolerance,
empathy, and on knowledge of history.

Curricula, materials, teaching methodologies, and dedicated time for developing socioemotional
skills. For example, Mexico’s Construye-T (Build Yourself) program gives daily microlessons of 10 minutes to
secondary school students, helping them to practice a range of socioemotional skills.

After-school activities across a range of countries provide youth with experiences that help them acquire
teamwork, problem-solving, and other skills.b

Source: Adapted from Acosta, Cunningham, and Muller 2016

a Collaborative for Academic, Social, and Emotional Learning. N.d. https://casel.org/
b Guerra, Cunningham, and Modecki 2014

Fourth, Indonesia could improve the quality of its training system by prioritizing
the development of quality-assurance mechanisms and by more effectively
engaging the enterprise sector. Indonesia’s training system has more than 45,000 TVET
institutions, which are managed by dozens of ministries and levels of government, and serve
millions of students each year. Some good practices are in place, but the complex system is
not conducive to efficiently providing the training services that are demanded by the labor
market, particularly in an environment in which the skills profiles of jobs are rapidly evolving.
The government has identified critical occupations71 and the fastest-growing jobs,72 and
this report has identified fast-growing sectors aligned with middle-class jobs. However,
these labor market trends are not shaping the publicly managed training sector. Efforts

71 World Bank and CMEA 2020. The 2018 Critical Occupations List is a list of 35 occupations that were identified as
being in shortage and of strategic importance to the Indonesian economy.
72 World Bank and Bappenas 2021

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 223

to develop quality-assurance mechanisms are in motion, but the quality-control system is

still fragmented, duplicative, and incomplete. Both the education and training systems are
largely divorced from the business sector, which further exacerbates the mismatch between
education and training services and good jobs.73

The public training system needs to develop quality-assurance tools and mechanisms to
crowd in TVET in critical and growing occupations that will move workers up the jobs ladder.
A recent report highlights nine areas where the TVET system could be improved (see Box
5.6). The government is already on track, and is developing competency standards for
TVET providers.74 The next immediate steps could include: (i) assigning a single body the
responsibility of overseeing the completion of the development of the competency
framework and standardizing it across the public training system; (ii)  strengthening the
accreditation standards, guiding all parties (that is, training providers of all types, prospective
students, and employers) on the meaning and use of different accreditation levels as
they pertain to the quality of training services, and enforcing accreditation requirements
across the public and private training systems; (iii) developing a centralized monitoring and
evaluation system and requiring all training institutes to report input, output, and (eventually)
outcome data;75 and (iv) establishing a National Vocational Skills Steering Committee that, in
the long run, is responsible for allocating resources to TVET institutions based on evidence
of success, as measured by the monitoring and evaluation system, and by market demands
(as defined through critical occupation lists, and occupational and employment surveys).

Only by engaging employers can Indonesia create a world-class TVET system. Employers
are the ultimate consumers of the skills development system. They hire, and benefit from,
a more productive labor force. They are best positioned to guide the training system to
provide services that will benefit the labor market. However, few employers in any country
engage with a training system due to a range of factors,76 and Indonesia is no exception.
The government can more effectively engage employers in the training system by offering
incentives for industry to provide and finance training, and guide the public sector on how
to improve services. Any such incentives will need to be contingent on the quality of the
firm engagements with the training system. Specifically, Indonesia could improve Ministry
of Finance Regulation 128/2019,77 which subsidizes worker training in firms, by linking
reimbursement to results, measured by the number of workers who achieve an increase
in their level of skill certification. Further, Indonesia could also use public resources to
incentivize exchanges of staff between industry and providers of TVET. The government
could finance internships in firms for TVET instructors and subsidize the wages of industry
experts to allow them to take time out of their firm to teach in TVET institutions.

73 World Bank 2020c

74 Standard Kompetensi Kerja Nasional Indonesia
75 Singapore has established enrollment targets for public providers of training in certain fields of study. The aim is to
align training outcomes with the country’s broader workforce plans. It also closely monitors attrition rates and graduates’
employment and links funding to these results.
76 Hoftijzer and Cunningham. Forthcoming
77 The Ministry of Finance issued a tax levy through Regulation 128/2019 on a “Company’s Brute Income Deduction
Based on the Implementation of Practical Work, Internship, Teaching and Learning for Competency-Based Human Resource
Development.” This regulation provides the basis for tax deductions of up to 200 percent for companies supporting
workforce development programs.

Pathways to Middle-Class Jobs in Indonesia June 2021

224 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

Box 5.6 Priority Reforms to Build a World-Class Workforce Development

System in Indonesia
A recent assessment of the technical and vocational education and training (TVET) system, based on an
internationally used assessment tool, concluded that Indonesia’s workforce development system has several
good elements, but does not operate efficiently. The report identified nine areas for improvement:

1. Reforming the institutional architecture to increase the impact of government efforts and investment to
improve the entire system;
2. Pursuing more avenues to ensure that TVET institutions teach the skills that employers need;
3. Expanding vocational education and training for the workforce, as a complement to the efforts to build
the prework TVET system;
4. Providing financial support to accelerate the development of competency standards and national
qualification certification schemes;
5. Strengthening accreditation practices in Indonesia’s workforce development system for better quality
assurance;
6. Providing tools and capacity building for employment services, providing services such as career
counseling, to better support the transition from training to employment;
7. Establishing a more comprehensive set of targets to improve the quality of public training provision;
8. Developing a national strategy to increase the number of instructors and administrators at TVET
institutions to ease existing shortages;
9. Strengthening efforts to assess the labor market outcomes of training programs, for example the number
of graduates getting jobs, and thereby better managing Indonesia’s workforce development system.

Source: World Bank 2020b

Reform Area 2: Support Job Transitions through Information

and Financing
Workers will transition to better jobs if they have information about job openings and the
means to move from one job to another. However, most people find jobs through informal
personal networks. As noted, internal migration is somewhat limited. Allocative inefficiency
is evident as seen in the skills mismatch between worker skills and job requirements.

First, Indonesia could upgrade its job matching tool. Information on job vacancies is
available through an array of platforms. The government hosts the KarirHub (and previously,
the Ayo Kita Kerja), which is populated with job vacancies identified through the public
employment service. Because KarirHub serves few clients (few employers post jobs and few
jobseekers use it to find jobs) and does not rely on advanced technology such as machine-
learning algorithms, its efficacy is very low.78 Consequently, the government announced a
new matching tool, the SiapKerja, that would: (i) rely on more advanced matching technology
(artificial intelligence [AI] and skills driven); (ii) carry out massive outreach campaigns to

78 World Bank 2020d

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 225

promote its use; (iii) establish partnerships with private jobs boards; (iv) provide personalized
services for clients lacking digital skills or internet access; and (v) link to new programs
such as the Job Loss Guarantee (Jaminan Kehilangan Pekerjaan; JKP). Private websites, such
as Indeed.com or LinkedIn, tend to provide information about middle-class jobs. However,
these sites are incomplete, often contain information only for niche occupations, and are
unconnected to each other, thus requiring users to separately register for and search
individual sites. These sites are also rarely used by the majority of jobseekers, who continue
to prefer informal networks for job search.

A recent assessment of the Ayo Kita Kerja provides recommendations that can be used to build
a strong KarirHub, as well as the next tool, the SiapKerja.79 In the short run, the government
could consider: (i)  fostering the ecosystem around the labor market information system
by signing agreements with private jobs portals and with public portals providing services
for job search and career development (Ministry of Education, Ministry of Social Affairs).
The agreements should cover information sharing, collaborating with potential institutional
users (job counselors and training centers, among others), and providing information and
support services to micro, small, and medium enterprises (MSMEs) and growth-oriented large
firms to help them use the system; (ii) improving and expanding the technical infrastructure
of the SiapKerja platform, including server capacity and the cloud, and monitoring systems
while strengthening the quality and security of data by encrypting personal information,
controlling access, and establishing security protocols; and (iii) developing a protocol
to systematize and populate KarirHub using real-time information. This would entail:
(i) developing AI to trawl information sites and pull job vacancy information into the public
platform; (ii) mapping and integrating existing labor market–related information, including
administrative data and economic analysis; and (iii) analyzing and curating information
to identify critical occupations,80 occupations with bright prospects through Indonesia’s
Occupational Employment Outlook 2020,81 skills demand through tailored surveys (à la O*NET
or via an analysis of online vacancy data) (Box 5.7), and other labor market trends to steer
jobseekers toward higher-value jobs.

In the long run, new tools and partnerships can be developed for the labor market information
system to provide frequent, fine-tuned, and comprehensive information. Specific reforms
could include: (i) regularly conducting field surveys to collect job preparation or search
information that still does not appear on public data sites; (ii) establishing coordination
mechanisms to manage information flows and synergies across the system; and (iii)
deepening the online services and analysis available for career guidance and labor market
information overall, including incorporating into the system information about education,
social services, and other auxiliary services that may be useful to jobseekers.

79 World Bank 2020d

80 CMEA (Coordinating Ministry for Economic Affairs). 2020.
81 World Bank and Bappenas 2021

Pathways to Middle-Class Jobs in Indonesia June 2021

226 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

Box 5.7 Workforce Development Tools for Detailed Jobs Information

Labor market information is a valuable input to career preparation and lifelong learning. Students (and their
parents), jobseekers, and those who are interesting in changing jobs can plan better for their education,
training, and job searches if they have detailed information about the ebb and flow of specific occupations. Job
counselors can provide data-based advice on job preparation and search. Education and training institutes can
better design their course offerings with such information. Policymakers use the information to define a list of
critical occupations that in turn define financing and labor policy.

Indonesia’s current surveys provide limited information for people preparing for or seeking a job. To fill the
gap, the government and the World Bank developed four pilot tools for workforce development. One of these
pilots is the Occupational Employment and Vacancy Survey. This is a firm-level survey designed to identify
occupations that are major sources of employment and observe occupational dynamics—including shortages,
hiring attempts and difficulties faced, current vacancies, and planned occupational employment growth and
declines. The pilot is designed to inform about job prospects in all occupations present in selected sectors and
regions in Indonesia.a The results are presented in Indonesia’s Occupational Employment Outlook 2020,b with ten
indicators that assess the current state of occupations in Indonesia and must be interpreted holistically. The
Indonesian government plans to adopt the pilot and run it nationally in the coming years.

The ten indicators are consolidated in an occupational score. Indicators are presented at up to four levels of
occupational grouping using the Indonesian Standard Classification of Occupations (KBJI-2014). The scores are
categorized into four groups: (i) “bright” occupations for which there is high demand based on all indicators; (ii)
“steady state” occupations that continue with normal trends of employment growth; (iii) “dim” occupations for
which there is either no demand or demand is shrinking; and (iv) “flagged” occupations for those cases where
there is insufficient data to make an accurate assessment. The findings of the first Occupational Employment
and Vacancy Survey are presented in Table B5.7.1.

The “brightest” occupations among the selected occupations and regions are roles in financial services, as well
as engineers, architects, designers, and environmental and occupational health. All are strongly associated with
middle-class jobs. Steady state jobs are in the social sectors (health, education) and, while not overwhelmingly
middle-class jobs, they tend to have job stability and greater social benefits; these are also occupations that
women cluster into. The dim occupations cut across occupations (professional, social, science, managers), some
being middle class and others less so.

a The pilot survey was fielded in selected high value-added services, low value-added services strategic for Indonesia’s growth, and selected activities of
the manufacturing industry in Java and Bali. It was also fielded for the manufacturing sector in North Sumatra, West Java, Banten, Central Java, East Java,
and South Sulawesi.
b World Bank and Bappenas 2021

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 227

Box 5.7 continued

Table B5.7.1 Findings of the First Occupational Employment and Vacancy Survey

Occupational score
Score Rationale Number of Examples of four-digit occupations
four-digit
occupations
Steady Occupations that continue with 122 2221-Nursing professionals
state normal trends of employment 2262-Pharmacists
growth.
2341-Primary school teachers
2342-Early-childhood educators
2412-Financial and investment advisers
Dim Occupations for which there is no 90 2422-Policy administration professionals
demand or for which demand is 2635-Social work and counseling professionals
shrinking.
2636-Religious professionals
3119-Physical and engineering science technicians
not elsewhere classified
3121-Mining supervisors
Flagged Occupations for which there 92 3312-Credit and loans officers
is insufficient data to make an 3313-Accounting associate professionals
accurate assessment.
3315-Valuers and loss assessors
3321-Insurance representatives
Source: World Bank and Bappenas 2021

Indonesia’s Online Vacancy Outlookc is designed to provide real-time analysis and granular information on
occupations in demand and the associated skills needed. It collects online data on job postings from up to 25
jobs boards and 100 employer sites in Indonesia, providing a real-time picture of the labor market situation
for occupations advertised online and a highly granular description of the skills they demand. This tool is
designed to complement other initiatives, since online job vacancies mainly cover high-skilled and formal jobs.
The most demanded skills category in 2020 was project management and business. Half the posts demanded
project management skills, with an average of two project management skills demanded per job post. Project
management and business skills are a broad category, with more than one thousand unique skills identified
within this group. Of these, planning, marketing, and business development are the most sought-after skills. This
result partly reflects the fact that online job postings are used for the high-skilled segment of the labor market.

