Unicorns FULL REPORt Harnessing Digital Technologies for Inclusion in Indonesia Beyond WORLD BANK INDONESIA FLAGSHIP REPORt © 2021 THE WORLD BANK 1818 H STREET NW, WASHINGTON DC 20433 TELEPHONE: 202-473-1000; INTERNET: WWW.WORLDBANK.ORG Some rights reserved This work is a product of the staff of The World Bank. 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. For information about WWW.WORLDBANK.ORG/ID the World Bank and its activities in Indonesia, @BANKDUNIA please visit: BANKDUNIA INSTAGRAM.COM/WORLDBANK WWW.LINKEDIN.COM/COMPANY/THE- W O R L D - B A N K All photos in this report are from Unsplash.com and are royalty free. ↳ HELP.UNSPLASH.COM/EN/COLLECTIONS/1463188-UNSPLASH-LICENSE) Satu Kahkonen� COUNTRY DIRECTOR World Bank Indonesia I ndonesia has seen a rapid rise in the number of people connected to the & Timor Leste internet over the last decade. A vari- ety of sources – including new ones documented in this report – show that Indonesians connected to the in- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA ternet use it quite intensively, spend- ing as many as six hours a day on aver- age to communicate with each other, participate in social media, stream content for leisurely consumption and increasingly also engage in com- mercial activity. This growing base of consumers of internet and internet enabled services has made Indonesia the largest and the fastest growing digital economy in South East Asia. The country is home to many of the sub-region’s largest digital platforms that are not only attracting large vol- ume of investments into the country but are also providing new and inno- vative solutions that are increasingly transforming the economic and social lives of Indonesians. Across the world, effective and smart use of digital technologies such as the internet and digital platforms, has emerged as a key marker of resilience for coping with the devasting impact of the COVID-19 pandemic on eco- nomic and social life. The same is true also for Indonesia. Firms have adopt- ed these technologies at unprece- dented rates to ensure the continuity of their operations. Households and individuals have increasingly relied on these technologies to work from home when they can, purchase con- sumer goods online, help children IV Foreword stay connected with schoolwork and achievements in expanding digital All of this would also require the even carry out medical consultations connectivity. Yet, almost half of the Government to lead from the front remotely. Government agencies have adult population is still without inter- and not only become more digital used these technologies not only to net access and, as such, automatically itself but also put guardrails for a try and address the public health chal- excluded from the country’s digital responsible digital economy. Taking lenges associated with the pandemic dividends. Even for the given level of a whole of government approach to but also to maintain the continuity connectivity, the quality of the inter- digitally integrate and streamline of their services and, importantly, to net experience in Indonesia is poorer systems and to create common and channel social assistance and other than in neighboring countries, which interoperable service platforms, support to those most affected. makes it a shaky foundation for more especially for the use and re-use of productive, high bandwidth applica- data, could generate payoffs across As Indonesia’s economy recovers, the tions, especially by firms but also by the whole society. For instance, accelerated utilization of digital tech- schools and medical facilities. developing a digital identification nologies can have a big impact on the framework to equip Indonesians country’s development. Firms can But digital divide does not have to be with the ability to securely prove become more productive, which can destiny. Taking stock of the current their identity when transacting on- drive growth. Workers can be pro- stage of digital technology adoption line would boost trust in the digital ductive and see their earnings grow. in Indonesia and identifying the var- economy, reduce fraud, and enable The Government too could utilize ious dimensions along which these more public and private sector ser- these technologies in various ways technologies could in fact widen vices to be delivered end-to-end to improve the quality and coverage gaps instead of closing them, the remotely. This would allow any In- of services it provides to its citizens, report identifies some crucial fronts donesian with internet, no matter which would be instrumental in re- on which policymakers could priori- how far away from the nearest gov- ducing inequality of opportunities. tize action to ensure that the risks of ernment office or bank branch, to For instance, mechanisms established these gaps could be mitigated. Uni- access such services. Similarly, the to identify, target and channel assis- versalizing internet is a key starting Government would benefit from tance to vulnerable citizens during point and should include interven- strengthening safeguards for digital the pandemic could become the ba- tions not just to address the hard economy transactions and Indone- sis for a permanent digitally powered infrastructure part, especially along sian in cyberspace by putting in place upgrade on social assistance targeting the last mile, but also softer inter- strong regulations and independent and delivery. ventions related to competition and enforcement regimes related to data sector regulation which have a major privacy, cybersecurity, and consum- A growing digital economy does not bearing on quality and affordabili- er protection. necessarily imply an inclusive one. ty of internet. These would need to As this report shows, the benefits of be complemented with policies and Our hope is that this report will pres- the digital economy cannot be tak- regulations that enable the digital ent a novel lens to look at this excit- en for granted by all segments of the economy to grow and active inter- ing landscape of digital technologies population. Some will thrive, some ventions to ensure that the workers in Indonesia and inform a broader may be left behind and some in fact across the country are equipped with public dialogue, especially on what is may stand to lose as well. For exam- the skills they need to make the most needed to make the country’s digital ple, the country has made remarkable of the opportunities. future more inclusive. V BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA VI Acknowledgements BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA B eyond Unicorns was prepared by a World Bank We express our sincere gratitude to Tokope- team led by Sailesh Tiwari, Senior Economist dia, Bukalapak, GoJek and Halodoc for their in the Poverty and Equity Global Practice of the collaboration and their generosity in sharing World Bank. The core team consisted of Rabia with us data that helped shed light on some key Ali, Erwin Ariadharma, I Gede Putra Arsana, questions addressed in the report. The team Lamiaa Bennis, Natasha Beschorner, Prasetya also benefited from excellent guidance and ad- Dwicahya, Jonathan Marskell, Ririn Salwa Pur- vice from the following peer reviewers at var- namasari, Laura Ralston, Jaffar al-Rikabi, Virgi ious points in the report preparation process: Agita Sari, Imam Setiawan, Bambang Suharnoko Carlos Rodriguez Castelan (Lead Economist), Sjahrir, Muhammad Ajisatria Suleiman, Maria Ana Paula Cusolito (Senior Economist), Na- Monica Wihardja and Putu Sanjiwacika Wibis- talija Gelvanovska-Garcia (Senior Digital De- ana. Other contributors included Arun Arya, velopment Specialist), Tim Kelly (Lead Digital Hamidah Alatas, Maarten Bosker, Massimiliano Development Specialist), Siddhartha Raja (Se- Cali, Aufa Doarest, Frederico Gil-Sander, Sean nior Digital Development Specialist), and Jane Hambali, Hanan Jacoby, Scott Minehane, Mo- Treadwell (former Lead Digital Development chamad Pasha, Josefina Posadas, Abror Tegar Specialist). We thank them without implicat- Pradana, Ibrahim Rahman, Anissa Rahmawati, ing them in any errors or omissions that may Jan van Rees, Mayla Safuro Putri, Husnul Rizal, remain. Helpful suggestions provided by Vivi Mark Roberts, Arya Swarnata, Michael Weber, Alatas, Wendy Cunningham, Camilla Holme- Noah Yarrow, and Yongmei Zhou. mo, Andrew Mason, and Rolande Pryce, espe- cially in the early stages of the work, are also The work was conducted under Satu Kahkonen gratefully acknowledged. (Country Director, Indonesia and Timor Les- te) with overall guidance from Hassan Zaman Funding for the report was provided through the (World Bank Regional Director, Equitable Partnership for Knowledge-based Poverty Re- Growth, Finance and Institutions) and Car- duction and the Australia World Bank Indonesia olina Sanchez-Paramo (World Bank Global Partnership (ABIP) by the Government of Aus- Director, Poverty and Equity Global Practice) tralia’s Department of Foreign Affairs and Trade. and technical guidance from Aaditya Mattoo Peter Milne edited the report, Muhammad Ka- (Chief Economist, East Asia, and Pacific Re- mal and Indra Inrawan designed the layout and gion), Rinku Murgai (Practice Manager, Pov- Dyah Nugraheni and Elisabeth Ekasari provided erty and Equity Global Practice) and Deepak impeccable administrative support through all Mishra (Practice Manager, Macroeconomics, stages of report preparation. Trade and Investments Global Practice). VII Abbreviations & Acronyms 2G Second Generation (basic mobile for SMS) 3G Third Generation 4G Fourth Generation 5G Fifth Generation AI Artificial Intelligence ASEAN Association of South East Asian Nations ASN Indonesian Public Officials (Aparatur Sipil Negara) B2C Business-to-Consumer Bappenas Ministry of National Development Planning (Kementeri Perencanaan Pembangunan Nasional/Badan Perencanaan Pembangunan Nasional) BEPS Base Erosion and Profit Shifting FRONT PAGES BIG Geospatial Information Agency (Badan Informasi Geospatial) BPNT Non-Cash Food Subsidy (Bantuan Pangan Non-Tunai) BPPT Agency for Assessment and Implementation of Technology (Badan Pengkajian dan Penerapan Teknologi) BSA Basic Saving Account BSSN National Cyber and Crypto Agency (Badan Siber dan Sandi Negara) BTS Based Transceiver Station C2C Consumer-to-Consumer BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA CDP Collection-and-Delivery Point CHIP Connect, Harness, Innovate and Protect CIT Corporate Income Tax CMEA Coordinating Ministry for Economic Affairs COVID-19 Coronavirus Disease 2019 DEHS Digital Economy Household Survey DFS Digital Financial Services DGT Directorate General of Taxes DKI Special Capital Region (Daerah Khusus Ibukota) DSL Digital Subscriber Line DTIA Digital Transformation Implementation Agency DTKA Unified Social Welfare Database (Data Terpadu Kesejahteraan Sosial) DTT Digital Transformation Taskforce Dukcapil Directorate General for Population and Civil Registration (Direktorat Jenderal Kependudukan dan Pencatatan Sipil) EdTech Education Technology EU European Union FSP Financial Service Provider FTTH Fiber to the Home G2P Government-to-Person GB Giga Byte GDP Gross Domestic Product GMV Gross Merchandise Value GNI Gross National Income GoI Government of Indonesia HealthTech Health Technology ICT Information and Communication Technology ID Identification IDR Indonesia Rupiah IIGF Indonesia Infrastructure Guarantee Fund IMT International Mobile Telecommunications IoT Internet of Things Kabupaten District Kartu Pra Kerja Pre-Employment Card Kecamatan Subdistrict KK Family Card (Kartu Keluarga) Kota City KPK Anti-Corruption Commission (Komisi Pemberantasan Korupsi) VIII KTP Indonesian Identity Card (Kartu Tanda Penduduk) KTP-el Indonesian Electronic Identity Card (Kartu Tanda Penduduk Eletronik) KYE Know-Your-Customer LAN State Administration Institution (Lembaga Administrasi Negara) LKD Digital Financial Services (Layanan Keuangan Digital), LP Branchless Banking (Laku Pandai) LSP Logistics Service Providers LTE Long-Term Evolution Mbps Megabits Per Second Menkominfo Ministry of Communications and Information MenPAN-RB Ministry of Administrative and Bureaucratic Reform (Kementerian Pendayagunaan Aparatur Negara dan Reformasi Birokrasi) MHz Mega Hertz MNE Multi-National Enterprise MNO Mobile Network Operators MoEC Ministry of Education and Culture MoF Ministry of Finance FRONT PAGES MoHA Ministry of Home Affairs MOOCs Massive Open Online Courses MRT Mass Rapid Transit MSME Micro, Small and Medium Enterprises NIK Unique Identity Number (Nomor Induk Kependudukan) O2O Online-to-Offline OER Open Education Resources BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA OKJ Financial Services Authority (Otoritas Jasa Keuangan) OTP One-Time Password Pembina Data Data Steward PIT Personal Income Tax PKH Conditional Cash Transfer Program (Program Keluarga Harapan) PLN State-Owned Electricity Utility (Perusahan Listrik Nasional) PMO Project Management Office PNG Papua New Guinea PPP Private-Public Partnership PNS Civil Servant (Pegawai Negeri Sipil) PSTN Public Switched Telephone Network QoS Quality of Service R&D Research and Development RAN Rural Access Network Raskin Subsidized Rice Program (Beras Miskin) Ro-Ro Roll-on, Roll-off SAKERNAS Indonesia Labor Force Survey (Survei Angkatan Kerja Nasional) SEA-US Southeast Asia–United States SFH Studying from Home SIAK National Population Registry System (Sistem Informasi Administrasi Kependudukan) SIM Subscriber Identification Module Sislognas National Logistics System (Sistem Logistik Nasional) SME Small and Medium Enterprises SMPCS Sulawesi-Maluku-Papua Cable System STEM Science, Technology, Engineering and Mathematics SUSENAS Indonesia Socioeconomic Survey (Survei Sosial Ekonomi Nasional) Tbps TeraBits Per Second, or Trillion Bytes per Second Telco Telecommunications Service Provider TVET Technical and Vocational Education and Training UNESCO United Nations Educational, Scientific and Cultural Organization USO Universal Service Obligation Fund VAT Value-Added Tax VoLTE Voice over LTE Wali Data Data Custodian Warung Street food vendor WFH Working from Home yoy Year-on-Year IX Contents FOREWORD v ACKNOWLEDGEMENT vi ABBREVIATIONS AND ACRONYMS vii FRONT PAGES EXECUTIVE SUMMARY 1 OVERVIEW 10 ACCESS TO THE MEDIUM 16 HARNESING THE MEDIUM TO BOOST INCOME 27 USING THE MEDIUM TO DELIVER BETTER SERVICES 44 POLICIES TO LEVERAGE DTs FOR GREATER INCLUSION 52 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA REFERENCES 68 CHAPTER 1: DIGITAL TECHNOLOGIES FOR INCLUSION – THE PROMISE AND THE CURRENT LANDSCAPE 74 1.1. INTRODUCTION 76 1.2. THE PROMISE AND AN ORGANIZING FRAMEWORK 78 1.3. DIGITAL TECHNOLOGY LANDSCAPE – WHAT IS THE STATE OF PLAY? 80 1.4. CONCLUSION 93 ANNEX 96 SPOTLIGHT 1: USE OF SOCIAL MEDIA IN INDONESIA AND IMPLICATIONS FOR INCLUSION 98 CHAPTER 2: AVAILABILITY, ACCESS AND ADOPTION: THE DRIVERS OF DIGITAL DIVIDE IN INDONESIA 104 2.1. INTRODUCTION 105 2.2. EVOLUTION AND THE STATE OF DIGITAL INFRASTRUCTURE 107 2.3. AFFORDABILITY AND QUALITY OF THE INTERNET EXPERIENCE 116 2.4. DRIVERS OF THE DIGITAL DIVIDE 120 2.5. CONCLUSION AND POLICY IMPLICATIONS 126 ANNEX 129 SPOTLIGHT 2: KEY FINDINGS FROM A MACROECONOMIC TAKE ON INDONESIA’S DIGITAL ECONOMY 130 X CHAPTER 3: LEVERAGING DIGITAL TECHNOLOGIES FOR BETTER JOBS AND ECONOMIC OPPORTUNITIES 134 3.1. INTRODUCTION 135 3.2. DIGITAL ADOPTION BY FIRMS 136 3.3. E-COMMERCE IN INDONESIA: PATTERNS, GROWTH AND THE BARRIERS TO ADOPTION 139 FRONT PAGES 3.4. JOBS IN THE GIG ECONOMY 148 3.5. THREE DISTRIBUTIONAL TENSIONS AND A LOOK TO THE FUTURE 150 3.6. CONCLUSION 154 ANNEX 160 SPOTLIGHT 3: EXPANDING THE REACH OF E-COMMERCE TO RURAL INDONESIA: LESSONS FROM TAOBAO VILLAGES IN CHINA 163 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA CHAPTER 4: DIGITAL TECHNOLOGIES FOR INCLUSIVE SERVICE DELIVERY 168 4.1. INTRODUCTION 169 4.2. EDTECH, HEALTHTECH AND THE ROLE OF DIGITAL TECHNOLOGIES IN EDUCATION AND HEALTH 171 4.3. INTRODUCTION OF DIGITAL IDENTIFICATION 176 4.4. DIGITAL TRANSFORMATION OF GOVERNMENT 180 4.5. CONCLUSIONS 186 SPOTLIGHT 4: THE LANDSCAPE OF INDONESIA’S PRIVATE EDTECH AND HEALTHTECH SECTOR 190 CHAPTER 5: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION – WHAT CAN INDONESIA DO? 196 5.1. IMPROVE DIGITAL CONNECTIVITY TO UNIVERSALIZE ACCESS 198 5.2. MAKE THE DIGITAL ECONOMY WORK FOR ALL 200 5.3. USE DIGITAL TECHNOLOGIES TO PROVIDE BETTER SERVICES AND UPGRADE CITIZEN-STATE INTERACTIONS 208 ANNEX 218 XI LIST OF FIGURES 1.1 Harnessing digital technologies for greater inclusion – three pathways 79 1.2 Access to the internet has been increasing over the past decade 80 FRONT PAGES 1.3 Some parts of the country are still lagging 81 1.4 Richer Indonesian’s have better access 81 1.5 Internet access is also higher among the younger, more educated and male segments of the population 81 1.6 Only 5 percent of population is connected to fixed broadband at home 82 1.7 Smartphone ownership rates have increased sharply over the years 82 1.8 Indonesians who are connected to the internet use it quite intensively 83 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA 1.9 Communications, social media and leisure activities account for 80 percent of average time spent online 83 1.10 One-quarter of Indonesians use ride-hailing apps, and almost one-fifth use it for their daily commutes 84 1.11 Strong complementarity with public transportation suggests that these services fill a crucial gap in urban mobility 85 1.12 Digital ride-hailing has also brought other conveniences 86 1.13 E-commerce more prevalent in populous and affluent provinces 87 1.14 E-commerce prevalence also remains higher among…. 88 1.15 Price and convenience are the most dominant reasons for buying online… 89 1.16 Fiber optic arrival in the eastern provinces has boosted internet access, but has otherwise had a minimal influence on 90 the underlying structure of the economy 1.17 Indonesia punches slightly above its weight on digital adoption index for the Government…. 92 1.18 …but the capability to provide digital public services is below its regional peers 92 2.1 Last mile connectivity challenges persist 105 2.2 Java-Bali and Sumatra still have the highest number of Indonesians not connected 106 2.3 Broadband infrastructure value chain 107 2.4 Submarine cables connecting Indonesia to the main international routes 108 2.5 Backbone networks are also sufficiently well deployed by service providers 110 2.6 Palapa Ring project 112 2.7 Mobile broadband access networks have also expanded 113 2.8 Indonesia lags behind many ASEAN neighbors on fixed broadband penetration 114 2.9 Fixed broadband penetration is also below countries at similar levels of development 115 2.10 Overall internet affordability in Indonesia is on the lower side… 116 2.11 …and the price of mobile broadband data is not the issue 116 2.12 Fifty-five percent of adults who do not use internet-enabled devices cite cost as the primary barrier to adoption 117 2.13 Affordability is a key barrier to the adoption of fixed broadband for many 118 2.14 Both fixed and mobile broadband download speeds in Indonesia are among the lowest in the ASEAN region 118 2.15 Quality of the internet experience varies significantly across the country 119 2.16 While the mobile broadband space is competitive…. 122 2.17 …the FTTH market is more concentrated, with Telkom dominating market share 122 2.18 Core digital skills among internet-using adults in Indonesia 124 2.19 Younger and more educated users have higher levels of digital skills 125 3.1 Share of firms adjusting their business process due to COVID-19, by type of adjustments 136 3.2 Share of firms reported adjusting in response to COVID-19 136 XII 3.3 Internet use by non-agricultural enterprises is still very low, with much higher incidence among medium and large 137 enterprises FRONT PAGES 3.4 Internet adoption is much higher among richer household enterprises 137 3.5 E-commerce has touched all parts of Indonesia but is still largely Java-centric 141 3.6 Adoption of digital financial services is extremely low 144 3.7 Access to and trust in digital financial transactions among the main reasons for not buying online 145 3.8 Among those who buy online, a majority still prefer cash on delivery 145 3.9 Not having accessed DFS is associated with lack of awareness and knowledge 146 3.10 Top 1% of sellers capture almost half of the sales 151 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA 3.11 Market competition is the dominant concern for vendors in platforms… 151 3.12 …with the crowded field being considered the most challenging aspect of market competition by over 60 percent 151 3.13 Indonesian firms face a deep skills mismatch, particularly for managers and professionals. 153 4.1 Better off households and those in Jakarta are more likely to do online learning during the COVID-19 pandemic 172 4.2 Distribution of Halodoc transactions across regions 172 4.3 Most EdTech services are concentrated in Jakarta 173 4.4 Only around 10 percent of all schools are connected to fixed broadband 175 4.5 Digital ID systems depend on having strong foundational ID systems in place 177 4.6 The Principles on Identification for Sustainable Development provide a guiding framework on how countries can 178 realize the transformational potential of digital ID systems, while mitigating the risks 4.7 Institutional structure of key e-government stakeholders in Indonesia 181 4.8 UK Government’s cross-sectoral digital transformation governance structure 182 4.9 Fragmented data and overly complex interdependence prevents BPJS Kesehatan from running efficiently 184 4.10 One Data as an opportunity to address data fragmentation 185 5.1 Transition from e-Government to digital transformation 210 5.2 A comprehensive national digital transformation agenda would require adjustments to the governance structure of 211 ongoing digital government and policy initiatives S1.1 Most recent use of various social media platforms (share of internet users) 99 S1.2 Profile of social media users in Indonesia (share of internet users) 99 S1.3 The most common activities on social media 99 S1.4 The most common topics in social media discussions 99 LIST OF TABLES 1.1 Greater internet access is benefiting higher-skilled workers more than the lower skilled ones 91 2.1 Current spectrum assignments for mobile services 120 3.1 E-commerce activity rate is higher when a job is a secondary rather than a main job, especially among women 140 3.2 E-commerce helps women return to work after giving birth 140 S4.1 Description of EdTech products with selected examples 191 S4.2 Description of HealthTech products with selected examples 192 XIII BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA EXECUTIVE SUMMARY XIV Chapter ⬎ ES Executive Summary EXECUTIVE SUMMARY BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA S imilar to many other countries around the world, of digital technologies. Businesses, both large the COVID-19 pandemic has hit Indonesia hard. and small, have flocked to digital technologies Latest estimates suggest that about 5.1 million to try to ensure the continuity of their opera- people—equivalent to 2.4 percent of the work- tions. School closures have forced students and ing-age population—have lost their jobs, while an teachers to adapt and explore digitally enabled additional 24 million have had to work reduced remote learning options, including the adoption hours due to the pandemic. As many as 50 per- of a variety of EdTech solutions. HealthTech apps cent of workers have experienced a reduction in enabling remote consultations and the delivery earnings. The impact on living standards has been of medicine have seen unprecedented growth in devastating, with more than 2.2 million Indone- adoption rates. Confined at home due to mobili- sians estimated to have been pushed into COVID- ty restrictions, Indonesians have switched to the 19-induced poverty in 2020. internet for their entertainment and social needs, driving sharp growth in the usage of digital media One unexpected silver lining from the crisis, (music and video streaming) and communica- however, has been the turbo-charged adoption tions applications. 1 5.1 With this pandemic-induced flight to digital ex- Indonesians who are connected to the internet pected to be permanent to a large extent, there use it intensively, with communication, social mil. is excitement about an even greater acceleration media and leisure applications dominating us- Indonesian in what was already the fastest growing digital age. On average, Indonesians spend around 6 have lost economy in Southeast Asia. But at the same time hours a day online, with the younger and more their jobs questions have also emerged about the possibil- educated segments more digitally engaged than ity of the differential access to and adoption of the relatively older and less educated demo- 24 digital technologies compounding existing in- graphic. Intensity of internet engagement is mil. equalities. For a country that considers achieving highest for the 16 to 25 age group, which on Had to work balanced development one of its key priorities, average spends 9.7 hours a day online. Digital reduced this is an important new challenge. ride-hailing services provided by companies hours such as Gojek—Indonesia’s homegrown de- 50% This report conducts an in-depth assessment of cacorn that epitomizes the country’s digital experienced digital technologies in Indonesia, examining the potential for policy makers—and its regional a reduction scale and extent of their current applications. competitor, Grab, are commonly used. And not in earnings Focusing specifically on how these technologies only for their mobility services to connect the are touching, shaping and influencing the lives otherwise fragmented labor markets in Indo- of Indonesians, the report also identifies some nesia’s large and sprawling metropolitan areas, policies that the Government of Indonesia (GoI) but also for several other conveniences, such as could adopt to ride the momentum of digital food delivery and logistics. Buying and selling adoption generated by the pandemic, not only online is also growing and, while still only prev- to power economic recovery in the short run but alent among a small share of the population, it BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA also to bring about greater inclusion and resil- is already enhancing consumer welfare by pro- ience in the medium run. viding cheaper options, and greater product 2.2 MIL. variety and convenience. Indonesia has made rapid progress in internet connectivity over the past decade but, despite In addition to consumers, these technologies this, half of the adult population is still without are also beginning to reshape the economic access. Moreover, the inequality in access to the opportunity landscape for workers, generating digital medium mirrors existing dimensions of newer forms of jobs in some cases and enhancing inequality—across groups, regions and income productivity of existing jobs in others. Howev- classes. While Indonesia compares well with its er, not all workers are positioned to capitalize regional peers on mobile broadband penetration on these opportunities. For example, the higher (at slower speeds), it is significantly behind on skilled workers, who as a group were already 4G/LTE and fixed broadband rollout. Issues of better rewarded in the labor market have edged availability and affordability constrain the adop- further ahead relative to lower skilled workers. tion of fixed broadband, while network conges- One of the reasons for this is that, despite the tion impinges on the quality of available mobile pandemic-induced increase in the adoption of broadband. Limited spectrum, unavailability of digital technologies, the overall level of digital specific bands—particularly the 700mHz digital adoption by firms remains low and of limited dividend band—limited regulatory clarity on in- intensity, which has naturally limited produc- Indonesians estimated to have frastructure sharing and the lack of competition, tivity gains for most workers. Another reason is been pushed into especially in the provision of fixed broadband that, while low-skill biased applications of digi- COVID-19-induced services, are the main barriers to universalizing tal technologies such as e-commerce and digital poverty access to good-quality internet throughout In- gig work are enabling certain segments of the 2020 donesia. workforce to boost income, their reach is also “Indonesia has made rapid progress in internet connectivity over the past decade, but half of the adult population is still 2 without access” limited. Digital gig work is largely concentrated among urban men. E-commerce is providing an income diversification pathway, particularly for women temporarily dislocated from labor markets and youth. However, e-commerce pen- ↳ Indonesians who are connected etration and intensity are both constrained by to the internet use it quite issues of trust in online transactions and pay- intensively, with communication, ments, logistics and internet connectivity. social media and leisure The ability of governments around the world to applications dominating usage.⬎ harness digital technologies to deliver services is perhaps one area on which the COVID-19 pan- demic has focused the strongest spotlight. Just EXECUTIVE SUMMARY as people and businesses have had to adapt to various measures taken to contain the spread of the virus, so too have governments. And the ability to use digital technologies to curb and manage the pandemic, as well as to ensure the continuity of essential services, has emerged as a key marker of resilience. Indonesia’s nascent EdTech and HealthTech BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA scene has received a major boost and, though these applications have clearly filled a void during the pandemic, their overall reach is limited to the more affluent clientele in urban centers, mostly within Java. Specifically, on ed- ucation, the lack of complementary inputs such as supporting ICT devices (smartphones and computers), and limited access to good-quality internet has limited the ability of students to access online learning during the pandemic. Ex- periences from around the world suggest that while some of these digital innovations help en- rich the menu of available solutions, the truly transformative impact can only come from the Government experimenting with, internalizing and eventually adopting some of these innova- tions at scale to enhance the delivery of services. In addition to the low connectivity in key ser- vice delivery nodes (e.g., schools and hospitals), two key challenges that Indonesia faces on this front are: (i) the absence of an officially-recog- nized and universally accessible digital ID for online transactions; and (ii) transitioning from a siloed structure of multiple, incompatible government information and data management systems to a platform-based whole-of-govern- ment approach to digital transformation, which has emerged as best practice in many economies globally. What can Indonesia do to address these chal- lenges and effectively leverage digital technol- ogies to achieve greater inclusion? The report emphasizes three key priorities for policy. 3 THREE KEY PRIORITIES The critical first priority is improv- will require investing in and facilitat- ing digital connectivity and univer- ing key enablers of the digital econo- 1 salizing access to the digital medium. my while at the same time unlocking This will require better management citizen capabilities to seize the oppor- of the telecom spectrum, regulato- tunities that are generated. To achieve ry clarity on passive infrastructure this, Indonesia will need to make sharing and improved competition progress on financial inclusion, facil- in the fixed broadband space. Under itate trust to promote greater take-up Improve Digital recently enacted Law No. 11/2020 on of digital payment services, improve Connectivity Job Creation (the so-called “Omni- logistics and scale up investment in & Universalize bus Law”), reform of spectrum man- digital skills as well as a broader set of agement is on the right track for the skills to thrive in the digital economy. Access optimization of spectrum allocation. The variety of efforts underway to im- EXECUTIVE SUMMARY The Omnibus Law also imposes a prove financial inclusion remain im- 2 mandatory two-year plan to convert portant entry-points. One promising analog television to digital television, digital solution is the reduction in ver- in order to secure a digital dividend ification costs that could come from in the 700 MHz spectrum band. This widespread availability of an e-KYC measure would be crucial in lowering option underpinned by a strong and capital costs require to provide afford- reliable digital ID system. For those able connectivity to remote parts of already banked, including some of the Make the the country. Release of the 2.6 GHz more sophisticated users, trust in on- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA band, currently used for satellite TV, line transactions and issues related to Digital would add capacity in urban centers data privacy, cyber security and finan- Economy Work and alleviate congestion. The Minis- cial integrity represent another barri- for All try of Communication and Informa- er that needs to be overcome. In this tion (Kominfo) should consider accel- regard, the passing of the draft Law erating its plans for the reallocation of on Personal Data Protection will be 3 this band from satellite broadcasting critical for promoting the adoption of to mobile broadband by the end of digital financial services in Indonesia. 2024. On citizen skills, modern tertiary ed- The Omnibus Law also mandates ucation needs to cultivate in students passive infrastructure sharing and the a minimum threshold of foundation- Adopt Digital next step for the GoI in this regard al “transferable” higher-order skills Technologies is to create an implementing regula- for the 21st century digital econo- to Upgrade tion. This will require effective lead- my, even in STEM fields. Technol- Citizen-State ership to coordinate the government ogy and integration have increased agencies involved, including, among the demand for higher-order gener- Interactions others, Kominfo, the Ministry of al cognitive skills—such as complex Home Affairs, the Ministry of Public problem-solving, critical thinking, Works and Housing, and local gov- and advanced communication—that ernment agencies. Indonesia should are transferable across jobs. Incorpo- also consider transitioning toward rating more general education in ter- the unified licensing of service pro- tiary programs is one way to do this. viders to enable each one to deliver An additional year of general edu- a larger portfolio of services. A fixed cation was added to undergraduate broadband provider should be able to programs in Hong Kong and China deliver a full portfolio of services, not in 2012 to focus on problem-solving, only broadband internet but also te- critical thinking, communication, lephony, TV, and other value-added leadership, and life-long learning services. Appointing an independent skills and that seems to have yielded regulator is critical to improving the some positive results.1 Another way overall competitiveness in the sector. is through innovative pedagogy that adopts learning strategies, including The second priority is ensuring that open-ended assessments, feedback the digital economy works for all. This opportunities, and a progressive 4 curriculum that balances academic the articulation of the vision as well as framework shields incumbents from challenges with student support. the formulation and implementation potentially more productive market of policies to drive the transformation entrants, while the unpredictable reg- The third policy priority is harness- agenda in an integrated manner. The ulatory environment further weakens ing the medium to provide better Government Digital Service, or GDS the business environment, inhibiting services and upgrade the quality of in the United Kingdom and GovTech competition and depressing invest- citizen-state interactions. This will re- in Singapore are successful examples ments. The super-efficiencies asso - quire, among others, a whole-of-econ- of this model. There could be oth- ciated with digital technologies in an omy national digital ID initiative and er models as well but success in this economic environment in which do- a whole-of-government approach to transformative agenda will rest cru- mestic firms, both large and small, are digital transformation driven by an cially on the adoption of a coordinat- shackled in this manner, and workers agency sufficiently empowered to re- ed whole-of-government strategy as lack the skills required to thrive in the solve crucial inter-agency coordina- opposed to a siloed one. second industrial revolution, let alone EXECUTIVE SUMMARY tion challenges. A national digital ID the fourth, could potentially amplify initiative would enhance and harness In addition to these three fronts there distortions and accentuate inequali- the existing strong national ID system is a crucial fourth area that should not ties. Thus, doubling down on reforms managed by the Ministry of Home Af- be forgotten. In order to truly make the to strengthen the analog foundations fairs to introduce an official digital ID digital economy more inclusive, efforts of the digital economy will be equally for online transactions and could rec- to universalize the digital medium and critical in ensuring that digital divi- ognize other forms of digital IDs, such stimulate digital innovations must be dends benefit all Indonesians. as those issued by the private sector, embedded in a broader ‘bricks-and- for interoperability purposes. In order mortar’ reform agenda that includes BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA to boost trust in online transactions greater openness and competitiveness, ↳ Indonesia’s digital and digital government, the passage stronger regulations, more compre- of the draft Law on Personal Data Pro- hensive social protection and invest- economy is growing and tection would be important to provide ments in skills for the future. The beginning to reshape safeguards and accountability for the long-standing non-digital challenges how Indonesians work, collection, use and sharing of personal that Indonesia has faced in its efforts data, as well as formalizing the rights to make growth inclusive continue to consume goods and of data subjects. An essential feature remain important constraints on how services and interact for the credibility and strength of such effectively Indonesia is able to harness with each other.⬎ a law would be independent oversight, the digital dividends for its poor. For similar to the Office of the Australian example, the high cost of logistics, Information Commissioner and the which has always prevented far flung Singaporean Personal Data Protec- markets in eastern Indonesia from tion Commission. being integrated with thicker mar- kets in Java, is precisely also one of the Beyond digital IDs, Indonesia could key binding constraints on growth of also consider reorienting from a e-commerce in these areas. Digitiza- narrow focus on e-government to a tion of the logistics sector itself can, comprehensive national digital trans- and indeed is, already beginning to formation agenda. Given the com- chip away at this. But a bigger challenge plexities, this can be implemented in lies in the analog domain and relates to phases. Strengthening the role of the the large deficit of connectivity infra- Ministry of State Apparatus and Bu- structure and the regulatory environ- reaucratic Reform (MenPAN-RB) to ment that keeps transit times long and orchestrate and deliver on the existing costs high. mandate on e-government could be coupled with parallel efforts to cre- Likewise, a whole host of other factors ate an institutional and regulatory that are binding constraints on the framework or apparatus necessary overall productivity of the economy for a more integrated and centrally are also relevant. It has been well doc- managed transformation agenda. This umented elsewhere, for example, that could be in the form of an agency that restrictive trade policies limit access to has multi-ministerial oversight and is key inputs and markets, restrictions attached to the Office of the President. on investments depress commercial Such an agency could be tasked with performance, a weak competition 5 What can Indonesia do to ensure an inclusive digital future for all? PRIORITY OBJECTIVE Improve Digital Connectivity I to Universalize Access EXECUTIVE SUMMARY 1 Optimize spectrum 2 Strengthen mechanisms to allocation for mobile ensure sharing of active and broadband passive infrastructure BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA → Implement the → Optimize the higher → Update the Telecom → Implement through “Analogue Switch Off” frequency capacity Law, PP 52/53, to regulations the by the end of 2022 in bands, starting with allow for sharing of passive infrastructure accordance with the the freeing up of the active infrastructure sharing mandated under Omnibus Law so as to 2.6 GHz band, followed on a B2B basis the Omnibus Law free up the 700MHz by the 3.4-3.8 GHz band for greater rural band and potentially RESPONSIBLE AGENCIES: RESPONSIBLE AGENCIES: connectivity using 4G the broader 3.3-4.2 Kominfo Kominfo and future 5G networks GHz band to facilitate MoHA 5G deployment in urban RESPONSIBLE AGENCIES: Ministry of Public Works areas. Consideration Kominfo should be given to the release of the mmWave spectrum bands for 5G in urban areas RESPONSIBLE AGENCIES: Kominfo → Review current → Appoint an 3 Strengthen competition licensing regime and independent regulatory along the broadband transition toward body for the telecom international best sector consistent with value chain practice of single/ exemplar practice unified licensing and Indonesia’s RCEP to allow a larger commitments number of providers to deliver the full RESPONSIBLE AGENCIES: portfolio of services Kominfo RESPONSIBLE AGENCIES: Kominfo 6 PRIORITY OBJECTIVE Make the Digital Economy II Work for All 1 2 Nurture digital skills and skills for the 21st century digital economy Support the development of logistics EXECUTIVE SUMMARY → Continue to develop → Align non-formal education → Build partnerships between basic physical services and lifelong learning tertiary institutions and infrastructure to with needs of the working-age the private sector to train improve connectivity adult population, adopting a vocational students in high- through strengthened modular approach to course tech areas using practice-based and/or rehabilitated offerings with an emphasis curricula, practitioner-led infrastructure on soft skills in addition instruction, and professional (roads, ports, to technical digital skills, certification and electricity), incorporating employment- leveraging private specific digital skills RESPONSIBLE AGENCIES: BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA sector capital relevant for Indonesia, Ministry of Education and Culture/Private and expertise as emphasizing a mindset of Sector appropriate through continuous learning, and PPPs offering opportunities to practice “self-managed” RESPONSIBLE AGENCIES: learning Ministry of Transport RESPONSIBLE AGENCIES: Directorate General of Highways Ministry of Education and Culture Ministry of Public Works Ministry of Manpower and Housing → Ensure that tertiary → Incentivize employers to → Deepen reforms education offers a minimum offer internships and off- to reduce entry threshold of foundational campus learning to students barriers to logistics “transferable” higher-order and transportation RESPONSIBLE AGENCIES: skills such as critical services and build thinking, problem-solving and Ministry of Education and Culture/ long-term investor Ministry of Manpower communication, even in STEM confidence, to expand fields warehousing outside of large metropolitan RESPONSIBLE AGENCIES: areas Ministry of Education and Culture RESPONSIBLE AGENCIES: Ministry of Transport, with → Continue to develop data systems that allow for identification support from Ministries of Trade and Finance and of occupations and skills in demand, monitor educational the Coordinating Ministry institutions’ compliance with quality standards, and ensure of Maritime Affairs and that information on employability of individuals with various Investments degrees, wage profiles, and occupation-specific courses is available to jobseekers and workers → Continue to modernize the National RESPONSIBLE AGENCIES: Post Service; Ministry of Education and Culture/Ministry of Manpower standardize addresses and postcodes RESPONSIBLE AGENCIES: Ministry of Communication and Informatics 7 PRIORITY OBJECTIVE Make the Digital Economy II Work for All CONTD. 3 Promote supply of DFS/digital payment solutions that cater to the unbanked and the underbanked → Streamline the → Standardize rules and licensing and procedures to enable registration processes interoperability of EXECUTIVE SUMMARY required to become a payment schemes provider of digital financial services RESPONSIBLE AGENCIES: Bank Indonesia RESPONSIBLE AGENCIES: Bank Indonesia OJK BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA → Incentivize the → Increase use cases → Strengthen the private sector to of DFS products legal and regulatory innovate and develop in the delivery of framework for managing new DFS products that government services risks related to cater to the rural (including Government- data governance population to-person payment) to and privacy, cyber sustain the commercial security and RESPONSIBLE AGENCIES: viability of DFS operational risk, and Bank Indonesia agents financial integrity, in order to build Directorate General of Taxes RESPONSIBLE AGENCIES: consumer trust in DFS CMEA (Financial Inclusion products Council) RESPONSIBLE AGENCIES: Ministry of Social Affairs Bank Indonesia OJK 4 Use tax policy instruments to ensure a level playing field → Complete the design of measures to reform → Lower the overly tax policy and modernize tax administration generous VAT threshold in relation to DE transactions, ensuring to expand the digital adoption and communication of consistent economy tax base rules, institution of efficient IT systems to administer registration, filing, payment RESPONSIBLE AGENCIES: and use of digital transaction data with Ministry of Finance other third-party data and taxpayer data for (Directorate General of Taxes, Directorate General of Customs strengthened compliance risk management and Excise, and Fiscal Policy Agency) RESPONSIBLE AGENCIES: Ministry of Finance (Directorate General of Taxes, Directorate General of Customs and Excise, and Fiscal Policy Agency) 8 Use Digital Technologies to Provide PRIORITY OBJECTIVE III Better Services and Upgrade Citizen- State Interactions 1 Develop a national digital ID framework → Pass the draft Law → Launch a whole- → Close the coverage → Introduce a national EXECUTIVE SUMMARY on Personal Data of-economy national gaps in the national digital ID system or Protection digital ID initiative population registry federated ecosystem to define the optimal database (SIAK), fit-for-purpose for RESPONSIBLE AGENCIES: model for Indonesia, streamline the new online transactions in Kominfo bringing together registration and the Indonesia context, government, private update processes, and building on the SIAK MoHA sector and civil introduce biometric- society based e-KYC processes RESPONSIBLE AGENCIES: Office of the President RESPONSIBLE AGENCIES: RESPONSIBLE AGENCIES: MoHA BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Office of the President MoHA Kominfo MoHA Kominfo 2 Reorient from a narrow 3 Implement a whole- focus on e-Government to of-government data a comprehensive national management policy digital transformation agenda → Strengthen the → Adopt a whole- → Implement the → Build alliances political and of-government One Data Policy with key institutions bureaucratic influence approach to digital comprehensively, in such as MenPAN- needed to move the transformation, priority sectors RB,Kominfo, and MoHA agenda by placing a coordinated and relevant for managing to instill One Data central government spearheaded by an the post pandemic principles into the authority like agency that has multi- recovery (e.g., implementation of the the Office of the ministry oversight. education, health, digital government President in the social protection or and digital ID driving seat RESPONSIBLE AGENCIES: MSMEs). initiatives. Office of the President RESPONSIBLE AGENCIES: RESPONSIBLE AGENCIES: RESPONSIBLE AGENCIES: Office of the President Bappenas Bappenas BPS BPS and other line agencies and other line agencies 9 Over 1.1 Access to the Medium: Who is connected, who is not and why? 1.2 HARNESSING THE MEDIUM TO BOOST INCOME: Who is winning, who is behind, and why? 1.3 USING THE MEDIUM TO DELIVER BETTER SERVICES: What holds Indonesia back? 1.4 POLICIES TO LEVERAGE DTS FOR GREATER INCLUSION: What can Indonesia do? rview 11 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA S imilar to many other countries around have switched to the internet for their enter- the world, the COVID-19 pandem- tainment and social needs, driving sharp growth ic has hit Indonesia hard. The global in the usage of digital media (music and video slowdown in economic activity, com- streaming) and communications applications. A bined with domestic measures to con- recent industry estimate shows that 37 percent tain the spread of the virus, dragged of all digital service consumers in Indonesia in 5.1 down the economy, which is estimated 2020 were new COVID-induced users and more mil. to have shrunk by around 2.2 percent than half (56 percent) were from conventionally in 2020. Latest estimates suggest that weaker markets such as non-metro areas.2 Indonesian have lost about 5.1 million people—equivalent to 2.4 per- their jobs cent of the working-age population—have lost This digital pivot is expected to be permanent their jobs, while an additional 24 million have and, as such, promises to provide a major boost 24 had to work reduced hours due to the pandemic. to what is already the largest and the fastest mil. As many as 50 percent of workers have experi- growing digital economy in Southeast Asia. enced a reduction in earnings. The impact on The behavioral change that has accompanied Had to work living standards has been devastating, with more this accelerated adoption of digital technolo- reduced hours than 2.2 million Indonesians estimated to have gies can be expected to be enduring. Estimates 50% been pushed into COVID-19-induced poverty from industry indicate that 93 percent of new in 2020. adoptees expect to stay with at least one of the experienced new digital services post-COVID-19. If true, this a reduction in earnings One unexpected silver lining from the crisis, how- would imply a major boost to Indonesia’s internet 2.2 MIL. ever, has been the turbo-charged adoption of dig- economy which, at an estimated US$44 billion in BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA ital technologies. Businesses, both large and small, gross merchandise value, is not only the largest have flocked to digital technologies to try to en- in Southeast Asia, but also the fastest growing.3 sure the continuity of their operations. Adoption of e-commerce has soared, with online purchases At the same time, the pandemic has also fo- proving a viable way of securing uninterrupted cused a spotlight on Indonesia’s inequalities. access to essential consumer goods, including One enduring challenge of Indonesia’s develop- even basic groceries. School closures have forced ment experience over the past two decades has students and teachers to adapt and explore dig- been making economic growth more inclusive. itally-enabled remote learning options, includ- While remarkable progress has been made in ing by adopting a variety of EdTech solutions. terms of poverty reduction, persistent welfare HealthTech apps enabling remote consultations disparities remain across regions, and growth and the delivery of medicine have seen unprec- in living standards experienced by those in the edented growth in adoption rates. Confined at bottom 40 percent of the income distribution home due to mobility restrictions, Indonesians has persistently lagged average growth.4 By dis- Indonesians estimated to have been pushed ↳Indonesia has made rapid progress “ into COVID-19- induced poverty 2020 in expanding internet access over the past decade, but half of the adult population is still without access→ 12 60% ↪ As school closures pushed students toward remote learning, around 60 percent of students were found to be unable to participate in online classes due to the lack of reliable internet and limited ownership of internet-enabled devices OVERVIEW proportionately hitting the poor and the vulner- Similarly, as school closures pushed students BOTTOM 40% able harder, the pandemic threatens to widen toward remote learning, around 60 percent of BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA these inequalities. Meanwhile, the susceptibility students were found to be unable to participate to infection, and the burden of disease among in online classes due to the lack of reliable inter- the poor and vulnerable in the most directly af- net and limited ownership of internet-enabled fected urban areas have been more intense than devices. A remedial measure was introduced in among the non-poor, likely due to risk factors the form of an IDR 7.2 trillion data package for such as poorer diets, lower access to quality students and teachers. But, given how costly health services, higher prevalence of smoking, mobile broadband data is for high bandwidth poorer hygiene and sanitation practices, and the applications such as video streaming, this is like- inability to afford inputs to preventive behaviors ly to only be a stop-gap solution to the real prob- (e.g., masks, hand sanitizers, etc.). At the same lem: namely, that only 10 percent of Indonesian time, sectors and forms of work that have been public schools are connected to fixed broadband most affected, especially in urban areas of the internet and around half of the population is still country, are also the ones in which the poorer not connected to any kind of internet. and less skilled segments of the population are more likely to be working. As a result, while the The pandemic has also put a spotlight on the pandemic has affected everyone, uninsured in- importance of digitized, integrated and interop- come and welfare losses in the bottom parts of erable data systems in achieving resilience. With the income distribution have been the highest.5 quarantines and the scaling-down or shutdown of many government and commercial opera- Differential access to and adoption of digital tions, countries have had to try rapidly shifting technologies could compound these inequalities. services to digital channels in order to keep them ↳ Growth in Digital solutions to achieve resilience during this going. Countries that had, for instance, invest- living standards crisis have not been available to all. For exam- ed in and built strong digital identification (ID) experienced by those in the bottom ple, compared with 88 percent of government systems before the pandemic were by and large 40 percent of the jobs and 40 percent of jobs in manufacturing, better able to do this than those countries that income distribution less than 10 percent of jobs in hotels and restau- had not. For example, Estonia, Singapore, the has persistently rants, construction, wholesale, and agriculture United Kingdom, and other digitally advanced lagged average were found amenable to being done remotely. countries were able to reduce disruption for growth Similarly, just 11 to 12 percent of casual and their public service delivery, as well as enable the own-account work could be performed from private sector to shift to trusted online transac- home (compared with 42 percent of wage work). tions. But the difference was particularly telling While 84 percent of high-skilled work could be in terms of whether countries were able to deploy done from home, 85 percent of work done by quick, accurate and effective social protection low-skilled workers required physical presence measures to mitigate shocks and safeguard liveli- in their place of work.6 hoods, including being able to reach populations 13 such as informal workers. These workers were an application has been incurred and the product especially difficult to target because they were tested, it is costless to replicate it as many times not present in either databases of people living as needed. Digital technologies also render moot in poverty or databases of people contributing place-based constraints on economic activities. to social security. Near costless transmission of information and digital services helps unlock opportunities for Thailand, for example, built a website to al- buyers and sellers anywhere in the country, lead- low informal workers to apply for emergency ing to both the deepening, as well as the stronger payments online, using its national ID system integration, of markets. Lower tracking costs, or to verify the identity of applicants, establish costs of connecting individuals with information their uniqueness and determine their eligibil- about them, facilitates the design and delivery of ity by cross-checking government databases. services better customized to personal needs (e.g., Since national IDs were already linked to bank government-to-person payments, credit provid- and mobile money accounts, there was added ed based on information on consumer behavior, assurance that cash transfers would reach the etc.). Finally, by lowering verification costs, digi- intended beneficiaries. Within just a few weeks, tal technologies can enable individuals and firms OVERVIEW more than 28 million Thai citizens had applied, to build up reputations and trust and participate with 15 million deemed to be eligible. Payments in markets assured of reasonable contestability. began for some within days of applying. In Chile Evidence, primarily from high-income countries, and Peru, social registries were used to expand suggests that the efficiency gains brought about existing social assistance programs and citizens by the adoption of these technologies can boost could use their digital ID to access a portal that aggregate productivity and growth.8 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA allowed them to check if they were included. In contrast, although Indonesia introduced signif- But there are risks as well. Without strong laws icant fiscal measures, there were critical delays on personal data and consumer protection, the in getting a couple of similar programs off the lowering of tracking and verification costs could ground on account of challenges related to the lead to price discrimination, fraud and violations updating and cross-checking of databases and of privacy. The lack of trust in government insti- the verification of identities. tutions and regulations to protect and safeguard citizen interests could retard the adoption of dig- Looking ahead, one key question policymakers ital technologies. Furthermore, the introduction in Indonesia are facing is how to ride the momen- of digital super-efficiency in an environment tum of digital adoption generated by the pan- in which there are significant structural ineffi- demic not only to power the recovery in the short ciencies in the bricks-and-mortar world could term but to also bring about greater inclusion potentially accentuate inequalities. This could and resilience in the economy in the medium play out in several ways. First, failure to univer- term. The crisis is clearly far from over, but some salize access to reliable and good-quality inter- signs of economic recovery are beginning to ap- net could result in the benefits of digitization pear on the horizon. As the economy rebounds being concentrated among a small segment of and adjusts gradually to the new post-pandemic the population, while a significant share gets left equilibrium, questions about ways in which In- behind. Skill premia may rise, and labor markets donesia could leverage digital technologies to could get polarized. Access to digitally enhanced better prepare for similar shocks in the future learning experiences and opportunities may be and address some of its long-term development limited to students of certain backgrounds, fur- challenges have become highly pertinent. ther widening intergenerational inequalities. The promise exists. Digital technologies signifi- Second, distributional tensions could arise also cantly reduce economic costs related to search, among online and offline domains on the one replication, transportation, tracking and verifi- hand, and domestic and foreign players on the cation.7 Lower costs for finding and comparing other, with gains for one coming at the expense of information can help reduce price dispersion, the other. Or between larger players and smaller improve matching, boost efficiency, and deepen ones, and between owners of capital and owners financial, labor and retail markets. The non-rival of labor. Most platform-based applications of nature of information (one person’s consumption these technologies are often propelled by strong does not lower the amount available for consump- network effects (or demand-side economies of tion by another) encoded digitally lowers the rep- scale); the larger the number of users, the greater lication cost; once the upfront cost of developing the value of the product for the users. This posi- 14 tive feedback loop generates advantages for the digital platforms in the country, this report: (i) first mover and can accord significant market investigates the key challenges that Indonesia power to the incumbent by making it difficult faces in maximizing its digital dividends in an even for more productive newcomers to enter inclusive manner; and (ii) identifies some ac- the market. As the market begins to mature, this tionable entry-points for investments and pol- concentration of some firms within an industry icies. The report is fundamentally about how can slow down innovation, hurt consumers and digital technologies touch, shape and influence workers, and lead to the redistribution of rents the economic and social lives of people, and the to the platforms themselves. diagnostic as well as policy thrust of the report is organized around three main dimensions along This report conducts an in-depth diagnostic of which the risks of ‘digital exclusion’ are the most digital technologies, and the scale and extent of pronounced: (i) access to the medium (digital their current applications in Indonesia. Fully technologies); (ii) ability to harness the medium exploiting Indonesia’s rich data landscape, new in private capacity to enhance incomes; and (iii) survey data collected specifically for this report, ability to benefit from public sector adoption of as well as anonymized data shared with the the medium to provide services more effectively OVERVIEW World Bank in confidence by some of the major and efficiently. BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA ↳ Digital technologies significantly reduce economic costs related to search, replication, transportation, tracking and verification⬎ 15 Access to the Medium →Who is connected, who is not and why? Indonesia has made rapid progress on internet connectivity, but about half of the adult population is still without access and the inequality in the access to the internet mirrors patterns of other inequalities between demographic groups, regions and income classes BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA 16 O ver the past decade, Indonesia has sustained steady growth in internet connectivity, driven primarily by rapid invest- ment in network infrastructure by the private sector. The share of the adult population connected to the internet increased almost four-fold, from 13 percent in 2011 to 51 OVERVIEW percent in 2019.9 This impressive growth notwithstand- ing, 49 percent of Indonesian adults are still not connect- ed to the internet and a significant digital divide persists across various spatial, economic and social dimensions. For example, the urban-rural divide in connectivity is large and appears to have been increasing over the years. BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA In 2019, 62 percent of Indonesian adults in urban areas were connected compared with just 36 percent in rural areas. Urban and rural internet connectivity was 20 and 6 percent, respectively, in 2011 (Figure O. 1). The Government of Indonesia (GoI) has made major efforts to close the digital divide, most notably with the implementation of the Palapa Ring project, which aimed to extend the fiber-optic backbone infrastructure of the country to the outer eastern islands. With the completion of the Palapa Ring project in 2019, all of Indonesia’s 514 kota/kabupaten (cities/districts) are now connected to the national backbone. This has led to a remarkable increase in the proportion of adults connected to the internet in all major island regions of the country. But there are still size- 1 able gaps across regions. For example, only about one-third of the adult population in Papua is connected, compared with about 55 percent in Java-Bali (Figure O. 2). At the Section ⬎ same time, the fact that almost half of the population, even in regions with relatively better infrastructure, remains without internet access points to major challenges in the middle and last mile connectivity segments. Income gaps in access are similarly huge. Adults in fam- ilies in the top decile of the income distribution are over five times more likely to be connected to the internet than adults in the poorest decile, only 14 percent of whom are connected (Figure O. 3). This sharp income gradient points to a possible affordability constraint in the access to internet. Likewise, there is also a sharp generational, education and gender divide. Younger adults are signifi- cantly more likely to be connected, as are adults who are better educated. Men are 8 percentage points more likely to be connected than women, suggesting possible inequal- ities in device ownership within households. 17 FIGURE O.1 The share of the adult population with access to the internet has been increasing over the past decade PROPORTION OF INDIVIDUALS 15+ WITH ACCESS TO THE INTERNET LEGEND Urban National Rural OVERVIEW BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA FIGURE O.2 FIGURE O.3 Some parts of the country are Richer Indonesians have better still lagging access PROPORTION OF ADULTS WITH ACCESS TO THE INTERNET, BY PROPORTION OF ADULTS WITH ACCESS TO THE ISLAND REGIONS IN 2011 AND 2019 INTERNET IN 2019, BY DECILE OF PER CAPITA CONSUMPTION Source: World Bank staff calculations based on various years of Susenas. Note: Connection to the internet is defined based on whether adults reported having access to the internet in the past three months (including accessing social media apps such as Facebook, 18 YouTube, Instagram, Twitter, and WhatsApp). Most Indonesians connect to the internet using FIGURE O.4 mobile devices, and private sector investment growth in mobile broadband infrastructure has Relative to other countries in the region powered the increase in internet connectivity over the past decade. Major players, such as Tel- Indonesia does well enough on mobile komsel, Indosat, XL Axiata, Tri and SmartFren, broadband penetration, but is lagging have accelerated the deployment of base stations severely on fixed broadband focusing increasingly on 4G/LTE stations. An estimated 95 percent of the population lives with- MOBILE BROADBAND PENETRATION in reach of the fastest, i.e., 4G/LTE networks, LEGEND though the number of actual 4G/LTE subscrib- 4G/LTE 3G GSM ers with access to those networks is significantly lower (just over 50 percent). Moreover, fixed broadband, or fiber-to-the-home (FTTH) ser- vice—particularly important for large volumes of data usage, for instance by schools, medical OVERVIEW facilities, government offices and businesses—is used by a very small segment of the population. Latest estimates from the industry put the total estimated number of fixed broadband subscrib- ers at about 9.7 million. This translates to fixed broadband penetration of just 4 percent of the BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA population, or 16 percent of households. Thus, while Indonesia performs relatively well com- pared with some of the regional peers in terms of mobile broadband penetration (particularly at slower speeds), it performs distinctly more poorly compared with some of the regional peers such as Singapore, Malaysia, Thailand and Vietnam on the fastest mobile broadband (4G/LTE) and fixed broadband rollout (Figure O. 4). → Adults in fam­ ilies in the FIXED BROADBAND PENETRATION top decile of the income distribution are over five times more likely to be connected to the internet than adults in the poorest decile LEGEND onnecting people is not just C about network coverage. Population Household Internet must be affordable as well, especially for the poorer and less affluent seg- ments of the population to be able to access and adopt it. Indonesia ranked 61st out of 100 countries Source: Telegeography analyzed by the Economist Intelligence Unit’s (2020). annual Inclusive Internet Index in 2020. Com- pared with its regional peers, Indonesia outper- forms Cambodia and the Philippines, but lags Thailand, Singapore, Malaysia and Vietnam in terms of overall internet affordability. 19 Affordability of mobile data is not a major con- FIGURE O.5 cern, particularly for relatively small volumes of data use. While mobile service providers in In- Fixed broadband subscriptions are donesia adopt differential tariff-pricing to cover prohibitive for many higher service provision costs in more sparsely populated and geographically challenging areas, REASONS FOR NOT SUBSCRIBING TO FIXED BROADBAND, BROKEN there is a wide range of mobile data packages for DOWN BY ISLAND REGIONS different budgets and needs, making internet mo- bile data relatively affordable for every segment of the population. Even for poor households, the average price per GB in a mobile data plan of US$0.64 is less than 1 percent of their monthly per capita expenditure, suggesting that mobile data costs should not prevent Indonesians from connecting to the internet. The price of mobile prepaid data in Indonesia—1GB at an average of OVERVIEW 0.95 percent of GNI per capita—is lower than the ASEAN average of 1.4 percent and the global av- erage of 5.5 percent. This is also well within the affordability criteria set out by ITU-UNESCO (1GB at 0.95 percent of GNI per capita). BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA In contrast to mobile broadband, setting up a fixed broadband service entails a wide array of costs that can include the cost of modem rental, an installation fee and the monthly subscription fee. In Indonesia, the price of a monthly broadband subscription cur- rently ranges from IDR 250,000 to IDR 800,000 (US$20 to US$55). The cost of installation and the first-month subscription of a fixed-line inter- net connection using even the cheapest internet package is estimated to be equivalent to around 1.2 times the monthly per capita expenditure of a typi- cal poor household. Indeed, Indonesia ranked 131st out of the 200 countries in the 2019 ITU rankings on fixed-line subscription fees, suggesting that cost could be a clear binding constraint for the adoption of fixed broadband. LEGEND Data from a survey conducted specifically for this report confirm this. Over 40 percent of house- Source: Digital Economy Household Survey, 2020. holds reported fixed broadband subscription costs to be prohibitive. This is distinctly higher than the proportion of households whose reasons for non-adoption were the use of mobile broadband as substitutes for fixed broadband (24 percent) and the proportion of households that reported Indonesia ranking not having access to a provider (14 percent) (Fig- out of the 200 countries in the ure O. 5). From a regional perspective, cost is the 131 2019 ITU rankings most salient barrier to adoption in places such as st on fixed-line Sumatra, Sulawesi and Maluku, while availability subscription fees of services is the bigger barrier in Papua and West and East Nusa Tenggara. In contrast, Java-Bali stands out with the highest proportion of house- holds that treat mobile broadband as a satisfactory enough substitute for the fixed broadband inter- net experience (Figure O. 5). 20 FIGURE O.6 →Both fixed and mobile broadband speeds in Indonesia are among the Both fixed and mobile broadband download speeds in Indonesia are among the lowest lowest in ASEAN, and the poorer in the ASEAN region quality in the more populous parts of the country suggests that network MOBILE BROADBAND THROUGHPUT (MBPS) congestion is a major challenge uality of service (QoS), or Q the overall quality and reli- ability of internet services, is also another challenge in different parts of the coun- OVERVIEW try. QoS impacts the speed of data transmission (up- loads, downloads), the quality of phone calls, and hence the ability of users to access online services, be it e-commerce or video streaming, for exam- ple, for remote health-care or learning services BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA FIXED BROADBAND THROUGHPUT (MBPS) that have become integral during the COVID-19 pandemic. This is also an area in which Indonesia lags critically behind some of its regional peers. Average mobile broadband download speed ex- perienced by Indonesian consumers is about 9.8 Mbps, the lowest speed in all of ASEAN. The mo- bile download experience in Indonesia is not just a mere one-quarter of the average speed available to Singaporean consumers but also a little over half of what is enjoyed by consumers in Myanmar. Incidentally, Myanmar is also the only country in the ASEAN region that Indonesia outperforms on fixed broadband download speeds. But the overall level is almost one-tenth of market lead- ers in the region such as Singapore (Figure O. 6). Source: Ookla Speedtest (March, 2020). Note: Mobile broadband speeds in the left There is considerable variation in download panel and fixed broadband speeds in the right panel. speeds within Indonesia as well, with, somewhat interestingly, places in West Papua (Sorong), Ma- luku (Ambon) and Papua (Jayapura) registering the best mobile download speeds. In contrast, more populous places particularly in the periph- ery of large metropolitan areas, such as Cimahi outside Bandung, and Tangerang and Bogor, just outside Jakarta, are the places with the poorest internet experience as measured by download speed. QoS is determined by several factors, including network congestion (linked to the availability of spectrum for data transmission), availability of cell sites, weather and geographic location. This observed pattern on the variation in the QoS across the country clearly points to network congestion being a key impediment to high quality download experience in the higher population density places in Indonesia. 21 →Limited spectrum, unavailability of specific bands, limited regulatory clarity on infrastructure sharing and lack of competition, especially in the provision of fixed broadband services, are the main drivers of limited access to good quality internet in Indonesia BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA 22 UPWARD OF 80% A ll Mobile Network Op- nities at significantly lower capital cost and erators (MNOs) need a therefore with lower ongoing operating costs. spectrum portfolio, and This would result in more affordable services the availability and overall to Indonesian consumers. quality of mobile broad- band hinges crucially on While the unavailability of low-frequency bands having adequate spectrum. Currently, Indonesia limits coverage expansion, the unavailability of has allocated a total of 467 MHz in total IMT specific high-frequency capacity bands retards spectrum to its active MNOs. In global and re- the preparation for 5G rollout. Indonesia has gional terms, this quantum of available spectrum recently made the 2,100 MHz and the 2,300 in Indonesian is low and has not increased for MHz bands available through an auction pro- some time except for some small additional spec- cess. But the continued unavailability of high- trum being made available in the 2.3 GHz band. er-frequency capacity spectrum, including par- Before the introduction of 5G services, 100-150 ticularly the 2.6 GHz and the 3.5 GHz bands, MHz in total IMT spectrum may have been has meant that the MNOs have been doing their enough for a successful MNO. However, this is best with their existing spectrum allocations. OVERVIEW unlikely to remain the case in the future. Poten- Arguably, they are over-investing in additional tially, MNOs will need more than 1,000 MHz cell sites for 4G/LTE services when they could of spectrum (including 3.5 GHz and mmWave be investing in new 5G technology. spectrum), although prices for spectrum cannot increase by that same factor as this would crowd As 5G is more efficient, it offers a much lower out investment. Three Indonesian MNOs do not cost per unit (and higher speed, better QoS ↳ Passive BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA have enough spectrum to be viable in the long and can be used for other innovations, such as infrastructure such as ducts, poles term, with Hutchison (Tri) having no low band 5G IoT), instead of continued capacity growth and rights of way spectrum at all. on current 4G/LTE systems. The sector needs account for up to clarity, perhaps in the form of a spectrum road- 80 percent of the At about 1.71 MHz per million people, Indone- map, in order to be better able to plan and direct investment in fixed sia’s total IMT spectrum allocation is on the low- investment in the network in the most optimal broadband rollout er side compared with other emerging countries. way. Without such clarity, investment becomes This lack of useable IMT spectrum available to significantly less efficient, and does not result in the MNOs has a direct negative impact on the improved broadband speeds and lower latency quality of mobile internet services in Indonesia. offerings to Indonesian consumers. In addition, This manifests itself in a number of ways, such as it is necessary to consider the future switch-off poor latency, call drops, inability to do voice over of legacy 2G and 3G networks as other Asian LTE (VoLTE), etc. However, the most import- and ASEAN markets are doing or have done in ant way is how it negatively affects broadband order to free up legacy spectrum for 4G and 5G speeds as discussed above, especially download services.11 For example, Singapore’s 2G services speeds as traffic to consumer mobile devices is are already switched off, Malaysia’s 3G networks typically asymmetrical. Higher spectrum pric- are scheduled to be switched off by the end of es also result in less capital being available for 2021, and Vietnam is scheduled to switch off MNOs to invest in the network. With height- legacy 2G services in early 2022.12 ened additional demands for bandwidth due to larger share of population working from home Major investment is required to expand fixed (WFH) and studying from home (SFH) due to broadband rollout in Indonesia. However, the COVID-19 pandemic, the need for further typically 70 to 80 percent of the investment in useable spectrum is more acute. fixed broadband is in the form of passive infra- structure, such as ducts, poles, rights of way, In addition to quality issues, this spectrum and civil works. Mobile broadband deployment scarcity also contributes directly to the digital across Indonesia has relied instrumentally on divide by constraining further rollout of mobile tower-sharing, which has been well-established broadband. The crucial 700 MHz band, which since the 2009 Regulation on Tower Sharing. has propagation characteristics that are par- Independent tower-sharing companies have ticularly well suited for remote connectivity, is brought efficiency to tower deployments. currently occupied by analog television. If the Cross-sectoral passive infrastructure sharing 700 MHz spectrum were to be made available between (rail)roads and fiber-optic and elec- to the MNOs in Indonesia, then 4G and/or 5G10 tricity poles has also occurred to some extent. services could be provided to remote commu- But sharing between the telecom operators is 23 65 not yet well established and lacks regulation. tion. This could be in the form of a single shared Duplication of passive infrastructure is costly RAN in the most remote areas, or two compet- mil. and unnecessary, and therefore making reg- ing RANs in rural areas that, through sharing, million ulatory provisions to avoid such duplication enable three or more mobile providers to offer homes would be a significant step toward expanding services without requiring a fully duplicated passed by fixed broadband access networks. RAN rollout. PLN 30 mil. Sharing of active network will also increase Unlike the mobile broadband market, the fixed million homes competition in mobile broadband in rural ar- broadband market is very concentrated, with passed by Telkom eas. This takes the sharing a step beyond the the current FTTH rollout dominated by PT current (passive) tower-sharing arrangements. Telkom (IndiHome), with some smaller pro- The more remote areas of the country often viders such as Biznet, MNC, My Republic, etc. have a single provider while, if active rural ac- in some major cities (Figure O. 7 and Figure O. cess network (RAN) sharing were to be allowed, 8). In most places there is no real competition this would create more options for competi- for fixed broadband. This lack of competition OVERVIEW FIGURE O.7 FIGURE O.8 While the mobile …the FTTH market is more broadband space is concentrated, with Telkom competitive… dominating market share BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA SUBSCRIPTION SHARES OF VARIOUS SUBSCRIPTION SHARES OF VARIOUS PROVIDERS PROVIDERS Telkom Telkom 46% 87% Source: Subscription data various sources. 