Improving Access to Financial Services in Indonesia

CONFERENCE EDITION

Conference Edition

1

Improving Access to Financial Services in Indonesia

Conference Edition

Conference Edition

2

THE WORLD BANK OFFICE JAKARTA Indonesia Stock Exchange Building, Tower II/12-13th Fl. Jl. Jend. Sudirman Kav. 52-53 Jakarta 12910 Tel: (6221) 5299-3000 Fax: (6221) 5299-3111 THE WORLD BANK The World Bank 1818 H Street N.W. Washington, D.C. 20433 USA Tel: (202) 458-1876 Fax: (202) 522-1557/1560 Email : feedback@worldbank.org Website : www.worldbank.org Printed in December 2009 Improving Access to Financial Services in Indonesia is a product of staff of the World Bank with the Financial Support from the Dutch Government, The findings, interpretation and conclusion expressed herein do not necessarily reflect the views of the Board of Executive Directors of the World Bank or the government they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denomination 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 of acceptance of such boundaries.

This Report is for discussion during the conference “Enhancing Access to Formal Financial Services in Indonesia”, 9-10 December 2009 in Jakarta, Indonesia. Do not cite without expressed permission of the World Bank.

Conference Edition

3

List of Abbreviations

A2F Access to Finance ADB Asian Development Bank AML Anti-Money Laundering APEC Asia-Pacific Economic Cooperation API Arsitektur Perbankan Indonesia (Indonesian Banking Architecture) ATM Automated Teller Machine Bapepam-LK Indonesia Capital Market and Financial Institution Supervisory Agency BEE Black Economic Empowerment BI Bank Indonesia BKD Badan Kredit Desa (Rural Credit Bank) BPD Bank Pembangunan Daerah (Regional Development Bank) BPR Bank Perkreditan Rakyat (People’s Credit Bank) BPS Badan Pusat Statistik (Central Bureau of Statistics) BRI Bank Rakyat Indonesia CGAP Consultative Group to Assist the Poor FAO Food and Agriculture Organization FATF Financial Action Task Force GDP Gross Domestic Product GoI Government of Indonesia GTZ Gesellschaft fur Technische Zusammenarbeit ICD Islamic Corporation for Development of the Private Sector IDIC Indonesian Deposit Insurance Corporation (LPS) IFAD International Fund for Agricultural Development IFC International Finance Corporation IFI International Financial Institution ILO International Labour Organization IPO Initial Public Offering Inpres Presidential Instruction IT Information Technology KSP Kelompok Sains Petani (Farmer's Science Group; Savings and Loan Cooperative) KTP Kartu Tanda Penduduk (National Identity Card) KUK Kredit Usaha Kecil (Small Scale Business Loan) Kupedes Kredit Umum Pedesan (Village Credit Program) KUR Kredit Usaha Rakyat (Peoples Business Credit) KYC Know Your Customer LDKP Lembaga Dana Kredit Pedesaan (Rural Credit Fund Institution) LPS Lembaga Penjamin Simpanan (Deposit Insurance Organization) MASS Microfinance Access and Services Survey MCSME Ministry of Cooperatives and Small and Medium Enterprises MoF Ministry of Finance MSME Micro Small and Medium Enterprises NBFI Non-Bank Financial Institution NGO Non-Government Organization NPWP Nomor Pokok Wajib Pajak (Taxpayer Number) OECD Organization for Economic and Cultural Development OJK Otoritas Jasa Keuangan (Financial Services Authority)

Conference Edition

4

PJTKI Perusahaan Jasa Tenaga Kerja Indonesia (Indonesian Workforce Service Company) PODES Statistik Potensi Desa (Village Potential Statistics) PP Perum Pegadaian (State-owned Pawnbroker) PPATK Pusat Pelaporan dan Analisis Transaksi Keuangan (Financial Intelligence Unit) ProFI Promotion of Small Financial Institutions Puskesmas Pusat Kesehatan Masyarakat (Community Health Center) RICA Rural Investment Climate Assessment SAF Survey of Access to Finance SEACEN South East Asian Central Banks SID Sistem Informasi Debitur (Debtor Reporting System) Simpedes Simpanan Pedesaan (Village Saving Program) SIUP Surat Ijin Usaha Perdagangan (Business Trading License) SKB Surat Keputusan Bersama (Joint Ministerial Decree) SME Small and Medium Enterprises SoB State-Owned Bank Susenas Survey Sosial Ekonomi Nasional (National Socio-Economic Survey) UNDP United Nations Development Programme

Conference Edition

5

Table of Contents List of Abbreviations ........................................................................................................................... 3 Table of Contents ................................................................................................................................ 5 List of Tables ........................................................................................................................................... 6 List of Figures .......................................................................................................................................... 7 List of Boxes ............................................................................................................................................ 8 Chapter 1: Executive Summary ........................................................................................................ 10 Chapter II: The Current Supply of Financial Services in Indonesia .......................................... 19

II.1 Overview of Indonesia’s Financial Sector ................................................................... 19 II.2 The Commercial Banking System ................................................................................ 20 II.4 Regional Reach of Indonesia’s Commercial Banks ................................................... 25 II.5 People’s Credit Banks (BPRs) ....................................................................................... 30 II.6 Microfinance Institutions .............................................................................................. 33 II.7 Sharia Finance & Banking ............................................................................................. 38 II.8 Non-Bank Financial Institutions (NBFIs) ................................................................... 41 II.9 Financial Services in Support of Remittances ............................................................ 42 II.10 Summary of Policy Issues, by Area ............................................................................... 47

Chapter III: Demand-side Aspects; What do People Want? ...................................................... 50 III.1 Introduction and Overview ........................................................................................... 50 III.2 Survey Results: Supply of Financial Services............................................................... 54 III.3 Survey Results: Demand for Types of Financial Services ....................................... 59 III.4.1 Survey Results: Demand for Savings Accounts ..................................................... 60

III.4.2 Barriers to Accessibility: Reasons for Savers Being Unbanked ...................... 63 III.4.3 Key Socio-economic Characteristics of Indonesia’s Savers ............................ 66 III.4.4 Econometric Analysis of Savings ........................................................................ 69

III.5.1 Survey Results on Demand for Loans .................................................................... 70 III.5.2 Who Does Not Borrow? ...................................................................................... 74 III.5.3 Barriers to Accessibility: Why Do They Not Borrow? ................................... 76 III.5.4 Key Socio-Economic Characteristics of Indonesia’s Borrowers .................... 77 III.5.5 Econometric Results for Loans ........................................................................... 82

III.6 The ‘Truly Financially Excluded’: No Loans And No Bank Accounts .................. 83 III.7.1 Survey Results on Demand for Insurance .............................................................. 85

III.7.2 Econometric Results for Insurance .................................................................... 90 III.8.1 Other Survey Results: Risks to Financial Well-Being ........................................... 90

III.8.2 Other Survey Results: Demand for Other Financial Instruments ................. 91 III.8.3 Other Survey Results: Expressed Demand for Financial Products ............... 93

III.9 Summary of Policy Issues .............................................................................................. 95 Chapter IV: Regulatory Impediments to Access ......................................................................... 99

IV.1 Introduction ..................................................................................................................... 99 IV.1.2 Indonesia’s Financial Regulators & Supervisors ............................................. 102

IV.2 Assessment of the Current Regulatory System, by Service Provider .................... 103 IV.2.1 Regulatory Barriers to Access: Commercial Banks ......................................... 104 IV.2.2 Regulatory Barriers to Access: Bank Perkreditan Rakyat (BPRs) ................ 108 IV.2.3 Regulatory Barriers to Access: Cooperatives ................................................... 114 IV.2.4 LPS’s Role in the Regulatory Environment ..................................................... 116 IV.2.5 Regulatory Barriers to Access: Finance Companies ...................................... 118

Conference Edition

6

IV.2.6 Regulatory Barriers to Access: Insurance Companies .................................... 119 IV.2.7 Barriers to Access: Pawnshops ......................................................................... 120

IV.3 Summary of Policy Issues, by Service Provider ....................................................... 122 Chapter V: Special Topics Concerning Access to Finance ....................................................... 126

V.1 MSMEs and Access to Finance .................................................................................. 126 V.1.1 Key Findings of Bank Indonesia’s Survey of MSMEs ................................... 127 V.1.2 MSME Lending & Policies in Indonesia .......................................................... 129

V.2 Migrant Workers and Access to Finance .................................................................... 133 V.2.1 A2F Findings on Indonesian Migrant Workers .............................................. 133

V.2.2 Migrants & Access to Financial Services ....................................................................... 136 V.2.3 The Pre-Departure Stage .................................................................................... 137 V.2.4 Financial Service Needs During Migration ...................................................... 143 V.2.5 The Post-Migration Period ................................................................................. 148

V.3 Mobile Banking ............................................................................................................. 149 V.3.1 The International Context .................................................................................. 149 V.3.2 The State of Mobile Banking in Indonesia ....................................................... 152 V.3.3 Mobile Banking to Improve Access to Financial Services.............................. 154

V.4 Policy Issues .................................................................................................................... 156 Chapter VI: Policy Recommendations for Improved Access to Financial Services .............. 159

VI.1 General Strategic Matters .............................................................................................. 159 VI.2 Regulatory Issues .......................................................................................................... 160

VI.2.1 Mobile Banking .................................................................................................... 160 VI.2.2 Commercial Banks ............................................................................................... 160 VI.2.3 BPRs ...................................................................................................................... 161 VI.2.4 LPS ......................................................................................................................... 161 VI.2.5 Cooperatives ......................................................................................................... 162 VI.2.6 Pawnshops ............................................................................................................ 162 VI.2.7 Other Financial Institutions ............................................................................... 162

VI.3 MSMEs ............................................................................................................................ 162 VI.4 Overseas Workers and Remittances........................................................................... 163 VI.5 Matters of More Limited Concern ............................................................................. 163

Annexes Draft .................................................................................................................................. 168 References ......................................................................................................................................... 198

List of Tables Table 1: International Comparison of Financial Sectors: Selected Financial Indicators ........ 19 Table 2: International Comparison of Bank Branch Density ..................................................... 20 Table 3: Indonesia, Number of Financial Institutions ................................................................ 22 Table 4: Number of Kabupaten/Kotamadya With and Without Branches of Commercial

Banks ........................................................................................................................................... 29 Table 5: Sharia Banking Offices ....................................................................................................... 38 Table 6: Sharia Rural Bank Performance Indicators ..................................................................... 39 Table 7: Official Overseas Worker Placement, 2004 .................................................................... 43 Table 8 : Cost of Domestic Remittances, April 2009 ................................................................... 46 Table 9: Travel to Banks, Time & Cost ......................................................................................... 56 Table 10: Saver’s Travel to Banks, by Urban/Rural & Java/Off Java ....................................... 57 Table 11: Non-savers' Travel to Banks, Urban/Rural & Java/Off-Java ....................................... 57

Conference Edition

7

Table 12: Summary of Indonesian Savers' Characteristics .......................................................... 69 Table 13: Household Indebtedness ................................................................................................ 77 Table 14: Borrowers’ Characteristics, by Gender .......................................................................... 79 Table 15: Summary of Insurance Holders’ Characteristics ........................................................ 90 Table 16: Summary of Households Use & Interest in Formal Financial Products ................. 96 Table 17: Financial Providers and Financial Services in Indonesia ......................................... 102 Table 18: BPR Start-up Capital Requirement ............................................................................. 109 Table 19: Access to Finance Indicators-Loans and Deposits ..................................................... 113 Table 20: Legal Basis of the Cooperative System....................................................................... 114 Table 21: Financial Services of Savings and Loan Cooperatives .............................................. 115 Table 22: Legal Basis of the Finance Companies ........................................................................ 118 Table 23: Sample Savings Product Offered by Banks and the Post Office ........................... 142 Table 24: Indicative Cost of Remitting Money from a Middle-East Country ........................ 146 Table 25: Key Policy Recommendations on Improving Access to Financing ....................... 164 Table 26: Statistical Comparison of Village-Level Variables in the A2F Survey and PODES

................................................................................................................................................... 176 Table 27: Statistical Comparison of A2F vs. Non-A2F Villages Using SUSENAS Data .... 177

List of Figures Figure 1: Share of the Population with formal financial access .................................................. 11 Figure 2: Savers’ Financial Inclusion ................................................................................................ 12 Figure 3: Borrowers’ Financial Inclusion ........................................................................................... 12 Figure 4: Financial Structure (December 2008) ............................................................................. 20 Figure 5: Total Banks by Type of Bank .......................................................................................... 21 Figure 6: Number of Bank Branches and ATMs, by Type of Bank ........................................... 23 Figure 7: Bank Head Offices & Branches, by Province ............................................................... 25 Figure 8: Bank Branches & Density Vs. Per Capita Income, by Province ................................ 26 Figure 9: Total ATMs & Density Vs. Per Capita Income, by Province ..................................... 26 Figure 10: Number of Kabupatens without Bank Branches, by Province ................................ 30 Figure 11: Financial Performance of BPRs .................................................................................... 33 Figure 12: Top 20 Developing-country Recipients of Remittances, 2005 ................................. 44 Figure 13: Bank Account Distribution, by Type of Bank ............................................................ 54 Figure 14: Bank Branch Locations ................................................................................................. 55 Figure 15: Travel Time to Nearest Bank Branch .......................................................................... 56 Figure 16: Average Time to Reach Select Institutions ................................................................ 58 Figure 17: Waiting Time to be Served in a Bank .......................................................................... 59 Figure 18: Financial Products Used by Households .................................................................... 60 Figure 19: Summary of Survey Results for Savers ........................................................................ 61 Figure 20: Overlap Among Savings Providers ............................................................................... 62 Figure 21: Reasons for Having a Bank Account, by Type of Bank ............................................ 63 Figure 22: Main Reasons for Not Having a Bank Account ......................................................... 64 Figure 23: Main Reasons for Not Opening a Bank Account, by Income Deciles .................. 65 Figure 24: Yearly Total Household Expenditure .......................................................................... 66 Figure 25: Savers’ Socio-economic Characteristics ....................................................................... 67 Figure 26: More Savers’ Socio-economic Characteristics ............................................................ 68

Conference Edition

8

Figure 27: Savings by Per Capita Expenditure ............................................................................. 69 Figure 28: Summary of Survey Results for Borrowers ................................................................. 71 Figure 29: Purpose of Loan, by Service Provider ........................................................................ 72 Figure 30: Loan Sizes, by Institution .............................................................................................. 73 Figure 31: Indicative Loan Interest Rates ....................................................................................... 74 Figure 32: Non-Borrowers by Per Capita Expenditure .............................................................. 75 Figure 33: Non-Borrowers’ Characteristics ................................................................................... 75 Figure 34: Characteristics of Borrowers ........................................................................................ 78 Figure 35: Borrowers’ Characteristics, Urban/Rural .................................................................... 79 Figure 36: Borrowers’ Characteristics, by Age ............................................................................... 80 Figure 37: Borrowers’ Characteristics, by Employment Status ................................................... 80 Figure 38: Borrowers’ Characteristics, by Type of Job ................................................................. 81 Figure 39: Borrowers’ Characteristics, by Enterprise Ownership ............................................... 81 Figure 40: Borrowers, by Region, by per Capita Expenditure..................................................... 82 Figure 41: Bank Loan, by Region, by per Capita Expenditure ................................................... 82 Figure 42: Characteristics of ‘Truly Financially Excluded’, by per Capita Expenditure .......... 83 Figure 43: Characteristics of the ‘Truly Financially Excluded’ ................................................... 84 Figure 44: Characteristics of the ‘Truly Financially Excluded’, continued ................................ 84 Figure 45: Insurance Holders, by per capita Expenditure ........................................................... 85 Figure 46: Types of Insurance Ownership ..................................................................................... 87 Figure 47: Characteristics of Holders of Insurance ..................................................................... 88 Figure 48: Characteristics of Insurance Holders, continued ........................................................ 89 Figure 49: MSME Lending ............................................................................................................. 130 Figure 50: MSME Characteristics .................................................................................................. 132 Figure 51: Characteristics of Migrant Workers ........................................................................... 134 Figure 52: Migrant Workers, by Type of Work and Legal Status ............................................. 135 Figure 53: Indonesian Migrant Workers and Access to Finance ............................................. 136 Figure 54: Migrant Workers in the Formal Sector, Financed by Employers .......................... 138 Figure 55: How Migrant Workers Finance Pre-departure Expenses ...................................... 139 Figure 56: Method of Receiving Remittances, by Host Country ............................................. 146 Figure 57: Access to Financial Services Versus Access to Credit ............................................. 169 Figure 58: Difference Between Access to and Usage of Services ............................................ 170 Figure 59: Samples in PODES 2005 vs. Access to Finance 2007, at Village Level .............. 174 Figure 60: A2F Sample Characteristics ........................................................................................ 178 Figure 61: Urban/Rural Split ........................................................................................................ 178 Figure 62: Respondents, by Province ............................................................................................ 179 Figure 63: Sample Characteristics, Age by Gender ..................................................................... 179 Figure 64: Respondents, by Employment & Education ........................................................... 180 Figure 65: Respondent Wage Composition ................................................................................ 180

List of Boxes Box 1: The Role of State Vs. Private Banks in Indonesia ........................................................... 24 Box 2: BRI’s MASS Survey: Insights to Banks’ Capacity to Reach the Poor .......................... 32 Box 3: The Shadowy Faces of Indonesia’s Lintah Darat ............................................................. 36 Box 4: BRI’s Unit Desa System ...................................................................................................... 37 Box 5: Wider Access to Financial Services through Sharia Mobile Banking ........................... 39 Box 6: What Do Household Surveys in Other Countries Tell Us? ........................................... 51

Conference Edition

9

Box 7: What Do Other Surveys for Indonesia Tell Us? ............................................................. 52 Box 8: Sampling Methodology of the Indonesian Survey .......................................................... 53 Box 9: Some Details on Insurance in the Survey ......................................................................... 86 Box 10: Survey Results on Mobile Banking ................................................................................... 92 Box 11: Survey Results on Financial Literacy ................................................................................ 93 Box 12: Evidence on Easing Constraints to Opening Bank Accounts ...................................... 94 Box 13: South Africa’s Approach to Improving Access to Financial Services ...................... 101 Box 14: Some History on Regulation of Other Microfinance Institutions ............................. 110 Box 15: Legal Framework for Microfinance Institutions in Indonesia .................................... 122 Box 16: Case Study: Migrant Workers Bound for Taiwan ......................................................... 140 Box 17: Perum Pegadaian (The State-owned Pawnshop) & Access to Credit ....................... 141 Box 18: Remittance Method and Frequency from a Middle East Country .......................... 144 Box 19: The Methods of Remittance: Formal vs. Informal ...................................................... 145 Box 20: Voices of Hard Experience .............................................................................................. 147 Box 21: Promising Technologial Advances – South Africa’s WIZZIT ................................... 150 Box 22: International Experience in Reducing the Cost of Mobile Banking .......................... 151 Box 23: Mobile Banking Product Development in Indonesia .................................................. 153 Box 24: Weaknesses in the Regulatory Environment for Branchless Banking in Indonesia ...... Sampling Methodology of the Indonesian Survey ...................................................................... 154

Conference Edition

10

Chapter 1: Executive Summary

Access to formal financial services is by now widely recognized as critically important to alleviating poverty around the world. An impressive literature has developed supporting the proposition that increased access to financial services has a significant impact on poverty. Financial inclusion is high on the policy agenda of several developing countries worldwide1 where banking and financial systems are usually underdeveloped, and often cater only to large firms and/or high-income individuals. This skewed distribution of finance hinders the growth and development of smaller firms and poorer households. There is a growing recognition that increasing access to formal financial services has both private and social benefits. Extending the breadth of financial service availability fosters economic growth and can improve income distribution. Providing access to financial services means mainstreaming people in many dimensions, fostering economic inclusion, and providing financial institutions with new and expanding markets. Improving access requires actions on both the supply and demand side, by both the public and private sectors as well as changes in the institutional environment. Recent experiences in several countries show that with the right information on who lacks access and for what reasons, policies can be adjusted and products can be designed to scale up access, especially with new technology.

The Government of Indonesia has also placed high importance on this issue acknowledging the significance of access to financial services as a constraint to development, and the authorities are initiating policies aimed at increasing access. One of the key constraints to concrete policy action in improving access to financial services, in particular at the household level, is concrete data and analysis on what exactly the demand-side view of the constraints is i.e. what do consumers and the currently un-banked population think of financial services, what access do they have and what products and services do they need. Such information would provide a solid basis that can inform policy and product development by the market to achieve the desired outcomes. While a significant amount of data and analysis is available on the issue of access to credit by Small and Medium Enterprises (SMEs) 2, there is a dearth of such work in the area of access to broader financial services.

The objective of this report – whose key feature is a nationwide household survey of access to financial services - is to provide data, analysis, and recommendations that can assist the authorities as well as other stakeholders such as the financial services industry in getting an insight into access to financial services in Indonesia. It begins with a review of the supply 1 See, for example, the work of Beck, Demirguc-Kunt and Martinez Peria (2004), (2005) and (2006). Banking the Poor (2009c), Access to Finance Study: Brazil (2004), India (2006c), Nepal (2007b), and Pakistan (2009b) 2 World Bank (2006e),“Making the New Indonesia Work for the Poor,”; World Bank (2006):“Revitalizing the Rural Economy: An assessment of the investment climate faced by non-farm enterprises at the District level” ; Significant work done by GTZ on rural banks: See http://www.profi.or.id/; FAO and IFAD on rural finance: http://www.ruralfinance.org/ and http://www1.deptan.go.id/kln/FAO%20in%20%20Indonesia.htm. ILO on migrant workers: See http://www.ilobkk-migration.org/, IFC/GTZ, and CGAP (2009d), ADB (2007): “Low Income Households’ Access to Financial Services” (2007)

Conference Edition

11

side of financial services from an access perspective, followed by an examination of the demand side of access to those services. It then looks at regulatory barriers that can help increase financial inclusion and certain other closely related topics that are of current policy interest, namely MSMEs, overseas migrant workers and mobile banking. The immediate purpose of the report is to inform policy-makers and the industry on who does (and does not) have access to financial services, including the constraining factors for broader access. The objective is to identify—as specifically as possible—measures that can lower barriers to access for poorer households, especially measures that work in cost-effective ways. Survey Results on the Demand for Financial Services

Just about half of Indonesia’s population has access to formal financial services. This is better than countries such as China, Pakistan, Bangladesh, and the Philippines. However, it is worse than countries such as Sri Lanka, Thailand, and Malaysia. Figure 1: Share of the Population with formal financial access

Sources: World Bank 2008; Nenova, Niang, and Ahmad (2009); Indonesia Access to Finance Survey

Commercial banks – that dominate the Indonesian financial sector - serve a relatively small proportion of Indonesian households. One-third of Indonesians don’t save at all, and can be considered ‘financially excluded’ (see Figure 2 below). Similarly, less than half of Indonesians save at banks, and of those who do save at banks, two-thirds also save at some other type of service providers. Considering the overlap between banks and the informal sector, informal institutions actually service more savers than do banks.

Conference Edition

12

Figure 2: Savers’ Financial Inclusion

68% Financially Included 32% Financially Excluded

Save at Banks: 47%

Save Only Informally:

18%

Don’t Save: 32%

0% 50% 100% Save at Other Formal: 3%

A mere 17% of Indonesians borrow from banks, and about 1/3rd more borrow from the informal sector. On this basis, roughly 40% of the population is ‘financially excluded’ from credit (see the Figure 3: below). The most important reason for exclusion appears to be inadequate documentation; evidence indicates that lack of collateral is a secondary reason.

Figure 3: Borrowers’ Financial Inclusion 60% Financially Included 40% Financially Excluded

Borrow from

Banks

Borrow Informally

Can’t Borrow

0% 50% 100% Borrow from Other Formal Sources Voluntarily Excluded

The single most important financial service that households would like to have is a bank account. The most important reason for having a bank account is ‘security’; by far the most important reason for not having a bank account is ‘lack of income’ or not having a job. Bank credit is important for households too, but it is considerably further down the list of priorities. Credit is still concentrated in the informal sector; and sources of credit are widely diversified among service providers.

Taken together, the above findings underscore the importance of expanding both the liabilities and assets (that is, deposits and credits) sides of financial services institutions, while raising depositors’ incomes by broader policies of economic development. They also underscore the challenge to Indonesia’s formal financial system, especially the banks, of significantly expanding its client base, to reach a much larger portion of the population.

Conference Edition

13

The ‘truly financially excluded’, which is to say, those who have neither a savings account nor a loan are predominantly poor, poorly educated, live off-Java in rural areas, and do not own non-farm enterprises. Off Java residents are more than twice as likely to have neither a bank account nor a loan, than are on Java residents.

Physical access to formal financial services is not a generalized problem; for the vast majority (some 95%) of Indonesians accessibility to banks is rated as convenient (or very convenient). The exception is the rural, off-Java regions, especially if water transport is involved. Nonetheless, it’s notable that average travel times to reach bank branches compare favorably to key public services like hospitals, schools and other health facilities.

One simple, low-cost solution for borrowers who want a lower interest rate on credit, is for them to open a bank account. Banks and MFIs both charge nominal interest rates of about 30% per annum, and both offer lower rates to borrowers who have a bank account. Nominal interest rates from other formal and informal sources of credit except for loans from employers, friends and neighbours tend to be higher. Therefore, enabling poor households to open a bank account would be a simple way to get them linked to the formal financial system and, among other benefits that it may offer, could help lower interest rates on their borrowings. Consumers do respond to more attractive pricing of financial services including lower charges on savings accounts. However, demand looks somewhat price inelastic which implies that banks’ need to carefully consider whether it is in their financial interest to reduce fees. This is consistent with identified bank policies that set deposit rates and administration fees in a way that discourages small savers. One policy option in this regard is to encourage banks to offer basic banking services or ‘no frills’ accounts (noting that some banks already do so and, under an agreement between several major banks and Bank Indonesia, ns are afoot to introduce more such accounts in 2010). Several countries in the world are implementing such schemes, although in different ways. Another option would be to encourage regulatory and technological advances (like mobile banking) that allow all service providers to reach more customers at lower cost, although international experience shows that while mobile banking has had a significant impact on payments services, it has had less impact on other financial services. Another innovation, especially given Indonesia’s geography would be to focus on bank partnerships with non-bank correspondent outlets of all forms to spread services. Some new products that would be of interest to consumers are contractual savings products for urban residents or mobile savings services for rural residents. As for extending the reach of formal bank services deeper into the lower strata of society, the most promising avenue looks like mobile banking – even if at first, it is likely that mobile banking is largely likely to be focussed on payments services. Even the poorest in remote villages have access to mobile phones these days, and the survey uncovers considerable interest in mobile banking among those with a mobile phone, but no bank account.

Conference Edition

14

Key Aspects of the Current Supply of Financial Services

Although the number of banks has declined substantially since the 1997/98 crisis, banks have significantly expanded the reach of their financial services through greater branching and use of ATMs. Other formal sector providers such as cooperatives and the state owned pawnshop have also expanded physical outreach. Per capita income and population (or land mass) go a long way towards explaining the reach of Indonesia’s commercial banking system, on a provincial basis. The only notable exceptions relative to the average are Jakarta (which is ‘over-serviced’) and East Kalimantan (which is large, resource rich, sparsely populated and ‘under-serviced’).

In considering banks, it is important to distinguish between the commercial banks and People’s Credit Banks (BPRs), which are regional in nature and much smaller in size. Among the former, only a relatively small number currently make a large, direct contribution to the financing of poorer households. And even among these, their focus tends to be on better-off clients. However, the commercial banks do make important contributions in other ways, for instance, indirectly through the so-called ‘linkage program’ with BPRs. In addition, their numbers include one of the largest micro finance institutions in the world (BRI’s Unit Desa system), and they are aggressive, opportunistic competitors who are quick to move into promising new markets. Such characteristics imply that the commercial banks are the institutions most likely to introduce new cost-cutting technologies and to put competitive pressure on other financial institutions. Still, they are only a part of the near-term answer to better access to financing because they do not currently reach deeply enough into the lower strata of Indonesian society. By contrast, the BPRs and other small financial institutions offer much more promise in the near- to medium-term. Despite a great deal of diversity, they are often on the frontline of the delivery of financial services to MSMEs and poorer households, including in very remote parts of Indonesia. As detailed below, much can be done on the regulatory front to extend their reach into lower segments of Indonesian society.

Sharia banking (and more generally, sharia financing) – though currently holding a small market share - has been expanding rapidly for about a decade. These institutions are particularly important because they cater almost exclusively to the lower end of the market, including in rural regions. Also, Indonesia’s first sharia bank (established in 1992) is a leading innovator in extending financial services to poor remote areas through mobile banking.

Among other financial institutions that provide access to financial services for the poor, three are especially notable: the state-owned pawnshop; cooperatives; and other micro finance institutions. Each has its peculiar impediments and possible solutions, which are discussed further below. Non-Bank Financial Institutions (NBFIs) have made much smaller inroads in being relevant for the poor – although there has been some encouraging progress in a few areas in recent years, for example, in micro insurance and leasing.

Conference Edition

15

Main Recommendations

Wider access to financial services by lower-income Indonesians needs both public and private sector interventions as well as some innovative public-private partnerships. The focus should be on broader access to financial services overall for the lower-income and poor as opposed to a narrow focus on access to credit. Credit is important for the poor, but savings requirements rank much higher. Much of the lower-income segments find existing financial products to be inappropriate to their needs. Designing and pilot testing appropriate products through partnerships could potentially open up more customers for the formal financial sector and vice versa, provide access to the formal financial system for a greater share of the Indonesian population.

From a public sector perspective, strengthening the existing legal and regulatory framework for various formal financial institutions would be a good first step in aiding the process. For every important service provider, there are aspects of the regulatory framework that could be eased for the sake of improving access to financial services, without compromising prudential safety.