Two other pilots were conducted to complement Indonesia’s Occupational Employment Outlook 2020 and measure
the skills and tasks demanded by occupations in demand. Indonesia’s Occupational Tasks and Skills (Indotask)d
is the first attempt to adapt to the Indonesian context a globally recognized instrument that measures skills
and tasks at a detailed level. Indotask adapts two modules of the globally recognized tool O*NET to describe
the tasks involved in performing an occupation and the skills needed to do so. More specifically, it retrieves
skills and task data of occupations in high demand relying exclusively on a questionnaire delivered online.
Indotask draws on the Occupational Employment and Vacancy Survey and Indonesia’s Online Vacancy Outlook

c Progress is reported in World Bank (2021b).

d Progress on Indotask is reported in World Bank (2021a).

Pathways to Middle-Class Jobs in Indonesia June 2021

228 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

Box 5.7 continued

to identify occupations in high demand. It is likely able to provide a more comprehensive view of labor
markets and skills in Indonesia than can big-data methodologies, given that online vacancies are only found
in certain large urban conglomerates and for certain occupations. A first key finding is that the list of tasks
and skills involved in occupations differ between Indonesia and the United States (from where the O*NET
originates),corroborating the importance of developing local instruments as the production functions vary
across countries. A second important key finding is that the skills most in demand in Indonesia—at least among the
selected occupations in demand—are basic skills such as speaking, reading comprehension, and active listening.

Note: The KBJI is the Indonesian Standard Classification of Occupations (Klasifikasi Baku Jenis Indonesia). Bappenas is the Ministry of National
Development Planning (Kementerian Perencanaan Pembangunan Nasional).

Second, Indonesia could support job mobility by helping workers to relocate. For
workers to benefit from the industrial revolution 3.0, many will have to change jobs. This
chapter argues that job switching is costly. Indonesia’s Labor Law limits job mobility.82
Monetary costs—being out of work without income, job-search costs—and the psychological
costs of moving may hinder job mobility. Even searching for a job while still working is costly
since there are few tools to facilitate the process. Physical relocation may be necessary
to acquire a middle-class job. Job opportunities, not least those created by foreign direct
investment (FDI), are geographically clustered in Indonesia, while the population is more
dispersed. Not surprisingly given the costs, migration for employment is relatively low. To
facilitate job mobility, policies need to support information about job opportunities across
the country, and provide financing and programs to help workers with physical relocation.
This may include: (i) ensuring the labor market information system provides national (and not
just local) information on job vacancies across Indonesia; (ii) training social and employment
workers in government offices (dinas; and especially in dinas that are far from Jakarta) to
provide information about work opportunities in other regions; and (iii) providing a stipend
for workers to move to jobs in critical occupations.83

Third, Indonesia could design a process to shift from a severance pay system that
rewards tenure to an unemployment insurance system that cushions income shocks
without creating disincentives to job separation. Severance payments have generally
not delivered the intended protections; few workers collect the benefits and the expected
costs may be internalized by firms, leading them to forego permanent hires for fixed-term
contracts.84 However, the recent creation of the Job Loss Guarantee (Jaminan Kehilangan
Pekerjaan; JKP) is a step toward greater support for workers to help them transition up the
jobs ladder. The process of implementing the JKP will need to be timed well so that workers
are not unprotected during the transition to the new system. An option is to mandate JKP
for workers with new contracts, while allowing workers with existing contracts to opt in

82 Weber and Kudo (2020) argue that regulating severance pay discourages worker mobility. Severance pay is awarded
to workers in Indonesia who are dismissed (but not those who voluntarily leave their job) and becomes quite generous
with tenure. As a result, voluntarily leaving a job can be quite costly for workers.
83 For example, a program in Bangladesh provided a transport subsidy for seasonal migrants to relocate to urban
centers. An impact evaluation found that the incentive led to an increase in migration, an increase in consumption by
the migrant’s household, and an increased likelihood of subsequent migration (Bryan, Chowdhury, and Mobarak 2014).
84 Brusentsev, Newhouse, and Vroman 2012

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 229

(or out). The opt-in arrangement fully acknowledges the prevailing rights of workers under
their old contracts, while giving them the choice to be part of the new scheme. A clear
communication to society about the benefits of the reform—it offers workers greater
certainty about their income, helps employers anticipate the costs of dismissals, and reduces
the risk of employers defaulting on their redundancy obligations—is crucial for a successful
implementation. Importantly, the JKP financing will need to be spread across employers
and employees, with careful consideration of the contribution rates to ensure there is no
introduction of perverse incentives (for example, firing workers or cutting wages, evasion)
into the employment relationship. To that end, it will be critical for Indonesia to monitor and
adjust the level and duration of benefits as it learns how the program incentives affect job
search and hiring choices. For the JKP to be effective, it should be accompanied by a strong
monitoring system to ensure JKP recipients are looking for and taking up jobs.

Reform Area 3: Provide Tailored Support to Special Groups

Some groups face additional challenges that raise their risk of being left behind in the
transition to middle-class jobs. Gender-related factors limit women’s access to middle-
class jobs. Youth are at a disadvantage due to the premium earned by work experience,
which they have yet to acquire. Targeted policies are necessary to bring these groups into
prosperity through jobs. Some, such as tailoring the services provided by the labor market
information system by group or earmarking budget for training, cut across gender and age.
Others, including legislation and enforcement of certain labor rights, focus on gender.

First, Indonesia could fine-tune its labor market information system and training
programs to address factors that limit the success of women and youth in acquiring
middle-class jobs. Youth use a range of job-search methods and take longer to find a job,
but their lack of experience (and the skills acquired therein) make it difficult to acquire
middle-class jobs. Relative to men, women engage in less job search and use more informal
methods, study in less lucrative professions, and have a higher tendency for urban self-
employment.

This range of challenges can be addressed through fine-tuning the labor market information
system and the overall support given to those undertaking job preparation and search.
In the short run, it will be important to: (i)  include gender- and age-relevant information
in the labor market information system, such as nonstandard fields of study that women
can succeed in, entry-level jobs with career progression to middle-class roles, and auxiliary
services such as childcare or student loans; (ii) earmark a portion of training vouchers for
women and youth to participate in courses that teach business practices and technology,
including those needed for engaging in the gig and green economies; and (iii) incorporate
business development training into entrepreneurship support programs for women,
teaching managerial skills, the skills needed to tap into markets and create market links, and
entrepreneurship behavioral skills to facilitate links with emerging value chains.85 In the long

85 Rizvi 2018

Pathways to Middle-Class Jobs in Indonesia June 2021

230 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

run, it will be important to: (i) include an interactive guidance portal targeted toward career
development to help youth and women, who are less experienced jobseekers, to identify
their interests and career ambitions, to walk them through the process needed to achieve
their job goals, and to direct them toward services, such as additional training, relocation
programs, active labor market programs, job-search sites, or vacancy information; and (ii)
provide scholarships and employment vouchers for women’s study and first job in science,
technology, engineering, or mathematics (STEM) fields, and incentives for employers who
hire women in STEM jobs; and (iii) provide vouchers to vulnerable youth, through systems
such as Kartu Prakerja, to purchase short-course training from accredited (virtual or face-to-
face) training providers that teach skills associated with critical occupations or middle-class
jobs.

Second, Indonesia could expand the access women have to social benefits through
their jobs. Women in middle-class jobs have greater access to family-friendly benefits than
do men. They may forego income in favor of such benefits, which include social benefits,
limited work hours, and paid leave. Indonesia’s current policy of financing maternity benefits
through the social security fund (rather than via employers) is noteworthy. However,
women have high labor force dropout rates at pregnancy and immediately after childbirth.
Indonesian women with access to affordable childcare, whether institutional or through
older family living in the household, are more likely to work as employees than are women
without such access. To further support social benefits and encourage the employment
of women in middle-class jobs, Indonesia could: (i) modify the Labor Law to specify that an
employer may not terminate the employment of, or take disciplinary measures against, a
woman during pregnancy, maternity leave, or in the period following her return to work;
and (ii) establish a national procedure for grievances related to firm payment of social
benefits. The procedure should provide a confidential and gender-friendly service through
which workers can report violations of the Labor Law, which guarantees social benefits
for workers. The provision of accessible and affordable childcare is also found to increase
women’s labor force participation.86

Third, Indonesia could enact legislation to set national norms for women’s access
to the labor market. National laws set standards that can change social norms. Indonesia’s
regulations on the labor market have specific provisions to protect men and women, but
some barriers are built in to them. The Labor Law defines generous maternity leave, but
specifies few days of paternity leave, implicitly reinforcing gender norms that place women
as the primary caregiver in the household. This pales relative to global standards and even
to practices within Indonesia where, for example, male civil servants are eligible for one
month’s paternity leave.87 Sexual harassment is not sufficiently defined in legislation, and
Indonesia is one of the few countries that does not specify equal pay for equal work. Recent
reviews identify policies that can improve the gender dimensions of the Labor Law.88

86 Halim, Johnson, and Perova 2017

87 National Civil Service Agency Regulation No. 24/2017
88 Aeberhard-Hodges 2020; Weber and Kudo 2020

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 231

The Labor Law can be modified to reestablish gender norms in the work sphere. Specifically:
(i) The language relating to “equal pay for equal work” could be clearly defined, and this could
include how to evaluate the value of work and the nature of “pay;” (ii) The government could
introduce an article that defines sexual harassment and clarifies the nature of unacceptable
workplace interactions; and (iii) The government could expand paternity benefits to 30 days,
thus putting the private sector on par with the public sector, while encouraging fatherhood
and reducing the cost differential of employing male versus female employees. It could
finance this through the social security program. In the long term, Indonesia could develop
legislation and legal precedents.

Annex Chapter 5
Annex 5.1 Definition of a Middle-Class Job 
The definition of “middle-class job” used in this report is subject to the data available in
the  Sakernas. The authors assume that a  middle-class  job pays the amount needed by a
family of four (comprising one full-time adult worker, one part-time adult worker, since only
50 percent of Indonesian women work, and two dependent nonworkers). A middle-class job
can be defined with three variables (or a mix of them): 

By wage/income. The Sakernas provides earnings income for wage employees, the self-
employed, and casual workers from 2001–18. This excludes unpaid workers (whose income
is difficult to impute) and employers (who earn profits, which are also difficult to impute).
Under this definition,  a  middle-class  job pays at least Rp3,752,000  monthly in 2018
prices. This estimate is derived as follows: 

Minimum MIDDLE-CLASS income threshold = (3.5*poverty line*4)/1.5 

where the national poverty line is Rp400,000 in 2018. Under this definition, 21.1 percent of

wage employees can be classified as holding middle-class jobs, while 15 percent of all paid
workers (all workers excluding the unpaid and employers) hold middle-class jobs.

By an indefinite work contract. Jobs with indefinite work contracts offer psychological

benefits since the fear of losing a job is significantly diminished. As of 2016, the Sakernas
includes information on the nature of the contract held by the worker. Approximately 19.1
percent of wage employees have an indefinite work contract. Half of them, however, do not
earn at least a middle-class income. 

By social benefits. Jobs are often accompanied by a range of social benefits that protect
workers from economic vulnerability. Some are mandated by law, and even informal firms
offer some of these. As of 2016, Sakernas allows us to define  middle-class  jobs as any
combination of: health insurance, work accident insurance, death benefits, old-age benefits,
pension benefits, and/or paid annual leave / sick leave / maternity leave. About 12.6 percent
of employees receive full benefits. Less than 2 percent of employees receive benefits but
not a middle-class wage or an indefinite work contract. 

Pathways to Middle-Class Jobs in Indonesia June 2021

232 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

Not all benefits are received equally by all  middle-class  (or non-middle  class) jobs.  Health
insurance is the most common benefit, with  86.4 percent of those in  middle-class  jobs
receiving it, as compared to 28.9 percent of those in a non-middle-class job. But only a little
more than half of workers with a middle-class job are paying into a pension system, which
may threaten their middle-class status once they stop working. 

The three definitions lead to very different designation of jobs as middle class. Under the
wage/income definition, 15 percent of jobs can be considered middle class while 21.1 percent
of wage employees (paid workers, excluding the self-employed) are earning at least Rp3.75m
monthly. If we use a more restrictive definition of holding a job and receiving benefits, only
7.1 percent of wage employment is considered middle class.89

Annex Table 5.1 Share of Workers Who Receive Each Benefit, by Middle-Class Job Status

Middle-class job No middle-class job

Health insurance 86.4% 28.9%

Work accident benefit 78.4% 24.5 %
Death benefit 68.5 % 16.6 %
Old-age benefit (nongovernment) 62.1 % 13.1 %
Pension (government) 56.3% 9.7%
Paid annual leave 75.2 % 23.0 %

Sources: World Bank staff calculations using Sakernas 2018

Note: Sakernas is the National Labor Force Survey (Survei Angkatan Kerja Nasional)

89 Since non-employees do not have work contracts or pay mandated social benefits, by definition, they would not
be included under the more restrictive definition. However, if they are included in the denominator, 4.1 percent of jobs
would be considered middle class. 