24 in the fixed broadband space constrains rollout forced to keep a Telkom connection to maintain and affects the quality of services, as well as the their telephony service. Number portability is affordability of tariffs, as it limits the incentive another issue. Even if another provider were to to upgrade networks in a timely manner and be allowed to offer fixed telephony services over remain competitive in terms of pricing. the broadband connection, customers would have to change telephone number to migrate to A fixed broadband provider should be able another provider. Not being able to take their to deliver a full portfolio of services, not only numbers to another provider locks in users with broadband internet but also telephony, TV and Telkom, as changing numbers can be a disin- many other value-added services. However, the centive, especially for institutional users (such current regulatory regime limits this by requir- as schools and businesses). ing providers to bid for service-specific licenses instead of a single uniform license for all ser- Although (fixed) telephony is a service under vices. Local telephone licenses are still formally pressure, the restrictive licensing regime re- limited to Telkom, Indosat and Batam-Bintan duces the competitiveness in the broadband Telekomunikasi, with Telkom effectively be- market and limits entry for other providers, OVERVIEW ing the only provider outside a specific area of since they cannot offer a full-service proposi- Batam-Bintan. This is a major hurdle for fixed tion to customers who remain locked in with broadband competition since customers are the incumbent provider. BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA “The lack of competition in the fixed broadband space not only constrains rollout but also affects quality and affordability of internet as the incentive to upgrade networks and remain competitive in terms of pricing is limited” 25 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA 26 Section ⬎ 2 Harnessing O n average, Indonesians spend around six hours on- the Medium line, with the younger and more educated segments more digitally engaged than to Boost the relatively older and less educated demographic. In- tensity of internet engage- Income ment is highest for the 16 to 25 age group, which on average spends 9.7 hours a day online. There are no significant differences →Who is between men and women in terms of intensity HARNESSING THE MEDIUM of internet use. Among the various online activ- ities, communication, social media and leisure winning, who dominate usage (Figure O. 9). According to in- dustry estimates, Indonesia is the fifth most in- ternet engaged country in the world, behind the is behind, Philippines, Brazil, Thailand and Colombia. The average intensity of internet usage in Indonesia is 28 percent above the global average, making it an & why? attractive market for content developers and ad- vertisers.13 Another industry estimate suggests BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA that daily time spent on the internet in Indone- sia could have spiked by 31 percent during the peak of the pandemic and leveled off at around Indonesians who are connected 20 percent more than pre-pandemic levels more to the internet use it recently, suggesting that as Indonesia emerges quite intensively, with from the crisis its population could have become communication, social media even more intensively engaged on the internet.14 and leisure applications Being an internet user in Indonesia is almost dominating usage synonymous with being a social media user. Over 85 percent of internet users were also users of social media, with the most popular platforms being WhatsApp, Facebook and Instagram. WhatsApp is principally used for communi- cation and sharing information while on other platforms (Facebook, Instagram and Twitter), users are active in seeking out news and infor- mation. Social media platforms are increasingly also important for commerce in Indonesia, with a sizeable proportion of users (about 20 percent) using Facebook, Instagram and WhatsApp for buying and selling. The most common topics of discussion across all platforms relate to hobbies 5 and lifestyles. However, religion, public policy th According to industry and politics are also widely discussed, highlight- estimates, Indonesia is the fifth most internet engaged ing the importance of social media as a channel country in the world, behind of communication and influence on these topics. the Philippines, Brazil, Thailand and Colombia. Though seemingly unproductive, social media and other digital applications used for leisure also generate value and can be stepping-stones to more sophisticated and ‘productive’ uses. A key question that is often asked is whether the consumption of these largely free digital ser- vices generates any value to the users, or are the 27 FIGURE O.9 Communications, social media and leisure activities account for 80 percent of average time spent online SHARE OF TIME SPENT ONLINE 36% 21% 21% OVERVIEW Leisure 11% 7% BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Others Communications Social Media Browsing Buying & Selling 3% Source: World Bank staff calculations using data from the Digital Economy Household Survey (2020). Note: Communication includes time spent on online texting and email. Leisure consists of time spent streaming content (audio and video), as well as playing online games. Social media includes time spent on Twitter, Facebook, and Instagram. Others consist of e-travel reservation, e-gig jobs, e-finance, ordering ride-hailing for transportation and food, and creating or downloading digital content. hours spent browsing pictures of friends on In- ↳ Though seemingly stagram and watching cat videos on YouTube simply time away from other more productive unproductive, social endeavors. If there is value generated, it is not media and other clear how this value can be measured. There- digital applications fore, these benefits are often also not captured in statistics on national accounts. Yet, the fact used for leisure also that so many consumers voluntarily choose to generate value and can spend so much time consuming these services be stepping-stones to must imply that there is some consumer surplus that they derive from it. There have been some more sophisticated and efforts, primarily in advanced economies, to try ‘productive’ uses to estimate the value of this consumer surplus and, depending on the methodology used, the numbers range from 3 to 25 percent of income in one estimate, and close to US$100 billion over the 2007–11 period in another.15 Using one of the approaches from the literature, we estimate that consumer surplus of free internet services in Indonesia averages around 19 to 21 percent of per capita income for users.16 28 21% ↳ of the urban population ordered food using the digital ride- hailing app HARNESSING THE MEDIUM BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA → Digital ride hailing is commonly used and not only provides mobility services to connect the otherwise fragmented labor markets in Indonesia’s large and sprawling metropolitan areas, but also offers several other conveniences D igital ride-hailing services 36.5 percent of urban dwellers reported using provided by companies the ride-hailing service provided by these com- 18% such as Gojek—Indonesia’s panies before the pandemic.17 A striking 18.4 homegrown decacorn that percent of users of these ride-hailing services of users of these ride-hailing epitomizes the country’s used them for their daily commutes. Other services used them digital potential for policy reasons for use included circumstances when for their daily makers—and its regional competitor, Grab, private vehicles could not be used (52 percent), commutes are potentially among the most frequently a travel option for odd hours (41 percent), and experienced digital transactions for many In- to travel to areas where public transportation donesians. The green jackets worn by the mo- services are not available (26.6 percent). With- tor-cycle taxi drivers of both companies are out the availability of these digital ride-hailing a distinctive feature of almost all Indonesian options, journey times would be longer for 55 cities. And this is not without reason: these percent of users but travel costs higher for a platforms not only provide efficient mobility negligible 1.9 percent of users. In addition to services to connect the otherwise fragmented ride-hailing services, these companies have also labor markets in Indonesia’s large and sprawling been providing a whole host of other services metropolitan areas, but also offer several other through their apps. Chief among these is food conveniences such as food delivery and logistics. delivery through their complementary Go- While the pandemic and the associated decline Food and Grab-Food services. Survey results in mobility has temporarily hit this segment show that around 21 percent of the urban pop- hard, one-quarter of all Indonesians and around ulation ordered food using these apps. 29 → Buying and selling online is growing, and while still only prevalent among a small share of the population, it is already enhancing consumer welfare by providing cheaper options, greater product variety and convenience OVERVIEW E -commerce is one of the ability of products drives 40 percent of online BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA largest components of In- purchases. In Bali, the reasons for buying online donesia’s digital economy are roughly split equally among price, conve- and has also been one of the nience and local non-availability of the goods major drivers of its growth that are purchased. In DKI Jakarta, however, during the pandemic. In- online purchases are driven almost exclusively dustry estimates suggest that the gross merchan- by price- and convenience-related factors. In dise value (GMV) of e-commerce in Indonesia is addition to lower prices, being able to consume likely to have increased by 54 percent yoy, from goods that were previously unavailable, having US$21 billion in 2019 to US$32 billion in 2020. a greater variety of the same products to choose This more than offsets the decline in travel over from and being able to conveniently procure are the same period, which contracted from US$10 welfare enhancing and Indonesian consumers billion to US$3 billion.18 The industry is buoyant who have adopted e-commerce are realizing about the stickiness of pandemic-driven adop- some of these benefits.19 tion from both the buyer and seller sides, and the possibility of this growth potentially setting off In addition to making cheaper goods available e-commerce on a different trajectory. In 2019, to those who buy online, greater penetration of the proportion of internet-using households e-commerce into a geography can also have an that reported buying and selling online was impact on the prices of similar goods sold offline, 12.8 and 5.1 percent, respectively. As a share of helping to lower overall inflation for the benefit the overall population, e-commerce-engaged of the larger community. This can happen be- households (those who buy or sell) only account- cause, with cheaper options online, including ed for around 7 percent. A rough back-of-the- from sellers in other parts of the country, buyers, envelope calculation using industry estimates on especially in remote places, can essentially arbi- the adoption of digital services during the pan- trage away price differentials across geographies. demic and their anticipated stickiness suggests Indeed, we find that price inflation of commodi- that this could grow by up to 10 to 11 percent of ties likely to be traded relatively more intensively the population as Indonesia begins entering the online was 0.8 of a percentage point lower in high economic recovery phase. e-commerce penetration provinces relative to provinces with lower e-commerce penetration. Indonesians who use e-commerce appear to be A similar impact does not exist for commodities benefiting from it. Price and convenience are the less likely to be traded online. In other words, most frequently cited reasons for online pur- greater penetration of e-commerce in a geogra- chases (Figure O. 10). Looking across regions, phy may be contributing to a broader slowdown other factors also come up in specific geogra- in inflation of a class of commodities more likely phies. For example, in Papua, local non-avail- to be traded online with greater intensity. 30 FIGURE O.10 Price and convenience are the most dominant reasons for buying online… …FOR THE COUNTRY AS A WHOLE HARNESSING THE MEDIUM …AND, WITH SOME HETEROGENEITY, FOR THE MAIN ISLAND REGIONS AS WELL BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA LEGEND Source: World Bank staff calculations based on data 40% from the Digital Economy Household Survey (2020). of e-commerce transactions in Papua entail purchase of goods not available in lower markets 31 →While most workers have benefited from digital technologies, the more educated BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA ones have benefited more than others 32 G HARNESSING THE MEDIUM rowth of the digital economy has clearly generated some Yet, evidence suggests that specific applications of digital BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA benefits for the Indonesian consumers who are connected technologies are benefiting some population sub-groups. and able to avail themselves of these services. A key ques- For example, e-commerce provides a viable pathway for tion of interest is whether the applications of these technol- Indonesian women re-entering the labor market after ogies can also bolster income-earning opportunities. This leaving jobs due to pregnancy/maternity or domestic can be both at the extensive margin by creating more jobs work. Similarly, digital gig jobs, including those in the of a certain type, and at the intensive margin by enhancing ride-hailing industry, are providing economic oppor- the returns to labor in the form of higher wages and salaries. tunities that appear slightly better than other informal Greater availability and access to fast internet has been options, especially to young, relatively better educated, found in the literature to have a positive impact on broad- male adults in urban areas. er employment outcomes, including increased (female) labor force participation and employment rates, net firm In an aggregate sense, however, we find that higher skilled entries and improved productivity.20 There is also evidence workers have perhaps benefited more than the lower from Nigeria to suggest that increased mobile broadband skilled ones from the expansion in internet access that coverage has led to an increase in labor force participation Indonesia has experienced in recent years. Specifically, and employment, particularly among women, and this has over the 15-year period between 2005 and 2019, educated enabled households to realize higher income and attain Indonesians have consistently enjoyed higher earnings higher living standards. Greater mobile broadband pen- relative to their less educated peers. For example, aver- etration is associated with lower aggregate poverty rates. age earnings for the most educated group (those with a college or university degree) have been almost 80 percent The evidence from Indonesia, however, is somewhat mixed. higher than those for the uneducated group (less than six In a sample of districts that were not connected to the fiber years of schooling) (Table O. 1).21 Looking at the differen- optic backbone until 2010 - 33 percent of all districts and tial impact of internet penetration on these skill premia, ones disproportionately outside of Java-Bali and Sumatra - we see that while greater internet penetration benefits being connected to the national backbone infrastructure in- all types of workers except the very low educated, the creased the number of internet users but did not fundamen- benefits are higher for the more skilled. The magnitude tally alter the structure of the economy. Patterns of overall of the coefficients implies that, if internet penetration labor force participation, female labor force participation, (measured by the proportion of individuals having access total employment and youth employment were similar be- to internet at home) increases by 10 percentage points fore and after the arrival of the fiber optic links. It is true that in a particular district, the earnings premium enjoyed a dominant share of Indonesia’s industrial activity is concen- by a college educated worker will, on average, go up by trated in Java-Bali. Close to 90 percent of jobs in manufac- 6 percentage points.22 Thus by benefiting higher-skilled turing, and 84 percent of all jobs in high-value or modern workers more relative to lower-skilled ones, the pattern of services, for example, are in Java-Bali and Sumatra, while diffusion of digital technologies in Indonesia is potentially the more recently connected parts of the country remain contributing to an increase in overall inequality. more dependent on agriculture and natural resources, etc. 33 TABLE O.1 Greater internet access is benefiting higher-skilled workers more than the lower skilled ones RETURN TO EDUCATION W.R.T. ADDITIONAL AVERAGE RETURN OF 1 <6 YEARS OF SCHOOLING PERCENTAGE POINT INCREASE IN INTERNET PENETRATION PRIMARY SCHOOL 18.50% 0.00% LOWER SECONDARY SCHOOL 32.80% 0.10% HIGHER SECONDARY SCHOOL 49.70% 0.30% COLLEGE/UNIVERSITY DEGREE OR HIGHER 79.40% 0.60% NON-PRODUCTION WORKERS (RELATIVE TO PRODUCTION WORKERS) IN MANUFACTURING 12.70% 0.20% Source: Jacoby et al. forthcoming. Note: Reported marginal returns to internet are coefficients on the interactions between education levels of individual workers and average internet connectivity within a district in fixed-effect panel data models that also control for all relevant individual characteristics, level of urbanization and economic development of the districts, as well as year effects to capture secular trends in returns to skills. A separate specification is used for the non-production worker result. The analysis uses data from Sakernas BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA (Indonesia Labor Force Survey) (1990–2019), Susenas (Indonesia Socioeconomic Survey) (1990–2019) and the Medium and Large Manufacturing Survey (1995–2015). The unit of the analysis is individual worker (production/non-production worker by sector-district in the case of the Manufacturing Survey → One of the reasons for this is that despite the pandemic induced increase in the adoption of digital technologies, overall digital adoption by firms and workers is still quite low T he need to ensure business five firms had closed and continued to remain continuity in a period of temporarily closed. Of the 46 percent of firms limited mobility during the that reported having to make some adjustments COVID-19 crisis has forced in their business processes to either remain many Indonesian firms to open or to reopen after a hiatus, 42 percent re- initiate and intensify their ported adoption of the internet, social media, digital transformation. A survey conducted specialized apps or digital platforms as having by the World Bank in June 2020 showed that been the main coping strategy (Figure O. 11). only 36 percent of firms had managed to re- The pandemic induced uptake of digital tech- main continuously open since March 2020. The nologies was found to be higher among larger same survey also showed that only about 40 firms (90 percent), but also not entirely neg- percent of firms had temporarily closed but had ligible for SMEs (58 percent) and micro firms reopened by June 2020, while roughly one in (32 percent) (Figure O. 12).23 34 FIGURE O.11 In addition to general business administration, marketing and sales have been the functions that SHARE OF FIRMS ADJUSTING THEIR BUSINESS PROCESS DUE TO COVID-19, have seen the strongest pivot to digital. Larger BY TYPE OF ADJUSTMENTS and more formal firms with a more conducive en- abling environment (e.g., access to internet, dig- ital knowhow) would have certainly been better positioned to take advantage of these opportuni- ties. Businesses that were already online before the pandemic also intensified their online activities. For example, digital merchants with an online and offline presence were found to increase their online activities during the pandemic.24 Indus- HARNESSING THE MEDIUM try estimates suggest that e-commerce volumes soared by as much as a factor of 2.1 during the pandemic and are expected to settle at around 1.7 times the levels relative to pre-pandemic levels.25 This soaring of demand can naturally be expected to draw more suppliers into the digital ecosystem. However, any increase in digital adoption that the pandemic brings will be from a very low base, es- Source: World Bank Survey. pecially for micro and small enterprises. Though BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA FIGURE O.12 somewhat dated now, the economic census from 2016 reveals that only 5 percent of all Indonesian non-agricultural enterprises used the internet, SHARE OF FIRMS THAT REPORTED ADJUSTING IN RESPONSE TO COVID-19 with significant heterogeneity by firm size. Larger establishments were significantly more likely to have adopted the internet (67 percent) to buy and sell, as well as to conduct other activities over the internet, while the adoption rate among micro-enterprises was significantly lower (4 per- cent) (Figure O. 13). More recent data suggest that internet adoption, especially by household enterprises, could have increased to about 11.1 percent in 2018 and 12.8 percent in 2020.26 How- ever, even among household enterprises, the rich- er ones exhibit a much higher level of adoption relative to the poorest ones (Figure O. 14). The low level of internet adoption by enterpris- LEGEND es—especially micro and small enterprises—also translates into low adoption among workers. This is especially true given the fact that these micro and small enterprises jointly account for almost Source: World Bank Survey. 75.3 percent of Indonesia’s non-agricultural em- ployment. In 2019, only 27.2 percent of all work- ers reported using the internet at their work. This Pandemic induced uptake of digital technologies was the highest among tertiary educated workers MICRO FIRMS (78.1 percent), followed by much lower use among LARGE FIRMS SMALL FIRMS MICRO FIRMS workers with just lower secondary education (17.7 90% 58% 32% percent).27 Thus, on the one hand, the overall level of internet adoption among workers and firms is low. On the other hand, there are sharp hetero- geneities in adoption between different types of workers and firms, with the higher skilled work- ers, and larger and more formal establishments having greater adoption rates. 35 FIGURE O.13 FIGURE O.14 Internet use by non-agricultural Internet adoption is far higher enterprises is still very low, with among richer household enterprises much higher incidence among medium and large enterprises LEFT AXIS SHARE OF NON-AGRICULTURAL ENTERPRISES SHARE OF HOUSEHOLDS WITH ENTERPRISES WHO USE THAT USE THE INTERNET, BY SIZE THE INTERNET FOR THEIR BUSINESS OPERATIONS, BY INCOME STATUS RIGHT AXIS THE SHARE OF ENTERPRISES OUT OF TOTAL NUMBER OF ENTERPRISES (DOT) OVERVIEW BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Source: Economic Census 2016, WB staff calculations. Source: DEHS 2020, WB staff calculations. → Another reason is that while low-skill biased applications of digital technologies such as e-commerce and digital gig-work are enabling certain segments of the workforce to boost income, their reach is limited. Digital gig work is largely concentrated among urban men and e-commerce growth is severely constrained by issues of trust, logistics and internet connectivity U nlike the narrative surrounding digital omy. Among the low-skilled workers who had a job before, economy jobs in more industrialized “unsatisfactory income” was the most common reason for countries, we see that digital gig jobs switching to digital gigs. However, these opportunities are in Indonesia are, on average, relatively largely limited to select demographic: men (85 percent of better paying than many informal jobs, all digital gig workers), living in urban areas (87 percent) though certainly not as well-paying as and working in the transportation, storage and communi- wage jobs. For example, internet-using gig workers make 6.2 cations sector (69 percent). In sum, digital gig workers in percent more per hour in earnings than otherwise identical Indonesia work significantly harder than all other workers informal workers in the same sector of employment. But but earn slightly better on average in comparison to informal they also work the longest hours among all other types of workers. Digital gigs are also a stepping-stone into the labor workers. Indonesians work 39 hours a week on average and market and a perceived ticket to higher incomes for many, informal workers typically work 38 hours a week. In con- but these opportunities are concentrated among a specific trast, digital gig workers average 49 hours a week. About 33 sub-group of the population. percent of digital gig jobs were being performed by workers for whom this was their first ever job, suggesting that these E-commerce is another promising source of employment were new opportunities being created by the digital econ- and income. In 2019, around 13.2 million out of a total of 36 38% around 127 million employed workers (10.4 percent) in Indonesia were engaged in e-commerce activities as either their primary or secondary job.28 In terms of its contribu- tion to overall employment, this number is higher than in China, where e-commerce accounts for 5 percent of total employment.29 A peculiar aspect of e-commerce in Indo- nesia is the dominance of social media and chat apps in the e-commerce landscape. Among those engaged in e-com- merce, about 71 percent reported selling exclusively via so- cial media and chat messaging apps, 3 percent sell via a plat- form only, while 26 percent use a blend of the two.30 This means that e-commerce in Indonesia remains dominated HARNESSING THE MEDIUM by consumer-to-consumer (or C2C in industry parlance) business, which makes it more similar to TaoBao Market- place, as opposed to the business-to-business (B2B) eco- system similar to Alibaba, or even a business-to-consumer (B2C) one, such as TMall. These C2C digital merchants are generally regarded as more casual market participants than the B2B or B2C ones. They also operate on a smaller scale and potentially also at a lower level of productivity. ↳ Among women who use the internet Although Indonesian men are almost twice as likely to be us- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA at work in their primary jobs, 37.8 ing the internet at work, internet-using women are slightly percent of them more likely than men to be engaged in e-commerce activ- are active online ities. Among women who use the internet at work in their sellers (as primary jobs, 37.8 percent of them are active online sellers opposed to 34.8 (as opposed to 34.8 percent of men). E-commerce is more percent of men) common as a secondary job, suggesting that it plays a useful role in supplementing family income, especially by women and youth. Among all women, e-commerce engagement (selling) is highest especially among those who are primarily engaged in housework. Moreover, e-commerce also appears ↳ 85 percent of to be providing a pathway for women re-entering the labor all digital gig market by providing opportunities to those who may not workers are men be in their previous jobs for a variety of reasons. About 58.1 and and 87 percent percent of internet-using women who had to leave their of them live in 85% previous jobs because of pregnancy/maternity or to return urban areas to domestic work were engaged in e-commerce, suggesting that e-commerce provides one avenue for women to stay productively engaged, given that they exit other forms of work to take on greater responsibilities at home. The female labor force participation rate in Indonesia has been persistently low, hovering at around 50 percent for the past three decades. This has been identified as one of the key challenges in meeting Indonesia’s development aspirations of becoming a high-income country.31 By expanding oppor- tunities for women in the labor market, e-commerce could help Indonesia to at least partially address this challenge. However, the opportunity to engage in and benefit from e-commerce is not currently available to all Indonesians. While e-commerce has increased and spread to all parts, its intensity remains concentrated in more populous and affluent parts of the country (Figure O.16). Analyzing the spatial and temporal evolution of e-commerce over the past five years, we find that, in addition to income and popula- 37 tion, internet connectivity and the cost of logistics FIGURE O.15 have also played an important role in explaining e-commerce growth. E-commerce penetration E-commerce is more prevalent in populous (the proportion of buyers and sellers in the popu- and affluent provinces lation) has increased more rapidly in provinces in which internet access expanded and the cost of lo- gistics declined over this period. E-commerce in- PROPORTION OF HOUSEHOLDS WITH ACCESS TO INTERNET ENGAGED IN tensity (value and volume of transaction per buy- E-COMMERCE, BY PROVINCE er and seller), on the other hand, is constrained more significantly by the cost of logistics. In other words, while expanding internet access has drawn more people into the ecosystem, the cost and ease of logistics constrain how intensively they are able to buy and sell online. Logistics as a bottleneck for e-commerce is con- OVERVIEW sistent with the broader challenges that Indo- nesia faces in moving goods and commodities across its far-flung geography. The availability and reliability of transport infrastructure is the first, and perhaps the major, impediment to a smooth logistics chain. There are considerable BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA imbalances of land, sea and air connectivity, for example, between well-developed cities around Jakarta, Surabaya, Medan and Denpasar (Bali), and to secondary, tertiary and rural areas, result- ing in high variability in logistics costs. A 2017 World Bank survey of manufacturers on logistics performance in Indonesia showed that compa- nies located in the Jabodetabek area (Greater Jakarta) managed to receive their goods in full and in good condition, and to maintain logistics costs at around 12 percent of the cost of sales, whereas companies in Kalimantan incurred up to 30 percent of cost of sales in logistics costs. In addition to direct logistics costs, delays in receiv- ing or sending products are another source of indirect logistics cost for manufacturers. Indonesia’s entire logistics performance is ham- pered by bottlenecks from the first to the last mile. In addition to weaknesses in transport in- frastructure, the availability and competitiveness of logistics service providers (LSPs) is a chal- lenge, especially outside of the main economic areas.32 On maritime connectivity, the lack of regular inter-island maritime routes affects the timeliness of goods transit, with domestic ship- ping lines channeling regular calls mostly on the most lucrative shipping lanes (between the main ports) and the development of roll-on roll-off shipping (ro-ro), which has proven cost effective in several maritime countries, being constrained. On air connectivity, airlines usually give priority to passengers rather than goods, which reduces the likelihood of goods being delivered on time Source: World Bank staff calculations using data from SUSENAS 2019 Note: Households are considered engaged in e-commerce if they report buying to the next destination. On land connectivity, or selling online 38 FIGURE O.16 Adoption of digital financial services is extremely low 0.5% 9% Advanced 43% DFS Users users of DFS 52% Personally own account HARNESSING THE MEDIUM With access to account All HHs 48% 9% 34% Access Only have Financially others' bank excluded account account 50% BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Source: World Bank staff calculations based on data from the Digital Economy Household Survey (2020). ↳ Even among those who buy online, 50 percent still prefer to pay cash on delivery. the lack of temperature-controlled trucks and quality purchased item will be delivered in good condition and warehousing services across Indonesia prevents expansion whether a return will be accepted in case the customer is to secondary and tertiary cities, especially for perishable not satisfied. Trust could also relate to having to make on- goods. More specifically on the last mile, a key challenge is line payments, and often in advance of the package being also the lack of standard addresses and precise postcodes in received. The general lack of trust in online transactions is Indonesia. This prevents the implementation of productiv- among the most frequently cited reasons for not engaging ity enhancement tools, such as the automation of sorting in digital transactions in Indonesia. Even among those who facilities to speed up routing of goods, and the utilization buy online, 50 percent still prefer to pay cash on delivery. of route optimization software to map the most efficient Lack of access to bank account is a binding constraint to routes for the courier. A complex regulatory framework on participation in the digital economy for those in the bot- logistics services and an uneven distribution of population tom 20 percent of the distribution. With only 52 percent across the archipelago drive these logistics challenges. of all households having access to a bank account, financial inclusion is a recognized challenge in Indonesia (Figure Another key bottleneck in the expansion of e-commerce O. 16). While 9 percent of households are users of DFS and the digital economy more broadly in Indonesia is the (which we define here as having access to online banking lack of trust in digital transactions and the consequent low and mobile money services), a much smaller proportion adoption of digital financial services (DFS). Trust can have of households (0.5 percent) are users of more advanced multiple dimensions and can, specifically in the context DFS products, which would include credit, remittances, of e-commerce, also include issues related to whether a insurance, etc.33 39 → Striking the right balance between promoting innovation and growth without according undue advantages to first movers in the new economy is critical from the perspective of equity OVERVIEW ↳ Concerns regarding unauthorized data disclosure by any of the several entities involved could n addition to universal- I discourage the more sophisticated would be-users from adopting DFS izing the medium by ex- panding access to reliable 52% and high-quality internet BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA For more sophisticated would-be users, issues of at affordable prices to all trust could also be rooted in the understanding of Indonesians, maximizing the risks related to data governance and privacy, economic inclusion benefits cyber security and operational risk, financial in- of digital technologies will also require enabling tegrity, and several others.34 Concerns regarding the population—especially those at the lower unauthorized data disclosure by any of the several parts of the income distribution or living in less entities involved could discourage this segment populous and prosperous parts of the country— from adopting DFS. And these are not entirely to harness the medium to boost their incomes. misplaced concerns; complaints coming from fi- Some of these efforts could be purely in the dig- nancial consumers, as well as the banking sector ital domain. For example, officially recognized itself, on these breaches have been increasing late- digital IDs backed by strong consumer and pri- ly. While Indonesia’s Financial Services Authority vacy protection mechanisms to foster greater (Otoritas Jasa Keuangan, OJK) has been trying to trust in digital transactions, greater adoption develop a stronger legal and regulatory frame- of digital payments and other digital financial work for this, consumers still regard the disclo- services. sure of personal ID to financial service providers as a major risk. Provision of a legal basis, such as But if the objective is to not just promote the digi- passing the current draft of the Law on Personal tal economy for its own sake but to also maximize Data Protection in the legislature is necessary to its impact on the population, then these efforts promote DFS adoption in Indonesia. must be nested in a broader reform agenda that ↳ only 52 percent of households have includes the analog, or the conventional bricks- There is a significant knowledge and awareness access to a bank and-mortar parts of the economy as well. Logis- deficit which, somewhat surprisingly, is almost account tics is one example. Though there are several digi- as prevalent among the financially excluded tal innovations disrupting the logistics landscape, as it is among traditional bank account users. the bigger challenges of logistics in Indonesia are Roughly half of banked, as well as unbanked, the ones related to the fundamental difficulty of households report not fully understanding the moving physical goods and commodities across benefits of DFS, or not knowing how to use the vast sprawling archipelago. These relate to them, or more surprisingly, especially for those the availability and quality of transportation in- already with bank accounts, never even having frastructure (air, land and sea), as well as the or- heard of these services. This highlights the key ganization of these markets, which has a bearing role for agent networks in helping to reach the on the cost structure. But a whole host of other unbanked, as the face-to-face personal interac- factors that are binding constraints on the overall tions are ideally suited to overcome the trust and productivity of the economy are also relevant. It knowledge gaps. has been well documented elsewhere, for example, 40 that restrictive trade policies limit access to key right balance between not according undue ad- inputs and markets, restrictions on investments vantages to the first movers in the new economy, depress commercial performance, a weak compe- while at the same time also not stymying inno- tition framework shields incumbents from poten- vation, is an important consideration from the tially more productive market entrants, while the perspective of equity. Policy makers in Indonesia unpredictable regulatory environment further are cognizant of this issue and have recently put weakens the business environment, inhibiting in regulation on e-commerce requiring digital competition and depressing investments. The merchants active on platforms to be eligible for introduction of super-efficiencies associated with value-added tax. However, Indonesia’s overly digital technologies in an economic environment generous VAT threshold also needs to be re- in which domestic firms, both large and small, viewed, not only to expand the digital tax base, are heavily shackled in this manner could amplify but also to make this an effective instrument for HARNESSING THE MEDIUM existing distortions and accentuate inequalities. leveling the playing field.39 One concrete example is cross-border trade on Finally, the third distributional tension is be- e-commerce which, in the current business en- tween workers and firms. Business models that vironment, could make importing significantly rely on platforms characterized by network ef- easier than exporting. Consumers would benefit fects essentially generate value by matching cus- from cheaper consumer goods from abroad en- tomers with complementary needs. The value of tering the domestic market. But if the econom- the platform to the marginal customer depends ic landscape forces domestic firms—especially on the number of vendors and service providers the smaller ones—to compete effectively with a already on board. Similarly, the value of the plat- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA hand tied behind their backs then this could end form to a marginal vendor also depends on the up eroding the country’s manufacturing base number of customers already on board. Thus, to and compounding inequalities. Indeed, with grow and become successful, platforms need to inbound e-commerce transactions increasing build a critical mass on both sides of the market. more than eight-fold from 6.1 million to 49 mil- But once a platform becomes dominant, it could lion in just two years between 2017 and 2019, this become too large for potential competitors to is an area that the Government is acutely con- dislodge, giving rise to a winner-takes-all market cerned about. And there has been some effort to structure. In developed countries, this network address the issue as well. For example, Minister effect has given rise to giant technology compa- of Finance Decree No. 199/2019, effective as of nies, such as Google, Facebook, Apple, Amazon, January 2020, lowered the taxable threshold on Uber and Airbnb, where just a few firms capture inbound shipments from US$75 to US$3 per a dominant part of the market. consignment, significantly increasing taxes on commodities such as foreign-produced textiles, The large amount of data these digital business- apparel, bags, and shoes.35 Lowering the de mini- es collect, and the increasing use of automated mis threshold to deal with low-value shipments machine-learning and artificial-intelligence (AI) puts Indonesia in the company of other early driven analytics to power continuous improve- movers in this reform space.36 However, anoth- ments in the products and services that they er complementary approach could be doubling provide, on pricing, personalization, and the down on addressing some the critical issues that targeting of ads, could further entrench their constrain the productivity of Indonesian firms.37 position. Data collection and sharing across platforms belonging to the same company could Another relevant distributional tension is the give rise to an exponentially higher benefit to one between online and offline players. If not all the company compared with those companies online purchases are catering to new demand, operating only under one platform. And once digital transactions will be displacing analog this is established, there is a risk that they could ones. More goods bought online necessarily exploit monopsonist power vis-à-vis vendors implies fewer goods are bought from bricks- to extract a larger share of value added of the and-mortar stores.38 A comprehensive welfare product. This increase in market concentration calculus needs to consider not only the benefits may lead to greater inequality, as labor receives a for consumers and producers who can partake in smaller share of value added in these segments.40 these digital transactions but also the potential Some have found these consequences of the dig- losses incurred by the bricks-and-mortar sell- ital economy to be the most important reason ers who are slower to adapt. Ensuring that the behind the recent increase in inequality in the policy and regulatory environment strikes the United States.41 41 The level of digitization in Indonesia has per- platforms is far from a realistic portrayal of the haps not reached levels where these effects would inclusive digital future Indonesia wants, so is a begin to manifest themselves and more work nation filled with data scientists, cloud-solution will be required specifically in the Indonesian architects and AI experts. Thus, in addition to context to see how these dynamics play out as low digital literacy and the severe shortage of the digital economy grows. Nonetheless, it is a digital talent at the high end, investments in potential risk with consequential distributional skills for a digital future should also include implications and a front to already start acting a broader-based effort to develop a pipeline on. Boosting digital entrepreneurship, including of workers who are generally better at prob- by addressing the glaring gaps in high-end digi- lem-solving, communication, teamwork and tal talent (developers, AI and machine-learning adaptability (i.e., high-order cognitive and so- specialists), leveling the uneven playing field for cial skills). innovation and preempting policies that protect players, while not being too restrictive on the growth of what is still a nascent digital economy, ↳ digital innovations can generate can be some elements of strategies to address opportunities, but the extent to which OVERVIEW this issue. Similarly, Indonesia does not current- workers can grasp these opportunities ly regulate working conditions and contracts of digital platforms and their workers and the to move up the economic ladder will platforms themselves set working conditions ultimately depend on worker skills; through their terms of service agreements. How- digital skills specifically as well ever, as the digital economy grows, and more and as a broader set of skills necessary BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA more workers enter these work arrangements, Indonesia may need to regulate this form of work to survive and thrive in the digital to provide workers the protection they need. economy Finally, digital innovations can generate oppor- tunities, but the extent to which workers can grasp these opportunities to move up the eco- nomic ladder will ultimately depend on worker skills; digital skills specifically, as well as a broad- er set of skills necessary to survive and thrive in the digital economy. Indonesian firms routinely identify the shortage of specialized professionals and managers in the local labor market as one of the most important bottlenecks in fostering innovation within firms. The share of firms re- porting inadequate skills as the top constraint in hiring professionals and managers is the highest in the ASEAN region. A shortage of qualified talent to develop new products and services has prompted well-funded digital platforms to look elsewhere for R&D capabilities. For example, Gojek has established an off-shore R&D division by acquiring three Indian tech companies, and its regional rival, Grab, has also set up an R&D center in India for similar reasons.42 This skill gap not only stunts innovation but also represents a major lost opportunity; these are good jobs going unfilled in a labor market in which millions are looking for a pathway to the middle class. It also explains the digital technology-induced widening of the skill premia discussed above (Table O.1). But just as a nation filled with smartphone-wielding micro-entre- preneurs all selling their wares through digital 42 “In addition to low digital literacy and the severe shortage of digital talent at the high end, investments in skills HARNESSING THE MEDIUM for a digital future should also include a broader-based BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA effort to develop a pipeline of workers who are generally better at problem-solving, communication, teamwork and adaptability.” 43 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA OVERVIEW 44 Using the Medium to Deliver Better Services →What holds Indonesia back? 3 he ability of governments Section ⬎ T around the world to har- ness digital technologies to deliver services is per- haps one area on which the COVID-19 pandemic has focused the strongest spotlight. Just as people and businesses have had to adapt to various measures taken to contain the spread of the virus, so too have governments. And the ability to use digital tech- USING THE MEDIUM nologies to curb and manage the pandemic, as well as to ensure the continuity of essential ser- vices, has emerged as a key marker of resilience. Digital technologies have also been useful in the rapid deployment of social assistance respons- es. In countries with extensive mobile phone or internet penetration, as well as strong existing social protection systems, digital technologies related to digital ID and digital financial services BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA have helped to facilitate the identification and ↪ Over 530,000 schools had to be registration of benefit recipients. Since it is the levels >530,000 closed, affecting 68 million students from pre-primary through tertiary poorer segments of the population that have been most affected by the pandemic—by job losses and income reductions, as well as by the disruption of education—countries that have been able to better harness digital technologies in this manner have also been successful in laying a more robust foundation for inclusive recovery. For Indonesia, the pandemic has laid bare the lack of readiness to fully capitalize on these dig- ital opportunities. Over 530,000 schools had to be closed, affecting 68 million students from pre-primary through tertiary levels. While all students still reported engaging in some form of learning activities at home, 64 percent of them faced critical constraints related to the lack of reliable connections and supporting devices. In the early stages of the pandemic, Indone- sia also struggled to obtain consistent data of COVID-19 cases due to challenges in integrating information systems across the various tiers of government, slowing down the response.43 The lack of digital and other alternatives caused widespread disruptions in non-COVID health care across the country.44 Some emergency so- cial assistance that was approved by the Govern- ment, such as Kartu Prakerja (pre-employment card, repurposed as temporary cash assistance to unemployed workers), faced critical delays in deployment due to difficulties in verifying the identities of recipients. More recently, efforts to accelerate vaccine rollout have also been ham- strung by lack of reliable data. 45 → Indonesia’s nascent EdTech and HealthTech scene has received a major boost and though these apps have clearly filled a void during the pandemic, their overall reach is limited to the more affluent clientele in urban centers, mostly within Java. For digital technologies to make a dent on the root causes of long-term inequalities, they need to be adopted and applied at scale by the Government. iscussions about the role services remotely (telemedicine) through apps D of digital technologies in and websites, as well as cloud-based solutions sectors such as education for hospital information management systems. and health often begin and While telemedicine generally covers various end with overwhelming health services, including consultations with optimism about the pow- doctors, the provision of diagnosis, treatment er of these technologies to and preventive care, telemedicine in Indonesia inspire a disjunctive break with the past. In the is relatively new and still limited primarily to context of developing countries that typically teleconsultation. For example, companies such face complex and deep-rooted challenges in as Halodoc and Alodokter—considered the delivering quality services, the “disruptive” most well-funded firms in the market—con- promise of these technologies can be especially nect patients to medical doctors to do online appealing. Indonesia is no different and, over consultations.47 Hospital information manage- the years, as the digital economy has grown, so ment systems have traditionally been offered by BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA has the number of new entrants in the EdTech software developers and vendors catering for and HealthTech scene. A World Bank study tailored solutions for hospitals. However, more estimated that close to 160 EdTech startups in recently, new firms such as Medico and Perik- Indonesia mostly established in the 2013–18 sa.id have started to offer cloud-based services period. Of these, 60 companies were still op- (software-as-a-service). erational and offering a variety of products and services before the pandemic struck.45 Similarly, The interest in these applications has soared the Indonesian HealthTech Association reports after the COVID-19 outbreak. In the second that around 250 registered firms are operating quarter of 2020, the number of new users on in the HealthTech space.46 Zenius jumped 12-fold over the previous year.48 Ruangguru, which had been growing steadily ↪ …while (EdTech) Indonesian EdTech companies provide a range of even before the pandemic, reported a jump in apps have clearly services and products: (i) targeted at students to web-hits from around an average of 7.5 mil- filled a void during help them with learning and upskilling; (ii) tar- lion to over 11 million per month. Industry the pandemic, their overall reach is geted at educators to assist them with student estimates suggest that the usage of telemedi- limited to the more management, communication and teaching; and cine apps in Southeast Asia increased by a fac- affluent clientele (iii) targeted at educational institutions to help tor of 4.5 in March 2020 (at the peak of the in urban centers, them with administration. For example, com- COVID-19 outbreak) compared with January mostly within panies such as Ruangguru, Zenius and Quipper 2020 usage. The number is consistent with Java and usage is higher among the develop and provide self e-learning content, in- growth seen by some of the prominent players relatively better teractive learning platforms and study tools that in Indonesia. Halodoc reported its monthly ac- off help K-12 students expedite the learning process, tive users increasing by a factor of 10 during the along with interactive online services that help pandemic compared with the fourth quarter of students with their assignments and test prepa- 2019, while Alodokter claimed to have experi- ration. Companies such as Arsa Kids, Digikids enced a 1.5-times increase compared with the and Educa Studio develop game-based and blend- pre-COVID-19-outbreak period.49 ed-learning experiences, including interactive storybooks and educational mobile apps, to help However, while these apps have clearly filled a improve early childhood educators’ effectiveness. void during the pandemic, their overall reach is These products and services are typically dissemi- limited to the more affluent clientele in urban nated using several approaches, such as web-based centers, mostly within Java and usage is higher and mobile-based applications. among the relatively better off. According to a nationally representative survey conducted a few On the HealthTech side, innovative products months after the pandemic outbreak, more than and services are geared toward providing health 83 percent of households in Jakarta were found 46 to be providing some form of learning activities protection, etc. However, as experience from using mobile apps and/or online learning to their the more developed parts of the world shows, ↳ more than children.50 The number was significantly low- and Indonesia's own recent experiences con- 83 percent of er in other parts of Java (43 percent), and even firm, these innovations should not be thought households in Jakarta were found lower outside of Java (38 percent). While on av- of as tools to displace the traditional modes to be providing some USING THE MEDIUM erage 54 percent of all Indonesian families took of service delivery, but ones to help enhance form of learning up some form of digital learning to minimize them.52 Especially from the perspective of using activities using the disruption in the education of their children, digital technologies to reduce intergenerational mobile apps and/or the option was available to only 38 percent of inequalities, bigger gains will come from the online learning to their children 83% families in the bottom 40 percent of the income Government’s readiness to experiment with distribution. Lack of complementary inputs and internalize some of these digital solutions such as supporting ICT devices (smartphones, to chip away at bricks-and-mortar challenges computers) and limited access to good-quality to enhance service delivery. In fact, given the internet were reported to be the main reasons differential adoption of digital technologies in- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA behind this observed pattern. duced by the pandemic in sectors such as educa- tion, with wealthier children in better endowed One interesting trend that has been observed schools having leapfrogged their peers from less throughout the world is how even with the wide- affluent backgrounds, it has become even more spread pivot toward online education induced urgent for the Government to catch up. In that by the pandemic, the specific technologies that sense, a vibrant innovation ecosystem that ex- have been adopted for the purpose have sought tends the frontiers of these solutions is some- to replicate the classroom environment instead thing Indonesia should certainly strive for as of deviating from it. Relatively better endowed these innovations expand the menu of options schools have taken to Zoom-classrooms, while available for the Government to experiment across the country WhatsApp has provided a with broader applications.53 medium for student-teacher interactions. Even as education has moved online, the pandemic A key challenge with all these efforts is to move has ended up underscoring the inherent irre- away from ad-hoc solutions to a more compre- placeability of student-teacher interactions.51 hensively thought out whole-of-government Likewise, in health, in-depth interviews with some doctors providing teleconsultations approach to the broader digital transformation of government and service delivery. Another 43% reveal that telemedicine is likely to be most foundational challenge is connectivity. As dis- Other parts useful for early-stage diagnosis and educating cussed above, fixed broadband penetration in of Java patients. It is also likely to be useful for certain Indonesia is very low compared with regional areas of medicine (e.g., mental health) that car- peers. This is an especially binding constraint 38% ry a certain stigma in Indonesia and on which for institutional users such as schools, hospitals, Outside patients may feel more comfortable discussing and other health facilities, which may want to of Java matters through the digital medium. But for intensify their use of digital technologies to en- a broad range of other medical consultations, hance service delivery. For example, of all the doctors regard being able to physically examine 219,000 schools under the Ministry of Educa- patients as an indispensable part of the diag- tion and Culture (MoEC), just 10 percent had nostic process. access to fixed broadband connections in 2019, 42 percent had some form of mobile broadband The EdTech and HealthTech sectors in Indo- connectivity, while a significantly large share (45 nesia face several challenges to their growth. percent) was not connected at all. Closing these These include difficulties with financing, lack connectivity gaps, not just in schools but also in of digital talent in the marketplace, and poor other important service delivery nodes, will be a regulatory clarity, including on sensitive topics crucial step in harnessing digital technologies for such as consumer protection and personal data greater inclusion in post-pandemic Indonesia. 47 → Indonesia does not yet have any officially- recognized digital IDs, but has a relatively strong national ID system that represents a significant asset on which to build a national digital ID ecosystem F or people to be able to ful- such as by using technologies or processes that ly participate in the digital are not compatible with the local context. Data economy and carry out of- protection and privacy breaches require a dual ficial and high-value trans- approach of developing comprehensive legal actions online, countries frameworks that ensure accountability, includ- need to introduce digital ing independent oversight. Vendor and technol- ID systems—a need that has been accentuated ogy lock-in can be mitigated somewhat through by the COVID-19 pandemic. Digital ID systems the adoption of open standards and open-source for online transactions are a natural progression software, competitive procurement, and strong OVERVIEW from national ID systems, which have been used contract and vendor management. For people to be able to fully predominately for in-person transactions, in- participate in the cluding in Indonesia, because a physical ID card While Indonesia does not yet have an official digital economy cannot be used remotely over the internet. digital ID system or ecosystem that would allow and carry out Indonesians to securely verify their legal identity official and high- Digital IDs make use of technologies such as smart- online, the existing national ID system is an asset value transactions BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA phones and cryptography to provide such security could be leveraged to create one relatively easily. online, countries need to introduce and assurance for remote interactions. They can be The national ID system, which is managed by the digital ID systems—a issued by a single entity (centralized) or by multi- Directorate General for Population and Civil Reg- need that has been ple entities in an ecosystem (federation). Emerging istration (Direktorat Jenderal Kependudukan dan accentuated by the standards are also creating opportunities for de- Pencatatan Sipil, Dukcapil), is well-established, COVID-19 pandemic centralized models where the digital ID is stored its database (Sistem Informasi Administrasi Kepen- on a personal device or digital wallet. dudukan, SIAK) has been digitized, and nearly the entire population has a unique identity numbers Well-designed and implemented digital ID sys- (Nomor Induk Kependudukan, NIK). In 2011, a new tems can unlock an enormous amount of econom- ID card (Kartu Tanda Penduduk Elektronik, KTP- ic value for countries, estimated by the McKinsey el) was introduced and biometric recognition Global Institute to be between 3 and 13 percent (fingerprint, iris and faces) was added to assist of GDP by 2030.54 The largest gains will be ex- with removing duplicates in the SIAK, as well as to perienced in the digital economy. First, digital ID enable identity verification. The national ID sys- systems can promote inclusion when they are uni- tem offers a strong foundation on which an official versally accessible and useable because they allow digital ID system or ecosystem should be built all people to do transactions, whether that means on. Therefore, digital ID is a ‘low hanging fruit’ opening a bank account or applying for a social for Indonesia to make services more inclusive, to protection benefit, online.55 Second, they enable promote trust in the digital economy and society, services to be expanded through online channels and to create new drivers of economic growth. and made more efficient. Third, they are part of a Doing so will also allow Indonesia to join all its ‘digital stack’ of platforms that promote innova- fellow middle and high income ASEAN Mem- tion and value added services, such as electronic ber States who have launched whole-of-economy signatures, digital payments, and allowing people digital ID frameworks. to exercise control over their personal data. Fourth, they facilitate cross-border digital transactions. In Both the Ministry of Home Affairs and Ministry the European Union, for example, the eIDAS reg- of Communications and Information (Kominfo) ulation and proposed regulation for decentralized have expressed strong interest in designing and digital identity wallets enable a digital ID issued by launching an official digital ID system or ecosys- one member state to be used in others, without tem. In the absence of such a system, online ser- needing to be physically present. vice providers are using unreliable and insecure mechanisms to verify the identity of Indonesians But there are also risks related to digital ID. Ex- online, such as requesting customers to take self- clusion prevents people not just from register- ies holding the national ID card (KTP-el). While ing for a digital ID but also using a digital ID, this is a practical workaround, some fintech pro- 48 ↪ Digital ID is a ‘low hanging fruit’ for Indonesia MYANMAR to make services more inclusive, to promote trust in LAO P.D.R the digital economy and society, and to create new drivers of economic growth. Doing so will also allow Indonesia to join all its fellow middle and high income ASEAN Member States who have launched whole-of-economy THAILAND digital ID frameworks VIETNAM CAMBODIA USING THE MEDIUM PHILIPPINES MALAYSIA BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA BRUNEI SINGAPORE INDONESIA TIMOR LESTE viders have reported that as many as 60 percent There is substantial demand for digital ID in of customers provide selfies that are unreadable Indonesia across the public and private sectors. or require manual intervention, such as a video E-KYC and digital ID remains one of the priority call, creating unnecessary expense and challenges. issues for the financial sector, from banks to fin- More recently, licensed e-signature providers have tech companies. Likewise, government agencies begun offering basic digital ID authentication ser- such as BPJS Employment and BPJS Health are vices (i.e., going lower in their value chain), but trying to transform the way they offer services to these are based on commercial relationships with citizens, including by shifting to online channels, the service providers and require the charging of but are hampered by the absence of an acces- higher fees in order to make a profit. Other third sible and affordable digital ID system. At the parties have emerged offering ID authentication regional level, there is an opportunity for mutual services, but again these charge substantial fees. recognition arrangements to enable cross-bor- For example, the Kartu Prakerja website had to der transactions as a stepping-stone toward a purchase facial recognition services from a private broader ASEAN-level arrangement, as has been company when it rolled out applications during alluded to in the ASEAN Digital Masterplan for the COVID-19 pandemic. Similar constraints 2025. Beyond ASEAN, there is also an opportu- exist for face-to-face transactions, as Dukcapil nity for mutual recognition arrangements with has not yet launched biometric authentication or the European Union (eIDAS), Australia, the electronic know-your-customer (e-KYC) services United States and Canada, among others. at a national scale. 49 A → Digital upgrade nother important and ing what digital technologies make possible may of government broader pathway to har- nessing digital technolo- be needed for transformation to be realized. can fundamentally gies for greater inclusion Global experience also shows that digital trans- transform the is by enabling these tech- nologies to fundamentally formation of government is a complex under- taking involving multiple stakeholders. In Indo- quality of citizen- transform the quality of citizen-state interac- nesia the public sector’s institutional structure state interactions tions. This could be for specific services, such as education, health and social protection as dis- is particularly complex and highly fragmented with multiple agencies having overlapping man- cussed above, but also for a whole host of other dates. This fragmentation naturally also makes services that could be improved significantly coordination a major challenge, which increases by a broader digital transformation of govern- the complexity of completing even the simplest ment. The key challenge will be to transition of tasks that require cooperation among differ- from the current siloed structure of multiple, ent stakeholders. For example, in more than two incompatible government information and years since the e-Government Regulation was data management systems to a whole-of-gov- promulgated, the e-government coordination ernment, platform-based approach, which has team, which consists of seven-line agencies with emerged as best practice in many economies MenPAN-RB in the coordinating role, has been globally. Related to that is the need to establish unable to convene and agree upon the shared clear leadership and coordination for govern- vision of e-government implementation. In the ment digital services. absence of coordinating body such as the Gov- ernment Digital Service (GDS) in the UK or BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Over the years, the Government has made GovTech in Singapore with authority over all several efforts to digitize government services. relevant stakeholders to overcome these coor- While some of these first-generation efforts dination challenges, the digital transformation have been moderately successful in their re- agenda in Indonesia has not been able to take spective objectives, they have been limited in any concrete shape and form, much less gather terms of their scope and often ad hoc in nature. any real momentum. These attempts to digitalize, often at the be- hest of an agency-head or a subnational entity Similarly, Indonesia does not have a clear lead championing the effort, have resulted in a whole-of-government data management poli- proliferation of information systems, websites, cy. Presidential Regulation No. 39/2019 on One apps and platforms that have very limited in- Data goes some of the way toward addressing teroperability at the front end and create sig- this, but fragmentation remains an issue and nificant duplication of effort and investment several implementation challenges remain to at the back end. be worked out. Each sector is responsible for its own data management, and guidance on In the past two years, the Government has made cross-sectoral data sharing and utilization is efforts to address these digital government is- often missing. The One Data regulation grants sues. One concrete step in this direction is the Bappenas greater authority to regulate, mon- issuance of Presidential Regulation No. 95/2018 itor, and enforce data governance across all on e-Government and Presidential Regulation government agencies. Recently issued imple- No. 39/2019 on One Data. The objective of menting regulations also shed some light on the the regulation on e-government is to imple- intersectionality of this regulation with other ment an integrated e-government system by, initiatives, such as the Regulation on E-Govern- among others, getting all government agencies ment/Digital Government (Presidential Regu- to adopt a common and interlinked enterprise lation No. 95/2018), the Electronic Transaction architecture, co-using IT systems and establish- Law and its implementing regulation (PP No. ing a national coordination team. The thinking 82/2012 and its revision PP No. 71/2019), the on digital transformation as embodied in the forthcoming Personal Data Protection Law,56 e-government initiative focuses to a large extent the Population Administration Law, the Digital on the digitization of existing internal govern- Payment Regulation, and the Omnibus Law.57 ment processes. However, the journeys taken by However, some of the key strategic datasets are countries such as the UK and Singapore, among maintained by various government institutions others, suggest a fundamental reimagination of and the One Data regulation is ambiguous on processes, procedures and structures consider- the role of these administrative entities.58 50 “Presidential Regulation No. 39/2019 on One Data goes some ways in strengthening Indonesia’s data management policy, but fragmentation remains an issue and several implementation challenges USING THE MEDIUM remain to be worked out” BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA 51 Policies to Leverage DTs for Greater Inclusion →What can Indonesia do? 4 Section ⬎ BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA 52 50% POLICIES TO LEVERAGE DTS ↪ Roughly half of the adult population is still not connected to the internet BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA T he analysis presented in this report points to three key policy principles to en- sure that no one is left be- hind: (i) improving digital connectivity and universal- izing access; (ii) enabling the medium to generate economic opportunities for all and unlocking citizen capabilities to seize these opportunities; and (iii) harnessing the medium to upgrade the quality of citizen-state interactions and improve service delivery. Some specific and actionable reforms to implement these principles are discussed below. 3 FIGURE O.17 What can Indonesia Do? policy priorities 1 2 3 Adopt Digital Improve Digital Make the Technologies Connectivity Digital to Upgrade & Universalize Economy Work Citizen-State Access for All Interactions 53 1 IMPROVE DIGITAL CONNECTIVITY & UNIVERSALIZE ACCESS A ddressing the digital divide Consistent with International Telecommunica- and making affordable and tions Union (ITU) recommendations, Indone- high-quality internet avail- sia should endorse an overall target for available able to all Indonesians will IMT spectrum of at least 840 MHz plus alloca- require regulatory reform tions of mmWave spectrum as soon as possible59 in three key areas: (i) spec- and certainly no later than 2024. Such alloca- trum management, specifi- tions should be made in larger contiguous blocks cally sequential freeing up of in accordance with future best practice. MNOs spectrum in specific bands; should also have the flexibility to use their allo- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA (ii) greater regulatory clari- cated IMT spectrum for mobile broadband and/ ty on provisions for active and passive infrastruc- or fixed wireless access (FWA) services. High ture sharing; and (iii) greater competitiveness and speed FWA services using 4G/5G technologies regulation of the telecom industry more broadly. are proving globally to be a very competitive product with fixed broadband services.60 Optimize spectrum allocation for mobile Strengthen mechanisms broadband. to ensure sharing of active and passive infrastructure. Under the recently enacted Omnibus Law on Job Creation No. 11/2020, reform of spectrum Tower sharing for mobile broadband networks management is on the right track for optimiza- was mandated in 2009 and has been deployed tion of spectrum allocation. The law also impos- at a large scale but sharing other passive infra- es a mandatory two-year plan to convert analog structure, such as ducts, poles, etc. (required television to digital television in order to secure for fiber optic networks) between provid- digital dividend in the 700 MHz spectrum band. ers will likely remain limited without further Release of the 2.6 GHz band, currently used for regulatory reform. The 2009 Tower Sharing satellite TV, would add capacity in urban centers Regulation brought efficiency improvements and alleviate network congestion. Menkominfo to deployment of towers with some sharing of should consider accelerating its plans for reallo- passive infrastructure across sectors, between cation of this band from satellite broadcasting (rail)roads and fiber optic, and electricity poles to mobile broadband before the current end of and fiber optic. But sharing between telecom the spectrum license in 2024. Securing 3.5 GHz operators lags behind, despite a joint letter is- band spectrum will require consultation with sued by Menkominfo and MoHA in 2018. The current C-Band satellite users. Over time, some Omnibus Law mandates passive infrastructure form of sharing should be considered with use sharing through a change to the Telecom Law of this band for mobile services in urban centers Article 34A and B, but implementation of the and for satellite services in those rural areas that associated provisions will require inter-agency still require C-band connectivity. This would add coordination among national agencies and local further capacity and enable introduction of 5G. governments. Separately, active infrastructure Finally, the GoI should prepare to make available sharing would promote competition in mobile the mmWave spectrum in the 24-29 GHz band broadband services in rural/remote areas but is to enable immediate deployment as soon as the currently not allowed under Telecom Law PP industry is ready for 5G mmWave services. 52/53. The Law requires in many cases separate 54 2 MAKE THE DIGITAL ECONOMY WORK POLICIES TO LEVERAGE DTS FOR ALL S deployment of (mobile) infrastructure by each everal crucial interventions will be required to and every telecom operator. The Omnibus Law make the digital economy work for all. does include a change to Article 34B that opens up the possibility of active infrastructure shar- ing, although that could have been made more Support the development explicit. Telecom Law PP 52/53 would have to of logistics. be updated to allow for sharing of active infra- structure on a B2B basis. Reducing the cost of logistics to enable an ef- ficient and cost-effective movement of goods BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Strengthen competition across the archipelago will enhance the impact e-commerce can have on boosting the produc- along the broadband tivity of small and micro-entrepreneurs, espe- value chain. cially those in the lagging places of the country. Continuing to develop basic physical infrastruc- To improve the competitiveness of service pro- ture to improve connectivity through strength- viders, Indonesia should consider transitioning ened and/or rehabilitated infrastructure such as toward the unified licensing of service provid- roads, ports and electricity will remain a crucial ers to enable each to deliver a larger portfolio foundation of this effort. This may also require of services. The current regulatory regime leveraging private sector capital and expertise, restricts the ability of operators to provide a as appropriate through public-private partner- full range of services by requiring them to ap- ships. Separately, fostering an innovation envi- ply for and maintain specific service licenses, ronment that is conducive to the scaling of the instead of a single uniform license for all ser- several e-logistics that are emerging to provide vices. This limits the issuance of telephony ser- customized solutions to connect enterprises to vice licenses and the portability of telephone market will also be important. numbers, etc. To address these issues, it is rec- ommended that the GoI reviews the current licensing regime and considers transitioning Nurture digital skills and toward a single/unified license to deliver the skills for the 21st century. full portfolio of services to a larger number of Institutional action and policies also need to service providers, so that effective competition recognize that digital skills are a subset of a for dual-play and triple-play fixed broadband broader skillset needed for the 21st century services will emerge. This should be comple- digital economy. The exponential pace of tech- mented by regulations enabling portability of nological change today makes it hard to antic- telephone numbers across providers. Neither ipate which job-specific technical, digital and the Telecommunication Law nor its amend- other skills will thrive and which will become ments in the Omnibus Law addresses these is- obsolete in the near future. As a result, the abil- sues. Telecommunications-related regulatory ity to adapt quickly to changes is increasingly commitments in the Regional Comprehensive valued by the labor market.63 As such, the most Economic Partnership (RCEP) and the ASE- sought-after trait globally is adaptability—the AN Digital Masterplan (ADM 2025),61 which ability to respond to unexpected circumstances was launched in January 2021 at the 1st ASEAN and to unlearn and relearn quickly. This trait Digital Ministers’ Meeting,62 should also guide requires a combination of certain cognitive further regulatory reforms. skills (critical thinking, problem-solving) and 55 2 socio-behavioral skills (curiosity, creativity). Indeed, the top five skills and skill groups that training with personal coaching to reshape the behavioral skills of university students.66 executives in Indonesia’s largest companies see as rising in prominence in the lead-up to Close collaboration between industry and ter- 2025 are creativity, originality and initiative, tiary education is critical. The low quality of complex problem-solving, active learning and TVET and tertiary education in Indonesia has learning strategies, emotional intelligence, and been linked to, among other factors, lack of com- analytical thinking and innovation.64 petency frameworks developed in consultation with the private sector. Inadequate labor-market Modern tertiary education needs to cultivate in information and intermediation make it hard to students a minimum threshold of foundational align curricula and teaching with the occupations “transferable” higher-order skills for the 21st and skills that are most needed. Private sector century digital economy, even in STEM fields. participation in tertiary education planning and BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Technology and integration have increased policy is thus a requisite both at a strategic and the demand for higher-order general cognitive technical, curricular level. In China, for exam- skills—such as complex problem-solving, crit- ple, Lenovo is working with tertiary institutes to ical thinking, and advanced communication— train vocational students in high-tech areas, such that are transferable across jobs. Therefore, the as cloud computing, that feature practice-based combination of general and technical skills is curricula, practitioner-led instruction, and pro- becoming highly valued (World Bank 2019). fessional certification.67 The GoI also needs to Tertiary education systems should therefore incentivize employers to offer internships and guarantee a minimum threshold of transferable off-campus learning to students. Second, fill- cognitive skills, which are also the best inocu- ing in information gaps during the job-search lation against job uncertainty. Incorporating process enables students to make better choic- more general education in tertiary programs is es between and within different paths. Chile is one way to do this. An additional year of general establishing online platforms where students education was added in 2012 to undergraduate can access information on the employability of programs in Hong Kong and China, focusing on individuals with various degrees, wage profiles, problem-solving, critical thinking, communi- and courses to take for certain occupations. cation, leadership, and life-long learning skills Colombia’s Jóvenes en Acción (Youth in Action) and with early assessments showing positive program combines classroom instruction with results.65 Another way is through innovative on-the-job training at private companies. Indo- pedagogy. The Faculty of Architecture and nesia needs to continue to develop data systems Environmental Design at the College of Sci- that allow for the identification of occupations ence and Technology, University of Rwanda, and skills that are most in demand, monitor ed- has promoted learning strategies that include ucational institutions’ compliance with quality open-ended assessments, feedback opportuni- standards, and ensure that the information on ties, and a progressive curriculum that balanc- employability of degrees, wage profiles, and es academic challenges with student support. occupation-specific courses is available to both These approaches have improved the criti- jobseekers and workers. cal-thinking skills of students. Forward-look- ing universities are finding ways for adult stu- dents to acquire a broad set of socio-behavioral Promote the use of skills. Vocational colleges in the Netherlands digital financial services/ are providing entrepreneurial courses aimed at payments, including improving noncognitive skills such as teamwork and self-confidence. Tunisia has introduced an among the unbanked and entrepreneurship track that combines business the underbanked. 