For commercial banks, among quick regulatory winners, the most promising avenue is mobile banking, which holds considerable promise. Mobile banking holds great promise for reducing costs and extending reach – although, in line with international experience, it is likely to initially focus on payments services and remittances. The main economic issue of access concerns finding low-cost ways to deliver the services that lower-end consumers want. BI has recently made notable regulatory advances in this regard, but more is possible. For instance, non-bank service providers can issue e-money, but only for payment purposes. If they want to offer person-to-person services, they need a remittance license and eligibility requirements are currently an (unintended) barrier to entry. Simpler ways are available to accomplish the same regulatory purpose, without restricting entry. To deliver mobile banking services cheaply, the economies of scale offered by a network of non-bank retail agents is vital. This would entail allowing banks the discretion to out-source services using a network of non-bank third parties, while holding the banks responsible for agent activities. For mobile banking to reach deeply into the ‘financially excluded’, there are also important KYC issues to be addressed. For example, simplified KYC requirements for low-risk, low-value accounts (and transactions) would permit the remote opening of bank accounts in isolated areas, and they would allow agents to facilitate the opening of new accounts. Other smaller steps on commercial banks would also be helpful. For example, an official policy on dormant accounts might help reduce banks’ monthly administration fees. Easier policies for banks to unilaterally close inactive, non-zero accounts could be offered to banks with institutional arrangements in place for the management of such accounts after they are closed. BI’s recent agreement with major commercial banks to introduce basic banking – such as the proposed launch of a new saving product called "TabunganKu" (My Saving) in early 2010 – is also a step in the right direction. On bank reporting, annual business plans could be combined with the banks’ annual reports, and elements of the

Conference Edition

16

business plan could be required only in fairly general terms. Regulations concerning relocations of branches and ATM machines look unnecessarily restrictive, and general descriptions of location should suffice. It would also be useful to ease official regulations on branching, at least to bring them into line with Bank Indonesia’s current, relatively liberal approach to implementation.

Concerning BPRs, there are several regulatory barriers that could be eased to improve access, and Bank Indonesia currently seems to be re-thinking policy in this area. Consideration could be given to a lower tier of minimum start-up capital for small BPRs in remote locations; NGOs and foreign investors could be allowed to take some ownership positions in BPRs that are looking for capital. Reporting requirements could be re-examined for small BPRs in locations without adequate communications services. Written disclosure requirements could be waived in areas of low financial literacy, and replaced by oral briefings for new customers, including in the local language, where appropriate. In a step that also applies to commercial banks, Know Your Customer (KYC) regulations could be simplified for small accounts and requirements for taxpayer numbers waived for small loans below a pre-specified threshold. Moreover, for the sake of regulatory transparency, the current tight branching requirements could be brought into line with BI’s liberal approach to implementation. To enforce regulations on BPRs, BI is already working hard to augment its capacity. As an additional interim step, BI might seek additional, temporary assistance by contracting firms that specialize in micro-finance.

Important regulatory steps could be taken concerning cooperatives, pawnshops and other microfinance institutions. On cooperatives, the most important issues appear to be prudential. These should be addressed on a sector-wide basis before any significant problems surface and potentially erode memberships’ existing access to financial services. Concurrently, there needs to be an upgrading of the MCSME’s regulatory and supervisory capacity. This could include temporary outsourcing of the function to firms specializing in micro-finance. Concerning pawnshops, the state-owned monopoly could be officially opened up to competition from the private sector – there are several privately run pawnshops operating anyway at present. In parallel, there needs to be a discussion on the extent to which these institutions needs to be brought under a formal regulatory umbrella, keeping in mind international experience. With regard to other microfinance institutions, the most productive way forward looks like restoring momentum to the drafting of a new Micro-Finance Law, and encouraging public debate on the issues during the process. It will be important that the new Law emphasize facilitation and access, taking into account emerging global experience regarding regulation and supervision of such institutions. In support, linkage programs between commercial banks and BPRs could be expanded to include non-bank MFIs, and it would be helpful if a similar role could be defined for NGOs. For most forms of insurance companies a stronger foundation is needed for healthy expansion of this industry. The industry faces several fundamental structural issues that need to be addressed such as the existence of several weak and unviable firms before the industry can play a large role in expanding access. An important exception is the micro-insurance business, which is currently expanding rapidly, with the benefit of a successful public-private partnership. This could serve as a model for other products aimed at the lower-end of the income spectrum. There are also emerging models globally in this area that can be explored.

Conference Edition

17

The report also addresses issues relating to MSMEs and migrant workers as special topics of interest to the Government at the time of the report. MSMEs’ issues of access to financial services are virtually one-dimensional, that is, they are only credit-related, with problems of access mainly arising at the micro level. Indonesia has been pro-active in MSME finance policies for many years, but there is general dissatisfaction with results to date, despite large expenditures by the government. This is due in large part to the past emphasis, which has been on subsidized credit programs; in line with international experience, these have largely not been successful. The Government continues to make access to credit for MSMEs a major policy issue and has initiated the Kredit Usaha Rakyat (KUR) program as a means to consolidate the existing programs and put in place an integrated credit guarantee scheme to bring previously unbanked MSMEs into the formal banking sector. While a formal review of this program was underway at the time of writing of this report, the Government has also announced a significant scaling up of the program. Depending upon the results of the assessment, the government may consider strengthening the modifying/existing KUR program; it may also be a model approach to consider assessments of its other subsidized lending programs.

Migrant workers’ issues are also high on the Government’s agenda. From an access to finance perspective, this group should be of particular interest to financial institutions, given the large remittances that these workers send home. In general, in several areas, to assist migrant workers, Indonesia also could ask to re-negotiate the terms of its Memoranda of Understanding on Migrant Workers with recipient countries 3 aimed at better balancing the interests of the workers themselves with interests of employers and recruitment agencies. From the perspective of increasing access to financial services, specific points of negotiation could include acceptable forms of identification (which would not limit access to formal sector services) and exempting small transfers from formal identification requirements. To convince banks of the commercial value of this market, it might be useful to explore possibilities for innovative public private partnerships to bring greater segments of these workers into the formal financial sector. One approach may be wider use of domestic guarantors (or co-signers) for pre-departure loans to migrant workers. Development partners (or NGOs) with particular interests in these workers might consider acting as the guarantor in pilot projects that can then be examined for their scaling-up potential. Another could be design of innovative savings instruments that will permit these workers to save their earnings for use over a longer period of time. To lower obstacles presented by Know Your Customer regulations, it may be possible to negotiate minimum documentation requirements for small transfers, that will not be a risk to the global AML/CFT efforts, but that also enhance access for migrant workers.

The way ahead

A major government focus on enhancing access to financial services for the poor and low-income segments (as opposed to a narrow focus on access to credit alone) is essential if financial inclusion in Indonesia is to be significantly increased. Several developing countries have adopted policy statements and strategies in this regard. Access to finance is an issue that cuts across several stakeholders : the authorities such as Bank Indonesia, 3 A recent World Bank study makes several practical suggestions in this regard; see The Malaysia-Indonesia Remittance Corridor (2008a).

Conference Edition

18

Bapepam-LK, Ministry of Finance, Ministry of Cooperatives and SMEs, etc. as well as the financial sector including state owned and private banks, non-bank financial institutions, as well as NGOs, foundations and think-tanks working in this area. Technology and education will play a key role in scaling up access rapidly – so actors such as telecom companies, academic institutions and financial literacy providers will also be important. To the extent useful, Indonesia’s international development partners can provide knowledge and financial inputs. Working together it is possible to scale up access to financial services for a greater share of Indonesians and provide a sound basis for sustained poverty reduction.

Conference Edition

19

Chapter II: The Current Supply of Financial Services in Indonesia

This Chapter looks at the current supply of financial services in Indonesia. It takes a broad perspective on issues, beginning with an overview of the Indonesian financial sector, to put the different service providers in context. There follows, a discussion of services, segmented by type of service provider, because different providers cater to different segments of the market for financial services.

II.1 Overview of Indonesia’s Financial Sector

By international standards, Indonesia’s financial sector is still small in relation to GDP (see Table 1 and World Bank (2006a)). On a rough measure of total financial assets, it represents the equivalent of just over 100% of GDP, way below other large Asian economies (India about 300% and China well over 500%) and barely comparable to countries like the Philippines and Pakistan. Among other indicators, Indonesian bank credit to the private sector is among the lowest of regional comparators (Table 1). In addition, Indonesia’s banking system does not provide a particularly dense coverage of the nation, on average (Table 2).4 Population per branch is high relative to most comparators; land area per branch is about average, although notably better than Mexico and Brazil.

Table 1: International Comparison of Financial Sectors: Selected Financial Indicators

Total Financial Assets

Credit to Private Sector

Equity Market Capitalization Private Bonds Public Bonds

GNI per capita

(% of GDP) (% of GDP) (% of GDP) (% of GDP) (% of GDP) (Atlas Method,

in US$) China 542.5% 114.5% 189.8% 0.2% 0.4% $ 2,360 Malaysia 383.5% 108.8% 180.2% 4.4% 7.1% $ 6,540 India 298.3% 47.4% 155.4% 0.7% 0.2% $ 950 Thailand 210.6% 84.2% 79.8% 1.4% 1.2% $ 3,400 Brazil 205.1% 49.8% 104.3% 2.9% 3.6% $ 5,910 Pakistan 150.2% 29.4% 48.9% 0.6% 1.3% $ 870 Philippines 128.7% 23.8% 71.6% 1.7% 11.3% $ 1,620 Indonesia 103.6% 25.4% 48.9% 2.1% 1.1% $ 1,650 Sri Lanka 60.8% 34.0% 23.3% 0.3% 0.2% $ 1,540 Bangladesh 54.8% 37.7% 10.0% … … $ 470 Source: World Development Indicators, 2008 (data for 2007) and Bringing Finance to Pakistan’s Poor, 2009, World Bank

4 The data in Table 2 include BRI’s Unit Desa system, which account for almost 1/3 of all bank branches in Indonesia.

Conference Edition

20

Table 2: International Comparison of Bank Branch Density

Population per Land Area per Branch Branch (sq km) India 14,888 44 Indonesia 12,547 110 Mexico 11,924 236 Brazil 9,331 470 U.S. 3,568 117 France 2,331 22 Japan 1,959 6 Germany 1,479 6 Source: “Brazil: Access to Financial Services”, World Bank, 2004.

By type of institution, Indonesia’s formal financial system is heavily dominated by banks (see Figure 4). Other types of financial institution remain small and are concentrated heavily in western Indonesia (Java, Sumatra and Bali).5 Some of these are discussed briefly in Section II-8 below. However, for issues of access for poorer households, only commercial banks have a wide geographical spread among formal institutions; even pawnshops (see Chapter IV) are tiny in comparison.

Figure 4: Financial Structure (December 2008)

II.2 The Commercial Banking System

Since the financial crisis of 1997/98, there have been important changes in the composition of financial institutions in Indonesia. In particular, the numbers of commercial banks (especially private banks; see Figure 5), people’s credit banks (BPRs) and multi-finance companies have all declined markedly (Table 3). This is largely a result of the government’s broader economic policy that is aimed at consolidation of the number of institutions in these

5 See World Bank (2006b).

Rural Inst, Pawnshop, VC

2%

Mutual Funds2%Securities Firm

2%Finance Companies

5%

Insurer6%

Pension Funds4%

Banks79%

Conference Edition

21

sub-sectors (see Chapter IV).6 By contrast, the numbers of bank branches, bank ATMs, pawnshop branches and cooperatives have all increased significantly (Table 3).

Figure 5: Total Banks by Type of Bank

This change in composition is an important development, with two key dimensions. First, although the number of banks (including BPRs) has fallen, their reach in the provision of services has increased significantly. For example, the number of bank branches has increased by 70% since 2000 and the number of ATMs has almost tripled (Figure 5). And second, the greatest expansion is by institutions that serve lower income groups, that is, cooperatives and pawnshop branches. For their part, the banks are expanding by low cost means (like branches, ATMs and mobile banking), which suggests that they, too, are increasingly better positioned to serve lower end markets.

6 For example, during 2007, Bank Indonesia approved 105 BPRs for consolidation into 19 entities and 5 others had their operating licenses revoked. BI issued approvals in principle for 27 new ones (Bank Indonesia (2007), p. 131).

0

20

40

60

80

100

120

140

160

1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008

Source: Bank Indonesia

Num

bers

State Ow ned Banks Private Com. BanksRegional Banks Foreign & Joint Banks

Conference Edition

22

Source: BI, Bapepam-LK, Annual Report of Perum Pegadaian, Ministry of Cooperatives & SMEs

Table 3: Indonesia, Number of Financial Institutions

Banks

Bank Branches, including cash

offices Bank ATMs BPRs

Venture Capital

Multi-finance

Pawnshop (branches)

Cooperatives

2000 151 6,374 7,114 2,419 59 245 692 103,0772001 145 6,657 8,997 2,355 60 245 722 110,7662002 142 6,844 10,613 2,143 60 244 739 118,6442003 138 7,554 11,837 2,141 60 239 774 123,1812004 134 7,808 13,772 2,158 60 237 806 130,7302005 131 7,918 15,862 2,009 60 237 840 134,9632006 131 8,980 16,991 1,880 60 214 869 141,3262007 130 9,535 18,596 1,817 60 211 899 147,0002008 124 10,868 20,792 1,772 60 210 1,331 149,913

Conference Edition

23

Figure 6: Number of Bank Branches and ATMs, by Type of Bank

All types of banks have participated in the expansion of branches (Figure 6). Foreign banks have increased the most rapidly in percentage terms, by some 350%, but from a relatively small base. Among the others, state-owned banks increased by about 80%; regional development banks (BPDs) by about approximately 70%; and private banks by 60%. In terms of ATMs, state-owned banks also expanded more rapidly than private banks; the state bank ATM network more than quadrupled from about 1,900 in 2000 to 8,225 in 2007; private banks’ ATMs more than doubled from nearly 4,700 to more than 11,350 over the same period (for geographical expansion, see Chapter II.4, below).

Number of Branches, by Type of Bank

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

Year

SOB

, BPD

, PC

B, &

BPR

0

50

100

150

200

250

300

350

400

FOB

Regional Bank State Owned Bank Private Com Bank BPR Foreign Owned Bank (Incl Joint Banks)

ATMs, by Type of Bank

0

2,000

4,000

6,000

8,000

10,000

12,000

1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008

Year

Num

bers

State Owned Banks Regional Banks Private Com. Banks Foreign & Joint Banks

Conference Edition

24

Box 1: The Role of State Vs. Private Banks in Indonesia Prior to the financial deregulation drive of the 1980s, Indonesia’s 7 state banks dominated the local financial landscape due to nationalization of all Dutch banks in the 1950s and 1960s. Each state bank had a somewhat specialized focus, with official statements underscoring their ‘social and development role’. For example, Bank Ekspor Impor (ExIm) emphasized international trade and finance; Bank Pembangunan Indonesia (Bapindo) was a development bank; Bank Tabungan Negara (BTN; a savings bank) specialized in housing; and Bank Rakyat Indonesia (BRI) concentrated on rural credit. The other three, Bank Dagang Nasional (BDN), Bank Bumi Daya (BDN) and Bank Negara Indonesia (BNI; the first and largest) leaned more towards the corporate sector. On a system-wide basis, licensing, credit and interest rates were tightly regulated by the central bank, and there were a large number of central bank re-financing schemes targeting politically important sectors. All of the state banks—some more than others—suffered internal governance problems. Much of this changed during the 1980s with a series of so-called deregulation packages that opened up the sector to entry and competition (see, for example, Cole and Slade (1996) and Binhadi (1995)). The state banks lost market share rapidly, but their official roles were little changed. Two subsequent developments changed the state banks irreversibly. The first was the financial crisis of 1997/98, which led to the merger of Bank ExIm, BDN, BBD and Bapindo into Bank Mandiri. The second was a series of partial (minority) privatizations via public stock issuance, beginning with 25% of BNI in 1996; within a decade or so, all except BTN were partially privatized. There were also critical management changes at some of the state banks that have improved internal governance appreciably. Today, only BTN and BRI retain specialized roles, and even BRI had been emphasizing corporate lending till recently (corporate lending was BRI’s main source of financial losses during the crisis of 1997/98). Bank Mandiri and BNI operate in direct competition with the private banks with little by way of special privileges. Such advantages as the state banks now hold, stem mainly from their size. Mandiri is the largest bank in the country (almost 16% of total assets) and BRI and BNI hold down third and fourth place, respectively, each with roughly 10% market share. BRI also holds an operational advantage over others in the micro-finance business because of its established, wide branch network that will be difficult for any newcomer to challenge. Nevertheless and despite large growth in numbers of branches and services like ATMs (noted in the main text), the share of the state banks’ assets in the banking sector continues on a long-term secular decline. By 2007, it was down to 37% compared with 49% in 2001 and almost 80% in the early 1980s. Another notable aspect of Indonesia’s state banks concerns their relatively impressive performance on some of this report’s criteria, for example, ATM and branch expansion. It should also be mentioned that the government, as majority shareholder, still makes occasional special demands of the state banks. For example, as the target of ‘moral suasion’ when particular policies prove difficult to implement by conventional means; the source of temporary expertise in times of crisis, for instance when weak banks need to be closed; and the leading edge of special programs, like the Kredit Usaha Rakyat (KUR) and the Linkage Program between commercial banks on the one side and BPRs and cooperatives on the other.

Conference Edition

25

II.4 Regional Reach of Indonesia’s Commercial Banks

In a country as large and diverse as Indonesia, it is very important to look at the regional distribution of commercial bank services when examining issues related to access to finance. Unfortunately, data on the dispersion of banking services is not available in sufficient detail to permit a robust spatial analysis across the country. For many purposes, data for provinces (and special regions, like DKI Jakarta) are the best available and these hide a great deal of residual diversity in, for example, Jakarta, Banten and East Java. Using a relatively coarse indicator (see Figure 7), it is obvious that bank outlets are heavily tilted in favor of the more heavily populated regions, namely, Jakarta, West Java, Central Java and East Java. An issue arises as to whether this unequal distribution of bank services is accounted for by population (or land mass) and income levels.

Figure 7: Bank Head Offices & Branches, by Province

To investigate this issue, the number of branches (and branch density) are regressed on real per capita GDP, on a provincial basis for end-2006. The results are presented in Figure 9 with branch density measured in two ways: on a population-adjusted basis (see the middle panel in Figure 8) and on a land areas basis (see the lower panel in Figure 8).7

7 The size of the bubble in the top panel of Figure 8 indicates the relative size of GDP in that province. In the middle panel, it indicates relative population; in the lower panel, relative land mass.

-100

100

300

500

700

900

1100

1300

1500

1700

1900

2100

2300

2500

2700

Aceh

Wes

t Sum

atera

Jambi

Beng

kulu

Bang

ka B

elitun

g

Jaka

rta

Banten

Yogy

akarta

Bali

East

Nusa T

engg

ara

Centra

l Kali

manta

n

East

Kalim

antan

Goronta

lo

Sout

h Sula

wesi

Sout

heas

t Sula

wesi

North

Maluku

Papu

a

Head Quarters Branch Offices Sub Branch Off ices Cash Offices Representative Office

Conference Edition

26

Figure 8: Bank Branches & Density Vs. Per Capita Income, by Province

Riau Riau Islands

Jakarta

West Java

Central Java

East Java

Bali East Kalimantan

-250

250

750

1250

1750

2250

2750

-1500 3500 8500 13500 18500 23500 28500 33500 38500

Bra

nche

s

Total Branches, at Dec 2006

Riau

Riau Islands

Jakarta

Central Java

East Java

Bali

East Kalimantan

-1

4

9

14

19

24

29

-1500 3500 8500 13500 18500 23500 28500 33500 38500

Tota

l Bra

nche

s pe

r 100

,000

pe

rson

s

Branch Density per 100,000 persons, at December 2006

RiauRiau Islands

Jakarta

East KalimantanNorth Maluku

Papua

-600

400

1400

2400

3400

-1500 3500 8500 13500 18500 23500 28500 33500 38500

Tota

l Bra

nche

s pe

r 1,0

00 k

m2

Constant GDP per Capita (IDR Thousand)

Branch Density per 1,000 sq km persons, at December 2006

Conference Edition

27

As indicated in Figure 8, there is a clear upward relationship between branches (and branch density) and real per capita income. The relationships are all statistically significant (t statistics on real capita GDP ranging from 2.3 to 5.1), with reasonably good fits (R2s ranging from .38 to .67). Statistical significance and fits are better for branch density (that is, for the lower 2 panels in Figure 8) than for branches. This implies that adjusting for land mass or population is helpful in explaining the relationship. Exactly the same conclusions hold, using ATMs as the metric instead of bank branches (see Figure 9). Indeed, the statistical fits are a bit tighter and the statistical significance a bit higher. This is important for policy purposes. Specifically, the relatively low-cost of service provided by ATMs is more responsive to income levels than are the relatively high-cost services provided by bank branches. By implication, if service costs can be further reduced—for example, by mobile banking—the banks would move further into regional markets. Looking at Figures 8 and 9, it is clear that Jakarta (the political, business and financial capital of Indonesia) is consistently a large, high-side outlier. That is, Jakarta has far more bank branches and ATMs than would be expected on the basis of population, land mass and income. All other provinces8 fall more or less in line with expectations after adjustment for income and population or land mass. How to explain this result for Jakarta? Two explanations look plausible. First, agglomeration theory (see the World Development Report 2009) postulates that economic development tends to occur disproportionately in clusters, for example, in relatively small, densely-populated regions like Jakarta. The second explanation concerns income inequality. The relatively large number of very wealthy people and businesses in Jakarta probably lead to a disproportionately large number of bank services in that region. The use of average income as an explanatory variable doesn’t capture the effect. And what can be said by way of policy conclusions? This analysis provides strong prima facie evidence that variations in the existing supply of financial services at the provincial level are largely explained by variation in population, land mass and income. Only Jakarta and East Kalimantan are clear exceptions, Jakarta being ‘over-serviced’ and East Kalimantan ‘under-serviced’. By implication, targeted efforts on ‘financially under-served’ areas will need to operate below the provincial level and they will need to take clear account of banks’ demonstrated capacity to respond to broad market conditions.

8 Another exception is the resource rich, but sparsely populated province of East Kalimantan. It is consistently a low-side outlier, implying that it has fewer bank branches and ATMs than would be expected on the basis of income, population and land mass.

Conference Edition

28

Figure 9: Total ATMs & Density Vs. Per Capita Income, by Province

Riau Riau Islands

Jakarta

West Java,

East Java,

BaliEast Kalimantan

-5000

5001000150020002500300035004000450050005500600065007000

-500 4500 9500 14500 19500 24500 29500 34500 39500

Tota

l ATM

sFTotal ATMs, at December 2006

Riau

Riau Islands

Jakarta

East Kalimantan

Papua-1500

1000

3500

6000

-500 4500 9500 14500 19500 24500 29500 34500 39500

Tota

l ATM

s pe

r 1,0

00 k

m2

Constant GDP per Capita (IDR Thousand)

ATMs Density per 1,000 sq km, at December 2006

Conference Edition

29

Some data are available on the number of local jurisdictions (kabupaten in rural areas and kotamadya in urban areas) with no commercial bank branches. Surprisingly, this number has gone up substantially from 16 in 1997 to 156 in 2007 (Table 4). Upon closer inspection, this larger number appears to be entirely due to the creation of new local government jurisdictions; the total number of such jurisdictions has increased from nearly 314 to 464 over the same period (Table 4). By contrast, the number of local jurisdictions with branches has remained fairly constant at around 300. On this basis, it seems clear that banks do not move into new regions, just because political boundaries have been re-defined; their locational decisions are based upon commercial viability, which is normally insensitive to political segmentation of existing jurisdictions.

Table 4: Number of Kabupaten/Kotamadya With and Without Branches of Commercial Banks

Year # of Kabupaten/Kotamadya

with Branches

Total Kabupaten/Kotamadya

Percentage # of Kabupaten/Kotamadya

with no Branches

Percentage

1997 298 314 94.9% 16 5.1%1998 304 314 96.8% 10 3.2%1999 299 316 94.6% 17 5.4%2000 306 334 91.6% 28 8.4%2001 268 349 76.8% 81 23.2%2002 304 362 84.0% 58 16.0%2003 303 384 78.9% 81 21.1%2004 311 424 73.3% 113 26.7%2005 305 448 68.1% 143 31.9%2006 309 448 69.0% 139 31.0%2007 308 464 66.4% 156 33.6%

Source: Bank Indonesia, Ministry of Home Affairs, Decentralization Support Facility and World Bank Estimates

The regional distribution of the local jurisdictions without a branch of a commercial bank provides additional evidence that Indonesia’s banks are sensitive to commercial opportunities as measured by population density. Most ‘unserved kabupatans’ are in the relatively remote, sparsely populated provinces of Papua, Kalimantan and Sulewesi (see Figure 10).9 However, there is no strong relationship between ‘unserviced’ locations and per capita income (see Figure 10).

9 The ‘unserviced’ jurisdictions were Sumatra (56); Sulawesi (23); Kalimantan (22); the Papua island group (19); the Malukus (11); and Nusa Tenggara (6). As of 2007, there were only a handful (6) on Java and none on Bali.

Conference Edition

30

Figure 10: Number of Kabupatens without Bank Branches, by Province

(GDP increases to the right)

II.5 People’s Credit Banks (BPRs)

As will be clear from the subsequent section and Chapter IV, BPRs10 play an important role in providing access to financial services for small customers. BPRs are a subset of Indonesia’s formal banking system, but they cater to smaller clients; they are not part of the payments system; they face strict branching restrictions; and they are regulated differently than commercial banks (see Chapter IV). In total, the BPRs are tiny compared to the commercial banks; as of end-2008, loans outstanding of rural banks were less than 2% of those for commercial banks. Also, there has been considerable consolidation in recent years (see Table 3) along with some failures of BPRs that has led to Bank Indonesia taking steps to strengthen the regulation and supervision of this sector. BPRs’ profitability, capital adequacy stability of funds, and loan-to-deposit ratios (LDR) are all comparable or better than commercial banks; this is important from an access perspective as it shows that these institutions are likely to continue to play an important role in the sector going forward11. However, BPRs as a group do have a higher rate of NPLs than commercial banks.

10 They are commonly called ‘Rural Credit Banks’ in English, but this is very misleading since many BPRs are in urban or peri-urban areas. 11 BPRs have fewer sources of funding than commercial banks. For example, they can’t borrow from capital markets or from off-shore; they aren’t publicly listed; and they can’t accept demand deposits. They do, however, have ‘linkage programs’ with commercial banks and some get relatively cheap financing from private foundations (yayasan).

East

Nus

a Te

ngga

raG

oron

talo

Nor

th M

aluk

u

Mal

uku

Wes

t Sul

awes

i

Wes

t Nus

a Te

ngga

ra Ben

gkul

u

Sout

heas

t Sul

awes

i Lam

pung

Sout

h Su

law

esi

Cen

tral

Jav

a

Jam

bi

Cen

tral

Sul

awes

i

D I

Yogy

akar

ta Wes

t Kal

iman

tan

Nor

th S

ulaw

esi

Ban

ten

Wes

t Jav

a

Bal

i

Cen

tral

Kal

iman

tan

Wes

t Sum

atra

Papu

a

Sout

h K

alim

anta

n

Nor

th S

umat

ra

East

Jav

a Sout

h Su

mat

era

Wes

t Pap

ua Ban

gka

Bel

itung

Isla

nd

Nan

ggro

e A

ceh

Dar

ussa

lam

Ria

u

Ria

u Is

land

East

Kal

iman

tan

D K

I J

akar

ta

0

2

4

6

8

10

12

14#

Kab

upat

ens

Conference Edition

31

The main criticisms of BPRs run parallel to those that are normally directed at commercial banks namely, that BPRs serve only down to the middle segment of the microfinance market; they do not directly target micro-entrepreneurs; publicly-owned BPRs cover different locations, operating as quasi-monopolies; and Bank Indonesia’s regulations push them to do conservative, collateralized lending.12 For its part, BI has been emphasizing their ‘community development’ role, and may be considering a requirement that 80% of their lending be into local areas. This is also discussed in Section II.10, below.

12 See, for example, http://www.bwtp.org/arcm/Indonesia/I_Country_Profile/Indonesia_country_profile.htm, in particular the discussion under ‘BPRs or People’s Credit Banks’.

Conference Edition

32

Box 2: BRI’s MASS Survey: Insights to Banks’ Capacity to Reach the Poor As noted in the main text, Bank Rakyat Indonesia (BRI) is Indonesia’s premier lender to MSMEs and it has built upon a long history of micro-lending in Indonesia (see Box 5). In the fall of 2002, BRI conducted a Microfinance Access and Services Survey (MASS) of 1438 households in 6 provinces (roughly half the size of the Survey of Chapter III) to map the financial landscape and to gauge potential markets. The enumerators were BRI loan officers and other professions, which provided a unique opportunity to assess the creditworthiness of households using BRI’s standard commercial procedures. The survey included other detailed information on assets, businesses and economic and social changes. Many of the results are useful for the purposes of this report, as summarized immediately below. First, loans for small businesses are important, but 30% of the time, households use the loans for consumption needs, like: school fees; medical treatment; household improvements; meeting daily consumption needs; and contributing to social and holiday expenses. This finding, which holds across a wide range of low income households, dilutes the case that lending to MSMEs will unleash a burst productive investment spending. Second, concerning the creditworthiness of the ‘unbanked’, results indicate that creditworthiness increases with income, but 40% of the unbanked poor and very poor were deemed creditworthy by the examiners. At all levels of income, more borrowers were judged to be creditworthy than are currently borrowing or saving. Third, the lack of collateral was cited as a deterrent by only 10% of households that are creditworthy and not borrowing, and the BRI examiners concurred. This confirms the insights of successful micro lenders, which is to lend against income, not assets. Fourth, and possibly most important, about half of creditworthy poor households are risk averse and do not seek credit. Some households simply don’t want it. The caveat here is that clearly household voluntarily exclude themselves at the prevailing price of credit.

Fifth, of those poor unbanked households who did seek credit, most desired relatively small loans, even by BRI’s standards. Calculations by the researchers showed that BRI (given its current fee structure and banking practices; emphasis not in original) would lose money when lending on the small scale desired by poorer households. On this particular result, the authors conclude that it remains commercially unviable for banks to lend to many poor households, even for BRI, Indonesia’s microfinance specialist.

Summarizing overall conclusions, evidence suggests that about 40% of the unbanked poor are creditworthy by MFI standards. However, roughly half of them do not want to borrow. And of the unbanked poor who did seek credit, about half are too small to be commercially viable even for BRI, given that bank’s current commercial standards. This survey, which was heavily tilted towards credit services, provides support for the findings of Chapter III that access to financial services covers much wider ground than access to credit, including for the poor.

Source: Johnston and Morduch (2007) and (2008).

.

Conference Edition

33

Figure 11: Financial Performance of BPRs

Total Assets & Loans Loan to Deposit Ratio

NPLs Rates of Return on Equity and Assets

Source: Bank Indonesia.