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs 233

Annex 5.2: Additional Tables

Annex Table 5.2 Determinant of Middle-Class Jobs, Average Marginal Effects

Dependent variable =1 if a middle-class jobholder

   (1) (2) (3) (4) (5) (6)

Control Control sector,
Control Control Control job
VARIABLES  sector and occupation, job
sector occupation type
occupation type
                    
years of schooling  0.022*** 0.021*** 0.015*** 0.02*** 0.015*** 0.0147***
  (0.00024) (0.00027) (0.00028) (0.00024) (0.00028) (0.000280)
=1 if receive training  0.043*** 0.04*** 0.03*** 0.039*** 0.032*** 0.0314***
  (0.0019) (0.0019) (0.0019) (0.0019) (0.00185) (0.00184)
potential experience  0.0062*** 0.0061*** 0.0058*** 0.0063*** 0.0061*** 0.00615***
  (0.00020) (0.0002) (0.00019) (0.0002) (0.00019) (0.000188)
-7.33e- -7.06e- -7.33e-
potential experience^2  -7.29e-05*** -7.45e-05*** -7.47e-05***
05*** 05*** 05***
  (3.83e-06) (3.73e-06) (3.58e-06) (3.74e-06) (3.56e-06) (3.56e-06)
=1 if male  0.051*** 0.05*** 0.048*** 0.051*** 0.0453*** 0.0441***
  (0.0016) (0.0016) (0.0016) (0.0016) (0.0016) (0.00163)
=1 if live in urban  0.069*** 0.061*** 0.061*** 0.062*** 0.058*** 0.0563***
  (0.0015) (0.0015) (0.0015) (0.0015) (0.0015) (0.00149)
=1 if disabled  -0.0043 -0.0038 -0.0023 -0.0041 -0.0025 -0.00252
  (0.003) (0.0029) (0.0029) (0.003) (0.0029) (0.00285)
=1 if mining and quarrying  0.12*** 0.1*** 0.0911***
  (0.0037) (0.0049) (0.00516)
=1 if manufacturing  0.0495*** 0.0430*** 0.0280***
  (0.00277) (0.00436) (0.00462)
=1 if electricity, gas, and water
0.0374*** 0.0113 -0.00222
supply 
  (0.00695) (0.00762) (0.00774)
=1 if construction  -0.00217 0.0125*** 0.0136***
  (0.00337) (0.00447) (0.00501)
=1 if wholesale and retail
0.0239*** 0.00626 -0.00454
trade, restaurants and hotels 
  (0.00270) (0.00437) (0.00462)
=1 if transportation, storage
0.0344*** 0.00968** 0.000800
and communications 
  (0.00325) (0.00473) (0.00500)
=1 if finance, insurance, real
0.0782*** 0.0330*** 0.0212***
estate and business services 
  (0.00353) (0.00487) (0.00505)
=1 if community, social and
0.0315*** -0.0254*** -0.0375***
personal services 
  (0.00255) (0.00425) (0.00451)

Pathways to Middle-Class Jobs in Indonesia June 2021

234 Workers, Jobs, and Barriers to Entry to Middle-Class Jobs

Annex Table 5.2 continued

=1 if armed forces occupations  0.196*** 0.229*** 0.224***
  (0.00579) (0.00641) (0.00641)
=1 if managers  0.0952*** 0.0962*** 0.0935***
  (0.00327) (0.00349) (0.00349)
=1 if professionals  0.0448*** 0.0676*** 0.0643***
  (0.00247) (0.00311) (0.00314)
=1 if technicians and associate
0.0798*** 0.0751*** 0.0712***
professionals 
  (0.00285) (0.00321) (0.00324)
=1 if clerical support workers  0.0574*** 0.0611*** 0.0570***
  (0.00252) (0.00293) (0.00298)
=1 if skilled agricultural,
-0.0195*** -0.0132*** -0.0184***
forestry and fishery workers 
  (0.00286) (0.00487) (0.00506)
=1 if craft and related trades
-0.0209*** -0.0423*** -0.0386***
workers 
  (0.00286) (0.00351) (0.00349)
=1 if plant and machine
0.0271*** 0.00792** 0.00747**
operators and assemblers 
  (0.00274) (0.00331) (0.00326)
=1 if elementary occupations  -0.0618*** -0.0679*** -0.0622***
  (0.00286) (0.00340) (0.00336)
=1 if employee  0.0255*** 0.0109***
  (0.00164) (0.00196)
=1 if casual worker in
-0.0901*** -0.0499***
agricultural sector 
  (0.00649) (0.00647)
=1 if casual worker in non-
-0.0769*** -0.0486***
agricultural sector 
  (0.00445) (0.00461)
Observations  330,709 330,709 330,709 330,709 330,709 330,709
Note: Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 
Heckman probit regression. The dependent variable for the selection equation is “working;” independent variables are: years of schooling,
training dummy, potential experience and the quadratic, sex dummy, urban dummy, and disability dummy. The unweighted estimates are
presented; weighted estimates are available from the authors.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 References 235

References
Acemoglu, Daron, and David Autor. 2011. “Skills, Tasks and Technologies: Implications for Employment
and Earnings.” In Vol. 4B of Handbook of Labor Economics, edited by Orley Ashenfelter and
David Card, 1043–71. San Diego, CA: North Holland 2011.

Acosta, Pablo, Wendy Cunningham, and Noel Muller. 2016. Minds and Behaviors at Work: Socioemotional
Skills for Latin America’s Workforce. Washington, DC: World Bank. https://openknowledge.
worldbank.org/handle/10986/24659.

Aeberhard-Hodges, Jane. 2020. “Technical Memorandum: Gender Analysis of Indonesia’s Manpower

Act No. 13/2003.” Technical memorandum. World Bank. Washington, DC.

Alatas, Vivi, and Lisa A. Cameron. 2008. “The Impact of Minimum Wages on Employment in a Low-
Income Country: A Quasi-Natural Experiment in Indonesia.” Industrial and Labor Relations
Review 61 (2): 201–23.

Artiawati. 2007. “The Work-Family Interface in Indonesia.” In The Work Family Interface in Global
Context, editd by Karen Korabik, Zeynep Aycan, and Roya Ayman. New York: Routledge.

Barrera-Osorio, Felipe, Marianne Bertrand, Leigh L. Linden, and Francisco Perez-Calle. 2011. “Improving
the Design of Conditional Transfer Programs: Evidence from a Randomized Education
Experiment in Colombia.” American Economic Journal: Applied Economics 3 (2): 167–95.
https://pubs.aeaweb.org/doi/pdfplus/10.1257/app.3.2.167

Barro, Robert J., and Jong-Wha Lee. 2010. “A New Data Set of Educational Attainment in the World,
1950-2010.” Working Paper No. 15902. National Bureau of Economic Research, Cambridge,
MA.

Bell, David, and David Blanchflower. 2010. “Young People and Recession: A Lost Generation?” Working
Paper. Dartmouth College, Hanover, NH.

Bratsberg, Bernt, Oddbjörn Raaum, and Knut Røed. 2006. “The Rise and Fall of Immigrant
Employment: A Lifecycle Study of Labor Migrants to Norway.” Oslo: Frisch Centre for
Economic Research.

Bryan, Gharad, Shyamal Chowdhury, and Ahmed Mushfiq Mobarak. 2014. “Underinvestment in a
Profitable Technology: The Case of Seasonal Migration in Bangladesh.” Econometrica 82 (5):
1671–748.

Brusentsev, Vera, David Newhouse, and Wayne Vroman. 2012. “Severance Pay Compliance in
Indonesia.” Policy Research Working Paper 5933, World Bank, Washington, DC.

Calì, Massimiliano, Taufik Hidayat, and Claire H. Hollweg. 2018. “Determinants of Labor Mobility Costs in
Indonesia.” World Bank

Cameron, Lisa, Diana Contreras Suarez, and William Rowell. 2019. “Female Labour Force Participation
in Indonesia: Why Has it Stalled?” Bulletin of Indonesian Economic Studies 55 (2): 157–92.

Chowdhury, Iffat, Elizaveta Perova, Aneesh Mannava, and Hillary C. Johnson. 2018. “Gender Gap
in Earnings in Vietnam: Why Do Vietnamese Women Work in Lower Paid Occupations?”
Policy Research Working Paper 8433, World Bank, Washington, DC. https://ssrn.com/
abstract=3176321.

Chun, Natalie, and Niny Khor. 2010. “Minimum Wages and Changing Wage Inequality in Indonesia.” ADB
Economics Working Paper Series No. 196, Asian Development Bank, Manila.

CMEA (Coordinating Ministry for Economic Affairs). 2020. “Kemenko Perekonomian Bersama World
Bank Meluncurkan Critical Occupation List (COL) Tahun 2018/2019.” April 29. https://ekon.
go.id/publikasi/detail/258/kemenko-perekonomian-bersama-world-bank-meluncurkan-
critical-occupation-list-col-tahun-20182019.

Pathways to Middle-Class Jobs in Indonesia June 2021

236 References

Crawford, Michael F., Amer Hasan, and Raja Bentaouet Kattan. 2018. Growing Smarter: Learning and
Equitable Development in East Asia and Pacific. World Bank East Asia and Pacific Regional
Report. Washington, DC: World Bank.

Cunningham, Wendy. 2007. Minimum Wages and Social Policy: Lessons from Developing Countries.
Washington DC: World Bank Group.

Del Carpio, Ximena, and Laura Pabón. 2014. Minimum Wage Policy: Lessons with a Focus on the ASEAN
Region. Washington DC: World Bank.

Demombynes, Gabriel, and Mauro Testaverde. 2018. “Employment Structure and Returns to Skill in
Vietnam: Estimates Using the Labor Force Survey.” Policy Research Working Paper 8364.
World Bank, Washington, DC. https://openknowledge.worldbank.org/handle/10986/29458
License: CC BY 3.0 IGO

Duryea, Suzanne. 2012. “Economic Crisis and School-to-Work Transition.” In Children and Youth in Crisis:
Protecting and Promoting Human Development in Times of Economic Shocks, edited by Mattias
Lundberg and Alice Wuermli. Washington, DC: World Bank.

Elder, Todd E. & Goddeeris, John H. & Haider, Steven J., 2010. “Unexplained gaps and Oaxaca-Blinder
decompositions,” Labour Economics 17(1): 284-290.

Fagertun, Anette. 2017. “Labour in Paradise: Gender, Class and Social Mobility in the Informal Tourism
Economy of Urban Bali, Indonesia.” Journal of Development Studies 53 (3): 331–45.

Gomez-Mera, Laura, and Claire H. Hollweg. 2018. “Firm Performance and Constraints in Indonesia.”
Unpublished paper, World Bank.

Gregg, Paul, and Emma Tominey. 2005. “The Wage Scar from Youth Unemployment.” Labour Economics
4 (12): 487–509. 

Gregory, Mary, and Robert Jukes (2001). “Unemployment and subsequent earnings: estimating
scarring among British Men 1984–94.” Economic Journal 111 (475): 607–25.

Guerra, Nancy, Wendy Cunningham, and Kathryn Modecki. 2014. “Developing Social-Emotional Skills
for the Labor Market: The PRACTICE Model.” Policy Research Working Paper 7123, World
Bank, Washington, DC.

Halim, Daniel, Sean Hambali, and Ririn Purnamasari. Forthcoming. “Assessing the Gendered Impacts of
the COVID-19 Pandemic on Indonesia’s Labor Market.”

Halim, Daniel, Hillary Johnson, and Elizaveta Perova. 2017. “Preschool Availability and Female Labor
Force Participation: Evidence from Indonesia.” Washington, DC: World Bank.

Harrison, Ann, and Jason Scorse. 2010. “Multinationals and Anti-sweatshop Activism.” American
Economic Review 100 (1): 247–73.

Herlyna, Mukhneri Mukhtar, and Eliana Sari. 2018. “The Evaluation of Paket C Equivalency Program
Implementation in Tangerang District, Indonesia.” Advances in Social Science, Education and
Humanities Research 178: 567–73.

Hoftijzer, Margo, and Wendy Cunningham. Forthcoming. “Promoting Employer Engagement in Skills
Development: A framework to Address Constraints to Promote the Effective Engagement
of Enterprises in Strengthening Workforce Skills.” Social Protection and Jobs Global
Practice/East Asia and Pacific Region. Washington, DC: World Bank.

Hohberg, Maike, and Jann Lay. 2015. “The Impact of Minimum Wages on Informal and Formal Labor
Market Outcomes: Evidence from Indonesia.” IZA Journal of Labor and Development 4 (1): 14.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 References 237

Hollweg, Claire, Maria Monica Wihardja, Massimiliano Calì, Milan Nedeljkovic, Milan, Julian Latimer
Clarke, and Brasukra Gumilang Sudjana. 2016. Indonesian Economic Transformation and
Employment: Policy Input for an Indonesia Jobs Strategy. Washington, DC: World Bank Group.
http://documents.worldbank.org/curated/en/894021469437474890/Indonesia-Economic-
transformation-and-employment-policy-input-for-an-Indonesia-jobs-strategy

Indraswari. 2006. “Women and Warung in an Urban Kampung.” PhD thesis, The Australian National
University.