56 Financial exclusion and low uptake of digital payments impedes the growth of the digital Use tax policy POLICIES TO LEVERAGE DTS economy and locks out a significant proportion instruments to ensure of the population from participating and bene- level playing field. fiting from it. For many Indonesians, especially those in the lower parts of the distribution, A well-designed system of taxation for the dig- not having a bank account is already a major ital economy can help level the tax playing field setback. The variety of efforts underway to between conventional and online businesses; improve the supply of digital payments and a within online businesses, between goods and broader suite of financial services remain im- services; and between resident and non-resident portant entry-points. One promising digital businesses. This will reduce the distortive impact solution is the reduction in verification costs taxation may have on the economy, helping to that could come from widespread availabili- ensure that sales, profits, and investment deci- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA ty of e-KYC option and a strong and reliable sions in the digital sector are driven by market digital ID system. Second, for those already dynamics and efficiency improvements, and not banked, including some of the more sophis- by advantages gained through uneven taxation ticated users, trust in online transactions and policy or taxpayer avoidance. In the Indonesian issues related to data privacy, cyber securi- context, two incidental benefits are equally rele- ty and financial integrity represent another vant. First, taxation of the digital economy will barrier that needs to be overcome. While the bring in a small but growing amount of revenue Financial Services Authority (Otoritas Jasa as digitalization accelerates within the Indonesian Keuangan, OJK) has been trying to strengthen economy,68 especially in the context of COVID-19 the legal and regulatory framework for DFS, pandemic. E-commerce in Indonesia, for exam- consumers still regard the disclosure of per- ple, is projected to have grown by 54 percent sonal ID to service providers as a major risk. during 2020, reaching US$32 billion, at a time In this regard, the passing of the draft Law on when private consumption in the overall economy Personal Data Protection will be critical for has suffered, falling by an estimated 2.7 percent promoting the adoption of DFS in Indone- overall.69 Second, taxation of the digital economy sia. Third, improving the interoperability of presents Indonesia with a unique opportunity to payment systems can stimulate use cases and boost formalization of businesses, particularly increase transactions volume. An incidental that of MSMEs. Registered businesses will enjoy benefit of this could also be an improvement in the benefits of having easier access to credit from the viability of the agent-based model, which the financial system, and of potential future fiscal is crucial to expanding DFS access, especial- support from the Government. ly among the underserved demographic for whom this model has shown to help overcome Indonesia is taking measures to reform its tax knowledge and trust deficits. Finally, invest- policy and modernize its tax administration ments in financial literacy programs should to deal with the challenge of intangible trans- continue to remain shared responsibilities of actions for VAT. Indonesia’s broad policy po- the regulator as well as market participants sition was reflected in Law No. 2/2020, which and should cover material on the variety of fi- provided the legal umbrella for imposing VAT nancial products and services but also of the on digital goods and services provided by for- risks: financial risks such as online fraud, dig- eign suppliers. Minister of Finance Regulation ital footprint, overborrowing; digital financial No. 48/ PMK.03/2020 (hereafter PMK-48) and risks associated with protection of personal DGT Regulation PER-12/PJ/2020 (hereafter information; and consumer risks and associ- PER-12) provided the next levels of policy and ated redress procedures. implementation details. They applied the ex- isting VAT statutory rate of 10 percent on all 57 2 intangible taxable goods and all taxable services that are provided by foreign suppliers through regard. First, the GoI will need to ensure adop- tion and communication of consistent rules, an electronic system. Both foreign and domestic leverage an efficient IT system to administer digital platforms are required to collect the VAT registration, filing, and payment, and combine on behalf of foreign suppliers, subject to being digital transaction data with other third-party appointed as “VAT Collectors” by the DGT. To data and taxpayer data to enhance compliance date, companies that have been asked by the risk management. Second, Indonesia’s overly DGT to serve in this capacity include Amazon, generous VAT threshold needs to be reviewed Facebook, Google, LinkedIn, McAfee, Netflix, to expand the digital economy tax base. Indo- Skype, Spotify, Twitter, and Zoom.70 The qual- nesia’s VAT threshold of IDR 4.8 billion stands ifying criteria for being appointed a VAT Col- out in international comparisons of VAT thresh- lector are based on: (i) annual sales volumes of olds to GDP per capita, a metric commonly used at least IDR 600 million for digital goods and to compare the generosity of VAT thresholds BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA services in Indonesia (or IDR 50 million in a worldwide. Less-developed economies tend to month); or (ii) online traffic volumes involv- have higher thresholds, in part because of weak- ing at least 12,000 users within Indonesia in 12 nesses in tax administration and in part because months, or 1,000 users in one month. To sup- they have larger informal economies.76 However, port compliance with the reforms, the DGT has even when compared against low-income econo- introduced several administrative measures.71 mies, Indonesia’s VAT ratio stands out as being too generous.77 A high threshold means that a Likewise, to deal with the challenge of low-val- vast share of businesses whose annual turnover is ue shipments Indonesia has lowered its mini- below the VAT threshold are excluded from the mum threshold from US$75 to US$3. In so VAT system, narrowing the base and distorting doing, Indonesia has become an early mover in the tax. This distortion is even greater when the this reform space, following on from Australia, threshold is applied to e-commerce in Indonesia, which was the first to slash its GST threshold which is characterized by a large share of small from AU$1,000 to zero from July 2018.72 The companies.78 EU has followed suit, with its previous exemp- tion of consignments of less than €22 abolished Overall, design and implementation of digital from January 2021 onwards.73 To allow VAT to economy taxation reform must be geared around be levied, all imports into the EU now must be the core principles of equity, efficiency, and sim- declared using an electronic customs declara- plicity, so that Indonesia’s economy and society tion. To ease the implementation burden, the can share in the benefits of digitalization. Over EU has also introduced a simplified customs time, as digitalization expands, the demarcation declaration form for all declarations of goods between the digital economy and the economy up to €150.74 In Indonesia, implementing the at large will become increasingly blurred and new rules without overwhelming the customs eventually dissolve. This evolving reality makes administration or creating unnecessary burdens it ever more important that the GoI gets tax- on business will be more challenging, requiring ation reform right. This means ensuring that effective risk management.75 Experience from tax policy and administration meet the princi- Australia and the EU may provide the GoI with ples of good taxation. Create an uneven playing some useful lessons. field of taxation, for example, by having different tax rates on online and offline businesses, or by The effectiveness of these instruments—in gen- enforcing rules on cross-border businesses but erating additional revenue, as well as in leveling not on domestic ones, and Indonesia will end up the playing field—will depend on its ability to with a tax system that detracts from the organic, effectively manage compliance with the new healthy growth of the digital economy. Revenue requirements. Two measures can help in this lost to special incentives that favor the select few 58 will mean less financing for the critical public investments need to enable inclusive growth of digitalization. Ultimately, taxation must not dis- tort business decisions on how to operate, and it should not alter consumer choice on what and where to buy—whether it be from a supermar- ket or a hypermarket, an online marketplace or via an online social media app. Taxes imposed 3 USE DIGITAL TECHNOLOGIES TO PROVIDE BETTER SERVICES AND UPGRADE CITIZEN-STATE must be fair and equal, and administered with a INTERACTIONS minimum burden on all. T his can start with a (i) concerted push on a BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA ↪ Overall, design and whole-of-economy nation- implementation of digital economy al digital ID initiative; (ii) a taxation reform must be geared thorough re-imagination of whole-of-government approach to digital trans- around the core principles formation driven by an agency empowered to of equity, efficiency, and resolve interagency coordination challenges; simplicity, so that and (iii) a special recognition and focus on the question of data integration which is an essen- Indonesia’s economy and tial bedrock of any effort to digitally modernize society can share in the government functions and services. benefits of digitalization Develop a national digital ID framework. Indonesia needs to launch a national digital ID initiative, bringing together various government stakeholders, the private sector and civil society to chart out the optimal path for Indonesians to be able to prove their identity over the internet and thus carry out trusted online transactions. The status quo constrains growth of the digital econ- omy and introduces additional costs and risks for service providers interacting with customers over the internet, including identity fraud. Thailand offers a useful example: its National Digital ID ini- tiative was born out of the National Digital Econ- omy Committee chaired by the Prime Minister. Similar digital ID initiatives have been launched out of agencies attached to heads of government, such as SingPass by the Singapore Government Technology Agency (GovTech) and United King- dom’s Verify by the Government Digital Service (GDS). An initiative in Indonesia would ideally be spearheaded by the Kantor Staf Presiden (KSP) 59 3 with central roles for the Ministry of Home Af- fairs as the authority responsible for population ID system and the digital economy more broadly, by providing safeguards and accountability for data (i.e., the legal identity of Indonesians), and the collection, use and sharing of personal data, the Menkominfo as the authority responsible for as well as formalizing the rights of data subjects. regulating electronic transactions, and also in- volving the Ministry of National Development Planning (Bappenas), the Coordinating Ministry Reorient from a narrow for Economic Affairs, Bank Indonesia and OJK. focus on e-Government to a comprehensive national Improvements to the national ID system will cre- ate a strong base on which to introduce a digital digital transformation ID system or ecosystem. It contains the popula- agenda. tion data on which digital IDs can be issued— by BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA the Government and/or the private sector—with The GoI is strongly recommended to consider high levels of assurance. This seamless onboard- a fundamental transition from the current nar- ing simplifies and significantly reduces the costs row focus on e-government to a comprehensive of introducing digital IDs in Indonesia. For the National Digital Transformation agenda. Given national ID system to realize its full potential the complexities, this can be implemented in to support service delivery and underpin a new phases. The current e-government initiative that digital ID system throughout Indonesia, the GoI is chaired by the Ministry of Administrative and will need to decide whether to build a central- Bureaucratic Reform (MenPAN-RB) focuses on ized digital ID system implemented exclusive- the digitization of internal government process- ly by the GoI (likely Dukcapil), or a federated, es. Looking at the scope in the e-Government multi-stakeholder ecosystem implemented by regulation, the role fits the main duties and re- multiple digital ID providers but still regulated sponsibilities of MenPAN-RB, which includes the by the Government through a trust framework management of government business processes. Doubling down of laws, rules and standards, with the possibility However, if the GoI aims to achieve truly com- on reforms related of limiting public sector transactions to a digital prehensive digital transformation of the nation, to the analogue ID issued by Dukcapil. it will need to shift the paradigm towards digital foundations of the government and rearrange its vision, governance, digital economy is crucial to ensuring One area that requires urgent attention is that and delivery model to better suit that aspiration. that digital Indonesia does not yet have a comprehensive Such a commitment would reorient Indonesia technologies do not data protection law with general applicability. in the same direction as that taken by digitally end up accentuating This undermines trust in any kind of collection, advanced countries, such as Singapore, the Unit- existing processing and sharing of personal data, which ed Kingdom, Australia, Estonia, the Republic of inequalities are core functions of a digital ID system, as well Korea and Canada.79 The digital government as the digital economy more broadly. A draft Law platforms would serve as the foundation of digital on Personal Data Protection was submitted by the services in the economy such that both public sec- President to Parliament at the end of 2019—a bill tor institutions and private enterprises can take that was largely modelled on the EU’s General advantage of these platforms to deliver services Data Protection Regulation (GDPR). An essen- to every citizen. tial feature for the credibility and strength of such a law would be an independent oversight entity, A short-term priority may be to continue to similar to the Office of the Australian Informa- strengthen the role of MenPAN-RB in orches- tion Commissioner and the Singaporean Personal trating and delivering the existing e-government Data Protection Commission. Such a law will help mandate. However, in parallel, Indonesia needs to ensure public trust and confidence in the exist- to create the regulatory and institutional frame- ing national ID system, as well as any future digital works necessary to imagine, articulate and imple- 60 ment a comprehensive transformation agenda. the One-Data policy comprehensively in priori- To start with, a presidential regulation could be ty sectors relevant in managing the COVID-19 POLICIES TO LEVERAGE DTS prepared to mandate: (i) the development of a pandemic and revitalizing the economy (edu- National Digital Transformation strategy; (ii) the cation, health, MSMEs, and social protection). establishment of a policy-making entity; and (iii) Appointing a data steward in each sector, assisting the establishment of an implementation entity. the stewards to develop a master data reference The policy-making entity—a Digital Transfor- for each sector and data standards, and to enforce mation Taskforce (DTT)—would ideally be at- the standards, and assisting policy makers in these tached to the Office of the President, overseeing sectors to draw insights from improved data man- and, if possible, consolidating and streamlining agement, which can then be used for expansion the governance structure of existing digital ini- to other sectors could be some concrete steps . tiatives, such as the e-Government agenda led by Building alliances with key institutions such as MenPAN-RB, One Data led by Bappenas, Online MenPAN-RB, Menkominfo and MoHA, to in- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Single Submission, and the One-Map Policy led still One-Data principles into implementation by CMEA. This would reduce the fragmentation of the Digital Government and Digital ID initia- that characterizes the governance structure of tives could be other measures. Bappenas could these initiatives, which has led to ineffective and also work closely with MoF during the execution unnecessarily complex implementation at a high of shared planning and budgetary roles to identify cost to coordination. Finally, the implementation and/or filter out programs and activities that lead arm of the DTT would be a Digital Transforma- to duplication or inefficient production of data. tion Implementation Agency (DTIA), similar to agencies in Singapore (GovTech) and the United Finally, doubling down on reforms related to the Kingdom (GDS). DTIA would be responsible analogue foundations of the digital economy is for delivery of cross-sectoral digital services, i.e., crucial to ensuring that digital technologies do platforms and supporting technologies for sec- not end up accentuating existing inequalities. toral ministries to host and operate their digi- A key finding of the report is that inequality in tal services. It would also serve as the ecosystem access to the digital medium and the ability to builder for the delivery of sector-specific digital benefit from this medium mirrors existing di- services, coaching and assisting sectoral minis- mensions of inequality—across groups, regions, tries to develop quality, and standardized digital income classes and skill levels. This implies that services for citizens. Both DTT and DTIA need the non-digital structural constraints that have not be newly established entities. Existing enti- long been realized as key barriers to inclusive ties within the public sector could be empowered development continue to remain important de- to play the role. Irrespective of the model that is terminants of how effectively Indonesia is able to chosen, the most crucial part of the reform would harness the digital dividends for its poor. Invest- be to ensure that the institutional arrangement ments and reforms directed toward a stronger facilitates a whole-of-government approach and economic integration of the country, reforms eliminates fragmentation. related to trade, competitiveness and the overall business environment, investments to improve Transition towards a the quality of Indonesia’s human capital, reforms to improve the effectiveness of the state will not whole-of-government only enhance Indonesia’s digital dividends but data management policy. also ensure that these are shared equitably across the population, In that sense, digital technologies To this end, Bappenas could strategize to ef- should not be seen as tools that can be used to fectively to develop and implement subsidiary leapfrog and circumvent conventional develop- regulations and also rally support for the One ment challenges, but rather as complementary Data agenda. Specifically, Bappenas could apply tools to effectively address some of them. 61 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA 62 destiny. does not h be Indone The digita esia’s have to tal divide 63 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA POLICIES TO LEVERAGE DTS Summary of Policy Recommendations � PRIORITY OBJECTIVE Improve Digital Connectivity I to Universalize Access 1 Optimize spectrum 2 Strengthen mechanisms to allocation for mobile ensure sharing of active and broadband passive infrastructure BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA → Implement the → Optimize the higher → Update the Telecom → Implement through “Analogue Switch Off” frequency capacity Law, PP 52/53, to regulations the by the end of 2022 in bands, starting with allow for sharing of passive infrastructure accordance with the the freeing up of the active infrastructure sharing mandated under Omnibus Law so as to 2.6 GHz band, followed on a B2B basis the Omnibus Law free up the 700MHz by the 3.4-3.8 GHz band for greater rural band and potentially RESPONSIBLE AGENCIES: RESPONSIBLE AGENCIES: connectivity using 4G the broader 3.3-4.2 Kominfo Kominfo and future 5G networks GHz band to facilitate MoHA 5G deployment in urban RESPONSIBLE AGENCIES: Ministry of Public Works areas. Consideration Kominfo should be given to the release of the mmWave spectrum bands for 5G in urban areas RESPONSIBLE AGENCIES: Kominfo → Review current → Appoint an 3 Strengthen competition licensing regime and independent regulatory along the broadband transition toward body for the telecom international best sector consistent with value chain practice of single/ exemplar practice unified licensing and Indonesia’s RCEP to allow a larger commitments number of providers to deliver the full RESPONSIBLE AGENCIES: portfolio of services Kominfo RESPONSIBLE AGENCIES: Kominfo 64 PRIORITY OBJECTIVE Make the Digital Economy II Work for All 1 2 Nurture digital skills and skills for the 21st century digital economy Support the POLICIES TO LEVERAGE DTS development of logistics → Continue to develop → Align non-formal education → Build partnerships between basic physical services and lifelong learning tertiary institutions and infrastructure to with needs of the working-age the private sector to train improve connectivity adult population, adopting a vocational students in high- through strengthened modular approach to course tech areas using practice-based and/or rehabilitated offerings with an emphasis curricula, practitioner-led infrastructure on soft skills in addition instruction, and professional (roads, ports, to technical digital skills, certification and electricity), incorporating employment- RESPONSIBLE AGENCIES: BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA leveraging private specific digital skills sector capital relevant for Indonesia, Ministry of Education and Culture/Private and expertise as emphasizing a mindset of Sector appropriate through continuous learning, and PPPs offering opportunities to practice “self-managed” RESPONSIBLE AGENCIES: learning Ministry of Transport RESPONSIBLE AGENCIES: Directorate General of Highways Ministry of Education and Culture Ministry of Public Works Ministry of Manpower and Housing → Ensure that tertiary → Incentivize employers to → Deepen reforms education offers a minimum offer internships and off- to reduce entry threshold of foundational campus learning to students barriers to logistics “transferable” higher-order and transportation RESPONSIBLE AGENCIES: skills such as critical services and build thinking, problem-solving and Ministry of Education and Culture/ long-term investor Ministry of Manpower communication, even in STEM confidence, to expand fields warehousing outside of large metropolitan RESPONSIBLE AGENCIES: areas Ministry of Education and Culture RESPONSIBLE AGENCIES: Ministry of Transport, with → Continue to develop data systems that allow for identification support from Ministries of Trade and Finance and of occupations and skills in demand, monitor educational the Coordinating Ministry institutions’ compliance with quality standards, and ensure of Maritime Affairs and that information on employability of individuals with various Investments degrees, wage profiles, and occupation-specific courses is available to jobseekers and workers → Continue to modernize the National RESPONSIBLE AGENCIES: Post Service; Ministry of Education and Culture/Ministry of Manpower standardize addresses and postcodes RESPONSIBLE AGENCIES: Ministry of Communication and Informatics 65 PRIORITY OBJECTIVE Make the Digital Economy II —CONTD. Work for All 3 Promote supply of DFS/digital payment solutions that cater to the unbanked and the underbanked → Streamline the → Standardize rules and licensing and procedures to enable registration processes interoperability of required to become a payment schemes provider of digital financial services RESPONSIBLE AGENCIES: Bank Indonesia RESPONSIBLE AGENCIES: Bank Indonesia OJK BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA → Incentivize the → Increase use cases → Strengthen the private sector to of DFS products legal and regulatory innovate and develop in the delivery of framework for managing new DFS products that government services risks related to cater to the rural (including Government- data governance population to-person payment) to and privacy, cyber sustain the commercial security and RESPONSIBLE AGENCIES: viability of DFS operational risk, and Bank Indonesia agents financial integrity, in order to build Directorate General of Taxes RESPONSIBLE AGENCIES: consumer trust in DFS CMEA (Financial Inclusion products Council) RESPONSIBLE AGENCIES: Ministry of Social Affairs Bank Indonesia OJK 4 Use tax policy instruments to ensure a level playing field → Complete the design of measures to reform → Lower the overly tax policy and modernize tax administration generous VAT threshold in relation to DE transactions, ensuring to expand the digital adoption and communication of consistent economy tax base rules, institution of efficient IT systems to administer registration, filing, payment RESPONSIBLE AGENCIES: and use of digital transaction data with Ministry of Finance other third-party data and taxpayer data for (Directorate General of Taxes, Directorate General of Customs strengthened compliance risk management and Excise, and Fiscal Policy Agency) RESPONSIBLE AGENCIES: Ministry of Finance (Directorate General of Taxes, Directorate General of Customs and Excise, and Fiscal Policy Agency) 66 Use Digital Technologies to Provide PRIORITY OBJECTIVE III Better Services and Upgrade Citizen- State Interactions 1 Develop a national digital ID framework POLICIES TO LEVERAGE DTS → Pass the draft Law → Launch a whole- → Close the coverage → Introduce a national on Personal Data of-economy national gaps in the national digital ID system or Protection digital ID initiative population registry federated ecosystem to define the optimal database (SIAK), fit-for-purpose for RESPONSIBLE AGENCIES: model for Indonesia, streamline the new online transactions in Kominfo bringing together registration and the Indonesia context, government, private update processes, and building on the SIAK MoHA sector and civil introduce biometric- society based e-KYC processes RESPONSIBLE AGENCIES: Office of the President RESPONSIBLE AGENCIES: RESPONSIBLE AGENCIES: BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA MoHA Office of the President MoHA Kominfo MoHA Kominfo 2 Reorient from a narrow 3 Implement a whole- focus on e-Government to of-government data a comprehensive national management policy digital transformation agenda → Strengthen the → Adopt a whole- → Implement the → Build alliances political and of-government One Data Policy with key institutions bureaucratic influence approach to digital comprehensively, in such as MenPAN- needed to move the transformation, priority sectors RB,Kominfo, and MoHA agenda by placing a coordinated and relevant for managing to instill One Data central government spearheaded by an the post pandemic principles into the authority like agency that has multi- recovery (e.g., implementation of the the Office of the ministry oversight. education, health, digital government President in the social protection or and digital ID driving seat RESPONSIBLE AGENCIES: MSMEs). initiatives. Office of the President RESPONSIBLE AGENCIES: RESPONSIBLE AGENCIES: RESPONSIBLE AGENCIES: Office of the President Bappenas Bappenas BPS BPS and other line agencies and other line agencies 67 References 1 Alibaba Group and World Bank. 2019.  E-commerce Development: Experience from China (English). Washington, DC: World Bank Group. 2 Al-Rikabi, Jaffar, Khalil Rohman, and Hambali. 2021. “On the Macroeconomics of the Digital Economy in Indonesia”, Working Paper, World Bank. Forthcoming. 3 Annur, Cindy Mutia. 2020b. “Layanan Telemedicine Diprediksi Tetap Berkibar Usai Pandemi Berakhir.” Katadata. Retrieved from: https://katadata.co.id/ameidyonasution/digital/5f45270a9ad78/layanan-telemedicine-diprediksi-tetap- berkibar-usai-pandemi-berakhir (accessed November 16, 2020). REFERENCES & ENDNOTES 4 ASEAN and Plum Consulting. 2021. “ASEAN Digital Masterplan 2025”. Retrieved from https://asean.org/storage/ ASEAN-Digital-Masterplan-2025.pdf (accessed March 25, 2021). 5 ASEAN. 2021. “Joint Media Statement of the 1st ASEAN Digital Ministers’ Meeting and Related Meetings”. Retrieved from https://asean.org/joint-media-statement-1st-asean-digital-ministers-meeting-related-meetings/ (accessed March 25, 2021). 6 Asosiasi Healthtech Indonesia. 2020. “About.” Asosiasi Healthtech Indonesia, accessed November 16, 2020, https:// healthtech.id/about/ 7 Australian Taxation Office. 2019. “Who charges GST”. Retrieved from: https://www.ato.gov.au/Business/International- tax-for-business/GST-on-low-value-imported-goods/Who-charges-GST/ (accessed November 16, 2020). 8 Autor, David, Dorn, David, Katz, Lawrence, Patterson, Christina, and Van Reenen, Jon. 2017. “Concentrating on the Fall of BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA the Labor Share.” American Economic Review 107(5): 180–85. 9 Bahia, Kalvin, Pau Castells, Genaro Cruz, Takaaki Masaki, Xavier Pedrós, Tobias Pfutze, Carlos Rodríguez-Castelán and Hernan Winkler. 2020. “Welfare Effects of Mobile Broadband Internet: Evidence from Nigeria” World Bank Policy Research Working Paper #9230. Washington DC: World Bank. 10 Banerjee, Abhijit V. and Esther Duflo. 2019. Good Economics for Hard Times. Allen Lane, United Kingdom. 11 Bhardwaj, Riaz; Yarrow, Noah; Cali, Massimiliano. 2020. “EdTech in Indonesia: Ready for Take-off ?”. World Bank, Washington, DC. 12 Brynjolfsson, Eric, Yu (Jeffrey) Hu, and Michael D. Smith. 2003. “Consumer Surplus in the Digital Economy: Estimating the Value of Increased Product Variety at Online Booksellers”. Management Science, 49(11):1580–1596. 13 Brynjolfsson, Erik, and JooHee Oh. 2012. “The Attention Economy: Measuring the Value of Free Digital Services on the Internet” Thirty Third International Conference on Information Systems, Orlando. 14 Burhan, Fahmi Ahmad. 2020a. “Strategi Anyar Ruangguru hingga Zenius Gaet Pengguna saat Pandemi,” Katadata. Retrieved from: https://katadata.co.id/desysetyowati/digital/5f6de371e728a/strategi-anyar-ruangguru-hingga-zenius- gaet-pengguna-saat-pandemi (accessed November 16, 2020). 15 Council of the European Union. 2017. “VAT on electronic commerce: new rules adopted”, Press Release, Council of the EU. Retrieved from: https://www.consilium.europa.eu/en/press/press-releases/2017/12/05/vat-on-electronic-commerce-new- rules-adopted/ (accessed November 16, 2020). 16 Dolfen, Paul, Liran Einav, Peter J. Klenow, Benjamin Klopack, Jonathan D. Levin, Laurence Levin, and Wayne Best. 2017. “Assessing the Gains from E-Commerce”. Stanford University, mimeo. 17 European Commission. 2021. “New form of customs declaration for low value consignment”. Retrieved from: https:// ec.europa.eu/taxation_customs/news/new-form-customs-declaration-low-value-consignments_en (accessed November 16, 2020). 18 Frank, Morgan R., David Autor, James E. Bessen, Erik Brynjolfsson, Manuel Cebrian, David J. Deming, Maryann Feldman et al. 2019. “Toward understanding the impact of artificial intelligence on labor”. Proceedings of the National Academy of Sciences, 116(14):6531-6539. 19 Global Mobile Suppliers Association. 2021. “4G-5G FWA Company Directory-2: Device Ecosystem – March 2021”. Retrieved from https://gsacom.com/paper/4g-5g-fwa-company-directory-2/ (accessed November 16, 2020). 20 Goldfarb, Avi and Catherine Tucker. 2019. “Digital Economics”. Journal of Economic Literature, 57(1):3-43. 21 Google, Temasek and Bain. 2020. “e-Conomy SEA 2020 – At Full Velocity: Resilient and Racing Ahead”. 22 Goolsbee, Austan, and Peter J. Klenow. 2006. “Valuing Consumer Products by the Time Spent Using Them: An Application to the Internet”, American Economic Review 96(2):108-113. 23 Hjort, Jonas, and Jonas Poulsen. 2019. “The Arrival of Fast Internet and Employment in Africa.” American Economic Review 109(3):1032-1079.   24 Hootsuite and We Are Social. 2019. Indonesian Digital Report 2019. Retrieved from https://datareportal.com/reports/digital- 2019-indonesia (accessed November 16, 2020). 68 25 International Telecommunication Union. 2019. “Digital Infrastructure Policy and Regulation in the Asia-Pacific Region”. Retrieved from https://www.itu.int/en/ITU-D/Regional-Presence/AsiaPacific/SiteAssets/Pages/Events/2019/ RRITP2019/ASP/ITU_2019_Digital_Infrastructure_5Sep2019FNL.pdf 26 Kop, Rita. 2020. “The challenges to connectivist learning on open online networks: Learning experiences during a massive open online course.” The International Review of Research in Open and Distributed Learning 12:19–38. 27 Minges, Michael. 2016. “Exploring the relationship between broadband and economic growth” World Development Report background papers. Washington DC: World Bank. REFERENCES & ENDNOTES 28 Pazarbasioglu, Ceyla, Alfonso Garcia Mora, Mahesh Uttamchandani, Harish Natarajan, Erik Feyen, and Matthew Saal. 2020. Digital Financial Services. World Bank, Washington, DC.  29 Reuters. 2020. “Indonesia adds Twitter, Zoom to tech companies that must pay 10% VAT”. Reuters. Retrieved from: https://www.reuters.com/article/us-indonesia-tax-digital-idUSKBN25Z2CU (accessed March 25, 2021). 30 Tiwari, Sailesh, Virgi Sari, Imam Setiawan and Juul Pinxten. 2020. “Pandemic, Poverty and Policy: Ex Ante Poverty and Distributional Impact of the COVID-19 Pandemic” World Bank Policy Research Working Paper, forthcoming. 31 Tracxn. 2019. “EdTech Startups in Indonesia”. Retrieved from: https://tracxn.com/explore/EdTech-Startups-in-Indonesia (accessed March 12, 2021) 32 Wayan Purnomo. 2020. “Example of data fragmentation problems in Indonesia,” Tempo. Retrieved from: https://majalah. tempo.co/read/nasional/160237/mengapa-data-korban-covid-19-pemerintah-pusat-dan-daerah-berbeda (Accessed BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA November 16, 2020). 33 Weil, David. 2014. The Fissured Workplace. Why Work Became So Bad for So Many and What Can Be Done to Improve It. Harvard University Press, Cambridge, Massachusetts, London, England.    34 White, Olivia, Anu Madgavkar, James Manyika, Deepa Mahajan, Jacques Bughin, Mike McCarthy, and Owen Sperling. 2020. “Digital identification: A key to inclusive growth.” McKinsey Global Institute. Retrieved from: https://www.mckinsey. com/business-functions/mckinsey-digital/our-insights/digital-identification-a-key-to-inclusive-growth 35 World Bank. 2018. March 2018 Indonesia Economic Quarterly: Towards inclusive growth. Retrieved from: https://www. worldbank.org/en/country/indonesia/publication/indonesia-economic-quarterly-march-2018 (accessed November 16, 2020). 36 World Bank. 2019. World Development Report 2019. Washington DC: World Bank Group. Retrieved from: https://www. worldbank.org/en/publication/wdr2019 (accessed March 25, 2021). 37 World Bank. 2020a. Indonesia Economic Prospects: The Long Road to Recovery. Jakarta: World Bank. 38 World Bank. 2020a. “How Covid-19 is affecting firms in Indonesia: Results from the 1st round of the Covid-19 Business Pulse Survey.” Policy Brief. World Bank. 39 World Bank. 2020c. Indonesia Systematic Country Diagnostic: Eliminating Poverty, Bringing Economic Security to All. World Bank Washington, DC. 40 World Bank. 2020b. “Impacts on Digital Merchants: Insights from the Bukalapak-World Bank Survey.” Policy Brief. World Bank. 41 World Bank. 2021a. Macro Poverty Outlook. Washington DC: World Bank Group. Retrieved from: https://www.worldbank. org/en/publication/macro-poverty-outlook (accessed March 25, 2021). 69 Endnotes 1 World Bank (2019) 2 Google, Temasek and Bain (2020). The same report also shows that two segments that have been affected negatively are online travel and transport, and ride-hailing services. 3 Indonesia’s internet economy measured by this metric grew five-fold between 2015 and 2019—a pace unmatched by any other country in the region. The estimated size of this economy, at US$44 billion in 2020, is roughly four times as large as Malaysia and five times as large as the Philippines and Singapore. 4 World Bank (2020a). REFERENCES & ENDNOTES 5 Tiwari et al. (2020). 6 World Bank (2020b). 7 Goldfarb and Tucker (2019). 8 Minges (2016). 9 There are various other statistics on the overall level of connectivity in Indonesia. Most estimates from the industry tend to rely on mobile and fixed-line subscription data coming from GSMA and ITU. Although there is some effort made to identify multiple connections, multiple SIM cards, etc., arriving at this statistic from subscription data alone appears quite challenging. For example, the January 2020 edition of Hootsuite reports internet penetration at 64 percent of the population. The data used here and throughout the report are based on what is perhaps the most comprehensive socioeconomic survey Indonesia conducts on an annual basis, interviewing around 300,000 households across the country BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA for each round. 10 This is a decision open to the MNOs, as all IMT spectrum is now technology neutral. Indonesia’s neighbors Malaysia and Singapore will use the 700 MHz band for 5G services. 11 Refer to www.gsma.com/spectrum/resources/legacy-mobile-network-rationalisation/ 12 Refer to https:/ /vietnamnet.vn/en/sci-tech-environment/no-new-2g-3g-phones-in-vietnam-from-july-1-708163.html and https://mic.gov.vn/Upload_Moi/VanBan/43TT.PDF (in Vietnamese). 13 Hootsuite We Are Social (2019). 14 Google, Temasek and Bain (2020) “e-Conomy SEA 2020”. 15 Goolsbee and Klenow (2006) and Brynjolfsson and Oh (2012). 16 The methodology essentially entails using the time value of leisure to estimate the opportunity cost and value of internet use. We use data from Indonesia’s labor force survey to estimate the earnings function and the time spent on the internet from the Digital Economy Household Survey conducted for this report. 17 Ride hailing suffered in the early days of the pandemic, but as social distancing also intensified e-commerce and food delivery services, some of the “ride-partners” in these platforms adapted by switching to delivery. 18 Google, Temasek and Bain (2020). 19 See Brynjolfsson et al. (2003) and Dolfen et al. (2017), for example. 20 Hjort and Poulsen (2018); Bahia et al. (2020). 21 Income-earners include wage employees, casual workers and self-employers whose income data are available. Income data are not available for employers with workers and unpaid workers. 22 Note that this is just the average increase for all workers for that type. The actual increase based on other worker characteristics, such as the industry, occupation, and whether the worker herself also uses the internet at work, could be much higher. 23 World Bank (2020a). The first round of the World Bank COVID-19 Business Pulse Survey (COV-BPS) was conducted in June 15 to 23 with phone interviews on a nationally representative sample of 850 formal sector firms. 24 World Bank (2020b). This is based on data from the first round of the World Bank-Bukalapak Digital Merchants Survey conducted between May 20 and June 27, 2020, with a total sample of 1,020 respondents. 25 Google, Temasek and Bain (2020). 26 The 2018 number is based on the national labor force survey of that year, while the 2020 number is based on the Digital Economy Household Survey, a special purpose survey conducted specifically for this report during February and March 2020. 27 SAKERNAS (2019). 28 Around 12.2 million workers alone were engaged in e-commerce activities as their primary job (SAKERNAS 2019). However, Indonesia’s National Social-Economic Survey 2019 (SUSENAS 2019) shows only 5.9 million online sellers (SUSENAS, March 2019). 29 Alibaba Group and World Bank (2019). 70 30 SAKERNAS (2019). 31 World Bank (2020c). 32 LSPs include transportation companies (shipping lines, ferry lines, airlines, trucking companies), as well as third-party logistics companies that coordinate multiple transportation modes, together with warehousing and other value-added activities such as packaging and sorting. 33 There are other measures of financial inclusion, including in Indonesia. These numbers from the World Bank Digital Economy Household Survey appear to align well with these numbers. For example, the Global Findex Database, 2017 found REFERENCES & ENDNOTES that 49 percent of Indonesian adults had transaction accounts. Similarly, the Financial Inclusion Insight (FII) Survey in 2018 found that 56 percent adults were owners of a bank account. Likewise, the Global Findex Database also showed that 3 percent of Indonesian adults owned mobile money accounts and there were at least 7 percent of adults who used some form of mobile banking. The numbers from the DEHS—which are based on households—appear to be in a reasonable ballpark of these numbers and, as such, provides a useful updated triangulation on the picture of financial access in Indonesia. 34 Pazarbasioglu et al. (2020). 35 Under this new regulation, foreign-produced textiles, clothes, bags, and shoes that cost a minimum of US$3 will be subject to a range of taxes and import duties with a total rate of 32.5 to 50 percent of their value. For other products worth at least US$3, the taxes and import duties will be lowered from 27.5 to 37.5 percent of their value to 17.5 percent. Goods worth below US$3 will be exempted from import duties but still be subject to some other taxes, such as value-added tax. 36 Australia was the first to slash its GST threshold from AUD1,000 to zero from July 2018. The EU is set to follow suit, with BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA its current exemption of consignments of less than euro 22 to be abolished from 2021 onwards. Indonesia is among the early movers in this space. But implementing these new rules without overwhelming the customs administration or creating unnecessary burdens on business will also be challenging and require effective risk management. 37 Existing roadmaps on e-commerce recognize the need to promote local products, catered both to the domestic as well as overseas market. However, currently these efforts are limited to co-branded promotion campaigns by the Government in partnerships with e-marketplaces to encourage Indonesians to buy local. 38 For example, data from the Digital Economy Household Survey show that 77 percent of app-based food deliveries in Indonesia are found to be substituting physical restaurant experiences. This was a survey that was done before the onset of the pandemic in March 2020. 39 Indonesia’s VAT threshold of IDR 4.8 billion stands out in international comparisons of VAT thresholds to GDP per capita, a metric commonly used to compare the generosity of VAT thresholds worldwide. Less-developed economies tend to have higher thresholds, in part because of weaknesses in tax administration and in part because they have larger informal economies. However, even when compared against low-income economies, Indonesia’s VAT ratio stands out as being too generous. A high threshold means that a vast share of businesses whose annual turnover is below the VAT threshold are excluded from the VAT system, narrowing the base and distorting the tax. 40 Frank et al. (2019), Banerjee and Duflo (2019). 41 Autor et al. (2017). See also Weil (2018), p.9: “Where lead companies once shared gains with their internal workforce, fissuring leads to growing inequality in how the value created in the economy is distributed.” And Banerjee and Duflo (2019) p.242: “The increase in concentration (among superstar firms) thus helps explains a part of why wages are not keeping pace with GDP. The rise of superstar firms also offers an explanation for why overall wage inequality has been rising…” 42 The human capital gap is magnified by restrictive government policy in hiring foreign workers, limiting the ability of employers to fill the skills gap with global talent in areas where local skills are in short supply. The recently promulgated Law on Job Creation (the “Omnibus Law”) includes a provision to relax this. 43 Tempo (2020). 44 Between April and May 2020, almost 80 percent of Posyandu (integrated maternal and child health and nutrition posts at the village level) were closed. Community-based outreach activities also reported significant disruptions with more than XXX percent ceasing operation. It is also reported that 86 percent of child growth monitoring activities, 55 percent of immunization services, 46 percent of Vitamin A distribution and ante-natal services were suspended or ceased due to these health posts closing down. 45 Bhardwaj, Yarrow, and Cali (2020). 46 Asosiasi Healthtech Indonesia (2020) 47 The most updated data on health-tech funding was prepared in September 2019 by Tracxn. (https:/ /tracxn.com/explore/ HealthTech-Startups-in-Indonesia). Since then, there has not been any major announcement of health-tech funding, except for Alodokter’s follow-on funding by MDI Ventures in November ’20 (https:/ /theinsiderstories.com/indonesias-alodokter- raises-series-c-funding-from-mdi-ventures/) . In October 2019, Grab and Ping An announced a joint venture to launch Grab Health / Good Doctors Indonesia, but the value of the investment to Indonesia was never disclosed. 71 48 Katadata (2020a). 49 Katadata (2020b). 50 World Bank conducted mobile phone based high frequency survey of households to track the evolving impact of the COVID-19 pandemic. 51 Education experts and practitioners argue that online learning cannot replace offline face-to-face learning. To avail of it one requires sufficient means and technology, which can be exclusive in some countries, and there are psychological and social REFERENCES & ENDNOTES factors that prevent online interactions from being as effective as offline (Dhawan 2020). 52 The experience of massive-open-online-courses (MOOCs) provides one illustrative example. Once regarded as one of the groundbreaking educational innovations of the past decade, they have had very limited success. Completion rates have remained below 5 percent for many years now, with lack of coordination and direction considered to be among the main contributing factors (Kop, 2020). 53 Indeed, the Government has been experimenting with several of these efforts in both education and health. Portal Rumah Belajar is an online learning platform equipped with complete multimedia education content managed by MoEC. Universitas Terbuka, a state university focused on distance learning also offers extensive open education resources. Similarly, in health, Sehatpedia is a Government tele-consultation platform. 54 McKinsey (2020). 55 Aadhar, India’s digital ID system, contributed to significantly expanding financial inclusion (from 35 percent in 2011 to 80 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA percent in 2017), by making it cheaper and easier for people to satisfy know-your-customer (KYC) requirements for bank account opening. 56 Although the law does not significantly affect One Data, it will provide guidance to data stewards in particular when managing personal information. Bappenas as the coordinator of One-Data implementation is expected to monitor closely the development and incorporate elements of the Personal Data Protection Law in the One-Data technical implementation guidance. 57 Although there is no specific mention of One Data in the Omnibus Law, there are several provisions regarding data that are relevant. For example, the decision on wages for MSMEs is mandated to be based on consumption aggregates calculated by BPS. This means that data published by one GoI agency, BPS, will be used to set wages. Lack of clarity on due process and transparency in data production may weaken the legitimacy of official statistics and leave them vulnerable to being politicized. 58 For example, institutions such as the Directorate General of Civil Registry of MoHA and the Directorate General of Tax of MoF traditionally control and establish relatively more mature data management for civil registry and taxpayer datasets. 59 ITU (2019). 60 Global Mobile Suppliers Association (2021). 61 ASEAN and Plum Consulting (2021). 62 ASEAN (2021). 63 World Bank (2019). 64 World Economic Forum (2020). 65 World Bank (2019) 66 Ibid 67 Ibid 68 The indirect revenue gains are potentially much greater, as data from the digital economy can be combined with other third- party data and used by the revenue authority, the Directorate General of Taxes (DGT), to boost compliance and raise higher revenues across Indonesia’s main taxes, i.e., value-added tax (VAT), corporate income tax (CIT) and personal income tax (PIT). 69 World Bank (2021a). 70 For some of the companies covered, see: “Indonesia adds Twitter, Zoom to tech companies that must pay 10% VAT”, Reuters (September 8, 2020), accessed online at: https://www.reuters.com/article/us-indonesia-tax-digital-idUSKBN25Z2CU 72 71 The newly appointed VAT Collectors are provided with a VAT Collector ID and a tax registration letter. These VAT Collector are required to create VAT collection slips providing information on VAT collection and payment, which can take the form of a commercial invoice, billing, or order receipt to ease the burden of complying. VAT payment is made electronically using a billing code that is provided by the DGT, and can be made using IDR, USD dollar, or all other foreign currencies accepted by the DGT system. VAT Collectors are required to file quarterly reports, with at minimum data on: (i) number of users in Indonesia; (ii) amount of payments (excluding the VAT); (iii) amount of VAT collected; and (iv) amount of VAT settled to the government. The DGT is permitted to request further detailed filings on an annual basis with transaction-level data. REFERENCES & ENDNOTES 72 Australian Taxation Office (2019). 73 Council of the European Union (2017). 74 European Commission (2021). 75 The large number of ‘informal’ MSMEs, not registered with government entities, and/or entrepreneurial households increasingly active on e-commerce represent significant challenges in the Indonesia context. Data from digital platforms can be a very powerful enabler of enhanced risk management when combined with customs data. Indonesia’s Directorate General of Customs is looking to do this, with an ongoing pilot involving several digital platforms focused on using e-commerce transactions data to tackle compliance risks including under-invoicing and missed declarations. 76 Revenue authorities with limited capacity to handle large numbers of monthly VAT filings (e.g., because they remain reliant on manual filing, and/or have limited ICT capacities to manage VAT e-filing nationwide) may set a higher VAT threshold so that fewer companies would need to register and file. An informal economy may be defined as one with characterized by firms BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA that are not registered, and/or whose employees work without a formal contract. Informal companies have been traditionally hard to tax, since they are often ‘hidden’ from government reach. 77 “Box B.1: Why is Indonesia’s tax-to-GDP ratio so low?” in The Indonesia Economic Quarterly: Towards Inclusive Growth” (March 2018), pp. 50. 78 Al-Rikabi, Khalil Rohman, and Hambali, “On the Macroeconomics of the Digital Economy in Indonesia”, Working Paper, World Bank (forthcoming, 2021). 79 The group of leading digital nationals: https://leadingdigitalgovs.org/ 73 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA CHAPTER — 01 74 Chapter ⬎ 1 for Inclusion Current Landscape →The Promise and the Digital Technologies 75 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA CHAPTER — 01 1.1. Introduction CHAPTER — 01 S imilar to many other coun- the largest and the fastest growing digital economy in the tries around the world, the sub-region. The behavioral change that has accompanied COVID-19 pandemic has this accelerated adoption of digital technologies can be hit Indonesia hard. The expected to be permanent. Estimates from industry in- global slowdown in eco- dicate that 93 percent of new adoptees expect to stick by nomic activity combined at least one of the new digital services post-COVID-19. with domestic measures to If true, this would imply a major boost to Indonesia’s in- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA contain the spread of the ternet economy which, at an estimated US$44 billion in virus dragged down the gross merchandise value, is not only the largest in South economy, which is estimat- East Asia, but also the fastest growing.2 ed to have shrunk by around 2.2 percent in 2020. Latest estimates suggest that an estimated 5.1 million people— At the same time, the pandemic has also put a spotlight equivalent to 2.4 percent of the working age population— on Indonesia’s inequalities. One enduring challenge of lost their jobs, while an additional 24 million have had to Indonesia’s development experience over the past two work reduced hours due to the pandemic. As many as 50 decades has been making growth more inclusive. While percent of workers experienced a reduction in earnings. remarkable progress has been made on poverty reduction, The impact on living standards has been devastating, with persistent welfare disparities remain across regions, and more than 2.2 million Indonesians estimated to have been growth in living standards experienced by those in the pushed into COVID-19-induced poverty in 2020. bottom 40 percent of the income distribution has per- sistently lagged average growth.3 By disproportionately One unexpected silver lining of the crisis, however, has hitting the poor and the vulnerable harder, the pandemic been the turbo-charged adoption of digital technologies. threatens to widen these inequalities. On the one hand, Businesses small and large have flocked to digital technolo- susceptibility to infection and burden of disease for the gies to ensure the continuity of their operations. Adoption poor and vulnerable in directly affected urban areas have of e-commerce has soared, with online purchases proving been more intense than among the non-poor, due to risk a viable way of securing uninterrupted access to essential factors such as poorer diets, lower access to quality health consumer goods, including basic groceries. School closures services, higher prevalence of smoking, poorer hygiene have forced students and teachers to adapt and explore and sanitation practices, and inability to afford inputs to digitally enabled remote learning options, including by preventive behaviors (e.g., masks, hand sanitizers, etc.). adopting a variety of EdTech solutions. HealthTech apps On the other hand, sectors and forms of work that have enabling remote consultations and the delivery of medi- been most affected are also the ones in which the poorer cine have seen unprecedented growth in adoption rates. and less skilled segments of the population are more likely Confined at home due to mobility restrictions, people have to be in. As a result, while the pandemic has been affected switched to the internet for their entertainment and social everyone, income and welfare losses at the bottom parts needs, driving sharp growth in the usage of digital media of the income distribution have been the highest.4 (music and video streaming) and communications appli- cations. A recent industry estimate shows that 37 percent Differential access to and adoption of digital technologies of all digital service consumers in Indonesia in 2020 were could compound these inequalities. Digital solutions to new COVID-19-induced users and more than half (56 achieve resilience during this crisis have not been available percent) were from conventionally weaker markets such to all. For example, in comparison to 88 percent of gov- as non-metro areas.1 ernment jobs and 40 percent of jobs in manufacturing, less than 10 percent of jobs in hotels and restaurants, con- This digital pivot is expected to be permanent and, as struction, wholesale or agriculture were found amenable such, promises to provide a major boost to what is already to be done remotely. Similarly, just 11 to 12 percent of 76 CHAPTER — 01 casual and own-account work could be done performed were already linked to bank and mobile money accounts, from home (compared with 42 percent of wage work). there was added assurance that the cash transfers would While 84 percent of high-skilled work could be done from reach their intended beneficiaries. In a few weeks, more home, 85 percent of work done by low-skilled workers than 28 million citizens applied with 15 million deemed required physical presence in their place of work.5 to be eligible. Payments began for some within days of applying. In Chile and Peru, social registries were used to Similarly, as school closures pushed students toward re- expand existing social assistance programs and citizens BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA mote learning, around 60 percent of students were found could use their digital ID to access a portal that allowed to be unable to participate in online classes due to lack of them to check if they were included. In contrast, even reliable internet and limited ownership of internet-en- though Indonesia introduced significant fiscal measures, abled devices. A remedial measure was introduced in the there were critical delays getting a couple of similar pro- form of an IDR 7.2 trillion data package for students and grams off the ground on account of challenges related teachers.6 But, given how costly mobile broadband data is to updating of databases, and verification of identities. for high bandwidth applications such as video streaming, this is likely to be a stop-gap solution to the real problem: Looking ahead, one key question that policy makers in only 10 percent of Indonesian public schools are con- Indonesia are facing is how to ride the momentum of nected to fixed broadband internet and about half of the digital adoption generated by the pandemic, not only population is still not connected to any kind of internet. to power the recovery in the short term but also to bring about greater inclusion and resilience in the economy The pandemic has also laid bare the Government’s own in the medium term. The crisis is clearly far from over, lack of readiness, especially with digitized, integrated, and but some signs of economic recovery are beginning to interoperable data systems in achieving resilience. With appear on the horizon. As the economy rebounds and quarantines and the scale-down or shutdown of many gov- adjusts gradually to the new post-pandemic equilibrium, ernment and commercial operations, countries had to try questions about ways in which Indonesia could leverage to rapidly shift services to digital channels to keep them go- digital technologies to better prepare for similar shocks in ing. Countries that had invested, for instance, in and built the future and address some of its long-term development strong digital ID systems before the pandemic were by and challenges have become highly pertinent. large better able to do this than those countries that had not. Estonia, Singapore, the United Kingdom and other This report conducts an in-depth diagnostic of digital digitally advanced countries were able to reduce disruption technologies, and the scale and extent of their current for their public service delivery. But the difference was par- applications in Indonesia. Fully exploiting Indonesia’s ticularly telling in terms of whether countries were able to rich data landscape, new survey data collected specifically deploy quick, accurate and effective social protection mea- for this report, as well as data shared with the World Bank sures to mitigate the shocks and to safeguard livelihoods, in confidence by some of the major digital platforms in including to reach populations such as informal workers the country, this report: (i) investigates the key challenges who were difficult to target because they were not present that Indonesia faces in maximizing its digital dividends in either databases of people living in poverty or databases in an inclusive manner; and (ii) identifies some action- of people contributing to social security. able entry points for investments and policies. The rest of this introductory chapter: (i) describes the mechanisms Thailand, for example, built a website to allow informal through which digital technologies can boost inclusion workers to apply for emergency payments online, using its and focuses on three main pathways along which the re- national ID system to verify the identity of applicants, es- port is organized; and (ii) presents some stylized facts tablish their uniqueness and determine their eligibility by on the current state of digital technology applications in cross-checking government databases. Since national IDs Indonesia to set the stage for the report. 77 1.2. The Promise and an Organizing Framework To delineate the outer boundary of the topics that this transportation, tracking and verification.10 Lower costs of report covers, it is important to first define what we mean finding and comparing information can help reduce price by the terms “digital economy” and “digital technologies”. dispersion, improve matching in financial, labor and retail CHAPTER — 01 Digital economy, an economy powered by digital technol- markets and help deepen these markets. The non-rival ogies, is often defined either too narrowly or too broadly. nature of information (one person’s consumption does Narrow definitions tend to equate the digital economy not lower the amount available for by consumption of to the “technology sector” or the “ICT sector”, missing another) encoded digitally lowers the replication cost; entirely the broader applications of these technologies once the upfront cost of developing an application has in other traditional sectors. One estimate, for example, been incurred and the product tested, it is costless to rep- shows that while the ICT sector value-added in GDP is licate it as many times as needed. Digital technologies around 7 percent up to 75 percent of the economic impact also render moot place-based constraints on economic BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA of the internet—the key enabler of the technology sec- activities. Near costless transmission of information and tor—comes through conventional firms harnessing this digital services help unlock opportunities for buyers and technology to improve productivity.7 On the other hand, sellers anywhere in the country, leading to both the deep- loose definitions tend to equate the digital economy with ening, as well as stronger integration, of markets. Lower fast-growing technology companies, such as Facebook, tracking costs, or costs of connecting individuals with Google or Amazon, and in the context of Indonesia, uni- information about them facilitates the design and delivery corns such as Go-Jek, Grab and Tokopedia. of services better customized to personal needs (e.g., G2P payments, credit provided based on information on con- This report adopts the definition of digital technologies sumer behavior). Finally, by lowering verification costs, as technologies that enable the “representation of infor- digital technologies can enable individuals and firms to mation in bits and significantly reduce the cost of storage, build up reputation and trust and participate in markets computation and transmission of data”.8 All applications assured of reasonable contestability. of this technology that mediate how individuals interact with marketplaces, interact with the Government and in- Evidence, primarily from high-income countries, suggests teract with themselves are relevant. This definition encap- that the efficiency gains brought about by the adoption of sulates broader applications of the internet and the ICT these technologies can boost aggregate productivity and sector, such as social media and e-commerce, and includes growth.11 What is of interest for this report, however, is the adoption of these technologies by the Government to how these aggregate benefits accrue to people at different strengthen accountability and service delivery. Given our parts of the income distribution, to workers who might focus on the low level of digitization in industrial activ- be traditionally excluded, especially women and youth, ities, especially manufacturing, higher-end applications workers at the lower end of the distribution of skills, and to of digital technologies, such as automation, robotics, IoT citizens in different parts of the country. The main channels etc., that are drivers of IR4.0 are not topics of central through which these benefits could accrue to people would focus for the report.9 Clearly, enabling Indonesian firms be through: (i) enhanced labor income, depending on how to innovate and adopt these technologies is extremely im- digital technologies alter the labor market landscape ex- portant for the country to become more productive and panding the economic opportunities in the form of more grow at a higher speed. These technologies can also have and better paying jobs; (ii) capital income, for micro-scale profound impact on inequality in the future by, among entrepreneurs who may be able to capitalize on opportu- others, fundamentally altering the labor market land- nities to innovate and grow; (iii) consumer surplus that scape. However, in this report we focus to a greater extent could be realized through lower prices, greater variety of on lower-skilled biased applications of these technologies consumer goods, etc.; and (iv) stronger delivery of public as they are currently being experienced by a significantly services, such as education, health and social protection larger segment of the population. programs, which could be enhanced by greater account- ability, as well as improved government effectiveness in What promise do digital technologies hold? Most of the generating tax and non-tax revenues to fund these pro- benefits of these technologies come in the form of a re- grams, better identification of targeted beneficiaries and duction in five key economic costs: search, replication, a wholesale digital transformation of government. 78 The report is fundamentally about how digital technologies scale); the larger the number of users, the greater the value touch, shape and influence the economic and social lives of of the product for the users. This positive feedback loop the people. As such, the diagnostic, as well as policy, thrust generates advantages for the first mover, according signifi- of the report is organized around three main dimensions cant market power to the incumbent and making it difficult along which the risks of ‘digital exclusion’ are the most pro- even for more productive newcomers to enter the market. nounced: (i) access to the medium (digital technologies); As the market begins to mature, this concentration of some (ii) ability to harness the medium in a private capacity to firms within an industry can slow down innovation, hurt enhance incomes; and (iii) ability to benefit from public consumers and lead to the redistribution of rents to the sector adoption of the medium to provide services more platforms themselves. Second, without strong laws on per- effectively and efficiently (Figure 1.1). sonal data and consumer protection, lowering of tracking and verification costs could lead to price discrimination, Digital technologies pose significant downside risks as well. fraud, and violation of privacy. A discussion of these risks CHAPTER — 01 Most applications of these technologies are often propelled and the measures that need to be taken to address these by strong network effects (or demand-side economies of appropriately is included throughout the report. FIGURE 1.1 Harnessing digital technologies for greater inclusion – three pathways BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA 1 Expand access to the medium 2 Enhance citizen capabilities to harness the medium to boost incomes 3 Accelerate government adoption of the medium to enhance services 79 1.3. Digital Technology Landscape– What Is the State of Play? To set the stage for the report, we start out first by de- FIGURE 1.2 scribing the current state of play on the adoption of digital technologies in Indonesia. How are these technologies currently touching the lives of the people in Indonesia? Access to the internet has been increasing over the past decade CHAPTER — 01 Who is connected and who is not? How are these technol- ogies shaping economic opportunities for those who have been able to plug in? What benefits are accruing to the PROPORTION OF INDIVIDUALS 15+ WITH ACCESS TO broader society? This section provides a few key stylized THE INTERNET facts on what we know about digital technologies, the level of adoption of these technologies and their imprint on Indonesian economy and society. BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA → Indonesia has made rapid progress in internet connectivity over the past decade, but half of the adult population is still without access Over the past decade, Indonesia has sustained steady growth in internet connectivity driven primarily by the rapid investment in network infrastructure by the private sector. The share of the adult population connected to the LEGEND internet increased almost four-fold from 13 percent of the adult population in 2011 to 51 percent in 2019.12 This im- Urban National Rural pressive growth notwithstanding, 49 percent of Indone- sian adults are still not connected to the internet. and a Source: World Bank staff calculations based on various years of SUSENAS. significant digital divide persists across various spatial, eco- Note: Connection to the internet is defined based on whether nomic and social dimensions. First, the urban-rural divide adults reported having access to the internet in the past three months (including accessing social media apps such as Facebook, in connectivity is large and appears to have been increasing YouTube, Instagram, Twitter, and WhatsApp). over the years. In 2019, 62 percent of Indonesian adults in urban areas were connected compared with just 36 percent the same time, the fact that almost half of the population, in rural areas. The urban and rural internet connectivity even in regions with relatively better infrastructure, re- was 20 and 6 percent, respectively, in 2011 (Figure 1.2). mains without internet points to major challenges in the middle and last mile connectivity segments. The Government of Indonesia has made major efforts to close the digital divide, most notably with the implemen- Income gaps in access are similarly large. Adults in families tation of the Palapa Ring project, which aimed to extend in the top decile of the income distribution are more than the fiber-optic backbone infrastructure of the country to five times more likely to be connected to the internet than the outer islands in eastern Indonesia. With the comple- adults in the poorest decile, only 14 percent of whom are tion of the Palapa Ring project in 2019, all of Indonesia’s connected (Figure 1.4). This sharp income gradient points 514 kota/kabupaten (cities/districts) are now connected to to a possible affordability constraint in accessing the inter- the national backbone. This has led to a notable increase net. Likewise, there is also a sharp generational, education in the proportion of adults connected to the internet in and gender divide in access to the internet. Younger adults all major island-regions of the country. But there are still are significantly more likely to be connected, as are adults sizeable gaps across regions. For example, only about one- who are more educated. Men are 8 percentage points more third of the adult population in Papua is connected com- likely to be connected than women, suggesting possible in- pared with around 55 percent in Java-Bali (Figure 1.3). At equality in device ownership within households (Figure 1.5). 80 FIGURE 1.3 FIGURE 1.4 Some parts of the country are still Richer Indonesian’s have better lagging access PROPORTION OF ADULTS WITH ACCESS TO THE INTERNET, PROPORTION OF ADULTS WITH ACCESS TO THE INTERNET BY ISLAND REGIONS IN 2011 AND 2019 IN 2019 CHAPTER — 01 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Source: World Bank staff calculations based on various years of SUSENAS. FIGURE 1.5 Internet access is also higher among the younger, more educated and male segments of the population PROPORTION OF ADULTS WITH ACCESS TO THE INTERNET BY AGE GROUP BY EDUCATIONAL ATTAINMENT BY GENDER Source: Source: World Bank staff calculations based on various years of SUSENAS. Note: Connection to the internet is defined based on whether adults reported having access to the internet in the past three months (including accessing social media apps such as Facebook, YouTube, Instagram, Twitter, and WhatsApp). 81 → Mobile phones are FIGURE 1.6 the predominant mode of internet connectivity Only 5 percent of population is connected to fixed broadband at home in Indonesia; very few Indonesians are connected PROPORTION OF POPULATION WITH FIXED BROADBAND CONNECTION AT HOME through fixed broadband at home Most Indonesians connect to the internet using CHAPTER — 01 mobile broadband and growth in mobile broad- band infrastructure is what has powered the in- crease in internet connectivity over the past de- cade. Only about 5 percent of the population is connected to fixed broadband at home (Figure 1.6). This is reflected also in the sharp increase in the ownership of 3G/4G enabled handsets and the corresponding decline in the ownership BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA of 2G handsets. Ownership of ICT devices such as computers that are common mediums for the use of fixed broadband internet at home have remained surprisingly level at 20 percent over the past decade (Figure 1.7). So, while the wide availability of mobile broadband has expanded access, it has probably not unlocked productive Source: World Bank staff calculations based on various years of SUSENAS. capabilities to the same extent, as it does not deliver the same capacity, and quality of ser- vice, and is not cost-efficient enough for high bandwidth applications as fixed broadband. FIGURE 1.7 The pandemic has laid this bare as millions of students and teachers across Indonesia found Smartphone ownership rates have increased their mobile data plans inadequate and cost- sharply over the years ly when it came to deal with high bandwidths needed for remote learning arrangements. The ICT DEVICE OWNERSHIP RATE challenges associated with universalizing access to affordable and high-quality internet are dis- cussed in Chapter 2. → Indonesians who are connected to the internet use it quite intensively, with communication, social media and leisure applications dominating usage On average, Indonesians spend around 6 hours Source: World Bank staff calculations using data from SUSENAS. online, with younger and more educated seg- Note: 3G/4G connection share is estimated as share of adults that access the internet ments more digitally engaged than the rela- using their phone, among adults with cellular phone ownership. 2G connection share is estimated as share of adults with cellular phone ownership that not accessing the tively older and less educated demographic.13 internet via their cellular phone. The 2018 figure is interpolated value. 82 FIGURE 1.8 Intensity of internet engagement is highest for the 16 to 25 age group, which on average spends as much as 9.7 hours a day online. There are no Indonesians who are connected to the internet significant differences between men and women use it quite intensively on the intensity of internet use (Figure 1.8). Among the various online activities, communi- HOURS PER DAY SPENT ONLINE cation, social media, and leisure dominate usage (Figure 1.9). According to estimates from the industry, Indonesia is the fifth most internet BY GENDER engaged country in the world, behind the Phil- ippines, Brazil, Thailand and Colombia. The average intensity of internet use in Indonesia CHAPTER — 01 is 28 percent above global average, making it BY AGE GROUP an attractive market for content developers and advertisers.14 Another estimate from the industry suggests that daily time spent on the internet in Indonesia could have spiked by 31 percent during the peak of the pandemic and leveled off at around 20 percent of pre-pan- demic levels more recently, suggesting that as BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Indonesia emerges from the crisis its popula- BY EDUCATIONAL ATTAINMENT tion could have become even more intensively engaged on the internet.15 The social media result is also not surprising. Being an internet user in Indonesia is almost synonymous with being a social media user. Source: World Bank Digital Economy Household Survey, 2020. Over 85 percent of internet users were also us- ers of social media with the most popular plat- forms being WhatsApp, Facebook and Insta- FIGURE 1.9 gram. WhatsApp appears to be used principally for communication and sharing information, while on other platforms such as Facebook, In- Communications, social media and leisure stagram and Twitter, users are active in seek- activities account for 80 percent of average ing out news and information. Social media time spent online platforms are also important for commerce in Indonesia, with a sizeable proportion of users SHARE OF TIME SPENT ONLINE (about 20 percent) using Facebook, Instagram and WhatsApp for buying and selling. The 36% 21% 21% most common topics of discussions across all platforms relate to hobbies and lifestyles. How- ever, religion, public policy and politics are also widely discussed, highlighting the importance of social media as a channel of communication Leisure and influence on these topics (see Spotlight 1 for more details on the patterns of social media 11% 7% use and some potential downside risks associ- ated with it). Others These free digital services generate sizeable Communications Social Media Browsing Buying & Selling 3% consumer surplus for users. A key question that is often asked is whether the consumption of these largely free digital services generates Source: World Bank Digital Economy Household Survey, 2020. any value to the users or are the hours spent Note: Communication includes time spent on online texting & email. Leisure consists of time spent streaming content (audio and video) as well as playing online games. browsing pictures of friends on Instagram and Social media includes time spent on Twitter, Facebook , and Instagram. Others consist of e-travel reservation, e-gig-jobs, e-finance, ordering ride- hailing for watching cat videos on YouTube simply time transportation and food, and creating or downloading digital content. away from other productive endeavors. If there 83 is value generated, it is not clear how this value FIGURE 1.10 would be measured. Therefore, these benefits are often also not captured in statistics on na- One-quarter of Indonesians use ride-hailing tional accounts. Yet, the fact that so many con- apps, and almost one-fifth use it for their sumers voluntarily choose to spend so much time consuming these services must imply that daily commutes there is some consumer surplus they derive RIDE-HAILING APPS USAGE from it. There have been some efforts, primarily in the United States, to try and estimate the val- ue of this consumer surplus and depending on the methodology used, the numbers range from 3 percent of income to 25 percent of income in CHAPTER — 01 one estimate and close to US$100 billion over the 2007–11 period in another.16 Using one of the approaches from the literature, we estimate the consumer surplus of free internet services in Indonesia to average around 19 to 21 percent of per capita income for users.17 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA → Digital ride hailing is commonly used and not only provides REASONS FOR USE mobility services to connect the otherwise fragmented labor markets in Indonesia’s large and sprawling metropolitan areas, but also offers several other conveniences Digital ride hailing services provided by compa- nies, such as Go-Jek—Indonesia’s homegrown decacorn that epitomizes the country’s digital potential for policy makers—and its regional competitor Grab, are potentially among the WITHOUT RIDE-HAILING APPS I WOULD… most frequently experienced digital transactions for many Indonesians. The green-jackets worn by the motor-cycle taxi drivers of both compa- nies are a distinctive feature of almost all Indo- nesian cities. And this is not for no reason. While the pandemic and the associated decline in mo- bility has temporarily hit this segment hard, one-quarter of all Indonesians and around 36.5 percent of urban dwellers reported using the ride hailing service provided by these companies be- fore the pandemic (Figure 1.10). A striking 18.4 percent of users of these ride-hailing services used it for their daily commutes. Other reasons for use included circumstances when private vehicles could not be used (52 percent), a travel option for odd hours (41 percent) and to travel Source: World Bank Digital Economy Household Survey, 2020. 84 FIGURE 1.11 Strong complementarity with public transportation suggests that these services fill a crucial gap in urban mobility IMPACT OF THE OPENING OF JAKARTA MRT ON GO-JEK DROP-OFFS, CHANGE IN DROP-OFF VOLUMES AROUND THE EVENT CHAPTER — 01 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Source: Bosker, Roberts, Tiwari, Wihardja and Wibisana, 2021 (forthcoming). Note: The plotted values are coefficients of an event study model that looks at the impact of the opening of the Jakarta MRT on ridership volume. The MRT locations are treated locations while a number of other locations with similar characteristics: (i) between two adjacent stations; and (ii) beyond the current MRT line where an extension is planned for the future are taken as control locations. The plotted figure shows that relative to these control locations, ridership volumes jump quite discontinuously in the treated locations after the event. to areas where public transportation services are not available to obtain the full picture on the social benefits of digital ride (26.6 percent). Without the availability of these digital ride hailing. One key question that underpins this discussion thus hailing options, journey times would be longer for 55 percent of is the interplay between these digital services and the available users, but travel cost higher for a negligible 1.9 percent of users. network of public transportation options. If these services complement—as opposed to substituting for—available public Recent work by the World Bank has shown that the lack of transportation options, then the cost of the congestion and within-city transportation options in Indonesia is one of the pollution externalities could potentially be counterbalanced primary reasons the country is struggling to realize the poten- by the positive externality on transit usership. tial of its cities to generate greater prosperity and inclusion.18 By providing commuting options for a sizeable segment of The opening of the Jakarta MRT in April 2019 provides a the population, helping reduce travel times and providing unique opportunity to investigate this question in one of the services in otherwise neglected parts of cities, these services largest and most mature markets for digital ride hailing in appear to be playing a useful function of connecting people Indonesia. If the availability of the public transit option ended to jobs. However, one potential negative externality of these up depressing ridership volumes around locations that even- services could be the congestion and pollution that is inevita- tually became MRT stations, then it could be inferred that bly associated with an increase in ridership volume.19 On the these services substitute for public transit. However, if rider- other hand, large public transportation systems such as metros ship volumes increase in these locations, then the two services in many developing countries are often not supported very likely complement each other. Relative to certain other parts of well by complementary feeder systems, potentially depressing the city that had characteristics similar to locations that were transit ridership. If by filling in this crucial “first mile” void, eventually chosen to become metro stations for this MRT, digital ride hailing services end up boosting transit ridership, Go-Jek pickups increased, peaked and eventually stabilized at then the improved social efficiency of public transit would be an average of 64 percent within three months of the opening a positive externality. From a public policy perspective, evalu- of the MRT (Figure 1.11). The impact on drop-offs similarly ating the magnitude of these externalities would be important was an overall increase of 58 percent over the same period. This 85 FIGURE 1.12 Digital ride-hailing has also brought other conveniences A SIZEABLE SHARE OF THE POPULATION PARTAKES IN …THOUGH DELIVERED FOOD IS A STRONGER SUBSTITUTE FOOD DELIVERED AT HOME… FOR RESTAURANT FOOD THAN HOME COOKED MEALS (% OF POPULATION WHO ORDERED FOOD VIA GO-FOOD OR GRAB-FOOD SERVICES) CHAPTER — 01 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Source: World Bank Digital Economy Household Survey, 2020. suggests that Go-Jek ride hailing services are strong comple- ments of the new MRT service that has been introduced and are probably playing an important role in filling the void on the first and last mile connectivity.20 In addition to ride-hailing, these companies have also been providing a whole host of other services through their apps. Chief among these is food delivery through their comple- mentary Go-Food and Grab-Food services. Survey results show that around 21 percent of the urban population ordered food using these apps (Figure 1.12). While home delivery of food is clearly a convenience that enhances utility for those who avail themselves of this service, the aggregate benefits could depend on the extent to which these services help ex- pand markets by increasing the proportion of population opting to dine out, for example. In Indonesia, only about 23 percent of those who order food using these services report a behavioral shift away from home-cooked meals (correspond- ing increase in restaurant food). An overwhelming majority (77 percent) appear to use deliveries simply as substitutes for physical restaurant experiences. Even if the overall mar- ket size has not increased by much, these platforms could help expand opportunities for specific vendors and warung that build a reputation within the system for consistently delivering high-quality products. The drivers or “ride-part- ners”, as they are regarded by the platforms, could also find delivery and ride services attractive employment options. To what extent are these platforms enhancing productivity of micro-scale enterprises? Are these digital ride and delivery services new jobs or existing ones in new form? How do they compare with other forms of informal work? These questions will be explored in depth in Chapter 3. 86 FIGURE 1.13 → Buying and selling online is growing but, E-commerce more prevalent in populous and affluent provinces though prevalent among only a small share of the PROPORTION OF HOUSEHOLDS WITH ACCESS TO INTERNET ENGAGED IN population, it is already E-COMMERCE, BY PROVINCE enhancing consumer welfare by providing cheaper options, greater product variety and convenience CHAPTER — 01 E-commerce is one of the largest components of Indonesia’s digital economy and has also been one of the major drivers of its growth during the pandemic. Industry estimates suggest that the gross merchandise value of e-commerce BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA in Indonesia is expected to have increased by 54 percent yoy from US$21 billion in 2019 to US$32 billion in 2020. This more than off- sets the decline in travel from US$10 billion to US$3 billion.21 While the industry is buoy- ant about the stickiness of pandemic-driven adoption from both the buyer and seller side driving growth during and post recovery, a sobering reality check on the transformative potential of this sector for the larger economy is the fact that, despite recent increases, the overall take-up rates still remain low. In 2019, the proportion of internet-using households that reported buying and selling online was 12.8 and 5.1 percent, respectively. As a share of the overall population, e-commerce engaged households (those who buy or sell) was only around 7 percent. A rough back-of-the-enve- lope calculation using industry estimates on the adoption of digital services during the pandem- ic and their anticipated stickiness suggests that as this could grow up to 10 to 11 percent of the population, as Indonesia begins entering the economic recovery phase. Moreover, there is significant heterogeneity, with e-commerce prevalence higher in more urbanized provinces and wealthier households. Large metropolitan provinces of Yogyakarta and Jakarta lead with over 20 percent of the inter- net-using population engaged in some form of buying and selling activity online, while the east- ern provinces of East Nusa Tenggara and Papua have more limited e-commerce reach (Figure 1.13). Urban areas, and especially metro cores, Source: World Bank staff calculations using data from SUSENAS 2019 Note: Households are considered engaged in e-commerce if they report buying or selling have higher e-commerce activity compared with online rural domains, which include rural peripheries 87 of larger cities. At the same time, there is also a FIGURE 1.14 distinct income gradient, with households in the top decile of the income distribution five times E-commerce prevalence also remains higher more likely to be engaged in e-commerce than among…. those at the bottom decile (Figure 1.14). Recognizing the potential that e-commerce holds to boost the productivity of micro and …HOUSEHOLDS IN URBAN AREAS (PROPORTION OF HOUSEHOLDS WITH THE small enterprises in smaller and more remote INTERNET ENGAGED IN E-COMMERCE ALONG THE URBAN GRADATION) parts of the country by expanding markets and helping achieve economies of scale, policy makers around the world have started getting CHAPTER — 01 interested in policies to push e-commerce into rural areas. China has taken this up as a national priority.22 The Ministry of Electronics and IT and India Post have been tasked with the job of pushing this agenda in India as part the ‘Digital India’ strategy.23 Other countries such as Viet- nam and Egypt have also prepared roadmaps and vision documents where this mission is ar- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA ticulated. Indonesia’s own e-commerce road- map (2017–2020) included a concerted effort to unlock the key enablers of e-commerce to grow the sector to US$130 billion in valuation by the end of 2020.24 What are the benefits of e-commerce and where is Indonesia in terms of realizing these benefits? The most robustly done analysis on the topic, especially in the context of developing coun- tries, has been done in China and it finds that …AND IN THE UPPER PARTS OF THE INCOME DISTRIBUTION (PROPORTION OF HOUSEHOLDS WITH THE INTERNET ENGAGED IN E-COMMERCE, BY there is little evidence of income gains to rural DECILES OF PER CAPITA CONSUMPTION) workers and producers. Instead, the paper finds most of the benefits of e-commerce to come from the consumption side: ‘gains from e-com- merce are driven by a reduction in local house- hold cost of living that is mainly due to the di- rect gains from access to the new e-commerce shopping option for local households. These gains are in the order of a 5 percent reduction in the cost of living for retail consumption among users, and about a 1 percent reduction for the average household living in these villages. For durable goods consumption, the estimated re- duction in the local cost of living is 17 percent among users and on average 3 percent among all households.’25 The data context does not permit a rigorous ex- amination of the impact of e-commerce on rural incomes in Indonesia, but there is some sugges- tive evidence that e-commerce might already be generating consumer surplus by lowering prices, providing greater product choice and convenience to buyers in Indonesia. Price and convenience are the most frequently cited rea- Source: World Bank staff calculations using data from SUSENAS 2019. Note: Households are considered engaged in e-commerce if they report buying or selling sons for online purchases (Figure 1.15). Look- online. 