II.6 Microfinance Institutions

In Indonesia, the term ‘micro-finance’ covers a wide range of institutions that can be very confusing. Unlike in some other countries, where micro-finance resides in the informal sector or at the edges of the formal sector, in Indonesia, the largest segments are well within the formal sector. Conceptually, Indonesia’s system of micro-finance may be envisaged as a pyramid (see the accompanying Figure and Box 14 in Chapter IV), 13 with different layers serving different segments and regulated differently. At the apex of the pyramid is BRI’s Unit Desa system (see Box 4);

13 The pyramid is reproduced from Bank Indonesia and GTZ (2000).

0

2

4

6

8

10

12

14

2001 2002 2003 2004 2005 2006 2007 2008

(%)

NPL Ratio

Conference Edition

34

formally, it resides well within the commercial banking system. Indeed, BRI is the third largest commercial bank (by assets) and micro-credit is its core business.14 In the mid-section of the MFI pyramid are people’s credit banks (BPRs; discussed above), which are also part of the formal banking system. The BPRs are not a homogeneous group of institutions: for instance, they vary widely by size (some are larger than the smaller commercial banks); market niche (urban versus rural); differential performance as between urban and rural;15 and operations (e.g., the former Badan Kredit Desa (BKDs; see Box 16) officially do not accept voluntary deposits). At the lower end of the pyramid, a bewildering array of institutions overlaps the formal and informal sectors. They include financial cooperatives; pawnshops; LDKPs (Lembaga Dana Kredit Perdesaan; see Box 16), which go by different names in different provinces and are mostly owned by provincial/district governments; 16 NGOs; Grameen Bank replicas; large numbers of unofficial moneylenders (see Box 4); and probably millions of small, informal savings and credit societies (arisan). Their main characteristics, which separate them from the regular commercial banks, are that they operate on a very small scale and they tend to lend against income whereas the commercial banks lend against collateral. Little formal data are available on most of these institutions, although one estimate (Morowijoyo, 2007) puts their number at 46,248 (excluding arisan) as of 2005. This report relies largely upon the Survey reported in Chapter III.17 Effectiveness of the players at the lower end of the pyramid in reaching the poorest depends in large part upon operational soundness of the MFIs, especially their sustainability. For example, if they are competing with subsidized government credit programs, their commercial viability is probably at risk. Likewise, if a significant portion of the MFIs’ financing originates with governments, international institutions or private donors, they are exposed to shifts in the internal policies of those collaborators. The lesson for any MFI at the lower end of the pyramid – in line

14 BRI units dominate micro borrowers’ deposits. They constitute 72% of deposits and 68.8% of total micro-deposits (Marutowijoyo, 2007), setting a high competitive standard for other players in this market. 15 See, for example, Bank Indonesia and GTZ (2000), p. 16. 16 In Bali and West Sumatra, they are owned by the villages in which they operate. LDKPs are concentrated on Java and Bali, and they are usually supervised and supported by the Regional Development Banks (BPDs). See http://www.profi.or.id/index.php?option=com_content&task=view&id=44&Itemid=55. 17 Other useful details on some micro institutions are provided in http://www.bwtp.org/arcm/.

BRI Unit Desa

BPRs

LDKP, Coops, BKDs, Pawnshops, NGOs, Etc.

Conference Edition

35

with global experience - is that independence and viability ultimately requires a self-sustaining profit niche, not dependence upon transfers, no matter how well-meaning the source of those transfers. As a general rule, the lower the segment in the customer pyramid, the larger is the size of the potential customer base and the lower its average income. Interest rates on credit follow the reverse pattern; effective rates of interest at BRI are the lowest, followed by BPRs and LDKPs. Cooperatives (Koperasi) and NGOs usually provide subsidized loans for their membership and target groups. Many of these lower-end micro-finance providers operate with little by way of formal legal basis. This incomplete legal and regulatory framework is limiting the reach of MFIs; restricting their financing (banks are penalized for lending to organizations without legal status); and providing an incentive for other players to operate in the least regulated segment. A micro-finance law has been under debate for the better part of a decade and has gone through various versions including submission to Parliament. However, it has not yet been approved and questions remain on the role of various institutions in implementation of any regulatory and supervisory framework for MFIs. In early 2009, under the leadership of the Coordinating Ministry of Economic Affairs (CMEA), a joint decree as signed between the Ministries of Finance, Cooperatives, Home Affairs and Bank Indonesia to provide a regulatory framework under the existing laws to the non-bank and non-cooperative MFIs that currently operate outside such a framework. The Joint Decree aims to formalize MFIs into one of four categories: (1) People Business Credit or BPR, (2) Saving and Loan Cooperatives, (3) Village-owned microfinance institutions or BUM Desa, or (4) Venture Capital Companies. This needs to be implemented. Also regarding the lower end of the pyramid, there are a number of subsidized government lending programs that attempt to reach into that strata. These programs are run by different government departments and state-owned enterprises, with different cultures, philosophies of economic development and supervisory skill sets. These compete with—and sometimes undermine—the micro-finance providers, often by replenishing schemes that aren’t working18,19:

18 ‘Finance for the Poor’, Indonesia Policy Brief, January 2005, available at http://www.worldbank.org.id. 19 Bank BRI’s Annual Report for 2007, p.57 notes its participation in two programs: the P4K program (Pembinaan Peningkatan Pendapatan Petani Nelayan or Assistance in Income Generation for Marginal Farmers and Fishermen) run by the Ministry of Agriculture; and Micro and Small Business Loans (KUMK-SUP). It’s notable that the P4K program has sometimes been identified as an exception to the general rule that subsidized credit programs don’t work (‘Rural Investment Climate Report’, World Bank, July 206). Other sources have criticized it as extremely expensive and heavily subsidized (Banking with the Poor, Asia Resource Centre for Microfinance; http://www.bwtp.org/arcm/Indonesia/)

Conference Edition

36

Box 3: The Shadowy Faces of Indonesia’s Lintah Darat

As in many developing countries, informal moneylenders are believed to abound in Indonesia, in both urban and rural areas. Understandably, they are reluctant to talk with World Bank researchers, even trusted ones, because they operate outside the law and shun public exposure. Even borrowers are hesitant to talk much about their experience with the lintah darat. The lenders’ collective name, lintah darat, literally means ‘land bloodsucker’, and it indicates the disdain that most Indonesians hold for the profession. Still, there is strong demand for their services because they provide a vital source of quick, short-term credit, despite usurious rates of interest. The lintah darat seem to fall into two broad categories: those who do it as a profession; and those—often women—who do it part-time on a small-scale basis, operating within their own peer group. The two mentioned below are in the former category. One person in the latter category declined to discuss her business with the Bank’s field team, knowing that the information would be used for this report. Instead, her input (among several others) helped to develop the characters described below. Johannes Hutabarat (JH is a fictional character, pieced together from partial, anecdotal information) lives in Jakarta. Originally, JH is from Medan and he maintains a strong local network among his regional ethnic group. He operates out of a middle-class residential neighbourhood in central Jakarta and he mainly serves retail clients. Typically, his clients come to him on a word-of-mouth basis; they are normally desperate for credit and need it fast (‘uang panas’), usually for personal emergencies. JH prefers salaried workers, especially government employees but they are hard to get because competition from the banks is tough; they, too, like the assured, regular flow of income for repayment. JH’s typical loan is small (maybe only a few million rupiah) and maturity varies between a few weeks and a few months. The effective annual rate is often well in excess of 50%. In the event that repayment difficulties develop, JH turns to his network of strong-arm friends from his hometown; their first resort is intimidation of the client, followed, only if absolutely necessary, by physical violence. JH needs his clients to repay—out of fear, if necessary—but he also values return business and needs to maintain his reputation within the community as a serious businessman. Djoko Tanuwijoyo (DT) is a real person, but not his real name. He lives in a mid-sized city on Java and he runs a different kind of operation than JH. His clients are usually SME businessmen who have problems with bridge financing, perhaps to meet a payroll or to cover an unexpected cost overrun. They come to him because ‘the banks ask too many questions’ and want collateral; the closely-held nature of local businesses puts a premium on confidentiality and quick response. Often DT’s clients are family members who fully understand the nature of his business and expect little by way of special favors in return. Actually, DT prefers family members (or best friends referred by them) because ‘they are more trustworthy than the rest’. This is important to DT because his business is largely based upon mutual trust. DT’s typical loan is much larger than JH’s and can run into the billions of rupiah. His interest rates are lower than JH’s, but they may still run up to 5% per month, according to his assessment of the risk involved in the loan. Because he lives in a smaller city and operates without a sizable band of enforcers, DT relies more on personal relationships and less on intimidation and violence. Still, his willingness to turn to such methods should not be tested because his patience with debt write-offs has its limits. Source: World Bank field interviews.

A special word is in order concerning NGOs, which tend to serve the very lowest levels in the pyramid, often on a quasi-humanitarian basis in the most remote corners of their host country. Their main advantage is a willingness to serve levels of society that are not commercially viable. However, historically NGOs were allowed only the narrowest space in which to operate and relations with the government were poor.20 Also, the 2001 Law of Foundations (Yayasan) put certain NGOs in a difficult position. According to the 2001 Law, foundations may only provide social, humanitarian and religious services, which prevents them from being involved in income-generating

20 See, for example, Schwarz (1994), p. 253-4, and p. 60 and Haggard (2000), p. 186.

Conference Edition

37

activities, like microfinance. They were required to cease micro-finance operations, or become BPRs or cooperatives. This has proved difficult for some NGOs that lack the funds or expertise to become BPRs or are reluctant to adopt the cooperatives model.21 Virtually no data on NGOs were available for this report. However, there are anecdotal bits of information on encouraging progress, including a major international foundation assisting with the purchase of a weak local bank, and its conversion into a wholesale bank (Bank Andara) for the financing of BPRs and cooperatives. Another, longer-standing example is the Rabobank Foundation, which focuses on the development of small cooperatives in rural regions.22 23 These trends should be encouraged and accelerated where ever possible, because of the depth of NGOs’ reach into the poorer strata of society. Box 4: BRI’s Unit Desa System

Bank Rakyat Indonesia (BRI) runs one of the largest micro-credit programs in the world with 3.52 million customers as of December 2007 versus 3.1 million at end-2003. Its genealogy is complicated owing to several name changes, but it can be traced back to1895, with the establishment of a small association to manage the funds of a local mosque in Central Java. In recent years, micro loans have comprised around 30% of BRI’s total loan portfolio, compared with 50% for small credits; medium (6%); and corporate lending (14%). As a matter of BRI corporate policy, MSME lending should be at least 80% of BRI’s total portfolio. The various programs are run out of some 5,200 branches, the vast majority of which comprise the ‘Unit Desa’ system, which manages the rural finance network. It reaches deeply into Indonesia’s villages (see Chapter III). BRI’s main savings product, Simpedes (Simpanan Pedesaan), is offered through the Unit Desa system, and is aimed at absorbing funds from middle-low income people, mainly in rural areas. During the 5 years ending 2007, Simpedes grew at more than 18% per annum, and it generates almost 2/3rds of BRI’s total deposits. It’s main features are: a low opening deposit (Rp100,000, without an ATM card); low follow-on deposits (Rp10,000); a low monthly admin fee for small deposits (Rp2,000); and regular lotteries based upon minimum monthly balance. As of April 2009, Simpedes paid 2-4% interest, depending upon the size of account. BRI’s premier micro loan product is Kupedes (Kredit Umum Pedesaan), which provides loans of up to Rp100 million through the Unit Desa network. Kupedes’ clients are individuals who are economically active poor and lower-middle class. Typically, they are small traders in rural and urban areas close to BRI units, borrowing to finance working capital, fixed assets and other purposes. During the past 5 years, loan growth has averaged more than 22% per year. By way of comparison, small loan growth has also averaged 22%; medium-scale loans 40%; and corporate lending almost 30%. Kupedes’ interest rates are currently around 1½% per month; in general, lending is against income and character, not collateral. Other features include convenient Unit locations; simple, quick procedures; flexible terms; and access to larger loans for borrowers who repay on time. Sources: BRI Annual Report for 2007; http://www.bwtp.org/arcm/; and field interviews.

21 See http://www.bwtp.org/arcm/ 22 See http://rabobank.com/content/about_us/rabobank_foundation/ 23 For other information on NGOs operating in financial services, see http://www.bwtp.org/arcm/.

Conference Edition

38

II.7 Sharia Finance & Banking

Sharia finance covers various forms of financing arrangements that are consistent with Islamic principles. At present, it holds a very small market share (roughly 2-3%) but it is much higher for MSMEs (about 10%)24 and it is growing very rapidly. On the basis of the available information, it’s difficult to determine if this expansion is coming at the expense of other institutions, or if new customers are being attracted to financial services for the first time. Both possibilities seem equally plausible in the local context. Still, it doesn’t really matter; the important point is that a very popular, new service is on offer; at a minimum, this puts pressure on existing providers to improve service and reduce prices. In recent years, Sharia products have expanded to cover all the major financial services, including: conventional banking; BPRs; money markets; the stock market; bonds; mutual funds; cooperatives; and insurance. Even the state-owned pawnbroker has a Sharia arm. Furthermore, the original Sharia commercial bank in Indonesia, Bank Muamalat, is a leading innovator in the outreach of financial services (see Box 6). For its part, Bank Indonesia has been supporting Sharia market development through its ‘Accelerated Islamic Banking Development Program’, including its ‘iB Campaign 2008’.25 As of end-2008, there were 5 commercial banks operating on Sharia principles and 27 conventional banks operating with a Sharia business unit (Table 5). Adding in rural banks and their service outlets, the total number of Sharia service outlets reached 2658 by end-2007, almost 6 times as many as 3 years previously. Their outreach has been extended to more than 70 regencies and municipalities in 31 provinces. Deposits and financing have been increasing at rates of around 45% per annum in the past 5 years, compared with some 60% during the previous decade and a half. Non-performing financing is running at a very respectable 4% at end-2007, but it is trending upwards (Chart 9.6 in Bank Indonesia 2007).

Table 5: Sharia Banking Offices

(number of banks, etc.) 2003 2004 2005 2006 2007 2008

Sharia Commercial Banks (SCB) 2 3 3 3 3 5 Sharia Business Units (SBU) 8 15 19 20 26 27 Total SCB & SBU network 253 355 458 531 597 790 Sharia Rural Banks 84 88 92 105 114 131 Sharia Service Outlets … … … 456 1,195 2,658 Source: Bank Indonesia

24 See p.13 of Bank Indonesia (2005). 25 This is essentially a public information and training program. See Chapter 9 of Bank Indonesia (2007), p. 146-147.

Conference Edition

39

For purposes of this report, it’s important to note that these institutions cater mainly to the smaller end of the market. For example, 70% of their total financing is to MSMEs, and the vast majority (almost 98%) of their customer accounts are held by individuals, the remainder by corporations (Bank Indonesia, 2007, p. 135 and 137). As for Sharia rural banks, which offer services almost exclusively to lower-income groups, their numbers have been increasing steadily with their funds mobilization and disbursement expanding at around 35% during 2007 (Tables 5 and 6). Their financing to deposit ratio is very high (over 120%) and their non-performing financing—which had been around 10% in 2005—is down to 8% (Table 6). Despite this rapid expansion in lending, their level of non-performing financing compares unfavourably with the conventional banks, which as mentioned previously have NPLs of around 4%.

Table 6: Sharia Rural Bank Performance Indicators

2005 2006 2007 chng 2007 Total Assets (Rp, billions) 605.0 906.3 1207.2 33.2% Depositor Funds (Rp, billions) 353.6 530.2 711.3 34.2% Financing Disbursed (Rp, billions) 435.9 636.3 879.7 38.3% Ratios: Financing/Deposits 123.3% 120.0% 123.7% 3.7% Non-performing financing (gross) 10.6% 8.3% 8.0% -0.3% Non-performing financing (net) 9.5% 7.1% 6.6% -0.5% Source: Bank Indonesia

Box 5: Wider Access to Financial Services through Sharia Mobile Banking

Bank Muamalat is Indonesia’s first Sharia bank, established in 1992. It is a leading innovator in development of financial products that reach out to poor, remote regions through the use of mobile banking. In January 2005, it launched Shar-e, a full banking services card that Bank Muamalat promotes as “the first investment card”. It is unique in that the card has no limit in its wallet size; it can be used as a savings account; and it functions as an ATM and debit card. Essentially, it is a “virtual bank account”. As of 2007, the number of Shar-e card holders reached 1.2 million with average deposit per customer of around Rp 700,000 (US$70). Plans include an additional 700,000 customers in 2008. Fifty percent of Shar-e card holders are active customers and they perform 15 transactions per month on average. Any single transaction above Rp90 million has to be reported to BI by the customer; otherwise, there is no limit on transactions per day. A potential Shar-e customer needs to go to a post office or a Shar-e agent, and complete a KYC form and present identification. It costs Rp125,000 (US$12.5) to open an account and the card holder will have an instant Rp100,000 (US$10) deposit. The Shar-e card holder needs to activate his/her account by contacting the Muamalat call center in order for the card to be used in ATMs. Shar-e card holders are expected to conduct transactions via phone banking, post offices, retail agents, or ATMs; rarely do they need to go to a Bank Muamalat branch. In 2007, Bank Muamalat worked with 1,800 post offices throughout the country and plans to add 500 more post offices next year. In total, there are 5,000 post offices in all over Indonesia, so they still have room to grow to reach customers in remote places. They struck an exclusive deal with the post office that basically does not allow other Sharia banks to work with post office. In addition to providing revenue to the post office, Bank Muamalat also provides computer terminals in each post office and trains postal workers in banking transactions.

Conference Edition

40

The Shar-e card can be used at any of the 5 ATM networks in the country, including BCA’s Prima network, which totalled some 6,000 units at end-2008 (roughly 30% of all ATMs in Indonesia; Table 3). Bank Muamalat will bear the cost of a balance inquiry. For other transactions (such as purchasing or bill payment) the card holder is charged between Rp2,000-4,000 (20-40 US cents). Phone banking, not mobile banking, is also provided for the cardholder to activate the account, check the balance, and transfer funds to other card holders. Among Shar-e card holders, 30% live in rural areas and 70% in cities. To promote Shar-e and educate its customers, Bank Muamalat uses TV ads and billboards. Also, representatives visit Islamic schools to promote savings and use of the card. It’s notable that many Shar-e card holders are not Moslems, for example, in Papua and North Sulawesi. Users cited convenience and access through many ATMs as the reason for using Shar-e. Bank Muamalat plans to extend its networks beyond the post office (e.g., to Fuji Image Plaza, a chain of film processing kiosks) as points for selling and transactions for Shar-e. Also, they already work with a switcher from Malaysia, so the customer can use a Shar-e card in 2000 ATMs in Malaysia. On remittances, they have discussed the possibility of allowing Indonesia migrant workers to send money home from Malaysia, Hong Kong and Saudi Arabia. Shar-e card is probably one of the success stories in Indonesian branchless banking to date, but Bank Muamalat appears to be cautious in extending further mobile banking capabilities to its customers as a more convenient and safer way to conduct transactions. They already use some features of mobile banking such as checking the balance, but no more than that. They plan to use mobile phone to allow people to top-up or deposit money conveniently. Source: IFC, Jakarta.

This sector is currently expanding very rapidly, and it’s difficult to imagine policy prescriptions that could further accelerate this expansion in a healthy way. Indeed, the risk seems to be on the other side, namely that expansion is currently too rapid to be sustained. In this regard, there appear to be two main risks to this sub-sector, which is very important to lower-end clients. First, the rapid gains in market share may be masking problems that will only come to light when the growth in assets and liabilities eventually stabilize closer to industry averages. In particular, relatively high NPLs is bothersome, especially since some loan restructuring was necessary26 to bring them down to current levels, which are still on the high side.27 The second risk is that the current political popularity of Sharia financing will become an obstruction for strong, effective regulation and supervision. 28 Bank Indonesia’s efforts to date (see Bank Indonesia 2007, p. 138-139) mainly emphasize industry promotion and expansion with virtually no reference to prudential compliance.

26 See Bank Indonesia 2007, p. 137. 27 See Bank Indonesia 2007, p. 137. 28 For a Malaysian view on the regulatory situation of Islamic financial markets in the Persian Gulf region, see http://www.reuters.com/article/IslamicBankingandFinance09/idUSTRE53EFE20090416

Conference Edition

41

II.8 Non-Bank Financial Institutions (NBFIs)29 NBFIs are not yet playing a big role in enhancing access to finance for the low income and poor segments. Indonesia’s non-bank financial sector is small, roughly 1/5 of the total financial system, and it’s much smaller in relative terms than in several other large developing countries and many countries in the East Asia region (World Bank (2006b), and Figure 4). At the present time, policy aims at weeding-out weaker finance and insurance companies through the instruments of minimum capital and capital adequacy. As can be seen from experience with commercial banks, this is not necessarily inconsistent with a deeper reach into the poorer strata of society. However, it does limit the effectiveness of an important instrument for greater access, namely, open entry. Until this phase passes, most of Indonesia’s NBFI sector (especially, pensions, insurance and finance companies) seem likely to continue to offer products that are mainly relevant for larger corporates and the better-off segments of society. The important exceptions are leasing companies and micro-insurance. The leasing/multi-finance industry is important for some MSMEs, in areas like construction. Its main distinguishing characteristics for the client are: simple collateral arrangements; availability of medium- to long-term financing to purchase equipment; more financing (relative to cost of the equipment) than provided by banks; flexible contracting arrangements; and tax incentives. For the multi-finance company, collateral is easier to repossess, despite being movable;30 capital requirements are lower; and there is less supervision than for banks. Presently, the Indonesian leasing industry is heavily dependent upon banks for its financing, and many finance companies have joint financing arrangement with banks. While leasing firms do reach the lower income segments through products such as motorcycle financing, As regards the insurance business, significant inroads have been made regarding micro-insurance.31 In 2004, GTZ undertook a public-private partnership on micro-insurance with German insurer Allianz, supported by the United Nations Development Program (UNDP). The partners aimed at developing a range of demand-based private micro-insurance products on a model basis. The resulting study concluded that the legal framework for insurance needed to be strengthened; capacity development was needed for insurance agents; and market education was necessary. 32 Results included a regulation that allows BPRs to sell insurance as agents of established insurers and a credit life insurance with additional benefits covering the loan client, named ‘Payung Keluarga’ (Family Umbrella).

After one year of pilot testing, over 30,000 policies were sold and the number of active policyholders was over 15,000. GTZ and Allianz were satisfied that micro-insurance can be a sustainable business segment for commercial insurers. They identified the prerequisites for 29 This Section draws heavily upon World Bank (2006b), using that report’s definition of NBFIs, which excludes MFIs. 30 Technological advances have helped, too. GPS monitors can now be secured inside the equipment, which allows the leasing company to monitor continuously the exact geographical location of the equipment. 31 This paragraph is drawn from GTZ (2007). 32 See http://www.microfinancegateway.org/content/article/detail/36433

Conference Edition

42

sustainability as: i) sound product design; ii) a conservative premium calculation; iii) a clear definition of benefits; and iv) efficient operations. ProFI continues to monitor regulatory developments and to lobby for a conducive regulatory framework for micro-insurance in Indonesia.

Since its launch, Allianz’s micro-insurance business has grown rapidly. In 2008, the number of policies quadrupled relative to 2007, rising to more than 178,000. As of 2008, Allianz had been cooperating with 21 MFIs, consisting of 4 cooperatives, 11 social foundations, 4 rural banks and 2 commercial banks. The main regions penetrated to date include: Sumatra; Jakarta; West and Central Java; Bali; Sulawesi; and Sumba Island. For 2009, Allianz expects at least 200,000 new customers.33

In a related development, Munich Re-insurance Company (the world’s largest re-insurance company) has teamed up with Indonesian insurer Asuransi Wahana Tata to bring innovative flood micro-insurance to Indonesia. The product, ‘Alert 1 Manggarai Protection Card‘, offers index-based flood insurance cheaply and easily to low income families in Jakarta.34 The new product began as a feasibility study between Munich Re (as partner and sole reinsurer), Asuransi Wahana Tata and GTZ. In mid-2007, a pilot was undertaken in 23 sub-districts of Jakarta.

II.9 Financial Services in Support of Remittances

As described in Chapter V, trade in labor services is potentially very important in the alleviation of poverty via many channels. On the international side, large numbers of Indonesians officially work overseas (Tenaga Kerja Indonesia, TKI; Table 7), with the total estimated at about 4.3 million (World Bank, 2008). This represents about 4% of Indonesia’s total work force, and many more work illegally. These workers send considerable amounts of money home (that is, remittances),

making Indonesia the second largest (after the Philippines) remittance receiver in East Asia and the Pacific. These remittances have a direct impact on the reduction of poverty in Indonesia insofar as they boost local consumption and investment. The extent of the impact is, however, in doubt. World Bank (2008a) argues that much of remittances are spent on repaying debt, and on consumption and luxury goods.

33 See www.allianz.co.id. 34 The product offers cheap and easily understood risk protection, specially designed for the local population. There is no lengthy policy document; rather, the insured receives a simple protection card. One card costs Rp50,000 and guarantees a one-time payment of Rp250,000 should the waters rise above 950cm (Alert 1) at the Manggarai Water Gate in Jakarta.

Conference Edition

43

Table 7: Official Overseas Worker Placement, 2004

Region/Country of Placement Formal Informal Total Male Female Male Female Asia-Pacific 68,022 47,284 497 45,167 160,970 o/w Malaysia 62,254 43,179 404 21,338 127,175 Hong Kong - - 2 14,181 14,183 Singapore 8 30 25 9,068 9,131 Brunei Darussalam 2,481 3,458 34 530 6,503 Middle East & Africa 609 312 14,930 203,848 219,699 o/w Saudi Arabia 432 191 13,724 189,099 203,446 Kuwait 29 9 1,206 14,745 15,989 Europe & America 17 3 - 1 21 Total 68,648 47,599 15,427 249,016 380,690 Source: Ministry of Manpower and Transmigration, cited in Manning and Cronin (2008)

Despite the large numbers working abroad, amounts of international remittances to Indonesia are small relative to international comparators (see Figure 12). Indeed, in relation to GDP Indonesia is nowhere near the top 20, using recent indicators. This raises an issue as to whether Indonesia’s financial infrastructure is adequately developed to support transfers from abroad, and whether there are barriers that can be removed through policy changes. This Section looks at these particular issues, which are essentially a subset of the broader issues of access to financial services for migrant workers (see Chapter V).

Conference Edition

44

Figure 12: Top 20 Developing-country Recipients of Remittances, 2005

Source: UNCTAD Handbook of Statistics, 2006-07, cited in Manning and Cronin (2008) For international remittances, this report relies upon a recent World Bank report, on these and other issues for Indonesians working in Malaysia (World Bank, (June 2008a)). As noted in Table 7, Malaysia is the second most important destination for Indonesians working abroad. Apart from geographic proximity, the issues and policy conclusions relevant for Malaysia are believed to be applicable for other major destinations, like in the Middle East and elsewhere in Asia. Consequently, the interested reader is referred to that document for a detailed review and analysis.

(In billions of US$)

0

5

10

15

20

25

India

Mex

ico

Philipp

ines

China

Leban

on

Mor

occo

Roman

iaEgy

pt

Pakist

an

Bangla

desh

Colombia

El Salv

ador

Brazil

Domini

can R

epub

lic

Russia

Indon

esia

Bosnia

Hondu

ras

Peru

Albania

Bill

ions

of U

S$

(As per cent of GDP)

0

5

10

15

20

25

30

35

Mold

ova

Leban

on

Hondu

ras

Bosnia

Tajiki

stan

El Salv

ador

Albania

Philipp

ines

Nicara

gua

Domini

can R

epub

lic

Mor

occo

Yemen

Armen

ia

Ugand

a

Bangla

desh

Georg

ia

Azerb

aijan

Roman

iaEgy

pt

Bulga

ria

(In

done

sia)

As

% o

f GD

P

Conference Edition

45

Briefly here, the Malaysian-Indonesian study argues that the formal financial sector in its current state has little to offer the migrant worker. For example, it is uncompetitive in terms of accessibility and instruments, and it offers little by way of financial literacy training and pre-departure credit, especially for illegal workers. This poor service accounts for the large number of informal agents, many of whom exploit migrant workers for their own purposes. The report concludes (p. 49) that migrants get caught in cycles of continuous migration as their families become dependent upon the remittances. Although these abuses at home dilute the value of having a family member working abroad, it’s hard to imagine that the family is financially worse off with a member working abroad. In discussing this issue with Indonesian bankers, credit risk is the main issue particularly if the worker is overseas. For commercial banks in the personal loan business, they would consider a loan co-signer in Indonesia (probably a family member in the home village),35 but the co-signer’s creditworthiness is the principal consideration, not the potential income flow from the migrant worker. In discussing remittances, a major Indonesian bank (with representative offices in Hong Kong and Kuala Lumpur) said that they value the remittance business of migrant workers (because of the fee income). The migrant worker can drop-off cash at the rep office and it is delivered immediately to any of their many branches in Indonesia; the recipient does not need to hold an account in that bank. That bank’s staff believe that they are very competitively priced, at least in Hong Kong. For its part, Bank BRI reports large incoming and outgoing remittance transfers with total flows increasing by more than 65% in 2007.36 On the side of domestic migrant workers, studies have noted the flexibility of Indonesia’s labor markets,37 and its inter-island and rural-urban mobility is high. Data on the volume of domestic transfers are difficult to track down,38 but they hardly seem necessary to demonstrate the importance of cheap, convenient methods for transferring money—perhaps in relatively small amounts—from workers in the cities to their families back in the villages. Data examined during this study indicate that good options exist for making domestic remittances (Table 8). If the sender values convenience and speed, Western Union is the best option, including because it has the most outlets among service providers in Indonesia. Indeed, its network is roughly twice as large as that of BRI. However, Western Union is expensive for small transfers costing more than 15% for amounts less than Rp¾ million. The cost drops off sharply as the size of transfer rises, but it still costs about 5% to transfer, say, Rp37 million.

35 Field reports indicate that BRI is conducting a pilot project for credit to migrant workers, to be co-signed by family members. Details could not be confirmed with BRI, but the development sounds encouraging and further collaboration could be explored by development partners with special interests in this area. 36 See Bank BRI Annual Report for 2007, p. 62. 37 See, for example, Manning (2000). 38 Work related to this area includes a study on internal migration, pending from Australian National University.