Kahn, Lisa. 2010. “The Long-Term Labor Market Consequences of Graduating from College in a Bad
Economy.” Labour Economics 17 (2): 303–16.

Lain, Jonathan, Hamidah Alatas, and Diahhadi Setyonaluri Setyonaluri. 2020. “Women at Work in
Indonesia.” Background paper, Pathways to Middle-Class Jobs in Indonesia, World Bank, Jakarta.

Montenegro, Claudio, and Harry Patrinos. 2014. ”Comparable Estimates of Returns to Schooling
around the World.” Policy Research Working Paper 7020, World Bank, Washington, DC.

Ministry of Social Affairs. 2016. “Pedoman Pelaksanaan Program Keluarga Harapan.”

Muro, Mark, Sifan Liu, Jacob Whiton, and Siddharth Kulkarni. 2017. Digitalization and the American
Workforce. Brookings India. http://hdl.handle.net/11540/7892.

Naidoo, Darian, Truman Packard, and Ilmiawan Auwalin. 2015. “Mobility, Scarring and Job Quality in
Indonesia’s Labor Market.” Policy Research Working Paper 7484, World Bank, Washington, DC.

OECD. 2016. Skills Matter: Further Results from the Survey of Adult Skills. OECD Skills Studies. Paris: OECD
Publishing.

OECD. 2019. SIGI 2019 Global Report: Transforming Challenges into Opportunities, Social Institutions and
Gender Index. Paris: OECD Publishing. https://doi.org/10.1787/bc56d212-en.

Oreopoulos, Philip, Till von Wachter, and Andrew Heisz. 2006. “The Short- and Long-Term Career
Effects of Graduating in a Recession: Hysteresis and Heterogeneity in the Market for
College Graduates.” NBER Working Paper No. 12159, National Bureau of Economic Research,
Cambridge, MA. https://www.nber.org/papers/w12159. 

Pradana, Abror Tegar, Grace Hadiwidjaja, and Josefina Posadas. 2021. “Labor Market Dynamics during
the COVID-19 Pandemic.” World Bank internal document, Jakarta.

Pradana, Abror Tegar, and Maria Monica Wihardja. Forthcoming. “Trends and Differentiated Returns
of Digital Task Content of Jobs in Indonesia.” Background paper, Pathways to Middle-Class Jobs
in Indonesia, World Bank, Jakarta.

Pratomo, Devanto Shasta. 2020. “Labor Supply and Policies in Expanding the Middle Class in
Indonesia.” Background paper, Pathways to Middle-Class Jobs in Indonesia, World Bank, Jakarta.

Rizvi, Anam. 2018. “Integrating Household Enterprises into the Industrial Landscape Through
Targeted Microenterprise Upgrading.” Unpublished manuscript, Social Protection and Jobs
Global Practice, World Bank, Washington, DC.

Roberts, Mark, Frederico Gil Sander, and Sailesh Tiwari, eds. 2019. Time to ACT: Realizing Indonesia’s
Urban Potential. Washington, DC: World Bank. doi:10.1596/978-1-4648-1389-4. License:
Creative Commons Attribution CC BY 3.0 IGO

Székely, Miguel, and Jonathan Karver. 2015. “Youth out of School and out of Work in Latin America:
A Cohort Approach.” Policy Research Working Paper Series 7421, Washington, World Bank,
Washington, DC.

Pathways to Middle-Class Jobs in Indonesia June 2021

238 References

Tehusijarana, Karina M. 2019. “ ‘The Main Thing is Not the Process, but the Result”: Jokowi’s
Full Inauguration Speech.” Jakarta Post. October 20. https://www.thejakartapost.
com/news/2019/10/20/the-main-thing-is-not-the-process-but-the-result-jokowis-full-
inauguration-speech.html.

United Nations, Department of Economic and Social Affairs, Population Division. 2019. “World
Population Prospects 2019, Volume II: Demographic Profiles.” https://population.un.org/
wpp/Graphs/1_Demographic%20Profiles/Indonesia.pdf

Weber, Michael, and Arvo Kudo. 2020. “Labor Regulations in Indonesia: An Assessment.” Internal
document, Social Protection and Jobs Global Practice, World Bank, Jakarta.

Witoelar, Firman, Alexander D. Rothenberg, T. Yudo Wicaksono, Tadeja Gracner, and Bondan Sikoki.
2017. “How Jakarta’s Traffic Affects Labor Market Outcomes for Women and People with
Disabilities: Results from a Baseline Survey.” Final Report, Australia Indonesia Partnership for
Economic Governance.

World Bank. 2020a. “The Jobs Response to Covid-19 in East Asia Pacific.” Note prepared by Social
Protection and Jobs Global Practice, World Bank. Version May 27, 2020.

World Bank. 2020b. No One Left Behind: Rural Poverty in Indonesia. Washington, DC: World Bank.

World Bank. 2020c. SABER-WfD Revised Indonesia Report. Washington, DC: World Bank.

World Bank. 2020d. “Toward a World-Class Labor Market Information System for Indonesia: An
Assessment of the System Managed by the Ministry of Manpower.” Washington, DC: World
Bank.

World Bank. 2020e. The Promise of Education in Indonesia. Washington, DC: World Bank.

World Bank. 2021a. Indonesia’s Occupational Tasks and Skills 2020: From Occupational Employment
Demand to Tasks and Skills Requirements: 2020.” Washington, DC: World Bank.

World Bank. 2021b. Indonesia’s Online Vacancy Outlook 2020: Technical Report: From Online Job Postings
to Labor Market Intelligence. Washington, DC: World Bank.

World Bank. Forthcoming 2021. Building Blocks for Advanced Labor Market Intelligence Services in
Indonesia. Washington, DC: World Bank.

World Bank. N.d. “Human Capital Index 2020.” https://databank.worldbank.org/data/download/hci/

HCI_1pager_IDN.pdf.

World Bank, and Bappenas (National Development Planning Agency; Badan Perencanaan
Pembangunan Nasional). 2021. Indonesia’s Occupational Employment Outlook 2020: Statistics
from the Occupational Employment and Vacancy Survey. Social Protection and Jobs Global
Practice. Washington, DC: The World Bank.

World Bank and CMEA (Coordinating Ministry for Economic Affairs). 2020. Indonesia’s Critical
Occupations List 2018. Social Protection and Jobs Global Practice, World Bank, Washington,
DC. http://documents.worldbank.org/curated/en/763611585857010121/Indonesias-
Critical-Occupations-List-2018-Technical-Report

June 2021 Pathways to Middle-Class Jobs in Indonesia

Workers, Jobs, and Barriers to Entry to Middle-Class Jobs
References 239

Pathways to Middle-Class Jobs in Indonesia June 2021

240 Report Summary and Policy Directions

Chapter 6

Report Summary and

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Report Summary and Policy Directions 241

Policy Directions

This report argues that Indonesia’s economy has yet to deliver the jobs and labor
income the country needs if it is to achieve its middle-class aspirations. From 2008
to 2018 Indonesia created an average of 2 million jobs each year. The employment rate
reached a two-decade high in 2018, with 63.7 percent of youth and adults working, while
unemployment rates were at a 20-year low of 5.3 percent.

Despite these encouraging signs, Indonesia is not creating the middle-class jobs
needed to fuel a middle-class country. The middle class accounted for 23 percent of the
total Indonesian population in 2018, up from 7 percent in 2002. This is slower growth than
experienced in neighboring countries—including Thailand and the Philippines—and less than
half the rate of expansion seen in China or Vietnam. Jobs are the primary source of income
in all these countries, but Indonesia has not been as successful at moving its workers up the
jobs ladder.

The question is where Indonesia went wrong, and what to do to get back on the
path toward middle-class jobs. Middle-class ambitions are at the top of the government’s
agenda, and good jobs are seen as the path to achieve them. In his inaugural speech in 2019,
President Joko Widodo spoke to the need to develop a “workforce that is hardworking,
dynamic, skilled . . .” and of the need to transform economically toward “a competitive
and modern manufacturing and service-based economy that has high added value for
the prosperity of the nation . . . for all Indonesian people.” The whole of government has
responded with policy making to address the constraints on the creation of productive jobs.

Pathways to Middle-Class Jobs in Indonesia June 2021

242 Report Summary and Policy Directions

This chapter summarizes the messages that cut across the analysis in this report
and the policy proposals that have emerged. It proposes that the central challenge
is to break out of the growth in low-productivity jobs that has characterized Indonesia’s
economy since 2008. This will require economywide efforts, targeted support to enable
workers to transition to higher-productivity jobs, and a deliberate and urgent strategy to
prepare workers to engage in a more sophisticated economic structure.

6.1 Summary of Messages

Indonesia’s modern history is a story of growth in low-productivity jobs. Indonesia
was growing its quality jobs through structural transformation in the 1980s and 1990s.
People were moving out of agriculture and into the rapidly growing manufacturing sector.
However, the Asian financial crisis interrupted the transition. Foreign direct investment (FDI)
slowed and manufacturing stalled.

Post Asian financial crisis, manufacturing did not resume its growth path; instead,
a commodity boom fueled the country. Since the commodities sector does not create
many jobs, the structural transformation became a movement out of agriculture and into
low-value service jobs, increasingly demanded partly as a result of urbanization and rising
income per capita. Indonesia sustained 5 percent economic growth and continued to reduce
poverty during this period, but its movement toward middle-class jobs stalled. A decade of
potential job improvements was lost. By the close of the commodity boom in 2012, low-value
services were deeply entrenched in the economy. Manufacturing growth resumed and high-
value services (such as finance and health) began to develop, but they were constrained by
a range of factors, including protectionism.

Through this turbulence, Indonesia continued creating jobs, but the quality was too
low to elevate enough people to the middle class, a level of prosperity Indonesians
were increasingly aspiring to. Only 12 million of Indonesia’s 85 million income earners have
an income that puts them in the middle class (the other 39 million workers of Indonesia’s
total workforce of 124 million are either unpaid or share profits). Two-thirds of jobs enjoyed
by the 124 million workforce are in agriculture or low-value services. Labor productivity is
low, resulting in unit labor costs that are both high and increasing relative to the East Asian
region, despite Indonesia’s low wages relative to that region. Only 3.5 million workers earn a
middle-class income and receive a full package of social benefits. Just 18 percent of jobs are
covered by work contracts.

To achieve jobs and incomes that provide economic security, Indonesia could move
toward a growth model that is driven by productivity and creates jobs connected
with higher value added and value-added growth, and that facilitates movement of
workers into more productive jobs, within and between sectors. This requires productivity
improvements, particularly in sectors and firms that create many good-quality jobs, and a
transformation of the highly capable workforce, which must be supported to prepare for
and integrate into productive jobs.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Report Summary and Policy Directions 243

The structure of the firm sector is not conducive to the creation of middle-class
jobs. Two-thirds of jobs are in household enterprises (HHEs). Indonesia’s HHE sector has 45
million HHE owners and 38 million workers, and nearly all owners and workers are informal.
Very few HHEs grow into small firms with employees. Formal firms are slow to create middle-
class jobs. The manufacturing sector—the source of many jobs in the 1980s and 1990s—
creates some openings, but most job creation is in old (more than 30 years) and very large
firms. Firm dynamism from 1990 to 2015 mostly came from surviving plants, instead of
from new firms entering the market and others closing down permanently. New firms found
it difficult to grow and bring on new staff, which limited competition and productivity
increases. Indonesia’s nonagricultural enterprises do not grow much over their 40-year life
cycle. Job churning is significant but this indicates that jobs are unstable, rather than that
people are moving up to better jobs; it occurs among existing firms rather than in new firms
entering the market and others closing down permanently.

Most of today’s workforce is not equipped to hold a middle-class job. Middle-class

jobs tend to be in high-skill occupations—managers, professionals, and high-level technicians
and associates in manufacturing and services industries—that require strong cognitive,
interpersonal,1 and digital skills, as well as knowledge in science, technology, engineering,
or mathematics (STEM), and business administration. However, 57 percent of the labor
force has a lower-secondary education (nine years of completed school) or less. The level
of learning is also low: Nearly 14 percent of firms interviewed in the Doing Business 2015
survey in Indonesia reported that  “workforce education” is a major or severe obstacle
to doing business, as compared to 8.4 percent of employers in the Philippines and 2.5
percent in Thailand. There is a particularly severe shortage of the skills needed for more-
skilled  occupations, including managerial roles.  Women and youth have particular trouble
converting skills into high-paying jobs, partly due to occupational and employment-type
sorting, gender wage gap (women), and lack of experience (youth).

Perhaps more worrisome is that Indonesia does not have the systems in place to help
the large workforce acquire the skills that a structurally transformed, productive
Indonesia will need. The education system for children and young adults (including
university, and technical and vocational education and training [TVET]) is relatively robust.
However, the average worker today is aged 40 years or older, is a household head, male, and
will continue working for at least another 20 years. The large stock of older workers that is
out of the reach of the education system limits the effectiveness of the education system
in upskilling the workforce. The firm sector does little to fill the gap as only a few firms
continuously provide training to ensure their workers’ skills remain relevant.