88 FIGURE 1.15 ing across regions, other factors also come up in specific geographies. For example, in Papua, local non-availability of products drives 40 per- Price and convenience are the most dominant cent of online purchases. In Bali, the reasons for reasons for buying online… buying online are roughly equally split among price, convenience and local non-availability of …FOR THE COUNTRY AS A WHOLE the goods that are purchased. In DKI Jakarta, however, online purchases are driven almost exclusively by price and convenience related factors. The literature suggests that, in addition to lower prices, being able to consume goods that were previously unavailable and being able CHAPTER — 01 to conveniently procure these goods generates sizeable welfare gains.26 In addition to making cheaper goods available to those who buy online, the penetration of e-commerce into a geography can have an im- pact also on the prices of similar goods sold of- fline, helping lower overall inflation for the ben- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA …AND, WITH SOME HETEROGENEITY, FOR THE MAIN ISLAND REGIONS AS WELL efit of the larger community. This can happen because with cheaper options online, including from sellers in other parts of the country, buy- ers, especially in remote places, could essen- tially arbitrage away price differentials across geographies. Indeed, we find that price inflation of commodities likely to be traded relatively more intensively online was 0.8 of a percent- age point lower in high-e-commerce penetra- tion provinces relative to provinces with lower e-commerce penetration. A similar impact does not exist for commodities less likely to be trad- ed online (see Annex 1.A for details). So far, we have looked at a few specific appli- cations of digital technologies in the economy. Here we zoom out and ask a broader question: what is all of this adding up to? Has the in- LEGEND creased availability and use of digital technol- ogies begun to fundamentally alter the econom- ic landscape in Indonesia, especially the most lagging regions, in a way that it might start showing up in aggregate statistics? Greater availability and access to fast internet has been found in the literature to have a positive impact on broader employment outcomes, including increased (female) labor force participation and employment rates, net firm entries and improved productivity.27 There is also some Source: World Bank Digital Economy Household Survey, 2020. evidence from Nigeria that suggests that mobile broadband coverage has had large and positive impacts on household consumption levels and → Digital technologies are beginning reduced the proportion of households below to bring about subtle shifts in the labor the poverty line. These effects were found to be at least partially due to an increase in labor market, but skilled workers appear to be force participation and employment, particu- better positioned to capitalize on the gains larly among women. 89 FIGURE 1.16 Fiber optic arrival in the Eastern provinces has boosted internet access, but has otherwise had a minimal influence on the underlying structure of the economy INTERNET ACCESS FEMALE LABOR FORCE PARTICIPATION CHAPTER — 01 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA YOUTH EMPLOYMENT NON-AGRICULTURAL EMPLOYMENT Source: Jacoby, Swarnata, Tiwari, Wihardja (2021) forthcoming Note: FE = district fixed effect; IV2SLS: 2SLS method proposed by Freyaldenhoven, Hansen, and Shapiro, 2019. Population density is a proxy of the confound variable. The closest lead of fiber optic deployment is the excluded instrument. A causal analysis of this kind must often contend with the po- structure increased internet access among the population, it tential confounding effect of reverse-causality: we would like translated into little else by way of broader labor market out- to examine the impact of internet on economic activity, but comes (Figure 1.16). Changes in non-agricultural employment, very often, the internet tends to be more prevalent precisely youth employment and female labor force participation rates in places that already hold greater economic promise. Using were practically indistinguishable before and after the arrival a panel event-study design and exploiting the staggered fiber of the internet. It is not shown here, but the results for other optic rollout in parts of Indonesia that were not already con- indicators such as total employment and total labor force par- nected in 2010, we examine the impact of increased internet ticipation rates are also largely similar. connectivity on several variables capturing the labor market characteristics of these places.28 Since large parts of Java and These results from the parts of the country that were largely Sumatra were already connected to the fiber optic backbone in the last to be connected to the fiber optic backbone may not 2010, the analysis is done on 33 percent of districts largely in necessarily generalize to the whole country. In fact, given that the lagging eastern parts of the country. The findings suggest a dominant share of Indonesia’s industrial activity is concen- that, while being connected to the national backbone infra- trated in Java-Bali, these results may in fact be understating the 90 TABLE 1.1 Greater internet access is benefiting higher-skilled workers more than the lower skilled ones RETURN TO EDUCATION W.R.T. ADDITIONAL AVERAGE RETURN OF 1 PERCENTAGE <6 YEARS OF SCHOOLING POINT INCREASE IN INTERNET PENETRATION PRIMARY SCHOOL 18.5% 0.0% LOWER SECONDARY SCHOOL 32.8% 0.1% HIGHER SECONDARY SCHOOL 49.7% 0.3% CHAPTER — 01 COLLEGE/UNIVERSITY DEGREE OR HIGHER 79.4% 0.6% NON-PRODUCTION WORKERS (RELATIVE TO PRODUCTION WORKERS) 12.7% 0.2% IN MANUFACTURING Source: Jacoby et al. forthcoming. Note: Reported marginal returns to internet are coefficients on the interactions between education levels of individual workers and average internet connectivity within a district in fixed-effect panel data models that also control for all relevant BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA individual characteristics, level of urbanization and economic development of the districts, as well as year effects to capture secular trends in returns to skills. A separate specification is used for the non-production worker result. The analysis uses data from Sakernas (Indonesia Labor Force Survey) (1990–2019), Susenas (Indonesia Socioeconomic Survey) (1990–2019) and the Medium and Large Manufacturing Survey (1995–2015). The unit of the analysis is individual worker (production/non-production worker by sector-district in the case of the Manufacturing Survey true impact. Close to 90 percent of jobs in manufacturing, These results suggest that the current pattern of digi- and 84 percent of all jobs in high-value or modern ser- tal technology diffusion is potentially widening, instead vices, for example, are in Java-Bali and Sumatra, while the of narrowing, development gap. They also foreshadow recently connected parts of the country remain more de- topics such as affordability of the internet, penetration pendent on agriculture and natural resources, etc. Indeed, of mobile vs. fixed broadband, the quality and reliability a national level analysis of the relationship between digital of mobile broadband, digital skills and skills required to technology penetration and wage premia confirms this. thrive in the digital economy that are discussed in some detail in the subsequent chapters. On average, over a 15-year period between 2005 and 2019, educated Indonesians have consistently enjoyed higher earnings relative to less educated Indonesians. For exam- ple, average earnings for the most educated group (those → Even though Indonesia with a college or university degree, or higher) have been punches slightly above its almost 80 percent higher than those for the uneducated group (less than six years of schooling) (Table 1.1).9 Ex- economic weight on the ploiting the spatial and temporal variation in the rollout digital adoption index of the of the internet at the district level over this period, we can Government, capability to examine the differential impact of internet penetration on these skill premia. While greater internet penetration provide digital public services benefits all types of workers relative to the uneducated is below regional peers and group, except the very low educated with only primary comparator countries education, the benefits are higher for the more skilled. The magnitude of the coefficients implies that, if internet penetration (measured by the proportion of individuals Digital technologies also present a great opportunity for having access to internet at home) increases by 10 per- the Government to transform the way it provides ser- centage points in a particular district, the earnings premi- vices and interacts with its citizens. These technologies um enjoyed by a college-educated worker will, on average, can unlock innovative solutions in services such social go up by 6 percentage points, while the earnings premium protection, education, health and other citizen needs, only went up by 3 percentage points for higher-secondary such as registration of life events, tax filing and payment, graduates and 1 percentage point for lower-secondary etc. How much has Indonesia taken advantage of these graduates (Table 1.1).10 opportunities? Based on an international benchmark- 91 ing of overall adoption of digital technologies by FIGURE 1.17 governments around the world, Indonesia appears to punch slightly above its economic weight (Fig- Indonesia punches slightly above its weight on ure 1.17). But in terms of providing some of these digital adoption index for the Government…. services digitally, Indonesia is well behind some of its regional peers such as Singapore, Malaysia, Vietnam and Thailand, and also behind countries (DAI GOVERNMENT SUB-INDEX) such as India and Mexico (Figure 1.18). There has been some increase in the supply of services provided by the Government using dig- ital technologies in the past decade. For example, CHAPTER — 01 most districts and all provinces in Indonesia now have a website that provides information on local investment opportunities, e-procurement, one- stop shops for local permits, and tourism potential. However, these websites are typically information- al and carry minimal interactivity. Big cities such as Jakarta, Surabaya, Bandung and Makassar have selectively adopted several smart city concepts, BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA such as the utilization of sensor networks (internet of things/IoT) and analytics to better manage their resources and infrastructure, and to become more responsive to their citizens’ needs. The Directorate General of Taxes launched e-filing and e-billing in 2014, which enabled taxpayers to submit their tax reports online. Payments to beneficiaries of social assistance programs such as the conditional cash transfer (Program Keluarga Harapan, PKH) Source: World Bank 2018. and food assistance (Bantuan Pangan Non Tunai, BPNT) have been digitized since 2017. FIGURE 1.18 But most of these fragmented and somewhat ad hoc initiatives at digitization have undermined …but the capability to provide digital public what ought to be a whole-of-government inte- services is below its regional peers grated approach. Meanwhile, the pandemic has exposed critical weaknesses as the Government has (ONLINE SERVICE INDEX 2018) struggled to mitigate disruptions of key services and roll out assistance to the affected population. The lack of digitized, integrated and interoper- able data systems, for example, made reaching households outside of the bottom 40 that are not covered in the country’s unified database diffi- cult. Specific programs such as Kartu Pra Kerja, that were repurposed to address the needs of this group, faced critical implementation delays due to issues related to identify verification. There are several significant challenges that need to be overcome for digital transformation of the Government to take off at scale. Some of the foundational elements of these challenges, such as integration of digital ID, cloud-first policy, government digital payment gateway, governance and institutional arrangements, are discussed in Chapter 4. Source: UN e-Government Survey 2018. 92 1.4. Conclusion CHAPTER — 01 This introductory chapter has described the overall mo- hold back a whole-of-government approach to digital tivation of this report, highlighting the various channels transformation. Finally, Chapter 5 presents a detailed and mechanisms through which digital technologies discussion of specific policy recommendations to address could be useful for Indonesia to achieve greater inclusion these challenges. and presenting an organizing framework to think about relevant policies. The chapter has also presented some In addition to these chapters, the report also includes stylized facts on the current state of adoption of digital four Spotlights that complement the overall narrative BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA technologies in Indonesia, emphasizing particularly the by presenting in-depth material on a selection of top- state of play on how these technologies and their various ics. Spotlight 1 is on the pattern of social media use and applications in the economy are touching and shaping some of the risks associated with it. Spotlight 2 presents the lives of Indonesians. Despite rapid progress achieved a summary of China’s e-commerce experience focusing in recent years on connecting its citizens, Indonesia still specifically on the rural revitalization aspects of the Tao- faces a gaping digital divide, with 49 percent of adults bao program. Spotlight 3 looks at innovations in digital still not connected to the internet. Specific digital tech- financial services in Indonesia and their potential to sup- nology applications such as e-commerce and digital ride port inclusion. Finally, Spotlight 4 presents a brief picture hailing have clearly begun to generate value in the form of of the EdTech and HealthTech landscape in the country. consumer surplus, not just to for those who use them but potentially also to others around them. But the level of adoption, and perhaps more importantly the geographic extent of these applications, have not reached levels that enable them to catalyze a more fundamental shift in the underlying structure of the economy and bring about transformative impact on the population. Recent expe- riences related to delays in getting some critical elements of pandemic response off the ground suggest that the po- tential for a broader digital transformation of government is largely unrealized. Overall, the stylized facts presented in the chapter suggest that the potential for digital technologies to bring about inclusion in Indonesia is largely unrealized in Indonesia. What constrains it and what can the Government do to eliminate these constraints? This question is the focus of the rest of the report. Chapter 2 examines the driv- ers of digital divide, looking at recent progress in digital infrastructure availability and identifying the barriers to making reliable and affordable internet accessible to all. Chapter 3 focuses on jobs and economic opportuni- ties that are currently being generated by applications of digital technologies in the labor market and diagnose the challenges that prevent these opportunities from be- ing available to more Indonesians. Chapter 4 looks at issues of digital adoption within the Government and identifies challenges related to digital ID, and factors that 93 Endnotes 1 Google, Temasek and Bain 2020. The same report also 16 Goolsbee and Klneow 2006; and Brynjolfsson and Oh shows that two sectors that have been affected nega- 2012. tively are the online travel and transport and ride hail- 17 The methodology essentially entails using the time val- ing services. ue of leisure to estimate the opportunity cost and val- 2 Indonesia’s internet economy measured by this met- ue of internet use. We use data from Indonesia’s labor ric grew five-fold between 2015 and 2019, a pace un- force to estimate the earnings function and the time matched by any other country in the region. The esti- spent on internet from the Digital Economy House- mated size of this economy of US$44 billion in 2020 is hold Survey conducted for this report. roughly four times as large as Malaysia and five times as CHAPTER — 01 18 Roberts, Gil Sander and Tiwari 2019. large as the Philippines and Singapore. 19 Though evidence in Figure 1.10 suggests that this 3 World Bank 2020a. may be less of an issue in Indonesia with significant 4 Tiwari et al. 2020. proportion of users of digital ride hailing services us- 5 World Bank 2020b. ing the services as substitutes for private vehicles. Ja- karta for example has an even-odd policy. As a result, 6 The program was initially for the September to De- many households with either only even- or odd-num- cember 2020 period and is discussed in further detail ber-plated vehicle cannot use their private vehicle to in subsequent chapters. BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA commute half of the total working days. 7 World Bank 2016. 20 Planned future work will attempt to look at the ex- 8 Goldfarb and Tucker 2019. tent to which the increase in ridership volumes around 9 This is confirmed in recent work that shows, for exam- metro stations is evidence of shifting commuting pat- ple, that only 6 percent of manufacturing firms in Indo- terns—which would imply corresponding decrease in nesia are using new technologies or other emerging dig- ridership in other parts of the city—or whether the in- itally enabled applications; for 64 percent of the firms’ troduction of the metro together with the complemen- spreadsheets and emails represent the frontiers of digital tary first and last mile service offered by digital ride technologies. ADB 2020. “Innovate Indonesia: Unlock- hailing have managed to lure in new users. ing Growth through Technological Innovation” 21 Google, Temasek and Bain 2020. 10 Goldfarb and Tucker 2019. 22 Expansion of e-commerce in the countryside has reg- 11 Minges, Michael 2016. ularly featured in the so-called “Number One Central Document” that lays out annual priorities for the coun- 12 There are various other statistics on the overall level try. of connectivity in the country. Most estimates from the industry tend to rely on mobile and fixed line sub- 23 MEITY. 2016. Digital India: Driving e-commerce in scription data coming from GSMA and ITU. Although rural and semi urban India. Ministry of Electronics and there is some effort made to identify multiple connec- IT Report tions, multiple SIM-cards etc., arriving at this statistic 24 The Government of Indonesia issued the 14th Eco- from subscription data alone appears quite challenging. nomic Policy Package on the Road Map for the Na- For example, the January 2020 edition of Hootsuite tional Electronic-Based Trade System (SPNBE) or the reports internet penetration at 64 percent of the pop- Road Map for e-Commerce in November 2016, which ulation. The data that are used here and throughout was later stipulated into Presidential Regulation No. 74 the report are based on what is perhaps the most com- of 2017 on the E-Commerce Roadmap. The roadmap prehensive socioeconomic survey Indonesia does on an aims to build a more efficient commercial institution annual basis interviewing around 300,000 households and ecosystem through seven steps, namely: (i) sim- across the country for each round. plifying and expanding access to finance; (ii) providing 13 The gradient on education, which can be expected to tax incentives; (iii) provide consumer protection; (iv) largely mimic the gradient on income is noteworthy. increasing human resource capabilities; (v) improving This is because in many other, especially higher income the national logistics system (Sislognas); (vi) accelerat- settings, it has been found that higher income individu- ing the development of communication infrastructure; als adopt internet quicker but conditional on adoption and (vii) improving cybersecurity. Through ecosystem the intensity of use (measured by time spent online) is development, it is targeted to create 1,000 technopre- actually higher for the low-income groups. In Indone- neurs with a business valuation of US$10 billion and an sia, it appears that the it is the high-income groups that e-commerce value of US$130 billion by 2020 are also spending more time online. See Godfarb and 25 Couture et al. 2020. Prince (2008) for example. 26 See Brynjolfsson et al. 2003; and Dolfen et al. 2017, for 14 Hootsuite We Are Social 2019. example 15 Google, Temasek and Bain, 2020. 27 Hjort and Poulsen 2018; Bahia et al. 2020. 28 The methodology closely follows Freyaldenhoven et al. 2019. 94 References 1 Asian Development Bank. 2020. Innovate Indonesia: Unlocking Growth through Technological Transformation. Asian Development Bank: Manila. Retrieved from: https://dx.doi.org/10.22617/SGP200085-2 (accessed November 16, 2020) 2 Bahia, Kalvin, Pau Castells, Genaro Cruz, Takaaki Masaki, Xavier Pedrós, Tobias Pfutze, Carlos Rodríguez-Castelán and Hernan Winkler. 2020. “Welfare Effects of Mobile Broadband Internet: Evidence from Nigeria” World Bank Policy Research Working Paper #9230. Washington DC: World Bank. 3 Bosker, Maarten, Mark Roberts, Sailesh Tiwari, Monica Wihardja and Putu Sanjiwacika Wibisana. 2021. “Complements or Substitutes? Insights from the Impacts of the Opening of the Jakarta MRT on Go-Jek Ridership” Background Paper for the Digital Economy and Inclusion Flagship report. CHAPTER — 01 4 Brynjolfsson, Eric, Yu (Jeffrey) Hu, and Michael D. Smith. 2003. “Consumer Surplus in the Digital Economy: Estimating the Value of Increased Product Variety at Online Booksellers”. Management Science, 49(11):1580–1596. 5 Brynjolfsson, Erik, and JooHee Oh. 2012. “The Attention Economy: Measuring the Value of Free Digital Services on the Internet” Thirty Third International Conference on Information Systems, Orlando. 6 Couture, Victor, Benjamin Farber, Yizhen Gu and Lizhi Lu. 2020. “Connecting the Countryside Via E-Commerce: Evidence from China” American Economic Review: Insights, forthcoming. 7 Dolfen, Paul, Liran Einav, Peter J. Klenow, Benjamin Klopack, Jonathan D. Levin, Laurence Levin, and Wayne Best. 2017. BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA “Assessing the Gains from E-Commerce”. Stanford University, mimeo. 8 Freyaldenhoven, Simon, Christian Hansen, and Jesse M. Shapiro. 2019. “Pre-Event Trends in the Panel Event-Study Design.” American Economic Review, 109(9):3307-3338. 9 Google, Temasek and Bain. 2020. e-Conomy SEA 2020 – At Full Velocity: Resilient and Racing Ahead. Retrieved from: https:// www.thinkwithgoogle.com/_qs/documents/10614/e-Conomy_SEA_2020_At_full_velocity__Resilient_and_racing_ ahead_bMmKO5b.pdf (accessed November 16, 2020) 10 Goolsbee, Austan, and Peter J. Klenow. 2006. “Valuing Consumer Products by the Time Spent Using Them: An Application to the Internet”, American Economic Review 96(2):108-113. 11 Goldfarb, Avi, and Jeff Prince. 2008. “Internet Adoption and Usage Patterns are Different: Implications for the Digital Divide”, Information Economics and Policy, 20(1):2-15. 12 Goldfarb, Avi and Catherine Tucker. 2019. “Digital Economics”. Journal of Economic Literature, 57(1):3-43. 13 Hjort, Jonas, and Jonas Poulsen. 2019. “The Arrival of Fast Internet and Employment in Africa.” American Economic Review 109(3):1032-1079. 14 Hootsuite and We Are Social. 2019. Indonesian Digital Report 2019. Retrieved from https://datareportal.com/reports/digital- 2019-indonesia (accessed November 16, 2020). 15 Jacoby, Hanan, Arya Swarnata, Sailesh Tiwari, and Maria M. Wihardja. 2020. “Welfare Impacts of Digital Infrastructure Development and Skill-Biased Technological Change in Indonesia.” Background Note for the Digital Economy and Inclusion Flagship report. 16 Ministry of Electronics and Information Technology, India. 2016. Digital India: Driving e-commerce in rural and semi urban India. Ministry of Electronics and IT Report. 17 Minges, Michael. 2016. “Exploring the relationship between broadband and economic growth” World Development Report background papers. Washington DC: World Bank. 18 Roberts, Mark, Frederico Gil Sander and Sailesh Tiwari. 2019. Time to ACT: Realizing Indonesia’s Urban Potential. Washington DC: World Bank. 19 Tiwari, Sailesh, Virgi Sari, Imam Setiawan and Juul Pinxten. 2020. “Pandemic, Poverty and Policy: Ex Ante Poverty and Distributional Impact of the COVID-19 Pandemic” World Bank Policy Research Working Paper, forthcoming. 20 World Bank. 2016. Digital Dividend: World Development Report. Washington DC: World Bank. 21 World Bank. 2020a. Indonesia Systematic Country Diagnostic Update. Washington DC: World Bank. 22 World Bank. 2020b. Indonesia Economic Prospects: The Long Road to Recovery. Jakarta: World Bank. 95 Annex. ANNEX 1.A Impact of e-commerce on prices of high and low-intensity commodities PANEL A: LOW-INTENSITY E-COMMERCE COMMODITIES CHAPTER — 01 VARIABLES D1. LOG (PRICE) 0.0143*** 0.0149*** POST 2015 (= 1) -0.00345 -0.00368 0.0204 0.018 HIGH E-COMMERCE PENETRATION IN 2019 (=1) -0.0181 -0.0172 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA -0.0342 -0.0344 POST 2015 (D=1) X HIGH E-COMMERCE PENETRATION IN 2019 (=1) -0.03 -0.0308 0.0318*** 0.156* CONSTANT -0.00295 -0.0764 ADDITIONAL CONTROLS NO YES OBSERVATIONS 214 214 R-SQUARED 0.012 0.014 PANEL B: HIGH-INTENSITY E-COMMERCE COMMODITIES VARIABLES D1. LOG (PRICE) 0.000459 0.000459 POST 2015 (= 1) -0.00319 -0.00267 0.00686** -0.0018 HIGH E-COMMERCE PENETRATION IN 2019 (=1) -0.00283 -0.00404 -0.00829* -0.00825* POST 2015 (D=1) X HIGH E-COMMERCE PENETRATION IN 2019 (=1) -0.00424 -0.00429 0.0416*** 0.175* CONSTANT -0.00188 -0.0916 ADDITIONAL CONTROLS NO YES OBSERVATIONS 217 217 R-SQUARED 0.022 0.094 Source: Authors’ estimation using data on prices from CEIC and e-commerce penetration data from platforms (Tokopedia and Bukalapak). Note: High e-commerce penetration are provinces that were in the top 40 percent in terms of the proportion of buyers in 2019 while low e-commerce penetration provinces are those in the bottom 40; low-intensity e-commerce commodities include items such as staple food, meats, processed foods, animal produce, vegetables and fruits; higher-intensity e-commerce commodities include building materials, household wares, clothes, convenience goods, and automotive spare parts; additional controls include province level logistics cost, per capita GDP, and the level of urbanization. Standard errors are clustered at the province level. 96 97 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA CHAPTER — 01 Spotlight 1. Use of Social Media in Indonesia and Implications for Inclusion CHAPTER — 01 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA → What are Indonesians doing on social media? Being an internet user in Indonesia is almost synonymous The popularity of social media may reflect the predilec- with being a social media user, as the use of social media tion of Indonesians to socialize and discuss. On the most is pervasive in the online population. Over 85 percent of popular platform, WhatsApp, this certainly appears to be internet users in the digital economy household survey the case, as users are principally using it for communica- (a nationally representative survey that was conducted tion and sharing information (Figure S1. 3). Likewise, on for this report) were also users of social media, with the other platforms (Facebook, Instagram and Twitter), users most popular platforms being WhatsApp, Facebook and are active in seeking out news and information. However, Instagram (Figure S1. 1). social media platforms are also important for commerce in Indonesia, with a sizeable proportion of users (about There are some gender differences, with females being 20 percent) using Facebook, Instagram and WhatsApp more avid social media users than males (93 percent fe- for buying and selling. males accessed social media in the last week vs. 88 percent of males surveyed) and particularly stronger users of the The most common topics of discussion across all plat- Instagram platform (32 percent females accessed Insta- forms relate to hobbies and lifestyles (Figure S1. 4). gram in the last week vs. 22 percent of males). There is However, religion, public policy and politics are also also an age dimension with social media use increasing widely discussed, highlighting the importance of social among younger cohorts (Figure S1. 2), particularly on media as a channel of communication and influence on certain platforms. For example, Instagram and Line are these topics. Relatedly, religious leaders are the most far more widely used by young adults aged between 15 followed public figure (38 percent of respondents follow and 29, while the use of Facebook and WhatsApp is wide- a religious leader on social media), followed by celeb- spread across all age groups. Indonesians of all education rities (30 percent), political leaders (21 percent) and levels are using social media, although there is more regu- government officials (18 percent). The level of trust lar use among those with tertiary education (97 percent) in the information received varies across platforms, compared with those with primary education (83 percent although over 90 percent of social media users across accessed in the last week). Overall, use of social media is all platforms indicate at least sometimes trusting the intensive, at over 25 hours per week, or 3.5 hours a day, information they receive. Information shared through spent on average using social media. This ranks Indonesia WhatsApp is considered most trustworthy, with 44 per- fifth in the world in terms of daily time spent using social cent always or often trusting the materials they receive, media. Intensity of social media use has a significant age while Facebook is considered a more uncertain source dimension to it, with younger cohorts of adults spending of information, with around 21 percent always or often significantly more time on social media. For example, trusting the materials they receive. adults aged between 15 and 24 spend 5.6 hours per day on average, while adults aged between 35 and 50 spend 2.7 hours per day. 98 FIGURE S1.1 FIGURE S1.3 Most recent use of various social media The most common activities on social media platforms (share of internet users) Source: World Bank Digital Economy Household Survey, 2020 Source: World Bank Digital Economy Household Survey, 2020 FIGURE S1.2 FIGURE S1.4 Profile of social media users in Indonesia The most common topics in social media (share of internet users) discussions Source: World Bank Digital Economy Household Survey, 2020 Source: World Bank Digital Economy Household Survey, 2020. 99 effect was found not only to be driven by the mechanical → What does the level of fact that influencers were likely to be connected to more adoption and pattern of use people, but also “because subsequent individuals are con- siderably more likely to respond to the message when it of social media imply for is sent by influential agents.”31 This suggests that social economic and social inclusion? media may generate opportunities to harness the brand value of influential celebrities to try and change offline This is an emerging research space and significant learn- beliefs and actions of large segments of the population ing remains to be done. But a couple of implications have to boost inclusion outcomes. started to crystalize based on evidence from around the world. One strand of research points to the potential A flipside to all of these is a potential risk of misinfor- coordination effects of social media, which can increase mation and polarization. By design, social media expos- CHAPTER — 01 voter participation, as well as coordinate citizens to ad- es users to unstructured information flows whereby the vocate for social and political changes, and influence information received is influenced by who is in your net- decision-making processes. A large-scale randomized work, the information they share, as well as algorithmic controlled trial (RCT) in the United States found that predictions of what is expected to interest you most. By messaging that encouraged citizens to “get out the vote” doing this, there is a risk that social media may create shared through Facebook was effective in increasing voter ideological “echo-chambers” and “filter bubbles”. In the turnout in the 2010 congressional elections, for voters echo-chamber phenomenon, individuals are exposed only of both major parties.29 Likewise, cross-sectional survey to information from like-minded individuals or people BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA analysis finds that social media usage coincides with in- with the same ideology.32 Meanwhile, the “filter bubble” creased civic participation (examples include voluntary conditions users only to see content that is selected by work for nonpolitical groups, raising money for charity, algorithms according to a viewer’s previous behaviors. attending a meeting to discuss neighborhood problems, These two conditions can amplify existing divisions in (not) purchasing products (mis-)aligned with personal society and cause further polarization.33 A second risk of social values), as well as political participation.30 how social media usage can cause polarization is through the emergence of fake news or misinformation. The in- Evidence of these coordination effects is visible within creasing number of available sources for news and infor- Indonesia. Over 55 percent of survey respondents indi- mation can deepen ideological differences, particular- cated that they participated in civic engagement activi- ly when users end up trusting unreliable ideologically ties (such as non-electoral and community organizations, aligned sources more than accurate neutral sources due fund raising, village and community meetings), with 67 to overconfidence in their own feedback.34 percent saying that social media had helped increase their participation. Likewise, 43 percent reported that social Empirically, there are a few studies providing evidence of media had helped increase their participation in polit- these risks, although primarily in the United States. For ical processes, such as voting, joining demonstrations example, in an RCT in the United States, temporary Face- and making financial contributions. While better access book deactivation reduced political polarization through to information is cited as the most common reason for lowering exposure to news from the political party Face- this impact (75 and 77 percent, respectively, for civic and book users were already most aligned with.35 However, political participation), the next most common reason is this same study also found users had a high willingness to because it allows social media users to find people with the pay for social media access and quickly returned to prior same beliefs as them (47 and 36 percent, respectively). levels of adoption at the end of the study, suggesting that user-directed moderation of content consumption is not One recent piece of evidence from Indonesia also suggests likely to have sustainable depolarizing impacts. A more that, in addition to the coordination function, social me- severe consequence of polarization is manifestations of dia and, especially social media actors with a following conflict, and some evidence highlights the risks that are within platforms who are known in the industry parlance becoming apparent even in Indonesia. Conflicts within as “influencers”, also have significant power to change villages increased on average by 0.6 percentage points offline beliefs and behaviors. For example, in an informa- relative to a baseline rate of 2.8 percent in villages that tion campaign on immunization in Indonesia, celebrity “received” BTS towers, communication technology that endorsement of the campaign had significant impact on provides internet and cellular network access with this offline knowledge about immunization among the popu- effect greater in more religiously polarized areas.36 lation exposed to it and greater take up of immunization among friends, neighbors and families of respondents Another emerging concern around the use of social media who had been exposed to this content. Oddly, citing a is on physical and mental health effects. Over 50 percent credible source of information in the endorsement mes- of survey respondents indicate some emotional attach- sage had an adverse effect on information diffusion. This ment to being online, with 25 percent reporting they 100 CHAPTER — 01 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA feel worried depressed, lonely, and 29 percent reporting they feel like they might be missing out on something, → What does this evidence and but it was “not a big deal”, when they were not online or these findings of social media able to use their phone for a day or more. Around 23 per- cent report making comparisons of their life with others usage imply for policy? through social media, while 43 percent indicate it has affected their family relationships, but mostly in a positive Social media is a tool for coordination. This can catalyze way (82 percent felt it brought family members closer greater civic participation, which can in turn strengthen together). Headaches and eye strain (44 percent), thumb/ accountability at all levels. However, there are fast-chang- wrist/hand/neck/shoulder pain (28 percent), insomnia ing dynamics of public opinion on current affairs given (22 percent), and forgetting to eat (21 percent) or exercise the ease of spreading information through social me- (18 percent) were the most common ailments that users dia. When this information turns more into the realm experienced. Little research has been done globally on of misinformation, gossip or simply polarizing content, the public health costs of internet and social media use, there is a heightened risk of greater social divisions and so it is hard to extrapolate out the welfare implications public sentiment evolving based on inaccurate or biased of these findings. information. This can be used to the advantage or dis- advantage of policy makers, and may call for stronger A final rising concern surrounding social media and oversight and regulation. Such oversight would need to broader internet access relates to exposure among chil- address when to correct “mistruths” and how, with tools dren. Thirty-two percent of survey respondents did such as de-bunking websites, correcting beliefs through not monitor their children’s online activities, while the building new narratives on social media, norm setting remaining respondents with children were highly con- by social media influencers and religious leaders, politi- cerned with their children’s online activities. The biggest cians speaking out against misinformation and not using concerns of parents were risk of negative content expo- negative smear approaches in their own campaigns, to sure (86 percent), interrupting study habits (59 percent), more severe forms such as controlling content or direct the risk of exploitation by other social media users (40 censorship and strengthening laws around the spread of percent) and teaching anti-social behavior, such as an false or misinformation. This would also call for policies inability to communicate face-to-face (28 percent). The and investments to boost the overall digital literacy of main means parents chose to guard against excessive use the internet-using population to enable them to discern of social media or online activities was through limiting better the veracity of the online information they con- access to ICT devices (85 percent) and directly monitor- sume through social media. Similarly, addressing health ing their online accounts and browser history (46 per- risks, particularly to children, relatively little is known cent). A minority resorted to installing software to mon- regarding the public health risks, but these will need to itor (8 percent) and block (4 percent) online activities. be studied and closely monitored in the future. 101 Endnotes 29 Bond et al., 2012. 30 Gil De Zuniga et al., 2019. 31 Alatas et al. 2019. 32 Sunstein, 2018; Baumann et al. 2019. 33 Bakshy, 2011. 34 Gentzkow, Wong, and Zhang, 2018. CHAPTER — 01 35 Alcott et al., 2019. 36 Boxell and Ralston, 2020. BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA 102 References 1 Alatas, Vivi, Arun G. Chandrasekhar, Markus Mobius, Benjamin A. Olken, and Cindy Paladines. 2019. When celebrities speak: A nationwide twitter experiment promoting vaccination in Indonesia (No. w25589). National Bureau of Economic Research. 2 Allcott, Hunt, Luca Braghieri, Sarah Eichmeyer, and Matthew Gentzkow. 2019. The welfare effects of social media (No. w25514). National Bureau of Economic Research. 3 Bakshy, Eytan, Solomon Messing, and Lada A. Adamic. 2015. “Exposure to ideologically diverse news and opinion on Facebook.” Science, 348(6239):1130-1132. 4 Baumann, Fabian, Philipp Lorenz-Spreen, Igor M. Sokolov, and Michele Starnini. 2020. “Modeling echo chambers and polarization dynamics in social networks.” Physical Review Letters, 124(4):048301. CHAPTER — 01 5 Bond, Robert M., Christopher J. Fariss, Jason J. Jones, Adam DI Kramer, Cameron Marlow, Jaime E. Settle, and James H. Fowler. 2012. “A 61-million-person experiment in social influence and political mobilization.” Nature, 489(7415): 295-298. 6 Boxell, Levi. and Laura Ralston. 2020. “Information Technology and Social Development in Indonesia”. Working Paper. 7 Gentzkow, Matthew, Michael B. Wong, and Allen T. Zhang. 2018. “Ideological bias and trust in information sources.” Working Paper. 8 Gil de Zúñiga, Homero, Nakwon Jung, and Sebastián Valenzuela. 2012. “Social media use for news and individuals’ social capital, civic engagement and political participation.” Journal of computer-mediated communication, 17(3): 319-336. BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA 9 Sunstein, Cass R. 2018. # Republic: Divided democracy in the age of social media. Princeton University Press. 103 Availability, Access and Adoption →The Drivers of Digital Divide in Indonesia CHAPTER — 02 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA 2 Chapter ⬎ 104 2.1. Introduction CHAPTER — 02 A s discussed in Chapter 1, being connected to high-quality internet is a pre-requisite to directly benefiting from the digital economy. The COVID-19 pandemic has further reinforced this by pushing work, education, and other es- sential social and economic interactions online. As also BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA pointed out in Chapter 1, despite rapid progress over the past decade or so, Indonesia faces several challenges in uni- versalizing access to the internet. In fact, reliable, fast and high-quality internet remains a luxury for many Indone- sians. Although by now connectivity infrastructure (network cov- erage) has reached almost all parts of the country, nearly 94 million Indonesians—more than one-third of total population—are still not plugged in. More than 60 percent of adults living in Nusa Tenggara, Maluku and Papua remain without internet (Figure 2.1). Even in Java-Bali, 45 percent of the adults are not connected to the internet. FIGURE 2.1 Last mile connectivity challenges persists PROPORTION OF INDIVIDUALS 15+ WITHOUT ACCESS TO INTERNET, BY REGION Source: SUSENAS, August 2019. 105 Moreover, owing to their large population size, the other- The objective of this chapter is to diagnose the drivers of wise relatively better-connected regions such as Java-Bali the various facets of the digital divide in Indonesia. Linking and Sumatra account for a total of 77 percent of all In- the supply- and demand-sides of connectivity, the chapter donesians who are not connected to the internet (Figure investigates the factors that hold back the universalization 2.2). As pointed out in Chapter 1, 58 percent of those who of internet access in the country. Reaping the full potential are not connected live in rural areas. These stylized facts of digital transformation is not only about the availability suggest that, while there is still work to be done to boost and accessibility of the hard infrastructure, but also about connectivity in parts of the archipelago where distance having the necessary knowledge and skills to fully exploit and low population density make investments costly and the opportunities that digital technologies generate in a demand relatively weak, there is a significant amount of safe and secure manner. Thus, in addition to the constraints work to be done also on going the last mile in places such on the broader availability of digital infrastructure and as Java-Bali and Sumatra. household level barriers to adoption of these technolo- CHAPTER — 02 gies, the chapter also utilizes a novel dataset from a survey There is heterogeneity in access even within places with implemented specifically for this report to describe and the same level of internet availability, with socio-de- characterize the level of digital skills in the population. mographic factors playing a key role. Offline adults in Indonesia are more likely to be female, relatively older, The chapter is organized in the following way: Section 2 have lower levels of education, and lower overall living describes the evolution and current state of digital infra- standards. Younger adults (16 to 25) in the same local- structure in the country, benchmarking Indonesia with ity are ten times more likely to be connected than older other regional peers where relevant. Section 3 discusses BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Indonesians, while those with tertiary education are five issues related to cost, affordability and the quality of the times more likely to be connected than those with junior internet experience. Section 4 carries out a deeper di- secondary school education. Individuals belonging to agnosis of the main drivers of the digital divide. It also the top 20 percent of the income distribution are three proposes a framework for measuring core digital skills times more likely to be connected than individuals from and describes the level of skills among internet using In- poor households. This is consistent with regional/global donesian adults. Finally, Section 5 concludes with a brief trends. discussion of potential entry points for policy. FIGURE 2.2 Java-Bali and Sumatra still have the highest number of Indonesians not connected PROPORTION OF INDIVIDUALS 15+ WITHOUT ACCESS TO INTERNET, BY REGION Source: SUSENAS, August 2019. 106 2.2. Evolution and the State of Digital Infrastructure FIGURE 2.3 Broadband infrastructure value chain CHAPTER — 02 CROSS BROADER NETWORK NATIONAL BACKBONE MIDDLE MILE NETWORK (INTERNATIONAL LINKS) (CORE NETWORK) (DISTRIBUTION) LAST MILE (ACCESS) Connecting the service Connecting bigger Bringing Internet Distributing Internet from centers to the world cities together and to a point in a the middle mile in the wide internet routing data to the community for broader community to individual service centers distribution homes & business Purpose BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA FIBER FIBER DSL COAX Technologies TERRESTRIAL FIBER FIBER OR MICROWAVE WIRELESS WIFI, LTE, TV WHITE SPACE SATELLITE LINK SATELLITE LINK (DTH) SUBSEA CABLE WIRELESS BALLOONS & DRONES Typical distance 1000s KM 100s-1000s KM 100s-100s KM <1-5s KM Source: World Bank, 2018. Indonesia’s digital connectivity infrastructure comprises Indonesia’s digital connectivity has been improving steadi- the following layers: the cross-border network or inter- ly over the past decade, driven primarily by private invest- national fiber optic connectivity (submarine cables); the ment in telecommunications/internet services by the ma- national fiber optic backbone consisting of terrestrial and jor telecommunications service providers (telcos). Mobile submarine cables to the main cities; the middle mile net- network coverage is near-universal, almost all Indonesians work that brings the internet to a point in a community have access to at least basic mobile (2G) and text messaging in the form of fixed and mobile broadband fiber network, (SMS) services, including in some of the remotest parts of and in the form of satellite links, and occasionally short- the country. Over the past five years, the telcos have also wave radio in very remote locations; and finally the last invested heavily in mobile broadband (3G and faster 4G) mile distribution to individual homes and businesses in infrastructure and services, and increasingly in fixed broad- the form of fiber, DSL and COAX cables, and other wire - band, notably fiber to the home (FTTH). The following less methods (Figure 2.3). sections review the current status of digital connectivity in Indonesia, and analyze critical access and quality gaps. 107 to more direct international submarine cables than Vietnam, → Though further cable builds Thailand and the Philippines, though there may be opportu- could be useful in improving nities for further cable builds in eastern Indonesia to provide redundancy/restoration in those regions. The two furthest resilience in some parts of the destinations on the Sulawesi-Maluku-Papua Cable System country, Indonesia has adequate (SMPCS) are Jayapura and Merauke. These end-tails of the international bandwidth to domestic submarine infrastructure do not have an alternative route since an overland cable between them is not yet possible service current and medium-term due to mountainous terrain and lack of roads. Further interna- demand tional extensions could be used to improve resilience. Jayapura has recently been connected to PNG, which opens the option Indonesia is connected to global telecommunications networks of an alternative route. Merauke would also benefit from an CHAPTER — 02 via multiple international submarine cables, such as Medan (2), onward international connection to PNG and/or Australia. Dumai (5), Batam (at least 4 routes), Bintan (1), Jakarta (at least 12 routes), Pontianak (1), Manado (1) and Jayapura (1). Figure 2.4 below shows the large number of international International bandwidth availability is adequate for current submarine cables passing through the Singapore Straits be- and medium-term demand given the very high capacity of each tween Singapore and Batam. Most of Indonesia’s international submarine cable route (for example, the cable from Manado to connectivity comes from Singapore and Malaysia though the the United States has a design capacity of 10 Tbps and initially SEA-US cable from Manado to the United States, and provides just activated less than 10 percent of this capacity, while ded- a high level of geographical diversity. Indonesian telecom pro- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA icated submarine cables to Malaysia and Singapore can carry viders also have capacity on various consortium cable systems very high traffic and capacity that is just lit-on as needed) and running directly to the United States and Europe. comparable to regional benchmarks. Indonesia also has access FIGURE 2.4 Submarine cables connecting Indonesia to the main international routess banda aceh samarinda sorong ambon jayapura Makassar Jakarta denpasar kupang Source: https://www.submarinecablemap.com/ Note: Only some domestic cables shown in the map. 108 “Backbone networks CHAPTER — 02 are also sufficiently well deployed, and BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA with the completion of the Palapa Ring project, all of Indonesia’s 514 districts are connected to the domestic backbone network.” 109 FIGURE 2.5 Backbone networks are also sufficiently well deployed by service providers TELKOM banda aceh CHAPTER — 02 sorong samarinda jayapura BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA ambon Jakarta Makassar denpasar kupang BIZNET banda aceh samarinda sorong jayapura ambon Jakarta Makassar denpasar kupang Source: Respective service providers. Note: In addition to the four providers shown here, XL and Indosat operate their own backbone networks across Java, Bali, Lombok, Sumbawa, Sumatra, Kalimantan and Sulawesi. 110 MORATELINDO CHAPTER — 02 banda aceh samarinda sorong jayapura ambon BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Jakarta Makassar denpasar kupang ICON + banda aceh samarinda sorong jayapura ambon Jakarta Makassar denpasar kupang 111 FIGURE 2.6 Palapa Ring project 57 Districts/Cities 13,000 Km Total lenght banda aceh 28 Interconnecting Cities CHAPTER — 02 samarinda sorong jayapura ambon Jakarta Makassar BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA denpasar kupang EXISTING OPERATORS’ FIBER OPTIC PLANNED OPERATORS’ FIBER OPTIC EXISTING PALAPA RING (FIBER OPTIC) PLANNED PALAPA RING (OPEN TECHNOLOGY) Source: Kominfo Note: This map is not fully up to date since the cable system to Merauke is already in operation as well. In Central Sulawesi also, the route from Bitung to Liwuk is already in operation and in Sumatra the islands to the west, such as Sabana-Aceh, Simeuleu-Bakongan, and Sibolga-Nias have already been connected Telcos have invested significantly in the domestic backbone, as 15 years) provided by the USO Fund (BP3TI, now called BAK- well as in middle mile networks that connect Indonesia’s main TI) to the three consortia that deployed the Palapa Ring proj- population centers. Figure 2.4 illustrates the backbone net- ects. The consortia obtained financing construction loans that works deployed by different service providers. As of end-2019, were backed by guarantees from the Indonesia Infrastructure 457 kabupaten/kota had been connected to high-speed optical Guarantee Fund (IIGF). The users (telecom providers leasing fiber backbone networks through private sector investments. backbone capacity) pay for the services and these revenues flow Of all the providers, the Telkom domestic backbone is the back into the USO Fund. most extensive (100,069 km at end-2019), but Indosat, XL, ICON+, Moratelindo and Biznet also have extensive fiber optic With the completion of the full Palapa Ring system, all 514 backbones on Java-Bali and Sumatra, with some extensions to kabupaten/kota have been connected. Almost all are on the fiber Sulawesi and Kalimantan (Figure 2.5). optic network, except for locations in the interior of Papua province where microwave was the only feasible connectivity The Palapa Ring project, a private-public partnership (PPP) led option. Since the completion of the Palapa Ring backbone, by the Ministry of Communications and Information Technolo- there have been subsequent private sector investments in mo- gy (Kominfo) and implemented during 2015–19, has succeeded bile broadband (LTE) and fixed broadband (FTTH) access in extending the domestic backbone to the entire country. The networks, and users in many of these more remote districts original Palapa Ring project was divided into a commercially have been able to avail themselves of broadband services sim- feasible part and a commercially non-feasible part. The com- ilar in performance to many urban centers. Nonetheless, fur- mercially feasible part (457 districts/cities) was undertaken ther extension of some middle mile and access networks to by the private sector and included the routes from Lombok to smaller population centers and villages within the districts is Kupang, SMPCS (Sulawesi-Maluku-Papua-Cable-System) up required. Linking the 514 kabupaten/kota on the main national to Jayapura and Merauke, and several smaller cable projects. The backbone to the 7,246 kecamatan (subdistricts) and further to commercially non-feasible part of Palapa Ring (57 districts) was 75,436 villages still has gaps. On Java-Bali and Sumatra most financed through the USO Fund (financed by a levy on the net kecamatan have been connected by middle mile connectivity, revenues of the telcos) and was also deployed in 2015–19. The mostly fiber optic and, in some cases, still through microwave business model is based on an availability payment (for about links, while in eastern Indonesia there is still significant middle 112 FIGURE 2.7 Mobile broadband access networks have also expanded (BASE STATION DEPLOYMENT OVER TIME) TELKOMSEL INDOSAT XL CHAPTER — 02 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Source: Respective company disclosures. mile connectivity to be improved. Some kabupaten are large and i.e., 4G/LTE networks (end of 2019; noticing major rollout middle mile distances of up to 100+ km are not uncommon in 2020, this should be higher now), while over 98 percent are in rural areas. within 2G network range. The 3G layer is estimated at over 95 percent but, as mobile operators migrate 3G radios by software upgrade to LTE, a decline in the number of 3G cell sites has → Mobile broadband access become visible over the course of 2020. The latest available networks have also expanded data from telcos, as well as public sources such as Telegeogra- phy (GSMA), put total mobile broadband (3G and 4G/LTE) rapidly to near universal coverage subscriptions in Indonesia at about 255 million (allowing for of 4G/LTE services, but fixed multiple SIM card ownership). Of these, about 157 million are broadband is lagging 4G subscribers (mid-2020), so 4G mobile broadband pene- tration is around 58 percent of the population. Access networks—the digital equivalents of feeder roads— A relatively small (1.3 percent) share of the population in are what connect the digital highways (fiber optic backbone) very remote locations has no mobile network access and re- to end users in the last mile. They can take the form of fixed lies entirely on satellite connectivity. The main project to broadband (ADSL or copper, and increasingly, optical fiber) close this remaining gap is the 5000 BTS blank spot program and mobile broadband networks (3G, 4G/LTE). managed by BAKTI which, as mentioned earlier, is also the entity in charge of the Universal Service Obligation (USO) Access networks for mobile broadband have expanded rap- Fund under Kominfo. idly over the past decade, with investments by some of the major players, such as Telkomsel, Indosat, XL Axiata, Tri and Telkom, the largest provider of fixed broadband or fiber-to-the- SmartFren, driving this progress. Over the past decade or so, home (FTTH) services in Indonesia, has been deploying ADSL the deployment of base stations has accelerated rapidly, driven services over the PSTN copper lines to provide basic broadband mainly by 3G and 4G/LTE stations (Figure 2.7). An estimated services. In 2014, Telkom launched IndiHome, a triple play fiber 95 percent of the population lives within reach of the fastest, optic proposition and, by December 2017, about 3 million out of 113 FIGURE 2.8 the 5.3 million fixed broadband customers were connected to FTTH. As of Q3 2020, Telkom had about 30 million homes passed (i.e., po- Indonesia lags behind many ASEAN neighbors on tentially connected, as optical fiber cables have fixed broadband penetration been deployed in the area) and 7.8 million homes connected.37 Other providers have also been ac- tive in this space. Linknet (the main cable TV provider in Jakarta) reported 2 million homes MOBILE BROADBAND PENETRATION passed, while MNC Play also reported 1.2 million homes passed. Biznet offers services in 100+ cit- LEGEND GSM ies but the extent of the access network in those 4G/LTE 3G places is limited, with a total number of reported CHAPTER — 02 homes passed at 450,000. As a new player PLN, through ICON+, has launched Stroomnet pro- vided jointly with electricity services. Considering the FTTH services provided by all these market players, the total estimated num- ber of fixed broadband subscribers is about 9.7 million. This translates to fixed broadband pen- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA etration of just 4 percent of the population, or 16 percent of households. Thus, while Indonesia performs relatively well compared with some of the regional peers in terms of mobile broad- band penetration, it performs distinctly less well compared with some of the regional peers, such as Singapore, Malaysia, Thailand and Vietnam, on fixed broadband rollout (Figure 2.8). From a more global perspective and relative to its level of development, Indonesia is behind on fixed broadband penetration (Figure 2.9). Fixed broadband access is essential for data-in- FIXED BROADBAND PENETRATION tensive transactions and very important for businesses, government, schools and health- care facilities, and limited progress in this is a key driver of the digital divide in Indonesia. Though mobile broadband coverage is near universal, and access to relatively high (as doc- umented in Chapter 1), the quality of service is highly variable, as discussed in the following sub-section. This imposes challenges on house- holds and businesses, particularly SMEs, which are compelled to rely on mobile broadband to support more sophisticated applications. LEGEND Population Household Source: Telegeography (2020). 114 FIGURE 2.9 Fixed broadband penetration is also below countries at similar levels of development CHAPTER — 02 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Source: World Development Indicators and ITU, 2020. 115 2.3. Affordability and Quality of the Internet Experience Connecting people is not just about network FIGURE 2.10 coverage. As discussed above, mobile broad- band (4G) network coverage now extends to Overall internet affordability in CHAPTER — 02 about 95 percent of the population. However, Indonesia is on the lower side… actual mobile broadband penetration (mea- sured in terms of subscriptions) is only about EIU AFFORDABILITY SCORE, 2020 55 percent of the population. Similarly, for fixed broadband, actual subscription is often just a fraction of the number of homes passed. For example, for PT Telkom, the actual number of subscribers is just 26 percent of the number of BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA homes passed. So improved network coverage does not necessarily lead to increased adoption/ uptake of services. We examine two key drivers of adoption decisions, namely cost and afford- ability, and the overall quality of service. → Mobile broadband data is affordable but the high cost of fixed broadband subscription weighs heavily on overall internet Source: EIU, Internet Inclusivity Index, 2020. Note: Scores are 0–100, with higher score representing greater affordability. affordability in Indonesia In terms of overall affordability of internet, FIGURE 2.11 Indonesia ranked 61st out of 100 countries analyzed by the Economist Intelligence Unit’s …and the price of mobile broadband annual Inclusive Internet Index in 2020. In com- data is not the issue parison to its regional peers, Indonesia outper- forms Cambodia and the Philippines, but lags PRICE OF MOBILE BROADBAND (1GB, Thailand, Singapore, Malaysia and Vietnam on PREPAID), AS % OF GNI ON 2017 overall internet affordability (Figure 2.10). Affordability of mobile data is not a major con- cern, since it is well within the affordability criteria (0.95 percent of GDP/capita versus the ITU-UNESCO target of below 2 percent). Al- though mobile service providers in Indonesia adopt differential tariff-pricing to cover higher service provision costs in more sparsely populat- ed and geographically challenging areas, there is a wide range of mobile data packages for differ- ent budgets and needs, making internet mobile data relatively affordable for all segments of the population. Even for poor households, the aver- Source: ITU 2017. 116 FIGURE 2.12 Fifty-five percent of adults who do not use internet-enabled devices cite cost as the primary barrier to adoption INTERNET ACCESS FOR THE POPULATION AGE 15 OR OVER CHAPTER — 02 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Source: World Bank Digital Economy Household Survey, 2020 age price per GB in a mobile data plan of US$0.64 is less than only 17 percent of poor households having the required ID and 1 percent of their monthly per capita expenditure, suggesting a bank account, this credit facility is unlikely to be helpful for that mobile data costs should not prevent Indonesians from this segment. Inability to afford internet-enabled devices is one connecting to the internet. From a regional perspective as well, of the main reasons some of the poorest in Indonesia are unable the price of mobile prepaid data in Indonesia—1GB at average to connect to the internet, despite near universal coverage of of 0.95 percent of GNI—is lower than the ASEAN average of mobile broadband. 1.4 percent and the global average of 5.5 percent. In contrast to mobile broadband, setting up a fixed broadband Affordability of mobile handsets, particularly internet-enabled service entails a wide array of costs that can include the cost of ones, appears to be somewhat an issue. Although mobile phone modem rental, installment fee and the monthly subscription penetration and smartphone ownership have been increasing fee. Indonesia ranked 131 out of the 200 countries in the 2019 at a significant pace over the past decade, many Indonesians ITU rankings on fixed line subscription fees, suggesting that still either do not have a mobile phone or own only a basic cost could be a clear binding constraint for the adoption of phone that lacks the necessary functionality to connect to the fixed broadband relative to countries around the world. Data internet. Around 55 percent of adults who do not own and op- from a survey conducted specifically for this report confirm erate internet-enabled mobile phones cite cost as the primary this. Over 40 percent of households reported fixed broad- impediment (Figure 2.12). band subscription costs to be prohibitive. This is distinctly higher than the proportion of households whose reasons for A variety of smartphones can be found at relatively low prices non-adoption were the use of mobile broadband as substi- in Indonesia. Handsets prices start at around IDR 450,000 tutes for fixed broadband (24 percent) and the proportion of (about US$31), or equivalent to 11 percent of average monthly households that reported not having coverage (14 percent) household expenditure. But for the poor, even these entry level (Figure 2.13). From a regional perspective, cost is the most devices can account for anywhere between 20 and 35 percent salient barrier to adoption in places such as Sumatra, Sulawesi of their monthly household expenditure, depending on where and Maluku, while availability of services is the bigger barrier they live. Setting aside such an amount for a one-time purchase in Papua and Nusa Tenggara. In contrast, Java-Bali stands out may not always be feasible. While credit schemes are increas- with the highest proportion of households that treat mobile ingly being offered for the purchase of ICT devices, availing broadband as a satisfactory enough substitute for the fixed of these schemes requires ID (KTP) and a bank account. With broadband internet experience (Figure 2.13). 117 → Both fixed and mobile FIGURE 2.13 broadband speeds in Indonesia Affordability is a key barrier to the are among the lowest in adoption of fixed broadband for many ASEAN, and the poorer quality in the more populous parts REASONS FOR NOT SUBSCRIBING TO FIXED BROADBAND, BROKEN of the country suggests that DOWN BY ISLAND REGIONS network congestion is a major challenge CHAPTER — 02 Quality of service (QoS), or the overall quality and re- liability of internet services, is also another challenge in different parts of the country. QoS impacts the speed of data transmission (uploads, downloads), the quality of phone calls, and hence the ability of users to access on- line services, be it e-commerce or video streaming, for example, for remote health-care or learning services that have become integral during the COVID-19 pandemic. BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA This is also an area in which Indonesia lags critical- ly behind some of its regional peers. Average mobile broadband download speed experienced by Indonesian consumers is about 9.8 Mbps, the lowest speed in all of ASEAN. The mobile download experience in Indonesia is not just a mere one-quarter of the average speed available to Singaporean consumers but also a little over one-half of that enjoyed by consumers in Myanmar. Incidentally, Myanmar is also the only country in the ASEAN region that Indonesia outperforms on fixed broadband download LEGEND speeds. But the overall level is almost one-tenth of market leaders in the region, such as Singapore. (Figure 2.14) Source: World Bank Digital Economy Household Survey, 2020 FIGURE 2.14 Both fixed and mobile broadband download speeds in Indonesia are among the lowest in the ASEAN region MOBILE BROADBAND THROUGHPUT (MBPS) FIXED BROADBAND THROUGHPUT (MBPS) Source: Ookla Speedtest (March, 2020). Note: Mobile broadband speeds in the left panel and fixed broadband speeds in the right panel. 118 FIGURE 2.15 There is also considerable variation in download speeds within Indonesia with, somewhat interestingly, places in West Papua (Sorong), Maluku (Ambon) and Papua Quality of the internet experience (Jayapura) registering the best mobile download speeds varies significantly across the country (Figure 2.15). In contrast, more populous places particu- larly in the periphery of large metropolitan areas, such as MOBILE BROADBAND THROUGHPUT (MBPS) Cimahi outside Bandung and Tangerang and Bogor, just outside Jakarta, are the places with the poorest internet experience as measured by download speed. QoS is deter- mined by several factors, including network congestion (linked to availability of spectrum for data transmission), the availability of cell sites, weather and geographic lo- CHAPTER — 02 cation. This observed pattern on the variation in the QoS across the country clearly points to network congestion be- ing a key impediment to high-quality download experience in the higher population density places in Indonesia. This issue is examined more closely in the following section. BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Source: Ookla Speedtest (March, 2020). 119 2.4. Drivers of the Digital Divide This chapter has so far established several key challenges related to internet connectivity in Indonesia. First, fixed → Spectrum and unavailability of specific bands CHAPTER — 02 broadband access is insufficient to meet rapidly growing de- mand for high-bandwidth applications, particularly for in- stitutional users. Second, mobile broadband is more widely available but does not deliver the same capacity, or QoS, All Mobile Network Operators (MNOs) need a spectrum and is not cost-efficient enough for high bandwidth appli- portfolio, and the availability and overall quality of mobile cations. Third, adoption of fixed broadband is constrained broadband hinge crucially on having adequate spectrum. by the availability of access networks, as well as the afford- Currently, Indonesia has allocated a total of 467 MHz in ability of the subscription plans. A sizeable proportion of total IMT spectrum to its active MNOs (Table 2.1). In glob- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA the population, particularly in Java-Bali, also appears to al and regional terms, this quantum of available spectrum regard mobile broadband as a satisfactory substitute to in Indonesian is low and has not increased for some time fixed broadband. Finally, the overall quality of the available except for some small additional spectrum being made internet service in Indonesia is significantly poorer than available in the 2.3 GHz band. Before the introduction its regional peers and varies substantially across the coun- of 5G services, 100-150 MHz in total IMT spectrum may try, with strong evidence of mobile network congestion, have been enough for a successful MNO. However, this especially in greater Jakarta and the larger urban centers. is unlikely to remain the case in the future. Potentially, MNOs will need more than 1,000 MHz of spectrum (in- This section goes deeper to examine the sectoral drivers cluding 3.5 GHz and mmWave spectrum), although prices of these challenges. Three key issues are discussed: spec- for spectrum cannot increase by that same factor as this trum and the unavailability of specific international mobile would crowd out investment. Three Indonesian MNOs do telecommunications (IMT) bands used to support mobile not have enough spectrum to be viable in the long term, broadband; infrastructure sharing as a constraint on invest- with Hutchison (Tri) having no low band spectrum at all. ments in fixed broadband access network; and the lack of competition, especially in the fixed broadband segment, At about 1.71 MHz per million people, Indonesia’s total which keeps costs high. IMT spectrum allocation is on the lower side in compar- ison to other emerging countries. (See Annex 2.A). This lack of useable IMT spectrum available to the MNOs has a direct negative impact on the quality of mobile inter- TABLE 2.1 Current spectrum assignments for mobile services 450 MHz / TOTAL NO OF OPERATOR 850 MHz 900 MHz SUB TOTAL 1800 MHz 2100 MHz 2300 MHz SUB TOTAL SPECTRUM SUBSCRIBERS Telkomsel - 2 x 15 30 MHz 2 x 22.5 2 x 15 30 105MHz 135 MHz 171.1 Million Indosat - 2 x 12.5 25 MHz 2 x 20 2 x 15 - 70MHz 95 MHz 59.3 Million XL - 2 x 7.5 15 MHz 2 x 22.5 2 x 15 - 75MHz 90 MHz 56.7 Million Hutchison - - 2 x 10 2 x 15 - 50MHz 50 MHz 30.4 Million Smartfren* 2 x 11 - 22 MHz - - 30 30MHz 52 MHz 13.3 Million Sampoerna 2 x 7.5 - 15 MHz - - - OMHz 15 MHz 0.2 Million Source: SDPPI, August 2020. Note: Coverage spectrum is below 1GHz and capacity spectrum is above 1GHz. As of April 2021, the SDPPI was running a spectrum acution for certain 2.3 GHz spectrum bands. Results have not been publicly announced as of this writing. 120 net services in Indonesia. This manifests itself in a number of particularly the 2.6 GHz and the 3.5 GHz bands, has meant ways, such as poor latency, call drops, inability to do voice over that the MNOs have been doing their best with their exist- LTE (VoLTE), etc. However, the most important way is how it ing spectrum allocations. Arguably they are over-investing in negatively affects broadband speeds as discussed above, espe- additional cell sites for 4G/LTE services when they could be cially download speeds as traffic to consumer mobile devices is investing in new 5G technology. typically asymmetrical. Higher spectrum prices also results in less capital being available for MNOs to invest in the network. As 5G is a more efficient technology, it offers a much lower With heightened additional demands for bandwidth due to cost per unit (and higher speed, better QoS and can be used larger share of population working from home (WFH) and for other innovations such as 5G IoT), instead of continued studying from home (SFH) due to the COVID-19 pandemic, capacity growth on current 4G/LTE systems. The sector needs the need for further useable spectrum is more acute. clarity, perhaps in the form of an IMT spectrum roadmap, to be better able plan and direct investments in the network in CHAPTER — 02 With the market move to 4G and future 5G services, the lack an optimal way. Without such clarity the investment becomes of available IMT spectrum in Indonesia is also putting pressure significantly less efficient, and does not result in improved on Indonesian MNOs to (i) merge or be acquired because their broadband speeds and lower latency offerings to Indonesian service offering is no longer competitive against the market consumers. In addition, it is necessary to consider the future leader and/or (ii) merge in order to ‘bulk up’ their IMT spec- switch-off of legacy 2G and 3G networks as other Asian and trum holdings in order to be able to provide a competitive ASEAN markets are doing or have done in order to free up service offering – namely higher speed broadband services. legacy spectrum for 4G and 5G services.40 For example, Sin- Knowing that there will be sufficient spectrum in the future gapore’s 2G services are already switched off, Malaysia’s 3G BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA to support 4G and 5G Service offerings, MNOs can confi- networks are scheduled to be switched off by the end of 2021 dently make the necessary long-term investments in digital and Vietnam is scheduled to switch off legacy 2G services in infrastructure. early 2022.41 In addition to quality, this IMT spectrum scarcity also con- The recently enacted Omnibus Law on Job Creation (No. tributes directly to the digital divide by constraining further 11/2020) includes a mandatory plan to implement the “Ana- rollout of mobile broadband. Harmonized spectrum below 1 logue Television Switch Off (ASO)” policy within a two-year GHz, especially the 700 MHz frequency band, is particularly timeframe. If successfully implemented, this would be a crucial useful. This is because the propagation characteristics of this step in securing the 700MHz spectrum band. The law also in- so-called digital dividend band provides greater geographic troduces reforms to optimize spectrum allocations and to allow reach/coverage and better in-building penetration relative to spectrum to be allocated to its most productive use. The actual higher IMT spectrum bands. With greater reach the num- implementation of these provisions in the law, however, will ber of cell sites needed to serve an area is significantly lower require significant inter-agency coordination and coordination than if higher frequencies such as 1800 or 2300 MHz are used, also among the various tiers of government. and this makes investments in rural rollout more attractive to MNOs. With just over 90 MHz in sub-1 GHz spectrum allo- cated to IMT purposes in Indonesia, insufficient spectrum is allocated for coverage purposes in the country. This crucial 700 → Infrastructure sharing, MHz band is currently occupied by legacy analogue television. competition and regulation If the 700 MHz spectrum were to be made available to the Major investment is required to expand fixed broadband MNOs in Indonesia, then 4G and/or 5G38 services could be rollout in Indonesia. However, typically between 70 and 80 provided to remote communities at significantly lower capital percent of the investment in fixed broadband is in passive in- cost and therefore with lower ongoing operating costs. This frastructure, such as ducts, poles, rights of way and civil works. would result in more affordable services to Indonesian con- Mobile broadband deployment across Indonesia has relied sumers. The use of 700 MHz would materially assist coverage instrumentally on tower-sharing, which has been well-estab- in rural areas. This is irrespective of the current use of LTE900 lished since the 2009 tower-sharing regulation. Independent because the coverage area (-100 dBm) for 900 MHz is materi- tower-sharing companies have brought efficiency to the tower ally smaller, namely 116.6 km² versus 161.7 km² for 700 MHz,39 deployments. Cross-sectoral passive infrastructure sharing albeit under theoretical and optimal deployment conditions. (between (rail)roads and fiber optic and electricity poles and fiber optic) has also occurred to some extent.42 But sharing While the unavailability of low frequency bands limits cov- between the telecom operators is not yet well established and erage expansion, the unavailability of specific high frequency lacks regulation. In fact, the Telecom Law (PP No. 52/53) pro- capacity bands retards preparation for 5G rollout. Indonesia hibited infrastructure sharing until recently. Duplication of has recently made the 2100 MHz and the 2300 MHz bands passive infrastructure is costly and unnecessary and, therefore, available through an auction process. But the continued un- making regulatory provisions to avoid this would be a signifi- availability of higher frequency capacity spectrum, including cant step toward expanding fixed broadband access networks. 121 FIGURE 2.16 FIGURE 2.17 While the mobile broadband space is …the FTTH market is more competitive… concentrated, with Telkom dominating market share SUBSCRIPTION SHARES OF VARIOUS PROVIDERS SUBSCRIPTION SHARES OF VARIOUS PROVIDERS CHAPTER — 02 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Source: Subscription data various sources. Source: Subscription data various sources. The dominance of an incumbent telecom operator (PT peting RANs in rural areas which, through sharing, enable Telekom), together with the lack of regulation, until re- three or more mobile providers to offer services without cently, to enforce sharing of ducts, poles and other passive requiring a fully duplicated RAN network rollout. infrastructure on fair, cost-based and non-discriminatory conditions, is a major factor in the lack of passive infra- Providers who want to overcome this by attempting to structure sharing among telecom operators. There are lay their own cables face additional challenges related to multiple fiber optic backbones across the main populated rights of way and local permits with local jurisdictions. islands such as Java-Bali and Sumatra, as well as subma- Local governments in Indonesia often view telecom ser- rine cables to Kalimantan and Sulawesi. While the fiber vice providers as a source of local revenues as opposed to optic backbone deployed by the USO Fund in the 57 most the providers of an important service that can drive local remote districts is accessible for all telecom operators, development by providing broadband connectivity. This the same is not true for the fiber optic networks owned is one of the major challenges, especially in going the last by PT Telkom. Lack of affordable access to those sections mile on connectivity. of the fiber optic backbone, including many districts in eastern Indonesia but also in rural parts of Sumatra, is a In addition to constraining rollout and keeping access key challenge for the newer telecom operators, and is also low, limited competition also affects other aspects of ser- a key impediment to the expansion of fixed broadband vice delivery by limiting market contestability. Unlike services in the country. the mobile broadband space, the fixed broadband mar- ket is very concentrated with the current FTTH rollout Sharing of active network will also increase competition dominated by PT Telkom (Indihome), with some smaller on mobile broadband in rural areas. This takes the sharing providers such as Biznet, MNC, My Republic, etc. in some a step beyond the current (passive) tower-sharing ar- major cities (Figure 2.16 and Figure 2.17). But for mobile rangements. The more remote areas of the country often broadband too, there are often just single providers in have a single provider, while if active rural access network many areas. This lack of competition limits incentives (RAN) sharing were to be allowed, that would create more to upgrade networks in a timely manner, and to remain options for competition. This could be in the form of a competitive in terms of pricing affecting quality and af- single shared RAN in the most remote areas or two com- fordability of services. 122 CHAPTER — 02 Although (fixed) telephony is generally a service under pres- sure, the restrictive licensing regime contributes to reducing → Digital Literacy and Skills competitiveness in the broadband market and limits entry. BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA A fixed broadband provider should be able to deliver a full In addition to the challenges related to hard infrastructure portfolio of services, not only broadband internet but also discussed above, another crucial dimension of the digital divide telephony, TV and many other value-added services. How- is digital literacy and skills. One can be plugged in and con- ever, the current regulatory regime limits this by requiring nected to the internet, but how effectively one can benefit from providers to bid for service specific licenses instead of a single it could depend crucially on the capacity for safe, secure and uniform license for all services. Local telephone licenses are productive use of the internet, and all the digital opportunities still formally limited to Telkom, Indosat and Batam-Bintan it can unlock. A detailed discussion of jobs and economic trans- Telekomunikasi, with Telkom effectively being the only pro- formation in the digital economy space, and the opportunities vider outside a specific area of Batam-Bintan. This is a major it has generated in Indonesia is presented in the next chapter. hurdle for fixed broadband competition, since customers are Here, we use data from a survey conducted specifically for this forced to keep a Telkom connection to maintain their tele- report to characterize the level of basic digital skills among the phony service. Number portability is another issue. Even if internet-using population in Indonesia. another provider were to be allowed to offer fixed telephony services over the broadband connection, customers would have Despite a growing literature on the topic, a consensus is yet to change telephone number to migrate to another provider to emerge on the conceptualization and measurement of dig- which is a major disincentive, especially for institutional users ital skills. Earlier literature tended to define digital skills in a (e.g., schools, hospitals, businesses). rather limited way as the ability to effectively use and navigate the internet.45 More recent work has expanded this domain Finally, all of this is fundamentally linked also to the regulatory to include elements that go beyond basic navigation skills, structure of telecom sector. The telecom sector in Indone- emphasizing, for example, the ability to understand and gen- sia used be regulated by the Indonesian Telecommunication erate content.46 Others have added the ability to communicate Regulatory Agency (Badan Regulasi Telekounikasi Indonesia, or and collaborate into the mix. UNESCO (2018) articulates a BRTI) set up by a 2003 ministerial decree. Even though the more holistic view of digital skills, defining it as the “ability to Ministry of ICT did play a role, this was considered to be a access, manage, understand, integrate, communicate, evaluate, and somewhat independent telecom regulator. In November 2020, create information safely through digital technologies for employ- however, the Government decided to dismantle BRTI and the ment, decent jobs and entrepreneurship.’ In order to operationalize establish this role within the Ministry of ICT. Post ratification, this definition, it also proposes six distinct competency areas: it is possible that this decision will need to be revisited or fur- information and data literacy, communication and collabo- ther institutional changes made given the commitment by ration, digital content creation, safety, problem solving, and Indonesia as a signatory to the Regional Comprehensive Eco- career-related competencies. Similarly, the EU defines digital nomic Partnership (RCEP) in late 2020 to inter alia support skills more broadly and includes more than basic or opera- competition in the provision of telecommunications services, tional skills of ICT devices and argues that digital skills have support access and use, competitive safeguards, and have an to encapsulate “the confident and critical use of information society independent telecommunications regulatory body.43 This will technology:” for work, leisure, learning, and communication.47 be crucial also from the perspective of recent legislative chang- es that allow for up to 100 percent foreign investment in tower In this report we adopt a measure of digital skills that draws business which was previously not possible.44 on both the UNESCO and EU frameworks and adapts it to 123 FIGURE 2.18 Core digital skills among internet-using adults in Indonesia CHAPTER — 02 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Source: World Bank Digital Economy Household Survey, 2020 the Indonesian context. It covers six components of core dig- applications, while only about two in three can compare in- ital skills: basic operation of ICT devices, applications and formation from different sources and discern the quality of internet, information and data literacy, communication and information. Similarly, on communications, nearly all inter- collaboration, digital content creation, digital security and net users are comfortable in communicating through instant safety, and problem solving. Within each of these areas, we messaging, but only one-half of the users reported being able assess skill at varying level of competency, going from basic to communicate via email and a little over one-third of the ones to more advanced capabilities. population has the more advanced competency of collaborat- ing through cloud sharing. Having defined digital skills in this way, we use a survey-based method to assess the level of skills among the internet-using Digital content creation, which covers not only the skill to population in Indonesia.48 Figure 2.18 summarizes the findings generate original local content and improve existing bodies of from this survey. Starting with operation of ICT devices, almost knowledge but also awareness of and good practice in applying all internet-using adults in Indonesia are comfortable operating copyrights and licenses, is a crucial dimension of digital skill. It some kind of ICT device, and over 80 percent are also able to also captures the ability to avoid fraudulent use of online content connect their devices to the internet, and to operate the brows- such as plagiarism. Among all internet-using adult respondents, ers and other applications to use the internet. A slightly lower around two-thirds know how to copy and forward information, share (72 percent) is also comfortable installing applications and while only about 46 percent know how to edit information from software on their ICT devices. Again, given the preponderance online sources. Consistent with the near-ubiquitous use of social of mobile broadband users among internet-using adults, it is media, around 81 percent of the internet-using adults reported reasonable to assume that most respondents are considering knowing how to compose and upload media (video, photos, etc.) mobile applications in responding to these questions. on the internet. In contrast, only one in five has advanced skills in the creation of original digital content, such as developing a The next competency related to information and data literacy website, putting in a watermark, etc. captures the ability of the users of digital technology to articu- late their information needs, search, use and share information Using digital technology and the internet comes with certain with a sense of judgement and responsibility. A little over four risks, such as the fraudulent use of personal information and in five internet-using adults appear comfortable using search breaches of data privacy and can potentially lead to many forms 124 FIGURE 2.19 Younger and more educated users have higher levels of digital skills CHAPTER — 02 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Source: World Bank Digital Economy Household Survey, 2020 of cybercrimes. Indeed, concerns about digital safety, security or unable to take the necessary actions to protect themselves and privacy are not just hurdles to digital technology adoption against them. For example, around 35 percent of respondents but also a key risk for all types of users. Indonesia has seen reported not reading data privacy policies when installing ap- an increasing number of cases involving breaches of data and plications and opening social media accounts. A much higher cyber-attacks. During January to April 2020, there were 88.4 share of social media users (75 percent) reported that they were million hacker attacks on Indonesian internet users49, a six-fold not well-informed about the terms and conditions of the social increase over same period last year.50 A recent high-profile media platforms they were using, with roughly half of the users case involved a security breach that led to the hacking and not being particularly concerned about it either. subsequent leaking of the personal addresses and financial in- formation of 15 million users of Tokopedia (one of the leading Digital literacy and skills also vary across different demograph- e-commerce platforms in the country). Similarly, in mid-2020, ic groups. Gender differences are not as large across all dimen- a hacker claimed to have breached the COVID-19 database and sions, except setting strong passwords, on which men were 10 stolen the COVID-19 patient data from various hospitals.51 percentage points more likely than women to be able to do so. Differences across socioeconomic status too were relatively Awareness of these evolving cyber risks and the ability to less pronounced, with individuals from wealthier households maintain basic ‘digital hygiene’ to protect and safeguard one’s slightly better able to search for information online, use email information and technology against these risks is also a key and also be aware of data privacy issues. Education and age pro- digital competency. This could include maintaining strong files, however, show a starker difference in literacy and digital passwords, using antivirus programs and firewalls to protect skills, with younger and more educated internet-using adults against scams and viruses, being aware of trustworthy web- being more digitally savvy than the relatively older and less sites and able to identify scams or phishing emails and other educated demographic (Figure 2.19). Even though the com- kinds of malware. About 70 percent of internet-using adults plementarity between level of education and digital skills has in Indonesia reported being able to restrict access to personal to be interpreted with caution, as it could be mediated by the information on their personal devices, and roughly the same differences in the kinds of work people do (with the tertiary ed- proportion also reported being able to identify and discern ucated more likely to be in white-collar office environments), phishing emails. This suggests that a sizeable share of the in- it also points to a potential opportunity to include digital skills ternet-using adult demographic is either unaware of these risks as part of the formal school curriculum. 125 2.5. Conclusion and Policy Implications CHAPTER — 02 This chapter has analyzed the underlying drivers of the ers, which accounts for as much as 80 percent of the invest- digital divide in Indonesia. The main findings of the ment in fixed broadband access networks, would provide a chapter are that, while investments in backbone infra- significant boost to fixed broadband rollout. This would be structure in recent years have translated into a remarkable critical support stable, good quality, and low latency fiber improvement in internet connectivity, a sizeable digital optic connections, which are crucial for high bandwidth divide still persists, with a significant share of the adult applications such as large-scale high-quality video, online population yet to have the basic wherewithal to connect education, business applications, school use and increasing to and benefit from the digital economy. The chapter also use of cloud-based solutions. These are all applications BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA finds that the lingering challenges on middle and last mile whose relevance has soared during the pandemic and is connectivity are related to spectrum scarcity/allocation, likely to remain high as Indonesia recovers and rebounds. limited sharing of active and passive infrastructure, and The recently enacted Omnibus Law on Job Creation (No. the lack of competition and regulation of the sector, which 11/2020) includes a stipulation to implement passive and affects not just access but also cost, affordability and the active infrastructure sharing. quality of the internet. Also, the recently enacted Omnibus Law on Job Cre- In recent years, the coverage of fiber optic network and ation includes a stipulation to implement the so called mobile broadband has expanded significantly reaching “Analogue Switch-Off ”, which would be a crucial step nearly all districts in Indonesia. However, fixed broad- in securing the 700MHz spectrum band. It also intro- band penetration is insufficient to meet rapidly growing duces reforms to optimize spectrum allocations and to demand for high-bandwidth applications. Mobile broad- allow spectrum to be allocated to its most productive use. band is more widely available but does not deliver the Similarly, the newly enacted law also introduces provi- same capacity, and QoS, and is not cost-efficient enough sions for infrastructure sharing. Speedy implementation for high bandwidth applications. While part of the chal- of the provisions of this law needs to be supported by lenge is boosting investment in network deployment in correct implementing regulations, and the provisions high cost, low population density parts of the country on infrastructure sharing are likely to require significant (including those areas where reliable electricity may be inter-agency coordination and coordination also among an issue), the inability to secure the digital dividend band the various tiers of government. (700 MHz) with the transition from analog to digital tele- vision remains a critical bottleneck in expanding coverage Finally, in addition to hard infrastructure, the chapter and access to rural and lagging areas. Spectrum scarcity has also presented some novel data on digital skills to also impinges on the quality of internet services for those highlight how the low level of digital literacy and skills already connected. The chapter reports measurable dif- constrain the ability of citizens who are connected from ferences in quality of experience, with strong evidence of fully reaping the benefits of digital technologies in a safe, mobile network congestion in the larger urban centers. secure and productive manner. Core digital skills related This reflects not only the need for additional investment to the basic operation of ICT devices, software and the in- in network infrastructure but also addressing spectrum ternet, data and information literacy, digital hygiene, etc., constraints. More spectrum below 1 GHz for cost-effi- are low among the internet-using adult population, and cient LTE coverage in rural areas and more spectrum highly correlated with educational attainment and overall above 1 GHz for cost-efficient LTE capacity would help socioeconomic status. This suggests that they represent a simultaneously address the challenges of broadening the critical and complementary frontier on closing the digital internet base and enhance the quality of the internet ex- divide and leveraging digital technologies for post-pan- perience of those already connected. demic recovery and addressing Indonesia’s medium run development challenge in an inclusive manner. Detailed Similarly, putting in place regulations to facilitate the shar- policy recommendations on digital infrastructure and ing of passive and active infrastructure among market play- digital skills are presented in Chapter 5. 126 Endnotes 37 Bisnis.com, 2019. 38 This is a decision open to the MNOs as all IMT spectrum is now technology neutral. Indonesia’s neighboring countries in Malaysia and Singapore will use the 700 MHz band for 5G services. Thailand has already launched 5G services utilizing the 700MHz spectrum band. 39 Refer to https://halberdbastion.com/technology/cellular/4g-lte/lte-frequency-bands/b8-900-mhz and https://halberdbas- tion.com/technology/cellular/4g-lte/lte-frequency-bands/b28-700-mhz 40 GSMA, 2020. 41 Vietnam.net, 2021. CHAPTER — 02 42 Sharing of the utility poles of PLN is of critical importance since PLN has a passive infrastructure of utility poles and electrifi- cation is over 98 percent. This implies the utility poles of PLN have over 65 million homes passed compared to Telkom which has about 30 million homes passed on fiber optic. This gap of about 35 million homes with access to electricity but beyond reach of the fixed broadband infrastructure can potentially be closed by sharing the utility poles of PLN. PLN has also entered the telecom market. 43 Refer to Chapter 8/Annex 8B of the RCEP, available at www.dfat.gov.au/trade/agreements/not-yet-in-force/rcep/ rcep-text-and-associated- documents BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA 44 Following the approval of the Omnibus Law, Presidential Regulation No. 10 of 2021 included provisions to eliminate the tele- com sector from the negative investment list (DNI). This is likely open up the sector to a range of new industry stakeholders able to invest in the sector, including deploying the digital infrastructure needed for 5G. 45 See examples at Krueger, 2006; Hargittai and Hsieh, 2012. 46 Van Deursen et al., 2014. 47 European Union, 2014. 48 There are some obvious limitations in this approach as it relies on the self-assessment of respondents. Ideally, these skills would be measured based on the actual testing of specific competencies. But implementing such tests would be expensive for the national scale in which we attempt this exercise here. 49 Indonesia State cyber and code agency (BSSN), 2020. 50 CNN Indonesia, April 2020. 51 Kompas, 2020. 127 References 1 Australian Department of Foreign Affairs and Trade. 2020. “Regional Comprehensive Economic Partnership (RCEP) text, Chapter 8/Annex8B”. Retrieved from https:/ /www.dfat.gov.au/trade/agreements/not-yet-in-force/rcep/rcep-text (Accessed: 03 December 2020). 2 BSSN. 2020. “Rekap Serang Siber (Januari – April 2020)”. Jakarta, Indonesia: Badan Siber dan Sandi Negara (Indonesia State cyber and code agency) (BSSN). Retrieved from: https:/ /bssn.go.id/rekap-serangan-siber-januari-april-2020/ (Accessed: 03 December 2020). 3 CNN Indonesia. 2020. “BSSN: Serangan Siber di RI selama Corona Naik 6 Kali Lipat”. Jakarta, Indonesia: CNN Indonesia. Retrieved from: https:/ /www.cnnindonesia.com/teknologi/20200423160048-185-496626/bssn-serangan-siber-di-ri- CHAPTER — 02 selama-corona-naik-6-kali-lipat (Accessed: 03 December 2020). 4 European Union. 2014. “Measuring Digital Skills across the EU: EU wide indicators of Digital Competence”. Retrieved from: https://ec.europa.eu/digital-single-market/en/news/measuring-digital-skills-across-eu-eu-wide-indicators-digital- competence (Accessed: 03 December 2020). 5 GSMA. 2020. “Network rationalisation – lessons from 2G and 3G migrations”. Retrieved from: https://www.gsma.com/ spectrum/resources/legacy-mobile-network-rationalisation/ (Accessed: 03 December 2020). 6 Halberd Bastion Radiofrequency Technologies. 2020. “4G LTE Frequency Bands”. Retrieved from: https://halberdbastion. com/technology/cellular/4g-lte/4g-lte-frequency-bands (Accessed: 03 December 2020). BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA 7 Hargittai, Eszter, and Yuli Patrick Hsieh. 2012. “Succinct survey measures of web-use skills.” Social Science Computer Review, 30(1), pp.95-107. 8 ITU. 2020. ICT Price Basket. Retrieved from https://www.itu.int/net4/itu-d/ipb/ (Accessed: 03 December 2020). 9 Bisnis.com Jatmiko, Leo Dwi. 2019. “The Next IndiHome Luncurkan Digital Movement #WujudkanDariRumah”. Jakarta, Indonesia: Bisnis.com. Retrieved from: https:/ /teknologi.bisnis.com/read/20190326/101/904302/the-next-indihome- luncurkan-digital-movement-wujudkandarirumah (Accessed: 03 December 2020). 10 Krueger, Brian S., 2006. “A comparison of conventional and Internet political mobilization”. American Politics Research, 34(6), pp.759-776. 11 Prabowo, Dani. 2020. “Data Pasien Covid-19, Dirahasiakan Pemerintah, Diduga Dijual Hacker...” Jakarta, Indonesia: Kompas.com. Retrieved from: https://nasional.kompas.com/read/2020/06/22/13191301/data-pasien-covid-19- dirahasiakan-pemerintah-diduga-dijual-hacker?page=all (Accessed: 03 December 2020). 12 Se, Chia. 2021. “No new 2G, 3G phones in Vietnam from July 1”. Vietnam.net. Retrieved from: https://vietnamnet.vn/en/ sci-tech-environment/no-new-2g-3g-phones-in-vietnam-from-july-1-708163.html (Accessed: 03 December 2020). 13 The Economist Inteligence Unit. 2020. The Inclusive Internet Index. London, United Kingdom: The Economist. 14 UNESCO. 2018. “A Global Framework of Reference on Digital Literacy Skills for Indicators”. Retrieved from: http:/ /uis. unesco.org/sites/default/files/documents/ip51-global-framework-reference-digital-literacy-skills-2018-en.pdf (Accessed: 03 December 2020). 15 van Deursen, Alexander J.A.M., Helsper, Ellen and Eynon, Rebecca. 2014. “Measuring digital skills: from digital skills to tangible outcomes”. Oxford, UK: Oxford Internet Institute, University of Oxford. 16 World Bank Group. 2018. Innovative Business Models for Expanding Fiber-Optic Networks and Closing the Access Gaps. World Bank, Washington, DC 128 Annex. ANNEX 2.A Indonesia in an International Comparison of Total IMT Spectrum Allocations TOTAL IMT MHz/ POPULATION SPECTRUM MILLION COUNTRY (M) ESTIMATE POPULATION COMMENTS AND PLANNED IMT SPECTRUM RELEASES CHAPTER — 02 March 2021 spectrum offered additional spectrum but not taken up India 1380 690 0.50 due to high reserve prices. Only 37 percent of spectrum offered at the auction sold Allocated 700, 2600, 3500 and 4800 MHz in late 2018 for 5G. China 1439.3 1237 0.86 Planning future release of mmWave spectrum. World Bank project on 5G Readiness including spectrum roadmap. Pakistan 220.9 256 1.16 Auction of vacant 1800/2100 MHz in 2021. Legislation to make 700 MHz available has passed Parliament. Indonesia 273.5 467 1.71 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Negotiations to clear 2.6 and 3.5 GHz bands. Plans to allocate spectrum in 3.5 GHz and vacant spectrum in 2.3 Bangladesh 164.7 453 2.75 and 2.6 GHz bands to support 5G in 2021 DR Congo 89.5 250 2.79 Not known Spectrum auction for 3.5 GHz band scheduled for June 2021. The Brazil 212.6 670 3.15 2.3, 3.5 GHz and mmWave bands are to be made available. Liberalisation of telecoms market with 2 new operators delayed due Ethiopia 114.9 370 3.22 to COVID-19. Now scheduled for 2021 Significant delays in releasing the 700, 2300 and 2600 MHz bands Vietnam 97.3 340 3.49 which are vacant and ready for use. 5G consultancy for the NCC recommends the release of additional Nigeria 206.1 820 3.98 IMT spectrum to support deployment. 2.6 GHz band was auctioned in November 2020. Plans exist for the Egypt 102.3 590 4.59 future release of 3.5 GHz band for 5G. mmWave auction was postponed in 2020 – but 400 MHz allocated as Russia 145.9 720 4.93 shared spectrum. 3.5 GHz is not available but 4.8 GHz band will be available for 5G. Iran 83.9 420 5.01 Currently preparing to auction 3.5 GHz band for 5G. An auction of 800MHz, 1.9GHz, 2.5GHz and AWS-3 bands is planned Mexico 128.9 800 6.21 for September 2021. While 600 MHz (70 MHz) and 3.5 GHz (150 MHz) are available for 5G. Turkey 84.3 570 6.76 Undertaking 5G tests in the 3.5 GHz and mmWave bands. After previously auctioning 600 MHz, mmWave etc, a successful USA 331.0 2780 8.40 auction for 3.5 GHz band was held in January 2021. Further spectrum releases are planned Third operator commenced March 2021. Plans for future IMT spectrum Philippines 109.5 985 9.00 reforms with World Bank support A successful 5G spectrum auction of 420 MHz (2.0 and 3.5 GHz Germany 83.7 1000 11.95 bands) was held in June 2019. Successful spectrum auction for 700 MHz and 3.6 GHz bands for 5G UK 67.8 1090 16.08 took place in March 2021. Successful 3.5 GHz reserved allocation and spectrum auction France 65.2 1100 16.87 undertaken in October 2020 Has already assigned 5G spectrum via way of a beauty contest in Japan 126.4 2800 22.15 2019 Italy 60.4 2000 33.11 Auctioned 5G spectrum bands in 2018 Thailand 69.8 3400 48.71 Successful 700, 2600 MHz and mmWave auctions in February 2020. 129 Spotlight 2. Key Findings from a Macroeconomic Take on Indonesia’s Digital Economy CHAPTER — 02 To assess the macroeconomics of the digital economy the digital sector as a whole for that decade. Since 2008, in Indonesia, a definition is first required. Consistent however, ICT manufacturing has seen a declining or flat with international approaches52, we focus our analysis performance, dragging on the performance of the digital on the ‘core’ of the digital economy, the digital sector, sector at large. As a result, Indonesia’s ICT manufacturing which includes: production of ICT consumer goods such is smaller than in many Asian peers, particularly in China, as computer hardware; the design, production, market- Japan, and Korea. ing of packaged and customized ICT software; digital BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA infrastructure, including 3G and 4G internet and other The contributions to value-added from digital media and telecommunications; and ICT services including consult- e-commerce are found to be smaller than that of the dig- ing and training. Next, given their relative importance ital sector but are growing over time. Using input-output in Indonesia and the availability of data, we look at two data, value-added from digital media is estimated to have components of the ‘digitalizing’ economy: digital media added in the range of 0.3-0.7 percent of GDP from 2000- and e-commerce. The macroeconomic impact of other 2015. Using a blend of firm-level survey-data, consumer aspects of the digitalizing economy, including industry survey data, and input-output data55, value-added from 4.0 and precision agriculture are outside the scope of our e-commerce is estimated to have added somewhere in analysis. the range of 0.3-0.5 percent of GDP in 2015, rising to 1.0-1.8 percent in 2019. This steady growth is projected to The contribution of the digital sector to the supply-side continue over the medium-term with the rapid expansion of the economy is estimated using input-output data. of e-commerce during and post the Covid-19 pandemic. This analysis, which is underpinned by a growth-ac- counting framework approach, looks at the ‘producers’ Looking at the demand-side, Indonesia has made im- of digital economy goods and services. On average, we pressive progress on digital adoption. At the turn of the find that the digital sector in Indonesia is slightly smaller 21st century, ICT represented less than 1 percent of the than that of advanced countries in the EU. Depending value of intermediate inputs of Indonesian businesses on on the input-output data source used and its breadth of average, approximately a seventh of that of India, South coverage53, we estimate the production side of the digital Korea, and Japan, and even lower compared to China and sector’s value-added was approximately in the range of Taiwan. While the digital economy’s production side re- 2.6-3.5 percent of GDP from 1998-2015. These Indo- mained flat as a share of GDP, Indonesia’s digitalization nesia estimates are about 1 percentage point lower than on the use-side grew seven folds from 2000-2010, reach- for the EU average. Compared to middle-income peers, ing 7 percent of the value of total intermediate inputs. Indonesia’s ratio of digital sector value-added to GDP is This share was higher than that of India, Japan, and Korea, approximately three fifths of the same ratio in China and at part with China, but still lower than Taiwan. However, India in 2014, and comparable to Brazil’s ratio54. Kenya, progress has recently stalled, with Indonesia’s share de- hosting one of the largest ICT sectors in Sub-Saharan clining in 2011, 2012, and 2014. Africa, provides another interesting case: value-added of ICT services including telecommunications reached 3.4 Indonesia’s strong digital adoption relative to peers percent of GDP in 2013 (World Bank, 2015). is driven by a few unique factors. In contrast to Asian peers, Indonesia’s digitalization has been driven by ICT The digital sector’s contribution to value-added has been telecommunications, with ICT manufacturing playing shaped by the contribution of ICT manufacturing. While a relatively much smaller role. Input-output data shows the intensity of telecommunications (share of value-add- that large investments in telecommunications were made ed to GDP) has remained largely flat from 1974-2015, a in Indonesia at a later stage than in Asian countries like rise in the intensity of ICT manufacturing in the 1990s India, Japan, and Korea, hence driving up Indonesia’s dig- was the main driver in the increase in the value-added of ital sector contribution. Other factors, including a higher 130 CHAPTER — 02 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA share of ICT imports in Indonesia, likely also drove up firms remain concentrated in West Java and Riau islands, the digital sector’s share of total intermediate inputs56. as measured both in terms of number of firms and the share of total ICT output. Economy-wide productivity-gains from digital adoption have been lower in Indonesia than in peer countries, due Econometric analysis finds strong cross-country evidence to both the characteristics of Indonesia’s digitalization of the productivity gains from the digital economy, but journey to-date and its concentration in a narrow set of our Indonesian findings are more equivocal. There is some sectors and provinces. Input-output analysis shows that evidence that phone ownership at the household level is investment in telecommunications is typically associated positively linked to GDP, GDP per capita, and GDP per with lower output growth effects than investment in ICT worker, but no statistically significant relationships for manufacturing or computer services. Thus, Indonesia’s other measures of digital economy adoption, including relatively higher share of telecommunications has result- teledensity (percent of PTSN-owning households), share ed in lower economy-wide output gains from the digital of internet users, and share of computer users. At the firm sector. Moreover, adoption of ICT telecommunications level, ICT manufacturing firms perform more strongly (and the even lower adoption of ICT manufacturing and than non-ICT manufacturing firms, recording substan- of computer services) has been heavily concentrated tially larger output, value-added, value-added-per-work- within a small number of ‘digital leaders’ at the sector er, and wages. However, ICT firms are also much more level. Here, there is clear evidence of a persistent de- likely to be foreign-owned and/or attract foreign invest- marcation between digital leaders and laggards, with ment. Once a set of firm characteristics are controlled Information and Communications, Finance, and Profes- for – in particular, the link with foreign ownership and sional Services, typically representing the leading digital investment – ICT firms are no longer robustly associated adopters. Indonesia’s digital economy similarly reflects with improved economic performance. Unfortunately, a great degree of concentration geographically. At the the limited availability of data (including small number top, ICT-use in DKI-Jakarta, Riau, D.I. Yogyakarta, East of ICT firms) precludes applying an experimental de- Kalimantan, and Banten, has been consistently higher sign to derive causal findings and makes it challenging than in other provinces. Firm-level evidence is similarly to statistically isolate the productivity impacts of foreign striking. Data from The Statik Industri surveys reveal knowhow and investment from the productivity impacts that just 1.2 percent of manufacturing firms in the sam- of digital technology. These data limitations do not imply ple engage in ICT manufacturing from the supply-side, that digitalization has not benefited the early adopters, and only 1.3 percent of manufacturing firms use ICT only that more research is needed with a larger sample of inputs in their production processes. These digitalized firms to isolate the magnitude of potential impact. 131 CHAPTER — 02 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Analysis of top-performing e-commerce firms provides sim- likelihood of selling both online and offline, and higher total ilarly mixed evidence of the impact of digitalization on pro- sales and profits. ductivity. Data from a recent World Bank-led survey of e-com- merce merchants that sold some of the most popular products Reviving the role of ICT manufacturing in the production-side, in 2016-18 is used to compare performance of e-commerce expanding digitalization across a broader set of sectors from firms against those firms surveyed in the World Bank Enter- the use-side, and investing in expanded national statistics prise Survey (WBES) of 2015. On the one hand, the results emerge as some of the main implications of this research. The show that the group of high-performing e-commerce firms growth in digital media and e-commerce, accelerated by the enjoyed higher profit-to-sales ratios than the average WBES Covid-19 pandemic, will add to the size of the broader digital firm and had higher export share of sales, despite being young- economy in Indonesia. But if the GOI seeks to enhance pro- er on average. There is evidence of a declining labor-cost share ductivity and output growth from the supply-side, historical of total-costs, and of growing productivity, as measured by val- Indonesian and cross-country experience to-date suggests ue-added-per-worker. Older firm groups are more productive that ICT manufacturing within the core digital sector can over time, and on average are more productive compared to play a significant role. From the use-side, accelerating digital younger firms. However, other findings from the survey call adoption within a broader set of sectors, a larger number of for a more nuanced understanding of the role e-commerce was firms, and across a wider geographic reach than West Java and playing in the Indonesian economy prior to the Covid-19 crisis. Riau Islands, is critical for carrying Indonesia’s digital journey The median value-added-per work estimated in the e-com- forward. The current Covid-19 crisis represents a unique op- merce sample in 2018 is lower than the median for the WBES portunity in this respect, helping to illustrate the potential firm in 2015. E-commerce sales, profits, and employment in gains from digital adoption in sectors such as education and the survey were all highly concentrated in a small number of health that have previously been reluctant to adopt technology. medium-sized firms that are mostly found in DKI Jakarta. Finally, upgrading Indonesia’s national statistics is critical to These firms are older, have higher sales and profit growth, and better monitoring and evaluating the size and contribution of combine high sales both online and offline, illustrating that a the digital sector and the broader digitalizing economy. Here, sharp dichotomy between online and offline commerce does Indonesia can benefit from a range of international, regional, not exist at the top. Indeed, the analysis finds that being fully and national efforts that are focused on this agenda.57 digitalized, defined as only buying inputs and only selling out- puts online, is associated with lower levels of sales and profits, while being larger (by employee size) and being older (years This spotlight draws on: Al-Rikabi, Jaffar, Ibrahim Khalil Rohman, since first established), is associated with a greater likelihood and Sean Hambali, “The Digital Economy in Indonesia: A Macro- to innovate, greater reliance on digital payments, a greater economic Perspective.” The World Bank (forthcoming, 2021). 132 Endnotes 52 The European Commission, the US Bureau of Eco- 56 Given Indonesia has one of the highest logistics costs in nomic Analysis, and the OECD estimate the ‘size’ of the region, importing a higher share of ICT intermedi- the digital economy through a focus on the digital sec- ate inputs than peers may have resulted in higher costs, tor, which is also referred to as ‘the ICT-sector’, ‘digi- driving up the share of ICT to intermediate inputs. tally-enabled infrastructure’, and other similar termi- Limits to current national account statistics may also nology. These definitions are supported by academic be playing a role in these figures. For example, while research, for example by Bukht and Heeks (2017). physical investments in the core digital sector goods 53 Our analysis uses three sources of input-output data such as telecommunications are well-captured by sta- for Indonesia: official statistics produced by the na- tistical agencies, other aspects of the digital economy CHAPTER — 02 tional statistics office, BPS; World Input-Output Data including ICT intangibles are difficult to measure and 2016; and EORA input output data produced by Len- so fall foul of ‘noise’ in the statistics. Since Indonesia’s zen, Kanemoto, Moran and Geschke (2012, 2013). digitalization journey has thus far had a higher share of Each of these three data sources bring different advan- telecommunications than peers, the mismeasurement tages and disadvantages. For example, official statistics problem arguably effects it less than in those other have a long coverage (1971 to 2010) and include data on countries. the digital sector and digital media, but the tables pro- 57 For example, in close collaboration with the OECD and duced are not consistent across years, making analysis other organizations, the G20 has produced a Toolkit BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA of historical trends difficult. for Measuring the Digital Economy, which outlines a 54 We present the size of the digital sector of China, India, measurement agenda for the digital economy, high- and Brazil based on the EC’s PREDICT database using lights statistical gaps, and suggests actions for im- the operational definition which is comparable to the provement. At the regional level, notable initiatives in- definition we use for our Indonesia estimate. In 2015, clude the EU’s digital scoreboard, the Digital Economy the average share of the digital sectors value-added in and Society Index, and the Monitoring Framework for each of these countries was 4.8%, 5.3%, and 2.8%, re- the Digital Economy and Society, and in Latin America spectively. and the Caribbean, The CAF Digital Ecosystem Obser- vatory. 55 Estimated of the output of e-commerce were tak- en from market data estimates provided by Statista, E-Marketer, and Google-Temasek. Estimates of the value-added of e-commerce firms are drawn from the World Bank 2019 e-commerce survey and cov- er 2016-2018 data. Given the survey sample design, which focused on e-commerce firms selling the most popular products on some of the largest Indonesian e-commerce platforms, this value-added estimate was taken to represent a ‘high’ scenario of the ratio of val- ue-added to output of the sector as a whole. Alterna- tive ‘low’ and ‘medium’ scenarios were estimated using input-output data on sectors relevant to e-commerce. The average range of value-added estimates included draws on these low, medium, and high scenarios, and represents an imperfect approximation of the sector’s contribution that nevertheless provides a useful start- ing point to understanding the size of e-commerce within the Indonesian economy as a whole. 133 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA CHAPTER — 03 134 Chapter ⬎ 3 Leveraging Digital Technologies for Better Jobs and Economic Opportunities CHAPTER — 03 3.1. Introduction BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA I n Chapter 1 we saw how some generally digital technologies can lower the cost of informa- applications of digital technol- tion and coordination, allowing firms and sectors to naturally ogies, particularly e-commerce extend the boundaries of the workplace, which is especially and digital ride-hailing services, common in sectors that create intellectual capital such as cre- are already beginning to gener- ative industries.58 ate notable consumer surpluses in Indonesia. A key question of A recent World Bank study in China shows that e-commerce interest is whether and how the has boosted income-earning opportunities for traditional- applications of these technolo- ly excluded workers such as youth, the semi-skilled, women gies are changing labor markets and the disabled. Imagine a homemaker selling pisang goreng and bolstering income-earning opportunities. This can be at madu (honey-coated friend banana) out of a small kiosk in one the extensive margin by creating more jobs, and at the intensive of Jakarta’s peripheries and reaching markets that she never margin by enhancing the returns to labor in the form of higher imagined through one of the food delivery platforms. Or a wages and salaries. Of particular interest is how these opportu- cluster of poor women weaving traditional bamboo baskets nities may differ for different segments of the population and in Indonesia’s eastern island of Flores who successfully brand how these forces may be compounding inequalities. We also their products and sell them online at scale in Jakarta with the already saw in Chapter 1 how, at the aggregate level, greater help of a social enterprise. Think of a low-skilled migrant who internet penetration is beginning to widen the skill premia, manages to use just his motorbike to gain a crucial foothold in by rewarding higher educated workers while leaving behind the urban labor market as a driver for one of Indonesia’s on-de- workers who are less well educated. This chapter delves more mand ride-hailing companies. Or a young family in Ubud who deeply into the topic to examine the underlying drivers of this succeeds in maintaining year-long occupancy in their apart- phenomenon, and examine the existing and future barriers ment through digital applications such as Airbnb. These are that will need to be overcome to enable the adoption of digital precisely also the kinds of micro-scale transformations ev- technologies to bring about economic transformation that is eryone, including policy makers in Indonesia, often have in truly inclusive. mind when thinking about the role of digital technologies in fostering greater inclusion. The potential clearly exists. Digital technologies can make business processes more efficient, and can boost the produc- This chapter looks at this landscape in depth, starting first tivity of firms and the workers that work in these firms. Digital by examining the level of adoption of digital technologies by technology applications such as e-commerce can, in principle, different types of firms, then homing in on two specific appli- also increase the productivity of household enterprises, which cations (e-commerce and gig jobs). This is followed by some account for a lion’s share of overall employment in Indonesia evidence and reflections on the winner-takes-all characteristic and are crucial new frontiers in expanding the middle class. of digital marketplaces and the implication that this may have In more urban areas, digital technologies can also help create for the realization of the promise of digital technologies, espe- more flexible forms of work that can activate vulnerable and cially in the low-skilled segments of the population. The final traditionally excluded workers such as women and youth. More section summarizes and concludes with reflections on policies. 135 3.2. Digital Adoption by Firms → The pandemic has FIGURE 3.1 accelerated the adoption CHAPTER — 03 Share of firms adjusting their business process of digital technologies, due to COVID-19, by type of adjustments though from a negligible base for small and micro firms59 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA The need to ensure business continuity during the period of limited mobility during the COVID-19 crisis forced many Indonesian firms to initiate and intensify their digital trans- formation. A survey conducted by the World Bank in June 2020 showed that only 36 percent of firms had managed to remain continuously open since March 2020. The same survey also showed that only about 40 percent of firms had temporarily closed but had reopened by June 2020, while roughly one in five firms had closed Source: World Bank Survey. and continued to remain temporarily closed. Of the 46 percent of firms that reported having to make some adjustments in their business pro- cesses to either remain open or to reopen after FIGURE 3.2 a hiatus, 42 percent reported that adoption of the internet, social media, specialized apps or Share of firms that reported adjusting in digital platforms had been their main coping response to COVID-19 strategy (Figure 3.1). The uptake of digital technologies was found to be higher among larger firms (90 percent), but also not entirely negligible for small to medium (58 percent) and micro firms (32 per- cent) (Figure 3.2 Figure 3.1).60 The same survey conducted in October 2020 shows that micro firms had begun to catch up with larger firms on digital adoption.61 The share of micro firms that used, started to use, or increased the use of internet, social media, specialized apps, or digital platforms increased by 25 percentage LEGEND points (pp), compared with 21 pp for small and medium firms, and 16 pp for large firms. How- ever, across both rounds of the survey, firms in low value-added services generally had a lower level of adoption relative to firms in higher val- ue-added services and in manufacturing. Source: World Bank Survey. 136 FIGURE 3.3 FIGURE 3.4 Internet use by non-agricultural Internet adoption is far higher enterprises is still very low, with among richer household enterprises much higher incidence among medium and large enterprises LEFT AXIS SHARE OF NON-AGRICULTURAL ENTERPRISES SHARE OF HOUSEHOLDS WITH ENTERPRISES WHO USE THAT USE THE INTERNET, BY SIZE THE INTERNET FOR THEIR BUSINESS OPERATIONS, BY INCOME STATUS RIGHT AXIS THE SHARE OF ENTERPRISES OUT OF TOTAL NUMBER OF ENTERPRISES (DOT) CHAPTER — 03 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Source: Economic Census 2016, WB staff calculations. Source: World Bank Digital Economy Household Survey, 2020 Part of this is likely to be driven simply by working from cent of all Indonesian non-agricultural enterprises used home arrangements. Workers in larger and more formal internet, with significant heterogeneity by firm size. Larg- firms with a more conducive enabling environment (e.g., er establishments were significantly more likely to have access to internet, digital knowhow) would have certainly adopted the internet (67 percent) to buy and sell, as well been better positioned to take advantage of these oppor- as to conduct other activities over the internet, while the tunities. However, evidence suggests that, among the firms adoption rate among micro-enterprises was significantly that intensified their digital footprints during the pandem- lower (4 percent). (Figure 3.3) More recent data suggest ic, many also used these technologies directly in marketing that internet adoption, especially by household enterpris- (82 percent), sales (75 percent), business administration es, could have increased to about 11.1 percent in 2018 and (51 percent), payments (42 percent), supply chain manage- 12.8 percent in 2020.64 However, even among household ment (33 percent) and other core business functions. Just enterprises, the richer ones exhibit a much higher level of on sales, for example, digital merchants with both an online adoption relative to the poorest ones (Figure 3.4). and offline presence significantly intensified their online activities during the height of the pandemic.62 Industry What determines technology adoption by firms? And estimates suggest that e-commerce volumes soared by as how do these drivers vary across firms of different type? much as a factor of 2.1 during the pandemic and are expect- These questions are relevant for policy, especially when ed to settle at around 1.7 times of pre-pandemic levels. 63 we think about the aggregate productivity benefits that This rising of demand can naturally be expected to draw these technologies might bring. Findings from a recent more suppliers into the digital ecosystem. survey show that, among household enterprises that have access to ICT devices to connect to the internet, a lack of aware- However, any increase in digital adoption that the pan- ness about the perceived benefits from using the internet demic brings will be from a very low base, especially for and a lack of digital skills are the most prominent reasons micro and small enterprises. Though somewhat dated now, for not actually doing so.65 This is true also for more spe- the economic census from 2016 revealed that only 5 per- cific forms of digital technologies such as e-commerce, as 137 CHAPTER — 03 Placeholder for Photo or Pull quote BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA discussed later in the chapter. The constraints faced by the higher-skilled workers and larger and more formal medium and large firms on the other hand may be entirely establishments having greater adoption rates. different, and could relate to, for example, the dynamism of the innovation ecosystem, management practices, the Overall, this observed pattern of digital technology adop- availability of talent and skills in local markets, and the tion by firms is one of the key reasons behind the stylized broader business environment in which these firms may fact noted in Chapter 1 that higher-skilled workers are be embedded.66 A careful assessment of these is left for perhaps benefiting more relative to workers with low- future work to tackle, but a macro-level analysis conduct- er-level skills. As such, this pattern of diffusion of digital ed for this report shows that aggregate productivity gains technologies in Indonesia could potentially be contribut- from digitalizing have been fairly limited in Indonesia due ing to an increase in overall inequality. This also reinforces to particular aspects of the country’s digitalization jour- the messages underscored in Chapter 2 on the need to ney, including its concentration within a narrow set of sec- address the critical challenges in making affordable and tors (see Spotlight 2 for key findings from this analysis). reliable internet available to all Indonesians. But it also raises important questions on the barriers to broader But, on the whole, the low level of internet adoption by application and utilization of these technologies in pro- enterprises, especially the micro and small enterprises, ductive endeavors. Technologies such as e-commerce, also translates into low adoption among workers. This is digital financial services, and the portfolio of other ser- especially true given the fact that these micro and small vices (ride-hailing, food delivery, etc.) offered by plat- enterprises jointly account for almost 75.3 percent of In- forms such as Go-Jek and Grab, do not require advanced donesia’s non-agricultural employment. In 2019, only reading and numeracy, as well as high-order cognitive 27.2 percent of all workers reported using internet at skills. As such, these low-skill-biased worker-enhancing their work. This was the highest among tertiary educat- technologies should be amenable to greater adoption by ed workers (78.1 percent), followed by much lower use market-size-constrained micro and small enterprises and among workers with just lower secondary education (17.7 less-skilled workers everywhere in the country. What percent).67 Thus, on the one hand, the overall level of constrains these enterprises and workers from adopting internet adoption among workers and firms is low. On and benefiting from these technologies? The rest of the the other hand, there are sharp heterogeneities in adop- chapter examines the specific case of e-commerce and gig tion among different types of workers and firms, with jobs in Indonesia. 138 3.3. E-Commerce in Indonesia: Patterns, Growth and the Barriers to Adoption CHAPTER — 03 As we saw in Chapter 1, some welfare benefits from e-com- e-commerce landscape. Among those engaged in e-commerce, merce are already starting to be realized, primarily through the about 71 percent reported selling exclusively via social media and consumption channel. Consumption baskets have expanded chat messaging apps, 3 percent sell via platform only, while 26 with heterogeneous product varieties made available by e-com- percent use a blend of the two.70 This means that e-commerce merce and, not only are consumers who buy online directly in Indonesia remains dominated by consumer-to-consumer (or able to avail themselves of better prices relative to their offline C2C in industry parlance) business, which makes it more similar options, but greater e-commerce intensity is also beginning to to TaoBao Marketplace, as opposed to the business-to-business affect price growth of more intensively traded commodities (B2B) ecosystem similar to Alibaba or even the business-to-con- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA differentially. Ideally, the next step would be to examine the sumer (B2C) one as TMall. These C2C digital merchants are extent to which this new digital economy also enhances welfare generally regarded as more casual market participants compared through the income channel. But as we also saw in Chapter 1, with the B2C or B2B ones. They also operate at a smaller scale not only is e-commerce intensity higher in more prosperous and potentially also at a lower level of productivity. parts of the country, but both buying and selling are also more prevalent among the relatively better-off segments of the in- Although Indonesian men are almost twice as likely to use inter- come distribution. Careful direct identification of the impact net at work, internet-using women are slightly more likely than of e-commerce on household income and consumption would men to be engaged in e-commerce activities. Among women require parsing out the confounding effect of higher incomes who use internet at work in their primary jobs, 37.8 percent of driving e-commerce growth. them are active online sellers (as opposed to 34.8 percent of men) (Table 3.1). E-commerce is more common as a secondary Instead of following this direct route, here we present a de- job, suggesting that it plays a useful role in supplementing family scriptive analysis of some of the more salient aspects of e-com- income, especially by women and youth. E-commerce incidence merce in Indonesia to highlight some suggestive evidence of is higher when it is a secondary job (50.4 percent), rather than how it could be benefiting different types of workers. In addi- when it is the main one (35.8 percent), especially for women (57.7 tion, we also identify the key barriers to growth of e-commerce vs. 46.8 percent for men in the secondary job) and youth below across Indonesia, and adoption by micro and small enterprises. age 25 (58.6 vs. 49.5 percent for those above 25 in the second- ary job) (Table 3.1). Interestingly, e-commerce engagement in → E-selling is much more social media is higher for the part-time/underemployed, while e-commerce engagement in platforms is higher for the full-time, commonly done through social confirming the fact that perhaps the more serious sellers prefer media and chat applications than to operate through platforms. In a recent survey, 65 percent e-marketplaces, providing a useful of merchants selling exclusively through social media cited its ease of use as the main attraction. In the same survey, limited source of supplemental income, understanding of digital marketing strategies and reservations especially for women and youth, regarding regulatory compliance71 were cited as the main reasons and a useful pathway for women for not joining the e-marketplaces by 39 and 33 percent of social media sellers, respectively.72 re-entering the labor market The female labor force participation rate in Indonesia has re- In 2019, around 13.2 million out of a total of around 127 million mained persistently low, hovering at around 50 percent for the employed workers (10.4 percent) in Indonesia were engaged in past three decades. It has also been identified as one of the key e-commerce activities as either their primary or secondary job.68 challenges to meeting Indonesia’s development aspirations of In terms of contributions to overall employment, this number becoming a high-income country.73 By expanding opportunities is higher than in China, where e-commerce accounts for 5 per- for women in the labor market, e-commerce could be helping cent of total employment.69 A peculiar aspect of e-commerce in Indonesia at least partially address this challenge. Evidence sug- Indonesia is the dominance of social media and chat apps in the gests that, among all women, e-commerce engagement (selling) 139 TABLE 3.1 E-commerce activity rate is higher when a job is a secondary rather than a main job, especially among women PERCENTAGE OF WORKERS WHO USE THE INTERNET AND SELL VIA SOCIAL MEDIA/ PLATFORM BY MAIN/SECONDARY JOB AND BY GENDER, AGE, LOCATION MAIN JOB SECONDARY JOB N SELLING VIA SELLING SELLING N SELLING VIA SELLING SELLING PLATFORM VIA SOCIAL VIA BOTH PLATFORM VIA SOCIAL VIA ONLY MEDIA ONLY ONLY MEDIA ONLY BOTH CHAPTER — 03 All Sample 34,113,645 1.2% 25.4% 9.3% 3,167,860 1.6% 37.5% 11.4% GENDER Female 12,562,412 1% 27.2% 9.6% 1,057,224 0.8% 43.4% 13.5% Male 21,551,233 1.3% 24.4% 9.1% 2,110,636 2% 34.5% 10.3% AGE GROUP 15-24 years old 5,914,634 1.2% 25.2% 11.6% 349,548 1.4% 41.6% 15.6% 25 years old and above 28,199,011 1.1% 25.5% 8.8% 2,818,312 1.6% 37.0% 10.9% BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA LOCATION Urban 25,997,782 1.3% 26.3% 10.6% 2,231,513 1.8% 39.4% 13% Non-urban 8,115,863 0.8% 22.8% 5.1% 936,347 1.2% 32.9% 7.7% WORKING HOURS Less than 20 hours per week 2,941,743 0.7% 32.2% 7.4% 2,523,141 1.2% 37.6% 10.8% 20 - 40 hours per week 12,471,051 0.9% 23.4% 8.1% 609,026 2.9% 37.1% 14.1% More than 40 hours per week 18,700,851 1.4% 25.7% 10.4% 35,693 5.6% 33.1% 8.7% Source: SAKERNAS, August 2019, World Bank staff calculations. Note: Samples are workers who use internet in their current respective job. TABLE 3.2 E-commerce helps women return to work after giving birth PERCENTAGE OF WORKERS WHO USE THE INTERNET AND SELL VIA SOCIAL MEDIA/PLATFORM BY REASONS OF QUITTING MAIN JOB SECONDARY JOB N SELLING VIA SELLING SELLING N SELLING VIA SELLING SELLING REASON OF QUITTING WORK PLATFORM VIA SOCIAL VIA BOTH PLATFORM VIA SOCIAL VIA ONLY MEDIA ONLY ONLY MEDIA ONLY BOTH WORK TERMINATION 104,062 2.9% 29.5% 5.4% 5,398 0.0% 71.7% 0.0% BANKRUPTCY 50,579 4.1% 23.7% 6.8% 6,557 11.5% 61.4% 4.5% UNSATISFACTORY INCOME 464,977 2.8% 29.6% 9.8% 51,286 2.8% 41.9% 12.0% UNSETTLED WITH WORKPLACE 282,299 1.2% 30.6% 12.2% 25,029 3.6% 32.4% 9.4% EXPIRED CONTRACT 341,104 1.5% 23.9% 11.8% 27,311 2.5% 26.7% 20.3% UNMATCHED SKILLS 49,065 1.8% 19.8% 4.6% 2,820 0.0% 70.1% 9.2% PREGNANCY/MATERNITY 14,160 1.3% 42.2% 14.6% 86 0.0% 0.0% 100.0% DOMESTIC WORK 47,414 4.5% 41.2% 14.6% 4,213 0.0% 45.4% 13.1% OTHERS 209,274 1.6% 29.9% 8.2% 23,614 3.1% 41.6% 20.5% Source: SAKERNAS, August 2019, World Bank staff calculations. Note: Samples are workers who are not in their previous main job within the past one year and use the internet in their current respective job. 140 FIGURE 3.5 E-commerce has touched all parts of Indonesia but is still largely Java-centric PERCENTAGE SHARE OF E-COMMERCE MERCHANTS IN THE POPULATION WHO SELL IN SOCIAL MEDIA/CHAT MESSAGING APPS AND PLATFORMS, AS THEIR MAIN JOB A. SOCIAL MEDIA CHAPTER — 03 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA B. PLATFORM Source: SAKERNAS, August 2019, World Bank staff calculations is highest especially among those who are primarily engaged in housework. Moreover, e-commerce also appears to be provid- → The geographic spread of ing a viable pathway for women re-entering the labor market e-commerce in Indonesia is by providing opportunities to those who may not be in their constrained largely by internet previous jobs for a variety of reasons. About one-third of women who were not working and who experienced a job termination in connectivity and challenges 2018 attributed their labor market exit to domestic work, and a related to logistics further 13 percent suggested pregnancy was the main reason.74 As such, ensuring those women with a labor market connection While e-commerce has spread to all parts of the country, it at some point in their lives stay engaged is a key policy entry remains Java-centric to a large extent. As discussed above, the point. About 58.1 percent of internet-using women who had proportion of e-sellers using social media and chat messaging to leave their previous jobs because of pregnancy/maternity or apps is much higher than that of digital platforms consistently to return to domestic work were engaged in e-commerce, sug- across all parts of the country (Figure 3.5). Focusing specifical- gesting that e-commerce provides one avenue for women to stay ly on the size and volume of transactions that occur through productively engaged as they exit other forms of jobs to take on digital platforms, we examine the question of what determines greater responsibilities at home (Table 3.2).75 the evolution of e-commerce penetration and intensity in the 141 country, over time as well as across space. We do this using data populous places, suggesting that e-commerce transactions made available to the World Bank by two of Indonesia’s largest may be smaller and more frequent in places better plugged e-commerce platforms, which allow for the construction of into the ecosystem where the logistics are both cost and time several measures that capture both the penetration and the efficient. Consumers in less populous (contentiously, more intensity of e-commerce, such as the number of buyers (or remote) places tend to ‘bundle’ or ‘bunch’ their purchases to sellers) per capita, number of purchases (or sales) per buyer save travel time and delivery costs, or to order higher-value (or seller), the gross merchandise value per buyer (or seller) products so that the delivery costs will be small in comparison and the gross merchandise value per purchase (or sale). to the value of the products. Based on some of the stylized facts already presented in Chap- Logistics as a bottleneck for e-commerce is consistent with the CHAPTER — 03 ter 1 and re-affirmed by Figure 3.5, one hypothesis is that the broader challenges that Indonesia faces in moving goods and more populous and prosperous provinces of Indonesia are the commodities across its far-flung geography. The availability places where e-commerce penetration and intensity are higher. and reliability of transport infrastructure is the first and per- A related hypothesis is whether this relationship is indepen- haps the major impediment to a smooth logistics chain. There dent or mediated through other relevant characteristics of are considerable imbalances of land, sea and air connectivity, places such as internet access and connectivity, which would for example, between well-developed cities around Jakarta, be associated with, among others, the cost of logistics to move Surabaya, Medan and Denpasar (Bali), and to secondary, ter- goods, human capital, etc. In other words, is it a question of tiary and rural areas, resulting in high variability in logistics BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA demand (which is a function of population size and average costs. A 2017 World Bank survey of manufacturers on logistics income levels), or is it a question of other factors, which could performance in Indonesia showed that companies located in be correlated with income and population size? Jabodetabek area (greater Jakarta) managed to receive their goods in full and in good condition, and to maintain logistics Full results of this analysis are presented in Annex Tables costs at around 12 percent of the costs of sales, whereas com- 3A-3B76. Here we summarize just the main findings. First, panies in Kalimantan incurred up 30 percent sales in logistics e-commerce penetration is clearly positively correlated with costs. In addition to direct logistics costs, delays in receiving both income and population; provinces with higher per capita or sending products are another source of indirect logistics consumption levels and those with larger populations have costs for manufacturers. higher proportion of buyers and sellers in the population. However, when we account for other factors, such as access to All of the Indonesia’s logistics performance is hampered by the level of internet, electricity, cost of logistics, and average bottlenecks, from the first to the last mile. In addition to human capital level of the population, the role of income disap- weaknesses in transport infrastructure, the availability and pears. This suggests that it is not income levels per se but other competitiveness of logistics service providers (LSPs)77 is a key digital as well as analog enablers also correlated with income challenge, especially outside of the main economic areas. On that drive e-commerce penetration. Furthermore, when we maritime connectivity, the lack of regular inter-island mari- look within provinces to see what explains the observed growth time routes affects timeliness of goods transit, with domestic in e-commerce penetration, we find that level of internet access shipping lines channeling regular calls mostly on the most among the population and the cost of logistics play decisive lucrative shipping lanes (between the main ports) and devel- roles. In other words, e-commerce penetration increased more opment of roll-on roll-off shipping (RoRo), which has proven rapidly in provinces in which internet access expanded and the cost effective in several maritime countries, being constrained. cost of logistics declined over this period. On air connectivity, airlines usually give priority to passengers compared with goods, which reduces the likelihood of goods Second, the e-commerce intensity measures broadly follow the being delivered on time to the next destination. On land con- same pattern with some differences. While logistics still emerge nectivity, the lack of temperature-controlled trucks and quality as important for measures such as gross merchandise value warehousing services across Indonesia prevents expansion to (GMV) per buyer (and seller), the association with changes secondary and tertiary cities, especially for perishable goods. in internet access is less pronounced. This suggests that, while expanding internet can draw more people into e-commerce at More specifically on the last mile, a key challenge is also the the extensive margin, it may not have a similarly strong effect lack of standard addresses and precise postcodes, which pre- at the intensive margin. Other factors might be important in vents implementation of productivity enhancement tools, expanding the number of sellers in an area. Also, population such as automation of sorting facilities to speed up routing size has a much stronger positive effect on all except one in- of goods, and utilization of route optimization software to tensity measures (GMV per order) related to sellers, poten- map the most efficient route for the courier. A complex regu- tially reflecting the fact that a bulk of the demand remains latory framework on logistics services combined with uneven concentrated in the more populous areas. Interestingly, GMV distribution of population across the archipelago drive these per order for both buyers and sellers is smaller in the more logistics challenges. 142 BOX 3.1 E-Logistics for Farmers: The Case of TaniSupply and TaniHub TaniHub, a leading agricultural start-up, enables farmers to requires temperature-controlled assets covering trucks, connect directly to buyers and supports them by providing warehouses and other temporary storage areas. This is es- key services such as access to finance and managing their pecially time-critical for fishery and livestock products. supply chain (by managing their orders and handling their TaniHub is planning to expand beyond Java,79 but the lack logistics). Its subsidiary, TaniSupply, handles the logistics of LSPs across the archipelago will lead it to open its own side, collecting, storing, packing and delivering products warehouse and sorting facilities. Second, a fast distribution CHAPTER — 03 from 30,000 small-holder farmers across Java-Bali. Tani- of products to the consumer is needed to ensure the fresh Supply is developing several key success factors influencing products arrive on time and in good condition. Earlier in the supply chain efficiency. First is the ability to handle 2020, TaniHub Group announced an MOU with FastPOS, temperature-controlled products across the supply chain an e-logistics startup subsidiary of Pos Indonesia. to reduce the loss from the farm to the final buyer. This BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA → Digital technologies are also consisting of mostly homemakers and micro enterprises, the availability of same-day delivery has helped sellers significantly beginning to disrupt the logistics increase their gross merchandise values by 30 percent, and business to overcome some increase their productivity and returns-on-investment by up of these challenges, but these to four times. 70 percent of Paxel customers have switched to same-day delivery. The same-day delivery between cities is innovative efforts are largely also providing a boost to ready-made food shipments between limited to large urban areas and major cities. The increasing availability of temperature-con- their peripheries trolled transportation and storage enables sellers to expand their markets, going beyond their district to other districts in a metropolis and to other cities. The proliferation of C2C and B2C e-commerce in Indonesia has given birth to a new type of “consumers” of logistic services, Some of these e-logistics solutions are being developed by the with no expertise in logistics themselves. Many of this new type platforms themselves. A case in point is Mitra Bukalapak, the of consumers are MSMEs. E-commerce buyers and sellers are online-to-offline (O2O) arm of the e-marketplace Bukalapak. increasingly ready to pay more for quicker deliveries.78 The Mitra Bukalapak is pitched as a dedicated solution to support quicker the delivery reaches the buyer, the quicker the payment warung (mom-and-pop stores/eateries) and provides a suite of is received, and the faster the business can grow. The challenge of integrated services that includes sourcing directly from brand low cost, with fast and reliable last mile delivery, is increasing the owners, storage, and other added-value services to support pressures on conventional brick-and-mortar logistics compa- cash-flow management, including utility bill payments. A key nies. The increasing number of small-sized, non-standard parcel aspect of this support to warung owners is facilitating next- shapes, the variety of pick-up and drop-off points, and the low day delivery, which is crucial for these small establishments accessibility of such points have led to higher cost per parcel. operating on tight budgets and for whom one day’s revenue All of these have led to new logistics business models and strate- determines the next day’s production capacity. While the exact gies, with the emergence of specialized e-logistics companies, an percentage of successful next-day deliveries is not available, adaptation of conventional freight-forwarding companies and Bukalapak reports that its e-logistics solutions have led to a express services to capture a share of the e-commerce market. substantial reduction of the rate of late deliveries. Not only has the evolution of e-logistics empowered MSMEs, but it has By adopting strategies such as the Uber-like model and the also empowered farmers (see Box 3.1). The next milestones 4th Party Milk Run, e-logistics companies are beginning to for e-logistics companies are to continue improving their ef- offer low-cost, same-day and next-day delivery services within ficiency, expand to same-day delivery, at the same cost across and across metropolitan cities. According to a survey under- Java, and further reduce delivery time outside of the Java-Su- taken by an e-logistics company, Paxel, of its own merchants, matra-Bali. 143 → Low adoption of digital payments, which is constrained on the demand side by issues of awareness and trust, and on the supply side by regulatory hurdles, lack of appropriate infrastructure and small margins for providers, is also another important factor limiting growth of e-commerce FIGURE 3.6 Another key bottleneck in the expansion of e-commerce in Indonesia is the low level of adoption of digital financial services (DFS) Adoption of digital financial services is among households. To begin with, financial extremely low CHAPTER — 03 exclusion is a major issue, with only 52 percent of households having access to a bank account (Figure 3.6). Only 43 percent of all households have a personal bank account and, while 9 per- cent in total are users of DFS (which we define All HHs here as online banking and mobile money ser- vices), 34 percent are exclusive users of offline banking services. Only one in five households BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA in the bottom 40 percent of the socioeconomic distribution are users of DFS. Half of those in the bottom 40 percent are unbanked. Overall, With access to only 0.5 percent of households are users of more 48% 52% account advanced DFS products, such as credit, remit- Financially tances, insurance, etc.80 excluded Access to and trust in digital transactions ap- pear to be among the main reasons that Indo- nesians are hesitant to get on board with e-com- merce. Consistent with our discussion above, a Personally own sizeable share cites the hassle of return (which is 9% 43% account related to the ease of logistics) and not having Access others' used the internet enough as the main reasons account for not buying online (Figure 3.7). But not hav- ing a bank account, especially among those in the bottom 20 percent, and the lack of trust in online transactions also show up as important barriers. While trust can have multiple dimen- DFS Users sions, including whether a purchased item will be delivered in good condition and whether a 34% Only have bank account 9% return will be accepted in case the customer is not satisfied, a bulk of it could also relate to having to make online payments, and often in advance of the package being received. This is certainly confirmed in the finding that over 50 percent of those who buy online still prefer to Advanced users pay cash on delivery (Figure 3.8). 0.5% of DFS For more sophisticated would-be users, issues of trust could also be rooted in the understanding of risks related to data governance and privacy, cyber security and operational risk, financial in- tegrity, and several others.81 Concerns regard- ing unauthorized data disclosure by any of the several entities involved could discourage this Source: World Bank Digital Economy Household Survey, 2020 segment from adopting DFS. And these are not 144 FIGURE 3.7 entirely misplaced; complaints coming from financial consumers, as well as banking sector itself, on these breaches have been increasing Access to and trust in digital financial lately. While the Financial Services Author- transactions among the main reasons for ity of Indonesia has been trying to develop a not buying online stronger legal and regulatory framework for this, consumers still regard the disclosure of PERCENTAGE SHARE OF HOUSEHOLD HEADS WHO DO NOT BUY personal ID to financial service providers as a ONLINE major risk. Provision of a legal basis, such as passing the current draft of the Personal Data Protection Policy Law in the legislature, is nec- essary to promote DFS adoption in Indonesia. CHAPTER — 03 In addition to the issue of trust, there is a sig- nificant knowledge and awareness deficit which, somewhat surprisingly, is almost as prevalent among the financially excluded as it is among the traditional bank account users. Roughly half of the banked, as well as unbanked, households report not fully understanding the benefits of BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA DFS, or not knowing how to use it, or more surprisingly, especially for those already with bank accounts, never even having heard of these services. This highlights the key role for agents in helping reach the unbanked as the face-to- face personal interactions are ideally suited to overcome the trust and knowledge gaps. There has been significant growth of agent net- works in Indonesia, with more than 4 million FIGURE 3.8 agents contracted under banks or e-money institution providing financial services access to the majority of population.82 The agents Among those who buy online, a majority of Laku Pandai (LP, or branchless banking) still prefer cash on delivery and Layanan Keuangan Digital (LKD, or dig- ital financial services) have played a signifi- PERCENTAGE SHARE OF HOUSEHOLDS WHO BUY ONLINE cant role in facilitating financial access to the underserved population, especially those on low incomes and/or residing in remote areas. These agents have contributed to some 27 mil- lion basic saving accounts (BSA) opening at LP agents and more than 12 million mobile money accounts registered at LKD agents, suggesting that these networks could have played a major role in expanding financial access, especially to the unbanked. While the use of agents has indeed great po- tential in boosting furthermore the adoption of DFS in Indonesia, there are several challenges in realizing the expected outcome. The first challenge is the small profits caused by low de- Source: World Bank Digital Economy Household Survey, 2020. mand of DFS transactions. The large expansion Note: Figure 3.7: Lack of trust in online transaction category can be of agent network is not automatically accom- further broken down into three categories: (i) discomfort with the concept of paying goods before receiving it; (ii) discomfort associated with panied with high transaction volume. The me- handing financial information; and (iii) discomfort rising from handing any personal private information. dian of one agent’s transactions per day is only four in Indonesia. This is very low compared 145 with other countries, which record a range 15 to 35 transactions per FIGURE 3.9 day. Low transaction volume generates low profits to agents. The median of agent’s monthly profit in Indonesia was US$15 for the Not having accessed DFS is Jabodetabek area and only US$5 for other rural and other urban associated with lack of awareness areas. According to the same survey, Tanzania and Bangladesh had the highest median at US$70 and US$57, respectively.83 The low and knowledge volume of DFS transactions in agents shows low demand for DFS. A 2016 survey highlighted that the low demand of DFS agent services PERCENTAGE SHARE OF HOUSEHOLDS was caused by a lack of awareness of the agent services and location THAT DO NOT USE DFS (40 percent), difficulty of use (21 percent), and the lack of relevant use cases (18 percent). CHAPTER — 03 Another challenge in expanding agent networks in Indonesia is related to infrastructure. The suboptimal infrastructure includes inadequate physical and ICT infrastructure, as well as limited in- teroperability of payment schemes. While DFS providers could have a strategic rationale to provide access to rural areas, the operational and economic challenges might deter them. Bank-led agent net- works might be interested in expanding to rural areas with a lower cost distribution and corporate social responsibility mindset, but BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA they still require proximity to an existing branch network to be able to manage the agents. Under the current setup of the agent-bank partnership, the bank is required to directly manage its agents. Banks are still not allowed to have a third-party service to manage their agents. The non-bank e-money issuers such as e-commerce platforms would like to also have customer acquisition and eco- system expansion, but view the operational challenges, especially connectivity, as being too great for deep rural growth. Source: World Bank Digital Economy Household Survey, 2020. 146 The third challenge is the legal and regulatory frame- November 2016, which was later stipulated in Presidential work that constrains the growth environment. While Regulation No. 74/2017 on the E-Commerce Roadmap. the current regulation has enabled new players and new The roadmap aimed to build a more efficient commercial approaches to offer DFS, several restrictive require- institution and ecosystem in seven steps, namely: (i) sim- ments exist. For example, Bank Indonesia Regulation plifying and expanding access to finance; (ii) providing No. 20/2018 regarding Electronic Money places an ex- tax incentives; (iii) providing consumer protection; (iv) plicit restriction on legal individuals becoming agents of increasing human resource capabilities; (v) improving non-bank financial service providers (FSPs). Lifting this the national logistics system (Sislognas); (vi) accelerating restriction could contribute significantly to the number the development of communication infrastructure; and of LKD agents, since half of the registered FSPs for LKD (vii) improving cybersecurity. Through ecosystem de- are non-bank. Financial Services Authority Regulation velopment, the roadmap targeted the creation of “1,000 No. 10/2014 on Branchless Banking is also considered too technopreneurs with a business valuation of US$10 billion CHAPTER — 03 detailed and overly restrictive for the BSA requirements, and an e-commerce value of US$130 billion by 2020.” as well as agent operations. For instance, the distinction That objective clearly was not met but, as the analysis in of seven different categories of agent could be simplified this chapter so far points out, some of the pathways to to avoid fragmenting the market and limiting business grow and develop e-commerce in Indonesia identified by autonomy. Also, the provisions for BSA are too detailed the roadmap, namely boosting internet connectivity, re- (both regarding what is permitted and what is prohibit- ducing the cost of logistics, expanding access to DFS and ed), with the result that banks are left with very limited putting in strong consumer protection, data protection autonomy to compete effectively in offering it, and find and cybersecurity laws to build trust in the ecosystem, BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA it unattractive as a profit-making business. remain just as valid. Finally, the licensing and reporting requirements for LKD The roadmap further highlighted more recent policy con- and LP agents are too complex and convoluted. Currently, versations on e-commerce in Indonesia concerning prod- the applicants of both LKD and LP must undergo two ucts dominated by cross-border transactions, and equal different and usually repeated administrative process- treatment of foreign and domestic actors. The Govern- es. During the process, applicants were asked to provide ment is rightly concerned about e-commerce becoming different documentations and reports without consistent the gateway for cheap foreign goods, flooding the domes- guidance from the two authorities. Moreover, there is un- tic market and crowding out domestic producers. The certainty on the timeframes for licensing individual and number of inbound e-commerce consignments increased legal entities. There needs to be an effort to harmonize more than eight-fold from 6.1 million in 2017 to 49.7 mil- the regulations of agents of LKD and LP, which fall under lion on 2019. Recently promulgated Minister of Finance the jurisdiction of the Bank Indonesia and the Financial Decree No. 199/2019, effective from January 2020, lowers Services Authority (Otoriti Jasa Keuangan, OJK). the taxable threshold on inbound shipments from US$75 to just US$3 per consignment to address this issue.84 This is a an important reform that needs robust implementa- tion, as it helps to level the playing field but, as Govern- → To harness the potential ment Regulation No. 80/2019 concerning Trade through Electronic Systems (“GR E-Commerce”) also points out, of e-commerce to boost greater effort is also required in identifying and promot- inclusion, a greater emphasis ing local products, catered to the domestic as well as is required on developing overseas markets. Currently, these efforts are limited to public-private joint co-branded promotion campaigns and implementing innovative to encourage Indonesians to buy local, but more can be solutions to identify and done by the Government, collaborating with the e-mar- promote local products, boost ketplaces to identify and nurture clusters of productivity, and provide training, coordination and a market linkages the productivity of small and function to enable more MSMEs to be able to produce and medium enterprises and also sell online. As China’s experience shows, while the first strengthen their ability to wave of e-commerce developed somewhat spontaneously and did so in places that were already well endowed with connect to the ecosystem the digital and analog enablers, the Chinese Government played a crucial role in the subsequent phases to ensure In response to the rapid development of e-commerce, the that e-commerce expanded opportunities for producers Government issued the 14th Economic Policy Package on in rural and less well-endowed parts of the country (see the Road Map for the National Electronic-Based Trade Spotlight 2 on China’s e-commerce experience and some System (SPNBE) or the Road Map for e-Commerce in lessons for Indonesia). 