Conference Edition

46

Table 8 : Cost of Domestic Remittances, April 2009

Service Provider Cost Notes

BRI (to BRI) Free within Jabotabek. Excludes Unit Desas in very remote locations (i.e. off-line). Rp2,000 for <5,000,000. Cost rises to Rp20,000 for Unit Desas not on-line. Rp3,000 for >5,000,000. Likely takes 3 days or more to reach off-line Unit Desa.

Small additional charge if above Rp 50,000,000. Receiver must have a BRI account.

BRI (to other bank) Rp25,000 if <12 PM. Cost rises to Rp100,000, if 2-3 PM. Money is transferred through the regular bank clearing system.

Western Union Rp112,500 for < Rp750,000. Instantaneous delivery to all 10,000 locations in Indonesia. Costs rise in 18 steps, to Rp6,000 stamp duty paid by receiver if receiving outlet Rp1,750,000 for transfers

from Rp37,500,001 to Rp41,250,000.

is physically located in a bank.

Larger amounts can be sent, but costs are not published locally. Money Gram Does not remit domestically. Post Office Rp11,000 for <200,000. Sender must return subsequently, if sending from a small post office. Cost rises to Rp21,000 if Cost rises by Rp2,000 to 12,000 for 'Instant' service, which Rp10 -25 million. is only available in large post offices. Recipient may live some distance from the local post office.

For regular remittances, special arrangements can be made for local delivery from large offices, but courier requires a 'tip'.

Money order is electronically, not by regular post. Source: World Bank field research.

There are other very cheap, fast options, if the recipient has a BRI account,39 or if the recipient lives near a large post office. BRI’s cost is extremely low (0.5% or so) and service is fast, if the Unit Desa branch is on-line. The post office service is also cheap, but more expensive than BRI (see Table 8). If the recipient lives in a very remote location, the options are BRI’s Unit Desa or the post office (if the recipient does not have a bank account). Financial costs are still low, but the wait-time increases, probably to 3 days or more and there may be additional, informal delivery charges (Table 8).

39 BRI offers transmission to branches of other banks, still at very low cost (roughly ½ of 1%) in the mornings. However, the cost rises sharply in the course of the clearing day (see Table 8).

Conference Edition

47

II.10 Summary of Policy Issues, by Area

This section summarizes the policy issues arising from the discussion of this Chapter. Policy recommendations are provided in Chapter VI.

Analytics of the Issue. The analytical evidence of this Chapter indicates that per capita income and population (or land mass) go a long way towards explaining the geographical reach of Indonesia’s commercial banking system, on a provincial basis. The only notable exceptions are Jakarta (which is ‘over-serviced’, being the political, commercial and financial capital) and East Kalimantan (which is large, resource rich, sparsely populated and ‘under-serviced’). Financial services in all other provinces are roughly in line with their levels of per capita income. In order to identify the ‘under-serviced’ regions, it is necessary to investigate below the provincial level, an issue that is taken up in the next Chapter. Conventional Commercial Banks. Indonesia’s commercial banks are opportunistic, profit-oriented institutions that will move aggressively into new, commercially viable markets. The evidence of this Chapter indicates that they have quite a wide regional reach, but they do not reach deeply into the poorer strata of Indonesian society. Virtually all of them seem to aim at the more lucrative middle income and higher segments. This issue is pursued in the next Chapter. This said, the commercial banks have two important roles to play in extending wider access to financial services. First, as innovators, the banks are the most likely institutions to introduce new technologies to deliver relatively high-cost services to low-income clients in costly, remote areas. Recent technological advances (like mobile banking, discussed later in this report) look to be a key in this regard; technology has the potential to get unit costs down to the point where services can be extended at affordable prices to the poorer strata of society even in remote regions of the country. Second, the banks will be helpful in extending access because they will put competitive pressures on the other service providers, thereby holding down prices and improving the quality of services. BRI’s Unit Desa. This is Indonesia’s premier micro-finance provider, and its client base is among the largest in the world. Nonetheless, there are questions as to whether the Unit Desa reaches deeply enough to service lower income clients. To extend its current reach—at least until lower-cost technologies are available—a policy decision on the part of bank management (and shareholders) may be needed as regards the current fee structure and lending policies. Higher interest rates on its major lending products to micro borrowers may be needed make lower-level lending commercially attractive. Alternatively, there could be additional administration fees, with the size of the admin fee roughly proportional to the additional costs of micro loan supervision. People’s Credit Banks (BPRs). These institutions offer considerable opportunities for wider access to finance for poorer households and MSMEs, essentially because they are relatively low-cost operations and because they have better local knowledge than regular commercial banks. In particular, there are significant opportunities on the regulatory front, a matter taken up in Chapter

Conference Edition

48

IV. In support, continued upgrades will be needed in Bank Indonesia resources and institutional capacity to protect the public against unviable banks, whether BPRs or regular commercial banks. Recent developments have been encouraging. For example, the 15 BPR failures since 2004 have been handled without any marked loss of public confidence. Also, by way of encouraging developments, Bank Indonesia seems to be considering changing its regulatory framework from one-size-fits-all-BPRs to a 3-tier approach based upon size of BPR. BI’s new approach looks to have merit and this report makes further suggestions in Chapter V. Other Micro Finance. The most productive way forward in this diverse area concerns progress on a new Micro-Finance Law, which has been in draft for many years. It would be helpful to restore momentum to the drafting process, in a direction that is supportive of MFI development. Also, there are several government programs that offer subsidized credit to special groups in the Micro-finance segment, and there is significant room for improvement in many of these programs. Regarding other issues this sub-sector, NGOs offer an important range of services at the lowest income levels, often in the country’s most remote regions. Since Indonesia’s shift to democracy, there has been considerable expansion of NGO services, but much more looks possible. Possibilities are explored in Chapter IV, together with steps to strengthen cooperatives and pawnshops. Sharia Banking. This market segment, although still small, is very important for issues of access to financial services in Indonesia. It has been expanding rapidly for the past decade or so; Sharia rural banks offer their services almost exclusively to the poor; and a major Sharia bank is leading technological innovations that extend financial services to the poor and remote areas. The key issue for policymakers is how to continue progress at a sustainable pace. Two issues arise in this context. First, rapid expansion may be masking serious problems (e.g., relatively high NPLs) that will only come to light when the growth in assets and liabilities eventually stabilize closer to industry averages. And second, the current political popularity of Sharia financing might undermine strong, effective regulation and supervision. Non-bank Financial Institutions. With some notable exceptions (e.g., leasing and micro-insurance), the majority of NBFIs currently offer products of limited relevance to poor households and MSMEs. Moving beyond traditional product lines will require innovative thinking, like that currently underway in micro-flood insurance and micro-life insurance. Public-private partnerships (like that by Alliance AG and BI supported by GTZ and the UNDP) offer promise in this regard, to be buttressed by a stronger legal framework; capacity development for insurance agents; and market education. Remittances. This report finds no evidence of any serious problems in this area. There are several good options available, depending upon the circumstances and interests of sender and receiver. Even for the most remote locations, there are good cheap options that simply entail a delay

Conference Edition

49

of a few days or more. Issues concerning international remittances are discussed at the end of Chapter V, in conjunction with more general issues associated with migrant workers.

Conference Edition

50

Chapter III: Demand-side Aspects; What do People Want?

III.1 Introduction and Overview

Surveys have become a preferred tool for analysis of issues concerning access to finance (see Box 7). Following this approach, the entire analysis of this Chapter is derived from two, nationally representative surveys on Access to Finance (A2F) in Indonesia. The first, large survey, of households, is summarized in Box 8 and its technical details are described in Annex A (including considerations regarding the choice between a household survey versus an individual survey). The second survey, of village heads, is much shorter and complements the material of Chapter II. The household survey was designed with several important goals in mind. First, we are interested in identifying and understanding the current sources of financial access for Indonesian households. This includes all institutional sources of access (e.g., banks, non-bank finance companies, micro-finance institutions, insurance companies, informal money lenders, community based welfare schemes and savings associations, family and friends) as well as all types of access. On the savings side, this includes everything from bank accounts to savings in community welfare schemes; on the borrowing side, from bank loans to borrowing from moneylenders. As a second goal, we are interested in identifying the existing constraints to accessing financial services. On the demand-side, these are income barriers, cost, financial literacy and so forth. Innovative public private partnerships, with collaboration of development partners if necessary, are required to identify the major, on-going supply-side constraints, and to find solutions that can be pilot-tested in the market.

Conference Edition

51

A third goal is to identify the potential, unmet demand for financial services. Specifically, this means measuring the scope of unmet demand for certain financial products (e.g., savings, loans, insurance, retirement plans, etc.) that households would like financial institutions to offer. The survey dedicates a significant number of questions to this goal, including a look at related pricing issues. Finally, looking further ahead another goal is to integrate the A2F Survey with existing household level socio-economic surveys in Indonesia, such as SUSENAS. The end result would be a continuous stream of annual data on financial access in Indonesia, which would allow researchers to track financial access indicators for the country on an on-going basis. By doing this, access to

Box 6: What Do Household Surveys in Other Countries Tell Us?

Access to financial services is increasingly recognized as critically important to alleviating poverty around the world. Banking and financial systems in developing economies are typically underdeveloped, and often cater mainly to large firms and wealthy individuals. This skewed distribution of finance hinders the growth and development of smaller firms and poorer households. Consequently, wider availability of financial services is a policy goal being advocated by development institutions worldwide.

A necessary step in improving financial access for households is to understand the constraints they currently face in accessing formal financial services. Household surveys are a useful tool in addressing this issue, and they have been conducted by the World Bank in several other countries, including Brazil, Colombia, India, Mexico, Nigeria and Pakistan. Data from these surveys indicate that banking services vary widely across countries. Population per bank branch ranges from 9,331 in Brazil to 14,888 in India, whereas the area covered per branch ranges from 470 square kilometres in Brazil to 44 square kilometres in India. Among survey respondents, 48% have bank accounts in India, 43% in Brazil, 41% in Colombia, 25% in Mexico, and only 13% in Pakistan. (By way of comparison, almost 90% of households in the United States have bank accounts). A majority of survey respondents identify high bank fees as the main constraint inhibiting their access, for instance 70% in Mexico. Other factors, such as lack of documentation and convenient bank location, are also hurdles to access. As for the financial characteristics of survey participants, more than 50% of respondents, on average, report no financial savings of any kind, while about 30% report using informal saving mechanisms, such as community savings associations. Among people who do use formal financial services, bank accounts are dominant with 96% use in Mexico; 95% in Brazil; 90% in India; and 85% in Colombia. Other formal institutions, such as Cooperatives and Post Offices, are much less popular. On the credit side, loan rejection rates are quite high, except for Pakistan. Out of all loan applicants, 10% are rejected in Pakistan, compared with 32% in Brazil, 26% in Colombia, 20% in Mexico, and an overwhelming 80% in India. According to applicants’ perceptions, 56% of respondents in Brazil feel that commercial banks reject their loan request due to a lack of earnings or lack of steady income. By contrast, 44% in Colombia and 41% in Mexico believe that it is the lack of loan guarantees or collateral. These data collection efforts at the household level are a useful step towards identifying key constraints faced by households in accessing financial services, and designing policies to reduce those constraints. Source: Various World Bank reports.

Conference Edition

52

finance issues and their links with poverty alleviation can be monitored on an ongoing basis and form the foundation for policy action. Box 7: What Do Other Surveys for Indonesia Tell Us?

Several other surveys in Indonesia touch upon access to finance issues.

BRI’s 2003 MASS Survey (see Box 3) is the most extensive alternative survey on access to financing in Indonesia. It estimates that less than less than 50% of rural households have access to banking services, and provides other important details that are summarized in Box 3.

Other studies on this topic are more specialized (usually concerning MSMEs’ access to credit) and they tend to yield information on access to financing as a by-product of other analysis. For example, a 2003 study by the Banking With The Poor network (see http://www.bwtp.org/arcm/Indonesia/) cites several secondary sources indicating that a majority of rural households do not have access to semi-formal or formal institutions as a source of funds. The same study asserts that less than ¼ of microenterprises have access to credit from a formal institution.

Bank Indonesia’s 2005 survey of SMEs showed that around 71% of micro business capital is raised through self-financing and 29% borrowed from friends, business partners and families. More details on BI’s survey are provided in Chapter V.

The World Bank’s Rural Investment Climate Assessment (RICA; 2006) devoted an entire chapter to constraints on accessing credit by rural non-farm enterprises. It found that access to credit was the second most important constraint for its clientele of 15.7 million micro and small enterprises in rural Indonesia. The interest rate they have to pay is a secondary concern. It recommends against subsidies to make borrowing cheaper, and encourages incentives for extending the reach of the commercial banking sector to connect with rural household enterprises. The RICA survey indicated that borrowers believed their major constraints to be: collateral requirements; the perceived complexity and cost of application procedures; and the legal standing of the enterprise in question (particularly the high costs of formal registration of a business required by formal lending institutions). For their part, banks feel constrained by the cost of information that allows them to identify which micro and small enterprises are good credit risks.

Conference Edition

53

Box 8: Sampling Methodology of the Indonesian Survey The sample selection methodology was designed to ensure national representation. Multi-stage random sampling was employed with population-weighted selection occurring first at the province level, next at the sub-province (Kabupaten/Kota) level, and finally at the village (Desa/Kelurahan) level. Overall, 4 provinces on Java and 6 provinces off Java were selected, namely Banten, West Java, Central Java, East Java (all provinces on Java); and Aceh, Jambi, West Kalimantan, North Sulawesi, Maluku, and Nusa Tenggara Barat (off Java provinces). On Java, 16 population-weighted sub-provinces (excluding Jakarta and Yogyakarta) were chosen at random. Within each of these 16 sub-provinces, 4 population-weighted villages were chosen at random, resulting in a total sample of 64 villages. The selection was stratified by urban/rural, with the final sample containing 34 urban villages and 30 rural villages. Off Java, the country was divided into 6 sub-regions, and one (population-weighted) province was chosen at random from each sub-region. Next, within each of these provinces, 2 population-weighted sub-provinces were chosen at random. Finally, within each of these 12 sub-provinces, 4 population-weighted villages were chosen at random, resulting in a sample of 48 villages. The urban/rural stratification resulted in a final sample of 18 urban villages and 30 rural villages, owing to the overwhelming rural nature of off-Java villages. Within each of the 112 villages in the survey, 30 households were randomly selected in each village for survey interviews, resulting in our final sample of 3,360 survey household respondents (1,920 on Java and 1,440 off Java). The interviews were conducted between July and December 2007.

Respondents’ characteristics are provided in Annex A under “Some Characteristics of Access to Finance Sample Respondents”.

Access to Finance Province Distribution Map (2007)

Given these multiple goals, the survey questionnaire is lengthy (some 50 pages) and detailed. In addition to financial access information, the survey contains information on household demographics, household financial preferences and special modules on financial literacy, household

Conference Edition

54

enterprises and households of migrant workers. As mentioned, there is also the much smaller (8 page) village head survey which was designed to capture some key supply side characteristics of every sample village, such as available infrastructure and number of financial service providers. The survey data is available on the World Bank’s Indonesia Office website.

III.2 Survey Results: Supply of Financial Services

As a complement to the preceding Chapter, this Sub-Section looks at the supply of financial services as indicated by the Access to Finance Survey. Results are based upon the survey of village heads, mentioned above. In terms of supply of financial services in Indonesia (Figure 13), Bank Rakyat Indonesia (BRI) clearly has the widest coverage in the country, confirming the conclusion of Chapter II. Of the 112 villages in the survey, 29 have at least one commercial bank branch, of which 23 are BRI Unit Desas. Indeed, BRI has branches in areas where institutions like credit cooperatives and pawnshops are absent. More than 40% of the population that has a bank account, does business with BRI. The private banks and the non-BRI state banks are also well-represented at the village level. In total, they account for over 50% of bank accounts, compared with about 45% for BRI (Figure 13).

Figure 13: Bank Account Distribution, by Type of Bank

This coverage at the village level (29 out of 112 villages) is not particularly impressive. However, follow-up questions indicated that 109 of the 129 villages used commercial bank services, even if the physical infrastructure was not present in their village. Also, the vast majority of villages believe that bank branches are well-located; all but 6% of respondents report that bank branches are conveniently (or very conveniently) located (Figure 14).

44.5

31.0

20.5

2.8 0.2

1.0

BRI Unit

S tate  Bank  non BRIUnitP rivate  Bank

BPR

S yariah Bank

Other

Conference Edition

55

Figure 14: Bank Branch Locations

Not surprisingly, physical access varies as between rural and urban areas (Figure 15); more sparsely populated areas are served more sparsely by bank outlets. For rural areas, the average travel time to the nearest branch is greater than 25 minutes for almost 35% of the respondents, versus less than 12% of urban respondents. Indeed, for the urban population, a bank branch is located within 10 minutes traveling time for 62% of the urban population compared with only 31% of the rural population, precisely half the value for urban areas.

10.4%

83.5%

6.1%

Very Convenient ConvenientInconvenient

Conference Edition

56

0

10

20

30

40

Percent

<= 10 min 11-15 min 16-25 min > 25 min

RuralTime Average for Traveling to Nearest Bank Branch

0

20

40

60

Percent

<= 10 min 11-15 min 16-25 min > 25 min

UrbanTime Average for Traveling to Nearest Bank Branch

Figure 15: Travel Time to Nearest Bank Branch

Looking in more detail at average travel to bank branches (see Table 9), public transport is the most common method of travel, followed by bike/motorcycle and by foot; the use of a personal car is quite infrequent. There is relatively little difference in time requirements as between (bank) savers and non-savers (see the middle columns of Table 9). However, the rupiah cost for savers is generally considerably higher for savers than non-savers. Since savers actually use banking services while non-savers do not, this indicates that both groups have a good idea of where the physical outlets are located, although non-savers have less of an idea about how much it actually costs to get there.

Table 9: Travel to Banks, Time & Cost

Respondent Average time (minutes) Average cost (rupiah) (One way trip) Savers* Non-savers Savers Non-savers Savers Non-savers On foot 148 342 13 15 1,193 926 Bike/Motorcycle 777 722 18 20 2,586 3,119 Personal car 71 6 17 20 8,830 3,208 Public transportation 293 603 22 22 3,781 2,988 Water transportation 4 6 356 263 186,531 4,830 Other (mainly becak) 13 18 72 67 42,904 2,615 *Bank savers

These data are available with a breakdown as between urban/rural and Java/off-Java locations (Tables 10 and 11). For bank savers (Table 10), average times for urban residents are very similar as between on and off-Java. However, differences are striking for rural residents; rural savers on-Java require roughly 50% more time than urban residents whereas those off-Java require far more time (2-10 times), especially if water transport is involved. Rupiah costs also vary markedly. The rupiah cost is lowest for on-Java urban residents; it’s higher for off-Java urban residents; and far higher for those who live in rural areas. Again, the cost is particularly high when water transport is required.

Conference Edition

57

Table 10: Saver’s Travel to Banks, by Urban/Rural & Java/Off Java

Average time (minutes) Average cost (rupiah) Urban Rural Urban Rural (One way trip) Java Off-Java Java Off-Java Java Off-Java Java Off-JavaOn foot 11 11 15 171 225 850 1,426 93,326 Bike/Motorcycle 12 14 22 28 2,142 3,181 2,669 5,109 Personal car 11 12 24 36 5,907 3,976 14,070 9,272 Public transportation 15 18 26 109 2,583 3,836 3,365 27,636 Water transportation NA 24 NA 392 NA 402,577 NA 137,535 Other 21 24 21 392 27,314 402,577 3,517 137,535 For non-savers, the travel pattern is similar to that of savers, but with less pronounced variation between regions (Table 11). It’s also notable that the rupiah cost estimates by non-savers are much lower than savers’ costs in rural locations, particularly off-Java, suggesting that the actual transportation cost may be contributing to exclusion in those areas.

Table 11: Non-savers' Travel to Banks, Urban/Rural & Java/Off-Java

Average time (minutes) Average cost (rupiah) Urban Rural Urban Rural (One way trip) Java Off-Java Java Off-Java Java Off-Java Java Off-JavaOn foot 14 13 15 69 783 661 1,017 4,485 Bike/Motorcycle 11 9 20 40 1,941 2,990 2,967 6,813 Personal car 15 NA 16 31 3,000 NA 2,356 5,556 Public transportation 15 15 19 50 2,541 2,227 2,565 5,834 Water transportation NA 10 NA 277 NA 5,000 NA 4,805 Other 5 10 21 277 3,000 5,000 2,300 4,805 The various travel times noted above may seem excessive. However, by local standards the averages actually compare favourably to key public services (see Figure 16). For example,40 some 48% of Indonesians can reach a bank branch in less than 10 minutes versus 7% to a public hospital; about 28% to a Puskesmas (a basic health facility); and some 57% to a school. At the other end of the scale, only about 7% of Indonesians require more than 30 minutes to reach a bank branch, which is similar to a school, and far better than 33% for a Puskesmas and 69% for a public hospital. These data indicate only modest differences in physical access as between those who already have a bank account (savers) and those who do not (non-savers; see Figure 16). This comparison is important for policy purposes when allocations of public expenditures are involved. In particular, it suggests that public monies would be well spent on improving physical 40 World Bank, Governance and Decentralization Survey. See http://www.dsfindonesia.org/apps/dsfv2/cgi-bin/dw.cgi

Conference Edition

58

access to key public services (like basic health and education) – physical access to financial services would improve in line with such overall improvements. Other instruments, like regulation and competition policies, would be more appropriate for improving access to financial services.

Figure 16: Average Time to Reach Select Institutions

Once inside a bank, the average waiting time to be served is fairly short. More than a third of clients are served within 10 minutes of arrival, and less than 18% of clients in rural areas have to wait for more than 30 minutes to be served. Wait times are fairly uniform as between urban and rural areas (Figure 17).

47.8

23.7

21.9

6.5

44.1

26.5

20.7

8.7

27.6

16.7

22.6

33.1

6.6

6.8

17.8

68.8

56.3

16.7

20.4

6.6

0 20 40 60 80 100

Up to 10 min

> 10 to 15 min

> 15 to 30 min

more than 30 min

School

Public Hospital

Puskesmas

Bank (Non‐savers)

Bank (savers)

Conference Edition

59

Figure 17: Waiting Time to be Served in a Bank

To summarize, these results on bank branch location suggest that in urban areas physical presence of bank branches is not a constraint to financial access. Once inside a bank, neither is processing or waiting time a problem in either urban or rural areas. In rural areas where distances are longer and transportation more difficult, it takes substantially longer to get to bank branches and hence this may pose something of a supply-side impediment. Nevertheless, it is not a significant barrier based upon users’ assessments of convenience and the travel times compare favourably to key public services like health care and schools. These results on physical accessibility have implications for what follows and for policy. In subsequent sections, the demand analysis by product is presented for urban and rural areas separately and the differences are statistically tested. For policy, it’s important to know why access is low, even if physical accessibility is not a problem, and whether something can be done about it.

III.3 Survey Results: Demand for Types of Financial Services

This Chapter now turns to results of the household survey. Beginning at a broad level, the survey results indicate demand for a wide range of financial services (see Figure 18). Of these, the single most important financial service is a bank account; more than 40% of respondents have a bank account compared with only 15% who have a bank loan. Other types of credit services rank high, too, but they are mainly from informal sources (e.g., friends and family (25%) and local shops

17.8

30.5

18.3

33.4

12.4

22.3

26.6

38.8

0 10 20 30 40 Percentage

Rural

Urban

>30min >15-30min >10-15min

<= 10min

>30min >15-30min >10-15min

<= 10min

Urban vs Rural

Conference Edition

60

(20%)) and they are widely scattered by type of institution. ATM and debit cards are relatively popular (17%), but mobile banking and credit cards rank at the very bottom of the list.

Figure 18: Financial Products Used by Households

III.4.1 Survey Results: Demand for Savings Accounts

Much of the higher-level survey information on Indonesia’s savers is summarized in Figure 19 (for borrowers, see Figure 28 in Section III.5.1). Roughly 1/3 of the population in Indonesia does not save at all, formally or informally, and might be considered ‘financially excluded’. Considering the remaining 68% of households who do save, 50% do so at a formal savings institution and the remaining 18% save informally via community welfare schemes or informal savings clubs. Decomposing the 50% who save at a formal institution, only 47% of the population saves using formal banks and 3% use other formal institutions (such as cooperatives and credit unions). These data point to an important weakness of the Indonesia banking system in providing access. Namely, that less than half (47%) of the total population save in a bank, and the majority of

2.2

2.6

3.4

4.6

4.9

5.9

9.3

14.7

17.3

20.6

25.4

40.6

0 20 40 60 80 100

C redit C ard

M‐banking

Loan from P awn S hops

Loan from E mployer

Loan from Daily  Bank

C ommunity  Welfare P rogram

Loan from MF I

Loan from Bank

ATM/Debit C ard

On C redit from Local S hop

Loan from Neighbors , F amily, & F riend

Bank  Account

Conference Edition

61

these save informally, too (see the following paragraph). This is striking evidence of the limited extent to which Indonesia’s banks actually deliver services to the great bulk of the country’s population.

Figure 19: Summary of Survey Results for Savers

Total Population

68% Save 32% Don’t Save Main Reason: i) No money (79%)

50% at formal 18% at only informal ii) No job (9%) institutions institutions iii) Don’t see the benefit (4%)

iv) Don’t understand banks (3%)

47% at banks 3% at other formal

41% use own bank 6% use others’ accounts accounts

Looking in more detail at the 68% who are ‘financially served’, there is a great deal of overlap among their use of formal institutions (Figure 20). In particular, savers who use banks, actually make much greater use other formal and informal providers (30.1% versus 16.6%). Also, among ‘exclusive use’ savers, informal vehicles (i.e., community welfare or informal savings clubs) are the most important category (18.2% versus 16.6% for banks and 1.2% for other formal).

Conference Edition

62

Figure 20: Overlap Among Savings Providers

Continuing with those who are ‘financially served’, the Survey asked respondents why they saved, including by type of bank (see Figure 21). The dominant reason in most cases is ‘security’, which underscores the importance of prudential supervision of savings institutions and the maintenance of financial stability. It’s notable that this is particularly important for Sharia customers, suggesting that the rapid growth of this sector may reflect new clients entering into the formal banking, having been attracted by the perceived safety of Sharia banking. For most depositors (other than BPRs, which often have very limited transfer capacity), it’s also important for clients that they can transfer money and borrow money. Other reasons tend to be relatively unimportant. There is some regional variation in these responses. Most notably, urban populations save in formal institutions primarily for predicted future needs whereas rural populations want to gain access to loans, significantly more so than urban populations. Also there are some gender differences. For instance, men are more likely to have a bank account in order to obtain a formal loan, whereas women perceive future needs as being significantly more important.

16.6%

2.6%

20%

18.2%

1.2%

7 5%1.9%

Banks

Formal Other

Informal

31.9% No Account

Conference Edition

63

Figure 21: Reasons for Having a Bank Account, by Type of Bank

III.4.2 Barriers to Accessibility: Reasons for Savers Being Unbanked Turning to those who are ‘unbanked’ (that is, the right-hand branch of Figure 19), the Survey asked households who do not currently have bank accounts to identify the major constraints to having a bank account (see Figure 22). By far the largest constraint identified is the lack of money (almost 80%), which is an obvious explanation. The second most common reason was ‘Do not have a job’ (about 10%), which is highly correlated with not having any money. Other constraints are tiny in comparison, for example, ‘Don’t understand banks’ (3.1%) and ‘Do not see the advantage of having a bank account’ (3.6%). It is striking that the least important reasons for being unbanked was ‘Bank fees are too high’; ‘Distance to the bank’; and ‘Deposit rates are too low’. This confirms the point, noted above, that longer distances to destinations is considered normal in rural areas; they are not judged to be an inconvenience. This is probably because the opportunity cost is lower for people in rural areas, their incomes being lower, on average. Also, these results suggest that the great majority of bank savers are not interest rate sensitive.

0.0 20.0 40.0 60.0 80.0 100.0

S tate Bank

P rivate B ank

S yariah B ank

BP R

S E CUR ITY  RE AS ONS

TRANS F E R  MONE Y

FOR  P R E DIC TE D  FUTURE  NE E DS  

TO  AB LE  TO  BORROW  MONE Y

GAIN ACC E S S  TO  OTHE R  F INANC IALS E RV IC E S  

FOR  EME RGE NC Y  NE E DS  

BE ING  AS K E D  BY  EMP LOY E R

E ARN  INTE RE S T

OTHE R

S AVE  MONE Y  FOR  A  S P E C IF ICP URCHAS E

P AY  B ILLS  OR  DE BTS

P RE VE NT S POUS E  F ROM S P E NDINGMONE Y

TO  WIN LOTTE RY

P RE VE NT F R IE NDS /RE LATIVE S  F ROMAS K ING  FOR   IT

Conference Edition

64

Figure 22: Main Reasons for Not Having a Bank Account

(Previously banked and never banked)

The Survey also looked at reasons why people don’t open a bank account by income level (see Figure 23)41 with all the respondents stratified into ten income deciles. Across all income deciles, the dominant reason (75% or more) was ‘Don’t have enough money’ (see the discussion immediately below). There is more variation among the secondary reasons for not opening a bank account. For example, respondents in the lowest deciles cite ‘No job’ far more often than the middle- and upper-income deciles, indicating that their reasons are almost entirely economic. Among the middle- and upper-income deciles, the more common reasons (5-10%) are: ‘Don’t see the advantage’; and ‘Don’t understand banks’. This may indicate the potential for a marketing or education campaign by banks to attract these potential customers. There are also indications that the higher deciles segment might be a bit more price sensitive (i.e., high bank fees) than low- and middle-income customers.

41 In this report, the measure of ‘income’ is actually household consumption, because estimates of expenditure are typically more reliable than estimates of income. In general, the results noted throughout this report also apply for household income, although the relationship isn’t quite as tight.

0.1

0.1

0.3

0.8

0.8

1.2

1.0

1.3

6.0

0.3

10.1

78.2

0.0

0.0

0.1

0.6

0.9

1.1

1.3

1.3

2.6

4.2

8.8

79.1

0 20 40 60 80 100

Interest rate offered on bank account are too low

Prefer to use the money for otherinvestment

Don't trust banks

Bank location is not convenient

Bank fees are to high

Bank staff are rude/unhelpful

Financial products are not suitable

Other

Don't see advantage of havingbank account

Don’t know how banks operate

Do not have a job

Do not have money

Previously Banked Never Banked

Conference Edition

65

Figure 23: Main Reasons for Not Opening a Bank Account, by Income Deciles

The finding that across income deciles, people report that lack of income is the most important constraint to opening a bank account is somewhat surprising and in contrast to other studies which show that even the poor save and also that most Asian countries have high savings rates. In order to explore the reasons for this finding further, Figure 24 provides a distribution of the sample across expenditure levels in Rupiah per year. This shows that the maximum household expenditure reported in the survey was about Rp. 50 million per year. 80 per cent of the sampled population reported household annual expenditure of Rp. 25 million per year or below. The sample is representative of lower-income households. This income distribution provides a clue as to why households across the income distribution report lack of income to be a constraint. It is likely that formal banking sector savings are driven by a small segment of households that are much wealthier than the mean household in this survey. While this latter issue is not explored more in detail in this report, it is clear that incomes are a major barrier to formal financial access. This is in line with the findings in several other countries – such as Brazil, India, and Pakistan – where similar surveys have been conducted.