1 Interpersonal skill is the ability to manage people and relationships in order to successfully perform a range of tasks.

Pathways to Middle-Class Jobs in Indonesia June 2021

244 Report Summary and Policy Directions

So, what can be done to deliver results soon and over the longer term? As this report is
being completed, the COVID-19 pandemic has eroded many of the gains Indonesia has made
toward creating middle-class jobs. However, with careful planning to build back better, the
disruption can be an opportunity to systematically and strategically map out a recovery
agenda that focuses on the creation of middle-class jobs and puts Indonesia more decisively
on the path of becoming a high-income county by 2045.

6.2 The Pathway to Middle-Class Jobs: A Policy Strategy

Indonesia’s economy has failed to deliver the jobs and labor income the country needs if
it is to achieve its middle-class aspirations. As argued in this report, this is partly due to a
disrupted structural transformation that resulted in the entrenchment of large and old
firms, the growth of low value-added services jobs, and a low-skilled labor force. To get
Indonesia back on track, this report proposes a three-pronged strategy (Figure 6.1).

First, Indonesia could promote productivity growth across the board, not least in
sectors with low levels of value added. The most promising route to higher productivity is
enacting competition-enhancing policies that will help more firms to enter the market and
grow, and that will spur innovation. Competition-enhancing policies include those that lower
the high cost of trade, that increase access to foreign talent to fill positions for which there
is currently a shortage of domestic workers, and that attract export-oriented, efficiency-
seeking FDI linked to global value chains (GVCs). Competition-enhancing policies also include
those that improve domestic regulations to create a more predictable and level playing
field for firms, and those that strengthen institutions that promote competition and sound
business regulation. This reform strategy will also require a concerted effort to support
productivity growth in HHEs and small and medium enterprises (SMEs). Such enterprises
account for a large share of jobs, but many operate at a low level of productivity.

Second, Indonesia could facilitate a more decisive shift in economic activity and
workers toward more productive and higher-paying sectors, firms, and jobs. Even
if reforms to increase productivity across the board are sector-blind, sectors that create
middle-class jobs could be targeted for priority reforms, such as strategies to attract FDI,
while existing SME and HHE policies could be applied with a middle-class-jobs lens. In addition,
Indonesia needs robust labor information systems, a more efficient job-matching mechanism,
and mobility support so that people can appropriately prepare for, and transition to, better
jobs in other sectors, firms, or occupations.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Report Summary and Policy Directions 245

And third, Indonesia could build a workforce that has the necessary skills to take on
the new jobs in higher value-added and internationally competitive sectors. This will require
changes to the education system to better prepare today’s youth for modern jobs. Perhaps
more challenging, it will also require innovations to upskill the current adult labor force. An
upskilled labor force will attract productive investment, will be able to meet the demands of
the new jobs, and will benefit more from potential technology transfers and spillovers of FDI
from multinationals. Additional support to crowd in women and youth is also increasingly
necessary as the share of those of working age within the total population begins to decline.

The government has started comprehensive policy interventions to close gaps in

infrastructure, human capital, and financial services.2 It is important that these continue as
they are the foundations of the pathway to middle-class jobs.

Figure 6.1 Three-Pronged Strategy Toward Middle-Class Jobs in Indonesia

Sources: World Bank

2 Indonesia needs an efficient financial sector that is less costly, faster, safer, and more transparent if middle-class jobs are
to be created. The financial sector must mobilize domestic savings and allocate them across time and space to the more
productive sectors, firms, and workers. An efficient financial sector also enables firms and households to cope with economic
uncertainties by hedging, pooling, sharing, and pricing risks, thereby helping savers and investors to make better decisions on
allocating their money and financial assets. The financial sector in Indonesia still, however, faces many challenges that could
impede allocative efficiency of resources to the most productive uses.

Pathways to Middle-Class Jobs in Indonesia June 2021

246 Report Summary and Policy Directions

6.2.1 Reform Strategy 1: Accelerate Across-the-Board

Productivity Growth
Indonesia’s increases in productivity are small, undermining the country’s growth
potential and the expansion of the middle class. Subdued growth in the manufacturing
sector has held back economywide productivity gains since 2000. Moreover, most
Indonesians work in agriculture or low value-added services, so raising productivity in these
sectors is necessary in order to make an impact on poverty, vulnerability, and economic
insecurity. Increasing labor productivity in agriculture will free up labor that was previously
underemployed in low-paying subsistence farming for more productive jobs in industry and
high-value services. There are high levels of poverty among those working in agriculture
and, if productivity is not further increased, poverty may force farmers (especially young
farmers) out of the sector. They will then accept nonagricultural jobs where the labor
productivity is slightly higher but insufficient to help Indonesia’s economic transformation.
Accelerating across-the-board productivity growth is critical and this requires Indonesia to
unlock firm entry and growth and attract FDI to create the competition and innovation that
spur productivity in the private sector. It also requires Indonesia to reduce the high cost
of trade and increase the access of firms to foreign talent to fill roles for which there is a
shortage of domestic workers. It will also require a concerted effort to support productivity
growth in SMEs and HHEs, sectors in which a large share of jobs exist. Reforms to accelerate
across-the-board productivity growth should be (almost) sector-blind.

First, Indonesia could unlock new firm entry and growth to create the competition
and innovation that spur productivity in the private sector. The private sector
accounted for 86.8 percent of all jobs in 2019 in Indonesia, and reducing the impediments
firms face in entering and growing are key to robust employment growth. But, to create
quality jobs and expand Indonesia’s middle class, Indonesia could do more. Integrating firms
into GVCs and leveling the playing field between, for example, domestic versus foreign firms
and small versus large firms are also key to ensuring firm dynamism and robust job creation
in the private sector. Along with reforms to close gaps in infrastructure, human capital, and
financial services, Indonesia needs a more predictable regulatory framework to enable firms
to enter and to grow.3

Four areas of policy intervention are critical if Indonesia is to unlock new firm entry and
growth, both of which are needed to create the competition and innovation that spur
productivity and create middle-class jobs in the private sector:4

(i) Improvement of FDI policies, increased access to inputs, markets, and foreign talent
to fill roles for which there is a shortage of domestic workers, and integration of firms
to GVCs: The government has begun one of the most ambitious investment and trade
reforms in decades (see Box 6.1). Further recommended reforms include reducing the
list of sectors reserved for micro, small, and medium enterprises (MSME); transforming
burdensome requirements across various sectors for a minimum local content

3 World Bank 2020a

4 See World Bank 2020a; Chang et al. 2019

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Report Summary and Policy Directions 247

into positive incentives to use local supply; eliminating the Expatriate Manpower
Utilization/Placement Plan as a precondition for the employment of foreign talent in
high-skill occupations for which there is a shortage of domestic workers; addressing
nontariff barriers to access to inputs of production—such barriers include preshipment
inspections, restrictions on the port of entry through which goods may enter Indonesia,
and requirements for ministerial letters of recommendation.5 The government is also
recommended to continue to pursue ambitious preferential trade agreements to lock
in domestic reform and attract investment, including the EU Comprehensive Economic
Partnership Agreement, which needs to be finalized.

(ii) Increased capacity of domestic firms and workers: To benefit from the technology
spillovers of FDI, Indonesia could increase the quality of human capital and domestic
workers (see Reform Strategy 3), promote firm links, provide incentives and advocacy to
promote firm-based training programs, build business networks, establish institutional
partnerships, especially between domestic and foreign-owned firms, and provide
information services covering, for example, the quality requirements of multinationals.
Improvements in these areas can increase both the absorptive and adaptive capacities
of local firms.

(iii) Provision of a more predictable regulatory framework to support firms: Indonesia could:
(a) introduce a regulatory oversight body to reduce regulatory uncertainty; (b) over
the longer term, introduce into the regulatory process compulsory public consultation
procedures that follow international best practices, and provide incentives to local
governments to streamline the process of business licensing; (c) consolidate and target
innovation policy instruments to better assist firms to perform innovative activities
that will spur productivity growth and new product development.

(iv) Strengthen the competition regulatory framework by: (a) strengthening the technical
capacity of the Business Competition Supervisory Commission (Komisi Pengawas
Persaingan Usaha; KPPU) to enforce competition laws and advocate pro-competition
policies; (b) revisiting the existing Competition Law (No. 5/1999) to improve the ability of
Indonesia’s competition framework to identify and sanction anticompetitive behavior;
(c) mainstreaming competition considerations in the policy process to ensure that
regulations do not unduly restrict competition.

Second, Indonesia could improve the productivity of its SMEs, which will help
them to grow and become better creators of middle-class jobs. Indonesia’s SMEs,
the “missing middle”—Indonesia has many micro and large firms, but not many medium-
sized firms—remain uncompetitive due to their low levels of productivity and innovation.
Ineffective government policies toward SMEs have contributed to this situation. The
government has issued a range of policies to improve productivity growth among SMEs,
including policies related to reducing the cost of and accelerating business registration,
technology transfer, and e-commerce; access to market; helping SMEs to export; and
financial assistance. However, SME policies continue to be framed with a welfare approach,

5 World Bank 2021

Pathways to Middle-Class Jobs in Indonesia June 2021

248 Report Summary and Policy Directions

instead of an approach based in competition, efficiency, and productivity. Existing policies

also reflect excessive protectionism. For example, some business sectors are reserved for
SMEs, and other programs also shield SMEs from competition.6

The report identifies five priority policy areas of intervention that would improve the
productivity of Indonesia’s SMEs, helping them to grow and become better creators of
middle-class jobs:

(i) Indonesia could develop specific measures and strategies to assimilate SMEs into GVCs,
including the integration of smallholders into agriculture value chains and home-based
workers into manufacturing and services GVCs. This will ensure that GVCs benefit the
poorest.7 Priority policy reform areas are: (a) addressing information gaps in order that
SMEs understand the quality standards they must meet to supply to multinationals or
their tier-three, or higher, suppliers; (b) supporting SMEs to meet these quality standards
and ensure they deliver on time; (c) providing tax incentives for multinationals that invest
in the upgrading of local suppliers through training, mentoring, or staff secondment;8
(d) reducing the transaction costs for SMEs using free trade agreements, simplifying
Certificate of Origin procedures and helping SMEs to arrange them, and making each
Certificate of Origin free of charge.9

(ii) Indonesia could promote the incremental introduction of innovative activities to boost
productivity. Reforms could focus on a mix of policy instruments that may include:
(a)  technology extension and diffusion programs; (b) early-stage infrastructure and
advisory services; (c) inducements; (d) vouchers for collaboration; and (e) direct grants
for business innovation (with embedded advisory services).10

(iii) Indonesia could close the digital infrastructure divide and ensure affordable, reliable, and
ubiquitous high-speed broadband internet access (fixed and mobile 4G) by: (a) reforming
spectrum management and infrastructure sharing that will stimulate further private
investment; and (b) targeting public-private partnerships to close internet access gaps.

(iv) Indonesia could expand access to financial services. Priority policy reform areas are:
(a) preparing a roadmap with a view to leveling the playing field for delivery of financial
services; (b) strengthening financial sector infrastructure such as credit registries and
secured transactions; and (c) promoting the use of dedicated liquidity and risk-sharing
facilities (especially relevant in the aftermath of COVID-19).

6 OECD and ERIA (Economic Research Institute for ASEAN and East Asia) 2018.
7 World Bank 2020c
8 OECD (Organisation for Economic Co-operation and Development) 2018a
9 Anas, Mangunsong, and Panjaitan 2017
10 Cirera et al. 2020

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Report Summary and Policy Directions 249

Third, Indonesia could target support to increase the productivity of HHEs, the
source of a large share of jobs.11 Household enterprises, which account for more than 98
percent of microenterprises, are significantly different to SMEs in many respects.12 Specific
policy reforms for HHEs should focus less on growth and more on increasing productivity,
and hence income, marginally and incrementally without requiring a significant amount of
additional resources (for example, time, workers, capital, or finance).

There are three policy areas for interventions that can improve the marginal productivity
of HHEs:

(i) Indonesia could improve basic business skills of HHE owners through training and
education. At least three skill sets need to be improved: financial literacy, management,
and entrepreneurship. Strategies to improve the business skills of HHE owners and
workers should be addressed in a short- or medium-term horizon and are: (a) creating
short training modules on simplified rule-of-thumb business skills, combined with training
in personal initiative (soft skills), business competition, and innovation; (b) providing
information to HHEs about the benefits of training and possible routes to acquiring it;
and (c) in the medium term, synchronizing and integrating government training and
support programs across ministries and bodies.

(ii) Indonesia could create inclusive ecosystems for HHEs. Priority policy reform areas to
create inclusive and more secure ecosystems for HHEs are: (a) encouraging the owners
of HHEs to join the social security system—Badan Penyelenggara Jaminan Sosial (BPJS)
Ketenagakerjaan—for nonwage workers by providing tailored information on the
benefits of doing so; (b)  providing government contributions to ease entry into and
sustain participation in the program; (c)  promoting the benefits of becoming a legal
entity by bundling legal registration with other government support programs; and
(d)  improving the delivery of support programs, and broadening and facilitating HHE
access to them, independently of their participation in social assistance programs.