147 3.4. Jobs in the Gig Economy After the in-depth examination the factors that constrain typical informal worker, and CHAPTER — 03 the ability of e-commerce to support greater inclusion, we turn next to another avenue through which digital mostly in the transport, storage technologies can bring about micro-transformations in and communications sectors the labor market. Instead of goods, we look at the pro- duction or delivery of a range of services specifically using We distinguish between two types of gig workers: those online platforms or, as these jobs are typically known, who use internet for their work and those who do not. “digital gig jobs”. The concept of “gig economy” is broadly We find that internet-using gig workers are, on average, defined as the portion of the economy accounted for by younger (35) than typical workers (41), mostly men (85 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA jobs with no long-term connection to a business by work- percent), and operate predominantly in urban areas (87 ers employed for a particular task or defined time. Gig percent). They are also much better educated than oth- workers are typically hired as “independent contractors”, er informal workers, with 60 percent having completed often defined as a natural person, business, or corporation upper secondary (compared with 25 percent for informal that provides goods or services to another entity under workers) and around 11 percent with some tertiary ed- terms specified in a contract or within a verbal agreement. ucation. Internet-using gig workers are less likely to be Unlike an employee, an independent contractor does not underemployed (20 percent) relative to other informal work regularly for an employer but works as and when workers (35 percent) and the average worker of any type demanded, during which time he or she may be subject to generally (29 percent). Only wage workers have lower legal provisions. Contractors often work through a limit- rates of underemployment (14 percent). Around 68.5 ed company or franchise, which they themselves own, or percent of these workers are in the transportation, storage may work through an umbrella company.85 and communications sectors, while another 16.7 percent are in the wholesale, retail, restaurants and hotels. Finally, As noted in Chapter 1, the ubiquitous green jackets worn digital gig workers work the longest hours. On average, by drivers of the homegrown ride-hailing decacorn Go- Indonesians work 39 hours a week. Informal workers typi- Jek and its regional competitor Grab are perhaps the cally work a 38-hour week. In contrast, digital gig workers most visible emblems of Indonesia’s digital economy. In clock in 49 hours a week (see Annex 3.C for a detailed addition to ride-hailing and delivery services, before the profile of gig workers). pandemic they provided a suite of other services such as cleaning, home repair, and massage. While the analysis of gig work in Indonesia has often tended to begin and → Digital gigs in Indonesia are end with the study of workers in the ecosystem of these mega-players, we zoom out and, exploiting nationally a stepping-stone into the labor representative data that was collected for the first time in market for many the national labor force survey in 2019, analyze this land- scape in a more general way.86 Who are the gig workers in Job flexibility is among the most touted attributes of a gig Indonesia? How do they fare in the labor market, relative job. Workers enjoy temporal flexibility, with platforms to other formal and informal workers in the country, in seemingly giving them full control over how to split their terms of earnings and overall job quality? time between working hours and leisure.87 Uber drivers benefit significantly from real-time flexibility, earning more than twice the surplus without flexible arrange - → Internet-using gig workers ments.88 Uber drivers in London reported a higher level of life satisfaction than other workers, despite remaining in Indonesia are young, mostly at the lower end of the London income distribution.89 male, live in urban areas, are These drivers have turned to the gig economy not due to much better educated than the the absence of jobs in the conventional labor market. In fact, a vast majority of them were employed either full- 148 (64 percent) or part-time (23 percent) prior to partner- comes). A study on Uber drivers in London found that ing with Uber, while 80 percent of them perceived their 45 percent of them self-reportedly said that their income previous job as a permanent job (until they left it, got fired increased “a little” or “a lot”, while only 19 percent of or laid off ).90 This reflected their preference for flexibility them said their incomes decreased.94 A key challenge in and the autonomy that the platform offered. Indeed, two interpreting results on income from these studies is that surveys on Grab and Gojek drivers in Indonesia show that they do not account for the possible increase in working the top reason for drivers partnering with the platform hours. In the case of gig workers working on digital plat- was flexibility of work.91 forms, collecting data on working hours of these workers is difficult because it is not possible to distinguish, for Besides providing flexible jobs, digital gig economy has example, between a driver “waiting for consumers” and also often been lauded for creating jobs, especially in cit- “not working” (taking a break) when she/he is shown to ies. In 2019, while around two-thirds of internet-using gig be active but not serving any customer. CHAPTER — 03 workers had a job before, while digital gigs were the first jobs for the remaining third. In other words, for around Again, applying a multi-variate regression technique on 33 percent of all internet-using gig workers in Indonesia the same labor force survey data, we find that monthly (of 1.2 million workers), 69 percent of whom were in the earnings for digital gig workers are 15.8 percent higher transport and logistics sectors, their digital gigs were the than other informal workers. This is after conditioning first stepping-stone into the labor market. This number on location, sector of employment, age, gender, educa- is very close to the findings from a 2019 study, which re- tion, internet use, and hours worked. In terms of earnings ported that 38 percent of GrabBike and 33 percent of per hour of reported work, digital gig workers earn 6.2 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA GrabCar drivers did not earn an income before joining percent higher than other informal workers. As noted the platform.92 Among those gig workers who had a job/ above, however, these workers put in longer hours than business before and left the previous job in the past one the average informal worker. year, 80 percent joined the transport and logistics sectors. Looking more closely into the reasons for leaving their Second, are gig workers more likely to have another job last job in order to join digital gigs, we find “unsatisfac- to diversify their earnings portfolio or to try and supple- tory income” was the most common reason among low- ment income? Many drivers on platforms such as Uber er-educated workers (lower secondary school graduates in the United States who had full- or part-time jobs prior or below), while “unsatisfactory working environment” to joining continued in those positions, which makes the and other involuntary factors such as “work termination” flexibility of working hours more valuable.95 Uber drivers and “out of contract/work arrangement” were the most also used their earning as drivers to smooth fluctuations common reasons among higher-educated workers. This in their income. We find that in Indonesia, digital gig suggests that, while digital gig jobs are a fallback option workers who are self-employed are less likely to have a for the higher educated who may be temporarily dislo- secondary job compared with other informal workers. cated from their existing jobs, they are potentially a step Indeed, a study on GrabBike drivers shows that only 25 up for the relatively lower-educated aspiring to boost percent of drivers have other secondary job(s), while the incomes. labor force survey 2019 reports that 11 percent of inter- net-using gig workers have other secondary job(s). In any case, these numbers are much lower than those of Uber → Digital gig workers earn driver-partners, one-third of whom had a part-time job and another one-third of whom had a full-time job, in more than otherwise similar addition to partnering with Uber.96 Finally, we also find informal workers, and are less that digital gig workers are no more or less likely than likely to be supplementing other informal workers to experience a high degree of job churning. incomes with a second job Overall, unlike some of the narrative around digital econ- Do digital gig workers end up earning more relative omy jobs in more industrialized countries around the to other informal workers? Various studies on online world, we see that gig jobs in Indonesia are, on average, ride-hailing drivers show that, on average, drivers’ in- relatively better than many informal jobs, though certain- come increased after they joined the platform.93 For ex- ly not better than wage jobs. Internet-using gig workers ample, a study on GrabBike drivers in five Indonesian earn relatively higher monthly income and income per cities shows that, on average, drivers experienced an 87 hour, and they are less likely to need a second job to sup- percent increase of their (total) income compared with plement their income. While these jobs are also the first their previous jobs after joining the platform, although jobs for many, they are available predominantly to urban some drivers never left their previous jobs (hence, the men. Finally, digital gig workers also work the longest stated income is a combination of their old and new in- hours among all other types of workers in Indonesia. 149 3.5. Three Distributional Tensions and a Look to the Future So far, this chapter has looked at two important applica- of market expansion, especially in its early years, caused tions of digital technologies (e-commerce and the digital large-scale disruptions in the market, wiping out not just CHAPTER — 03 gig-economy) from the lens of economic opportunities the inefficient players. Consumers ended up better off, that they represent for Indonesians to improve their in- but the losses, which were often localized, did not enter comes. The focus has been on the individual experience the aggregate welfare calculus. of the workers or e-sellers active on these platforms, how they interact with the market, what constraints they As the digital economy expands in Indonesia and plat- face coming on board, and what kind of gains they make form affiliations generate opportunities for places and once they are on board. But do gains accruing from the small enterprises with the digital and analog readiness adoption of digital technologies to one set of actors in to capitalize on these opportunities, this will inevitably BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA the economy come at the cost of losses to others? If they affect offline sellers. As we show in Chapter 1, consumers do, is there a need then to calibrate our views regarding are already benefiting in the form of greater convenience, these technologies? More importantly, how should policy a broader range of product choices and better prices. A makers respond? In this section we take a slightly broader, more vibrant digital economy is likely to further enhance ecosystem-wide view and identify a few additional con- these benefits. But as we also show in the case of digital siderations that are uniquely important for the digital food delivery market, only about 22 percent of all con- economy, especially from a distributional perspective. sumers who order food are ‘new customers’ who are tak- ing fewer home-cooked meals. The rest are simply using deliveries to substitute for physical restaurant experienc- es. What this implies for the broader digital economy is → Not everyone is a winner that not all digital transactions are likely to be new de- mand resulting in the growth of the aggregate market. Limiting ourselves entirely to the realm of digital plat- There is a limit to this growth and some, and perhaps a forms, as has been our focus throughout the chapter, we significantly larger parts, of digital transactions will be identify three potential distributional tensions: (i) online displacing existing offline ones. Indonesia has recognized vs offline; (ii) among online market participants within this issue and has initiated some measures to level the the same platform; and (iii) at a higher level of aggrega- playing field between online and offline actors. The new tion, more generally between workers and firms. Given its regulation on e-commerce takes a few concrete steps on level of digitization, some of these distributional tensions leveling the playing field by requiring digital merchants may look less relevant to Indonesia now than others, but active on platforms to be eligible for VAT. More on this all three are important considerations for policy makers is discussed in Chapter 5. to be aware of as they shape the digital economy landscape in the coming decades. The second distributional tension is among merchants within platform ecosystems. One of the defining char- The first distributional tension is between online and of- acteristics of commercial activities conducted through fline domains. As e-commerce grows and more and more digital platforms is that they make an incredible amount people start buying through digital marketplaces, offline of information available to consumers. Not only can sellers (including the mom-and-pop stores) and the local consumers learn everything there is to learn about the economies they reside in could be adversely affected. A product of their interest, but they can also compare pric- prominent example of this in the context of the United es offered by a multitude of other sellers offering the States is the disruption Amazon brought about in the same product, often in the same platform. What is more, brick-and-mortar book industry, driving many local consumers can also learn about the online reputations booksellers out of business. Some of this could be the built by these sellers in the form of rating systems, as well classic Schumpeterian creative destruction at play, with as written feedback provided by others who the sellers inefficient firms being displaced by more efficient digital may have interacted with in the past. The availability ones but, as has been noted in the literature, Amazon’s of this type of information allows marginally superior aggressive growth strategy that sacrificed profits in favor products and marginally superior vendors to essentially 150 FIGURE 3.10 capture a large share of the market, creating a winner-takes-all type of dynamic within the marketplace.97 This is not just within e-com- Top 1% of sellers capture almost half of merce, but any type of platform that relies on the sales network effects. % OF TOTAL SALES REVENUE ACCOUNTED FOR BY THE TOP 1%, THE NEXT An e-commerce vendor survey conducted by 9%, THE NEXT 40% AND THE BOTTOM 50% the World Bank in 2019 provides some sugges- tive evidence confirming this. Among the thou- sand or so respondents of this online survey of vendors in multiple platforms, the top 1 percent of sellers accounted for almost half of the total CHAPTER — 03 sales (46.7 percent), while the top 10 percent accounted for almost 80 percent of all sales (Figure 3.10). The share of sales for the entire FIGURE 3.11 bottom half of the sample was just 4 percent of the total. In the same survey, vendors were also asked to identify their top challenge for devel- Market competition is the dominant concern oping e-commerce in Indonesia. In addition to for vendors in platforms… other issues that we have discussed in the chap- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA ter, such as societal trust in digital transactions (11 percent), payments and financing (10 per- cent), and the state of development of the ICT infrastructure (7 percent), an overwhelming 60 percent of the respondents identified market competition as the key challenge (Figure 3.11). Specifically, when asked about what aspect of competition they were most worried about, over 70 percent of respondents singled out the fact that there were simply too many players in the industry (Figure 3.12). This fact that the ecosystems within platforms are crowded fields and marginally superior vendors/sellers can command a large share of the market has implications for the gains that a marginal ven- FIGURE 3.12 dor can expect to make by joining. …with the crowded field being considered Finally, the third distributional tension is be- the most challenging aspect of market tween workers and firms. The “winner-takes-all” competition by over 70 percent characteristic of the digital economy at the mer- chant level could play out also at a more macro, industry level.98 Business models that rely on multi-sided platforms essentially generate val- ue by matching customers with complementary needs. The value of the platform to the marginal customer depends on the number of vendors and service providers already on board. Similarly, the value of the platform to a marginal vendor also depends on the number of customers already on board. Thus, in order to grow and be success- ful, platforms need to build a critical mass on both sides of the market. But once a platform becomes dominant, it could become too large for potential competitors to dislodge giving rise Source: World Bank staff calculations based on data from the E-Commerce Survey, 2019. Note: The survey was implemented as an online survey and included modules on company to a winner-takes-all market structure. In devel- characteristics, customer reach, sales and innovation, challenges and opportunities and other relevant information. A total of 1073 vendors/merchants responded to the oped countries, this network effect has given rise survey. to giant technology companies, such as Google, 151 Facebook, Apple, Amazon, Uber and Airbnb, where a few firms on-boarding issues (awareness, knowledge and trust) discussed capture a dominant part of the market. here, significant room may still exist for digital technologies to further boost the productivity of Indonesia’s low-skilled and Given the large amount of data these digital businesses col- low-productivity workers and small-scale household enterpris- lect, and the increasing use of automated machine learning es, 75 percent of whom reported being growth oriented. How- and AI-driven analytics to drive continuous improvements in ever, given the winner-takes-all dynamics that are inherent to products and the services they provide, on pricing and person- platform-based business models, even if Indonesia is able to alization, so targeting of ads could further entrench their po- digitally on-board all of the lower segment of the workforce, sition. Data collection and sharing across platforms belonging it may not be enough to catapult the 115 million aspirants into to the same company could give rise to an exponentially higher the middle class. benefit to the company compared with those companies oper- ating only under one platform. And, once this is established, This puts the spotlight back on quality wage jobs that have CHAPTER — 03 there is a risk that they could exploit monopsonist power vis-à- been identified as the most viable pathway to the middle class vis vendors to extract a larger share of value added of the prod- for most Indonesians in the long term.102 The challenges that uct. This increase in market concentration may lead to greater Indonesia faces in generating quality jobs, including stable and inequality as labor receives a smaller share of value added in protected wage jobs, at scale have been well recorded in other these segments.99 Some have found these consequences of the recent World Bank publications and they will not be repeat- digital economy to be the most important reason behind the ed here.103 However, with broader digital adoption by firms, recent increase in inequality in the United States.100 and increased sophistication and automation of production technologies, the landscape of jobs could also begin to shift BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA The level of digitization in Indonesia has perhaps not reached fundamentally in the medium and longer term. More research levels where these effects would begin to manifest themselves is required on the topic, but preliminary analysis shows that and more work will be required specifically in the Indone - this shift may already be underway. Digital tasks—or tasks sian context to see how these dynamics play out as the digital that require the ability to work with computers and electronics economy grows. Nonetheless, it is a potential risk with con- used on the job—have been the fastest growing task-content sequential distributional implications and space that needs of jobs relative to others since 2007.104 These have also grown to be closely monitored. Boosting digital entrepreneurship, fastest in sectors that are relatively more productive to begin including by addressing the glaring gaps in high-end digital with, namely manufacturing and high value-added services. talent (developers, AI and machine-learning specialists), lev- The second fastest growing task-content of jobs is (routine and eling the playing field for innovation and preempting policies non-routine) cognitive tasks, while (routine and non-routine) that protect players, while not being too restrictive on the manual task-content is in decline, albeit still dominating the growth of what is still a nascent digital economy can be some labor market at the moment. Executives in Indonesia’s largest elements of strategies to address this issue. companies increasingly list data analysts and scientists, big data specialists, AI and machine learning specialists, and digital marketing and strategy specialists among leading positions in → Between a nation full of growing demand, as well as job roles such as process automa- tion specialists and IoT specialists. smartphone wielding micro- entrepreneurs and AI whiz kids How ready are Indonesian workers to embrace these changes and thrive in the modern economy? Not so much, based on So, what do these distributional tensions at the aggregate level available evidence. The quality of the labor force has been a imply for the potential of digital technologies to bring about major hinderance even for the brick-and-mortar economy; and transformative changes in the jobs and economic opportunities without significant investments to bridge the human capital gap landscape in Indonesia? In this chapter, and indeed this entire and boost skill levels, it is likely to be so also for the digital econ- report, the focus has been on low-skilled biased elements of omy. Indonesian firms have identified the shortage of specialized digital technologies, which might be more relevant to the low- professionals and managers in the local labor market as one of skilled Indonesian workforce, 60 percent of which has educa- the most important bottlenecks in fostering innovation within tion below secondary level (Indonesia’s education profile in firms. The share of firms reporting inadequate skills as the top 2018 is similar to that of the United States in 1940).101 There constraint in hiring professionals and managers is the highest is evidence that the digital economy is already providing bet- in the region (Figure 3.13). A 2018 joint assessment by the Gov- ter opportunities to the workers who have been able to enter ernment of Indonesian and the World Bank highlights critical the growing gig economy and, likewise, e-commerce is also shortages of skills in dozens of managerial and professional po- helping to boost productivity, especially by activating women sitions critical to technology innovation, such as cloud-solution and youth in the labor market. Given the barriers to entry and architects and UI/UX designers. Shortage of qualified talent to adoption related to affordable and good-quality internet con- develop new products and services has prompted well-funded nectivity discussed in Chapter 2, and other constraints related digital platforms to look elsewhere for R&D capabilities. For to logistics, digital payments and access to finance, as well as example, Go-Jek has established an off-shore R&D division 152 FIGURE 3.13 by acquiring three Indian tech companies and its regional rival, Grab, has also set up an R&D center in India for similar reasons.105 Indonesian firms face a deep skills mismatch, particularly for managers and This skill gap not only stunts innovation, but professionals it also represents a lost opportunity; these are good jobs going unfilled in a labor market in SHARE OF FIRMS THAT CITED INADEQUATE SKILLS AS THE KEY BARRIER which millions are looking for a pathway to the TO HIRING EACH TYPE OF WORKER, PERCENTAGE middle class. It also explains the skill premium and the digital technology-induced widening of the skill premium discussed earlier in the chapter. But just as a nation filled with smart- CHAPTER — 03 phone-wielding micro-entrepreneurs all selling their wares through digital platforms is far from a realistic portrayal of the inclusive digital fu- ture that Indonesia wants, so is a nation filled with data scientists, cloud-solution architects and AI experts. Thus, in addition to low digital literacy and the severe shortage of digital talent at the high end, investments in skills for a dig- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA ital future should also include a broader-based effort to develop a pipeline of workers that are generally better at problem-solving, com- munication, teamwork and adaptability (i.e., Source: Gomez-Mera and Hollweg, 2018 based on WBES data. high-order cognitive and social skills). 153 3.6. Conclusion CHAPTER — 03 This chapter has analyzed some aspects of jobs and eco- between the Government (central, as well as local) and nomic opportunities in Indonesia’s digital economy. Four digital marketplaces to identify, nurture and support key findings emerge from the analysis. First, despite the pockets of growth-oriented micro and small enterpris- accelerated adoption induced by the pandemic, the use es. This could take the form of outreach and training, of digital technologies among the small and micro firms, as well as the provision of complementary services, to as well as low-skilled workers, remains highly limited. As enable them to better harness opportunities presented a result, at the aggregate level, the increased availabili- by e-commerce. Beyond e-commerce and micro-enter- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA ty of and access to the internet has benefited the bigger prises, major investments will also be required to boost firms and higher skilled workers more than the smaller the overall skill levels of Indonesian workers to prepare and lower skilled ones. Second, e-commerce—a specific them for the opportunities and challenges as a result of application of digital technologies in the economy—is an the broader application of digital technologies by larger important source of not only primary but also supple- and more sophisticated firms. Some concrete recommen- mental income, especially for women and youth, and a dations for policy makers are summarized in Chapter 5. useful pathway for women re-entering the labor market. Third, internet access and challenges related to logistics Finally, by highlighting some distributional tensions that are major impediments to the expansion of the geographic could arise as inevitable upshots of platform-based busi- extent of e-commerce with logistics, in particular, stand- ness models, the chapter also flags the importance of tax ing out as a binding constraint also on the intensity of policies to ensure a level playing field for online and offline these transactions. Low adoption of digital payments, actors, and policies and regulations related to competition which are constrained on the demand side by issues of and innovation that ensure a level field for incumbents awareness, knowledge and trust, and on the supply side and innovators. Just as important are worker protection by regulatory hurdles, a lack of appropriate infrastructure policies. Currently, there is no government regulation on and small margins for providers, is also another important digital gig economy platforms and their workers. Rather factor limiting the growth of e-commerce in Indonesia. the platforms themselves set working conditions through their terms of service agreements. As the digital economy The chapter also presents some novel evidence showing deepens and platforms start seeking profitability, it will that digitally-enabled gig jobs in Indonesia are not only a be important to ensure that the welfare of workers in the stepping-stone into the labor market and offering flexible digital economy is protected. work hours for many, but they also pay better than other informal jobs. However, these opportunities are available primarily to urban men and digital gig workers work on average 10 hours a week more than every other type of worker in Indonesia. These findings suggest that, in addition to expanding access to affordable and reliable internet to almost half of the country’s population that is not yet connected, addressing challenges related to logistics and the greater adoption of digital payments, as well as the on-boarding issues of awareness, knowledge and trust, could have sig- nificant payoffs in terms of the growth of e-commerce in Indonesia. Specifically, from the perspective of enhancing the impact of e-commerce on inclusion, the findings of the chapter point to the need for a stronger partnership 154 155 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA CHAPTER — 03 Endnotes 58 Weil, 2014. 71 “Regulatory compliance issue” category includes “will 59 Also see Box 1.3, p.28-29, World Bank, 2021. have to pay tax” and “will have to comply with some regulations”. If we break down “regulatory compli- 60 World Bank, 2020a. The first round of the World Bank ance”, “will have to pay tax” alone does not seem to be COVID-19 Business Pulse Survey (COV-BPS) was the reason for not joining digital platforms, contrary to conducted in June 15-23 with phone interviews on 850 the assertation that social media and chat messaging representative firms. apps are used to avoid taxes. This has policy implica- 61 The second round of the World Bank COVID-19 Busi- tion in regard to digital taxation, as the main reason ness Pulse Survey (COV-BPS) conducted in October the government has been hesitant to tax merchants on CHAPTER — 03 2020. The percentage point differences already reflect digital platforms is because digital platforms have ar- the lower attrition rate in the second round of the sur- gued that imposing tax on merchants who are on dig- vey. ital platforms will deter those who want to join digital 62 World Bank, 2020b & 2021b. This is based on data platforms from joining. This reason seems to be un- from the World Bank-Bukalapak Digital Merchants grounded based on this finding. Survey conducted between May 20 and June 27, 2020, 72 Digital Economy Household Survey (DEHS) done by with a total sample of 1,020 respondents and World the World Bank during February and March of 2020. Bank-Shopee Digital Merchants Survey conducted 73 World Bank, 2020c. BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA between December 21-25 2020 with a total sample of 15,238 respondents. 74 Lain, Alatas and Setyonaluri, 2020. 63 Google, Temasek and Bain, 2020. 75 The same survey also finds that e-commerce is found to help the disabled graduated from the disability spe- 64 The 2018 number is based on the national labor force cial schools to be economically active. Having a severe survey of that year while the 2020 number is based disability reduces engagement in the labor force (Wi- on the Digital Economy Household Survey, a special hardja and Cunningham, forthcoming). Although peo- purpose survey conducted specifically for this report ple with a moderate disability represent a greater share during February and March 2020. labor force (6.9 percent) and workers (7.1 percent) 65 “Do not need it” is a more common reason for house- than expected given that they are only 6.2 percent of holds in the highest quintile (69 percent) reporting not the working age population, in contrast, those with a to use internet compared with households in the low- severe disability represent 0.9 percent of the working est quintile (15 percent). This difference is statistically population but only 0.3 percent of labor force partici- significant at 95 percent ci. Meanwhile, “do not know pants or workers. The e-commerce incidence of those how to use the internet/lack of knowledge and skills” graduating from the upper-secondary-level disability is a more common reason for households in the lowest special school is highest (61.4 percent) among workers quintile (86 percent) compared with households in the with any educational background. richest quintile (40 percent), although this difference 76 Tiwari, Wibisana and Wihardja, 2020. is not significantly significant at 95 percent ci. 77 LSPs include transportation companies (shipping 66 This is consistent with findings from in-depth inter- lines, ferry lines, airlines, trucking companies), as well views conducted by the World Bank with some of the as third party logistics companies that coordinate mul- larger digital economy players in Indonesia. tiple transportation modes together with warehousing 67 SAKERNAS, August 2019. and other value-added activities such as packaging and 68 Around 12.2 million workers alone were engaged in sorting. e-commerce activities as their primary job (SAKER- 78 Paxel, 2019. NAS 2019). However, Indonesia’s National Social-Eco- 79 TaniSupply has warehouses outside of Jakarta, such as nomic Survey 2019 (SUSENAS 2019) shows only 5.9 Bogor and Surabaya, and is planning to open another million online sellers (SUSENAS March 2019). ones in Jabodetabek areas, Bali, Kalimantan and Su- 69 Alibaba Group and World Bank, 2019. lawesi. 70 SAKERNAS, August 2019. 156 80 There are other measures of financial inclusion, includ- 87 See Hall and Krueger, 2016; McGuire, 2018. ing in Indonesia. These numbers from the new World 88 Chen et al., 2016. Bank survey appear to align well with these numbers. For example, the Global Findex Database, 2017 found 89 Berger et al., 2018. that 49 percent of Indonesian adults had transaction 90 Ibid. accounts. Similarly, the Financial Inclusion Insight 91 CSIS and Tenggara Strategics, 2019; Puskakom, 2017. (FII) Survey in 2018 found that 56 percent adults were owners of a bank account. Likewise, the Global Fin- 92 CSIS and Tenggara Strategics, 2019. dex Database also showed that 3 percent of Indonesian 93 CSIS and Tenggara Strategics, 2019; Puskakom, 2017. CHAPTER — 03 adults owned mobile money accounts and there were at 94 Berger et al., 2018. least 7 percent of adults who used some form of mo- bile banking. The numbers from the DEHS—which 95 Hall and Krueger 2016. are based on households—appear to be in a reasonable 96 Ibid. ballpark of these numbers and, as such, provides a use- 97 See for example Guellec and Paunov, 2018. ful updated triangulation on the picture of financial ac- cess in Indonesia. 98 See, for example, Banerjee and Duflo, 2019, p.241. 81 Pazarbasioglu et al., 2020. 99 Frank et al., 2019; Banerjee and Duflo, 2019. BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA 82 CICO Economics in Indonesia, BCG, 2019. 100 Autor et al., 2017. See also Weil, 2018, p.9: “Where lead companies once shared gains with their internal 83 Agent Network Accelerator (ANA) Research, Indone- workforce, fissuring leads to growing inequality in how sia Country Report: December 2017. the value created in the economy is distributed.” And 84 Under this new regulation, foreign-produced textiles, Banerjee and Duflo 2019, p.242: “The increase in con- clothes, bags, and shoes that cost a minimum of US$3 centration (among superstar firms) thus helps explains will be subject to a range of taxes and import duties a part of why wages are not keeping pace with GDP. with a total rate of 32.5 to 50 percent of their value. The rise of superstar firms also offers an explanation For other products worth at least US$3, the taxes and for why overall wage inequality has been rising…” import duties will be lowered from 27.5 to 37.5 percent 101 Wihardja and Cunningham, forthcoming. of their value to 17.5 percent. Goods worth below US$3 will be exempted from import duties as well as val- 102 World Bank, 2019. ue-added tax. 103 See World Bank, 2020c; World Bank, 2019. 85 Berg, 2016. 104 Wihardja and Pradana, 2020. 86 The labor force survey defines a ‘gig worker’ as a worker 105 The human capital gap is magnified by restrictive gov- who is either self-employed or an employer with tem- ernment policy on the hiring of foreign workers, lim- porary/family/unpaid workers, but does not have a full iting the ability of employers to fill the skills gap with control of his/her business because of several things global talent in areas where local skills are in short sup- related to his/her business activities, namely having ply. The recently promulgated Omnibus Law on Job a third party (individual/business/firm) regulate the Creation includes a provision to relax this. price/tariff of goods and services, control the raw ma- terials/machine and tools/capital goods OR control the access to market/clients. This definition may exclude other types of gig workers, such as contract firm work- ers (outsourced workers), on-call workers and fixed- term or temporary workers. 157 References 1 Alibaba Group and World Bank. 2019. “E-commerce Development: Experience from China” (English). Washington, DC: World Bank Group. Available at: http://documents1.worldbank.org/curated/en/823771574361853775/pdf/Overview.pdf (accessed on December 14, 2020). 2 Autor, David, Dorn, David, Katz, Lawrence, Patterson, Christina, and Van Reenen, Jon. 2017. “Concentrating on the Fall of the Labor Share.” American Economic Review 107(5): 180–85. 3 Banerjee, Abhijit V. and Esther Duflo. 2019. Good Economics for Hard Times. Allen Lane, United Kingdom. 4 Berg, Janine. 2016. “Income Security In the On-Demand Economy : Findings and Policy Lessons from a Survey of Crowdworkers.” Comparative Labour Law and Policy Journal 37(3). CHAPTER — 03 5 Berger, Thor, Carl Benedikt Frey, Guy Levin, Santosh Rao Danda. 2018. “Uber Happy? Work and Well-being in the ‘Gig Economy’.” Presented at the 68th Economic Policy Panel Meeting, October 4-5, Vienna. 6 Boston Consulting Group, MicroSave Consulting. 2019. “CICO Economics in Indonesia”. Available at: https://connect.bcg. com/cicoeconomics/6-cico-economics-in-indonesia_aug-2019-2/ (accessed on December 14, 2020). 7 Centre for Strategic and International Studies (CSIS) and Tenggara Strategics. 2019. “Economic Research: How GRAB help increase the potential of the informal sector in Indonesia”. Slide presentation. 8 Chen, Cecilia, and Marcus Haymon. 2016. “Realizing the Potential of Digital Job-Seeking Platforms.” Brookings Institute. BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA A paper prepared for the 2016 Brookings Blum Roundtable. Available at: https:/ /www.brookings.edu/wp-content/ uploads/2016/07/Global_20160720_Blum_ChenHaymon.pdf 9 Frank, Morgan R., David Autor, James E. Bessen, Erik Brynjolfsson, Manuel Cebrian, David J. Deming, Maryann Feldman et al. 2019. “Toward understanding the impact of artificial intelligence on labor”. Proceedings of the National Academy of Sciences, 116(14):6531-6539. 10 Google, Temasek and Bain. 2020. “e-Conomy SEA 2020 – At Full Velocity: Resilient and Racing Ahead”. 11 Gomez-Mera, L. and C. Hollweg. 2018. “Firm Performance and Constraints in Indonesia.” Mimeo. World Bank. 12 Guellec, Dominique, and Caroline Paunov. 2018. “Innovation policies in the digital age.” OECD Working Paper Series. 13 Hall, Jonathan V., and Alan B. Krueger. 2016. “An Analysis of the Labor Market for Uber’s Driver-Partners in the United States,” NBER Working Paper Series 22843. 14 Kapoor, Raunak, Alfa Pelupessy, Linggo Kusuma, Vera Bersudskaya, Sidra Mughal, Grace Retnowati and Aakash Mehrotra. 2017. Agent Network Accelerator Research: Indonesia Country Report 2017. MicroSave Consulting. Available at: https://www. microsave.net/2017/12/04/agent-network-accelerator-research-indonesia-country-report-2017/ (accessed on December 14, 2020). 15 Lain, Jonathan, Hamidah Alatas, and Diahhadi Setyonaluri. 2020. “Women at Work in Indonesia.” Background paper to Indonesia’s Jobs Report: Pathways to Middle-Class Jobs Indonesia. 16 McGuire, Robert. 2018. “Ultimate Guide to Gig Economy Data: A Summary of Every Freelance Survey We Can Find”. Nation 1099. Available at: https://nation1099.com/gig-economy-datafreelancer-study/. 17 Paxel. 2019. “Introduction Paxel”. Presentation at the World Bank’s Digital Economy BBL on December 4, 2020, at World Bank Office Jakarta, Indonesia. 18 Pazarbasioglu, Ceyla, Alfonso Garcia Mora, Mahesh Uttamchandani, Harish Natarajan, Erik Feyen, and Matthew Saal. 2020. Digital Financial Services. World Bank, Washington, DC. 19 Puskakom (Pusat Kajian Komunikasi Universitas Indonesia). 2017. “Research findings on the social benefit of on-demand applications: Case Study on Gojek Indonesia.” A slide presentation. 20 Tiwari, Sailesh, Putu Sanjiwacika Wibisana and Maria Monica Wihardja. 2020. “Drivers and Welfare Impacts of E-Commerce Growth in Indonesia: Analyses Using Digital Platform Data.” Background Note for the Digital Economy and Inclusion Flagship report. 21 Weil, David. 2014. The Fissured Workplace. Why Work Became So Bad for So Many and What Can Be Done to Improve It. Harvard University Press, Cambridge, Massachusetts, London, England. 22 Wihardja, Maria Monica and Abror Tegar Pradana. 2020. “Digital Skills and Occupation in Indonesia.” Background study for Indonesia’s Digital Economy and Inclusion Flagship report. 23 Wihardja, Maria Monica, and Wendy Cunningham. Forthcoming. Indonesia’s Jobs Report: Pathways to Middle-Class Jobs in Indonesia. World Bank. 24 World Bank. 2019. Indonesia’s Aspiring Middle: Expanding the Middle Class. World Bank, Washington, DC. 25 World Bank. 2020a. “How Covid-19 is affecting firms in Indonesia: Results from the 1st round of the Covid-19 Business Pulse Survey.” Policy Brief. World Bank. 158 26 World Bank. 2020b. “Impacts on Digital Merchants: Insights from the Bukalapak-World Bank Survey.” Policy Brief. World Bank. 27 World Bank. 2020c. Indonesia Systematic Country Diagnostic: Eliminating Poverty, Bringing Economic Security to All. World Bank Washington, DC. 28 World Bank. 2021. “Uneven Recovery”. East Asia and Pacific Economic Update (April), World Bank, Washington, DC. Doi: 10.1596/978-1-4648-1702-1. License: Creative Commons Attribution CC BY 3.0 IGO CHAPTER — 03 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA 159 Annex. ANNEX 3.A Drivers of E-commerce Growth (Buyers’ Side) DEPENDENT VARIABLE PERCENTAGE OF BUYERS IN THE NUMBER OF PURCHASES PER BUYER NUMBER OF PURCHASES PER POPULATION (LN) POPULATION (LN) (1) (2) (3) (1) (2) (3) (1) (2) (3) CHAPTER — 03 11.10*** 2.984 -1.612 1.277*** -0.609* -0.166 5.070*** -0.298 0.186 Real per capital consumption (ln) (3.819) (2.524) (2.874) (0.278) (0.307) (0.256) (0.705) (0.624) (0.545) 2.533** 1.564** 0.413 0.008 0.145 0.054 0.393** 0.295 0.163 Population (ln) (1.086) (0.645) (0.425) (0.052) (0.089) (0.057) (0.155) (0.189) (0.125) 0.454*** 0.291* 0.159*** 0.011 0.350*** 0.071* Internet access (0.127) (0.165) (0.016) (0.018) (0.028) (0.037) 0.058** 0.018 -0.016*** 0.002 -0.012** 0.016*** Share of Urban Population (p.p) (0.028) (0.022) (0.003) (0.002) (0.005) (0.005) BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA -1.834** -2.168** 0.196* -0.081* 0.144 -0.220** Cost of logistics (0.860) (0.945) (0.098) (0.046) (0.147) (0.086) Additional controls: Schooling years NO YES YES NO YES YES NO YES YES + electricity access Island-region + Year effects NO NO YES NO NO YES NO NO YES Observations 170 170 170 170 170 170 170 170 170 R-squared 0.359 0.635 0.782 0.122 0.547 0.895 0.449 0.839 0.959 *Significant at the 10 percent level; ** Significant at the 5 percent level; *** significant at the 1 percent level. Note: The analysis exploits province-year level variation using data on e-commerce volumes and intensity provided by e-commerce platforms and other data from SUSENAS. Internet access is measured by proportion of adults with access to internet. GMV OF PURCHASED GOODS PER GMV PER PURCHASE (LN) GMV PER BUYER (LN) DEPENDENT VARIABLE POPULATION (LN) (1) (2) (3) (1) (2) (3) (1) (2) (3) 4.538*** 0.208 0.309 -0.532** 0.506 0.123 0.744*** -0.104 -0.044 Real per capital consumption (ln) (0.585) (0.457) (0.563) (0.250) (0.346) (0.295) (0.236) (0.293) (0.313) 0.357** 0.118 0.036 -0.035 -0.178*** -0.127** -0.027 -0.033 -0.073 Population (ln) (0.144) (0.149) (0.130) (0.052) (0.065) (0.055) (0.048) (0.060) (0.060) 0.281*** 0.086** -0.069*** 0.015 0.090*** 0.026 Internet access (0.023) (0.041) (0.012) (0.030) (0.013) (0.028) -0.005 0.015** 0.006** -0.002 -0.010*** 0.000 Share of Urban Population (p.p) (0.005) (0.005) (0.003) (0.004) (0.003) (0.003) -0.329* -0.557*** -0.473*** -0.337*** -0.276*** -0.418*** Cost of logistics (0.164) (0.153) (0.113) (0.103) (0.100) (0.097) Additional controls: Schooling years No Yes Yes No Yes Yes No Yes Yes + electricity access Island-region + Year effects No No Yes No No Yes No No Yes Observations 170 170 170 170 170 170 170 170 170 R-squared 0.463 0.805 0.908 0.043 0.295 0.647 0.069 0.392 0.522 *Significant at the 10 percent level; ** Significant at the 5 percent level; *** significant at the 1 percent level. Note: The analysis exploits province-year level variation using data on e-commerce volumes and intensity provided by e-commerce platforms and other data from SUSENAS. Internet access is measured by proportion of adults with access to internet. 160 ANNEX 3.B Drivers of E-commerce Growth (Sellers’ Side) DEPENDENT VARIABLE PERCENTAGE OF SELLERS IN THE NUMBER OF ORDERS PER POPULATION NUMBER OF ORDERS PER SELLER (LN) POPULATION (LN) (1) (2) (3) (1) (2) (3) (1) (2) (3) CHAPTER — 03 1.476* 0.478 -0.474 -0.023 -1.685* -0.184 4.171*** -2.865** 0.775 Real per capital consumption (ln) (0.774) (0.514) (0.595) (0.652) (0.912) (1.037) (1.490) (1.400) (1.340) 0.471* 0.273* -0.023 1.112*** 0.816*** 0.817*** 2.386*** 1.326*** 1.224*** Population (ln) (0.235) (0.138) (0.051) (0.112) (0.155) (0.210) (0.278) (0.294) (0.336) 0.044 0.038 0.0586** -0.00803 0.281*** 0.115 Internet access (0.027) (0.029) (0.0260) (0.0493) (0.040) (0.083) 0.015*** 0.002 0.0175* 0.0207 0.038** 0.037* Share of Urban Population (p.p) (0.005) (0.005) (0.00928) (0.0129) (0.016) (0.020) BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA -0.378* -0.429** -0.0281 0.0412 -0.714** -0.520** Cost of logistics (0.191) (0.205) (0.168) (0.162) (0.290) (0.239) Additional controls: Schooling years No Yes Yes No Yes Yes No Yes Yes + electricity access Island-region + Year effects No No Yes No No Yes No No Yes Observations 170 170 170 170 170 170 170 170 170 R-squared 0.262 0.527 0.737 0.596 0.670 0.735 0.668 0.850 0.914 *Significant at the 10 percent level; ** Significant at the 5 percent level; *** significant at the 1 percent level. Note: The analysis exploits province-year level variation using data on e-commerce volumes and intensity provided by e-commerce platforms and other data from SUSENAS. Internet access is measured by proportion of adults with access to internet. DEPENDENT VARIABLE GMV OF ORDERED GOODS PER GMV PER ORDER (LN) GMV PER SELLER (LN) POPULATION (LN) (1) (2) (3) (1) (2) (3) (1) (2) (3) 4.972*** -2.247 0.951 0.801** 0.618 0.175 0.779 -1.068 -0.009 Real per capital consumption (ln) (1.382) (1.441) (1.458) (0.362) (0.501) (0.547) (0.611) (0.962) (1.051) 1.908*** 1.066*** 1.004*** -0.478*** -0.260*** -0.220** 0.633*** 0.556*** 0.597** Population (ln) (0.287) (0.300) (0.342) (0.067) (0.084) (0.093) (0.129) (0.181) (0.230) 0.364*** 0.157* 0.083*** 0.042 0.142*** 0.034 Internet access (0.046) (0.085) (0.021) (0.029) (0.031) (0.050) 0.015 0.028 -0.023*** -0.010 -0.005 0.011 Share of Urban Population (p.p) (0.019) (0.023) (0.005) (0.006) (0.012) (0.016) -0.882*** -0.999*** -0.168 -0.478*** -0.196 -0.437** Cost of logistics (0.291) (0.256) (0.128) (0.113) (0.197) (0.180) Additional controls: Schooling years No Yes Yes No Yes Yes No Yes Yes + electricity access Island-region + Year effects No No Yes No No Yes No No Yes Observations 170 170 170 170 170 170 170 170 170 R-squared 0.607 0.832 0.892 0.252 0.465 0.672 0.336 0.502 0.623 *Significant at the 10 percent level; ** Significant at the 5 percent level; *** significant at the 1 percent level. Note: The analysis exploits province-year level variation using data on e-commerce volumes and intensity provided by e-commerce platforms and other data from SUSENAS. Internet access is measured by proportion of adults with access to internet. 161 ANNEX 3.C Profiles of Gig-Workers Indonesia GIG WORKERS CHAPTER — 03 Employ- Self- Non- All er with Inter- Em- Inter- infor- All All Temp. net ployed net mal workers Workers User Only User workers Only AGE 39 46 38 44 35 42 41 GENDER (1=MALE) 76% 74% 76% 65% 85% 61% 61% LOCATION (1=URBAN) 71% 46% 77% 52% 87% 48% 55% BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA EDUCATION Primary education, lower 31% 52% 26% 56% 10% 51% 40% Lower secondary education 19% 20% 19% 23% 16% 21% 18% Upper secondary education 41% 22% 45% 20% 59% 25% 30% Diploma I/II/III 3% 1% 3% 1% 4% 1% 3% University/Diploma IV or greater 7% 5% 7% 1% 11% 3% 10% Underemployed 25% 27% 25% 32% 20% 35% 29% SECTOR OF EMPLOYMENT Agriculture, forestry, livestock and fishing 12% 39% 7% 25% 2% 35% 27% Mining and quarrying 1% 1% 1% 1% 0% 1% 1% Manufacturing 15% 18% 15% 27% 6% 12% 15% Electricity, gas, and water supply 1% 1% 1% 1% 0% 0% 1% Construction 2% 3% 1% 3% 1% 8% 7% Wholesale and retail trade, restaurants and hotels 20% 32% 18% 25% 17% 29% 25% Transportation, storage and communications 43% 2% 51% 11% 69% 5% 5% Finance, insurance, real estate and business services 2% 2% 2% 1% 2% 1% 3% Community, social and personal services 4% 2% 4% 5% 3% 7% 15% WORKING HOURS PER WEEK 44 39 45 38 49 38 39 HAS SECONDARY JOB? 16% 30% 13% 21% 11% 19% 17% STOPPED WORKING IN THE PAST 1 YEAR? 9% 4% 10% 4% 13% 4% 4% Source: SAKERNAS 2019, World Bank staff calculations. 162 Spotlight 3. Expanding the Reach of e-Commerce to Rural Indonesia: Lessons from Taobao Villages in China CHAPTER — 03 E-commerce has grown rapidly in China. The Chinese government incentives and investments in incubation e-commerce market alone accounts for 45.8 percent of parks and the creation of e-commerce specialty towns the global market share. This is larger than the combined also helped, generating spillover effects in surrounding markets of North America, Europe, Latin America, Mid- towns and villages. dle East, and Africa. Total e-commerce trade volume in- creased from less than ¥1,000 billion (US$120.8 billion) Clustering also occurred in terms of product specializa- in 2004 to nearly ¥30,000 billion (US$4.44 trillion) in tion. For example, more industrialized Taobao Villages BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA 2017. Against the backdrop of this impressive growth, along or closer to the eastern coastline have tended to online retail sales in rural areas have grown faster than the focus on manufactured goods, such as clothing, furni- national average. From 2014 to 2017, online retail sales in ture, and shoes. Meanwhile, more rural villages in the rural China increased from ¥180 billion to ¥1.24 trillion, west have specialized in local products, such as matrimo- a compound annual growth rate (CAGR) of 91 percent, ny vine (Lycium barbarum, the source of goji berries) in compared with 35 percent nationally.106 Ningxia, preserved meat in Guizhou, and black bee honey in Xinjiang. In 2018, more than 100 Taobao Villages sold The expansion of e-commerce in the rural parts of the locally produced agricultural products online. Flowers China can serve as a useful reference for Indonesia’s own from Shuyang, pecans from Lin’an, tea from Anxi, apples pursuit of expanding the reach of the digital economy to from Yantai, and Longjing tea from Hangzhou are exam- more rural parts of the country. Rapid growth of rural ples of products with unique local characteristics. The e-commerce in China can be attributed, in large part, same also applies to handicrafts, such as peony painting to the rise of Alibaba’s Taobao Villages. Taobao Villag- in Henan, Xuan paper in Anhui, birdcages in Shandong, es are rural villages in China where e-commerce is most and silverware in Yunnan. developed or concentrated. A village will automatical- ly be labeled as a Taobao village if at least 10 percent of This has led to a higher intensity of online activities also the households actively engage in e-commerce, or where within Taobao Villages. For example, the number of active there are at least 100 active e-shops, with annual online online shops in Taobao Villages has seen a sharp increase sales of at least ¥10 million (or US$1.5 million). A Tao- in recent years. Taobao Villages had about 70,000 active bao Town is a town that consists of at least three Taobao online shops in 2014 and more than 660,000 in 2018, a Villages. The number of Taobao Villages grew from just CAGR of 75 percent. The number of active online shops 20 in 2013 to 3,202 in 2018, and as Taobao Villages ex- ranges between 100 and 300 in most Taobao Villages, panded they clustered into Taobao Towns. The number though a few have more than 1,000 or even 10,000. While of Taobao Towns grew from 19 in 2014 to 363 in 2018.107 connectivity infrastructure, logistics, and broadband internet have driven growth, Taobao Villages have also taken advantage of the skills and enterprise of workers, → What are the characteristics many of whom were returnee migrants, who established successful e-commerce businesses, created role models of Taobao Villages? for fellow villagers, and shared knowledge within circles of friends and relatives.109 Taobao villages were mostly developed in clusters. Clus- tering mainly occurred initially in places where there was Overall, numerous studies show that Taobao Villages a well-developed industrial base and with rapid develop- have vitalized China’s rural economy and e-commerce ment of e-commerce related services.108 There was also in general and has made people better off. According to a significant demonstration as the success of one Tao - the Taobao Village Survey,110 online shop workers (other bao Village attracted its surrounding village to also start than the self-employed shop-owners) in Taobao Villages online businesses and become a Taobao Village. Local have wage levels equal to or higher than workers in ur- 163 ban private industries. The average monthly wage workers in system, offer tailored support to villagers, and link the local online shops ranged from some ¥3,800 for unskilled workers merchants to e-commerce platforms. to nearly ¥6,800 for technicians, compared with the nation- al average of ¥3,569 for urban workers in private industries. Well-developed express delivery services played a key catalytic The most popular jobs in these shops are non-technical, such role. The Taobao Village Survey finds that nearly all Taobao as customer service, marketing and promotion, or product Villages have package pickup and express delivery services, supplier and delivery. Rural e-commerce development also with a wide range of logistics service providers. Forty-eight increased household consumption growth, which stemmed percent had logistics centers in the village at the time of the from lower search costs and transaction costs that resulted in survey in 2018, compared with about 25 percent of villages in lower prices. Couture et al. (2018) show that the expansion China that had e-commerce logistic service distribution sites of e-commerce to the Chinese countryside is associated with in 2016. Sixty-two percent of Taobao Villages have more than lower costs of living, and for the goods that are available at five logistics or express delivery companies providing services CHAPTER — 03 both the Rural Taobao online terminal and in the village, the and 27 percent have more than 10 companies; all large Taobao median price from the online terminal is 15 percent cheaper. Villages (where GMV larger than ¥197 million) receive logis- The same study also shows that 62 percent of goods bought tics services from at least three companies. through Alibaba’s Rural Taobao platform were not available in the village.111 But the Government has also played a key role by providing various incentives and support to promote e-commerce in Despite the positive impact, firms, and workers in Taobao Vil- less developed areas. Important components of government lages still face many challenges, most notably from their small support in China have included collaboration with e-platform BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA size. The Taobao Village Survey finds that 64 percent of e-shop companies and logistics companies and subsidizing logistic owners are self-employed without employees, 30 percent have costs for packages to and from targeted poor areas. The Gov- fewer than five workers, 3 percent have five to nine workers, ernment has also provided subsidies and incentives to house- and only 3 percent have 10 workers or more. When an owner holds and individuals, including free or subsidized e-commerce starts hiring employees, 56 percent of the employees are family training, and free or subsidized e-commerce post-training members. While this allows flexibility of production and may support through e-commerce service firms, as well as awards make it easier to change products to meet evolving demand, it for high volume of online sales. limits the ability to scale up or upgrade the technology when demand surges. In addition, the Government has also launched integrated “demonstration programs” to target specific areas. The Rural Another challenge faced by the online shops is the fact that E-commerce Demonstration Program, launched jointly by most end up selling the same or similar products within their the Ministry of Finance and Ministry of Commerce in 2014, market. Under the framework of “One Village One Product” aims to contribute to poverty reduction and the modernization campaign, enterprises within one village may produce and of rural areas. Its main activities consist of establishing and sell similar products. This replication can result in low-end improving rural e-commerce public service, fostering rural competition within the same village or with an adjacent village, e-commerce supply chains, promoting connectivity between driving down profit margins or leading to disputes where there agriculture and commerce, and enhancing e-commerce train- are conflicting ownership rights for products, or where the ing. The program grew quickly and, by 2018, had supported process innovations are unclear. 1,016 demonstration counties, covering 737 poverty-stricken counties (89 percent of the total), including 137 counties with extreme poverty (41 percent of the total). According to Min- → What were the drivers of istry of Commerce, the program had created 120,000 jobs for poor households by the end of 2016. The number of online Taobao Villages? stores registered in rural areas increased from 8.2 million in 2016 to 9.9 million in 2017 and resulted in employment for 28 In general, the formation of Taobao Villages proceeded million people. At least 15 e-commerce companies have joined through three broad stages. The first stage entailed grassroots the effort to reduce poverty in rural areas.112 development, whereby villagers who were often returned mi- grants with entrepreneurial skills, led the establishment of In some cases, the Government also partnered with the private online businesses and created models for other villagers to sector to accelerate digital adoption in rural areas. In 2014, follow. In the second stage, the Government provided direct the Alibaba Group, in collaboration with the Government, support for infrastructure, e-commerce training, and finance. launched the Alibaba Rural Taobao Program (which is differ- The third stage was characterized by the involvement of spe- ent from the general Taobao Villages, which developed organi- cialized e-commerce service providers in areas where the in- cally without government intervention) to help give rural resi- dustrial base is weaker and human capital is more limited (i.e., dents greater access to a broader variety of goods and services, in central or western areas of China). These consulting firms and help farmers earn more by selling agricultural products were subsidized by local governments to foster the digital eco- directly to urban consumers in online platforms. The activi- 164 CHAPTER — 03 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA ties under this partnership include: setting up an e-commerce The public and private sectors in Indonesia have been seeking service network in counties and villages; improving logistical inspiration from the success of Taobao Villages to drive rural connections for villages through shipping packages between e-commerce in the country. Some existing rural economic re- county centers and villages; providing training in e-commerce vitalization plans already include efforts to encourage village and promoting entrepreneurship; and finally, developing rural enterprises to adopt digital technologies on a larger scale. Local financial services through the Ant-Financial subsidiary of Al- agencies at subdistrict and village levels, supported by the na- ibaba. The Rural Taobao Program expanded from 212 villages tional ministry of villages and disadvantageous regions, have in 12 counties in 2014 to more than 30,000 villages in 1,000 initiated “digital villages” program through internet broad- counties in 2018, spreading from the coast inland.113 band installation in villages, and the introduction to digital technologies through training and knowledge exchange. There are several success stories where the villages managed to sell → What are the lessons for their local products at the national marketplaces, or where local farmers deployed automated smart sensors to manage their Indonesia? fish farms. More notable achievements came from villages that successfully turned social media to promote their local tourism To a large extent, the success of Taobao Villages can be at- sites and attract local visitors tributed to the right blend of a natural progression along the existing value chain and the right government interventions to However, overall, the uptake remains small; only around 10 make e-commerce more inclusive by helping it spread to more percent of the total 74,957 villages in Indonesia have been rural parts of the country. In the initial stages, Taobao Villages registered in the Ministry of Village’s information system shared certain characteristics of market proximity, connec- (Sistem Informasi Desa dan Kawasan, SIDeKa). While more ad- tivity, and a pool of returned migrants with the necessary en- vanced villages have started utilizing social media platforms trepreneurial skills provided the spark. Digital technologies to promote their local products and services, most villages were the enablers that accelerated online commerce in these that have registered have just some basic information about areas. Naturally, areas with more solid connectivity, better their locality online. Meanwhile, some of the larger e-com- market access, and stronger industrial base, as those in the merce platforms have also started opening branches or dis- coastal areas, were the first to reap the digital dividends. In the tribution centers in rural areas to expand their market reach. later stages, government interventions stepped in to provide Though these efforts are still in their infancy, opportunities additional support for scaling up. Government interventions exist for governments—both central and local—to collabo- ranging from infrastructure provision to value-chain support rate with some of these platforms to provide complementary and entrepreneurial education were useful accelerators, with services, such as training, branding, value-chain support, etc., some e-commerce activities already in place. This support to enable small and medium enterprises in rural and more enabled the “winners” selected by the market to continue their remote places to capitalize on the opportunities generated expansion by upgrading products, exploring new clusters and by digital technologies. agglomerations to achieve economies of scale, and innovating. 165 Endnotes 106 Mofcom China, 2018. 107 World Bank and Alibaba Group, 2019. 108 AliResearch, 2016. 109 Zhang et al., 2015. 110 Taobao Village Survey is a between the World Bank, Alibaba Group, and Peking University and Nankai University joint re- search team that focuses on rural e-commerce clusters. It is representative for the universe of 2,118 Taobao Villages in 2017, conducted between August and September 2018. CHAPTER — 03 111 Couture, Faber, Gu, and Liu, 2018. 112 World Bank and Alibaba Group, 2019. 113 For discussion on public-private partnerships for e-commerce development, see Xubei Luo. 2019. E-commerce for poverty alleviation in rural China: from grassroots development to public-private partnerships. https://blogs.worldbank.org/easta- siapacific/e-commerce-poverty-alleviation-rural-china-grassroots-development-public-private-partnerships BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA 166 References 1 AliResearch. 2016. China Taobao Village Report 2016 (中国淘宝村研究报告 2016). 2 Couture, Victor, Benjamin Faber, Yizhen Gu, and Lizhi Liu. 2018. “E-Ccommerce Iintegration and Eeconomic Ddevelopment: Evidence from China.” National Bureau of Economic Research. 3 Ministry of Commerce, People’s Republic of China. 2018. “Regular Press Conference of the Ministry of Commerce (January 25, 2018)”. Retrieved from: http://english.mofcom.gov.cn/article/pressconferenceinyears/2018/201802/20180202708848. shtml 4 World Bank and Alibaba Group. 2019. “E-commerce Development: Experience from China” Vol.2 (English). Washington, DC: World Bank Group. Retrieved from: http://documents.worldbank.org/curated/en/552791574361533437/E-commerce- CHAPTER — 03 Development-Experience-from-China 5 Xubei Luo. 2019. “E-commerce for poverty alleviation in rural China: from grassroots development to public-private partnerships.” World Bank Group. Retrieved from: https:/ /blogs.worldbank.org/eastasiapacific/e-commerce-poverty- alleviation-rural-china-grassroots-development-public-private-partnerships 6 Zhang, Xiaobo, and Wu Zhu. 2015. “The spatial patterns of e-commerce in China.” Paper presented at the annual Allied Social Science Associations, January 3:5. BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA 167 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA CHAPTER — 04 168 Delivery Chapter ⬎ for Inclusive Service Digital Technologies 4 4.1. Introduction CHAPTER — 04 T he ability of governments around the world to harness digital technologies to deliver services is perhaps an area on which the COVID-19 pan- demic has focused the strongest spotlight. Just as people and businesses have had to adapt to the various measures taken to contain the spread of BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA the virus, so too have governments. The ability to use digital technologies in an effort to curb and manage the pandemic, as well as to ensure the continuity of essential services, has emerged as a key marker of resilience, along with capacity and political will. Countries such as the Republic of Korea, Singapore, Ger- many, Australia and Malaysia have been at the forefront, using these technologies not only to track the disease in real time, to conduct contact tracing and to implement quarantines as necessary, but they have also made use of digital technologies to coordinate and manage data, and organize their policy response. For example, the Rep. of Korea moved 470,000 classrooms into the digital space, enabling almost 4 million users, both teachers and stu- dents, to continue learning activities during the pandemic. Malaysia launched a digital-learning platform for teachers, students and educators that, by October 2020, had 1.7 mil- lion monthly active users, with 10,000 schools, 370,000 teachers and 2.5 million students in the system.114 Digital technologies have also been useful in the rapid deployment of social assistance responses. In countries with extensive mobile phone or internet penetration, as well as strong existing social protection systems, digital technologies related to digital ID and digital financial services have helped to facilitate the identification and registration of benefit recipients. For example, Brazil enabled workers not already registered in the national social protection registry to register through a website or phone application and start receiving emergency benefits for those who had become unemployed.115 Because Thai- land has enabled people to link their national ID to bank and e-money accounts through the highly interoperable PromptPay system, greater assurance was achieved that payments to beneficiaries who registered over the internet were indeed the right ones, and payments could be made more quickly and with higher confidence. Likewise, Chile 169 CHAPTER — 04 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA leveraged the ID-linked basic account, Cuenta-Rut, to bile learning apps and/or online schooling for distance deliver emergency relief. learning.121 Other countries are using this opportunity to do far more. For example, Singapore, already one of At the same time, COVID-19-related disruptions have the global leaders in this area, has ramped up its spend- laid bare the lack of readiness on the part of Indonesia to ing on digital technologies focusing on the development fully capitalize on these digital opportunities. Early ef- of new technology tools to respond to the COVID-19 forts by the Ministry of Education and Culture to deploy pandemic, the development of citizen- and business-cen- educational TV, as well as online resources, along with tric digital services, system development on the cloud, efforts from the Ministry of Religious Affairs and indi- modernization of government ICT infrastructure, and vidual teachers, meant that all students reported engaging the advanced use of digital analytics, AI and IoT sensors in some form of learning activities at home.117 Important within the public sector122 Vietnam is digitalizing two of challenges remain, however, with 64 percent of students its important national databases to enable e-identification reporting critical constraints related to the lack of reli- and authentication.123 Brazil has laid out an ambitious able connections and supporting devices.118 Similar to digital government transformation strategy for the next many other countries, Indonesia also struggled to obtain two years, which includes, among other things, plans to: consistent data on COVID-19 cases due to challenges in (i) strengthen digital identification with a target of is- integrating information systems across the various tiers suing 40 million identities by 2022; (ii) complete the of government, slowing down the response.119 The lack of consolidation of 1,500 government websites onto a sin- digital and other alternatives has caused widespread dis- gle portal; and (iii) develop cloud adoption and provide ruptions in non-COVID health care across the country.120 broader access to open government data.124 Some emergency social assistance that was approved by the Government, such as Kartu Prakerja (pre-employ- Since it is the poorer segments of the population that ment card, repurposed as temporary cash assistance to have been most affected by the pandemic—by job loss- unemployed workers), faced critical delays in deployment es and income reductions, as well as by the disruption due to difficulties in verifying the identities of recipients. of education—countries that have been able to better harness digital technologies in this manner have also Indonesia has taken some measures to apply digital tech- been successful in laying a more robust foundation for nologies to deal with the crisis at hand, but opportunities inclusive recovery. This chapter examines the landscape exist to further accelerate the adoption not just to build on digital technology adoption by the Government of resilience for future shocks but also to strengthen the Indonesia, looking at specific service delivery areas such Government’s ability to support a more inclusive future. as education and health, which have become more salient For example, to support students likely to fall behind during the pandemic, as well as cross-cutting foundation- in their learning due to limited connectivity, the Gov- al areas such as the introduction of digital IDs for online ernment introduced internet data packages (“airtime transactions and more wholesale digital transformation subsidy”) for students and teachers to support online of government. The focus is on identifying transformative learning tools. However, these subsidies do not appear opportunities and diagnosing some of the key challenges to have had an impact on the share of students using mo- that need to be addressed to realize these opportunities. 170 4.2. EdTech, HealthTech and the Role of Digital Technologies in Education and Health CHAPTER — 04 → Indonesia’s nascent EdTech adoption by educational institutions during the pandemic, the Government and some EdTech companies have also and HealthTech scene has created teacher-training modules to help teachers upskill received a major boost during by covering pedagogical and professional subjects, such the pandemic as classroom management, implementing project-based learning, tracking student performance, etc.128 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA In some policy circles, conversations regarding the role On the HealthTech side, innovative products and ser- of digital technologies in sectors such as education and vices are geared toward providing health services remote- health often begin and end with overwhelming optimism ly (telemedicine) through apps and websites, as well as about the power of these technologies to inspire a dis- cloud-based solutions for hospital information manage- junctive break with the past. In the context of developing ment systems. While telemedicine generally covers vari- countries that typically face complex and deep-rooted ous health services, including consultations with doctors, challenges in delivering quality services, the “disruptive” the provision of diagnosis, treatment and preventive care, promise of these technologies can be especially appealing. telemedicine in Indonesia is relatively new and still lim- Indonesia is no different and, over the years, as the digital ited primarily to teleconsultation. For example, companies economy has grown, so has the number of new entrants such as Halodoc and Alodokter—considered the most in the EdTech and HealthTech scene. According to an well-funded firms in the market—connect patients to online analytics portal125 (Tracxn) there are about 195 medical doctors to do online consultations.129 Hospital EdTech startups in Indonesia mostly established during information management systems have traditionally been the 2013–19 period. Of these, 60 companies were still offered by software developers and vendors catering for operational and offering a variety of products and services tailored solutions for hospitals. However, more recently, before the pandemic struck.126 Similarly, the Indonesian new firms such as Medico and Periksa.id have started to HealthTech Association reports that around 250 reg- offer cloud-based services (software-as-a-service). istered firms are operating in the HealthTech space. 127 The interest in these applications soared after the Indonesian EdTech companies provide a range of services COVID-19 outbreak. In Q2 2020, the number of new and products: (i) targeted at students to help them with users on Zenius jumped twelve-fold over the previous learning and upskilling; (ii) targeted at educators to as- year.130 Ruangguru, which had been growing steadily sist them with student management, communication and even before the pandemic, reported a jump in web-hits teaching; and (iii) targeted at educational institutions to from around an average of 7.5 million to over 11 million help them with administration. For example, companies per month. Industry estimates suggest that the usage of such as Ruangguru, Zenius and Quipper develop and telemedicine apps in Southeast Asia increased by a factor provide self e-learning content, interactive learning plat- of 4.5 in March 2020 (at the peak of the COVID-19 out- forms and study tools that help K-12 students expedite the break) compared with January 2020 usage. The number learning process, along with interactive online services that is consistent with growth seen by some of the prominent help students with their assignments and test preparation. players in Indonesia. Halodoc reported its monthly active Companies such as Arsa Kids, Digikids and Educa Studio users increasing by a factor of 10 during the pandemic develop game-based and blended-learning experiences, compared with the fourth quarter of 2019, while Alo- including interactive storybooks and educational mobile dokter claimed to have experienced a 1.5 times increase apps, to help improve early childhood educators’ effec- compared with the pre-COVID-19-outbreak period.131 tiveness. These products and services are typically dissem- Some of the larger digital platforms previously not en- inated using several approaches, such as web-based and gaged in the space have also expanded their offerings to mobile-based applications. As a response to elevated digital include health services (e.g., Grab-Health). 171 → While these apps have FIGURE 4.1 clearly filled a void during Better off households and those in the pandemic, their Jakarta are more likely to do online overall reach is limited to learning during the COVID-19 pandemic the more affluent clientele in urban centers, mostly SHARE (%) OF HOUSEHOLDS ENGAGED IN LEARNING ACTIVITIES within Java WITH MOBILE LEARNING APPS AND/OR ONLINE SCHOOLING In addition to an increase in volume, there CHAPTER — 04 have also been changes in the way people have started using some of these apps. For exam- ple, whereas respiratory issues accounted for just 24.5 percent of all consultations through Halodoc before the pandemic, these now ac- count for half of all consultations, suggesting that pandemic-related health issues themselves, as well as possible fear of exposure during a BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA physical visit to health facilities, have driven this growth. Growth in EdTech apps has also been driven by teachers and schools increas- ingly experimenting with various products to Source: World Bank Round 2 HiFy Household Survey 2020, World Bank staff calculations. try and figure out remote learning tools that Note: Samples are households with students who had been engaged work for them. in learning activities at home since March 2020. While these services are clearly filling an im- portant gap, especially during the pandemic, FIGURE 4.2 the reach of these digital solutions remains highly urban and Java-centric, and usage is Distribution of Halodoc higher among the relatively better off. The transactions across regions pandemic has exposed a digital divide that limits the penetration of the EdTech products and their potential to complement tradition- al teaching methods. This digital divide has been widening the education gap in Indone- sia, since many students from lower socioeco- nomic groups not only lack the resources to access online learning solutions but also lack supportive home environments.132 According to a nationally representative survey conduct- ed a few months after the pandemic outbreak, more than 83 percent of households in Jakarta were found to be providing some form of learn- ing activities using mobile apps and/or online learning to their children. The number was sig- nificantly lower in other parts of Java (43 per- cent), and even lower outside of Java (38 per- cent) (Figure 4.1). While on average 54 percent of all Indonesian families took up some form of digital learning to minimize the disruption in the education of their children, the option was available to only 38 percent of families in the Source: World Bank staff calculations based on Halodoc data. bottom 40 percent of the income distribution. Note: Pre-COVID refers to three months before the first COVID-19 case was identified in Indonesia (i.e., November to January Lack of complementary inputs such as support- 2020); Post-COVID refers to three months after the first COVID-19 ing ICT devices (smartphones, computers) and case was identified (i.e., February to April 2020). 172 FIGURE 4.3 limited access to good-quality internet are re- ported to be the main reasons behind this ob- served pattern. To avoid widening this educa- Most EdTech services are concentrated in Jakarta tion gap further, the Ministry of Education and Culture (MoEC) needs to create differentiated DISTRIBUTION OF FIRMS BY GEOGRAPHIC PENETRATION OF THEIR solutions for students and teachers with limited PRODUCTS (PERCENT OF SURVEY RESPONDENTS) access to digital infrastructure during the pan- demic. For example, it could invest in teacher training modules to help teachers adapt to new learning methods, deliver support and guidance for parents to provide more a supportive learn- ing environment at home, and establish more CHAPTER — 04 open-air schools or schedule regular teacher visits.133 Following the pandemic, connecting all schools to reliable, high-speed internet is both possible and essential.134 The regional breakdown of all Halodoc trans- actions before and after the pandemic shows a similarly limited geographic reach. Before the BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA pandemic, around half of all transactions were accounted for by Jakarta and West Java. With the exception of Sumatra, which accounted for about 11 percent of all transactions, only a small proportion of transactions were in islands off-Java (Figure 4.2). Similarly, most EdTech services are also largely concentrated on Java, with Jakarta being the largest market (Figure 4.3). As large population centers with higher levels of aggregate human capital, cities in Java also have large concentrations of high school students who are among the demographic most likely to be users of EdTech products to, among others, prepare for university entrance exams. Given that they live in the relatively more af- fluent parts of the country, the willingness to pay for these products is also higher in these places.135 → Digital innovations led by the private sector help enrich the menu of options, but the truly transformative impact, specifically from the perspective of inclusion, is likely to come from the Government experimenting with and internalizing some of these innovations to Source: World Bank EdTech Report 2020. enhance service delivery 173 CHAPTER — 04 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Massive-open-online-courses (MOOCs), once regarded ties with financing, lack of digital talent in the market- as one of the groundbreaking educational innovations of place, and poor regulatory clarity, including on sensitive the past decade, have had limited success. Completion topics such as consumer protection and personal data rates have remained below 5 percent for many years now, protection, etc. A vibrant innovation ecosystem that ex- with lack of coordination and direction considered to be tends the frontiers of these solutions is something Indo- among the main contributing factors.136 Even with the nesia should certainly strive for. However, as experience widespread pivot toward online education induced by the from the more developed parts of the world shows, as well pandemic, an interesting trend that has been observed as the current level of coverage and usage of these private throughout the world is how the specific technologies that sector-led initiatives within Indonesia indicate, digital have been adopted for the purpose have sought to replicate technologies should perhaps not be thought of as tools the classroom environment instead of deviating from it to displace the traditional modes of service delivery, but or improving upon it. Relatively better endowed schools, ones to help enhance them. So, rather than a multitude of even in Indonesia, have taken to Zoom-classrooms, while options with EdTech and HealthTech initiatives bringing across the country WhatsApp has provided a medium about disruptive transformations in these sectors, bigger for student-teacher interactions. Even as education has gains are likely to come from the Government’s readiness moved online, the pandemic has ended up underscoring to experiment with and internalize some of these digital the inherent irreplaceability of student-teacher interac- solutions to chip away at brick-and-mortar challenges. tions.137 Despite the potential promise of adaptive and personalized learning, one of the hurdles that EdTech Similarly, in health, in-depth interviews with some doc- firms need to overcome in the short term is to come up tors providing teleconsultations reveal that telemedicine with solutions that help to replicate/substitute/comple- is likely to be most useful for early-stage diagnosis and ment the classroom experience. educating patients. It is also likely to be useful for certain areas of medicine, such as mental health issues, that carry The EdTech and HealthTech sectors in Indonesia face a certain stigma in Indonesia and where patients may feel several challenges to their growth. These include difficul- more comfortable discussing matters through the digital 174 FIGURE 4.4 Only around 10 percent of all schools are connected to fixed broadband SCHOOLS CONNECTION BY TYPE, % OF SCHOOLS CHAPTER — 04 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Source: Dapodik, MoEC 2019. medium. But for a broad range of other medical consul- health-care services, such as the one introduced by Kota tations, doctors regard being able to physically examine Makassar in 2015, which provides 24-hour homecare and patients as an indispensable part of the diagnostic process. tele-medicine services to its residents. The Government has already been experimenting with A key challenge with all these efforts—and a theme we some of these technologies. In education, there have been will return to at the end of this chapter—is to move away multiple efforts to make material available online. Portal from ad-hoc piecemeal solutions to a more comprehen- Rumah Belajar is an online learning platform equipped sively thought out whole-of-government approach to the with complete multimedia education content managed by broader digital transformation of government and service MoEC. Universitas Terbuka, a state university focusing delivery. Another foundational challenge is connectivity. on distance learning, also offers extensive open educa- As discussed in Chapter 2, fixed broadband penetration in tion resources (OER). More recently, the Government Indonesia is very low compared with Indonesia’s regional has begun thinking about a back-office online tool for peers. This is an especially binding constraint for insti- schools to manage procurement, self-learning modules tutional users such as schools, hospitals and other health for teachers, and modularized online curricula to enable facilities, which may want to intensify their use of digital teachers to teach children to the right level of competency. technologies to enhance service delivery. For example, of Similarly, in 2018, the Ministry of Health launched its all the 219,000 schools under MoEC, just 10 percent had own teleconsultation platform, called Sehatpedia. BPJS access to fixed broadband connections in 2019, 42 percent Health, which manages Indonesia’s national health insur- had some form of mobile broadband connectivity, while a ance scheme, launched its own smartphone application, significantly large share (45 percent) was not connected JKN Mobile, in 2017, which initially served as a mem- at all (Figure 4.4). Closing these connectivity gaps, not ber management tool to help all citizens manage their just in schools but also in other important service delivery membership, including for invoicing, payment, and the nodes, will be a crucial step in harnessing digital technol- verification of claims. Another example is district-based ogies for greater inclusion in post-pandemic Indonesia. 