0.0

0.0

0.1

0.1

0.2

0.2

0.0

0.0

0.2

0.3

0.8

1.3

0.9

1.5

1.6

0.4

1.4

2.0

0.8

0.7

2.0

0.6

1.3

2.7

3.3

2.9

3.1

16.8

5.1

3.8

1.4

4.6

6.2

75.3

81.5

76.7

0 10 20 30 40 50 60 70 80 90

Low

Middle

Upper

Do not have money

Don't s ee advantage for having  bank account

Do not have a job

Don’t know  how  bank operate

Financ ial product are not s uitable

Bank fee are to high

Bank s taffs  are rude/unhelpful

Other

Bank location is  not convenient

Interes t reate offered on bank account are toolowDon't trus t banks

Prefer to us e the money  for other inves tment

Conference Edition

66

Figure 24: Yearly Total Household Expenditure

These results have two important implications for policy. First, the dominant reasons for being ‘unbanked’ are economic. This implies that higher income (i.e. stronger, sustained economic growth) is the main driver for alleviating constraints to having a bank account. Second, because results hold across all deciles levels, these results suggest that even when incomes go up, people still feel that they don’t have enough money to open a bank account. These findings also point to the important issue of threshold income levels at which people start to become ‘financially included’, an issue that is pursed in Chapter III.6. These results also suggest that there might be a role for consumer education aimed at convincing people of the advantages of saving, especially those who have never had a bank account. Section III.8.1 explores the possibilities for lowering some of these barriers.

III.4.3 Key Socio-economic Characteristics of Indonesia’s Savers There is a great deal of interest as to the main social and economic characteristics of Indonesia’s savers. This Section describes their salient features according to the information provided by the Survey. The statistically significant features are summarized in Section III.4.4. Beginning with education, the use of bank savings account is clearly related to the education level of the respondent (see the upper left panel in Figure 25). For example, respondents who are university graduates are almost four times more likely to have a savings account than respondents who did not complete primary school (and eight times more likely than a respondent who never went to school). As regards an urban/rural split (the upper right panel of Figure 25), urban populations make greater use of formal and informal savings instruments, in marked contrast to rural households which exclusively use informal sources significantly more than urban households. Whereas 53% of

Conference Edition

67

the urban population saves at a formal bank, only 32% do so in rural areas. By contrast, while 16% of the urban population saves exclusively in informal institutions, only 20% of the rural population does so. Extending this regional breakdown to inter-island comparisons raises two important observations. First, urban areas are very similar in terms of their access to savings accounts both on and off Java (Figure 25). However, the rural differences are striking. Rural Java has almost 35% banked population versus only 20% off Java. Further, almost 70% of off Java rural residents have never had a bank account, compared to 47% for Java. This is additional evidence of unmet demand for financial services in rural, off Java areas. Considering gender dimensions, there are no significant differences between male and female respondents (Figure 25). Female respondents are just as likely to have a savings account (68%) as males; 41% have bank accounts (compared to 40% for male respondents); and 49% save informally (compared to 47% for male respondents). These results stand in a stark contrast with, some other countries, for example, Pakistan, where women have dramatically less overall access to financial services.

Figure 2: Savers’ Socio-economic Characteristics

By occupation, government employees are the most likely to have formal savings accounts, very likely because their salaries are paid into these accounts. Freelance workers (18%) and unpaid

Bank  Ac c ount Owners hip  by  E ducation

11.023.3 27.7

43.264.4

87.8

9.712.3

21.820.5

20.7

11.2

79.264.4

50.436.2

14.91.0

0

20

40

60

80

100

Never go tos chool

Did notcomplete

primary  s chool

Primary  S chool S econdaryS chool

High S chool Univers ity

C urrently Banked P revious ly Banked Never Banked

%

Us age  of S av ing s  Ac c ountB y  Type of Oc cupation

88.8

57.3

39.1

36.2

18.3

4.9

21.4

18.8

12.8

20.2

21.3

42.1

51.0

61.5

6.3

0 20 40 60 80 100

Government

Private

S elf Employed

Unpaid HH Worker

Freelance Worker

C urrently Banked P revious ly BankedNever Banked

%

34.4 18.3 47.3

19.9 10.9 69.2

52.7 18.4 28.9

52.8 13.8 33.4

0 20 40 60 80 100Percentage

Rural

Urban

Java

Off-Java

Java

Off-Java

Urban vs. Rural in Java and Off-JavaUsage of Savings Account

Currently Banked Previously BankedNever Banked

40.1

18.9

41.0

41.1

16.0

42.9

40.6

17.5

41.9

0

20

40

60

80

100

Percent

Male Female Total

By GenderUsage of Savings Account

Currently Banked Previously BankedNever Banked

Conference Edition

68

family workers (36%) are substantially less likely to have formal bank accounts (the lower left panel of Figure 25). On a different definition of occupations, the survey finds that households involved in agriculture are substantially less likely to have either formal or informal savings. Indeed, salaried employees are more than twice as likely to have a bank account as agricultural workers (Figure 26). From another angle, a respondent with household enterprise ownership is substantially more likely to have a savings account and to have one in a bank (Figure 26). This suggests that enterprise ownership is simply acting as a proxy for income and/or wealth, which is normally associated with savings. This strong positive relationship is discussed below. The analysis finds no substantial variation in savings account usage by age of respondent. Those in the 40-51 age bracket have only slighter higher usage rates than the other 10-year age brackets.

Figure 26: More Savers’ Socio-economic Characteristics

Turning to the all-important metric of income, the survey analysis shows that income is positively and strongly correlated with having a savings account (Figure 27). This result holds across all income deciles with no indications of a critical threshold level of income, including below the official poverty line. The result also holds independent of region (that is, urban or rural), although the gains accrue more quickly (that is, at lower income levels) for urban residents (Figure 27).

Do  you  have a  s av ing s  ac c ount of any kind?

Ag ric ulture S ec tor Workers  vs . S alaried  E mployees

60.0

87.0

0

20

40

60

80

100

Agriculture S ec tor Workers S alaried E mployees

%Do  you  have a  bank  ac c ount?

Ag ric ulture S ec tor Workers  vs . S alaried  E mployees

28.8

70.9

0

20

40

60

80

100

Agriculture S ector Workers S alaried E mployees

%

Do  you  have a  s av ing s  ac c ount of any kind?

B y  Hous ehold  E nterpris e Owners hip

61.3

78.7

0

20

40

60

80

100

Non‐owners Owners

%Do  you  have a  bank  ac c ount?B y  Hous ehold  E nterpris e Owners hip

34.4

50.4

0

20

40

60

80

100

Non‐owners Owners

%

Conference Edition

69

Figure 27: Savings by Per Capita Expenditure

III.4.4 Econometric Analysis of Savings In this Sub-section, the preceding analysis is taken a step further by attributing the likelihood of holding savings accounts to various several socio-economic characteristics covered by the Survey. The analysis is in two parts: i) the probabilities of households having a savings account of any kind, and ii) the probabilities of formal and informal savings accounts. The regression analysis, which uses a weighted linear probability model, is presented in detail in Annex B. 42 These indicate that households are more likely to have a savings account of any kind (formal or informal) if the respondent attended school and is financially literate (as measured by his/her score on the financial literacy questions of the A2F Survey). Specifically, controlling for all observed and unobserved variation across villages, schooled respondents are 11% more likely to have a savings account as compared to respondents who did not attend school; financially literate respondents are 8% more likely to have savings accounts than financially less literate respondents.

42 This determines the baseline model with follow up by non-parametrically controlling for all province and then all village fixed effects. These non-parametric controls are included to absorb all unobserved variation across provinces and villages, respectively. For analytical purposes, the interest in including these additional controls is to check the robustness of regression coefficients on our main covariates of interest. The regressions prove robust to alternate limited dependent variable regression specifications, such as probit and logit specifications. The choice of a linear probability specification was made for ease of interpretation of the regression coefficients.

0.5

1

0 5 10 0 5 10

Urban Rural

Wei

ghte

d M

eans

Monthly Per Capita Expenditure DecilesNote: Vertical line corresponds to poverty line.

By Monthly Per Capita ExpenditureDo you have a savings account of any kind?

Table 12: Summary of Indonesian Savers' Characteristics

i) Who Saves Formally?

--Smaller, older, urban households --Individuals who are financially literate --Households (HH) with higher income --HHs with larger houses --HHs with basic amenities --HH that own an enterprise --HH with a migrant worker abroad

ii) Who Saves Informally? --Individuals who are financially literate --Respondent is not head of HH --Respondent is younger & risk-taker --Wealthier HHs with some amenities --Own an enterprise --HHs without a migrant worker abroad

Conference Edition

70

Other characteristics help explain the likelihood of having savings accounts, specifically: higher income households (as measured by consumption expenditure); households that reside in larger houses (with more rooms); and households that have basic amenities such as electricity, telephone and tap water. Also, households that own enterprise and have migrant workers abroad are substantially more This indicates a low savings rate at all income levels for reasons that are not pursued in the Survey. likely to have savings accounts. These findings point to overall household income and wealth as positively correlated with having a bank account, which is fully consistent with previous findings. Similarly, households that own an enterprise or have migrant workers abroad likely require savings accounts in order to conduct business and send/receive remittances. The results of the breakdown as between those who save formally and informally are summarized in Table 12 (regression results are in Annex C and D, respectively. The main difference concerns those households with a migrant worker abroad. They either save formally, or they do not save at all, probably to assist with money transfers from abroad.

III.5.1 Survey Results on Demand for Loans

This Section turns to the loan side of financial access, following the same logical sequence as for savings in the previous Section. First, we look at ‘Why people borrow’, including by type of borrowing institution; the rates at which they borrow; and loan sizes by institution. The issue of ‘Why people do not borrow’ is pursued in some detail in Section III.5.2, including using other sources of information, owing to certain limitations of the access to finance survey. Subsequent Sections focus on socio-economic characteristics of borrowers (Section III.5.3) and econometric analysis that pinpoints the statistically significant factors (Section III.5.4). The higher-level survey results for borrowers are summarized in Figure 28. As indicated, a fairly large proportion (60%) of the Indonesian population borrows money, but only 27% do so from a formal bank or microfinance institution. A much larger proportion (43%) borrows from informal sources (such as neighbourhood schemes, friends and family); smaller proportions borrow from community welfare schemes (6%) and pawnshops (3%).

Conference Edition

71

Figure 28: Summary of Survey Results for Borrowers

Total Population

60% Borrow 40% Don’t Borrow at Present Why Not?43 i) 60% are not creditworthy ii) 20% don’t want it.

8% Applied in iii) 4% no collateral Informal Banks MFIs Community Pawnshops past 12 months iv) 16% other 43% 17% 10% Welfare 3%

Schemes 6% 1% pending 1% rejected Why?44

i) Lack of documentation (45%)

ii) No collateral (32%) iii) Insufficient income (22%) iv) Too much debt (10%)

The data on borrowers in Figure 28, like those for savers (Figure 20), highlight the shortcomings of the formal banking sector as regards the extent of its outreach to the Indonesian population. Namely, a remarkably low proportion of the population (17%) currently borrow from banks. Far more—roughly 2 ½ times as many—borrow from the informal sector, and more borrow from institutions like MFIs, pawnshops and community welfare schemes. Concerning the summary uses of household borrowing, formal sources are used primarily for business loans, whereas informal sources are tapped mainly for consumption purposes. This simple observation has important policy implications. Namely, policies that aim at different users of credit (e.g., investment versus consumption) need to aim at different financial services providers (e.g., rural banks versus pawnshops). Continuing with those who do borrow (that is, moving down the left branch of Figure 28), the most popular service providers are in the informal sector. The loans are generally used for consumer expenditure (some 40-50% of respondents) and business & investment spending (25-40%). Borrowings for other purposes are quite small and the pattern is relatively uniform across informal providers (the left panel of Figure 29).

43 Source: BRI MASS Survey (see Box 3), which refers to the poor and very poor. 44 Based upon a very small sample; see main text.

Conference Edition

72

L oan  fromPawn  S hop, C ommunity  Welfare  S c hemes  & Informals

0.0

3.3

0.0

2.2

4.2

25.5

5.7

18.6

40.5

1.5

2.1

2.2

0.0

4.3

12.4

7.8

31.6

38.2

0.2

0.6

0.7

1.3

4.1

16.5

19.5

19.7

53.8

0 20 40 60 80 100

To  purchase  animals or  livestock

Other

P arty, traditiona l ceremony

Repay  loan

P ay school fee

To  pay for unforeseen  expenses

Investment ex penditure

Business ex penditure

C onsumption  ex penditure

Informal

Community WelfareS c hemePawn S hop

L oan  fromB anks  & MF Is

0.1

0.2

1.1

1.1

3.5

9.1

14.5

31.7

38.8

2.4

1.1

0.9

2.3

8.6

20.6

23.8

14.7

25.6

0 20 40 60 80 100

Other

To  purchase  animals orlivestock

Repay  loan

P arty, traditiona l ceremony

P ay school fee

To  pay for unforeseen  expenses

C onsumption  expenditure

Investment ex penditure

Business expenditure

MFI

Bank

Purposes of borrowing are somewhat different for those who borrow from banks and formal MFIs (the right panel of Figure 29).45 Business and investment purposes dominate these institutions’ lending, with consumer financing occupying a much smaller role. Banks dominate when it comes to business and investment purposes, whereas the MFIs lend a relatively large amount for consumer spending.

Figure 29: Purpose of Loan, by Service Provider

Loan sizes vary significantly by institution (Figure 30). Not surprisingly, banks provide the largest loans to households (1/4 are greater than Rp20 million, roughly US$2,000), and the amounts shrink dramatically deeper into the informal sector. Indeed, it is striking that among community welfare schemes, almost ¼ of loans are less than Rp100,000 (about US$10).

45 Definitions are important. This Survey defines a Micro Finance Institution (MFI) as a small credit program to finance small business to produce income for oneself and the family. For the survey, micro credit is disbursed by non-bank finance institutions, for example: BKD (Village Credit Body); savings and loan cooperatives (KSP); savings and loan unit (USP); village credit fund institution (LDKP); baitul mal wattanwil (BMT); non-governmental organization (NGO); arisan; Grameen financing programs; ASA financing programs; self-help groups (KSM); and credit unions. This definition overlaps with that discussed in Chapter II.6.

Conference Edition

73

Figure 30: Loan Sizes, by Institution

Interest rates charged by various institutions also vary widely (Figure 31). At the lower end, employers charge relatively low rates (less than 10% per annum); friends & neighbors are also relatively cheap, charging about 17%. Banks come next at about 25%, compared with over 40% for MFIs and Community Welfare Schemes. The most expensive are shops (that is, store credit) and daily banks (see the upper panel in Figure 31). The latter are particularly expensive because of the very short-term maturities (1 day) and these build up enormously with compounding at annual rates. For policy purposes, it’s important to note that most institutions—especially banks and to a lesser extent MFIs—charge markedly less interest, if the borrower has a bank account (see the lower panel in Figure 31). By implication, one of the simplest ways to reduce interest costs would be for borrowers to open a bank account, probably because this is a low-cost indicator to lenders of creditworthiness. Three qualifications are in order concerning these interest rate data. First, the survey respondents often did not know the interest rate that they paid for their loans or the maturity. Usually, they seemed more concerned with the size of the monthly payments, and if they had the cash flow to carry the payments. Second, within institutions, there is a wide range of rates charged on loans, with the upper end typically 2-3 times the mean (or median). And third, some of the categories have very small numbers of observations, which makes a strict interpretation of these rates problematic.

Bank  L oan  S ize

25.4

27.523.1

23.9

<=  4  million >4‐10  million >10‐20  million >20  million

MF I L oan  S ize

26.4

16.636.7

20.2

<=  0.5  million >0.5‐1  million >1‐3  million >3  million

Pawn  L oan  S ize22.2

26.928.7

22.2

<=  0.2  million >0.2‐0.5  million >0.5‐1.3  million >1.3  million

C ommunity Welfare  S c heme  L oan  S ize

23.4

29.1

29.3

18.1

<= 100 thous ands>100‐300 thous ands>300‐500 thous ands>500 thous ands

Informal L oan  S ize

25.3

25.224.5

25.0

<=130  thous ands >130‐600  thous ands >600  thous ands ‐9  million >9  million

Conference Edition

74

Figure 31: Indicative Loan Interest Rates

III.5.2 Who Does Not Borrow? The Survey identified several socio-economic characteristics of Indonesians who have never borrowed, namely:

• They are relatively concentrated in the lowest income deciles (Figure 32).

8.4

17.1

25.6

42.3

42.6

73.5

312.1

0 50 100 150 200 250 300 350

Employer

Neighbors, Family & Friends

Bank

MFI

Community Welfare Scheme

On Credit From Shop

Daily Bank

% per annum

By Savings Account Ownership

13.0

22.6

29.9

34.7

38.2

39.7

428.9

4.7

38.3

11.8

113.0

45.8

45.7

276.2

0 50 100 150 200 250 300 350

Employer

Bank

Neighbors, Family & Friends

On Credit From Shop

Community Welfare Scheme

MFI

Daily Bank

With savings account Without savings account

% per annum

400

Conference Edition

75

• Concentration in the lowest income deciles is more pronounced for rural areas (Figure 32). However, on average there is virtually no urban-rural difference (see the lower left panel in Figure 33).

• They tend to be poorly educated and live off-Java (see the upper panels in Figure 33). • They tend to be women somewhat more often than men (see the middle lower panel

in Figure 33). And, • Usually, they are not owners of non-farm enterprises (see the lower right panel of

Figure 33).

Figure 32: Non-Borrowers by Per Capita Expenditure

Figure 33: Non-Borrowers’ Characteristics

0.1

.2.3

0 5 10 0 5 10

Urban Rural

Wei

ghte

d M

eans

Monthly Per Capita Expenditure DecilesNote: Vertical line corresponds to poverty line.

By Monthly Per Capita ExpenditureNever Loan Respondent

Non‐Borrowersby  J ava‐Off J ava

14.3

24.4

0

5

10

15

20

25

J ava Off J ava

%

Non‐Borrowersby  E duc ation

14.5

11.2

15.7

13.1

18.3

27.0

0 5 10 15 20 25 30

Univers ity

S enior High S chool

S econdary  s chool

P rimary  s chool

Did not complete primary  s chool

Never go to s chool

%

Non‐Borrowersby  Urban ‐Rural

15.5 15.7

0

5

10

15

20

25

Urban Rural

%

Non‐Borrowersby  Gender

14.516.8

0

5

10

15

20

25

Male F emale

%

Non‐Borrowersby  Non ‐farm  enterpris e  

owners hip

11.7

18.1

0

5

10

15

20

25

Owners Non‐owners

%

Conference Edition

76

III.5.3 Barriers to Accessibility: Why Do They Not Borrow? As suggested by the right-hand branch of Figure 28, the survey did not pursue the reasons as to why respondents have not borrowed. However, the Survey did probe non-borrowers who had applied for a loan within the last 12 months; if they were rejected, the Survey asks ‘Why rejected?” As indicated in Figure 28, the principal reasons cited were:

• Lack of documentation (45%), for example: personal identity; land ownership certificate; or proof of permanent address;

• Insufficient collateral (32%); • Insufficient income (22%); and • Too much debt (10%).

It should be noted that the number of data points in these categories is very small, which undermines their statistical reliability. For instance, only 1% of population has had a loan application rejected in the past year, and each of the rejection reasons, noted immediately above, is a subset of this 1%. To provide more insights into this critical issue, this report turns to the BRI MASS Survey, which provides the following critical results (Box 3):

• 40% of the unbanked poor and very poor are creditworthy by the commercial standards of BRI’s examiners;

• By implication, 60% of the unbanked poor and very poor are not creditworthy by the standards of BRI’s examiners;

• Of the creditworthy poor and very poor, 50% do not want credit, that is, they are voluntarily excluded – at the prevailing price of credit; and,

• Of those creditworthy, 10% were excluded owing to a lack of collateral.

What can be inferred on this very important issue? First, we can make a very rough estimate of why 25% of the poor or very poor households (that is, those in the 2 lowest income deciles; see Figure 32) that have never borrowed. Using the MASS results, this 25% breaks down as follows:

• 15% are not creditworthy (by BRI Unit Desa standards). • 5% don’t want credit and are therefore are voluntarily excluded. • 1-2%, are excluded because they have no collateral. And, • The remainder, roughly 4%, are excluded for other, unknown reasons.

Second, lack of collateral does not look like a major problem, certainly not according to the BRI MASS Survey. Evidence of the access to finance Survey is weaker on this issue, but does not

Conference Edition

77

contradict the MASS conclusion, considering the small sample size involved (about 1/3 of 1% of the population). Third, inadequate identification and documentation are more important than collateral, a conclusion inferred from both the A2F and MASS Surveys. Fourth, other problems (like insufficient income and too much existing debt) are roughly consistent as between the two Surveys.

III.5.4 Key Socio-Economic Characteristics of Indonesia’s Borrowers As with savers, there is a great deal of interest in the main social and economic characteristics of Indonesia’s borrowers. The remainder of this Section describes their salient features according to the information provided by the Survey. The statistically significant factors are summarized in Section III.5.5. One important point needs to be made at the outset of this Section. Namely, that average household indebtedness is moderately high in Indonesia (see Table 13). For most socio-economic groups, debt is some 25-30% of annual income. For the poor (that is, for those below the poverty line) total household debt is 19% of annual income. These levels of debt already constitute a sizable burden for households. For example, at typical rates of interest (say 30%; see Figure 31), the average household needs to spend about 1/10th of its income on interest payments, if it can continually rollover its short-term debt. If the debt has to be re-paid, almost 40% of an average household income will be needed in that year.

Table 13: Household Indebtedness

(in US$ @ 9056 Rp/US$)

Total Indebtedness Average Annual As % of Expenditure Expenditure

Average Indebtedness $ 783 $ 2,829 27.7%

- Rural $ 680 $ 2,400 28.3% - Urban $ 926 $ 3,422 27.1%

- Male $ 796 $ 2,830 28.1% - Female $ 771 $ 2,828 27.3%

- Above poverty line $ 881 $ 3,143 28.0% - Below poverty line $ 155 $ 806 19.3% For the poor (that is, those below the poverty line; see Table 13) the situation is similar despite much smaller amounts of debt, because they will probably need to pay considerably higher rates of interest, say 50% (or more, if they need to access the services of a lintah darat; see Box 4). If the short-term debt cannot be continually rolled over, they would need about almost 1/3rd of a year’s

Conference Edition

78

household income. One important limitation of the survey is that it did not explore the non-financial assets of the households; an estimate of the degree of indebtedness as a function of overall household wealth is therefore not available. Therefore, the above findings will need to be interpreted with caution. However, the implication of the above finding is the need to explore much more closely the degree of indebtedness of lower-income households along with policies to further focus on increasing access to credit. Turning to socio-economic characteristics, 50% of household with savings accounts also borrow from a bank (the left panel of Figure 34). By contrast, less than 10% of households without bank savings accounts borrow from a bank. Several explanations could account for this large difference. For example, it could potentially reaffirm an established finding of loan financing, namely that having a savings account acts as guarantee/collateral against which banks can lend money to clients. Or, it could also reflect an unwillingness of certain segments of the population to engage with the formal financial sector, for religious or other reasons. Alternatively, from an economic point of view, if poorer segments of society don’t have enough income to open a savings account, neither do they have enough income to be a good credit risk, so they don’t get loans from formal providers. The evidence, presented earlier (Figure 31), supports the first explanation, namely that formal lenders (banks and MFIs) charge less interest on loans to customers who have a bank account.

Figure 34: Characteristics of Borrowers

Has HH ever borrowed  money from   ...By Urban/Rura l

0.1

0.1

0.1

0.1

0.1

0.3

0.6

0.2

0.1

0.1

0.1

0.1

0.2

0.3

0.4

0.3

0.0 0.2 0.4 0.6 0.8 1.0

Daily  B ank

E mployer

C ommunity  Welfare S cheme

P awnshop

Microfinance

C redit from S hop

Neighborhood C ommunity

F ormal  Ins titution

Rural Urban

Has your HH ever used  ... to  borrow  money?By S aving  Account Ownership

0.1

0.1

0.1

0.1

0.1

0.3

0.6

0.1

0.0

0.1

0.1

0.1

0.2

0.4

0.4

0.5

0.0 0.2 0.4 0.6 0.8 1.0

Daily  B ank

E mployer

C ommunity  Welfare S cheme

P awnshop

Microfinance

C redit from S hop

Neighborhood C ommunity

F ormal  Ins titution

Without S avings  Account With S avings  Account

Conference Edition

79

On a regional basis (Figure 35), the Survey shows that loans from banks, MFIs, and pawn shops are more common in urban areas. Rural households rely substantially more on informal sources of loans; loans from Community Welfare Scheme are very similar in both regions.

Figure 3: Borrowers’ Characteristics, Urban/Rural

By gender (Table 14), there are no significant differences, although women are slightly more likely to borrow from informal sources. As noted earlier, these findings are in sharp contrast to surveys of other developing Muslim-dominant countries.

Table 14: Borrowers’ Characteristics, by Gender

In contrast to gender, age of the respondent does matter, and there is an opposite relationship between formal and informal loans. Older age brackets are more likely to have loans from banks or MFIs (except for the oldest category); younger age brackets are more likely to have informal loans. These results suggest that formal financial institutions take experience (proxied by age) into account when making loan decisions. The drop off in bank loans for the highest age category (which is faster for MFI loans) is not surprising as this category comprises many retirees.

17.3

11.7

5.3 5.5 4.6 5.1

18.6 19.5

12.5

6.44.3

2.84.5 5.6

27.5

21.3

0

5

10

15

20

25

30

Bank MFI Employer Pawn Daily Bank Community Welfare Scheme

Neighbour, Family

On Shop

Urban Rural

%

Pawn Community 

Shop Welfare Scheme

Male 17.4 11.0 2.5 6.8 41.9Female 17.3 9.4 3.4 5.4 44.2

Bank MFI Informal

Conference Edition

80

Figure 4: Borrowers’ Characteristics, by Age

By type of employment, government employees are substantially more likely to have formal bank loans, whereas freelance workers and unpaid households (the ‘other’ category) are highly likely to borrow from informal sources. These findings point towards both supply and demand side factors. On the supply-side, formal bank institutions may consider government employment as a significant, positive factor in making loan decisions. On the demand-side, government employees, by virtue of receiving their salaries through banks, have greater interaction with banks and are more likely to be better informed on bank products that are on offer.

Figure 37: Borrowers’ Characteristics, by Employment Status

As was the case for savings, salaried employees are almost three times more likely to have a formal bank sector loan (four times for an MFI loan) than agricultural workers. The probabilities are

0

5

10

15

20

25

<=32 >32‐40 >40‐51 >51

Do you have a ... loan?by Age Bracket

Bank Loan MFI Loan

%

0

10

20

30

40

50

60

<=32 >32‐40 >40‐51 >51

Do yo have an informal loan?by Age Bracket

Informal Loan

%

0 10 20 30 40 50 60

Freelance worker/ Casual Labor

Self Employed

Private Employed

Other

Government employed

Do you have a … loan?

MFI Loan

Bank Loan

% 0 10 20 30 40 50 60

Freelance worker/Casual Labor

Self Employed

Private Employed

Other

Government employed

Do you have informal loan?

Informal Loan

%

Conference Edition

81

similar as regards borrowing from the informal market (Figure 38). This implies that the two types of workers have equal access to the informal market, but formal financial institutions prefer regular versus irregular income flows, which would be a low-cost proxy for risk.

Figure 38: Borrowers’ Characteristics, by Type of Job

Households that run non-farm enterprises are more likely to borrow from banks, MFIs, and informal sources (Figure 39).

Figure 39: Borrowers’ Characteristics, by Enterprise Ownership

Finally, looking at the critical metric of income, there is no pronounced relationship between borrowing from any source and total household expenditure, apart from the lowest decile of income (Figure 40). The observation is very similar for both urban and rural areas. However, the financial institution of choice does change with income (or expenditure); as income rises, borrowers (urban and rural) make markedly greater use of banks (see Figure 41). By contrast, their use of pawnshops and community welfare schemes drops off sharply at higher incomes. The drop-off at higher incomes tends to be less pronounced for MFIs and informal sources.

0

5

10

15

20

25

30

35

Agriculture Sector Workers Salaried Employee

Do you have ... loan?Agriculture Sector Workers & Salaried Employees

Bank Loan

MFI Loan

%

0

10

20

30

40

50

Agriculture Sector Workers Salaried Employee

Do you have informal loan?Agriculture Sector Workers & Salaried Employees

Informal Loan

%

0

5

10

15

20

25

Non‐Owners Owners

Do you have ... loan?by Non‐farm Enterprise Ownership

Bank Loan

MFI Loan

%

0

10

20

30

40

50

Non‐Owners Owners

Do you have an informal loan?by Non‐farm Enterprise Ownership

Informal Loan

%

Conference Edition

82

Figure 40: Borrowers, by Region, by per Capita Expenditure

Figure 41: Bank Loan, by Region, by per Capita Expenditure

III.5.5 Econometric Results for Loans Based upon econometric analysis (see Annex E), borrowing of any kind (formal, semi-formal (MFIs and cooperatives) or informal) is more likely if: the respondent is married; attended school; high income; and employed. For formal sources (Annex F), older respondents and those with better

0.5

1

0 5 10 0 5 10

Urban RuralW

eigh

ted

Mea

ns

Monthly Per Capita Expenditure DecilesNote: Vertical line corresponds to poverty line.

By Monthly Per Capita ExpenditureDo you have a loan?

0.5

0 5 10 0 5 10

Urban Rural

Wei

ghte

d M

eans

Monthly Per Capita Expenditure DecilesNote: Vertical line corresponds to poverty line.

By Monthly Per Capita ExpenditureDo you have a bank loan?