(iii) Indonesia could improve product competitiveness. Specific policy reforms to improve
product competitiveness are: (a) promoting on-the-job training and apprenticeships; (b)
providing shared factories or machines that can be used by HHEs; and (c) supporting
product standardization and modernization of HHEs.

While this review does not explore factors that constrain productivity in the agricultural
sector, the recommendations for non-farm HHEs are also relevant for the farm sector.

11 This recommendation is largely drawn from a policy background paper written for this report by Purnagunawan (2020).
12 See Table 4.15 in Chapter 4 of the Technical Report.

Pathways to Middle-Class Jobs in Indonesia June 2021

250 Report Summary and Policy Directions

Box 6.1 Creating Better and More Jobs: Recent Reforms to Investment,
Access to Skilled Foreign Workers in Shortage, and Manufacturing Inputs
Until recently, Indonesia had some of the tightest restrictions on foreign direct investment (FDI) among middle-
income countries, which has weighed down on the creation of middle-class jobs. These restrictions on FDI
have been compounded by restrictions on the hiring of foreign high-skilled workers to occupations for which
Indonesia has a shortage of domestic workers, and high costs for sourcing manufacturing inputs from global
markets.

The government has sought to address some of these issues by implementing one of the most ambitious
investment and trade reform programs in decades. The key reform was a presidential regulation on investment,
which reduced the number of business activities subject to at least one investment restriction from 813 to
260.a This liberalization of investment, particularly FDI, moves Indonesia from having one of the most restrictive
FDI regulatory regimes in the region to having one of the most open. According to World Bank estimates,
this liberalization may generate between US$4.1 billion and US$6.0 billion in additional foreign and domestic
investment in the liberalized sectors. The investment flows would translate into an estimated increase in GDP
growth of between 0.12 and 0.17 percentage points, raise productivity, employment, and wages, and ensure
lower prices and/or better-quality products and services for consumers.

To facilitate a more adequate supply of high-skilled professionals in shortage to meet the needs of the labor
market, the government also passed Government Regulation No. 34/2021 on Foreign Workers. This complements
a regulation issued by the Minister of Manpower in 2019b in two ways. First, it removed the requirement for an
employer to have an Expatriate Manpower Utilization/Placement Plan in order to employ an expatriate worker
for a number of positions. Second, it has eliminated the requirement for an employer to obtain a written
license (an Expatriate Manpower Employment License) from the Minister of Manpower or an appointed official
to employ an expatriate worker. This is the most ambitious set of reforms on work permits for foreigners that
Indonesia has enacted in recent decades. It is likely to provide more certainty to the approval process, eliminate
the discretion of sectoral ministries, and expand the range of skills gaps that firms can fill through global
markets. The immediate expected outcome is that a higher share of high-skilled jobs in shortage in Indonesia
will be covered by expatriates.

To facilitate access to manufacturing inputs, the government reduced the cost of compliance with national
product standards (SNI) by removing compulsory SNI certification for steel, stove and wire products used by
producers. This was carried out in Minster of Industry Regulation No. 35/2019, which revoked 14 individual
regulations.c The inputs covered by the 14 Ministry of Industry regulations revoked by the reform are worth
about US$3.9 billion, or 3.5 percent of total intermediate imports in 2019. This is equivalent to almost 10
percent of the intermediate imports the SNI certification is applied to. This exemption should not affect the
safety of production processes using these inputs as the products do not present high risks for users. By
lowering the costs and increasing the certainty of supply of key inputs, this reform is expected to increase the

a The comparison of the number of business activities subject to investment restrictions in the older regulation (Presidential Regulation No. 44/2016)
versus the new is not straightforward as the business activities in the older regulation are defined using different levels of aggregation.
b This 2019 regulation has more than trebled the number of occupations in which high-skilled expatriate workers (managerial, professional, technical)
may be employed. It also allows the minister to permit expatriate employment on a case-by-case basis for positions that are not specified in the regulation.
Previously, specific permission was needed from the relevant ministries.
c For producers the exemptions are granted automatically. For general importers, further consideration is required before an exemption is granted.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Report Summary and Policy Directions 251

Box 6.1 continued

competitiveness of producers, particularly in manufacturing. It is expected to reduce producers’ fixed costs of

switching inputs, thereby facilitating innovation, which is currently limited among firms in Indonesia.

Source: This box is largely drawn from World Bank 2021.

6.2.2 Reform Strategy 2: Structural Transformation 3.0:

Transition to Sectors and Firms That Create Good-Quality
Jobs
Indonesia’s continued dependence on commodities and nontradable sectors and
its inward-looking economic policies have not created enough jobs in higher-
productivity sectors; this has left many workers stuck in low-productivity, low-
wage work. Workers can earn higher wages when the productivity of the jobs currently
available increases (Reform Strategy 1), or when they shift from lower-productivity
jobs, firms, or sectors such as traditional agriculture or low value-added services to
higher-productivity jobs, firms, or sectors such as tourism, health, and education. So far,
Indonesia’s growth model has failed to deliver a significant transition of workers from lower-
productivity to higher-productivity sectors and activities, and there has been insufficient
structural transformation to push labor productivity higher. With a more skilled workforce
and implementation of the policies in Reform Strategy 1, the economy will naturally create
more jobs in higher-productivity sectors. Indonesia can accelerate the transition from
lower- to higher-productivity jobs with policies to increase labor demand in sectors that can
deliver middle-class jobs. This will include prioritizing sectors such as tourism, information
communication technology (ICT), finance, health, education, and manufacturing for
investment promotion efforts, as well as for relieving infrastructure bottlenecks.

First, Indonesia could prioritize investment promotion strategies to attract FDI to

sectors and projects (including infrastructure projects) that can create middle-
class jobs. The goal is to inform multinationals of the investment possibilities in Indonesia,
reduce gaps in knowledge, and identify and overcome potential barriers such as market
failures or the existence of externalities.13 Prioritizing foreign investment promotion in
a select set of sectors can improve the overall effectiveness of investment promotion
agencies.14 Literature shows that countries obtain a higher level of FDI in sectors that are
being prioritized by investment promotion agencies.15 While prioritization of FDI promotion
may increase investment in the targeted sectors, it may also involve risks. Prioritization
strategies by investment promotion agencies are varied and complex, and must be carefully
balanced across many aspects of investment promotion, including allocation of staff and

13 OECD 2018b
14 OECD 2018b
15 See, for example, Harding and Javorcik (2011)

Pathways to Middle-Class Jobs in Indonesia June 2021

252 Report Summary and Policy Directions

resources and services offered as part of the prioritization.16 Global practices show that
promotion strategies can focus on sectors or projects, or particular countries and/or
investors from which funding is sought. For example, based on a survey in 2017, 71 percent
of investment promotion agencies from countries of the Organisation for Economic Co-
operation and Development have established criteria for prioritizing investment projects,
where “job creation” (79 percent) is the third most commonly used criterion after “impact
on innovation” and “priority sectors.” These criteria are broadly in line with the criteria used
to target countries and sectors. For example, countries that prioritize investment from
nations that could be a source of high technology and support to sectors that will have
a positive impact on employment choose “innovation” and “job creation” as the criteria
for choosing which of their investment projects to promote.17 Tax incentives are often
considered in this context, but they are an inferior instrument, especially in Indonesia, which
already collects less tax than most of its peers and as a result has a revenue base that is
insufficient to sustainably fund the public services, such as infrastructure, that attract high-
quality investments.

The strategic roadmap of the Indonesian Investment Coordinating Board (BKPM)18 is

prioritizing: (i) low-hanging fruit, meaning optimizing Indonesia’s natural resources (a short-
term focus); (ii) soft and hard infrastructure; (iii) investment that will lay the foundations
for industrialization, including education to create a highly skilled and educated workforce;
(iv) support for a knowledge-based economy and further development of a globally
competitive and educated workforce. This prioritization strategy broadly focuses on sectors
that create middle-class jobs, such as tourism, high value-added agriculture, manufacturing,
health, education, finance, and ICT. There is, however, room to strengthen the criteria for
creating middle-class jobs. One of the background papers to this report found that while
there are significant benefits from all manufacturing FDI, the aggregate labor market effects
are most widespread for the low-skill manufacturing sectors,19 such as food and beverages,
tobacco and textiles. However, in the BKPM’s strategic roadmap there is no explicit focus on
low-skill manufacturing sectors. Moreover, since structural transformation must go hand in
hand with infrastructure development,20 to which the government is strongly committed,
a prioritization criterion for infrastructure development could use a lens of middle-class job
creation.

To further strengthen the government’s investment promotion strategy, policy interventions

could focus on:

(i) Revisiting the current strategic roadmap of the BKPM (including provisions for
infrastructure development) and applying a lens of middle-class job creation (from
both the demand and supply sides) using global best practices on prioritization of

16 OECD 2018b
17 OECD 2018b
18 BKPM (Indonesian Investment Coordinating Board). N.d.
19 See Annex 4.2 for manufacturing subsector classification
20 Lin 2012

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Report Summary and Policy Directions 253

investment, such as the OECD Policy Framework for Investment.21 Also, revisiting how
prioritization is being implemented. For example, assessing whether there is a need for
a BKPM unit dedicated to priority sectors, projects, countries, or investors, and also
assessing the types of services provided to priority sectors, projects, countries, and
investors (including addressing investor grievances), and assessing human resources
and budgets allocated to priority sectors, projects, countries, and investors. Indonesia
could evaluate in the medium term whether the prioritization strategy has achieved the
intended results, that is, whether it has created more middle-class jobs.

(ii) Moreover, to the extent that reforms are not sector-blind, Indonesia could apply a lens
of middle-class job creation when selecting sectors and enterprises to target with
policy interventions. Policy interventions on SMEs and HHEs could target sectors and
enterprises that have greater potential to create middle-class jobs.

Second, Indonesia could close information gaps to help workers transition to

higher-quality jobs. When new firms enter (exit), new jobs are advertised and/or new
sectors emerge (disappear). Information has to be available to allow workers to relocate. If
productivity gains are to be achieved, new firms must enter, more productive firms must
outcompete less productive firms, and labor must be able to move into more productive
sectors, firms, and jobs. Indonesia needs policies to help workers move to better jobs:

(i) Upgrade KarirHub and the overall labor market information system by: (a) including
within the system private job portals and public portals providing services for job search
and career development; and (b) building out the labor market information system in
the short run with real-time information. This can be achieved by developing artificial
intelligence (AI) to trawl sites and pull information to the platform, fielding surveys to
collect information on how people search for and prepare for jobs, and analyzing and
curating information to identify both critical occupations and high-growth occupations.22

(ii) Providing relocation support for hard-to-fill vacancies by: (a) ensuring the system
provides information on nationwide job vacancies; and (b) providing stipends to support
relocation (travel, housing, and other social services) for workers who are relocating to
fill critical occupations.

(iii) Continuing to develop the Job Loss Guarantee Program (Jaminan Kehilangan Pekerjaan;
JKP) to support the transition of workers up the job ladder by: (a) allowing current
workers to opt in (or out) of the JKP (during the initial implementation phases, while
mandating all newly hired workers into the system); (b) working with social partners
to define the program parameters—including financing mechanisms—and to implement
a communications campaign about the benefits of the reform for workers and firms;
(c) developing a strong monitoring system to avoid misuse of the program; and (d)
establishing strong links with the entirety of the ecosystem around the labor market
information system.

21 OECD 2018b
22 See also World Bank and CMEA (Coordinating Ministry of Economic Affairs) (2020).

Pathways to Middle-Class Jobs in Indonesia June 2021

254 Report Summary and Policy Directions

6.2.3 Reform Strategy 3: Build a Middle-Class Workforce23

The workers of today and tomorrow do not have the skills they need to attract and
succeed in middle-class jobs. Only 40.6 percent of youth have completed upper secondary
school, a prerequisite for a middle-class job. There is a significant mismatch between learning
and job-relevant skills, especially higher-order cognitive skills and behavioral skills, which
should be honed at secondary school, and professional studies. Perhaps more worrisome
are the 124 million workers who will comprise the labor force for many decades, but who
will not return to school and who have few options for developing their skills. Indonesia
needs an upskilling and reskilling policy agenda in the short run to meet emerging industry
demands on labor. This leads to three areas of policy interventions.

First, Indonesia could facilitate learning. The skills imparted by existing secondary
school and training programs do not meet market demands. This is partly due to the
mismatch between the curriculum and the demands of the labor market, but is also due to
a lack of access to education and training. Results-based teaching can both expand access
and enhance the relevance of the teaching materials to the world of work. In addition to
improving the general education system,24 Indonesia could:

(i) Implement targeted incentives for students at risk of dropping out of secondary school.
This will require the risk factors for at-risk populations to be mapped at the local level
and interventions designed accordingly. It will also require reform of the curriculum and
delivery mechanism of Paket C (a diploma equivalency program for those aged over 18
who are unable to gain a diploma through traditional schooling), with greater use of
distance learning to facilitate program participation by the target audience.