175 4.3. Introduction of Digital Identification → For people to be able to fully be able to access all available services online. Citizens and e-residents of Estonia can use a digital ID combined with participate in the digital economy, electronic signature capabilities to carry out more than and carry out official and high- 99 percent of government transactions completely on- value transactions online, CHAPTER — 04 line, without visiting an office or using any paperwork. For many developing countries, this can be a key cost-efficient countries need to introduce digital strategy for making services available to more people, es- ID systems—a need that has been pecially those in remote and underserved communities. accentuated by the COVID-19 • A platform for other digital economy transactions: Be- cause they identify people, digital IDs can be the basis pandemic of dependent value-added functions, such as electronic signatures for entering into contracts, digital payments, BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA and allowing people to exercise control over their personal The ability of a person to verify her/his legal identity is often data. Singapore’s MyInfo service, which allows citizens and a prerequisite to their ability to access services and exercise residents to manage and share their personal data from many rights. A service provider expects to have trust that the trusted government sources, is built on the national digital person in front of them is indeed who she/he claims to be. ID system, SingPass, to link a person to their own data and New mechanisms known as digital ID systems138 are needed to allow them to provide consent for its use and re-use. to enable people to establish and verify their legal identity • Enabling cross-border transactions: Digital IDs can be in a reliable, secure, privacy-preserving and remote manner, recognized across borders for doing business. In the Eu- without the need for a physical intervention, such as having to ropean Union, the eIDAS regulation and proposed reg- go to an office or posting a document. Digital ID systems for ulation for decentralized digital identity wallets govern online transactions are a natural progression from national ID a framework in which a digital ID issued by one member systems, which have been used predominately for in-person state is recognized for all official transactions in other transactions, including in Indonesia, because a physical ID member states, meaning that trade and business can be card cannot be used remotely over the internet. carried out without needing to be physically present in another country. Digital IDs make use of technologies such as smartphones and cryptography to provide such security and assurance for remote interactions. They can be issued by a single entity (cen- → The COVID-19 pandemic has tralized) or by multiple entities in an ecosystem (federation). Emerging standards are also creating opportunities for decen- revealed that digital ID systems tralized models where the digital ID is stored on a personal stand alongside accessible and device or digital wallet. affordable internet connectivity as Well-designed and implemented digital ID systems can unlock a core national infrastructure, and an enormous amount of economic value, estimated by the McK- are central to achieving resilience insey Global Institute to be between 3 and 13 percent of GDP by 2030.139 The benefits of digital ID can be described as follows: With quarantines and the scaling-down or shutdown of many government and commercial operations, countries had to try • Promoting inclusion: Digital ID systems can facilitate to rapidly shift services to digital channels to keep them going. inclusion when they are universally accessible and usable. Countries that had invested in and built strong digital ID sys- Aadhaar, India’s digital ID system, contributed to increas- tems before the pandemic were by and large better able to do ing financial inclusion from 35 percent in 2011 to 80 per- this than countries that had not. Estonia, Singapore, the United cent in 2017, by making it cheaper and easier for people Kingdom and other digitally advanced countries were able to to satisfy know-your-customer (KYC) requirements for reduce disruption for their public service delivery. But the differ- bank account opening. ence was particularly telling in terms of whether countries were • Shifting services to digital channels: A person with in- able to deploy quick, accurate and effective social protection ternet access and an officially recognized digital ID should measures to mitigate the shocks and to safeguard livelihoods. 176 FIGURE 4.5 Digital ID systems depend on having strong foundational ID systems in place CHAPTER — 04 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Source: Analysis by the World Bank Group’s Identification for Development (ID4D) Initiative. This most critically included reaching populations such as in- While many countries have introduced digital ID systems, formal workers who were difficult to target because they were there has been mixed success in terms of realizing systems that not present in either databases of people living in poverty or are truly inclusive and trusted. Success comes in the form of databases of people contributing to social security. widespread adoption by the population and by service pro- viders, and impacts such as enhanced access to services, cost Thailand, for example, built a website to allow informal work- savings and productivity gains. Low- and middle-income econ- ers to apply for emergency payments online, using its national omies have a great opportunity to introduce digital ID systems, ID system to verify the identities of applicants, establish their since they have fewer legacy systems to overcome, which have uniqueness and determine their eligibility by cross-checking constrained the development of digital ID systems in some government databases. Since national IDs were already linked high-income countries. to bank and mobile money accounts, there was added assurance that the cash transfers would reach their intended beneficiaries. The challenge is that country experience will vary, but there Over a few weeks, more than 28 million citizens applied, with are some commonalities. First, issues around identity and ID more than 15 million deemed to be eligible. Payments began for systems often antagonize social, cultural, political, econom- some within days of applying. In Chile and Peru, social registries ic and legal fault lines. It is therefore crucial that design and were used to expand existing social assistance programs and implementation reflect the local context, and are informed citizens could use their digital ID to access a portal that allowed by meaningful consultations with a wide set of stakeholders, them to check to see if they were included. including civil society, and that ‘analog’ enablers, such as a comprehensive legal framework for the digital ID system and data protection more broadly, are in place. Second, since digital → Digital ID systems can provide ID systems are ‘new’ and may be replacing existing systems or processes, including potentially previously failed digital enormous benefits, but they are ID system projects, there can be inertia or legacy challenges. not easy to build. If not well- Third, the absence of compelling use cases across both the pub- designed and implemented, they lic and private sectors will lead to weak demand from the pop- ulation, and thus low adoption and impact. Fourth, the system come with risks that include design, including chosen technologies, may not be compatible exclusion and privacy breaches with the local context. For example, a country that has low rates of smartphone ownership should not expect widespread 177 FIGURE 4.6 The Principles on Identification for Sustainable Development provide a guiding framework on how countries can realize the transformational potential of digital ID systems, while mitigating the risks INCLUSION 1. Ensure universal access for individuals, free from discrimination 2. Remove barriers to access and use. CHAPTER — 04 DESIGN 3. Establish a trusted—unique, secure, and accurate—identity. 4. Create a responsive and interoperable platform. 5. Use open standards and prevent vendor and technology lock- in. 6. Protect privacy and agency through system design. BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA 7. Plan for financial and operational sustainability GOVERNANCE 8. Protect personal data, maintain cybersecurity, and safeguard people’s rights through a comprehensive legal and regulatory framework. 9. Establish clear institutional mandates and accountability. 10. Enforce legal and trust frameworks through independent oversight and adjudication of grievances. Source: World Bank Group’s Identification for Development (ID4D) Initiative 2021. adoption of a digital ID that depends on such devices. Fifth, Sustainable Development (Figure 4.6), which have been endorsed a country may not have the pre-requisites of trusted identity by 25 international organizations, offer a guiding framework data and mechanisms to verify those data (Figure 4.5), and may on how these risks can be mitigated and how the full benefits instead need to focus on building those. Finally, the knowledge of digital ID systems can be realized. asymmetry that sometimes exists between governments and technology vendors can lead to unfavorable contracts and PPPs that may have long-term consequences. → While Indonesia does not yet The three most common risks related to digital ID systems have an official digital ID system are: (i) exclusion; (ii) data protection and privacy breaches; or ecosystem that would allow and (iii) vendor and technology lock-in. Exclusion refers to Indonesians to securely verify preventing people not just from registering for a digital ID but also using a digital ID, such as by using technologies or their legal identity online, the processes that are not compatible with the local context, as existing national ID system is an described above. Data protection and privacy breaches require asset could be leveraged to create a dual approach of developing comprehensive legal frame- works that ensure accountability, including independent one relatively easily oversight. Vendor and technology lock-in comes in various forms, such as the use of proprietary data formats or depen- The Ministry of Home Affairs and the Ministry of Commu- dence on a technology provider to maintain a system without nications and Information have expressed strong interest in knowledge transfer. This can be mitigated somewhat through designing and launching an official digital ID system. In the the adoption of open standards, open-source software where absence of such a system, online service providers are using appropriate, competitive procurement, and strong contract unreliable and insecure mechanisms to verify the identity of and vendor management. The ten Principles on Identification for Indonesians online, such as requesting customers to take sel- 178 fies holding the national ID card (KTP-el). While this is a practical workaround, some fintech providers have reported that as many as 60 percent of customers provide selfies that are unreadable or require manual intervention, such as a video call, creating unnecessary expense and challenges. More recently, licensed e-signature provid- ers have begun offering basic digital ID authentication services (i.e., going lower in their value chain), but these are based on commercial relationships with the service providers and require the charging of higher fees in order to make a profit. Other third parties have emerged of- fering ID authentication services, but again these charge CHAPTER — 04 substantial fees. For example, the Kartu Prakerja web- site had to purchase facial recognition services from a private company when it rolled out applications during the COVID-19 pandemic. Similar constraints exist for face-to-face transactions as, despite some pilots, the Directorate General for Population and Civil Registra- tion (Direktorat Jenderal Kependudukan dan Pencatatan Sipil, Dukcapil) under the Ministry of Home Affairs has BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA not yet launched biometric authentication or electronic know-your-customer (e-KYC) services nationwide. There is substantial demand for digital ID in Indonesia across the public and private sectors. E-KYC and digi- Coverage of the existing national ID system is high at the tal ID remains one of the priority issues for the financial national level, with some geographic disparities and gaps sector, from banks to fintech companies. Likewise, gov- in registering the births of children. According to the 2017 ernment agencies such as BPJS Employment and BPJS ID4D-Findex survey, 96 percent of Indonesians aged 16 Health are trying to transform the way they offer services and over have a KTP or a KTP-el. The 2018 SUSENAS to citizens, including by shifting to online channels, but found that 94 percent of Indonesians have an ID number are hampered by the absence of an accessible and afford- (Nomor Induk Kependudukan, NIK), but with coverage gaps able digital ID system. At the regional level, there is an in Papua (where only 52 percent of the population have a opportunity for mutual recognition arrangements to en- NIK) and among children (of which only 75 percent have able cross-border transactions as a stepping-stone toward a NIK). Notwithstanding these gaps, which Dukcapil has a broader ASEAN-level arrangement, which has been committed to swiftly close, achieving this high level of called for in the ASEAN 2025 Digital Masterplan. Beyond coverage is a considerable accomplishment, considering ASEAN, there is also an opportunity for mutual recog- the population size, diversity and geography of Indonesia. nition arrangements with the European Union (eIDAS), Australia, the United States and Canada, among others. Indonesia’s national ID system, which is managed by → Digital ID is a ‘low hanging Dukcapil, is well-established and provides a strong basis fruit’ for Indonesia to make for introducing a digital ID system or framework of sys- services more inclusive, to tems. It builds on a long tradition of civil registration— promote trust in the digital the recording of births, deaths, and marriages—that was introduced in various ways during the colonial period economy and society, and to and continued after independence in 1945. The most re- create new drivers of economic cent modernization was a large IT project that started in growth. Doing so will also 2011 and launched the new KTP with an electronic chip (KTP-el), which was provided free of charge to citizens, allow Indonesia to join all introduced biometrics (fingerprint, iris and facial rec- its fellow middle and high ognition), and enhanced the population registry (Sistem income ASEAN Member Informasi Administrasi Kependudukan, SIAK) and other Dukcapil functions. The biometrics system was able to States who have launched assist with removing duplicates in the SIAK, as well as to whole-of-economy digital ID enable identity verification. frameworks. 179 4.4. Digital Transformation of Government In addition to digital ID, another important and broader income tax reports submitted, were made through the pathway to harnessing digital technologies for greater in- e-Filing system.140 Lapor!, which has now been integrated CHAPTER — 04 clusion is by enabling these technologies to fundamentally within 623 government institutions (including 34 minis- transform the quality of citizen-state interactions. This tries, 74 line agencies and 515 subnational governments), could be for specific services, such as education, health also has 790,000 registered users and has been receiving and social protection, but also for a whole host of other around 560 daily reports over the past two years, spiking services that could be improved significantly by a broader to 600 during the pandemic.141 digital transformation of government. As discussed in Chapter 1, Indonesia is behind regional peers and compar- There have also been several attempts by local governments ator countries in terms of its readiness to provide effective to harness digital technologies in their respective domains BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA digital services to its citizens. There are many challenges and for specific purposes. For example, West Java has en- hindering government digitalization and, as discussed abled motor-vehicle tax payment through an e-commerce in Chapter 2, limited connectivity, particularly to fixed platform. Similarly, the province of Jakarta has developed broadband, is chief among them. Connectivity in public an app (Alpukat Jakarta) that purports to help citizens facilities such as schools, health posts and government obtain civil registration documents. While some of these agencies at the central and local levels is low, and mak- first-generation efforts have been moderately successful ing progress on this is a crucial necessary step. The key in their respective objectives, they have been fairly limited challenge will be to transition from the current siloed in terms of their scope and often ad-hoc in nature. These structure of multiple, incompatible government infor- agency and subnational entity level attempts to digitalize mation and data management systems to a whole-of-gov- have resulted in a proliferation of information systems, ernment, platform-based approach, which has emerged as websites, apps and platforms that have very limited in- best practice in many economies globally. Related to that teroperability at the front end and create significant du- is the need to establish clear leadership and coordination plication of effort and investment at the back end. for government digital services. In the past two years, the Government has made a strong effort to address these digital government issues. Two → The Government has taken concrete steps to back up this commitment have been the issuance of Presidential Regulation No. 95/2018 on some concrete policy measures e-Government and Presidential Regulation No. 39/2019 in recent years to advance its on One Data. The objective of the regulation on e-gov- digital transformation agenda ernment is to implement an integrated e-government sys- tem by, among others, getting all government agencies to Over the years, the Government has made several efforts adopt a common and interlinked enterprise architecture, to digitize government services. Among some of these co-using IT systems and establishing a national coordina- first-generation efforts, Lapor! and e-Filing are probably tion team to resolve common issues. The One Data regu- the most well-known. Lapor! (which translates to the lation seeks to facilitate the production and dissemination word “report” in English) is a complaints-handling sys- of high-quality data with a high degree of integrity by tem that allows the public to submit reports and complain requiring agencies to adhere to data governance standards about any services. Initially developed in 2011 by the Pres- (e.g., standardized metadata, interoperable formats, etc.). ident’s Delivery Unit, it is now operated by the Ministry of In addition, the President recently announced five direc- Administrative and Bureaucratic Reform (MenPAN-RB) tives to accelerate digital transformation that included, with support from the Office of the President. Meanwhile, among others, the preparation of sectoral digital trans- e-Filing is a system to file annual income tax returns on- formation roadmaps and accelerating the development line managed by the Directorate General of Taxes (DGT) of a National Data Center.142 These regulations and the under the Ministry of Finance. It was officially launched in presidential directives set a firm foundation for Indonesia 2015 to make the reporting process simpler for taxpayers. to get started on its digital transformation agenda. In 2019, 10.6 million, or roughly 97 percent of all annual 180 FIGURE 4.7 Institutional structure of key e-government stakeholders in Indonesia POLICY STRATEGY PLANNING FUNDING PROCUREMENT DELIVERY OPERATIONS Setting of Define Scope & write Review & Purchase Design, Operate & policies & priorities business approve of project test & maintain standards & roadmap cases funding needs implement systems RESPONSIBLE Execute & MEN PAN RB COORDINATION MEN PAN RB complete TEAM CHAPTER — 04 activities KOMINFO Men PAN RB BAPPENAS KOMINFO KOMINFO MOF BAPPENAS MEN PAN RB MOHA KOMINFO MOF BSSN KOMINFO BPPT MOHA LKPP BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA BPK BSNN WANTIKNAS, KSP, KPK, BPK and rest of the Government BPKP BPPT AGENCIES / MINISTRIES / SNGs INFORMED Informed of Rest of Government (as and when required) activity & progress Source: World Bank analysis, BAKTI, BCG experience, BCG analysis. → Experience in implementing the e-government regulation highlights severe challenges related to the fragmentation and coordination of institutions, stakeholders with overlapping mandates, and the lack of central leadership with the authority to drive the agenda This is a vast landscape and there are several challenges related is particularly complex and highly fragmented with multiple to common standards, policies and an interoperable framework agencies having overlapping mandates. (Figure 4.7) This has for core back-office information systems, a government cloud several adverse implications. First, resources to deliver digital (for sensitive data), a government cloud-first policy to enable services are thinly spread across the various agencies, with each the use of commercial cloud solutions for non-sensitive data, a agency having an incentive to underspend on foundational, in- government digital payments platform, and technology acqui- teroperable common platforms and overspend on agency spe- sition and financing. But learning from efforts to implement cific applications. This leads to a waste of resources, as some of the e-government and One Data regulations, this chapter fo- the back-end functionalities are likely to be duplicated across cuses on just three high-level issues, namely: (i) leadership and the various agencies. coordination; (ii) digital talent within government; and (iii) the fragmentation of government data. Each of these is a foun- Second, fragmentation naturally also makes coordination a dational challenge that needs to be overcome for Indonesia to major challenge, which increases the complexity of completing further crystallize and realize its digital transformation vision. even the simplest of tasks that require cooperation among different stakeholders. For example, MenPAN-RB must work Global experience shows that digital transformation of gov- with at least three different deputy ministers within Bappe - ernment is a complex undertaking and involves multiple stake- nas when discussing and formulating e-government policies. holders. In Indonesia, the public sector’s institutional structure MenPAN-RB must communicate with the Directorate of State 181 FIGURE 4.8 UK Government’s cross-sectoral digital transformation governance structure POLICY STRATEGY PLANNING FUNDING PROCUREMENT DELIVERY OPERATIONS Setting of Define Scope & write Review & Purchase Design, Operate & policies & priorities business approve of project test & maintain standards & roadmap cases funding needs implement systems RESPONSIBLE Execute & complete activities GDS CROWN CHAPTER — 04 GOVERNMENT DIGITAL SERVICES COMMERCIAL GDS (GDS) SERVICES ACCOUNTABLE (CCS) HM sign off TREASURY & approve activities GDS BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA AGENCIES / Rest of Government MINISTRIES AGENCIES / MINISTRIES INVOLVED INFORMED CABINET OFFICE Informed of activity & progress AGENCIES / MINISTRIES INVOLVED Source: World Bank analysis, BAKTI, BCG experience, BCG analysis. Apparatus to talk about overall planning and budgeting of Italy have some variant of a similar model. Japan is currently e-government, with the Directorate of Systems and Procedure discussing a proposal to establish a similar digital service unit to talk about the integration of planning and budgeting, and within its parliament.143 with the Directorate for Monitoring and Evaluation when dis- cussing the design of national data architecture. A coordinating body, such as the GDS in the UK, with au- thority over all relevant stakeholders is necessary to overcome Countries that have made good progress on digital transfor- these challenges and drive the agenda. In the absence of such mation of government have significantly simpler institutional an agency, the digital transformation agenda in Indonesia structures. The UK model, for example, is designed to stream- has not been able to garner any real momentum. One of the line decision-making processes for major digital government integral first steps in the deployment of coordinated digital initiatives. (Figure 4.8) The institutional arrangement grants services as envisioned in the e-government regulation is the the Government Digital Service (GDS) the autonomy to scru- streamlining and integration of various applications related tinize and assess digital service initiatives that are above cer- to the planning, budgeting and archiving functions used by tain threshold in terms of financial value. GDS also plays a various agencies. In addition to the planning and budgeting role in implementation, especially on digital services that cut application used at the national level, there are more than 50 across multiple sectors and agencies such as Gov.UK (one-stop planning and budgeting applications that are used by subna- service), Gov.uk Pay (payment service), Gov.uk Verify (ID tional governments. More than two years since the regulation service), and Government Cloud. In addition to policy and im- was promulgated, the e-government coordination team, which plementation roles, GDS is also responsible for managing the consists of seven line ministries and agencies (Menkominfo, digital, data and technology (DDaT) stream for professionals MenPAN-RB, Bappenas, MoF, MoHA, the Agency for As- within the civil service. As the primary agency responsible for sessment and Implementation of Technology [Badan Pengka- driving the digital transformation, GDS has the authority to jian dan Penerapan Teknologi, BPPT] and the National Cyber streamline and simplify the deployment of digital public ser- and Crypto Agency [Badan Siber dan Sandi Negara, BSSN], vices across the entire government machinery. Other countries with MenPAN-RB in the coordinator role) has been unable such as Singapore, the United States, Australia, Canada, and to convene and agree upon the shared vision of e-government 182 CHAPTER — 04 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA implementation. While this could be partially attributed services to every citizen. Examples of a common data to the limited capacity to produce alternative solutions exchange platform in Estonia,145 digital single sign on and service standards for planning and budgeting applica- ID in Singapore,146 and digital payments in the United tions, lack of leadership plays a key role as well. This kind Kingdom147 show how pivotal digital government plat- of integration requires visionary leadership that possesses forms can be in allowing nations to operate effectively strong technical knowledge but also political influence to in the digital space and to stimulate innovations in the drive change. market. The reform would need to be spearheaded by a central government authority such as the Office of the The current thinking on digital transformation as embod- President, to strengthen the political and bureaucratic ied in the e-government initiative focuses to a large extent influence needed for such far-reaching reforms. on the digitization of internal government processes. Dig- ital transformation should be limited not just to this, but a If the Government were to consider taking concrete steps fundamental reimagination of processes, procedures and in this direction, another key challenge that may emerge structures considering what digital technologies make would be the level of digital skills within the civil service. possible. Looking at the scope in the regulation, the role Currently, there are around 4 million civil servants (pegawai fits the main duties and responsibilities of MenPAN-RB, negeri sipil, PNS) in Indonesia, with over 50 percent com- which includes the management of government business prising specialist or functional roles (such as teachers). ICT processes. However, if the Government aims to have a specialists account for just 0.17 percent of the total civil comprehensive digital transformation of the nation, it servants. This small number of ICT specialists indicates has to gradually depart from e-government and slowly that the Government has been relying largely on external rearrange its vision, governance, and delivery model to resources, such as contract workers or third-party vendors, better suit that aspiration. Such a commitment would to deliver ICT solutions. While collaboration with external reorient Indonesia in the same direction as that taken parties is to be expected and welcomed, without strong by digitally advanced countries, such as Singapore, the in-house capability, the sustainability of ICT initiatives United Kingdom, Australia, Estonia, the Rep. of Korea remains at risk. Going beyond ICT specialists, if a broad- and Canada.144 The digital government platforms would er digital transformation strategy is to be conceived and serve as the foundation of digital services in the economy, implemented, the Government may need to nurture an so that both public sector institutions and private enter- entirely new cadre of programmers, software developers prises can take advantage of these platforms to deliver and data scientists to support these efforts. 183 FIGURE 4.9 Fragmented data and overly complex interdependence prevents BPJS Kesehatan from running efficiently CHAPTER — 04 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Source: World Bank analysis, → Fragmentation of the Unified Social Welfare Database (Data Terpadu Kese- jahteraan Sosial, DTKA) of the Ministry of Social Affairs. government data is another Furthermore, to ensure compliance, BPJS Health needs important challenge that needs to be connected to the tax ID system of the Ministry of to be overcome to bring about Finance. To verify enrolment, the system must also have interconnectivity with health facilities and health records. digital transformation This does not necessarily mean that the Government needs to unify everything. What it could do is ensure a minimum Data are foundational. Data fragmentation keeps govern- level of inter-connectedness and interoperability of these ment institutions from functioning efficiently and impinges data systems (Figure 4.9). The use of digital technologies on their ability to deliver effective services. The COVID-19 can facilitate this process and, indeed, digitally intercon- pandemic has focused a spotlight on this issue.148 A case in nected and interoperable data systems are a cornerstone point is the national health insurance system managed by of digital transformation of government. BPJS Health. The Government needs to link BPJS Health participants with their national ID number (NIK), which The presidential regulation on One Data aims to address is issued and managed by the Ministry of Home Affairs. the challenges of data fragmentation across government For those who are poor and eligible for social assistance, agencies by regulating agencies to comply with data gov- the database of BPJS Health also needs to be linked with ernance, ensuring data adhere to standards, standardized 184 FIGURE 4.10 One Data as an opportunity to address data fragmentation STEWARD FOR STEWARD FOR STEWARD FOR STATISTICAL DATA FINANCIAL DATA GEOSPATIAL DATA STREERING COMMITTEE CHAPTER — 04 SECRETARIAT INDONESIA ONE DATA BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA DATA STEWARD DATA CUSTODIAN K/L/D DATA CUSTODIAN K/L/D DATA PRODUCER K/L/D DATA PRODUCER K/L/D Source: KSP 2019. metadata, and interoperability format. One Data follows in- the standards that are defined by the data stewards. As head of ternational practices in data management and establishes the the steering committee, Bappenas is tasked with the regulation, high-level institutional arrangement and division of roles among monitoring and enforcement across the entire government. institutions in relation to government data management. The regulation also specifies a governance structure that includes a While the One Data regulation has laid the necessary foun- One Data steering committee, data stewards, data custodians, dations to enable better data governance, several challenges and data producers (Figure 4.10). The steering committee, have emerged in its implementation. First, One Data appears chaired by Bappenas, is responsible for laying out the broader silent on the role of administrative data, which is an important data policy, while data stewards (pembina data) are tasked to omission given the foundational nature of these kind of data work as business owners of the data, and responsible for defining generated by government agencies in a variety of relevant sec- and enforcing data standards in their respective sectors. The tors. Second, the intersectionality of the One Data regulation regulation identifies three data stewards: BPS for statistical data, with other relevant initiatives, such as e-Government/Digital MoF for financial data, and the Geospatial Information Agency Government (Presidential Regulation No. 95/2018), the Elec- (Badan Informasi Geospatial) for geospatial data. Data custodians tronic Transaction Law and its implementing regulation (PP (wali data) are supposed to be technical custodians and the focal No. 82/2012 and its revision PP No. 71/2019), the forthcoming points for data-sharing and exchange. Finally, data producers Personal Data Protection Law149, the Civil Registry Law, the are the frontline of data collection in their respective agencies, Digital Payment Regulation, and the Omnibus Law is another and are responsible for collecting the data in accordance with area that requires further clarity.150 185 4.5. Conclusions CHAPTER — 04 This chapter has looked at the opportunities and chal- the private sector to come up with innovative tools, true lenges that Indonesia faces in using digital technologies transformation will ultimately depend on the extent to to deliver better services to its citizens. Unlike many other which the Government itself can experiment with and countries around the world that have been able to utilize adopt some of these tools to bring about incremental digital technologies to deal with the public health aspects improvements in efficiency. One barrier to doing this is of the pandemic, as well as provide rapid economic re- limited internet connectivity, especially the more robust lief to the affected population, Indonesia has not had the connectivity that comes in the form of fixed broadband, BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA digital readiness to do so as effectively as it could. The in key service delivery nodes. increased adoption of digital technologies and lessons learned from the pandemic create a great opportunity The chapter also identifies the adoption of robust, reli- for Indonesia to invest in the foundational elements of a able and highly interoperable digital IDs as a potential broader digital transformation of government, not only to game-changer in the way the Government and the pri- build resilience for future shocks, but also to address the vate sector provide a broader set of services to its people. fundamental challenge of economic and social inclusion Assessing progress on the recent the government regu- in the medium term. lation on e-government, the chapter identifies coordi- nation across a variety of fragmentated initiatives and On critical services such as education and health, the na- stakeholders with overlapping mandates, a lack of digital scent EdTech and HealthTech sectors have seen an ex- skills and awareness within the bureaucracy, and the lim- plosion of interest, which has also been followed by an ited integration and interoperability of government data infusion of new capital. However, this chapter finds that systems to be key bottlenecks for progress. Rather than a the reach of these innovative solutions from the private heterogeneous agency-by-agency approach to digitaliz- sector remains limited, as they cater primarily to the rel- ing government, the chapter calls for a ‘whole-of-gov- atively more affluent households in the more prosperous ernment’ approach that establishes common standards, places within the country that have more access to digital practices, and interoperability across a variety of use cases. infrastructure. Even as learning has moved online and Assessing the current institutional landscape, the chap- the fear of exposure has led many health-care seekers to ter also concludes that the coordination and leadership take up teleconsultations, the pandemic has in a way also required to embark on such an ambitious agenda would underscored the fundamental irreplaceability of the face- need to come from a higher-level authority, and propos- to-face interactions in the delivery of key services such as es a digital transformation agency, potentially reporting education and health. What this implies for policy is that, directly to the President. Details on what this agency while the challenges in the innovation ecosystem should might look like and other recommendations for policies be addressed to continue to foster an environment for are presented in Chapter 5. 186 Endnotes 114 Open Gov Asia, 2020. 135 Bhardwaj, Yarrow, and Cali, 2020. 115 ILO, 2020. 136 Rita Kop, 2019. 116 Rutkowski, et.al., 2020. 137 Education experts and practitioners argue that 117 Yarrow et al. 2020. online learning cannot replace offline face-to-face learning. To avail of it one requires sufficient means 118 Indonesia High Frequency Monitoring of Covid-19 and technology, which can be exclusive in some Impacts Round 4, World Bank 2021. countries, and there are psychological and social 119 Tempo, 2020. factors that prevent online interactions from being CHAPTER — 04 120 Between April and May 2020, almost 80 percent as effective as offline (Dhawan 2020). of Posyandu (integrated maternal and child health 138 For brevity, the term “digital ID system” is used to and nutrition posts at the village level) were closed. describe both a single system and an ecosystem/ Community-based outreach activities also report- framework that may comprise multiple digital ID ed significant disruptions with more than 40 per- systems (e.g., a federation of digital ID systems). cent ceasing operation. It is also reported that 86 See more on White et al, 2020. percent of child growth monitoring activities, 55 139 White et al, 2020. percent of immunization services, 46 percent of 140 This is nevertheless just one-quarter of all regis- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Vitamin A distribution and ante-natal services were suspended or ceased due to these health posts tered taxpayers, highlighting the larger problem of closing down. tax compliance in the country. 121 Indonesia High Frequency of Monitoring of 141 Lapor! statistics indicate that in 2019, out of Covid-19 Impacts Round 4, World Bank 2021 196,437 complaints, 28,773 (14 percent) were from Java, followed by Sumatra (2 percent). Only around 122 W Media, 2020. 0.15 and 0.12 percent of the total complaints came 123 Vietnam News, 2020. from Papua and Maluku. Most complaints came 124 Forbes, 2020. from men (87 percent), moderately well educated (60 percent with more than nine years of educa- 125 Tracxn, 2019. tion), and those in the private sector (47 percent). 126 Bhardwaj, Yarrow, and Cali, 2020. 142 The other three directives relate to expanding dig- 127 Asosiasi Healthtech Indonesia, 2020. ital connectivity, preparing and nurturing digital talent and mobilizing financing to commence dig- 128 KrAsia, 2020. ital transformation. 129 The most updated data on health-tech funding was 143 Japan Times, 2020. prepared in September 2019 by Tracxn. (https:/ / tracxn.com/explore/HealthTech-Startups-in-In- 144 The group of leading digital nationals: https:// donesia). Since then, there has not been any ma- leadingdigitalgovs.org/ jor announcement of health-tech funding, except 145 Estonia’s Data Exchange Platform or widely known for Alodokter’s follow-on funding by MDI Ven- as X-Road: https:/ /e-estonia.com/solutions/in- tures in November ’20 (https:/ /theinsiderstories. teroperability-services/x-road/ com/indonesias-alodokter-raises-series-c-fund- ing-from-mdi-ventures/) . In October 2019, Grab 146 Singapore ID Service or Singpass: https://www.sing- and Ping An announced a joint venture to launch pass.gov.sg/spauth/login/loginpage?URL=%2F&- Grab Health / Good Doctors Indonesia, but the TAM_OP=login value of the investment to Indonesia was never dis- 147 UK Payment Service Platform: https://www.pay- closed. ments.service.gov.uk/ 130 Katadata, 2020a. 148 Christy Pettey. 2019. 131 Katadata, 2020b. 149 Bisnis.com, 2020. 132 East Asia Forum, 2020. 150 For example, data residency and data sovereignty 133 ibid. related issues in the Personal Data Protection Law are likely to have implications for implementation 134 World Bank, 2020. regulations for One Data. 187 References 1 Alifia, Ulfah. 2020. “COVID-19 is widening Indonesia’s education gap”. East Asia Forum. Retrieved from: https://www. eastasiaforum.org/2020/10/22/covid-19-is-widening-indonesias-education-gap/ (accessed November 16, 2020). 2 Annur, Cindy Mutia. 2020b. “Layanan Telemedicine Diprediksi Tetap Berkibar Usai Pandemi Berakhir.” Jakarta, Indonesia: Katadata. Retrieved from: https:/ /katadata.co.id/ameidyonasution/digital/5f45270a9ad78/layanan-telemedicine- diprediksi-tetap-berkibar-usai-pandemi-berakhir (accessed November 16, 2020). 3 Asosiasi Healthtech Indonesia. 2020. “About.” Asosiasi Healthtech Indonesia, accessed November 16, 2020, https:// healthtech.id/about/ 4 Bhardwaj, Riaz, Noah Yarrow, and Massimiliano Cali. 2020. “EdTech in Indonesia: Ready for Take-off ?” Jakarta: World CHAPTER — 04 Bank. 5 Burhan, Fahmi Ahmad. 2020a. “Strategi Anyar Ruangguru hingga Zenius Gaet Pengguna saat Pandemi,” Jakarta, Indonesia: Katadata. Retrieved from: https://katadata.co.id/desysetyowati/digital/5f6de371e728a/strategi-anyar-ruangguru-hingga- zenius-gaet-pengguna-saat-pandemi (accessed November 16, 2020). 6 Crowley, Stuart. 2020. “Singapore Government looks to accelerate digitalisation and tackle COVID-19 with $3.5 billion ICT spend increase,” Singapore, Singapore: W Media. Retrieved from: https:/ /w.media/news/singapore-accelerate- digitalisation-covid-19-3-5-billion-ict-spend/ (accessed November 16, 2020). BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA 7 Dhawan, Shivangi. 2020. “Online Learning: A Panacea in the Time of COVID-19 Crisis.” Journal of Educational Technology Systems, 49(1):5-22. 8 Evandio, Akbar. 2020. “RUU PDP Wajib Rampung di November Ini.” Jakarta, Indonesia: Bisnis.com. Retrieved from: https://teknologi.bisnis.com/read/20200921/84/1294235/ruu-pdp-wajib-rampung-di-november-ini (accessed November 16, 2020). 9 ILO. 2020. “Social protection responses to the COVID-19 pandemic in developing countries: Strengthening resilience by building universal social protection.” Geneva: ILO. 10 Japan Times. 2020. “Japan to speed up work to launch digital agency”. Tokyo, Japan: Japan Times. Retrieved from: https:// www.japantimes.co.jp/news/2020/09/18/national/japan-speeds-digital-agency/ (accessed at November 16, 2020). 11 Kop, Rita. 2020. “The challenges to connectivist learning on open online networks: Learning experiences during a massive open online course.” The International Review of Research in Open and Distributed Learning 12:19–38. 12 Lapor. 2019. “Laporan Tahunan Pengelolaan SP4N-Lapor! Tahun 2019”. Yearly Report. 13 Mari, Angelica. 2020. “The Brazil Tech and Innovation Round-Up: Tech Sector Develops Immunity Passport, Government Digital Strategy Launched, Data Protection Regulations Postponed”, New Jersey, United States: Forbes. Retrieved from: https://www.forbes.com/sites/angelicamarideoliveira/2020/05/01/the-brazil-tech-and-innovation-round-up- tech-sector-develops-immunity-passport-government-digital-strategy-launched-data-protection-regulations-p- ostponed/?sh=52fc93b62881 (accessed November 16, 2020). 188 14 Mulia, Khamila, 2020. “Edtech platforms step up while Indonesia shuts downs schools in major cities”, Singapore, Singapore: KrAsia, Retrieved from: https:/ /kr-asia.com/edtech-platforms-step-up-while-indonesia-shuts-downs-schools-in-major- cities (accessed March 12, 2021). 15 Pettey, Christy. 2019. “Why Data and Analytics Are Key to Digital Transformation.” Retrieved from: https://www.gartner. com/smarterwithgartner/why-data-and-analytics-are-key-to-digital-transformation/ (accessed November 16, 2020). 16 Purnomo, Wayan. 2020. “Example of data fragmentation problems in Indonesia,” Jakarta, Indonesia: Tempo. Retrieved from: https:/ /majalah.tempo.co/read/nasional/160237/mengapa-data-korban-covid-19-pemerintah-pusat-dan-daerah- berbeda (Accessed November 16, 2020). CHAPTER — 04 17 Rutkowski, Michael, Alfonso Garcia Mora, Greta L. Bull, Boutheina Guermazi, Caren Grown. 2020. “Responding to crisis with digital payments for social protection: Short-term measures with long-term benefits.” World Bank Blog. Retrieved from: https:/ /blogs.worldbank.org/voices/responding-crisis-digital-payments-social-protection-short-term-measures- long-term-benefits (accessed November 16, 2020). 18 Sharon, Alita. 2020. “Malaysia Sets New Digital Learning Ambitions.” Open Gov Asia. Retrieved from: https://opengovasia. com/malaysia-sets-new-digital-learning-ambitions/ (accessed November 5, 2020). 19 Tracxn. 2019. “EdTech Startups in Indonesia”. Retrieved from: https://tracxn.com/explore/EdTech-Startups-in-Indonesia (accessed March 12, 2021) BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA 20 Vietnam News. 2020. “50 million citizen identification cards to be granted,” Retrieved from: https://vietnamnews.vn/ society/772149/50-million-citizen-identification-cards-to-be-granted.html (accessed November 16, 2020). 21 White, Olivia, Anu Madgavkar, James Manyika, Deepa Mahajan, Jacques Bughin, Mike McCarthy, and Owen Sperling. 2020. “Digital identification: A key to inclusive growth.” McKinsey Global Institute. Retrieved from: https://www.mckinsey. com/business-functions/mckinsey-digital/our-insights/digital-identification-a-key-to-inclusive-growth 22 World Bank. 2020. “Indonesia High Frequency Monitoring of Covid-19 Impacts Round 4”. Jakarta: World Bank. 23 World Bank. 2020. “The Promise of Education in Indonesia”. World Bank. Retrieved from: http://documents1.worldbank. org/curated/en/658151605203420126/pdf/The-Promise-of-Education-in-Indonesia.pdf (Accessed November 16, 2020). 24 Yarrow, Noah; Masood, Eema; Afkar, Rythia. 2020. “Estimated Impacts of COVID-19 on Learning and Earning in Indonesia: How to Turn the Tide”. Jakarta: World Bank. 189 Spotlight 4. EdTech and HealthTech Landscape CHAPTER — 04 Improving access to quality health-care and education EdTech and HealthTech firms in Indonesia are very di- services is essential in expanding opportunities for all verse, with most of them offering more than one product citizens. In addition to key reforms in the public edu- or service, so that they can offer more comprehensive cation and health-care systems, the use of information, support to their target groups. EdTech firms in Indonesia communications and technology (ICT) for education offer a wide variety of products and services, targeting BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA (EdTech) and for health (HealthTech) provision by the different users, including parents, educators, students, private sector holds considerable promise in improving educational institutions’ managements and corporations. educational and health outcomes. A focus on comprehensive support provision was clear- ly evident soon after it became clear that COVID-19- In line with growing internet penetration and the trend induced lockdowns would be in place for an extended toward increasing private investment in Indonesia’s period of time. School closures due to the COVID-19 digital economy,151 the number of new entrants in both crisis pushed students to adopt online learning tools and EdTech and HealthTech has grown significantly during pushed schools to adopt digital solutions to deliver edu- the past six years. According to an online analytics plat- cation services. To improve online learning experiences, form,152 there are about 181 EdTech startups in Indonesia. in addition to creating solutions for students, it was im- The majority of these firms were established in the period perative for EdTech firms to enable teachers and parents of 2013–19, which follows the trend of increased internet to understand, utilize and adopt digital solutions effec- penetration in Indonesia.153 The number of HealthTech tively. Therefore, EdTech companies such as Ruangguru firms are slightly higher. According to the Indonesian and Zenius went beyond offering solutions exclusively for HealthTech Association around 250 firms have been reg- students to extending support and providing training for istered with the organization since its inception in 2018. parents and teachers to be able to participate in the digital transformation of the education sector.158 Despite the fact Although HealthTech and EdTech have been around for that the digital-adoption learning curve has been harder several years, the COVID-19 pandemic has accelerated for teachers and parents than students, the pandemic has the adoption of HealthTech and EdTech products, while boosted digital literacy because adopting technology now also facilitating increased private sector investment in is a ‘need’ versus merely a luxury. these two sectors. In the pre-COVID-19 era, EdTech and HealthTech firms were not prioritized by the ma- For the most part, Indonesian EdTech products aim to jority of private investors in Indonesia, such as venture help students with learning and upskilling, educators capital firms, private equity firms, etc. However, more with student management, communication and teach- recently, especially following the forced surge in demand ing, and educational institutions with administration. For for HealthTech and EdTech products, both sectors have example, some of the core products that companies such started to attract higher investments in more mature, as Ruangguru, Zenius and Quipper develop and provide later-stage funding rounds. In EdTech, the most notable are e-learning content, interactive learning platforms and deal was Ruangguru’s reported Series C investment of study tools that help K-12 students expedite the learning US$150 million in December 2019,154 while in HealthTech process, along with interactive online services that help Halodoc completed a funding round close to US$100 students with their assignments and test preparation. million in July 2019 from strategic investors such as Pru- Companies such as Arsa Kids, Digikids and Educa Stu- dential, Allianz, UOB Venture Management, and the Bill dio offer game-based and blended learning experiences, and Melinda Gates Foundation.155 Other notable deals in- including interactive storybooks and educational mo- clude Zenius’ US$20 million investment in EdTech,156 and bile apps, to help improve early childhood educators’ Alodokter’s US$33 million investment in HealthTech.157 effectiveness. These products and services are typically 190 TABLE S4.1 Description of EdTech products with selected examples159 PRODUCTS / SERVICES DESCRIPTION SELECTED EXAMPLES Classroom tools Help teachers to implement interactive teaching methods, offering features Google Classroom, Microsoft Teams, such as live lectures, discussion forums, cloud-based student response Circledoo CHAPTER — 04 tools and other classroom communication tools. Learning Management Improve parent-teacher-student communications, monitor student performance Quintal, Sikad, Codemi, Scola System and progress, track assignments, share educational content, help teacher auto-generate reports and provide facilities for collaboration amongst teachers and students. Online Learning The most common and largest in terms of products in Indonesia, providing Kelaskita, IndonesiaX, HarukaEDU. (Broad, mainly Higher educational content across diverse subjects ranging from traditional HarukaEDU is one of the largest OPM Education/ Vocational subjects (math, science, IT, business) to more vocational subjects (such providers that supports universities Courses) as photography, entrepreneurship, music), through massively open online to convert their offline courses BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA content—MOOCs, and MOOC-based OPM (online program management) models. into online learning. Online Learning Develop and provide content, multiple products and learning material Ruangguru, Zenius, Quipper, Wardaya (specific to K-12) for K-12 students, including self-learning content, interactive learning College platforms and study tools that help students expedite the learning process and interactive online services that help students with their assignments. Many companies also place special focus on providing exam preparation solutions for students to prepare for standardized exams, such as national exams (UN and SNMPTN) and entrance exams for state universities. Digital books / convert textbooks and print books into digital books. PesonaEdu (oldest edutech in interactive content Indonesia) Education events Provide tickets and information about various events (such as seminars, Ngampooz marketplace trainings, competitions, etc.) around university and other educational institution campuses. Hardware Provide low-cost digital devices (such as laptops, notebooks, etc.) or Google Chromebook, Endless Computer enabling hardware technology solutions with built in educational tools. Language learning Facilitate language learning for non-native speakers either by providing Cakap by Squline, Bahaso, Kelas pre-developed content or by connecting the learners with native speakers. Bahasa Such platforms cater to learners across the age spectrum, whether they are in school or adults looking to develop new language skills. Early childhood Create educational games and educational toys for children, including Arsa Kids, Digikids, Educa Studio, education interactive storybooks and educational mobile apps. Rabbit Hole Technology learning offline and online learning solutions that cater specifically to teaching Hactiv8, Inspira. Academy, Koding programming and other engineering disciplines. Next, Bali Cipta Innovator Learning Application Provide digital learning solutions and educational content for students Lexipal (learning platform for for Special Needs with special needs. children with reading difficulties. Children Offline-to-online (O2O) Platforms for learners or students to find face-to-face or online tutoring, Maubelajarapa, Sukawu, Ngampooz classes or workshops. School Administration Simplify the administrative tasks of schools, such as digitizing InfraDigital Nusantara, edConnect, transcripts, school fee management and online payments, facilitating 7Pagi school-wide communication, online examination and assignments and admissions support and application tracking. Social/ Collaborative Social media-like features to improve peer-to-peer (P2P) communication and Kelase, UtakAtikOtak Learning Platform collaboration both during class and outside class. Career Planning and Career planning and counselling tools to help students find the relevant Rencanamu (formerly Youthmanual), Counselling universities and majors based on their personality. Ikigai Student Loans and Provide lending platforms to offer affordable loans for students to fund Danacita, DANAdidik, Pintek Finance their tertiary education and training. Technology Adoption Provide support educational institutions to adopt and integrate with Websis, PesonaEdu Consulting available digital solutions to become more productive. 191 TABLE S4.2 Description of HealthTech products with selected examples PRODUCTS / SERVICES DESCRIPTION SELECTED EXAMPLES Hospital Information Help clinic or hospital process and integrate more seamless service flow Medico, Periksa.id, trustmedis, System (Patient and delivery (often cloud based), including for document management, doktersiaga.com, klinikgo, data management/ record, claim management, and medical record management. sisfomedika administration) Healthcare marketplace Help the general public to discover and select services from the available Kuponsehat, Klikhospital, healthcare providers Planetmedika CHAPTER — 04 Tele-consultation The most common and largest in terms of products in Indonesia, offer Halodoc, Alodokter, Good Doctor, consultation by remote communication, generally for the purpose of SehatQ, medis sehat diagnosis or treatment of a patient at a site remote from the patient or primary physician. On-demand healthcare Help patients arrange house call / home visit from healthcare professionals Prosehat, perawat.id, Homecare24, (i.e., doctors, nurses, midwives, lab test officers). Medicall Health media and Provides general health news and information to the public to facilitate Pasienia, Klikdokter community the promotion of healthy lifestyle, and often peer exchange. BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Health education Provide online learning tools for the general public or aspiring healthcare Jago Preventif, Docquity professionals on health / medical subjects (content is more in-depth than popular articles). AI, IoT, bio-medics Implement ICT and bioinformatics support for healthcare stakeholders. Nalagenetics, Neurabots Clinical decision Health information technology (installed at healthcare worker or patient Sehati - Tele-CTG support facilities side) designed to provide health care professionals with clinical decision support. Tele-pharmacy and Delivery of pharmaceutical care via telecommunications to patients in Halodoc, Klik24 e-prescribing locations where they may not have direct contact with a pharmacist, and through the use of electronic prescriptions. Mobile health (wearable Smart electronic devices that can be worn on the body as implants or Fitbit, Samsung Gear, Apple Watch devices) accessories. Appointment scheduling Application that helps patients to manage appointment across clinics and RSPI Mobile health-care professionals. Pharmacy network Facilitates supply and outreach in the pharmaceutical sector among mClinica pharmacies, pharmacists and health-care professionals. Financial support Support patients or healthcare facilities with financial offerings, from Qoala (loan, insurance) claim/insurance management to credit/loan options disseminated using several approaches, such as web-based could prevent this type of services to scale (see Table S4. 2 for and mobile-based applications (see Table S4. 1 for complete complete products and services). products and services). Both EdTech and HealthTech sectors are still in their infancy, As for HealthTech, innovative products and services are with the majority of firms engaging in product/market exper- geared toward providing distant communications of health- imentation, while responding to changing market needs and care services (tele-medicine), while firms that offer hospital digital environment. Ninety percent of EdTech firms surveyed information management systems gradually shift to cloud- by the World Bank have changed their original business mod- based systems. For example, companies such as Halodoc and els after identifying new gaps in the sector and/or to achieve Alodokter, considered the most well-funded firms based on greater cost efficiency. In addition, most Indonesian EdTech the publicly available data, connect patients to health-care and HealthTech firms initially offer some features or content professionals (medical doctors) to receive consultations and for free (freemium pricing model), to attract more users, better may lead to drug prescription. Hospital information man- understand the market needs and experiment with differential agement systems have been traditionally offered by software pricing. developers/vendors that built tailored solutions for hospitals. Recently, new firms such as Medico or Periksa.id have started The COVID-19 pandemic has propelled many schools, especial- to offer cloud-based services (software-as-a-service), but strict ly those located in urban areas with better internet connections, practices on the management and transfer of medical records along with the support of MoEC’s free internet quota, to adopt 192 Placeholder for Photo or Pull quote CHAPTER — 04 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA global tools such as Google Classroom or Microsoft Teams. In tele-medicine service in 2015 to enable health-care services for April 2020, 31.6 percent of teachers reported using learning home visits to Makassar residents. management platform, including MoEC’s virtual class in Rumah Belajar platform, Quipper School, Quiziz, Google Classroom or The fast-tracked digital transformation in various sectors due Microsoft Teams, or schools own platforms.160 Wearable tech- to the COVID-19 pandemic has exposed a digital divide across nologies in health care (smart devices) have also been dominated various socioeconomic groups. Low-income students, as well by global brands such as Fitbit, Samsung, and Apple. as students outside the Jakarta Metropolitan and Bali region, lack the required infrastructure to fully utilize the digital prod- Understanding the importance and potential of EdTech and ucts offered by the various EdTech and HealthTech firms. To HealthTech, many government agencies have also created free address this digital divide, many EdTech firms, such as Zenius, products. For instance, Portal Rumah Belajar is an online learn- Quipper school, and Ruangguru, have partnered with govern- ing platform equipped with complete multimedia education ment and telecommunication providers to provide free services content managed by The Ministry of Education and Culture and free data packages so that students can easily access the (MoEC) since 2011. Universitas Terbuka, a state university programs and products on their platforms. focusing on distance learning, also offers extensive open educa- tion resources (OER). MoEC has a specialized unit, the Center To ensure efficient allocation of resources between the public for Data and Information Technology (Pusat Data dan Teknologi and private sectors, and to minimize the digital divide across Informasi, Pusdatin) that develops and manages ICT for edu- Indonesia and maximize the benefits of EdTech and HealthTech cation programs, such as the provision of ICT infrastructure products, both parties need to engage with each other more ef- and facilities, ICT-based learning content, development of fectively. Government, education providers, the telecom sector knowledge management systems to facilitate the sharing of and EdTech/HealthTech firms need to collaborate to ensure information and knowledge among learners and educators, that the digital divide is minimized161 and that digital products development of ICT-based learning resource centers in ele- are easily accessible. Private firms need to better understand the mentary and secondary education, and increasing human re- needs of teachers, schools and parents, while the public sector source capacity to support the efficient use of ICT at the central needs to become more effective at engaging with the private and local levels. The Government is also active in introducing sector, clarifying its governance structure and promoting pub- technology solutions for the health sector. The Ministry of lic-private partnerships (PPPs) for product development. Health in 2018 launched its own tele-consultation platform called Sehatpedia. BPJS-Health, which manages Indonesia’s As a final recommendation, the Government should continue single national health insurance scheme, launched its own to invest in improving digital infrastructure and connectivi- smartphone application, JKN Mobile, in 2017. This initially ty, particularly in underdeveloped areas and for underserved served as a member management tool to help all citizens (who communities. The inequalities in access to online learning are enrolled in a mandatory scheme) manage their member- and online health services mean that those without connec- ship, including for invoicing, payment, and claim verification. tivity have fewer opportunities to learn, or to receive health Another example is district-based health-care services, such as treatment. The situation has become more acute due to the the case of Makassar, which launched a 24-hour homecare and COVID-19 outbreak. 193 Endnotes 151 Google, Temasek, Bain & Company, 2019. 152 Tracxn (https://tracxn.com/) is an online analytics platform that tracks startups and private companies 153 Bhardwaj, Yarrow and Cali, 2020. 154 Ruangguru, 2019. 155 Halodoc, 2019. 156 TechInAsia, 2020. 157 Mobi Health News, 2019. CHAPTER — 04 158 Credit Suisse, 2021. 159 Bhardwaj, Yarrow and Cali, 2020. 160 MoEC, 2020. 161 Credit Suisse, 2021. BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA 194 References 1 Bhardwaj, Riaz; Yarrow, Noah; Cali, Massimiliano. 2020. “EdTech in Indonesia: Ready for Take-off ?”. World Bank, Washington, DC. 2 Credit Suisse. 2021. “Education Technology. Coronavirus and Beyond”. Available at: https://www.credit-suisse.com/media/ assets/microsite/docs/responsibleinvesting/cs-education-technology-spread.pdf (accessed March 12, 2021) 3 Google, Temasek, Bain & Company. 2019. “e-Conomy SEA 2019: Swipe up and to the Right: Southeast Asia’s $100 billion Internet Economy”. Available at: https:/ /www.blog.google/documents/47/SEA_Internet_Economy_Report_2019.pdf (accessed December 14, 2020). 4 Halodoc. 2019. “Halodoc Completes Strategic Funding Round as it Continues to Expand its Portfolio of Services”. CHAPTER — 04 Available at: https://www.halodoc.com/media/press-release-halodoc-completes-strategic-funding-as-it-continues-to- expand (accessed December 14, 2020). 5 Koh, Dean. 2019. “Indonesia’s healthcare superapp Alodokter raises $33M in Series C funding”. Mobi Health News. Available at: https:/ /www.mobihealthnews.com/news/asia-pacific/indonesia-s-healthcare-superapp-alodokter-raises- 33m-series-c-funding (accessed December 14, 2020). 6 MoEC. 2020. “Analisis Survei Cepat: Pembelajaran dari Rumah dalam Masa Pencegahan COVID-19”. Available at: https:// puslitjakdikbud.kemdikbud.go.id/assets_front/images/produk/1-gtk/buku/1__Hasil_Survey_Cepat_BDR-Kepala_ Sekolah_dan_Guru_(1).pdf BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA 7 Ruangguru. 2019. “Ruangguru Completes $150M Series C Transaction Led by General Atlantic and GGV Capital”. Ruangguru Press Release. Available at: https://blog.ruangguru.com/ruangguru-completes-150m-series-c-transaction (accessed December 14, 2020). 8 Tracxn. 2019. “EdTech Startups in Indonesia”. Available at: https://tracxn.com/explore/EdTech-Startups-in-Indonesia (accessed March 12, 2021) 10 Yu, Doris. 2020. “Indonesia’s Zenius bags $20m, appoints former Gojek exec as CEO”. TechInAsia. Available at: https:// www.techinasia.com/zenius-bags-20m-appoints-ceo (accessed December 14, 2020). 195 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA CHAPTER — 05 196 Chapter ⬎ Harnessing Digital →What Can Indonesia Do? Technologies for Inclusion 5 T CHAPTER — 05 his report has documented how the COVID-19 pandemic turbo-charged the adoption of digital technologies and e-commerce among business- es and consumers. School closures forced stu- dents and teachers to adapt and explore digital- ly enabled remote learning options. HealthTech apps enabling remote consultations and the de- livery of medicine saw unprecedented growth. BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA At the same time, this report has argued that the potential for digital technologies to bring about inclusion in Indonesia has remained largely unrealized. Despite rapid progress achieved in the years prior to the pandemic in connecting its citizens to the internet, the country faces a gaping digital divide, with just under half of adults still not connected. The level of adoption and geographic reach of applications such as those used for e-commerce and digital ride-hailing have not reached levels needed for catalyzing a fundamental shift in the underlying structure of the economy or its inclusiveness to bring about a transformative impact on the population. The ongoing pandemic demonstrates the power- ful role that digital technologies can play in mitigating impacts of crises and building resilience, reducing disruption of critical services, and delivering quick and well-targeted social and health system responses to impacted pop- ulations. But with low adoption and low digital readiness at the start of the pandemic, even the rapid growth documented in this report was not sufficient to allow the GoI to take full advantage of digital technologies in its pandemic response, evident in delays in deployment of response measures. To accelerate the diffusion of digital technologies and their use for productive purposes in an inclusive Indonesian economy, this report identifies constraints that need to be lifted in three broad areas. First, despite improvements in connectivity over the past decade, a sizeable digital divide persists because of lingering challenges in middle and last mile connectivity. The divide con- strains the adoption and use of digital technologies for e-commerce, as well as for the provision of efficient, quality government services to Indonesian citizens. Second, harnessing the potential of digital technologies to generate economic opportunities for all and unlocking citizen capabilities to seize these opportunities will require greater financial inclusion, nurturing of trust to promote uptake of digital payment services, and investing in skills needed for the 21st century digital economy. Third, the potential of digital technologies to improve the efficiency of government administration and the quality of services to all citizens cannot be realized until the fragmentation of regula- tions, policies and their implementation across government institutions is resolved through stronger leadership and a clear vision for comprehensive whole-of-government digital transformation. This chapter describes the key regulatory and program reforms needed to lift the constraints in each of these areas. 197 5.1. Improve Digital Connectivity and Universalize Access CHAPTER — 05 Despite improvements in connectivity over the past de- the 1 GHz band for cost-effective coverage in rural areas cade, a sizeable digital divide persists in Indonesia as a and above the 1 GHz band for large spectrum assignments result of lingering challenges on middle and last mile con- for cost-efficient capacity. The main bands that are not nectivity. Fixed broadband penetration is insufficient to yet available for mobile broadband (4G and, in the future, meet rapidly growing demand for high-bandwidth appli- 5G) in Indonesia are the 700 MHz and 2.6 and 3.5 GHz BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA cations. Mobile broadband is more widely available but bands. Release of the 700 MHz band requires the digital does not deliver the same capacity or quality of service, TV migration to be completed. This band will enable nor is it cost-efficient for high-bandwidth applications. cost-efficient rollout of (4G/LTE and beyond) mobile The divide constrains the adoption and use of digital tech- broadband services in rural areas. Under the recently en- nologies for e-commerce and its enablers such as e-logis- acted Omnibus Law on Job Creation No. 11/2020, reform tics services that require last mile connectivity for services of spectrum management is on the right track for optimi- such as tracking and tracing. Low internet penetration zation of spectrum allocation. Omnibus Law No. 11/2020 in schools, hospitals and health posts threatens to widen also imposes a mandatory two-year plan to convert analog human capital disparities as large segments of the popu- television to digital television to secure the digital divi- lation get left behind, with learning and health-care deliv- dend in the 700 MHz spectrum band. Release of the 2.6 ery migrating online for some but not all. Addressing the GHz band, currently used for satellite TV, would add digital divide will require universalizing access to reliable capacity in urban centers and alleviate network conges- and good quality internet as a first step. tion. Kominfo should consider accelerating its plans for the reallocation of this band from satellite broadcasting to As identified in Chapter 2, there are three key regulatory mobile broadband before the end of the current spectrum bottlenecks that are holding Indonesia back: (i) spectrum license in 2024. Securing the 3.5 GHz band spectrum will management, specifically unavailability of spectrum in require consultation with current C-Band satellite users. specific IMT spectrum bands; (ii) the lack of regulatory Over time, some form of sharing should be considered mechanisms to facilitate active and passive infrastructure with the use of this band for mobile services in urban cen- sharing;and (iii) limited competitiveness among service ters and for satellite services in those rural areas that still providers. Thus, optimizing spectrum allocation for mo- require C-band connectivity. This would add further ca- bile broadband and strengthening mechanisms to ensure pacity and enable the introduction of 5G. Finally, the GoI the sharing of active and passive infrastructure among should prepare to make available the mmWave spectrum market participants, and streamlining licensing proce- in the 24-29 GHz band to enable immediate deployment dures to support greater market contestability, are among as soon as the industry is ready for 5G mmWave services. the most important policy measures that can be taken to universalizing internet access and use in Indonesia. Consistent with ITU recommendations, Indonesia should endorse an overall target for available IMT spectrum of at → Optimize spectrum least 840 MHz plus allocations of mmWave spectrum as soon as possible,162 and certainly no later than 2024. Such allocation for mobile allocations should be made in larger contiguous blocks in broadband accordance with future best practice. MNOs should also have the flexibility to use their allocated IMT spectrum The freeing-up of additional spectrum would reduce for mobile broadband and/or fixed wireless access (FWA) network congestion in urban areas and facilitate more services. High-speed FWA services using 4G/5G technol- widespread rollout in rural areas. Mobile broadband is ogies are proving globally to be a very competitive product highly dependent upon sufficient spectrum, both below with fixed broadband services.163 198 → Strengthen mechanisms to including deploying the digital infrastructure needed for high quality 4G services and future 5G services. ensure the sharing of active and passive infrastructure Tower sharing for mobile broadband networks was man- dated in 2009 and has been deployed on a large scale, but As noted in Chapter 2, between 70 and 80 percent of the the sharing of other passive infrastructure, such as ducts, cost of fixed broadband is typically accounted for by passive poles, etc. (required for fiber-optic networks) between infrastructure such as ducts, poles, rights of way and civil telecom operators lags behind, despite a joint letter is- works, and the limited provisions for active and passive sued by the Ministry of ICT and the Ministry of Home infrastructure sharing make the costs of investment in fixed Affairs in 2018. The Omnibus Law mandates passive in- broadband prohibitive. Infrastructure sharing would bring frastructure sharing through a change to Telecom Law CHAPTER — 05 benefits in terms of reduced capital expenditure and op- Article 34A and B, but implementation of the associated erating expenditure, increased competition between pro- provisions will require inter-agency coordination among viders, lower prices and, as a result, increased take-up and national agencies and local governments. Separately, ac- connectivity. Sharing is also necessary for the extension of tive infrastructure sharing would promote competition optical fiber networks down to the kecamatan level, and to in mobile broadband services in rural/remote areas but is improve access among households and businesses (“fiber currently not allowed under Telecom Law PP No. 52/53. to the home” or FTTH). FTTH is currently dominated This law requires in many cases separate deployment of by PT Telkom (IndiHome) across Indonesia, and some (mobile) infrastructure by each and every telecom op- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA smaller providers such as Biznet, MyRepublic, MNC and erator. The Omnibus Law does include a change to Law Oxygen (Moratelindo) in some cities. The recent entry Article 34B that opens up the possibility of active infra- of the electricity service provider, PLN, into the FTTH structure sharing, although this could have been made market could potentially alter the low competition status more explicit. Telecom Law PP No. 52/53 will have to be quo, as PLN has passive infrastructure to each house with updated to allow for the sharing of active infrastructure an electrical grid connection.164 on a B2B basis. Omnibus Law No. 11/2020 mandates passive infrastruc- → Strengthen competition ture sharing and the next step for the GoI is to create an implementing regulation. This will require effective lead- along the broadband value ership to coordinate the government agencies involved, chain including, among others, the Ministry of Communication and Information, the Ministry of Home Affairs, the Min- To improve the competitiveness of service providers, In- istry of Public Works, and local government agencies. The donesia should consider transitioning toward the unified GoI will need to modify infrastructure planning approval licensing of service providers to enable each to deliver a procedures to include obligations for the sharing of pas- larger portfolio of services. As noted in Chapter 2, the sive infrastructure (e.g., ducts, free space for an additional current regulatory regime restricts the ability of operators network operator, poles, rights of way, and civil works) for to provide a full range of services by requiring them to deployment of fiber-optic networks, include in the licens- apply for and maintain specific service licenses, instead ing process clear guidelines for network sharing, create of a single uniform license for all services. This limits the a single entity responsible for setting tariffs for passive issuance of telephony service licenses and the portability infrastructure sharing, and establish a rapid and effective of telephone numbers, etc. To address these issues, it is dispute resolution mechanism. A planning database con- recommended that the GoI reviews the current licens- taining detailed information on infrastructure available ing regime and considers transitioning toward a single/ for sharing at the national and kota/kabupaten levels also unified license to deliver the full portfolio of services to a needs to be created. Local governments would contribute larger number of service providers, so that effective com- by deploying passive infrastructure for shared use, for ex- petition for dual-play and triple-play fixed broadband ample, ducts along roads and streets during routine civil services will emerge. This should be complemented by works. Operators should be required to publish relevant regulations enabling portability of telephone numbers information for infrastructure sharing, while those bene- across providers. Neither the Telecommunication Law fiting from the Universal Service Obligation (USO) fund nor its amendments in the Omnibus Law addresses these should be required to share subsidized infrastructure. issues. Telecommunications-related regulatory commit- Changes arising from the Omnibus Law now allow up to ments in the Regional Comprehensive Economic Partner- 100 percent foreign investment in the tower and trans- ship (RCEP) and the ASEAN Digital Masterplan (ADM mission business, which was previously on the negative 2025),165 which was launched in January 2021 at the 1st investment list (DNI). As a result, there will be a range ASEAN Digital Ministers’ Meeting,166 should also guide of new industry stakeholders able to invest in the sector, further regulatory reforms. 199 5.2. Make the Digital Economy Work for All Indonesians who are connected to the internet tend to Deepening of reforms to address entry barriers to logistics CHAPTER — 05 use it intensively, but largely for communication, social and transportation services would lower costs, decrease media and leisure, as opposed to economically productive transit times and improve reliability of e-logistics ser- purposes (Chapter 1). Digital technology adoption in the vices. Entry barriers to logistics and transportation ser- workplace by household enterprises and among low-edu- vices in Indonesia include the fragmentation of licensing cated individual workers is still very limited, while returns requirements, limits to foreign equity investment, min- to internet use at work are higher for the better educat- imum capital requirements, barriers to development of ed, leaving the less- and least-educated workers further inter-island shipping (port tariffs, roll-on roll-off regula- behind than they already were. While e-commerce has tions, cabotage), barriers to development of inter-modal BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA grown and micro and small merchants could, in principle, logistics and the cost of red tape (licensing, regulated benefit from its growth, including by opening jobs and agents in airports, etc.). The GoI has issued several reg- economic opportunities for otherwise-vulnerable women ulations to reduce these barriers, but additional reform and youth workers, its reach is constrained by uneven is needed and should be covered in the implementing internet access, high logistics costs, and poor digital skills. regulations being designed for the recently passed Om- Finally, taxation of the digital economy would support in- nibus Law.169 Removal of barriers to reverse logistics, in clusive growth by providing revenues for public spending, particular the restrictions on re-export, would help to reducing some of the distortive impact of taxation on the provide certainty over unsold or deficient stocks, especial- economy, and encouraging formalization of businesses. ly for cross-border e-logistics. Attracting private sector investment in warehousing outside of Indonesia’s large metropolitan areas would be more efficient than the status → Support development of quo where traders set up warehouses and distribution centers for their own use, paying a high cost to ensure logistics compliance and meet licensing requirements. Building investor confidence in turn requires a minimum level of Development of physical infrastructure to improve con- demand for handling products, reliable electrical supply nectivity in and outside urban areas needs to be continued. and skilled labor. Fragmentation in licensing discourages A strong transport and logistics infrastructure network investors from moving outward beyond the Jakarta-Sura- (roads and electricity) in and outside of urban areas is a key baya-Medan-Bali network. enabler of e-logistics. UNCTAD’s Rapid e-Trade Readiness Assessments of Least Developed Countries recommends Modernization of PT Pos Indonesia, the national postal the rehabilitation of physical infrastructure as central to service, should be continued. The national postal service support e-trade. Areas in China that are well connected has unused capacity available linked to a sharp decline in by road and other key infrastructure (electrical, telecom- mail traffic. In addition to adapting the sorting facilities munications) have benefited most from the e-commerce to efficiently handle packages, there is potential for postal ecosystem and have developed the most.167 Public-private facilities to be transformed into collection-and-delivery partnerships (PPPs) with local and international investors points (CDPs) to support the e-commerce supply chain. have been shown to speed up infrastructure development In Estonia, for example, the National Post Omniva has and improve efficiency. These are urgently needed in Indo- already well running locker points across the country to nesia, where growth in infrastructure developed has lagged enable any user to drop or collect a parcel using SMS mes- behind growth in GDP.168 The e-logistics ecosystem is saging. Elsewhere, PPPs have been developed between also held back by low efficiency of airports and ports, and the national postal services and private sector e-logistics limitations in the services they provide. Measures such as companies, such as the Cainao network in China (Alib- the development of temperature-controlled facilities and aba’s logistics arm). Finally, standardized addresses and improvements in the risk management capacity of airport postcodes would enable the automation of parcel sorting personnel would help to improve turnaround times and and route optimization in delivery, reducing handling support a smooth e-logistics chain. costs per parcel, the damage rate, and delivery times. 