Conference Edition

83

math skills are more likely to have a loan. Household level characteristics include household wealth size and basic amenities.

Regarding informal sources, households are more likely to borrow if the respondent is

married; employed; and if the household as a whole is wealthier and does not own property (Annex H). Borrowers that own property and are not in the formal loan market probably have other avenues for funding or chose not to borrow at all.

III.6 The ‘Truly Financially Excluded’: No Loans And No Bank Accounts

This Section looks at those Survey respondents with neither a bank account nor a loan, loosely referred to here as ‘the truly financially excluded’. In total, they constitute about 17% of the (weighted) A2F sample. As for their socio-economic characteristics--and beginning with the critical metric of income--the rural poor stand out from the rest (the right panel of Figure 42); more than 20% fall into this category. Indeed, improved access is almost exponential as income rises, right up to the seventh income decile. For the urban poor (see the left panel of Figure 42), the improvement is only pronounced at the lowest levels of income; after that it is relatively flat through the fifth income decile, when access improves again. At the middle and higher income levels, there is very little difference between urban and rural regions.

Figure 42: Characteristics of ‘Truly Financially Excluded’, by per Capita Expenditure

Looking at other characteristics of this important group (Figure 43), the ‘truly financially excluded’ tend to be the poorly educated; they live off Java, in rural regions; and they are not owners of non-farm enterprises. Gender differences are relatively minor (the lower left panel of Figure 43).

0.5

0 5 10 0 5 10

Urban Rural

Wei

ghte

d M

eans

Income DecilesNote: Vertical line corresponds to poverty line.

By IncomeRespondent Without Loan & Savings

Conference Edition

84

Figure 43: Characteristics of the ‘Truly Financially Excluded’

Among other indicators (see the final panels of Figure 44), the incidence of respondents without a loan or savings account tends to increase in a pronounced way with age. By type of employment, they tend to be the unemployed, self-employed and freelance workers. Furthermore, they tend to score low on the math and financial literacy scores (see Section III.8.2), probably reflecting the relatively low levels of education just noted. These results are important for policy. They are clear evidence that policies to improve access to financial services should target the rural poor and poorly educated in off-Java regions.

Figure 44: Characteristics of the ‘Truly Financially Excluded’, continued

11.9

20.9

0

5

10

15

20

25

Urban Rural

Respondent Without Loan & Savingsby Urban‐Rural%

17.2 17.0

0

5

10

15

20

Male Female

Respondent Without Loan & Savingsby Gender%

10.1

21.5

0

5

10

15

20

25

Owners Non‐owners

Respondent Without Loan & Savingsby Non‐farm Enterprise Ownership%

14.8

32.5

0

10

20

30

40

Java Off Java

Respondent Without Loan & Savingsby Java‐Off Java%

3.7

5.6

12.8

17.5

24.9

43.2

0 10 20 30 40 50

University

Senior High School

Secondary school

Primary school

Did not complete primary school

Never go to school

Respondent Without Loan & Savingsby Education

%

0.4

0.50.6

0.7

0.80.9

0.0

0.2

0.4

0.6

0.8

1.0

Without any loan & savings

with informal account With bank account

Math & Financial  Literacy Level

Financial Literacy Level Math ability level

7.5

23.7

34.5 34.3

0

10

20

30

40

Other Freelance Self employed

Unemployed

Respondent Without Loan & Savings

by Types of Employment%

9.8 10.6

20.7 22.4

36.5

0

10

20

30

40

<18‐25 26‐32 33‐40 41‐51 >51

Respondent Without Loan & Savingsby Age%

Conference Edition

85

III.7.1 Survey Results on Demand for Insurance

At the most aggregate level, the Survey showed that approximately 50% of Indonesian households have some kind of insurance (see Figure 45). This is a remarkably high use of insurance products, considering the evidence on penetration and density from previous studies.46 However, the composition of coverage by type of insurance (Figure 46) is very important in this regard. It indicates that the great bulk was accounted for by travel insurance and public sector health insurance. After further consideration (Box 10), subsequent analysis focussed on only four categories of insurance (asset, education, life and private health).

Figure 45: Insurance Holders, by per capita Expenditure

46 Compare, for example, with World Bank (2006d).

Conference Edition

86

Box 9: Some Details on Insurance in the Survey The Questionnaire asked about seven different types of insurance as indicated in Figure 46, namely: i) public medical; ii) private medical; iii) home owners’; iv) education; v) travel/accident; vi) vehicle/asset insurance; and vii) life insurance. Beyond these 7 categories, the questionnaire was not specific about types of insurance, leaving detailed interpretation to the discretion of the respondent. Further questions asked about ownership; availability; willingness to pay; reasons for not having insurance; and risks. As mentioned in the main text, overall results indicated a surprisingly high level of insurance coverage, around 50% of all households had some form of insurance. However, the details of composition indicated that almost 33% of respondents have (or had) government medical coverage and almost 30% have (or had) travel or accident insurance (see the upper half of Figure 46). This high coverage by public health insurance is almost certainly accounted for by the compulsory coverage of government workers by PT ASKES and by Askeskin (which is social insurance for the identified poor).a/ As for travel/accident insurance, such a high incidence was not envisaged at design of the questionnaire, and they look to be inflating the overall numbers. To avoid distortions of overall results and conclusions, these two categories were excluded from further aggregate analysis. By contrast, home owners’ insurance had miniscule (about 1%) coverage, which undermines detailed statistical reliability; information of this category was rolled into the much broader ‘assets insurance, which also includes ‘home or vehicle insurance’. Further detailed analysis in the main text focussed on the four remaining categories, which are more representative of insurance in Indonesia.

a/ For further information on these types of insurance, see Annex S of World Bank (2008b).

Looking only at the four, more representative types of insurance, Indonesia has low take-up rates and they are heavily tilted towards the urban, upper income groups (Figure 47). The take-up is comparable for the lowest consumption (income) deciles for rural and urban regions, but take-up for urban residents rises much more quickly as income increases. The effect is especially noticeable for asset (e.g. home or car) insurance and for life insurance. In rural regions, education insurance is particularly uncommon. The main socio-economic characteristics of insurance are presented in Figure 47. The self-employed are the most likely to buy insurance of all for four kinds, except for life insurance. For that kind of insurance, private employees are far more likely make the purchase. By gender, women have a strong preference for education insurance, whereas men prefer life insurance and, to a minor degree, asset insurance. There is no difference as regards private health insurance.

Conference Edition

87

Figure 46: Types of Insurance Ownership

There is a dramatic difference in insurance take-up between agricultural workers and salaried employees. The latter are usually more than 10 times more likely to buy insurance—of all types—than the former. Similarly, the owners of non-farm enterprises are much more likely to buy insurance than non-owners. Further issues related to insurance are included in Section III.8.1, below.

26.49.4

0.93.5 10.3

25.69.0

6.0

2.0 0.5 2.03.0

3.3

3.0

67.7

88.698.6 94.4

86.771.2

88.0

0.00

20.00

40.00

60.00

80.00

100.00

Gov't medical ins urance Private medicalins urance

Home ow ner'sins urance

Education ins urance A s s et ins urance Travel ins urance L ife ins urance policy

C urrently  Owned P revious ly  Owned Never Owned

Ins uranc e  Owners hip  by Employment S ta tus

69.53 68.1856.83

68.65 66.50 72.68 76.58

30.47 31.8243.17

31.35 33.50 27.32 23.42

0.00

20.00

40.00

60.00

80.00

100.00

Gov't medicalins urance

P rivate medicalins urance

Home owner'sins urance

E ducationins urance

As s et  ins urance Travel  ins urance L ife  ins urancepolicy

E mployed Unemployed

Conference Edition

88

Figure 47: Characteristics of Holders of Insurance47

47 Note difference in vertical scales in the various panels of this Figure.

Conference Edition

89

Figure 48: Characteristics of Insurance Holders, continued

Ins uranc e  Owners hip

by  Gender

56.2

51.8

50.1

32.8

43.8

48.2

49.9

67.2

0 20 40 60 80 100

L ife ins urance

A s s etins urance

Private Healthins urance

Educationins urance

Female

Male

%

Ins uranc e  Owners hipby  Types  of Work

0.5

2.0

0.9

4.0

15.7

20.7

9.1

0.8

55.9

24.5

22.5

15.8

6.1

22.3

26.6

37.0

21.8

30.4

41.0

42.5

0 20 40 60 80 100

L ife ins urance

Private Healthins urance

A s s et ins urance

Educationins urance

Self Employed

Government Employee

Private Employee

Freelance

Other

%

Ins uranc e  Owners hipAg ric ulture S ec tor Worker & S alaried  E mployee

0.3 2.6 1.2 2.48.119.9 24.2 25.1

0

20

40

60

80

100

Educationins urance

A s s et ins urance Private Healthins urance

L ife ins urance

Agriculture Sector Worker Salaried Employee

%Ins uranc e  Owners hip

by  Non‐farm E nterpris e Owners hip

2.5 8.5 8.0 9.35.1 9.8 11.5 12.9

0

20

4060

80

100

Educationins urance

L ife ins urance Private Healthins urance

A s s et ins urance

Non-Owners Owners

%

Conference Edition

90

III.7.2 Econometric Results for Insurance The econometric analysis for insurance (see Annex H) finds that households are more likely to have insurance of any kind if they are located in urban areas; have mathematical skills; and are relatively well-off with a larger house size with some amenities. It’s notable that results for migrant workers are negative and highly significant, indicating that households with a migrant worker abroad are less likely to hold insurance. This is a confusing result since all (legal) migrant workers are required to have insurance (see Section V.2.1). One simple explanation is that the households with family member abroad, aren’t aware of that the migrant worker has insurance, which is in line with the results of Chapter V concerning the financial literacy of migrant workers and the families. The income from abroad may serve as insurance for many migrant worker families, substituting for formal insurance products. This would be consistent with earlier observations about lower informal savings by families of migrant workers (Section III.4.4).

III.8.1 Other Survey Results: Risks to Financial Well-Being

The A2F Survey also asked respondents about their perceptions of risks to their financial well-being. The results could provide insights into the type of insurance products of interest to households. Among all categories, illness (79%) and loss of employment (56%) are the strongest perceived risks to financial well-being. Damage to dwelling (33%), poor business performance (30%), death (28%) and harvest failures (26%) are other notable risks. By region, illness is the greatest perceived threat to financial well-being in both rural and urban areas. However, urban households fear loss of employment significantly more than do rural households (62% vs. 52%). Similarly, poor business performance is significantly more of a risk factor in urban areas than rural ones (34% vs. 26%). Rural households worry far more about harvest failures (39% vs. 7%). In terms of gender differences, men and women access risks differently. Women assign more weight to illness (82% vs. 77%) and loss of employment (60% vs. 53%); men assign more weight to harvest failures (32% vs. 19%). Richer households worry less about illness (75% vs. 84%) and more

Table 15: Summary of Insurance Holders’ Characteristics

Who Has Insurance? --Households (HH) in urban areas --HHs that are well-off --HHs without a migrant worker abroad --Civil Servants (public health insurance) --Male & older --Less likely if employed, married & head of HH

--Some HH amenities, electricity detracts.

Conference Edition

91

about poor business performance (33% vs. 26%). For their part, poorer households worry more about harvest failures (30% vs. 22%). The Survey did not ask respondents how they protect themselves against such risks, or how they manage the crisis when it occurs. However, if we link these survey results to those that came earlier, some answers do emerge. For example, a rural family would protect itself against crop failure by diversifying sources of income, by having one or members as migrant workers, either overseas or in domestic urban areas. Similarly, during good times rural households often save non-financial assets, for example, by building up their personal stocks of basic foods, like rice; they may buy small animals, like goats; or they may buy small pieces of jewellery, preferably gold. In the event of personal crisis (like illness, death in the family, flood damage, etc.) when these accumulated assets prove inadequate, the poor access the informal market: employers; friends or family; pawnshops; or the local moneylender. Overall, these results suggest demand for formal insurance products that are geared towards low-cost protection from illness, especially among poorer populations. Competitively priced asset insurance products may also be of interest to richer households in urban areas, many of whom may operate a business of their own. Results on international migrant workers suggest possible interest among them in commitment savings products (see Chapters II and V).

III.8.2 Other Survey Results: Demand for Other Financial Instruments The survey finds significant demand for banking services for the purpose of making financial transactions. For example, 22% of households who have bank accounts mainly use the account to transfer funds, either to send or receive money; 12% use their bank accounts to gain access to other banking services such as loans. This probably reflects the growing popularity of relatively new services offered by banks such as cash payment, bills payments and transfer facilities through ATM machines, internet and mobile banking (Box 11 and Chapter II).

Conference Edition

92

Box 10: Survey Results on Mobile Banking As discussed in Chapter V, mobile banking (that is, conducting banking transactions through cell phones, commonly called hand phones in Indonesia) has great potential as a cost-effective means of expanding financial access in many markets. The use of such technology allows banks to provide better services to existing customers and to reach out to new customers without establishing bank branches and/or ATM locations. Moreover, mobile banking presents a promising avenue for future expansion of financial services to the currently underserved populations. Almost 95% of respondents to the household surveyed said that they have access to a mobile phone. However, as mentioned in the main text, only about 3% currently use mobile phones for conducting financial transactions. The survey asked respondents if they would be interested in using their mobile phones to conduct such transactions in the future. 22% responded positively, which is promising interest in such products. More encouraging, the Survey found that 29% of people who do not have a bank account, have a mobile phone and would be interested in mobile banking.

Among the respondents who did not express an interest in mobile banking, the greatest (almost 50% of respondents) concern was difficulty in understanding how the product would work. Some did not have mobile phones (32%), while some were concerned with costs (21%). Overall, these identified constraints are promising for banks and other financial institutions, because the major constraints seem to be lack of knowledge about the product. Potentially, this can be remedied through marketing and product orientation initiatives.

22.06%

77.94%

Yes No

Are you interested in banking services which can use mobile phone?

0.1

.2.3

.4W

eigh

ted

Mea

n

Why are you not interested in banking services which can use mobile phone?

Cost SecurityDifficulty ReliabilityNo handphone Never heard about this serviceNo handphone available Do not need it yetOther

Conference Edition

93

Box 11: Survey Results on Financial Literacy

In recent years, there has been considerable attention in policy circles to providing financial literacy training to poor segments of the population. However, this push for financial literacy programs has not yet been backed up by solid evidence on precise needs and outcomes. The survey attempted to determine current levels of financial literacy. As part of this, all respondents were administered a set of questions on mathematics and logic (cognitive ability), which yielded a math skills score. Among all survey respondents, 74% of households claim they are interested in financial matters and respondents scored well on tests of mathematical ability (81% correct). Also, a large proportion (78%) understand the concept of compound interest. However, only 52% managed to answer all financial literacy questions correctly; fewer (44%) comprehend loan portfolio choice; and even less (28%) grasp the concept of risk diversification. It’s particularly noteworthy that the worst scores were on a question concerning an interest rate calculation on alternative loan repayment schemes; less than half chose the correct answer. Urban households are considerably more financially literate than rural households (60% vs. 47%) and urban respondents score higher in all categories of financial literacy. In terms of gender differences, men are a bit more financially literate than women (54% vs. 50%), and men do better in the mathematics test (83% vs. 79%). In terms of income levels, better off households are markedly more financially literate than poorer households (57% vs. 47%) and are better at math (85% vs. 76%). Finally, households with formal savings accounts are also significantly more financially literate (60% vs. 41%) and score better at math (88% vs. 71%) than do households without a bank account. Taking the analysis a step further, Annex I presents econometric results where households’ financial literacy score is regressed on a range of potential explanatory variables. The regression analysis shows that household heads are more likely to be financially literate if they are: male; older; have attended school; and score well on the math test. Further, urban, wealthier and smaller household heads score better on financial literacy, as do households that own an enterprise. Owing a house is inversely related to financial literacy, but this is roughly offset by certain amenities in the house. Investing in housing by those with limited financial literacy is not surprising, because it is a simple, useful way to save and invest, especially by lower-income groups who (as noted in Section III.4.2) do not understand banks or do not see the advantage of having a bank account.

III.8.3 Other Survey Results: Expressed Demand for Financial Products As preliminary market research, the Survey asked respondents whether they would be interested in specific financial products. In response, 43% of the population would be interested in a commitment savings product (wherein money is set aside, usually for 6 months to a year, without the option of withdrawal).48 Also, 50% of the population would be interested in a retirement savings product. Furthermore, 25% would use the services of a deposit collector who, for a small fee, would come to their village/town to collect monthly savings and then deposit them at a local financial institution. Urban populations are more interested in both the commitment savings and retirement savings products than rural populations (49% vs. 39% and 54% vs. 47%, respectively). Income also

48 Ashraf, Karlan and Yin (2006) find that such programs do lead to significant gains in savings, especially for women.

Conference Edition

94

plays a role as better-off people are more interested than poorer people in both commitment savings and retirement savings products (50% vs. 41% and 53% vs. 47%, respectively). Respondents who already have bank savings accounts express a greater desire for both commitment savings and retirement savings products as compared to the unbanked population (49% vs. 35% and 60% vs. 38% respectively). There are no significant differences based on the gender of the respondent. These results are potentially useful for commercial purposes by banks and other financial services companies. Pilot testing of some of these types of products would be a useful way to try to increase access to financial services for the currently under-served population. Box 12: Evidence on Easing Constraints to Opening Bank Accounts As noted in the main text, the survey indicated that lack of funds and limited financial literacy were the main constraints to opening a bank account. A question arises as to whether relaxing these constraints would induce households to open bank accounts. Using a randomized evaluation design, Cole, Sampson and Zia (2009) use survey data and unbanked respondents from the A2F Survey to answer this question. The researchers directly tested the efficacy of financial education by offering a course on bank accounts to a set of unbanked households that were survey respondents on Java. Among the sample of unbanked households, half were randomly invited to a group-based, 2-hour financial education seminar in which participants were educated on the benefits of savings; of having bank accounts; and the procedures to open bank accounts. The results indicate that the financial education program did not have any significant overall impact on the likelihood of opening bank accounts. However, the researchers found significant positive impacts for financially illiterate and unschooled households i.e. providing financial literacy training to people with low levels of education has significant benefits. A separate experiment was designed to investigate whether lower fees would increase of likelihood of opening bank accounts. Three levels of incentives were offered to randomly selected households with the lowest amount representing a fraction of the bank account opening and maintenance fees. The largest amount was greater than these fees. In stark contrast to the results on financial education, the research found a strong, significant impact on the likelihood of opening bank accounts by offering households financial incentives to cover bank account fees. For example, more than 1/3 would be interested in opening an account, if fees were cut by 50%. This result is important because the sample size is quite large (all those without a bank account, which is almost half of the total population), which boosts statistical reliability. These results suggest that lowering banking fees could be a useful mechanism for increasing financial participation and that well-targeted financial literacy training would be useful.

Conference Edition

95

Another part of the survey questionnaire focuses on the unbanked population in order to identify deterrents to opening a bank account. To determine how best to alleviate two of these constraints (cost and financial literacy), hypothetical questions were posed to households who currently did not have bank accounts. The survey results show that 37% of the unbanked population would be interested in opening a bank account if the bank account fees were cut by 50%. A substantial 58% would be interested in opening a bank account if these fees were completely eliminated. In addition, 74% of the unbanked would be interested in attending financial literacy training. Moreover, these results are relatively homogeneous, except that: urban populations are significantly less interested in financial literacy training programs; and men seem to be more willing to open bank accounts if fees are cut by 50%, although this difference is no longer significant if fees are cut by 100% (or more). These matters are discussed further in Box 13.

III.9 Summary of Policy Issues

This section summarizes the policy issues arising from the discussion of survey results in this Chapter. Policy recommendations are provided in Chapter VI. The discussion below is by major service area, taking special note of a recurring survey result that there are few significant gender differences uncovered by this survey; in general, gender differences are barely noticeable. Limited Reach by the Formal Financial System. A recurring theme of this Chapter is the limited reach of the formal financial system in Indonesia, especially among banks (even including BPRs and BRI’s Unit Desa system). By way of examples, less than half of Indonesians save at a bank and, of them, 2/3rds also save at some type of informal institution. Similarly, only 17% borrow from banks versus 43% that borrow from informal sources. This is striking evidence of the banks’ potential for further contributions to poverty alleviation, if sufficiently low-cost means can be found for these markets to be commercially attractive to the banks. In looking at the reasons for limited reach, physical accessibility is not a generalized problem for bank customers. To be sure, customers in rural off-Java regions require long times to reach bank branches, especially if water transport is involved. But, for the vast majority (over 95%) of survey respondents, bank branches are conveniently (or very conveniently) located. To underscore this point, bank outlets are, on average, already more conveniently located than key public services like basic health and education facilities. To identify the impediments to financial services, it’s necessary to look beyond physical access. Important Financial Services. On the basis of this survey, the single most important financial service is a savings account and the primary reason for holding one is ‘security’ (Table 17). As a policy issue, this highlights the importance of policies to maintain overall financial sector stability as well as promote savings products targeted at lower-income households such as basic

Conference Edition

96

banking services. Credit is important, too, but usage is currently concentrated in the local informal sector, and it is widely diversified among different types of institutions. Savings Accounts. As mentioned, roughly 2/3rds of Indonesians save but mainly through informal providers. For those who do save at banks, they actually make greater use of non-bank service providers; among ‘exclusive users’, the most important are informal providers. For policy, this implies that there is a large demand for formal financial services, if conditions are right for Indonesians to move their savings out of the informal sector and into the formal financial system. The primary reason for saving in banks is ‘security’ (Table 17), this underscores the vital importance of continuing to maintain confidence in the banking sector.

Table 16: Summary of Households Use & Interest in Formal Financial Products

Use of Existing Bank Products Reasons for Using Bank Services Formal Product % use % use

Savings Account 41% Provides Security 53% ATM Card 20% Saving for Future Needs 42% Loans 17% Transfer Money 37% Mobile Banking 3% Emergency Needs 31% Credit Card 2% Access to other Financial Services 26% Interest in Prospective Financial Products Formal Product % interest Notes: Retirement Savings 50% Mainly of interest to urban residents; the better-off; and

those already holding a bank account. Contractual Savings 43% Deposit Collector Service 25% Mainly of interest to those in rural areas. Mobile Banking 22% Cheaper products If bank fees cut in half 37% For those who don’t have a bank account. If bank fees eliminated 58% For those who don’t have a bank account.

As for reasons for not saving, economic reasons are by far the most important. People believe they don’t have enough money to make good use of a bank account (or they don’t have a job) which points to the importance of broad-based policies to raise incomes. These reasons also tend to hold across all income segments, which provide some room for targeted policy interventions in the form of public education programs.

Credit. 60% of Indonesians borrow, but the great majority from informal sources; a mere 17% borrow from a bank. Borrowers tend to be married; relatively well-off; educated and older. They also to be relatively moderately indebted, which indicates that policies need to focus more on broader provision of financial services than just narrowly designed to raise borrowing.

Conference Edition

97

Nominal interest rates tend to be high, except for loans from employers, friends and neighbours. Informal sources are the most expensive, whereas banks and MFIs are about middle of the road. Both banks and MFIs charge lower rates to borrowers who have a bank account. For policy purposes, this suggests that one simple, low cost solution for borrowers who want a lower interest rate, is to open a bank account. Lack of collateral does not look like a generalized problem. It is an issue for some banks, but these currently serve a minority of Indonesian borrowers. Among many providers (including informal providers, MFIs and BRI’s Unit Desa program), the principle of lending against income, not collateral, is already well-established. Other problems, like lack of documentation to access loans, appear more important than lack of collateral. The ‘Truly Excluded’. This report explores one further group, referred to here as the ‘truly financially excluded’, which is to say, those who have neither a savings account nor a loan. Dominantly, they are poor (all the lower income deciles, with incidence dropping steady as income rises) and uneducated; they live off-Java in rural areas; and they do not own a non-farm enterprise. There are slightly more women in this category than men, but the difference is relatively minor. For policy purposes, this group should be the target for policy interventions--that is, the rural, uneducated poor in locations off-Java. Insurance. The survey finds a surprisingly high use of insurance, but mainly in the form of travel insurance and public health insurance of one form or another (e.g., for government employees). Other types of insurance tend to be a high-income product for the urban elite. Among the poor, it appears that having a household member working abroad serves as an elementary form of insurance, effectively substituting for formal insurance. New Products. Some new products of the formal sector would be of interest to consumers (see Table 17). For example, contractual savings products for urban residents or mobile savings services for rural residents. As for extending the reach of formal bank services deeper into the lower strata of society, the most promising avenue looks like mobile banking. Even the poorest in remote villages have access to mobile phones these days, and the survey uncovers considerable interest in mobile banking among those with a mobile phone, but no bank account. Pricing and Training. On pricing of banks accounts, the Survey results show that consumers will respond to attractive pricing of financial services. However, the results also suggest that demand is marginally price inelastic, which means that it may be against the banks’ financial interests to reduce fees. One policy option – that is being implemented by Bank Indonesia in partnership with several banks – is to offer basic banking services or ‘no frills’ accounts (some banks already do). Another option is to encourage technological advances (like mobile banking) that allow banks to reach more customers at lower unit cost. On financial literacy training, results of the survey suggest caution regarding recent global trends towards financial literacy training programs. The experimental results show that financial

Conference Edition

98

literacy training induces only subsets of the trainees to participate in the financial sector. The policy lesson is that careful targeting of such programs on specific demographic groups would be needed to yield significant positive results.

Conference Edition

99

Chapter IV: Regulatory Impediments to Access

IV.1 Introduction

This Chapter assesses the extent to which the Indonesian regulatory environment promotes or restricts access to financial services. Its goal is to identify—as specifically as possible—policy interventions that can improve access while taking into account the critical importance of maintaining the prudential safety of the financial system. As noted in Chapter I, access to financial services is a broad concept covering the presence of price and non-price barriers to the use of those services.49 From the regulatory side, the idea is to establish “rules of the game” that improve the degree to which appropriate financial services can be offered to meet demand at a fair, market-driven price. This includes due allowance for potential customers who are voluntarily and involuntarily excluded, as discussed early in the report. Consumers who choose not to use financial services do not pose a regulatory issue because their lack of demand dictates their decision rather than lack of supply. However, in the case of Indonesia, there are large areas of unmet demand, where consumers want formal financial services but are unable to obtain them, owing to issues like lack of appropriate products and geographic isolation.

At the outset, it’s also important to note that conflicts often develop between prudential regulation and issues like broader access and in making policy recommendations, it is very important to balance such conflicting objectives. With specific reference to low-income households and MSMEs, different countries have taken different approaches. South Africa, for example, has adopted an ‘affirmative action’ approach (see Box --). Supporters see it as a solid policy framework for the future development of the industry, and one that will underpin sound business practice and maintain the strength and stability of the financial sector. This report follows the philosophy that regulatory recommendations should foster sustainable practices within the broader Indonesian financial system. These are aimed at integrating low-income households and MSMEs into the greater financial system, including by following certain lessons based on international experience: 50

a) Subsidized credit, policy-directed lending and debt forgiveness in general undermine economic development, including for poor households; b) Interest rates should be set to cover costs of funding, risk, loan losses and transaction costs;

49 Demirguc-Kunt, Asli; Beck, Thorsten; and Honohan, Patrick (2008), p. 27. 50 Ledgerwood (1998), p. 2-3.

Conference Edition

100

c) Interest rate caps usually limit service and encourage high arrears, especially in areas of poor financial literacy; and c) Regulation should encourage scale, efficiency and enterprise to integrate microfinance into the formal financial sector over the long-term.

Absent these, regulation may favor low-income households and MSMEs to such an extent that the supply of financial services actually deteriorates. For instance, poor regulation could disrupt informal markets that are not mature enough for formal regulations, but that reliably provide financial services.

Conference Edition

101

Box 13: South Africa’s Approach to Improving Access to Financial Services In October 2003, the Banking Council of South Africa released its “Black Economic Empowerment” (BEE) charter, providing for increased access to financial services for poor households and communities. It aimed to direct billions of rand of investment into transformational infrastructure, agricultural development, low-income housing and small- and medium-sized black businesses. Among other things, the charter provides for significant increases in black ownership and control, management and skills development over the next 10 years.

The charter was developed by the financial sector as a whole, representing banks, long and short-term insurers, black professionals and black business, unit trusts, fund managers and brokerage firms. It is a voluntary commitment; it was agreed unanimously by 10 industry associations in the financial sector. Targets are to be achieved consistent with sound business practise.

The broad-based thrust of the charter is reflected in the fact that 81% of the charter targets relate to the employment, training and promotion of black people; improvement of access to financial services by poor people; targeted investments in projects that address backlogs, underdevelopment and support job creation; and the procurement of services and goods from black businesses. The financial sector committed itself to fostering new and developing BEE companies. This includes, where appropriate, referring business opportunities to, and procuring financial services from, black-owned financial institutions.

With specific reference to access to financial services, there was provision for effective access of 80% of people in LSM 1-5 (the lower half of South Africa’s measure of living standards) to products and services in life assurance, collective investments, short-term risk insurance and banking services. There was also provision for 0.2% of the post-tax operating profits of the financial sector for consumer education.

Concerning empowerment financing, some R75bn of total empowerment financingwas directed towards the provision of financing for or investment in targeted investments and BEE transactions. Of this, approximately two thirds will be targeted investments in low-income housing, transformational infrastructure, agriculture and black SMEs; one third would be BEE transactions.

The targets of the charter came into effect on 1st January 2004, and remain in effect until the end of December 2014. Companies are to publish annual BEE reports, including audited scorecards. Achievements, encapsulated in an annual scorecard and charter rating, were to be one of the factors measured by government in the award of state tenders, and by businesses in their commercial dealings with each other.

A mid-term review is due in 2009 and a comprehensive review in 2015. The intention is that, while the charter targets have a finite 10-year life, the principles and commitment to empowerment will live on after 2014.

Source: See http://www.fscharter.co.za/page.php?p_id=1 and http://www.banking.org.za/

Conference Edition

102

IV.1.2 Indonesia’s Financial Regulators & Supervisors

In Indonesia, the main regulatory and supervisory bodies are:

i) Bank Indonesia (BI) as the central bank oversees the banking system and maintains the stability and value of rupiah.51 To achieve this, BI implements monetary policy, governs the smooth functioning of the payment system and regulates deposit-taking institutions, that is, banks.

ii) Ministry of Finance (MoF) has a large role in economic and fiscal policy and it regulates capital

markets and most non-bank financial institutions through Bapepam-LK. The MoF is also responsible for granting certain licenses that establish legal status and the issuance of tax regulations for, among others, all financial institutions.

iii) Ministry of Cooperatives and Small-and-Medium Enterprises (MCSME) has jurisdiction over the

establishment of cooperatives services outlets.52 In Indonesia, different financial providers are covered by different regulators and regulations

(Table 18), as much reflecting local historical precedent as reflecting differences in the products. Often the financial providers--including those that serve low-income segments--are a product of the regulatory framework and historical considerations.