(ii) Support development of and access to online distance learning in short TVET courses
by accrediting online training courses (offered by public and private sector providers)
and including a database of accredited online training courses on the labor market
information system, as well as providing training vouchers for vulnerable groups or
those seeking training in skills relevant to critical occupations. One of the best ways to
upskill the current workforce is by providing more and better short training courses.

(iii) Modernize curricula and introduce new pedagogical methods to develop nonroutine
interpersonal, analytical, and digital skills. Interventions must be tailored to the age
(from preschool through adult education) and context of the learners.

(iv) Improve the quality of the training system by prioritizing the development of
quality-assurance mechanisms and more effectively engaging the enterprise sector.
Regarding quality-assurance mechanisms, Indonesia could place a higher priority on the
development of strictly defined and enforced competency frameworks and accreditation
standards. In the long run, it will be necessary to develop a centralized monitoring and
evaluation system in which all training institutes report input, output, and (eventually)
outcome data. It will also be necessary to establish a National Vocational Skills Steering

23 This section is partly drawn from a policy background paper written for this report by Pratomo (2020)
24 World Bank 2020b

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Report Summary and Policy Directions 255

Committee to make funding and other systemwide decisions based on quality data.
Regarding the enterprise sector, Indonesia could amend Ministry of Finance Regulation
128/2019 to ensure the focus is on subsidizing results-based worker training in firms.
Public resources can be used to incentivize exchanges of staff between industry and
institutions providing TVET by financing TVET instructors to undertake internships in
firms and subsidizing the wages of industry experts to allow them to take time out from
their firm to teach in TVET institutions.

Indonesia could coordinate the reforms to build a middle-class workforce (as laid out
in Reform Strategy 3) with reforms to unlock new firm entry and growth (as laid out in
Reform Strategy 1), especially reforms designed to attract FDI. This will avoid any mismatch
between the demand and supply sides of middle-class jobs. It will also tilt the scale of net
job creation slowly toward sectors and firms that can create middle-class jobs. Singapore is
an example of a country that regularly maps short- to long-term economic investment and
the demand created for certain skills or types of worker. It uses the data to upskill or reskill
current workers with updated educational and training curricula.

Second, Indonesia could provide tailored support to particular groups. The above
measures would benefit all learners, but some groups have additional constraints that
limit their success. The factors underlying the significantly lower wages earned by women
compared to men are difficult to identify, but there is some indication that women tend
to study and work in lower-paying social and other services and are overrepresented in
urban self-employment. Women tend to engage in less job search than do men and use
more informal means, and the conflict between work and home leads to suboptimal work
decisions. Youth also experience a significant wage gap relative to older adults, which is
primarily attributable to their lack of experience. Both women and youth would benefit
from tailored job support, as well as shifting social norms, including:

(i) Fine-tuning of the labor market information system and training programs to address
factors that limit their success in acquiring middle-class jobs; including gender- and
age-relevant information in the labor market information system, such as nonstandard
fields of study that women can succeed in and auxiliary services such as childcare or
student loans, and providing complementary face-to-face information interventions
for girls in secondary school. In terms of training, Indonesia can earmark a portion
of training vouchers for women and youth to enable them to participate in courses
on business practices and the technology needed for engaging in the gig economy.
Indonesia can also incorporate business development training into entrepreneurship
support programs for women.

(ii) Expanding women’s access to social benefits through their jobs. Indonesia could
use legislation to facilitate the formal employment of women by securing women’s
jobs during pregnancy, maternity leave, or in the period following the return
to work. Indonesia could establish grievance procedures enabling workers to
report violations of the Labor Law, which guarantees social benefits for workers.

Pathways to Middle-Class Jobs in Indonesia June 2021

256 Report Summary and Policy Directions

(iii) Amending the Labor Law to set national norms for women’s access to the labor market,
including a robust definition of “equal pay for equal work” (see Box 6.2), a robust
definition of sexual harassment, clarification of the nature of unacceptable workplace
interactions, and extending of paternity benefits to 30 days.

Box 6.2 Finding a Balance between Labor Market Flexibility and Worker
Protection
The Omnibus Law on Job Creation (Omnibus Law) aims to boost competitiveness, including through significant
reforms to Labor Law 13 of 2003. The government aims, with the reforms, to improve labor market flexibility,
while maintaining worker protection by modernizing existing labor institutions to better align with modern
labor markets. The goal is to make Indonesia more attractive to investors, thereby potentially creating more
good jobs, and also to improve the currently low level of compliance with labor regulations. As is standard
practice in Indonesia, the Omnibus Law was followed by a series of government regulations on the reforms.
Reforms have been made to areas including:

Fixed-term contracts (articles 56–61):  The Omnibus Law modifies work agreements, particularly fixed-
term contracts. Most notably it stipulates that fixed-term contracts cannot be (automatically) transformed
into permanent work. The implementing regulation, Government Regulation No. 35/2021, allows for a longer
duration of fixed-term contracts (the maximum length of a fixed-term contract has been increased from
three to five years) and introduces compensation for termination that is proportional to the length of the
employment relationship.

Employment agencies for temporary or outsourced staff (articles 64–66): The Omnibus Law makes
these employment agencies responsible for providing the same benefits to its workers as those provided by
firms to their permanent workers. These agencies must be licensed by the Ministry of Manpower. The previous
restriction that limited outsourcing to noncore business processes has been eliminated.

Working time (articles 78 and 79): The Omnibus Law expands overtime to a maximum of 4 hours per day (1
additional hour) and 18 hours per week (4 additional hours). It also removes regulations on extended leave and
other types of leave, instead allowing this to be agreed by the worker and the company, or the union and the
company. Government Regulation No. 35/2021 provides further details on how to calculate overtime wages
and other overtime provisions.

Minimum wage (article 88): Micro and small firms are exempt from compliance with the minimum wage
if they rely on traditional resources and do not operate in high technology or in capital-intensive business
(definitions will be introduced in  implementing  ministerial regulations). Instead, Government Regulation No.
36/2021 stipulates that the employer needs to pay above a subfloor  equivalent of  50 percent of average
public consumption and 25 percent above the provincial poverty line. The minimum wage formula has been
modified—sectoral minimum wages have been eliminated, as have provincial wage councils—with the objective
of pacing  minimum wage growth in regions where minimum wages are already high, and decreasing wage
inequality across regions where the ratio of minimum wage to median wage is low. Government Regulation
No. 36/2021 introduces a formula for district minimum wages, and a more complex formula for provincial
minimum wages that depends on economic growth, inflation, purchasing power parity, manpower absorption
level, average consumption per capita, average number of household members who work in the region, and
median wages.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Report Summary and Policy Directions 257

Box 6.2 continued

Wage policy (articles 88 and 88B): The Omnibus Law introduces equal wages for equal pay, legislation that
is commonly used by countries to address gender wage gaps. The Omnibus Law introduces a minimum hourly
wage, further regulated in Government Regulation No. 36/2021. 

Contract termination and severance (articles 150–90): The Omnibus Law eliminates the health and housing
allowance component of compensation pay, which had accounted for 15 percent of total severance pay and
service pay. Government Regulation No. 35/2021 reduces the reward pay for termination of short-tenure
employment (for example, the requirement to pay two months’ wages for up to three years of work has been
removed). It also adopts the concept of notice of termination, similar to countries throughout much of Asia
and common law jurisdictions, such that an employer can unilaterally issue a written notice of termination with
reason at least 14 working days prior to the planned termination date. If the employee accepts the termination,
the employer must report the termination to the local labor office.

6.3 Implementation Challenge

This chapter has presented an ambitious reform program. It has outlined a broad, yet
curated, range of policy areas that need reform. The policies proposed only brush the
surface of many of the intervention areas; the report selects a few key policies that are
necessary and could be prioritized, but are far from sufficient. Implementing the agenda is
a national task, requiring the efforts of a broad range of ministries, levels of government,
the private sector, and citizens themselves. More often than not, in a country as diverse and
decentralized as Indonesia, there is no one answer on the “how to” of implementing reforms.
For example, a reform may be more easily implemented in some sectors or regions than in
others. Hence, discussion of each of the above policy recommendations needs to be further
disaggregated into subsets of “how to.” Reformers must also take into account institutional
and political constraints as reforms do not happen in a vacuum.25 This means that reforms
must be endogenous to secure political will and support. “Quick wins” or success stories
might be useful as stepping stones to gain political support and credibility along the way
to a more complex reform that may only bear fruit in the distant future.26 Given the need
for reforms to be endogenous, the policy recommendations would benefit from more
elaborated discussions on how to achieve these quick wins, especially for those reforms
that will only show results in the longer term.

Indonesia has shown that it can take on the challenge of implementing a multifaceted reform
agenda for jobs. It is one of few countries that has approached the jobs agenda holistically,
engaged stakeholders across the spectrum, and deliberated a law such as the Omnibus
Law for Job Creation (Omnibus Law). The accountability structure and implementation
monitoring (as well as enforcement in some reform areas) will be critical and, perhaps, even
more of a challenge than the preparation of the Omnibus Law itself. Unlike sector-specific

25 Basri 2017
26 Basri 2017

Pathways to Middle-Class Jobs in Indonesia June 2021

258 Report Summary and Policy Directions

reforms, the jobs reform agenda requires engagement across almost all vertical and
horizontal government actors and must operate above the sectoral ministerial level, in the
form of a supraministerial coordinating body. More importantly, the government cannot do
this alone. The reform efforts must include nongovernmental actors: employer associations
and the private sector; labor, trade, and farmers unions; the judiciary; international
development partners; civil society organizations; and academics, including a consortium of
academic institutes and think tanks such as the Employment Policy Forum (Forum Kebijakan
Ketenagakerjaan) established in 2014.27

Indonesia is aiming to become a high-income country by 2045. The message from the highest
level is that jobs are a fundamental part of realizing this vision.

Significant analytical work is still needed to further refine the policy responses. The policies
presented in this chapter are drawn from in-depth Indonesia studies of some topics and
global literature on others. However, there is a notable shortage of deep-dive empirical
studies on the topics of high value-added services sectors (and the services sector more
generally), agricultural jobs, how worker protection laws do (or do not) contribute to middle-
class jobs, HHEs, and the effect of global megatrends including digital development, among
many others.

27 Forum Kebijakan Ketenagakerjaan. N.d. https://laborforum-id.org/

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Report Summary and Policy Directions 259

Table 6.1 Summary of Policy Directions

Policy Selected Policy Actions Actors

Reform Area

Reform Strategy 1: Accelerate Across-the-Board Productivity Growth

Unlock new Improve foreign direct investment (FDI) policies, increase Office of the President; Coordinating
firm entry access to inputs, markets, and foreign talent for occupations Ministry of Economic Affairs
and growth where there is a shortage of domestic workers, and integrate (Kementerian Koordinator Bidang
to create firms with global value chains (GVCs) by: (i)  reducing the list Perekonomian; CMEA); Ministry of
competition of sectors reserved for Indonesian micro, small, and medium Finance; Investment Coordinating
and innovation enterprises (MSMEs) and turning burdensome minimum local Board (Badan Koordinasi Penanaman
that spur content requirements across various sectors into positive Modal; BKPM); Ministry of Trade;
productivity incentives to use local supply; (ii) eliminating the Expatriate Ministry of Industry; Ministry of
in the private Manpower Utilization/Placement Plan as a precondition for Manpower; Ministry of Cooperatives
sector. employment of expatriates in high-skill occupations for which and SMEs; Ministry of Foreign Affairs;
there is a shortage of domestic workers; (iii) reviewing, reducing, Ministry of Law and Human Rights;
and streamlining nontariff measures including measures Directorate General of Immigration
around product certification, preshipment inspections, and
requirements for ministerial letters of recommendation;
(iv) pursuing ambitious preferential trade agreements to lock
in domestic reform, and attract investment by finalizing the EU
Comprehensive Economic Partnership Agreement.

Increase the capacity of domestic firms and workers to benefit Ministry of Trade; BKPM; Ministry of
from the technology spillovers of FDI by: (i)  increasing the Research and Technology; Ministry of
human capital of domestic workers (see Reform Strategy 3); Industry
(ii) promoting links between firms; (iii) providing incentives and
advocacy to promote firm-based training programs; (iv) building
business networks, (v) establishing institutional partnerships,
especially between domestic and foreign-owned firms; and
(vi) providing information services covering, for example, the
quality requirements of multinationals.

Support firms with a more predictable regulatory framework Office of the President; Cabinet
by: (i) introducing a regulatory oversight body; (ii) introducing Secretariat; Ministry of National
compulsory public consultation procedures; and (iii) Development Planning; Ministry of
consolidating and targeting policy instruments that better Law and Human Rights
assist firms to innovate.

Strengthen the competition regulatory framework by: CMEA; KPPU

(i)  strengthening the technical capacity of the Commission
for the Supervision of Business Competition (Komisi
Pengawas Persaingan Usaha; KPPU); (ii)  revisiting the existing
Competition Law (No. 5/1999); (iii) mainstreaming competition
considerations in the policy-making process.