200 E-logistics can be leveraged to unlock the economic They also limit the development of EdTech and Health potential of micro, small and medium enterprises (MS- Tech initiatives and the utilization of services that do exist MEs), as well as agricultural smallholders. The MSME in these sectors, and they stand in the way of efforts to sector continues to underperform relative to its poten- digitize government for improved delivery of services tial and suffers from consistently low productivity. The across a whole swath of sectors. general recommendations described above would bene- fit MSMEs by making logistics service providers (LSPs) Interventions to nurture digital skills need to be contex- more competitive and PT Pos Indonesia more efficient. tualized in an environment of rapid changes in the nature Reduced transit times and logistics costs support MS- of jobs, even in Indonesia. Worldwide, automation—and MEs’ cashflows and increase their profitability, including the adoption of technology more generally—is making through an expanded customer base. More targeted sup- some jobs obsolete at the same time as innovation is cre- port could include training on logistics and supply chain ating new ones (World Bank 2019). E-commerce in Indo- CHAPTER — 05 management, such as inventory and order management. nesia has a long way to go before it can catch up with other Warung entrepreneurs, for example, lack these skills and economies, but change is already here: executives in Indo- rely on same-day sales to estimate input orders for the nesia’s largest companies list data analysts and scientists, next day, leading to out-of-stock products (missed sales) big data specialists, AI and machine-learning specialists, or obsolete inventory (wasted cash). The GoI could: (i) digital marketing and strategy specialists, among leading partner with the private sector to support entrepreneur positions in growing demand, as well as job roles such as access to e-logistics solutions offered by initiatives such process automation specialists and Internet-of-Things as Warung Pintar and Mitra Bukalapak170 in large cit- specialists (World Economic Forum 2020). The emer- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA ies like Jakarta and Surabaya, including data analytics to gence of these roles reflects the acceleration of automa- support small entrepreneurs in decision-making when tion, as well as the resurgence of cybersecurity risks.173,174 placing orders171 and same day/next day deliveries; (ii) The COVID-19 pandemic may accelerate structural shifts facilitate partnerships between marketplaces, the national in the economy that exacerbate the digital skills deficit, if post, and private sector logistics companies to create a it leads to permanent changes in the nature of work due sustainable logistics ecosystem such as Cainiao in China’s to accelerated adoption of technologies, greater reliance Taobao rural program that supports distribution between on home-based work and online meetings, or changes in cities and rural areas;172 (iii) support basic training via global value chains resulting in rebalancing of economic vocational training programs at the local level and offer sectors. incentives to encourage entrepreneurs to get trained; and (iv) explore supporting e-logistics services that aim to Institutional action and policies need to recognize that improve village-to-city connectivity, such as the farm-to- digital skills are a subset of a broader skillset needed for table startups creating marketplaces connecting farmers the 21st century digital economy. The exponential pace to buyers (SayurBox, TaniHub) and financing raising of technological change today makes it hard to anticipate funds to support smallholders (CROWDE). which job-specific technical, digital and other skills will thrive and which will become obsolete in the near future. As a result, the ability to adapt quickly to changes is in- → Nurture digital skills and creasingly valued by the labor market.175 The sought-after trait globally is adaptability—the ability to respond to skills for the 21st century digital unexpected circumstances, and to unlearn and relearn economy quickly. This trait requires a combination of certain cog- nitive skills (critical thinking, problem-solving) and so- This report’s findings that core digital skills in the adult cio-behavioral skills (curiosity, creativity). Indeed, the population are low, limiting the ability of Indonesians top five skills and skill groups that executives in Indone- from participating in the digital economy safely, se- sia’s largest companies see as rising in prominence in the curely and productively, align with a broad literature run-up to 2025 are creativity, originality and initiative, that concludes that skill levels in the workforce have not complex problem-solving, active learning and learning kept up with employer demand, that there is a mismatch strategies, emotional intelligence and analytical thinking between skills needed and those supplied, and that the and innovation (World Economic Forum 2020). workforce lacks the core skills needed for 21st century jobs. Shortages of workers qualified in engineering and ICT professions, and those with general computer skills, → The working-age population are particularly acute. Poor digital skills constrain many of the enablers of the digital economy examined in this Digital skills training needs to be embedded in non-for- report, and they also constrain the expansion of e-logis- mal education services and lifelong learning tailored to the tics services outside Indonesia’s metropolitan areas and needs of the working-age adult population, the majority the adoption of digital financial payments, in particular. of which is either informally employed or in self-employ- 201 BOX 5.1 A fundamental transformation of lifelong learning services is needed to meet the digital skill needs of the working-age adult population What sets adult learning apart from traditional tertia- solutions for problems of everyday life. This in- ry or TVET education is that, for effective implemen- volves abilities such as solving technical problems tation, these programs need to rely on more systemat- from troubleshooting to solving more complex ic diagnoses of the specific constraints that adults face, problems, identifying needs and technological pedagogies that are customized to the adult brain, and responses through critical evaluation of possi- flexible delivery models that fit with adult lifestyles. ble solutions, and using technology creatively CHAPTER — 05 Becoming a digitally competent individual is a moving for multimedia production and self-expression. target, requiring cognitive flexibility and openness to- Learning-by-doing approaches are recommended ward change. Rather than focusing solely on technical to promote these skills. Digital media is especially knowledge or specific IT tools, which risk becoming useful for engaging adult learners in the creative rapidly obsolete, adult training programs for digital process of multimedia production, through tasks skills should encourage learners to approach “the ma- such as making multimedia resumes. chine” by trial and error and promote: BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Additional design features that improve effective- • Abductive processes, which involve making in- ness of adult learning programs in general include ferences and drawing insights from information diagnosis and evaluation to understand the needs found on the web, leading to formation of ex- of the target population, i.e., those employed in- planatory theories and generation of new ideas. formally or self-employed, or unemployed; insights Browsing the web is not a linear process, and ser- from neuroscience and behavioral economics (use of endipity—the instance of something interesting practical exercises and visual aids to help memory, or pleasant happening by chance—often charac- motivational tools such as financial rewards, work terizes how resources are accessed or information experience, or frequent feedback); flexibility in pro- discovered online. This requires a positive atti- gramming to allow learning at own convenience; ex- tude toward the unknown, enjoyment of random plicit links to employment opportunities, through discoveries, and the ability to carry out abductive apprenticeships or internships; and mechanisms inferences to make sense of these discoveries. to address multiple constraints at the same time • Development of problem-solving skills associated (combining training with cash or capital, improve with use of digital technologies. Problem-solving job searches through peer support, text message re- skills are needed not just to solve technological minders, and action planning, etc). issues, but also to conceptualize technological Source: Extracted from World Bank 2019; and Ranieri 2020. ment. Central to this is Indonesia’s TVET system, which the future. This could be addressed by: (i) adult learning includes both a formal and a non-formal path. The lat- programs that focus on adult literacy, supplemented with ter targets out-of-school initial jobseekers and workers basic digital literacy; (ii) skills training for wage employ- without formal vocational education. Both paths suffer ment, delivered in collaboration or partnership with the from deep fragmentation, with management and regula- private sector; and (iii) entrepreneurship programs that tion spread across 20 government ministries, depending embed digital skills. In each case, program effectiveness on the level and type of education and training provider, will depend on the extent to which lessons are integrated and limited coordination and incomplete feedback mech- into everyday life (Box 5.1). anisms lowering effectiveness and efficiency.176,177 The low quality of both TVET and tertiary education178 has failed Digital skills programs need to cultivate a mindset of to prepare the workforce with those skills demanded by continuous learning and offer opportunities to practice employers, and needs to undergo intensive reform if it “self-managed” learning. Given the pace of technological is to train the workforce for 21st century jobs. A broad change, becoming a digitally competent individual is a literature describes the needed reforms; the focus in this moving target, requiring workers to adapt to changes over section is on measures for nurturing digital skills, as well time, and to unlearn and relearn quickly. One example as complementary skills needed for the digital economy of of a program that anticipates this is the Ikanos project, 202 promoted in 2012 by the Basque Country Government careers relevant to AT&T’s future workforce and digi- to build a learning support platform for the digital com- tal strategies, including web and mobile development, petence needs of citizens, employed and/or unemployed data science and machine learning.182 To date, AT&T people (Kluzer et al. 2020). In addition to a self-assess- has spent over US$200 million annually to design the ment tool for career and training guidance, as well as curriculum and has already achieved over 4,200 career to increase the employability of unemployed people, a pivots with 70 percent of jobs filled internally by those Personal Learning Environment (PLE) was developed who were reskilled. Similarly, Shell launched an online to help “learners take control of and manage their own education effort titled the Shell.ai Development Program, learning”. This included providing support to set own which focuses on teaching AI skills to its employees. Both learning goals, manage own learning, and communicate programs have created customized versions of Udacity’s with others in the process of learning, and managing both Nanodegree programs to reskill and upskill employees the content and process to achieve learning goals. Import- with hard-to-source, in-demand skill sets. Another ex- CHAPTER — 05 ant concepts in PLEs include integration of formal and ample is Coursera for Government. In order to respond informal learning episodes into a single experience, use to the surge in unemployment during the COVID-19 of social networks that can cross institutional boundar- pandemic, governments in over 100 countries provided ies, and use of networking protocols (peer-to-peer, web access to the platform to citizens looking to gain new skills services, syndication) to connect a range of resources and and credentials to re-enter the workforce. The programs systems within a personally-managed space. connected graduates directly with local companies who agreed to accept those credentials as the basis for hiring. In addition to focusing on the priority knowledge and Since April, the program has reached 650,000 unem- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA skill gaps identified in this report, the content of digital ployed workers who enrolled in over 2.5 million courses skills programs will need to recognize Indonesia’s em- that provide the skills needed for fast-growing jobs in IT, ployment-specific needs. Priority areas include: (i) basic health care and business.183 digital hygiene on issues related to data privacy, such as identification of phishing, social engineering, and access Many countries have developed innovative funding to personal information on devices; (ii) digital financial mechanisms to finance the reskilling of workers, with ad- services and their use to promote consumer trust and ditional provisions to mitigate the adverse impacts of the confidence in online financial tools; (iii) entrepreneurship COVID-19 pandemic. Globally, reskilling and upskilling training that includes topics such as digital marketing policies that have been utilized to date span the condi- strategies, taxation and regulation of the digital economy, tionality of unemployment benefits on taking up new and operations in integrated e-commerce ecosystems that re-skilling and up-skilling, providing wage subsidies to include payment, logistics and insurance services; and (iv) companies that extend reskilling and upskilling to work- e-logistics, in particular knowledge, standards and skills ers, providing online learning accounts to citizens, and needed for warehouse operations. Second, the content for starting to fund online learning in addition to university digital skills programs will need to be based on a frame- degrees, TVET and school tuition. Singapore recently work that recognizes employment-specific differences in complemented its pioneering “Skills Future Initiative” skills relevant for Indonesia. An example is the EU Digital through the deployment of Enhanced Training Support Competence Framework.179 Package to support workers and organizations in sus- taining investment in reskilling and upskilling during As formal sector employment grows over time, improved the COVID-19 pandemic. The package includes a signif- incentives for employers to provide on-the-job training icant increase in funding for Absentee Payroll Support will become critical for nurturing digital skills among and Course Fee Support among industries severely hit workers. A mere 5 percent of Indonesians in the labor by the pandemic. At the end of 2019, France created force receive training beyond the formal school system.180 an individual skills account with an integrated mobile The only financial incentive offered by the GoI to employ- application dedicated to vocational training and life- ers to provide skills training is the recently adopted “super long learning. Under the “moncompteformation.gouv. tax deduction” that allows firms to deduct from taxes fr” (“MySkillsAccount”) scheme, 28 million eligible twice the amount spent on training, but only large firms full- and part-time workers will receive €500 annually can benefit from this provision.181 The share of workers directly into their skills account to spend on upskilling in formal sector employment will grow over time, albeit and continuous learning, with low-skilled workers and slowly, and the GoI should plan to partner with the pri- those with special needs receiving up to €800 annually, vate sector in re- and upskilling workers, especially for capped at a total of €5,000 and €8,000, respectively. nurturing of digital skills. The experience of several com- The Danish Ministry of Employment has introduced panies demonstrates the potential of the private sector measures aimed at providing additional opportunities in this regard. Telecommunication company AT&T has for upskilling and job-focused education for workers worked with Udacity to create 50 training programs, the furloughed during the pandemic. First, both skilled and “T University”, to prepare individuals for the technical unskilled workers who pursue a vocational education 203 are being provided with 110 percent of their usual un- corporating more general education in tertiary programs employment benefits. In addition, the Danish Govern- is one way to do this. An additional year of general edu- ment expanded the scope of its current apprenticeship cation was added in 2012 to undergraduate programs in scheme, at the same time as prolonging its job rotation Hong Kong SAR, China, focusing on problem-solving, scheme, making it possible for more unskilled workers critical thinking, communication, leadership, and life- to have access to upskilling and reskilling opportunities. long learning skills. Another way is through innovative pedagogy. The Faculty of Architecture and Environmen- tal Design at the College of Science and Technology, the → Future entrants to the University of Rwanda, has promoted learning strategies that include open-ended assessment, feedback opportu- workforce nities, and a progressive curriculum that balances aca- demic challenge with student support. These approaches CHAPTER — 05 Education technology (EdTech) initiatives at all edu- have improved the critical-thinking skills of students. cation levels need to be mobilized via partnerships with Forward-looking universities are finding ways for adult the private sector, complementing public sector online students to acquire a broad set of socio-behavioral skills. learning resources. Embedding ICT into the learning Dutch vocational colleges are providing entrepreneurial experience in schools and universities serves not only to courses aimed at improving noncognitive skills, such as enhance teaching, learning and assessment, and optimize teamwork and self-confidence. Tunisia introduced an learning processes, but also to expose students to digital entrepreneurship track that combines business training technologies from an early age, easing the transition to with personal coaching to reshape the behavioral skills BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA picking up occupation-specific digital skills later on in of university students. In China, a combination of coop- life. In addition to improving digital infrastructure and erative learning and role play enhanced self-educational connectivity in schools, particularly in underserved com- abilities and communication skills among undergraduate munities, this will require: (i) reforming of curricula to students in pharmacology classes. nurture the aforementioned 21st century skills needed for Industry 4.0, including digital skills, taught early on Close collaboration between industry and tertiary edu- and reinforced throughout the lifelong learning process; cation is critical. The low quality of TVET and tertiary and (ii) training existing teachers and future entrants education in Indonesia has been linked to, among other to the profession on use of ICT to enhance teaching and factors, a lack of competency frameworks developed in preparing them for effective delivery of the new curricular consultation with the private sector. Inadequate labor content.184 A clear vision is needed for deployment of market information and intermediation makes it hard to curriculum-aligned EdTech with a focus on lower-income align curricula and teaching with occupations and skills and rural, remote areas based on evaluations that iden- needed. Private sector participation in tertiary education tify cost-effective approaches for use at scale. Additional planning and policy is thus a requisite both at a strategic measures needed include: (i) development of standards and technical, curricular level. In China, for example, Le- for data privacy and security related to EdTech products; novo is working with tertiary institutes to train vocational (ii) partnerships between EdTech firms, academia and students in high-tech areas, such as cloud computing, government to establish clear standards for performance that feature practice-based curricula, practitioner-led and cost-effectiveness, and to transparently and rigorous- instruction, and professional certification (World Bank ly evaluate the leading products; (iii) support to EdTech 2019). First, the GoI also needs to incentivize employ- startups through use of startup assistance organizations ers to offer internships and off-campus learning to stu- and promotion of PPPs for product development; and dents. Second, filling in information gaps during the job (iv) partnerships with EdTech firms to improve teachers’ search process enables students to make choices between ability to deliver technology-focused content.185 and within different paths. Chile is establishing online platforms where students can access information on the Modern tertiary education needs to cultivate in students a employability of individuals with various degrees, wage minimum threshold of foundational “transferable” high- profiles, and courses to take for certain occupations. Co- er-order skills for the digital 21st century economy, even in lombia’s Jóvenes en Acción (Youth in Action) program STEM fields. Technology and integration have increased combines classroom instruction with on-the-job training demand for higher-order general cognitive skills—com- at private companies. Third, Indonesia needs to continue plex problem-solving, critical thinking, and advanced to develop data systems that allow for identification of communication—that are transferable across jobs. As occupations and skills in demand, monitor educational a consequence, the combination of general and techni- institutions’ compliance with quality standards, and en- cal skills is becoming highly valued (World Bank 2019). sure that information on the employability of degrees, Tertiary education systems should therefore guarantee a wage profiles, and occupation-specific courses is available minimum threshold of transferable cognitive skills, which to both jobseekers and workers. are also the best inoculation against job uncertainty. In- 204 not benefiting from the G2P digitization due to profit → Promote the use of digital and liquidity challenges faced by agents when serving PKH beneficiaries. The GoI is recommended to revise financial services and payment the G2P model to drive sustainable provider revenue and solutions, including among the incentivize expansion. The digitization of private sector unbanked and the underbanked wage needs also has significant potential to foster DFS adoption. In Indonesia, 30 million banked adults receive private wages in cash, 72 percent of whom have a mobile. Incentive schemes can be introduced for the private sec- Moreover, another 15 million unbanked adults receive tor to innovate and develop new digital financial services private wages in cash. (DFS) products that cater for wider group populations, especially the rural population. The underserved and un- Stronger legal and regulatory frameworks for managing CHAPTER — 05 banked population is mostly located in rural areas. The risks related to data privacy, cyber security and finan- regulators could provide a new enabling environment cial integrity are needed to build consumer trust in DFS designed only for rural areas. While DFS can be provided products. Lack of consumer trust stems from perceptions via channels and instruments from various types of insti- of risks, including risks related to data governance and tutions, including bank and non-bank payment service privacy, cyber security and operational risk, and the in- providers, agent banking services are particularly promis- tegrity of financial transactions. Concerns revolve around ing for reaching Indonesia’s largely unbanked population. unauthorized data disclosure by the various entities in- Allowing non-bank e-money provider to recruit indi- volved, and are reflected in the growing number of com- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA vidual agents has the potential to create new rural access plaints registered by consumers in Indonesia, including points with a full suite of e-money services, including in the banking sector. While Indonesia’s Financial Ser- (cash-in-cash-out) CICO access points. The regulator vices Authority (Otoritas Jasa Keuangan, OJK) has been can also permit third-party agent network managers for trying to strengthen the legal and regulatory framework efficient agent services across segments. This will then for DFS, consumers still regard the disclosure of personal ease the operational challenge faced by DFS providers, ID to service providers as a major risk. In this regard, while at the same time improve the customer services of the passing of the draft Personal Data Protection Policy the agents. Another change in regulation that could bring Law will be critical for promoting the adoption of DFS more income for agents is by allowing non-exclusivity in in Indonesia. ‘frontier’ rural areas. A non-exclusive model has brought 40 percent higher income for agents in Bangladesh and Kenya. Subsidies and other tax inducements have also been introduced by several countries to encourage both → Use tax policy instruments to businesses and consumers to adopt DFS. For example, ensure level playing field for all Uruguay provided the subsidies for digital payments accepted by small businesses, while reducing tax with- A well-designed system of taxation for the digital econ- holding requirements for firms and lowering value-added omy can help level the tax playing field between conven- taxes for consumers. The Republic of Korea also allows tional and online businesses; within online businesses, wage earners to claim tax deductions for purchases made between goods and services; and between resident and using digital payments. Early research suggests the re- non-resident businesses. This will reduce the distortive forms have helped increase the number of digital payment impact taxation may have on the economy, helping to transactions. ensure that sales, profits, and investment decisions in the digital sector are driven by market dynamics and efficien- The GoI can also play a role in sustaining the commercial cy improvements, and not by advantages gained through viability of current DFS agents by maintaining transac- uneven taxation policy or taxpayer avoidance. In the In- tion volumes at scale. Increase use cases or transaction donesian context, two incidental benefits are equally rel- volume is essential so that prices are more aligned with evant. First, taxation of the digital economy will bring in cost to serve. Improving the interoperability of the pay- a small but growing amount of revenue as digitalization ment system can result in higher use cases or transaction accelerates within the Indonesian economy,186 especially volume. The GoI could increase demand for DFS through in the context of COVID-19 pandemic. E-commerce in various government programs. This has been started with Indonesia, for example, is projected to have grown by 54 the digitization of the conditional cash transfer program percent during 2020, reaching US$32 billion, at a time of Program Keluarga Harapan (PKH). PKH alone has when private consumption in the overall economy has brought more than 10 million people to the formal fi- suffered, falling by an estimated 2.7 percent overall.187 nancial services. However, the use cases of DFS among Second, taxation of the digital economy presents Indo- the beneficiaries are still low, with most of the beneficia- nesia with a unique opportunity to boost formalization ries still cashing out all the money. The agent network is of businesses, particularly that of MSMEs. Registered 205 businesses will enjoy the benefits of having easier access to lection mechanisms for the new tax.195 The United States has credit from the financial system, and of potential future fiscal reacted strongly against these unilateral measures. In January support from the Government. 2021, the US Trade Representative (USTR) announced that investigations into digital services taxes adopted by Austria, But the digital economy also creates special challenges for tax- India, Italy, Spain, Turkey, and the United Kingdom were sub- ation. For direct taxation (CIT), the main challenge relates to ject to “possible trade actions” under Section 301, “because the mismatch between the intangibility of value-drivers within they discriminated against U.S. digital companies, were incon- the digital economy and traditional taxation rules’ reliance on sistent with principles of international taxation, and burdened tangible features of a business to determine taxing rights and U.S. companies.”196 Indonesia, Brazil, the Czech Republic, and profit allocation (World Bank 2021).188 For indirect taxation the European Union were dropped from this investigation, (VAT), three main issues have attracted policy attention: (i) the USTR noted, only because they “have not adopted or not the intangibility of transactions that relate to the consumption implemented the DSTs under consideration when the investi- CHAPTER — 05 of online media and other digital goods and services; (ii) the gations were initiated.”197 In response, the Indonesian Ministry digital economy enabling a very sharp increase (in 2019, of a of Finance has clarified that its priority is indirect taxation reported 254 percent yoy189) in imported e-commerce parcels, and, while maintaining its right to impose direct taxes, it has the majority of which are low-value shipments that fall below delayed implementing this policy in the hope that a global the traditional minimum threshold;190 and (iii) the business solution can be found.198 landscape of the digital economy featuring a large amount of businesses below the VAT threshold, making the current Indonesia is taking measures to reform its tax policy and threshold increasingly less suitable.191 modernize its tax administration to deal with the challenge BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA of intangible transactions for VAT. Indonesia’s broad policy A long-term solution to the challenge of direct taxation on the position was reflected in Law No. 2/2020, which provided the digital economy requires a global consensus, but this is thus legal umbrella for imposing VAT on digital goods and services far lacking. Negotiations on reforming international taxation provided by foreign suppliers. Minister of Finance Regulation rules are taking place under the leadership of the Inclusive No. 48/ PMK.03/2020 (hereafter PMK-48) and DGT Regu- Framework (IF)192 on Base Erosion and Profit Shifting (BEPS) lation PER-12/PJ/2020 (hereafter PER-12) provided the next and could increase global corporate income tax revenues by up levels of policy and implementation details. They applied the to 4 percent, equaling US$100 billion annually.193 However, existing VAT statutory rate of 10 percent on all intangible these negotiations are proving to be politically contentious taxable goods and all taxable services that are provided by and agreement has been delayed. In the absence of a consen- foreign suppliers through an electronic system. Both foreign sus, countries including Indonesia, are considering interim and domestic digital platforms are required to collect the VAT measures. These rely on a simple measure of gross revenue on behalf of foreign suppliers, subject to being appointed as to define the tax base for multi-national enterprises (MNEs). “VAT Collectors” by the DGT. To date, companies that have For example, they could include the gross revenue arising from been asked by the DGT to serve in this capacity include Am- the sale of advertising or data to Indonesian businesses, or the azon, Facebook, Google, LinkedIn, McAfee, Netflix, Skype, amounts collected from Indonesian users for provision of a ser- Spotify, Twitter, and Zoom.199 The qualifying criteria for being vice, regardless of the physical presence of the MNE providing appointed a VAT Collector are based on: (i) annual sales vol- such goods and services (WDR 2021).194 While continuing umes of at least IDR 600 million for digital goods and services to participate in international consensus-building on direct in Indonesia (or IDR 50 million in a month); or (ii) online taxation of the digital economy, the next step is for the GoI to traffic volumes involving at least 12,000 users within Indo - complete the design of its proposed interim measures. nesia in 12 months, or 1,000 users in one month. To support compliance with the reforms, the DGT has introduced several As part of its COVID-19 fiscal response, the GoI passed Law administrative measures.200 No. 2/2020 claimed taxing rights on non-resident firms with a “significant economic presence” in Indonesia and this is set As discussed in Chapter 3, to deal with the challenge of low-val- to rely on a simple measure of gross revenue as the tax base. ue shipments Indonesia has lowered its minimum threshold The interim measure is described only in broad terms, with from US$75 to US$3. In so doing, Indonesia has become an the Ministry of Finance given powers to define its parameters. early mover in this reform space, following on from Australia, Issues still to be worked out include: (i) the scope of revenues, which was the first to slash its GST threshold from AU$1,000 namely whether to limit the scope to B2B transactions (as to zero from July 2018.201 The EU has followed suit, with its is the case in the interim measures proposed by Austria and previous exemption of consignments of less than €22 abolished Hungary, and those implemented by India), or to encompass from January 2021 onwards.202 To allow VAT to be levied, all B2C transactions as well (as Chile proposes); (ii) incorporating imports into the EU now must be declared using an electronic a minimum provision (often linked to the sales revenue of the customs declaration. To ease the implementation burden, the worldwide group, and/or sales in the domestic market); (iii) EU has also introduced a simplified customs declaration form rate of levy (e.g., 2 percent for the United Kingdom, 3 percent for all declarations of goods up to €150.203 In Indonesia, imple- for Italy, and 5 percent for Austria); and (iv) defining the col- menting the new rules without overwhelming the customs ad- 206 CHAPTER — 05 Placeholder for Photo or Pull quote BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA ministration or creating unnecessary burdens on business will to e-commerce in Indonesia, which is characterized by a large be more challenging, requiring effective risk management.204 share of small companies (Al-Rikabi et al. 2021).207 Experience from Australia and the EU may provide the GoI with some useful lessons. Overall, design and implementation of digital economy taxa- tion reform must be geared around the core principles of equi- The amount of revenue the Government is successful in col- ty, efficiency, and simplicity, so that Indonesia’s economy and lecting will depend on its ability to effectively manage com- society can altogether share in the benefits of digitalization. pliance with the new requirements. Two measures can help in Over time, as digitalization expands, the demarcation between this regard. First, the GoI will need to ensure adoption and the digital economy and the economy at large will become communication of consistent rules, leverage an efficient IT increasingly blurred and eventually dissolve. This evolving system to administer registration, filing, and payment, and reality makes it ever more important that the GoI gets taxation combine digital transaction data with other third-party data reform right. This means ensuring that tax policy and admin- and taxpayer data to enhance compliance risk management. istration meet the sound principles of good taxation. Create an Second, Indonesia’s overly generous VAT threshold needs to uneven playing field of taxation, for example, by having differ- be reviewed to expand the digital economy tax base. Indonesia’s ent tax rates on online and offline businesses, or by enforcing VAT threshold of IDR 4.8 billion stands out in international rules on cross-border businesses but not on domestic ones, and comparisons of VAT thresholds to GDP per capita, a metric Indonesia will end up with a tax system that detracts from the commonly used to compare the generosity of VAT thresholds organic, healthy growth of the digital economy. Revenue lost to worldwide. Less-developed economies tend to have higher special incentives that favor the select few will mean less financ- thresholds, in part because of weaknesses in tax administration ing for the critical public investments need to enable inclusive and in part because they have larger informal economies.205 growth of digitalization. Ultimately, taxation must not distort However, even when compared against low-income economies, business decisions on how to operate, and it should not alter Indonesia’s VAT ratio stands out as being too generous.206 A consumer choice on what and where to buy—whether it be high threshold means that a vast share of businesses whose from a supermarket or a hypermarket, an online marketplace annual turnover is below the VAT threshold are excluded from or via an online social media app. Taxes imposed must be fair the VAT system, narrowing the base and distorting the tax. and equal, and administered with a minimum burden on all. This distortion is even greater when the threshold is applied 207 5.3. Use Digital Technologies to Provide Better Services and Upgrade Citizen-State Interactions CHAPTER — 05 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA The COVID-19 pandemic has laid bare the low level of recognized, yet there is substantial demand and significant “digital readiness” on the part of the GoI to deliver ser- opportunities (see Chapter 4). However, Dukcapil is devel- vices online and in a paperless manner. Governments oping prototypes and pilots for a mobile ID equivalent of around the world are adapting rapidly to use digital tech- the national ID card. A key question is the optimal model nologies to cope with the crisis and ensure business conti- for the Indonesian context (i.e., centralized or federat- nuity, curb and manage the health contagion, and target ed), building on the national ID system (Box 5.3). The economic and social relief programs to vulnerable popu- status quo constrains growth of the digital economy and lation sub-groups. The GoI, in contrast, has struggled to introduces additional costs and risks for service providers mitigate disruptions of services due to low and unequal interacting with customers over the internet, as evidenced utilization of digital technologies across sectors, despite by increased fraud. It will therefore be critical that such a an increasing variety of online services provided in both national digital ID initiative is a whole-of-country effort the public and private sectors, particularly for education that links with other relevant initiatives, such as the BI and health. In addition to limitations in connectivity, the Payment Systems Roadmap and Satu Data. Thailand offers absence of a national digital ID system and a comprehen- a useful example: its National Digital ID initiative was born sive digital government system are the key factors under- out of the National Digital Economy Committee chaired mining Indonesia’s response to the COVID-19 pandemic. by the Prime Minister. Similar digital ID initiatives have been launched out of agencies attached to the head of gov- ernment, such as SingPass by the Singapore Government → Develop a national digital ID Technology Agency (GovTech) and the Australian Trusted Digital Identity Framework by the Digital Transformation framework Agency (DTA). An initiative in Indonesia would ideally be spearheaded by the President’s Office (KSP) with central Indonesia needs to launch a national digital ID initiative, roles for the Ministry of Home Affairs as the authority bringing together various government stakeholders, the responsible for population data and the Ministry of Com- private sector and civil society, among others, to create munications and Information as the authority responsible a framework that would allow Indonesians to securely for regulating electronic transactions. prove their identity online. Although there is a relatively strong national ID system managed by Dukcapil, Indonesia Improvements to the national ID system will create a currently lacks an official digital ID system or framework strong base on which to introduce a digital ID system or to bring coherence to a myriad of systems to be officially framework. It offers population data that would allow an 208 BOX 5.2 Designing Indonesia’s Digital ID system: Centralized or federated? A centralized system involves government developing ploring a shift to a federated approach that would build and implementing the digital ID system, commonly on the existing SingPass. A key benefit of federations is built as a feature of an existing national ID system or that: (i) they can leverage a dynamic private sector and a layer on top. A famous example of this is Estonia’s multiple public sector actors; (ii) they provide people system, which leverages the national ID smartcard with choice; and (iii) they should promote innovation and making widely available card readers that can plug through competition between digital ID providers. into personal computers, as well as complementing However, they can be more complex, as the trust frame- CHAPTER — 05 that with mobile phone versions. Singapore’s Sing- work has to account for multiple actors and require Pass is similar. Centralized systems may be simpler to institutional capacity to effectively regulate and super- launch, especially when there is an existing national ID vise the participants. Indonesia would not necessarily system, but can be quite rigid, and require substantial have to recreate the wheel and leverage documentation capital and sustained operating investment by govern- from some of the examples mentioned above. Dukcapil ment. A smartphone application or digital credential would play an essential role in any federation, not only could be introduced as a natural centralized extension as the trusted source of population data but also poten- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA of the NIK and KTP-el. Integrating digital credentials tially as a digital ID provider (e.g., through a mobile ID on the KTP-el is another possibility. However, con- equivalent to the national ID card). sidering the requirement for specialized readers and the costs of upgrading all KTP-els in circulation, this Either approach has merits in the local context. A likely could be burdensome and would unlikely have centralized system could be straightforwardly built on a high return on investment. top of, or as part of, the existing national ID system, but the rigidity might limit use cases and adoption. Federated models are where multiple digital ID provid- A federated system could also be relatively easily be ers that can come from the public and/or private sec- developed to leverage the national ID system, while tors are governed by a trust framework and supervised also harnessing Indonesia’s dynamic private sector, by government. The trust framework includes laws, but would need substantial work to develop the ca- rules and standards for how the digital ID providers pacity needed to effectively regulate and supervise can access trusted data when people register to create the federation. A federated approach would provide a digital ID, and how service providers interface with more flexibility and opportunities for innovation. Im- the digital ID providers to verify the identity of their portantly, a government-issued digital ID would not clients. There may be a central hub that links the var- be excluded in a federated model and could even be ious components: people, digital ID providers, trust- the exclusive credential used for government trans- ed data sources, and relying parties. Federations have actions. The optimal model for Indonesia requires been adopted by the United Kingdom, France (France further study by a cross-governmental national digital Connect), Australia (Trusted Digital Identity Frame- ID initiative, taking into account the unique circum- work), Canada (Pan-Canadian Trust Framework) and stances of Indonesia, global lessons and examples, and Thailand (National Digital ID). Singapore is also ex- the national and cross-border use cases to be pursued. official digital ID system (managed by the Government) hanced in terms of policies, processes and scalability, in- or officially-recognized digital ID system (managed by cluding to introduce national-scale biometric verification. the private sector but regulated by the Government) Dukcapil has piloted one-to-one authentication, such as with high levels of assurance. This onboarding process through biometric verification or one-time passwords simplifies and significantly reduces the costs of introduc- (OTP), and an e-KYC service that shares selected data ing digital IDs in Indonesia. For the national ID system after a successful authentication, and these need to be to realize the full potential to support service delivery scaled up nationwide. Finally, the GoI will need to decide and underpin a digital ID system in Indonesia, two gaps on whether to build a centralized digital ID system im- will need to be addressed. First, coverage of the national plemented as a single system by the GoI, or a federated ID system, including continuous registration of births, one implemented by multiple digital ID providers from deaths and marriages, will need to be universal. Second, the public and private sectors and regulated by the Gov- the utilization of the national ID system needs to be en- ernment (Box 5.2). 209 FIGURE 5.1 Transition from e-Government to digital transformation How re-aligning ...to embracing From focusing on the vision address National Digital eGovernment only.. identified problems Transformation Providing tech Providing tech CHAPTER — 05 platform and platform and governance to improve governance to improve government initiated government, public and Attaining high services private services level political Objectives leadership which is currently absent due to the limited Tech-platform for e-Government scope Cross-government multi-stakeholders agency platform use (e.g.: payment integration (e.g.: gateway, digital ID, BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA performance-based public-private data Typical budgeting) exchange, etc) Wider digital initiatives transformation not only will unlock true potentials Ministry responsible of tech, but also overseeing government strengthen the business processes Centre of Government and organizations (Office of hte current eGovernment Ideal (currently MenpanRB) President) agenda Coordinator Passage of a draft Personal Data Protection Law is a criti- cal next step. One area that will need particular attention → Reorient from a narrow is that Indonesia does not yet have a comprehensive data focus on e-Government to a protection law with general applicability. This under- comprehensive national digital mines trust in any kind of collection, processing and shar- ing of personal data, which are core functions of digital transformation agenda ID systems. However, a draft Personal Data Protection The GoI could consider committing to a transition from Law was submitted by the President to parliament at the the current narrow focus on e-Government to a compre- end of 2019, and this draft law was largely modelled on hensive National Digital Transformation agenda. The the EU’s General Data Protection Regulation (GDPR). current e-Government initiative that is chaired by Men- An essential feature for the credibility and strength of Pan-RB focuses on the computerization of government such a law would be independent oversight, similar to internal processes. Looking at the scope in the regulation, the Office of the Australian Information Commissioner, the role fits the main duties and responsibilities of Men- Philippines National Privacy Commission, and the Sin- Pan-RB, which includes the management of government gaporean Personal Data Protection Commission. Such a business processes. However, if the GoI aims to achieve law will ensure public trust and confidence in the existing comprehensive digital transformation of the nation, it national ID system, as well as any future digital ID sys- has to gradually depart from e-Government and slowly tem, and the digital economy more broadly by providing rearrange its vision, governance, and delivery model to safeguards and accountability for the collection, use and better suit that aspiration as depicted in Figure 5.1. Such sharing of personal data, as well as formalizing the rights a commitment would reorient Indonesia in the same di- of data subjects. rection as that taken by digitally advanced countries, such as Singapore, the United Kingdom, Australia, Estonia, the Rep. of Korea and Canada.208 The digital government platforms would serve as the foundation of digital services in the economy, so that both public sector institutions and private sector enterprises can take advantage of these plat- 210 FIGURE 5.2 A comprehensive national digital transformation agenda would require adjustments to the governance structure of ongoing digital government and policy initiatives GOVERNANCE STRUCTURE TO ALIGN WITH TRANSFORMATION AGENDA Additional Notes on National Digital Transformation AGENDA FOCUS eGOVERNMENT NATIONAL DIGITAL TRANSFORMATION Governance • DTT to be led by Digital Transformation the Office of The CHAPTER — 05 eGovernment taskforce (DTT) President to provide Policy, Coordination strategic directions strategy Team (Office of The on national digital and President) directives PMO transformation agenda • DTT will consolidate Digital the work of various AGENCIES AGENCIES INSTITUTIONAL Transformation government digital SETUP Implementation related initiatives BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Agency (DTIA) aside from Implementation SNGs SNGs eGovernment, such as Satu Data, OSS, One Map policy, etc. • Strengthen eGovernment • Issue new PR on Digital Transformation Agenda • Implementation ENABLERS implementation by equipping replacing PR 95/2018 agency to be formend the eGovernment Coordination • Establishing DIgital Transformation Taskforce as an Independent team with a PMO function (DTT), led by the Office of the President Agency, with the • Establish Digital Transformation • Prepare and implement the option of being a Implementation Agency (DTIA) transition strategy and plan BLU (Public Service • Consolidate government digital transformation initiatives under DTT Agency) Existing functions(s) New functions(s) to establish forms to deliver services to every citizen. Examples of a ated regulation mandates government agencies to deliver common data exchange platform in Estonia,209 digital sin- geospatial data from over 80 sectors across 34 provinces gle sign on ID in Singapore,210 and digital payment in the within four years. Given the heavy coordination work United Kingdom211 show how pivotal digital government required, CMEA decided to establish a PMO and staff it platforms can be to allow nations to operate effectively in with top talent. Foreseeing that the required talent may digital space and to stimulate innovations in the market. not exist yet in the civil service, CMEA drew on the ser- The reform would need to be spearheaded by a central gov- vices of an external consulting firm to manage the PMO. ernment authority, for example, the Office of the President, CMEA successfully achieved the milestones in the One to provide the political and bureaucratic influence needed Map policy action plan in a timely manner. for such far-reaching reforms. In parallel, creation of regulatory and institutional At the same time, a short-term priority is to strength- frameworks necessary for the comprehensive digital en the role of MenPan-RB in orchestrating and deliv- transformation agenda can commence. To start with, a ering the existing e-Government mandate. This could presidential regulation could be prepared to mandate: be achieved by setting up a Project Management Office (i) the development of a national digital transformation (PMO) under the institution that could help address the strategy; (ii) the establishment of a policy-making entity; disconnects and administrative burdens that currently and (iii) the establishment of an implementation entity. exist in governance of the e-Government initiative. The The policy-making entity, a Digital Transformation Task disconnects stem from ineffective communication be- Force (DTT), would ideally be attached to the Office of tween the various institutions involved and disconnected the President, overseeing and, if possible, consolidating escalation/delegation processes between technical and and streamlining the governance structure of existing leadership actors. The practice of setting up a PMO is digital initiatives such as the e-Government agenda led not unprecedented. Strong PMOs have played import- by MenPan-RB, One Data led by Bappenas, Online Single ant roles in several national strategic programs, such as Submission, and the One Map policy led by CMEA (Fig- the delivery of the One Map policy by the Coordinating ure 5.2). This would reduce the fragmentation that char- Ministry for Economic Affairs (CMEA).212 The associ- acterizes the governance structure of these initiativies, 211 and that leads to ineffective and unnecessarily complex Payment regulation, and Omnibus Law No. 11/2020.214 implementation at a high cost of coordination. Finally, One of the recently issued One Data implementing reg- the implementation arm of the DDT would be a Digital ulations, Permen Bappenas No. 16/2020, has addressed Transformation Implementation Agency (DTIA), sim- the relations between One Data and e-Government. The ilar to agencies in Singapore (GovTech) and the United regulation outlines key elements of data management in CHAPTER — 05 Kingdom (GDS). DTIA would be responsible for deliv- view of e-Government, such as: (i) data architecture; (ii) ery of cross-sectoral digital services, i.e., platforms and master data and reference data; (iii) databases; and (iv) supporting technologies for sectoral ministries to host data quality. If well implemented, this can significantly and operate their digital services. It would also serve as improve government data governance in Indonesia. The the ecosystem builder for delivery of sector-specific dig- master data and reference data management principles ital services, coaching and assisting sectoral ministries to in the regulation are critical to reducing duplications, develop quality, standardized digital services for citizens. improving quality, and enabling access of key datasets, es- Both DTT and DTIA need not be newly established en- pecially the key datasets that are traditionally maintained BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA tities. Existing entities within the public sector could be by government institutions that administer important empowered to play the role. Irrespective of the model service.215 However, it is must be noted that the access that is chosen, the most crucial part of the reform would to these key datasets must observe the other prevailing be to ensure that the institutional arrangement facili- regulations such as privacy and the data protection law. tates a whole-of-government approach and eliminates Therefore, Bappenas, in its role as the leading ministry fragmentation. for the One Data policy, needs to strategize how to assert its leadership to effectively implement these subsidiary regulations by rallying support for the new government → Transition toward a data management policy. whole-of-government In a politically challenging environment, Bappenas needs data management policy to not only strategize the implementation of One Data to undergird digital subsidiary regulations, but also to rally support for the One Data policy. Specifically, Bappenas can: (i) Apply transformation the One Data policy comprehensively in priority sectors relevant for managing the COVID-19 pandemic and re- The One Data Presidential Regulation strengthens vitalizing the economy (education, health, MSME, and somewhat Indonesia’s data management policy, but social protection). Bappenas could do this by appointing fragmentation remains an issue and several implemen- a data steward in each sector, assisting stewards in devel- tation challenges need to be worked out. Until recently, oping a master data reference for each sector and data Indonesia did not have a clear whole-of-government data standards and in enforcing the standards, and assisting management policy. Each sector was responsible for its policy makers in these sectors to draw insights from im- own data management, and guidance on cross-sectoral proved data management, which can then be used for data-sharing and utilization was missing. Presidential expansion to other sectors; (ii) Build alliances with key Regulation No. 39/2019 on One Data issued in 2019 institutions such as MenPan-RB, Kominfo and Dukcapil and its subsidiary implementing regulations try to ad- to slowly instill One Data principles into implementa- dress these fragmentation issues by granting Bappenas tion of the digital government and digital ID initiatives; greater authority to regulate, monitor, and enforce data and (iii) Work closely with MoF during the execution of governance across government agencies. However, the shared planning and budgetary roles to identify and/or associated implementing regulations not only need to filter out programs and activities that lead to duplication specify the implementation sequence but also address the or inefficient production of data. Bappenas and MoF can intersectionality of the regulation with other initiatives promote evidence-based and accountable policy-making such as e-Government/Digital Government (Presidential by requiring program/activity proposals from line min- Regulation No. 95/2018), the Electronic Transaction Law istries to meet certain standards of cost-benefit analysis and its implementing regulation (PP No. 82/2012 and that use reliable data. This could encourage line ministries its revision PP No. 71/2019), the forthcoming Personal to actively use data for policy-making and invest in their Data Protection Law,213 the Civil Registry Law, the Digital data systems. 212 Placeholder for Photo or Pull quote 213 BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA CHAPTER — 05 Endnotes 162 Refer to Recommendation 6 on Releasing More IMT Spectrum 179 The typology in the EU Digital Competence Framework, for ex- for Wireless Broadband and 5G Deployment as contained in the ample, recognizes four categories: (i) more or less broadly defined ITU White Paper, Digital Infrastructure Policy and Regulation existing occupations such as administrative worker in the public in the Asia-Pacific Region, September 2019. administration, primary school and early childhood teacher etc.; (ii) generic business functions such as operations and industrial 163 Global Mobile Suppliers Association, 2021. services, marketing and sales, etc.; (iii) generic work conditions 164 PLN has 67 million homes passed with the power utility (mostly such as entrepreneur, virtual office worker, employment services poles) while the largest FTTH network, Telkom, reaches only staff; and (iv) new IT-intensive jobs in different economic sectors about 30 million homes passed. (Industry 4.0 jobs in manufacturing, new digital jobs in muse- ums, etc.), and distinct from IT specialist job profiles. CHAPTER — 05 165 ASEAN and Plum Consulting, 2021. 166 ASEAN, 2021. 180 Underlying factors include limited capacity among small firms to allocate resources for training, and talent poaching that per- 167 Alibaba Group and World Bank. 2019. petuates high turnover rates, especially among high-skilled and 168 Draft. The World Bank Group. 2019. Indonesia’s Infrastructure young workers (World Bank, 2020) Sector Assessment Program (InfraSAP). 181 Regulation No.128, Ministry of Finance, 2019. 169 With regard to transportation sectors, the law includes broad 182 World Economic Forum, 2020. provisions to reduce or remove foreign equity restrictions, sim- plify licensing processes by transitioning from specific licenses to 183 In one example, the Government of Costa Rica has worked with BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA general business licenses, and transferring licensing responsibili- local employers across the country to identify current job open- ties from line ministries to the central government. ings and skill demand and tailored the program offering to that local demand. Similar structures of collaboration have been es- 170 Bukalapak’s analysis of its own data shows that same day and tablished across local government in the United States, specifi- next day delivery within cities have helped warung performance, cally across a network of job centers. leading to a substantial reduction of the rate of late deliveries and complaints (Bukalapak presentation November 2019). Similar- 184 World Bank, 2020. ly, Paxel’s analysis of its own data shows that same day and next 185 Bhardwaj et al, 2020. day inter-city delivery is helping housewives turned female mi- 186 The indirect revenue gains are potentially much greater, as data cro-entrepreneurs producing food products expand their mar- from the digital economy can be combined with other third-par- ket to other big cities in Java. ty data and used by the revenue authority, the Directorate Gen- 171 Warung Pintar offers member warung owners with data analytics eral of Taxes (DGT), to boost compliance and raise higher reve- to support decision-making when placing orders, such as an inte- nues across Indonesia’s main taxes, i.e., value-added tax (VAT), grated supply chain solution to manage their stock. corporate income tax (CIT) and personal income tax (PIT). 172 The arrangement drove down logistics costs and enabled next-day 187 World Bank, 2021a. delivery to many secondary and tertiary cities. The model does 188 For example, a guiding principle for income taxation has tradi- carry the risk of monopolistic behavior stemming from a single tionally been the principle of ‘permanent establishment’, which marketplace and logistics ecosystem, and thus calls for strict reg- looks at tangible features of a business including the location ulatory oversight by the government to protect MSME interests. of staff and assets to determine taxing rights and profit alloca- 173 At the opposite end of the scale, the roles that are set to be in- tion. However,, digital companies can generate large sales in a creasingly redundant by 2025 include roles being displaced by country with little or no physical presence. The intangibility of new technologies: accounting and bookkeeping and payroll value-drivers of digital businesses has facilitated aggressive tax clerks, data entry clerks, and assembly and factory workers, and avoidance, resulting in a much lower tax burden on digital busi- administrative and executive secretaries. nesses compared to traditional businesses, and large revenue 174 Elsewhere in low- and middle-income countries, many people losses for governments. For a more comprehensive discussion, are employed in jobs that did not exist three decades ago (World see “Chapter 7. Creating value in the data economy: the role of Bank 2019). India has nearly 4 million app developers; Uganda competition, trade, and tax policy.” World Bank, 2021b. has over 400,000 internationally certified organic farmers; and 189 “Creating Fair Tax Treatment and Protecting Domestic Small China has 100,000 data labelers. and Medium Industries, Government Changes Provisions on 175 World Bank 2019. Import of Shipments (E-Commerce)”, Ministry of Finance Press Release (December 23, 2019), accessed online at: https://www.ke- 176 See Table 1 of World Bank 2020. menkeu.go.id/publikasi/siaran-pers/siaran-pers-ciptakan-per- 177 Ibid. lakuan-perpajakan-yang-adil-dan-lindungi-industri-ke- cil-dan-menengah-dalam-negeri-pemerintah-ubah-ketentu- 178 Poor performance is linked to lack of qualified faculty members an-impor-barang-kiriman-e-commerce/ and infrastructure, low capacity and reach of accreditation sys- tems, limited number of certification providers, and lack of com- 190 According to data from the Ministry of Finance, low-value ship- petency frameworks developed in consultation with the private ments below the previous USD 100 de minimis threshold have sector. Inadequate labor market information and intermediation increased significantly in recent years, accounting for 84.4 per- makes it hard to align curricula and teaching with occupations and cent of total shipments in 2018 (as measured by looking at the skills needed. Workers find jobs through informal networks rather number of consignment notes issued). According to Indonesia’s than through formal messaging boards, job announcements, or job tax regulations, shipments that fall below the minimum thresh- matching services. Most TVET providers have no mechanisms to old are exempt from import duties and import VAT. Thus, the assess the labor market performance of their graduates. expansion of the digital economy has resulted in a significant re- duction in tax potential for taxes collected at the border. 214 191 Data from Indonesia’s economic census and several surveys show 202 Council of the European Union. 2017. “VAT on electronic com- that the vast majority of businesses within Indonesia have sales merce: new rules adopted”, Press Release, Council of the EU (De- below the current VAT threshold of IDR 4.8 billion. Analysis cember 5, 2017). of Indonesia’s e-commerce landscape shows a similar pattern 203 Taxation and Customs Union. “New form of customs declara- (Al-Rikabi 2020). tion for low value consignments”, The European Commission, 192 The IF has 130 country members (including all OECD members) accessed online at: https://ec.europa.eu/taxation_customs/news/ and is housed at the OECD. International organizations such as new-form-customs-declaration-low-value-consignments_en the World Bank Group have an observer role. 204 The large number of ‘informal’ MSMEs, not registered with gov- 193 See the OECD’s preliminary estimate on revenue impacts here: ernment entities, and/or entrepreneurial households increasingly CHAPTER — 05 https://www.oecd.org/tax/beps/webcast-economic-analysis-im- active on e-commerce represent significant challenges in the Indo- pact-assessment-february-2020.htm nesia context. Data from digital platforms can be a very powerful enabler of enhanced risk management when combined with cus- 194 For example, India introduced a 6 percent charge on digital ser- toms data. Indonesia’s Directorate General of Customs is looking vices linked to online advertising in 2016, which it considered to do this, with an ongoing pilot involving several digital platforms an “equalization levy” charged as a proxy for corporate income focused on using e-commerce transactions data to tackle compli- tax on foreign suppliers that did not have a permanent establish- ance risks including under-invoicing and missed declarations. ment in the country. The list of countries that have announced that they plan to introduce interim measures include: Austria; 205 Revenue authorities with limited capacity to handle large num- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA France; Hungary; Italy; India; Indonesia; Kenya; Malaysia; Mex- bers of monthly VAT filings (e.g., because they remain reliant ico; Pakistan; Singapore; Turkey; Uruguay; Vietnam; the United on manual filing, and/or have limited ICT capacities to manage Kingdom; and Zimbabwe. See WDR 2021, Ibid. VAT e-filing nationwide) may set a higher VAT threshold so that fewer companies would need to register and file. An informal 195 “Indonesia: Direct Taxation of the Digital Economy”, Internal economy may be defined as one with characterized by firms that Note, The World Bank (April 2020) are not registered, and/or whose employees work without a for- 196 The USTR announcement on the next steps into its Section 301 mal contract. Informal companies have been traditionally hard to Digital Services Taxes Investigation can be found here: https://ustr. tax, since they are often ‘hidden’ from government reach. gov/about-us/policy-offices/press-office/press-releases/2021/ 206 “Box B.1: Why is Indonesia’s tax-to-GDP ratio so low?” in The march/ustr-announces-next-steps-section-301-digital-ser- Indonesia Economic Quarterly: Towards Inclusive Growth” vices-taxes-investigations. In addition to the US formal action, (March 2018), pp. 50. an opinion-editorial was published by the former US Ambassador to Indonesia which expressed similar sentiments, recalling the US 207 Al-Rikabi, Khalil Rohman, and Hambali, “On the Macroeco- import retaliatory action against “France’s discriminatory digital nomics of the Digital Economy in Indonesia”, Working Paper, services tax which was targeted at US tech firms”, and advising World Bank (forthcoming, 2021). that “Indonesia must tread carefully” or else “would likely invite 208 The group of leading digital nationals: https://leadingdigital- similarly retaliatory action”. See The Jakarta Post, 2020. To read govs.org/ the original trade action from the Office of the United States Trade Representative, see: “Docket No. USTR-2020-0022. Initiation of 209 Estonia’s Data Exchange Platform or widely known as X-Road: Section 301 Investigations of Digital Services Taxes”. /e-estonia.com/solutions/interoperability-services/x-road/ https:/ 197 See USTR announcement, Ibid. 210 Singapore ID Service or Singpass: https://www.singpass.gov.sg/ spauth/login/loginpage?URL=%2F&TAM_OP=login 198 Adrian Wail Akhlas and Eisya Eloksari, “Indonesia defends digital tax policy despite US scrutiny,” The Jakarta Post (June 211 UK Payment Service Platform: https://www.payments.service. 16, 2020), accessed online at: https://www.thejakartapost.com/ gov.uk/ news/2020/06/16/indonesia-defends-digital-tax-policy-de- 212 Indonesia One Map Portal: https://portalksp.ina-sdi.or.id/ spite-us-scrutiny.html 213 Although the law does not significantly affect the One Data pol- 199 For some of the companies covered, see: “Indonesia adds Twitter, icy, it will provide guidance to data stewards in particular when Zoom to tech companies that must pay 10% VAT”, Reuters, 2020. managing personal information. Bappenas as the coordinator of 200 The newly appointed VAT Collectors are provided with a VAT One Data implementation is expected to monitor closely the de- Collector ID and a tax registration letter. These VAT Collector velopment and incorporate elements of the Personal Data Pro- are required to create VAT collection slips providing informa- tection Law in the One Data technical implementation guidance. tion on VAT collection and payment, which can take the form of 214 Although there is no specific mention of One Data in the Om- a commercial invoice, billing, or order receipt to ease the burden nibus Law, there are several provisions regarding data that are of complying. VAT payment is made electronically using a billing relevant. For example, the decision on wages for MSME is man- code that is provided by the DGT, and can be made using IDR, dated to be based on consumption aggregates calculated by the USD dollar, or all other foreign currencies accepted by the DGT Central Statistics Agency (BPS). This means that data published system. VAT Collectors are required to file quarterly reports, by one GoI agency, BPS, will be used to set wages. Lack of clarity with at minimum data on: (i) number of users in Indonesia; (ii) on due process and transparency in data production may weaken amount of payments (excluding the VAT); (iii) amount of VAT the legitimacy of official statistics and leave them vulnerable to collected; and (iv) amount of VAT settled to the government. being politicized. The DGT is permitted to request further detailed filings on an annual basis with transaction-level data. 215 For example, institutions such as the Directorate General of Civ- il Registry of the Ministry of Home Affairs and the Directorate 201 For more on Australia’s GST, see: https://www.ato.gov.au/Busi- General of Tax of the Ministry of Finance traditionally control ness/International-tax-for-business/GST-on-low-value-im- and establish relatively more mature data management for civil ported-goods/Who-charges-GST/ registry and taxpayer datasets. 215 References 1 Akhlas, Adrian Wail and Eisya Eloksari, 2020. “Indonesia defends digital tax policy despite US scrutiny,” Jakarta: The Jakarta Post. Retrieved from: https://www.thejakartapost.com/news/2020/06/16/indonesia-defends-digital-tax-policy- despite-us-scrutiny.html (accessed March 25, 2021). 2 Alibaba Group and World Bank. 2019. E-commerce Development: Experience from China (English). Washington, DC: World Bank Group. 3 Al-Rikabi, Jaffar and Jan Loeprick. 2021. “Direct and indirect VAT revenue potential from taxing the digital economy in Indonesia”. The World Bank. Forthcoming. 4 Al-Rikabi, Jaffar, Khalil Rohman, and Hambali. 2021. “On the Macroeconomics of the Digital Economy in Indonesia”, CHAPTER — 05 Working Paper, World Bank. Forthcoming. 5 ASEAN and Plum Consulting. 2021. “ASEAN Digital Masterplan 2025”. Retrieved from https://asean.org/storage/ ASEAN-Digital-Masterplan-2025.pdf (accessed March 25, 2021). 6 ASEAN. 2021. “Joint Media Statement of the 1st ASEAN Digital Ministers’ Meeting and Related Meetings”. Retrieved from https://asean.org/joint-media-statement-1st-asean-digital-ministers-meeting-related-meetings/ (accessed March 25, 2021). 7 Australian Taxation Office. 2019. “Who charges GST”. Retrieved from: https://www.ato.gov.au/Business/International- BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA tax-for-business/GST-on-low-value-imported-goods/Who-charges-GST/ (accessed November 16, 2020). 8 Bhardwaj, Riaz; Yarrow, Noah; Cali, Massimiliano. 2020. “EdTech in Indonesia: Ready for Take-off ?”. World Bank, Washington, DC. 9 Blake, Robert. 2020. “Indonesia must tread carefully in considering new digital tax”. Jakarta: The Jakarta Post. Retrieved from: https://www.thejakartapost.com/academia/2020/05/06/indonesia-must-tread-carefully-in-considering-new- digital-tax.html (accessed March 25, 2021). 10 Council of the European Union. 2017. “VAT on electronic commerce: new rules adopted”, Press Release, Council of the EU. Retrieved from: https://www.consilium.europa.eu/en/press/press-releases/2017/12/05/vat-on-electronic-commerce-new- rules-adopted/ (accessed November 16, 2020). 11 European Commission. 2021. “New form of customs declaration for low value consignment”. Retrieved from: https:// ec.europa.eu/taxation_customs/news/new-form-customs-declaration-low-value-consignments_en (accessed November 16, 2020). 12 Global Mobile Suppliers Association. 2021. “4G-5G FWA Company Directory-2: Device Ecosystem – March 2021”. Retrieved from https://gsacom.com/paper/4g-5g-fwa-company-directory-2/ (accessed November 16, 2020). 13 International Telecommunication Union. 2019. “Digital Infrastructure Policy and Regulation in the Asia-Pacific Region”. Retrieved from https://www.itu.int/en/ITU-D/Regional-Presence/AsiaPacific/SiteAssets/Pages/Events/2019/ RRITP2019/ASP/ITU_2019_Digital_Infrastructure_5Sep2019FNL.pdf (accessed November 16, 2020). 216 14 Ministry of Finance Press Release. 2019. “Creating Fair Tax Treatment and Protecting Domestic Small and Medium Industries, Government Changes Provisions on Import of Shipments (E-Commerce)”. Retrieved from: https:/ /www. kemenkeu.go.id/publikasi/siaran-pers/siaran-pers-ciptakan-perlakuan-perpajakan-yang-adil-dan-lindungi-industri- kecil-dan-menengah-dalam-negeri-pemerintah-ubah-ketentuan-impor-barang-kiriman-e-commerce/ (accessed March 25, 2021). 15 Office of the United States Trade Representative. 2021. “USTR Announces Next Steps of Section 301 Digital Services Taxes Investigations”. Retrieved from: https:/ /ustr.gov/about-us/policy-offices/press-office/press-releases/2021/march/ustr- announces-next-steps-section-301-digital-services-taxes-investigations (accessed March 25, 2021). CHAPTER — 05 16 Organization for Economic Cooperation and Development webcast. 2020. “Webcast: Update on Economic Analysis and Impact Assessment”. Retrieved from: https:/ /www.oecd.org/tax/beps/webcast-economic-analysis-impact-assessment- february-2020.htm (accessed March 25, 2021). 17 Reuters. 2020. “Indonesia adds Twitter, Zoom to tech companies that must pay 10% VAT”. Reuters. Retrieved from: https:// www.reuters.com/article/us-indonesia-tax-digital-idUSKBN25Z2CU (accessed March 25, 2021). 18 World Bank. 2019. World Development Report 2019. Washington DC: World Bank Group. Retrieved from: https://www. worldbank.org/en/publication/wdr2019 (accessed March 25, 2021) 19 World Bank. 2018. March 2018 Indonesia Economic Quarterly: Towards inclusive growth. Retrieved from: https://www.worldbank. BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA org/en/country/indonesia/publication/indonesia-economic-quarterly-march-2018 (accessed November 16, 2020). 20 World Bank. 2019. World Development Report 2019. Washington DC: World Bank Group. Retrieved from: https://www. worldbank.org/en/publication/wdr2019 (accessed March 25, 2021). 21 World Bank. 2020. Closing the Gap between Workforce Development and Labor Markets: An Assessment of Indonesia’s Workforce Development System Based on the SABER Tool. Washington DC: World Bank Group. Retrieved from: https:/ /openknowledge. worldbank.org/handle/10986/35379?show=full&locale-attribute=en (accessed November 16, 2020). 22 World Bank. 2021a. Macro Poverty Outlook. Washington DC: World Bank Group. Retrieved from: https://www.worldbank. org/en/publication/macro-poverty-outlook (accessed March 25, 2021). 23 World Bank. 2021b. World Development Report 2021. Washington DC: World Bank Group. Retrieved from: https://www. worldbank.org/en/publication/wdr2021 (accessed March 25, 2021). 217 Annex. Policy Recommendations RECOMMENDED SOME PRIORITY ACTIONS RESPONSIBLE AGENCIES POLICY DIRECTION – HOW? – WHO? – WHAT? PRIORITY OBJECTIVE I: IMPROVE DIGITAL CONNECTIVITY TO UNIVERSALIZE ACCESS OPTIMIZE SPECTRUM ALLOCATION Implement the “Analogue Switch Off” by the end of 2022 in KOMINFO FOR MOBILE BROADBAND accordance with the Omnibus Law so as to free up the 700MHz band for greater rural connectivity using 4G and future 5G networks CHAPTER — 05 Optimize the higher frequency capacity bands, starting with the freeing up the 2.6 GHz band, followed by the 3.4-3.8 GHz band and potentially the broader 3.3-4.2 GHz band to facilitate 5G deployment in urban areas. Consideration should be had to the release of the mmWave spectrum bands for 5G in urban areas STRENGTHEN MECHANISMS TO Update the Telecom Law, PP 52/53, to allow for sharing of KOMINFO ENSURE SHARING OF ACTIVE AND active infrastructure on a B2B basis PASSIVE INFRASTRUCTURE Implement through regulations the passive infrastructure KOMINFO, MOHA, MINISTRY OF PUBLIC WORKS sharing mandated under the Omnibus Law BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA Require service providers benefiting from Universal Service KOMINFO Obligation fund to share subsidized infrastructure, and create a single entity responsible setting tariffs for passive infrastructure sharing STRENGTHEN COMPETITION ALONG Review current licensing regime and transition toward KOMINFO THE BROADBAND VALUE CHAIN international best practice of single/unified licensing to allow a larger number of providers to deliver the full portfolio of services Appoint an independent regulatory body for the telecom sector consistent with exemplar practice and Indonesia’s RCEP commitments PRIORITY OBJECTIVE II: MAKE THE DIGITAL ECONOMY WORK FOR ALL SUPPORT THE DEVELOPMENT OF Continue to develop basic physical infrastructure to improve MINISTRY OF TRANSPORT, DIRECTORATE LOGISTICS connectivity through strengthened and/or rehabilitated GENERAL OF HIGHWAYS, MINISTRY OF PUBLIC infrastructure (roads, ports, and electricity), leveraging WORKS AND HOUSING private sector capital and expertise as appropriate through PPPs Deepen reforms to reduce entry barriers to logistics and MINISTRY OF TRANSPORT, WITH SUPPORT FROM transportation services and build long-term investor MINISTRIES OF TRADE AND FINANCE AND confidence, to expand warehousing outside of large metropolitan THE COORDINATING MINISTRY OF MARITIME areas AFFAIRS AND INVESTMENTS Continue to modernize the National Post Service; standardize MINISTRY OF COMMUNICATION AND addresses and postcodes INFORMATICS NURTURE DIGITAL SKILLS AND Align non-formal education services and lifelong learning with MINISTRY OF EDUCATION AND CULTURE AND SKILLS FOR THE 21ST CENTURY needs of the working-age adult population, adopting a modular MINISTRY OF MANPOWER DIGITAL ECONOMY approach to course offerings with an emphasis on soft skills in addition to technical digital skills, incorporating employment- specific digital skills relevant for Indonesia, emphasizing a mindset of continuous learning, and offering opportunities to practice “self-managed” learning Build partnerships between tertiary institutions and the MINISTRY OF EDUCATION AND CULTURE/ private sector to train vocational students in high-tech areas PRIVATE SECTOR using practice-based curricula, practitioner-led instruction, and professional certification Ensure that tertiary education offers a minimum threshold MINISTRY OF EDUCATION AND CULTURE of foundational “transferable” higher-order skills such as critical thinking, problem-solving and communication, even in STEM fields Incentivize employers to offer internships and off-campus MINISTRY OF EDUCATION AND CULTURE/ learning to students MINISTRY OF MANPOWER Continue to develop data systems that allow for identification MINISTRY OF EDUCATION AND CULTURE/ of occupations and skills in demand, monitor educational MINISTRY OF MANPOWER institutions’ compliance with quality standards, and ensure that information on employability of individuals with various degrees, wage profiles, and occupation-specific courses is available to jobseekers and workers 218 RECOMMENDED SOME PRIORITY ACTIONS RESPONSIBLE AGENCIES POLICY DIRECTION – HOW? – WHO? – WHAT? PROMOTE SUPPLY OF DFS/DIGITAL Streamline the licensing and registration processes required to BANK INDONESIA AND OJK PAYMENT SOLUTIONS THAT CATER become a provider of digital financial services TO THE UNBANKED AND THE UNDERBANKED Standardize rules and procedures to enable interoperability of BANK INDONESIA payment schemes Incentivize the private sector to innovate and develop new DFS BANK INDONESIA AND DIRECTORATE GENERAL CHAPTER — 05 products that cater to the rural population OF TAXES Increase use cases of DFS products in the delivery of CMEA (FINANCIAL INCLUSION COUNCIL) AND government services (including Government-to-person payment) to MINISTRY OF SOCIAL AFFAIRS sustain the commercial viability of DFS agents Strengthen the legal and regulatory framework for managing BANK INDONESIA AND OJK risks related to data governance and privacy, cyber security and operational risk, and financial integrity, in order to build consumer trust in DFS products BEYOND UNICORNS: HARNESSING DIGITAL TECHNOLOGIES FOR INCLUSION IN INDONESIA USE TAX POLICY INSTRUMENTS TO Complete the design of measures to reform tax policy and MINISTRY OF FINANCE (DIRECTORATE GENERAL ENSURE A LEVEL PLAYING FIELD modernize tax administration in relation to DE transactions, OF TAXES, DIRECTORATE GENERAL OF CUSTOMS ensuring adoption and communication of consistent rules, AND EXCISE, AND FISCAL POLICY AGENCY) institution of efficient IT systems to administer registration, filing, payment and use of digital transaction data with other third-party data and taxpayer data for strengthened compliance risk management Lower the overly generous VAT threshold to expand the digital economy tax base PRIORITY OBJECTIVE III: USE DIGITAL TECHNOLOGIES TO PROVIDE BETTER SERVICES AND UPGRADE CITIZEN-STATE INTERACTIONS DEVELOP A NATIONAL DIGITAL ID Pass the Personal Data Protection law KOMINFO FRAMEWORK Launch a whole-of-economy national digital ID initiative to OFFICE OF THE PRESIDENT, MOHA, AND define the optimal model for Indonesia, bringing together KOMINFO government, private sector and civil society Close the coverage gaps in the national population registry MOHA database (SIAK), streamline the new registration and update processes, and introduce Introduce a national digital ID system or ecosystem fit-for- OFFICE OF THE PRESIDENT, MOHA, AND purpose in the Indonesia context KOMINFO REORIENT FROM A NARROW Strengthen the political and bureaucratic influence needed to OFFICE OF THE PRESIDENT FOCUS ON E-GOVERNMENT TO A move the agenda by placing a central government authority like COMPREHENSIVE NATIONAL DIGITAL the Office of the President in the driving seat. TRANSFORMATION AGENDA Create an appropriate institutional structure in the form of a dedicated digital transformation implementation agency to consolidate various disparate initiatives and to drive the agenda in an integrated and coherent manner. IMPLEMENT A WHOLE-OF- Implement the One Data Policy comprehensively, in priority BAPPENAS GOVERNMENT DATA MANAGEMENT sectors relevant for managing the post pandemic recovery (e.g., POLICY education, health, social protection or MSMEs) Build alliances with key institutions such as MENPAN-RB, BAPPENAS Kominfo, and MoHA to instill One Data principles into the implementation of the digital government and digital ID initiatives 219 WORLD BANK INDONESIA FLAGSHIP REPORt