51 Article 7 of Bank Indonesia Act No. 23 of 1999, as subsequently amended. 52 Appendix to Minister Decision No. 351/KEP/M/XII/1998. The authority and objectives of the MCSME are detailed out in Decree No. 9/M/2005, Article 94 and 95.

Table 17: Financial Providers and Financial Services in Indonesia

Financial Provider Organizational Format

Ownership Regulatory Status

Range of Financial Services Permitted

1st and 2nd Tier Commercial Banks - (Bank Umum)

Limited Liability Company, Government Enterprise

Private sector entities, and public sector ownership.

Regulated and licensed by Bank of Indonesia

Lending, small scale business loans (KUK), business development service providers (BDSP), savings facilities

People’s Credit Banks - (Bank Perkreditan Rakyat or BPR)

Limited Liability Company, Government Enterprise

Private sector entities or stakeholders, no foreign ownership

As above. As above.

Village Credit Banks - (Badan Kredit Desa or BKD, currently a sub-set of BPRs)

Limited Liability Company, Government Enterprise

Private sector entities or stakeholders

Regulated and licensed by Bank of Indonesia; supervised by BRI

Lending, small scale business loans (KUK), business development service providers (BDSP), savings facilities

Rural Credit Fund Institutions – (Lembaga Dana Kredit Pedesaan or LDKP)

Limited Liability Company, Government Enterprise

Mostly owned by provincial governments

Mainly regulated and licensed by provincial govts; supervised through BPDs.

Lending, small scale business loans (KUK), business development service providers (BDSP)

Non-Bank Financial Institutions (NBFI)

Limited Liability Company, Government Enterprise

Private sector entities or stakeholders; some public sector ownership

Ministry of Finance Leasing, consumer financing, credit cards and provision of funds and capital goods

Conference Edition

103

Source: World Bank staff research.

IV.2 Assessment of the Current Regulatory System, by Service Provider

In reviewing the current regulatory framework to assess its impacts on access, background work for this report investigated issues on seven broad functional areas, namely:

(i) Entry into the industry: capital requirements; management qualifications and certification; and ownership.

(ii) Activities related to asymmetric information between clients and financial institutions: Know Your Customer Principles; the debtor information system (Credit Bureau); customer protection; transparency of financial conditions; transparency of financial products; and deposit insurance.

(iii) Prudential supervision: regular on-site supervision; capital adequacy; statutory reserves; legal lending limits; etc.

(iv) Pricing of financial products: deposit and lending interest rates.

(v) Servicing of financial products: saving products; loan products; accessibility; and funding.

(vi) Exercising claims/remedies: debt restructuring; and execution of collateral. And,

(vii) Sustaining the industry: unsound ratings/conditions; mergers and acquisitions; single presence policy; and liquidation.

Content summaries of the relevant regulations are presented in Annex J-L. They cover the three most important service providers for current purposes, namely: banks; finance companies; and cooperatives. Similar background work was done for the pawn shops, but the details are not presented in an Annex owing to simplicity of the issues.

Village Unit Cooperative – ( KUD)

Membership-based cooperative or credit union

Members Ministry of Cooperatives and SMEs (MCSME)

Lending to members only

Saving and Loan Cooperative (KSP)

Membership-based cooperative or credit union

Members MCSME Lending, savings, time deposits to members only

Islamic Cooperatives (BMT)

Membership-based cooperative or credit union

Members MCSME Lending, savings, time deposits to members only

Pawnshop State chartered institution with branches

Government/Ministry of Finance

Government of Indonesia

Micro-loans, emergency loans and transactions-for-cash, certifications for minerals and stones

NGO MFIs Non-profit association, trust or foundation

Private sector entities or organization

No regulation Savings and lending facilities and social services

Conference Edition

104

IV.2.1 Regulatory Barriers to Access: Commercial Banks

With the notable exception of Bank Rakyat Indonesia (BRI; see Box 5) and a handful of other large banks that have a large branching network,53 commercial banks in Indonesia have yet to play a major role in direct access to financial services for low-income households. Entry Restrictions are Currently Tight On regulatory barriers, the most important at present concerns entry. In principle, licensing is open, but minimum capital is fairly high and it’s difficult to get a new license. For all practical purposes newcomers must takeover an existing bank.54 As part of the API, this is explicit policy on the part of Bank Indonesia, designed to reduce the number of banks by encouraging mergers and industry consolidation. Monthly Fees and Minimum Balances Vary Considerably by Bank: Most Look High

Based upon the Survey results of Chapter III, the more serious impediments to access arise from high monthly fees and high minimum account balances. For example, monthly administration fees can overwhelm interest paid on small deposits.

Field investigation of these issues indicates that minimum balances and monthly

administration fees are matters of internal corporate policy, not BI regulation. Minimum balances vary considerably by bank, with two of the most popular (BRI’s popular Simpedes account and Bank Danamon’s DSP Savings accounts) set at Rp100,000 (about US$10),55 which looks low. By contrast, BCA’s advertised opening minimum balance on their ‘Tahapan’ account for individuals is considerably higher (Rp500,000),56 and industry analysts believe that this is typical of other large commercial banks. On the basis of this evidence, minimum monthly balances do not look like a major, generalized problem on rupiah accounts.57 However, some banks are certainly higher than

53 As of late 2008, BRI had over 5,200 branches. The next largest, Bank Danamon, had around 1,175, including its small, regional (DSP) outlets. The others with large networks are: Mandiri (almost 1,000); BNI (962); and BCA (819). Another, Panin-ANZ, is currently implementing a major expansion in its branching network. 54 For example, HSBC has bought some 89% of Bank Ekonomi; Maybank has acquired Bank Internasional Indonesia; and Barclay’s has bought Bank Akita. 55 The rules vary slightly by bank. In general, the Rp100,000 is the minimum opening balance for a basic savings account and minimum subsequent deposits are Rp10,000. 56 See http://www.klikbca.com/individual/silver/ind/rates.html?s=3 as of 12 May 2009. Subsequent minimum deposits are Rp50,000 and the minimum balance is Rp10,000. BCA staff note that there is no minimum balance requirement for 2 years for their main TKI product. 57 As for foreign currency accounts (which are little used by poorer households), one upper-tier bank charges a monthly admin fee of US$1 on its US dollar accounts. At rates of interest on deposits in early 2009, (which were unusually low at that juncture), these charges exceed interest earnings on accounts of less than some US$950.

Conference Edition

105

others, which requires households to be selective in choosing services appropriate to their needs (also see comments regarding fees at BPRs in the following Section).

Some Monthly Admin Fees Hurt Small Savers

Monthly admin fees are often a different matter.58 Most banks intentionally structure their

interest payments on deposits and monthly fees in a way that discourages small deposits. Indeed, the pricing structure usually ensures that the balance eventually falls to zero.59 Banks do this because small accounts are administratively costly to maintain and because unilaterally closing a non-zero dormant account entails (contingent) financial liabilities. (Also see the continued discussion concerning BPRs, immediately below). It would be helpful if BI were to establish a clear policy in this area, perhaps making it much easier for banks to close non-zero dormant accounts after an appropriate period of time of lying unused – say two years. One possibility would be to transfer funds from inactive accounts to a centralized account to be administered by a relevant centralized authority – that would therefore have low-cost funds available – while having the on-going responsibility to return funds to the depositor through the bank upon presentation of adequate identification. Further work on this issue including investigation into the number of dormant accounts, amounts of money lying in such accounts, and potential institutional arrangements to manage this issue would be helpful.

Another approach to increase access by lowering monthly administrative fees is by requiring

banks to offer a basic or “no frills” account that include a zero (or very low) monthly admin fee with simple application forms and minimum identity requirements60. This would be in line with basic banking approaches being tried in several countries - a package of free or low-cost services with strict restrictions on the menu of services and the amounts that can be eld in the accounts. For eg: “no frills” accounts may enjoy free transactions up to a limit provided they are carried out at ATMs and not at tellers, fees and charges on some transactions up to certain limits could be waived etc. And all these could be plainly disclosed and publicized.

Basic banking has been offered in some countries for a long time – for eg: Canada, France,

Sweden, and the United States. In the US, “lifeline” banking was introduced in the 1980s and 1990s. In Sweden, banks are not allowed to refuse applicants wishing to open a savings or a deposit account. Similar provisions exist in France and Canada.

58 The monthly fee on BRI’s Simpedes account (without an ATM card) varies between Rp2,000 and Rp5,000, depending upon the size of the account. The smaller fee is for an account less than Rp300,000 (US$30); the larger amount is for an account above Rp1,000,000 (US$100). At BCA, the monthly admin fee is Rp5,000 for basic accounts. Bank Danamon’s DSP has no monthly admin fee for a Savings Account, if the average monthly balance is greater than Rp100,000. 59 BRI’s Simpedes account, for example, has an implicit breakeven point of Rp3 million; above this amount the balance will increase because interest paid exceeds admin charges. But below Rp3 million, the account will eventually drop-off to zero (because admin charges exceed interest paid). 60 World Bank, (2008). “Banking the Poor”.

Conference Edition

106

Although there has been no rigorous evaluation of the efficacy of these schemes, in developed countries, several developing countries are trying such policies to expand access. Pakistan introduced a basic-account regulation in 2005. Mexico’s 2007 regulation provides a list of services that banks must offer free of charge to anyone who meets the requirements and accounts must be below a certain size. India mandated the Zero Balance account in 2005. South African banks voluntarily introduced the Mzansi accounts with no minimum-balance requirements and no maintenance fees but with a minimum set of services offered. A recent evaluation by the World Bank distinguished between regulations mandating basic accounts and the existence of such account offerings by banks (whether mandated by regulation or offered voluntarily). The finding was that basic account availability is positively associated with access as measured by accounts per thousand adults; however regulations mandating such accounts seemed to have less of an effect on access (although the study caveats this latter result by stating that not enough time may have passed between the regulations being implemented and the assessment for the impact to have been felt). The bottom line therefore appears to be that having basic accounts in banks helps increase access.

In October 2009, Bank Indonesia along with the participation of 22 banks, agreed to launch a new saving product called "TabunganKu" (My Saving) in early 2010. The rationale for this initiative is to help providing access to millions of people in Indonesia who are believed to have the financial capacity to save money in banks but in small amounts. As discussed above, for these people the monthly administrative fees that banks normally charge are too hefty and will cause the banks to close their saving account not long after they opened it due to inadequate balance. The features of the TabunganKu is as follows: the initial deposit will be IDR 20,000 (US$2) with no administrative fees and interest rate will be calculated based on daily balance, Along with the introduction of the TabunganKu, Bank Indonesia is also promoting a new theme called 3Ps which stands of "Pastikan Manfaatnya (ensure the benefits), Pahami Risikonya (understand the risk), Perhatikan Biayanya (be careful of the costs)" as part of its overall financial literacy initiative.

Restrictions on New Branches Look Difficult, But Aren’t so Onerous in Practice Two aspects of Bank Indonesia’s regulations regarding new branches look onerous.

• First, BI will only issue the license for a new domestic branch office after an analysis of the respective bank capacity and market. In principle, the criteria for BI’s assessment are the current level of competition among banks; saturation level of the market; and role of the bank in national economic development.

• Second, expansion of a commercial bank’s branching needs to have been identified in the bank’s business plan of the previous year. This looks like it might restrict banks’ ability to respond quickly in the face of changing market conditions.61

61 Indeed, the time horizon could be 18 months or longer, depending upon the timing of the bank’s internal decision to go ahead with a new branch.

Conference Edition

107

Feedback from banks indicates that neither of these is a serious problem in practice, because BI has adopted a relatively liberal attitude in implementing these regulations. Representatives of Bank Indonesia do inspect the location, but their interests seem to be limited to security and IT capacity (which is used as an indicator of management capability). Bankers comment that, in many cases, the site inspections do not pose serious hurdles. Regarding the second bullet above, banks with large branching networks indicate that this is not a major problem for them. They comment that one year is not a long planning horizon, and they note the opportunity at mid-year to modify their plans (with the qualification that they can down-scale their plans at mid-year, but not up-scale them). Apparently, for a bank with good corporate governance, this limitation on branching is not a problem; for others, it may impose a useful discipline. Banks make different points about the cost of branching regulations. For example,62 they say that their major problem arises in the negotiation of rental contracts for branch locations. Potential landlords know that the banks have very limited flexibility in location and timing, and the rental rates are priced accordingly, which unnecessarily inflates banks’ costs. To the extent that these costs are passed on to customers, it has an adverse impact on pricing and access.

Restrictions on Relocations of Branch & ATM are More Problematic

Since early 2009, Bank Indonesia (Regulation #11/1/PBI/2009, Concerning Public Banks) has required detailed information on the locations of existing branches and ATMs. Even re-locations of ATMs to different floors of the same building must be reported to BI. It would be helpful for BI to reassess the need for these sorts of detailed information requirements. It seems more reasonable to require descriptions of the locations of branches and ATMs in relatively general terms, which would allow the banks more flexibility in their extension of services to customers. This is especially helpful as branchless banking models – including ATMs are a prime means to expand access globally. Currently, ‘Directed Lending’ is Not a Substantive Policy Instrument

Following an unsatisfactory experience with directed, small-scale lending in the 1990s, in

recent years Bank Indonesia has been using ‘moral suasion’ to encourage bank lending to MSMEs (see Section 5 of Annex J). Specifically, self-determined targets for MSME lending must be published in the bank’s annual business plan, which looks like the opportunity for BI to encourage greater small-scale lending. As a matter of regulatory reform, it looks more cost effective for banks—and would serve the same purpose for BI—if this information were published in the banks’ audited accounts at year-end; for purposes of the business plan, a general statement of intentions should suffice. Publication at year-end affords banks the

62 There may also be a problem of policy consistency as between Bank Indonesia’s head office and BI’s regional offices, which do most of the on-site inspections for new branch office locations. Bankers indicate that exact procedures for local inspections vary markedly by BI regional office.

Conference Edition

108

flexibility to respond in the course of the year to changing market conditions. Feedback from banks indicates that Bank Indonesia is generally implementing this policy in a liberal way and in practice, it is not a significant issue for access to financial services.

Despite only a few banks playing a significant direct role in access for low-income

households and MSMEs, the commercial banks have a pivotal role in ensuring open, competitive financial markets. In particular, the commercial banks are most likely to be the aggressive newcomers in any promising markets, and they will put competitive pressure on existing entities to lower prices and improve service. In addition, commercial banks are developing important linkages with BPRs (see Section 5.5 of Annex J), thereby providing indirect financing for medium- and small-scale loans. Most importantly, the commercial banks are profit-oriented institutions that will actively pursue opportunities in the smaller end of the market, if those opportunities are financially attractive. As technology advances (see Chapter V, Box 25) and the cost of servicing small customers declines, the commercial banks—including foreign-owned banks—will move into commercially viable segments of this sub-sector, if the regulatory environment permits.

IV.2.2 Regulatory Barriers to Access: Bank Perkreditan Rakyat (BPRs)

In contrast to commercial banks, rural banks (BPRs) have a large, direct role in providing access to financial services for low-income households and MSMEs in Indonesia. By definition, they are relatively small and serve local areas. Often BPRs are the frontline of financial services between this segment and the wider financial system. Since they are rarely—if ever—large enough to pose systemic risk, 63 the regulation of BPRs is light in comparison with the commercial banks and adequate depositor protection is in place (see Section IV.2.4). Minimum Capital is Reasonable, but High for Some Areas

The minimum paid-up capital for BPRs was increased significantly (by a factor of 10 or more) in 1999, but it is still reasonable (up to US$1/2 million, for Jakarta; see Table 19). Also, it varies significantly by location with less populous areas (i.e., smaller markets) requiring smaller amounts of start-up capital (Table 19). This graduated system of minimum capital requirements looks reasonable from an access perspective because it ensures low entry costs into small markets in remote regions of the country. Sharia BPRs enjoy even lower minimum capital requirements in certain areas (Table 19), although the economic rationale for this is unclear. It would be appropriate to reconsider the capital requirement for Sharia BPRs on the basis of risks that they face and the regions in which they operate – in line with those for non-Sharia BPRs. In general, strengthening the existing BPR sector, as well as easing entry for BPRs would likely contribute to increased access and

63 This said, particular supervisory attention would be appropriate, if the BPR has business relationships with commercial banks, other BPRs and/or cooperatives.

Conference Edition

109

attention to this area would be in line with ongoing discussions about local area banking and two-tier banks in several countries, including for eg: in India. Minimum Balances and Monthly Admin Fees are Variable Field interviews indicate that minimum balances and monthly administrative fees at BPRs are quite low. Indeed, they are roughly half—or less—than those at commercial banks, and some accounts have no admininistrative fee whatsoever (e.g., for student savings accounts).64 Regarding ‘no frills accounts’, it looks reasonable to argue that typical accounts at BPRs are all ‘no frills’ relative to those at regular commercial banks. They offer passbook savings accounts with no minimum number of transactions per month. Surprisingly, some BPRs also have forms of ‘contractual savings accounts’. Under these arrangements, every month a customer deposits a certain amount and receives the cash at the end of the term of the contract. There are also ‘mandatory savings accounts’ for loan customers; when a new loan is approved, the money is deposited in the account and the customer withdraws only the amount needed at the moment, leaving the rest to earn interest until it is needed. Issues concerning dormant accounts are similar to those at commercial banks.65 Significant Ownership Restrictions

There is an important regulatory restriction on ownership of BPRs (see Annex J), namely, they may only be owned by Indonesian citizens. In particular, foreign ownership and ownership by

64 A typical opening minimum deposit for a BPR is Rp50,000 (US$5) which is half that at a regular commercial bank (Section IV.2.1). Often there are no minimums on subsequent deposits. Monthly admin fees are normally Rp1,000 to Rp2,000 which is less than half of commercial banks. 65 For example, the BPR will close an account unilaterally if the balance has fallen to zero and if there has been no activity for 6 months. For non-zero accounts, if the account has been inactive for a long period of time, the account will be temporarily de-activated. It will be re-activated if the customer wants to perform a transaction, but the customer must present adequate personal identification to prevent fraud.

Table 18: BPR Start-up Capital Requirement

Conditions BPR (conventional) Sharia BPR

• Capital City Territory of Jakarta Rp5 billion Rp2 billion

• The districts/municipalities of Bogor, Depok, Tangerang and Bekasi

Rp2 billion Rp2 billion

• A provincial capital in Java and Bali

Rp2 billion Rp1 billion

• Provincial capital outside the area referred above

Rp1 billion Rp1 billion

• Outside the area referred to above Rp500 million Rp500 million

Conference Edition

110

foundations (i.e., NGOs) is not allowed. This has worked to the disadvantage of BPRs, which often need capital, management skills and technical knowledge. It’s not clear that many foreign investors would be interested in this business but it is difficult to see why these important options should be closed-off to BPR managements. Relatively Easy to Open New Accounts Currently customers can only open new accounts at the BPR’s heads office, branches offices or cash offices. Independent of KYC regulations (see immediately below), BPRs are cautious in this regard because of the risk of fraud, that is, fake employees of BPRs soliciting new customers and absconding with the proceeds.66 Nonetheless, opening new accounts appears fast and convenient, and special arrangements can be made. For example, upon request BPR account officers will come to new depositors (e.g., to their home) to open a new account and pick up the money. Box 14: Some History on Regulation of Other Microfinance Institutions

Microfinance in Indonesia dates back more than a century to the establishment of the “Priyayi Bank of Purwokerto” in 1895 and the “Poerwokertosche Hulp- Spaar en Landbouwcredietbank” established one year later. The early “Volkscredietwezen” (popular credit system) fostered the emergence of thousands of small village banks with millions of micro-borrowers. The movement culminated in the foundation of the “Algemeene Volkscredietbank” (AVB-Bank) in 1934, which later became Bank Rakyat Indonesia (BRI). It is worth noting that the founders of BRI’s Unit Desa system consulted extensively the credit manuals of the AVB-Bank to design the now famous Kupedes loan and Simpedes savings products. a/

Other financial activities of MFIs have traditionally been undertaken by Village Credit Banks (BKDs) and

Rural Fund and Credit Institutions (LDKPs). BKDs were formed at the end of the 19th century under Dutch colonial rule in Java and Madura as ‘village banks’ and ‘paddy banks’; officially, they do not take voluntary deposits. LKDPs refer to various regional MFIs, most of them established between 1970 and 1990 by provincial administrations.

The Banking Act of 1992 led to a fundamental change in microfinance activities in that BKDs and

LKDPs were required to convert into BPRs. For BKDs, the conversion was relatively simple; the Ministry of Finance granted them a collective BPR licence, with certain exemptions from BPR regulatory requirements (e.g., minimum capital requirements, compliance with soundness ratings and the submission of standardized reports). Also supervision was turned over to Bank Indonesia, which out-sourced the function to BRI (which had supervised BKDs). BI continues to supervise commercial banks with micro-finance operations.

For LKDPs, the transition was far more difficult, owing to: confusing regulations; special exemptions;

and splintered supervisory authority, including involvement by the Ministry of Home Affairs and some provincial governments (e.g., Central Java, through its Regional Development Bank (BPD)). In the end, an estimated 27% of LKDPs converted to BPRs (or cooperatives or other permitted institutions); others closed down or became illegal operations, effective October 1997. At the present time, an unknown number of these institutions continue to operate on the fringes of the financial system.

There are lessons in this experience because these were some of the first institutions in Indonesia to

66 One BPR owner specifically made the point that their customers are mainly low-income with little or no educational background. This segment of customers looks especially vulnerable to this kind of crime.

Conference Edition

111

conduct microfinance, sustainably and commercially, without subsidy from the government. Also, KYC principles were inherent in their operations since the managers of BKDs and LKDPs were prominent village members who knew their customers well. Often this substituted for a lack of formal bank training. The fact that managers were well known in villages also improved transparency of product information in that borrowers were told the terms of loans in a manner that was easily understood. Finally, a strong BKD or LKDP in a village was often well-positioned to open effective branches, thereby expanding sustainable outreach. Nonetheless, regulatory practices have pushed many of them into an unclear, probably illegal status. At this point, there are legitimate issues as to whether it is worth trying to formally clarify this situation until a new MFI Law can be drafted and passed by Parliament. A ‘hands off’ approach is one serious possibility, and probably the default case. Other options might include:

• Acknowledgement of the situation by the authorities, licensing them as semi-formal institutions that abide by non-prudential regulation.

• Adding a new licensing tier for MFIs that accommodates the functions performed by these institutions. a/ This paragraph is borrowed from Bank Indonesia and GTZ (2000). The same source is used elsewhere in the Box.

KYC Requirements May be Problematic In 2001, BI mandated Know Your Customer Principles (KYC) for commercial banks and BPRs in an effort to bolster information and transparency and to compliance with international Anti-Money Laundering (AML). Implementation of KYC may have reduced access to financial services for low-income households. 67 By way of examples that apply equally to commercial banks, new depositors must sign-up in a branch office, which limits outreach including for mobile banking (see Chapter V). Also, problems arise because of identification requirements, which cover identity cards (KTP), driver’s license or passport, but must include information such as date of birth, permanent address, nationality and taxpayer identification number.68 These regulations adversely affect financial access by low-income households, especially as regards taxpayer numbers (NPWP) because obtaining a taxpayer number means burdensome process and probably informal payments for those with precious few resources. On the other side of this issue, some observers suggest that the KYC and AML regime in Indonesia has had no significant negative impact on financial inclusion. From the demand side, there seems to be no real issue concerning identification cards,69 which are widely available and come in variety of obtainable forms. On the supply side, many of the KYC and AML standards coincide with normal business practices, so that the added costs to financial services providers has been minimal in either case. As evidence, they point to Bank Indonesia’s Survey of MSMEs, which indicates that most MSMEs do not have a problem with legal business requirements like licenses and taxpayer numbers when requesting credits.70

67 If the amount is greater than Rp100 million, remittances are subjected to KYC principles, usually by way of identification of both the recipient and sender of monies. 68 Notably, village credit banks, cooperatives, LKDP and other MFIs are exempt from these requirements. 69 For example, everyone has a KTP card. Indeed, many people have more than one. 70 See p. 34 of Bank Indonesia (2005).

Conference Edition

112

To ensure that KYC principles are not obstructing access to financial services, it might be helpful to take certain simple steps, which could potentially help increase access while ensuring that Indonesia maintains a sound KYC regime. For example, BI could exempt small business loans and personal loans under, say, Rp50 million from the NPWP requirement. For larger loans, BI might try linking NPWPs into a public credit registry, preferably working with the Directorate General of Taxation of the Ministry of Finance. Burdensome Reporting Requirements for Small BPRs in Remote Regions BPRs of all sizes are subject to relatively heavy reporting requirements. For example,71 there are monthly financials to BI; quarterly budgets and progress reports against those budgets to BI; publication of quarterly financials in local newspapers (or posted in some public place); semi-annual reports from the Board of Commissioners to BI; an annual business plan and budget to BI; and an annual report. There are also KYC reports to BI and STR reports to PPATK (the financial intelligence agency). For BPRs with good IT systems, these requirements may not be a serious problem. But for smaller BPRs with limited access to on-line services, they do look burdensome. This is particularly the case if the BPR operates in a remote location, because management may need to deliver soft copies by hand to the regional office of Bank Indonesia.72 Clearly there is a trade-off between easing reporting or disclosure requirements, which could be a slippery slope to tread, and prudential requirements. It would be important to look at this issue closely and identify any scope for steps in this area that would help increase access by reducing some of the burdens on BPRs. Some Demanding Management Requirements Bank Indonesia requires a minimum of 2 persons on a BPR’s Board of Commissioners and 2 on its Board of Directors, with certain minimum banking background and education requirements. Industry sources say that these are difficult for some BPRs to meet, especially in remote locations of the country. Bank Indonesia could examine ways to address these issues in regions where management resources are hard to find. Doubtful Benefits from Disclosure Regulations As mentioned, information and transparency regulations in Indonesia require BPRs (and commercial banks) to publish their financial statements and product information (see Annex J). The effectiveness of these regulations for BPRs is highly questionable because BPR clients are often unable to understand the publications due to low financial literacy rates. For clients, reputation, experience and personal relationships are far more important than written information. BI could

71 Source: field interviews by World Bank staff. 72 To be sure, there are very few BPRs in this category, perhaps only 1% of the total. Still, they are exactly the ones that are most needed to extend financial services into Indonesia’s most isolated regions.

Conference Edition

113

however, still usefully retain minimum disclosure requirements in which it receives adequate information from BPRs.

Adding Metrics of Access to CAMEL

Prudential supervision of BPRs (and commercial banks) uses the CAMEL method, (see Box

15) but does not include any metrics that measure access. Notwithstanding recent crisis-driven concerns over international standards for supervision, the current CAMEL methodology is a reasonable starting point for these institutions. For purposes of this report, the key point is that the system does not incorporate any measurement (or proxy) pertaining to access to financial services, which is a negative incentive for banks to expand access to financial services. Future reviews of this framework might consider including indictors of access, while preserving the prudential nature of the Soundness Rating System. Some suggestions are provided in Table 20.

Table 19: Access to Finance Indicators-Loans and Deposits

Branching Restrictions Look Very Restrictive, But Aren’t So Onerous in Practice

BPRs are subject to more branching regulations than are commercial banks, described above. For instance, branches may only be opened in the same province as the BPR’s main office; the BPR must have been financially sound over the past year while maintaining a capital adequacy of at least 10 percent; and the BPR must have current information technology. Again, Bank Indonesia seems to be interpreting these rules is a liberal manner. By way of a few examples, BPRs located in border areas are allowed to conduct operations (that is, to have clients and supervise lending to clients) in neighbouring provinces, even if they can’t open branches. Also, for the purposes of this regulation the entire area of Jabotabek (greater Jakarta) is considered one province, despite covering parts of two provinces and the entire Jakarta Special Region. Likewise, when new provinces are created, BPRs that were formerly allowed to operate in the old province are now allowed to operate in the new provinces created from the old one. While acknowledging BI’s positive approach to managing this situation, it

  Financial Service Physical Access Affordability (% of GDP per capita) Eligibility

Deposits Number of banks

Deposit market share of bank

Locations to open a deposit

Minimum amount to open a checking/savings account

Minimum account balance in checking/savings account

Minimum account balance in checking/savings account

Annual fees in checking/savings account

Number of documents to open checking/savings account

Loans Number of banks

Loan market share

Locations to submit loan applications

Minimum amount consumer loan

Fees consumer loan

Minimum amount mortgage loan

Fees mortgage loan

Days to process consumer/mortgage loan

Source: Finance for All (2208) by Asli Demirguc-Kunt, Thorsten Beck and Patrick Honohan

Conference Edition

114

would be helpful if these restrictions could be officially eased, at least for the sake of regulatory transparency.

IV.2.3 Regulatory Barriers to Access: Cooperatives

The function of cooperatives is to extend loans to members, member-candidates, other cooperatives and/or their members, and to do so in that order of importance.73 Indonesia has roughly 150,000 cooperatives (see Table 3) that serve 28 million members. With this size of customer base, cooperatives play a sizable role in financing low-income households in Indonesia.

A wide variety of

cooperatives are regulated under the Cooperative Law No. 25/1992. Of these, this report will focus primarily on Savings and Loan Cooperatives (KSP) and Village Unit Cooperatives (KUD). This focus is due to KSP being a deposit-taking institution and the large market share KUD holds, being the traditional disbursement tool for subsidized lending programs of the government. As another historical matter, cooperatives have so far played a relatively minor role as financial intermediaries due to traditionally repressive regulation.

The legal basis for specific types of cooperatives is outlined in Table 21. A summary of the

regulatory framework is provided in K, noting that the Ministry of Cooperatives, Small and Medium Enterprises (MCSME) is the regulatory authority for all cooperatives. Cooperatives’ main financial services are noted in Table 22.74

73 PP 9/1995, Art. 19 b. 74 Cooperatives have a business strategy towards membership that supports either a “closed” or “open” bond, although the regulatory regime doesn’t seem to distinguish between the two. An “open” bond cooperative allows a more broadly defined membership than a “closed” bond. Another distinction concerns the difference between primary cooperatives (which are founded by individuals and have a membership that consists of only individuals) and secondary cooperatives (which are founded by other cooperatives to serve a membership of cooperatives).