Pathways to Middle-Class Jobs in Indonesia June 2021

260 Report Summary and Policy Directions

Improve Develop specific measures and strategies to assimilate SMEs CMEA; Ministry of Industry; Ministry
productivity into GVCs. Smallholders should be integrated into agriculture of Cooperatives and SMEs; Ministry of
of small and value chains and home-based workers into manufacturing and Trade
medium services GVCs by: (i) addressing information gaps between SMEs
enterprises and multinational corporations regarding quality standards; (ii)
(SMEs) to help supporting SMEs to meet these quality standards; (iii) providing
them grow tax incentives for multinationals that invest in the upgrading
and become of local suppliers; and (iv) lowering transaction costs for SMEs
better creators using free trade agreements.
of middle-class
jobs. Promote incremental increases in innovative activities to Ministry of Research and Technology;
boost productivity with a mix of policy instruments including Ministry of Cooperatives and SMEs;
(i) technology extension and diffusion programs, (ii) early- Ministry of Trade
stage infrastructure and advisory services, (iii) inducement
incentives, (iv) vouchers for collaboration, (v) direct grants for
business innovation.

Close the digital infrastructure divide and ensure affordable, Ministry of Information and
reliable, and ubiquitous high-speed broadband internet Communications Technology, in
access (fixed-broadband and 4G) by: (i)  reforming spectrum collaboration with telecoms/internet
management and infrastructure sharing to stimulate further service providers (private sector);
private investment; and (ii) closing remaining gaps in internet Ministry of Public Works; Ministry of
access through targeted public-private partnerships. State-Owned Enterprises

Expand access to financial services by: (i) preparing a roadmap Bank Indonesia; Financial Services
targeting a leveling of the playing field for the delivery Authority (Otoritas Jasa Keuangan;
of financial services; (ii) strengthening financial sector OJK);
infrastructure such as credit registries and infrastructure Ministry of Finance; financial
related to secured transactions; and (iii) promoting the use of intermediaries
dedicated liquidity and risk-sharing facilities.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Report Summary and Policy Directions 261

Target support Increase basic business skills through training and education Ministry of Cooperatives and SMEs;
to increase by: (a) creating short training programs with simplified rule- Ministry of Manpower; Ministry of
productivity of-thumb modules in business skills combined with personal Education; OJK
in household initiative (soft skills) training and innovation training;
enterprises (b) providing information to HHEs about the benefits of training
(HHEs), which and possible routes to enrolling in training; (c) in the medium
house a large term, synchronizing and integrating government training and
share of jobs. support programs across ministries and bodies.

Create inclusive ecosystems for HHEs by: (a) encouraging HHE CMEA; BPJS Ketenagakerjaan; Ministry
owners to join the social security system Badan Penyelenggara of Manpower; Ministry of Finance;
Jaminan Sosial (BPJS) Ketenagakerjaan for nonwage workers Ministry of Social Affairs; Ministry of
by providing tailored information on the benefits of doing so; Cooperatives and SMEs
(b) providing government contributions to ease entry into and
sustain participation in the program; (c) promoting the benefits
of becoming a legal entity by bundling legal registration with
other government support programs; and (d) improving the
delivery of support programs, and broadening and facilitating
HHE access to them, independently of their participation in
social assistance programs.

Improve product competitiveness by: (i) promoting on-the-job Ministry of Cooperatives and SMEs;
training and apprenticeships; (ii) providing shared factories or Ministry of Manpower; technical
machines that can be used by HHEs; (iii)  supporting product ministries, including Ministry of
standardization and modernization of HHEs. Agriculture and Ministry of Industry;
National Standardization Body of
Indonesia (Badan Standardisasi
Nasional)

Pathways to Middle-Class Jobs in Indonesia June 2021

262 Report Summary and Policy Directions

Reform Strategy 2: Transition to Sectors and Firms That Create Good-Quality Jobs

Prioritize Prioritize investment promotion strategies to draw FDI to CMEA; BKPM; Ministry of Finance;
investment sectors and projects (including infrastructure projects) Ministry of Trade; Ministry of
promotion that are amenable to the creation of middle-class jobs by Cooperatives and SMEs; Ministry of
strategies to (i) revisiting the current strategic roadmap of the BKPM Manpower
attract FDI (including the goals for infrastructure development) and
to sectors applying a lens of middle-class jobs creation using a global best-
or projects practice framework, such as the OECD Policy Framework for
(including Investment, when considering which investment to prioritize;
infrastructure also, (ii) revisiting how prioritization is implemented and
projects) that evaluating in the medium term whether the prioritization
are amenable strategy has achieved the intended result of more middle-
to the creation class jobs. Also, to the extent that reforms in the SME and HHE
of middle-class sectors are not sector-blind, select sectors and enterprises to
jobs. target with policy interventions by considering their potential
for creating middle-class jobs.

Close Upgrade KarirHub and the overall labor market information Ministry of Manpower; Ministry of
information system by: (i) including private jobs portals and public portals Social Welfare; Ministry of Education
gaps to help providing services for job search and career development
workers within the system; and (ii)  populating the labor market
transition to information system platform with real-time information.
higher-quality
jobs. Provide relocation support for hard-to-fill vacancies by:
(i)  ensuring the labor market information system provides
information on nationwide job vacancies; and (ii)  providing
stipends to support relocation (travel, housing and other
social services) for workers who are relocating to fill critical
occupations.

Continue to develop the Job Loss Guarantee Program (Jaminan

Kehilangan Pekerjaan; JKP) to support workers’ transitions up
the job ladder by: (i) allowing current workers to opt in (or out)
of the program while mandating all newly hired workers to
join the system; (ii) working with social partners to define the
program parameters—including financing mechanisms—and
to implement a communications campaign about the benefits
of the reform for workers and firms; (iii) developing a strong
monitoring system to avoid misuse of the program; and (iv)
establishing strong links with the rest of the labor market
information ecosystem.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 Report Summary and Policy Directions 263

Reform Strategy 3: Build a Middle-Class Workforce

Facilitate Provide support to at-risk students to encourage them to Ministry of Education, Culture and
learning. complete secondary school by: (i) mapping risk factors for at- Higher Education (Directorate of
risk populations and designing interventions accordingly; and Community Education and Special
(ii) expanding the reach of and modifying the service delivery Education); local education authorities
mechanism of the Paket C diploma equivalency program. (Dinas Pendidikan)

Support the development of short TVET courses for adult Ministry of Manpower; Ministry of
workers and also support their access to these courses by: Information, Communication and
(i) accrediting web-based training courses; (ii) providing training Technology; Ministry of Education,
vouchers for adult learners in critical occupations or skills; and Culture and Higher Education; National
(iii) developing a database of accredited online training courses. Agency for Professional Certification
(Badan Nasional Sertifikasi Profesi);
National Accreditation Committee
(Komite Akreditasi Nasional)

Teach nonroutine interpersonal, analytical, and digital skills Ministry of Education, Culture
to students and adult learners through revised curricula and and Higher Education; Ministry of
pedagogical methods. Manpower; Ministry of Information,
Communication and Technology

Enhance the quality of TVET programs by: (i)  prioritizing the Coordinating Ministry of Human
development of the competency framework and the process Development and Cultural Affairs;
for accrediting institutions and programs; (ii) developing CMEA; Ministry of Finance; Ministry
a centralized monitoring and evaluation system for TVET; of Education, Culture and Higher
(iii) establishing a National Vocational Skills Steering Committee; Education; Ministry of Manpower,
and (iv)  more effectively engaging the enterprise sector by National Agency for Professional
amending Regulation 128/2019 to subsidize results-based Certification; National Accreditation
worker training, and earmarking public resources to incentivize Committee
exchanges of staff between industry and institutions providing
TVET.

Provide tailored Fine-tune the labor market information system and training Ministry of Manpower; Ministry
support to programs to address factors that limit the success of women of Education, Culture and Higher
particular and youth in acquiring middle-class jobs by: (i) including gender- Education; and technical ministries
groups. and age-relevant information in the labor market information
system; (ii) earmarking a portion of training vouchers for
women and youth to participate in courses that teach business
practices and the technology needed for engaging in the gig
economy; and (iii) incorporating business development training
into entrepreneurship support programs for women.

Expand women’s access to social benefits through their jobs CMEA; Ministry of Manpower; Ministry
by: (i) modifying the Labor Law to protect against dismissal or of National Development Planning,
disciplinary measures during pregnancy or maternity leave; Healthcare and Social Security Agency
and (ii) establishing a grievance procedure process related to (BPJS)
firm payment of social benefits.

Enact legislation to set national norms for women’s access to Ministry of Manpower; Ministry of
the labor market by: (i) clearly defining the “equal pay for equal Women’s Empowerment and Child
work” language in the Labor Law; (ii) introducing an article Protection
that defines sexual harassment; and (iii) extending paternity
benefits to 30 days.

Pathways to Middle-Class Jobs in Indonesia June 2021

264 References

References
Anas, Titik, Carlos Mangunsong, and Nur Afni Panjaitan. 2017. “Indonesian SME Participation
in ASEAN Economic Integration.” Journal of Southeast Asian Economies 34 (1) 77–117.

Basri, M. Chatib. 2017. “Reform in an Imperfect World: The Case of Indonesia.” Asian Pacific
Economic Literature 31 (2) 3–18. https://onlinelibrary.wiley.com/doi/abs/10.1111/
apel.12195.

BKPM (Indonesian Investment Coordinating Board). N.d. “Strategic Roadmap.” https://

www2.bkpm.go.id/en/about-bkpm/strategic-roadmap.

Chang, Philip Han, Rachita Daga, Gautam Vishram Mehta, and Massimiliano Calì. 2019.

Creating Markets in Indonesia. Unlocking the Dynamism of the Indonesian Private
Sector: A Country Private Sector Diagnostic. Washington, DC: World Bank Group.

Cirera, Xavier, Jaime Frias, Justin Hill, and Yanchao Li. 2020. A Practitioner’s Guide to
Innovation Policy: Instruments to Build Firm Capabilities and Accelerate Technological
Catch-Up in Developing Countries. Washington, DC: World Bank. https://
openknowledge.worldbank.org/handle/10986/33269 License: CC BY 3.0 IGO.

Harding, Torfinn, and Beata S. Javorcik. 2011. “Roll Out the Red Carpet and They Will Come:
Investment Promotion and FDI Inflows.” Economic Journal 121 (557): 1445–76.

Lin, Justin. 2012. New Structural Economics: A Framework for Rethinking Development
Washington, DC: World Bank.

OECD (Organisation for Economic Co-operation and Development). 2018a. SME

and Entrepreneurship Policy in Indonesia 2018. OECD Studies on SMEs and
Entrepreneurship. Paris: OECD Publishing.

OECD. 2018b. Mapping of Investment Promotion Agencies in OECD Countries. Paris: OECD
Publishing. www.oecd.org/investment/Mapping-of-Investment-Promotion-
Agencies-in-OECD-Countries.pdf.

OECD and ERIA (Economic Research Institute for ASEAN and East Asia). 2018. SME Policy
Index: ASEAN 2018: Boosting Competitiveness and Inclusive Growth. Paris: OECD
Publishing; Jakarta: ERIA.

Pratomo, Devanto Shasta. 2020. “Labor Supply and Policies in Expanding the Middle Class
in Indonesia.” Background paper, Pathways to Middle-Class Jobs in Indonesia, World
Bank, Jakarta.

Purnagunawan, Raden Muhamad. 2020. “Household Enterprise: Challenges and Policies to

Improve Productivity.” Background paper, Pathways to Middle-Class Jobs in Indonesia,
World Bank, Jakarta.

June 2021 Pathways to Middle-Class Jobs in Indonesia

 References 265

World Bank. 2020a. Indonesia Systematic Country Diagnostic. Indonesia: Eliminating Poverty,
Bringing Economic Security to All. World Bank.

World Bank. 2020b. The Promise of Education in Indonesia. Washington, DC: World Bank.

World Bank. 2020c. World Development Report 2020: Trading for Development in the Age of
Global Value Chains. Washington, Washington, DC: World Bank. doi:10.1596/978-1-
4648-1457-0. License: Creative Commons Attribution CC BY 3.0 IGO.

World Bank and CMEA (Coordinating Ministry of Economic Affairs). 2020. Indonesia’s Critical
Occupations List 2018: Technical Report (English). Washington, DC: Social Protection
and Jobs Global Practice, World Bank. http://documents.worldbank.org/curated/
en/763611585857010121/Indonesias-Critical-Occupations-List-2018-Technical-
Report.

World Bank. 2021. Program Document for a Proposed Loan: Indonesia Investment and
Trade Development Policy Loan. World Bank, Washington, DC.

Pathways to Middle-Class Jobs in Indonesia June 2021

266

June 2021 Pathways to Middle-Class Jobs in Indonesia

 267

Pathways to Middle-Class Jobs in Indonesia June 2021

Pathways
268 to Middle-Class Jobs in Indonesia

June 2021 Pathways to Middle-Class Jobs in Indonesia