Table 20: Legal Basis of the Cooperative System

Regulation/Law Mandate:Law No. 25 of 1992 CooperativesGovernment Regulation no. 9 of 1995

Implementation of Saving and Lending Activities by Cooperatives

Decree of the Minister of Cooperatives, Small and Medium Enterprises No. 351/KEP/M/XII/1998

Operational Guidance for Saving and Lending Activities by Cooperatives

Decree of the Minister of Cooperatives, Small and Medium Enterprises No. 194/KEP/M/IX /1998

Operational Guidance for Assessment on the Soundness Rating of Saving and Loan Cooperatives (KSP) and Saving and Loan Unit of Cooperatives (USP)

Decree of the Minister of Cooperatives, Small and Medium Enterprises No. 96/Kep/M.KUKM/IX /2004

Standard Operational Manual for Management of Saving and Loan Cooperatives (KSP) and Saving and Loan Unit of Cooperatives (USP)

Conference Edition

115

Background work for this report has taken note of several prudential weaknesses in the regulatory framework of cooperatives although these weaknesses are not discussed in this report in detail. These should be strengthened in order to ensure a healthy cooperative sector that provides greater access to financial services to its members. Special note is taken of the danger that widespread insolvency of cooperatives poses to the poor and to MSMEs.75 The main regulatory barriers to access are as follows.`

• The establishment of cooperatives services outlets (branch offices, sub-branches or cash offices) must comply with strict licensing standards of the MCSME (see Annex K). The requirement that a primary cooperative must be in operation for two years looks excessive. For example, if a financially sound cooperative wants to open a branch in to serve a broader membership, the two year time limit may allow competitors to move into that market space, thereby creating an investment opportunity lost. The regulation should also consider a distinction in time line depending on the function and size of the branch that is opening.

• In an effort to improve transparency, cooperatives are required to observe substantial

reporting requirements in Indonesia by standardizing financial reporting and publishing product information.76 It is unlikely that this documentation is effective, considering the high rate of

75 Recent reports of widespread insolvencies in East Java and high NPLs are particularly worrisome in this regard. See ‘Hundreds of cooperatives in Malang liquidated’, The Jakarta Post 05/04/2009. 76 Under the Operational Standard for Financial Reporting (Attachment of Decree No. 96/2004), financial statements must be understandable, reliable, relevant and up to industry standards in order to provide a reliable source of information for members, potential creditors and regulatory bodies. More specifically, the regulation states that annual financial statements for KSP that disburse more than Rp1 billion per year must be audited by an independent auditor or by the Auditing Department of the MCSME. Cooperatives are mandated to publish product information on savings facilities and loans (Decree 96/2004 para III.1.2.3 and Decree 96/2004 para III.1.3.8, respectively). Particular to savings, cooperatives are required to detail how deposit interest rates are determined, reward policy for participation and what – if any deposit protection practices are used for members. On the lending

Table 21: Financial Services of Savings and Loan Cooperatives

Type of Service Purpose Decree DescriptionLoans Consumption Loan • Term

• Short-term (less than one year) • Medium-term (1-3 years) • Long-term (more than 3 years)

• Decree 96/2004, Art. 21

Amount of consumption loan is 3 times of savings and not more than 30% of income of the applicant

Working Capital and Investment Loan • Sectoral

• Trade, industry, agriculture,

livestock or services

• Decree 96/2004, III.1.3.2

Amount of productive loan backed up with collateral is 75% of the value of collateral

Savings Deposits • Time • Share

• Mobilize savings for members

• Art. 19 a.PP No. 9/1995

• Decree 96/2004, III.1.1.2

Interest rate on savings for members is higher than non-members

Conference Edition

116

financial illiteracy. A stronger role for the regulator in informing the public about potential problems with specific institutions would be helpful.

• The management of cooperatives and Board of Supervisors establish the interest rates and administrative fees on loan products in the Member Meetings. These are done according to three prescribed methodologies.77 A regulatory framework to better promote access to financing would endorse flexible and market-based interest rates that provide a sound business strategies for cooperatives.

• Cooperative regulations set the criteria for lending.78 Currently, it is difficult for a cooperative to

know the credit history of a loan applicant and cooperatives are not yet to the scale that they are able to participate in the commercial banks’ Debtor Information Systems. More importantly, in a group lending model, it is the strength of the relationship among the members and not necessarily the creditworthiness of an individual that determines the success of cooperative operations.

In summary, cooperatives in Indonesia seem to face a large number of issues that extend well beyond matters concerning broader access to credit. These problems are mainly prudential in nature, and they pose the risk of insolvencies, which would reduce existing access to credit among cooperative members. These issues need to be addressed as a matter of priority. Once these matters are addressed, there are several smaller steps that could be taken to broaden access, including a regulatory framework that endorses more flexible, market-based interest rates; weaker criteria for increasing service outlets; and less onerous reporting and disclosure requirements, which are unlikely to be effective in the light of low rates of financial literacy among members.

IV.2.4 LPS’s Role in the Regulatory Environment The Indonesian Deposit Insurance Corporation (IDIC, or LPS by its Indonesian initials) occupies a very important role in the regulatory framework of the Indonesian financial system. Law No 24 of 2004 provides the legal basis for LPS’s operations, and detailed processes and

side, information includes a written policy on interest rate calculation, membership rewards for participation, types of loans, preconditions for applications and terms of loan. 77 These are: a) expected interest margin based on computation of total costs borne by the KSP; b) an indexed market pricing mechanism; and c) rates to attract new members or restrict current membership. 78 In terms of borrowers, cooperatives screen loan applicants based on the stipulations in Decree No. 96/2004, Art. 19, which state a) members and member candidates must reside in the operational area of a cooperative; b) have a business/fixed income; c) have active savings deposits; d) have no arrears in lending from other cooperative or other party; e) have never been involved in crime; f) have good morality; and f) take a part in a pre-disbursement initiation program.

Conference Edition

117

responsibilities are well-defined.79 Issues regarding LPS and the commercial banks are discussed in a separate document; 80 this report focuses on LPS’s relationship with the BPRs and other microfinance institutions. Several points are notable in this regard:

• LPS’s responsibilities cover only commercial banks and BPRs; they do not include, for example, cooperatives that take deposits.

• Because of their size, BPRs are all considered ‘non-systemic risks’, so the authority for problem resolution rests entirely with LPS.

• Decisions regarding resolution problems lay with LPS’s 6-person Board of Commissioners. • Since September 2005 (when LPS was established), it has closed 16 BPRs, including PT BPR

Tripanca Setiadana (see Chapter II), which was the third largest BPR in the country.81 • To date, LPS reports that all rules (regarding, for example, maximum deposit size and

interest rates) have been observed during liquidation processes. Dissatisfaction among uninsured depositors looks like it has been relatively limited and well managed.

• LPS staff acknowledge that their rules regarding deposit and interest rate limitations are probably not well understood among depositors at the village level. However, they believe that the BPRs have the primary responsibility for providing information and that BI has responsibility for enforcement.

• Significant numbers of BPR closures have been in populous urban locations. For example, 2 around Jakarta and 7 in Bandung.

• LPS’s larger, potential problems appear to lie with the financial cost of resolving failures that might occur among much larger commercial banks.82

Field interviews indicated that LPS has a good track record of liquidating BPRs without the threat of triggering wider problems in the overall banking system. This includes experience with relatively large BPRs and in major urban areas. There is no evidence of regulatory issues that could compromise LPS’s capacity to close insolvent BPRs and to pay off eligible depositors. Issues in this regard relate more to the financial costs of failures among the commercial banks which are much larger and therefore more expensive to liquidate; such problems for LPS could spill-over unto the BPRs. There may also be an issue with adequacy of information being provided to depositors with limited financial literacy.

79 ‘The Bank Resolution Process Undertaken by the Indonesia Deposit Insurance Corporation’ Powerpoint presentation by LPS in April, 2009. 80 See Hanson (2009). 81 The closure of Tripanca created temporary, local spill over effects on other BPRs in the Lampung area, but LPS was able to calm the situation by direct discussions with depositors. 82 See Hanson (2009).

Conference Edition

118

IV.2.5 Regulatory Barriers to Access: Finance Companies

In Indonesia, finance companies (sometimes referred to as ‘multi-finance companies’) cover a wide range of ground, including leasing, consumer financing, mutual funds, factoring, credit card financing, and securities trading (see World Bank (2006)). Their common characteristic is that they are non-deposit taking,83 and they fall under the jurisdiction of Bapepam-LK, the regulatory arm of the Ministry of Finance (for their legal basis, see Table 23). The regulatory objectives of finance companies are similar to those of commercial banks: to improve the role of finance companies as engines of national economic growth and support the efficiency of economic activities in Indonesia. It should also be noted that operations of finance companies in Indonesia are relatively sophisticated operations, which are not normally of interest to MSMEs and low-income households (with the exception of leasing companies and some insurance, as mentioned in Chapter II).

Table 22: Legal Basis of the Finance Companies

• Presidential Decree No. 61 of 1988

• On Financial Institutions (including Security Traders and Venture Capital Companies),

• Minister of Finance Decree No. 1251/KMK.013/1998

• On the Provisions and Implementation Procedures of Finance Companies (including Security Traders and Venture Capital Companies),

• Minister of Finance Decree No. 448/KMK.017/2000

• On Finance Companies (excluding Security Traders and Venture Capital Companies),

• Minister of Finance Regulation No. 64/PMK.012/2006

• On Finance Institutions (excluding Security Traders and Venture Capital Companies),

Traditionally, start-up owner’s equity for finance companies was low and roughly proportional to the size of its outreach and scope of operations (Annex L). However in 2006, regulations were amended (MoF Regulation No. 84/PMK.012, Article 13) that increased minimum paid-up capital in a major way, by a factor of 10 or more, for prudential reasons. This probably squeezed out many smaller players (or at least rendered them inactive), which looks to have reduced services in some areas, but mainly services of interest to mid- and upper-end clients. Similarly, the industry was effectively closed from 1995 to 2000 and then again from 2002-2006 (Annex L). From an access to finance point of view, these actions likely hampered the expansion of formal financial services, including to some small-scale borrowers. As further concerns this consolidation among finance companies (and as was the case for commercial banks; see Chapter II), the underlying issue is whether the regulator has the institutional capacity and resources to weed-out the weaker institutions—big and small—in an even-handed way 83 It is notable that venture capital activities are no longer regulated as finance companies as of 2000 (Decree No. 448/KMK.017).

Conference Edition

119

on a timely basis. If not, the blunt instruments of high minimum capital and closed entry look justified as second-best options (see a similar discussion below, for insurance companies). As mentioned in Chapter II, Indonesia’s leasing industry is one component of the finance sector that has particular interest for issues of access because leasing is important for some MSMEs. 84 Currently, the main regulatory issue concerns industry consolidation. As with the rest of the finance sub-sector, the leasing industry has too many (some estimates place it at half) small companies of questionable viability. Nonetheless, there are regulations that are not conducive to consolidation. For example, foreign investment is limited to 85% of owner’s equity, and another investee company. There are also relatively restrictive limits (40% of equity) on the size of certain investments (see World Bank (2006b) for details). IV.2.6 Regulatory Barriers to Access: Insurance Companies85 Indonesia’s insurance industry is small and highly fragmented with a large number of relatively small players in both the life and non-life segments. By international standards, penetration (premiums relative to GDP) and density (premiums per capita) are low, but the industry is only moderately concentrated (World Bank (2006b)). It is significantly less concentrated than Thailand and the Philippines, and a bit less concentrated than Malaysia. The industry includes a rapidly-growing Sharia component. The legal framework for the insurance industry is the Insurance Law of 1992, significantly modified in 1999/2000 to introduce risk-based capital requirements. Bapepam-LK, the regulatory arm of the Ministry of Finance, is the regulator and supervisor of the insurance sector. There are also important issues of transparency in the industry, and public disclosure requirements are particularly important in the insurance business. 86 At present, companies are required to publish their financial information (including their risk-based capital ratio), but a complete set of all risk-based capital ratings is not available from the regulator and there are no private sector ratings of insurance companies available in Indonesia. The regulator monitors company information and validates the companies’ reports, but it does not publish those ratios. Addressing this issue—possibly in concert with the industry associations—would be a useful step in broadening access to insurance services.

84 See Chapter 7 of World Bank (2006b), which looks at leasing companies. Some data in that report are somewhat outdated, but the issues remain the same, including as regards the regulatory framework. 85 This Section draws heavily upon Chapter 6 of World Bank (2006b). Details comparable to those in Annex L are available from that report, and not repeated here. Some data in that report are somewhat outdated, but the main issues are the same. 86 Unlike other finance companies (where the public can react to worsening news by slowly withdrawing its business), clients of insurance companies enter into long-term arrangements that normally cannot be broken without significant penalties. Consequently, disclosure and transparency at the outset are especially important for consumer protection.

Conference Edition

120

The weak prudential state of the industry is a significant complication for issues of broader access to insurance services in Indonesia. Some large companies in the industry are widely considered to be insolvent and could potentially pose a systemic risk. For their part, many smaller companies are undercapitalized and look unlikely to withstand stiffer market competition in the future. In this situation, the top priorities for policy are: industry consolidation, in particular the resolution of weak and bankrupt firms; and upgrading the institutional capacity of the regulator and improving enforcement. Until matters like these are addressed, issues of broader access to insurance services are of secondary importance; it’s more important to have more reliable insurers, even if that means fewer of them for the time being. With regard to micro-insurance (discussed in Chapter II.8), there is considerable uncertainty in that business partly owing to the virtual absence of an appropriate regulatory framework. Resolution of this uncertainty would give the industry a sizable boost.

IV.2.7 Barriers to Access: Pawnshops

Pawnshops in Indonesia have been an important source of financial services for low-income households and MSMEs since 1928 (Government Regulation No. 81, Staatsblad Nomor 81/1928). They are often the most convenient and trusted source of quick, short-term liquidity for low-income households and MSMEs, usually for personal purposes in times of individual or family crisis. Relatively little business spending is financed by pawn shops,87 and the only official pawnshop, Perum Pegadaian (PP), is a state-owned monopoly.

Since 2004, PP has scaled up micro-lending and its outreach efforts by opening significant

numbers of new branches. As of December 2007, it had 900 branches throughout Indonesia, concentrated in urban areas. These 900 branches (of which 52 are sharia-based) serviced 17 million people in 2007, versus 797 branches and 10.8 million customers at end-2004. In early 2008, PP started money transfer services in cooperation with Western Union. The Government has been discussing the possibility of letting PP list on the Indonesian Stock Exchange in the near future,88.

Regulatory and supervisory oversight of the state pawnshop is the responsibility of the Bapepam-LK under the authority of Government Regulation No. 103 of 2000. As a state-owned enterprise, the management of pawnshops is given considerable independence by the MoF. In general, financial services are lending based on the State Pawnshop Law; fiduciary guarantees; the provision of deposit boxes; and certification of precious minerals and stones.

87 About 25% of customers use the financing for business activities; another 15% use the financing for both personal and business purposes. Women use much more of the financing for personal purposes than men. (World Bank (2008c)). 88 The Jakarta Post, 15 May 2009.

Conference Edition

121

To oversee financial soundness and performance of PP, management is required to document that 45 percent of profit for the fiscal year is allocated for: a) general reserve until this amount reaches 200 percent of owner’s equity; b) social purposes and education; c) bonuses; d) contributions to pension funds; and e) provisions for loan losses. The owner’s equity and capital adequacy level of PP is a decision of the Finance Minister (Art. 10, PP No 103/ 2000). Operationally, the most vital information is the valuation of the goods at hand rather than the creditworthiness of the customer; the main risk for PP is that pawned articles are stolen. Local pawnshop procedures dictate penalties on the client for failure to repay on time.89 As mentioned, the pawnshop business is officially a state-owned monopoly. However, unlicensed (i.e., illegal) private outlets have operated in relatively open competition with Perum Pegadaian for a long time, usually working out of local gold and jewellery stores. For example, a 1986 SEACEN study on informal markets estimated that Indonesia’s pawnshop industry was 42% government and 58% private.90 The state-owned pawnshop has expanded its branch network steadily over the past decade, at about 4% per year from 2000 to 2007, and it has plans for a major expansion in the number of its branches in the next year or so. This is an important step towards improved access for the poor. However, its branches are concentrated in urban areas; they number far less than BRI’s (5,214 branches as of September 2008); and even BRI has come under criticism for not reaching deeply enough into the nation’s remote kabupatens. If the state-owned pawnshop maintains its rate of branch expansion of recent years, it will be another decade or more before it has the same regional reach as BRI, even allowing for a major expansion this year and next. To accelerate the process, other options could be considered, for example, officially opening the pawnshop business to private participation.91 Considering the successful steps taken by state-owned banks over the past two decades, there is little justification for not opening up this sector in the same way, to provide more competitive pricing (that is, higher appraisal values and lower interest rates) and better services for the poor. The sector might even be opened up to a broader set of private investors.92 Needless to say, a stronger supervision framework would need to be put in place concurrently.

89 For instance, late payment usually results in the borrower losing ownership of the pawned good, which is essentially collateral. With this in mind, precise records of transactions are important. In most cases, procedures for delinquent borrowers are quickly handled with collateral already in custody. A unique feature of PP concerns official policy regarding proceeds from auctions. If these proceeds exceed the value of the outstanding loan, the difference is returned to the delinquent borrower. However, no data are available on how often this happens and actual practice has not been confirmed with defaulting clients or with bidders in the auctions. 90 The work is cited in Skully (1994). 91 In 2007-08, a Law was being drafted to open up this business to private participants. 92 Weak management and low profitability of the state-owned pawnshop do not appear to be major issues, which reduces the case for steps like partial privatization of PP. Likewise, surveys indicate high levels of satisfaction among customers (World Bank (2008c)).

Conference Edition

122

Box 15 : Legal Framework for Microfinance Institutions in Indonesia There are number of factors which prevent micro-finance institutions (MFIs) in Indonesia to provide financial services to the poor. Three primary reasons are as follows; firstly, the legal framework for non-bank and non-cooperative type of MFIs is uncertain and this could cause such existing MFIs reluctance to expand their services to new and potential clients and markets. Second, is the high cost of reaching out to many small and potentially high risk customers. Thirdly, many MFIs lack sufficient access to capital. The Government of Indonesia has recognized the constraints the MFIs are facing and is eager to address the issue of legal framework for MFIs as part of a vibrant and comprehensive national financial system. It also recognises that along with other partners, the government has a key role in providing guidance, support, and strategies to ensure that MFIs play their role in providing needed financial services to those at the bottom of the pyramid. Moreover, it is clear that there will need to be a strong policy framework to create an environment in which MFIs can flourish with sustainability and a diversity of products, and such policy that will protect the small depositors and create stability and integrity of the microfinance sectors. A micro-finance law has been under debate for the better part of a decade and has gone through various versions including submission to Parliament. However, it has not yet been approved and questions remain on the role of various institutions in implementation of any regulatory and supervisory framework for MFIs. In early 2009, under the leadership of the Coordinating Ministry of Economic Affairs (CMEA), a joint decree (Surat Keputusan Bersama) was signed between the Ministries of Finance, Cooperatives, Home Affairs and Bank Indonesia to provide a regulatory framework under the existing laws to the non-bank and non-cooperative MFIs that currently operate outside such a framework. The Joint Decree aims to formalize MFIs into one of four categories: (1) People Business Credit or BPR, (2) Saving and Loan Cooperatives, (3) Village-owned microfinance institutions or BUM Desa, or (4) Venture Capital Companies. To implement the above decree, detailed activities for improving legal framework for MFIs for 2009 – 2011 have been developed. A working group consisting of the representatives from the technical ministries, BI, and CMEA has been established and is expected to work on implementing the detailed activities so that MFIs could improve financial inclusion in Indonesia.

IV.3 Summary of Policy Issues, by Service Provider

There are several aspects of Indonesia’s regulatory framework where reform would assist with access to financial services for low income households are MSMEs, as discussed below. At present, the biggest obstacle is entry for: commercial banks; BPRs; insurance; other finance companies; and pawnshops. In most cases, little can be done about this, because official

Conference Edition

123

policy is one of industry consolidation, to force reductions in the numbers of weak companies. This policy is a second-best solution, but it does recognize supervisors’ difficulties in weeding-out the weaker institutions—large and small—in an even-handed, timely manner. Until this policy of consolidation runs its course, new entrants will need to take up ownership in existing—and probably failing—failing companies. Commercial Banks. Current banking policy is controversial for at least two reasons: i) licensing is designed to reduce the number of banks; and ii) minimum capital requirements are intended to force banks to become larger. It is debatable whether either of these policies increases access to financial services by poorer households and MSMEs. However, it looks like it will be some time before until these policies run their course. In the meantime—and with the exception of M-banking (which is covered in Chapter V) —there are no big regulatory winners for access to financial services via commercial banks. For example, oversight concerning ‘directed lending’ to MSMEs and the rules on opening new branches look restrictive, but BI is enforcing them in a relatively liberal manner. Regulatory easing in both areas would be helpful, but it would simply bring official policy into line with actual practice. Recent regulations regarding relocations of branches and ATM machines are somewhat more problematic. Current regulations look unnecessarily tight in terms of location. If the original locations of branches and ATMs were defined in more general terms (especially for ATM machines), it would reduce banks’ reporting and regulatory burden. Similarly, to ease reporting requirements, some reports look like they could be doubled-up. For example, the annual business plan could be part of the company’s annual report. Banks’ minimum balances and monthly administration fees are not a matter of regulation, because those fees are set according to internal corporate policies of individual banks. For both minimum balances and admin fees, there is considerable variation among the commercial banks, but good low-cost options are available through BRI and BPRs.

The lack of official policy on dormant accounts may be contributing to higher-than-necessary admin fees on small accounts, because banks are reluctant to close dormant accounts unilaterally unless the balance falls to zero. Consequently, banks set pricing formula to ensure that the balance in small accounts eventually falls to zero. It might be useful if Bank Indonesia were to set an official policy in this area with a view to making it easier for banks unilaterally to close dormant, non-zero accounts. As part of the new BI policy, the responsibility for administration could be transferred to a centralized account after a certain well-defined time period, and the central authority would assume responsibility for administration of the account.

Another possibility is to introduce basic banking or ‘no frills accounts’, which would include very low monthly admin fees. Bank Indonesia has already launched this and the first accounts are

Conference Edition

124

expected to be available in early 2010. International experience has shown that where commercial banks offer basic banking, more people use banks. BPRs. There are several issues that Bank Indonesia could address to ease the regulatory burden on BPRs, especially for small ones in remote regions. For example, ownership only by Indonesian citizens looks too restrictive. The sector could usefully be opened up to NGOs and foreign ownership, possibly on a joint venture basis. Also, a new, lower tier of minimum start-up capital could be set for particularly small BPRs. Numbers and qualifications for management of BPRs could be eased, especially for regions where management expertise is in short supply. Reporting requirements look burdensome for BPRs in regions without good IT services, and BPR managers may actually have to hand-deliver reports to the closest Bank Indonesia regional office. As part of reporting, disclosure requirements should be examined for their effectiveness due to low levels financial literacy and better alternatives are needed. As for commercial banks, official branching requirements look too tight, but BI is implementing the regulations in a liberal fashion. Still, these could be eased for the sake of regulatory transparency. For both commercial banks and BPRs, KYC requirements may be somewhat problematic. This is particularly the case for credit to MSMEs who have their own reasons for potentially not wanting a taxpayer number. A relatively low threshold could be set to exempt credits from these requirements. Bank Indonesia is currently re-thinking its regulatory policies on BPRs, which looks like the ideal opportunity to put in place a framework that will allow these institutions to enhance access on a sustainable basis. Other MFIs. Many of these institutions operate with no formal legal status. Probably, the most productive thing that can be done is to restore momentum to drafting of a Micro-Finance Law, including an invitation to public debate on the issues. But it will be important that the new Law adopt a facilitating approach aimed at increasing access and learning from emerging lessons based on international experience. In the meantime, it is important to ensure that this essential service continues to function. Cooperatives. Cooperatives play a key role in providing access to a large segment of the population and should be further strengthened to ensure sustainable operations. There are some regulatory impediments in this area, for example in branching, pricing and the effectiveness of reporting. However, more fundamental reforms are needed in many areas, preferably before the sector faces any major difficulties that may jeopardize members’ existing access to financial services.

Insurance and Finance Companies. By and large, these sectors currently cater to the higher income and urban segments. Despite some promising areas (for example, leasing and micro-insurance) that can have a significant impact on access, it looks likely to be a long time before the sub-sector’s reach extends in any significant way into lower-income households. In the meantime, if industry consolidation could be implemented rapidly, then the foundation would be laid for these

Conference Edition

125

sectors to provide enhanced access to a larger share of Indonesians. The micro-insurance business has significant potential among lower-end households, and a regulatory framework more focussed on micro issues, would give the industry a boost.

Conference Edition

126

Chapter V: Special Topics Concerning Access to Finance This Chapter looks at three special topics that are closely connected to issues of access to financial services, namely: Micro- Small- and Medium-sized Enterprises (MSMEs); international workers’ remittances; and mobile banking. The issues in each area are quite different, as are the policy options. V.1 MSMEs and Access to Finance93

MSMEs are widely recognized as very important in any country’s economic development. The underlying reasons are clear: international evidence indicates that they account for some 98% of all enterprises; they employ about 60% or more of the private sector workforce; they create about 50% of value added; and they account for about 30% of direct exports.94 In rural areas of large countries like Indonesia, where large enterprises are few in number, development of MSMEs is almost synonymous with sustainable poverty alleviation. Moreover, the micro enterprise component is the backbone of the informal sector, which accounts for some 70% of total employment in Indonesia.95

Several international surveys regularly find that aspects related to financing are some of the critical constraints faced by MSMEs – ‘Cost and Access to Credit’ are usually a major problem. In the specific case of Indonesia, a Bank Indonesia survey, described in Section V.1.2, took a broader perspective and found variation within MSMEs.96 For its part, The World Bank/IFC’s ‘Doing Business’ indicator97 for 2009 ranked Indonesia 109th out of 181 countries in terms of ‘Getting Credit’, on a par with countries like Uganda, Nepal, Bolivia and Belarus. The OECD, in its

93 On Government of Indonesia definitions, a Micro Enterprise is a productive business unit owned by an Indonesian family or individual and with turnover up to Rp100 million per annum (Decree of the Minister of Finance No. 40/KMK.06/2003, issued on January 29, 2003). A Small Enterprise is any kind of business unit with net assets up to Rp200 million (excluding land and building for business site), or with turnover up to Rp1 billion per year, owned by an Indonesian, independent and not a branch or subsidiary of certain firms (Law No. 9 Year 1995 on Small Businesses). A Medium Enterprise is a business that has annual turnover of between Rp1 billion to Rp50 billion and with net assets between Rp200 million and Rp10 billion, excluding land and building for business location (Presidential Instruction No. 10/1999 on the Empowerment of Medium Businesses). 94 See APEC (2002). 95 For August 2008 on OECD definitions. See footnotes to Table 3.1 of OECD (2008) and http://wwww.bps.go.id/sector/employ/table3.shtml. 96 For example, almost all micro and small businesses, and all medium businesses, have used banking services. For credits, the main considerations are the interest rate and location; the main obstacle for micro businesses is collateral whereas for small and medium businesses, it is high interest rates. 97 See http://www.doingbusiness.org.

Conference Edition

127

pioneering Survey of Indonesia,98 asserted that access to credit is particularly difficult for SMEs, especially those operating in the informal sector. It took special note of the collateral issue.

As suggested by the preceding discussion, most financial studies of MSMEs focus almost exclusively on access to credit (that is, loans). By contrast, this study considers access to a much wider range of financial services, especially for Micro enterprises, which often operate in the informal economy at the household level (as indicated in the pyramid diagram, above).

V.1.1 Key Findings of Bank Indonesia’s Survey of MSMEs

The survey reported in Chapter III focused on households’ access to financial services rather than MSMEs’. However, much of the data reported in Chapter III (especially those for non-farm enterprises, freelance workers and the self-employed) are relevant for Micro and Small enterprises, because they mainly operate at the household level. To broaden the results—so as to include Medium enterprises or to ensure that only enterprises are included—it is useful to examine a survey targeted directly on MSMEs.

Bank Indonesia has done such a survey of MSMEs and their characteristics in Indonesia.99 The survey covered 11,000 respondents in 11 provinces and pursued many topics (such as 98 OECD (2008). 99 See Bank Indonesia (2005).

Large

Medium Enterprises

Small Enterprises

Micro EnterprisesHousehold Level (informal Economy)

Conference Edition

128

managerial issues, inputs and production, marketing, and manpower), and it includes a review of 6 other research studies on MSME problems in Indonesia. For the purposes of this report, the BI study is the main source of descriptive information on MSME issues in Indonesia. The highlights of the study are summarized immediately below, by topic area of special interest to this report; the interested reader is referred to Bank Indonesia (2005) for extensive details.

BI Survey Results: Legal Issues & Business Development

• Most MSMES do not have an official legal status, especially at the micro level. • Most micro businesses don’t have business licenses (TDI/TDP and SIUP) or a Tax Registry

Number (NPWP); half of small businesses have them; most medium-scale businesses have them.

• Costs, procedure and administrative requirements are not obstacles to obtaining business licenses and NPWP.

• The main reasons for having business licenses and NPWP is to obey government regulations and to obtain credit.

• Few micro and small businesses have cooperation agreements with big businesses, but they can see the advantages.

• Few MSMEs believe that they benefit from existing government programs. • Their greatest needs from the government are credit, training and market information.

BI Survey Results: Financing Issues

• Most MSMEs transact in cash. Nonetheless, almost all micro and small businesses, and all medium businesses, have used banking services.

• In choosing banks, safety is their main consideration for deposits; for credits, the main considerations are the interest rate and location.

• The main obstacle for micro businesses is collateral; for small and medium businesses, it is high interest rates.

• Most micro and small businesses believe that banks allocate insufficient credit for MSMEs. Medium businesses find the allocations fair.

BI Survey Results: Banks’ Policies towards MSMEs

• To increase credits to MSMEs, commercial banks believe that better credit analysis is the internal key, including training of their staff. BPRs believe the key is supervising and collecting payments.

• The external key for banks is linkage programs with BPRs and cooperation with the credit guarantee institution.

Conference Edition

129

• For micro credits, most banks do not require licenses (SIUP, NPWP, etc.). For small businesses, requirements are tighter; for medium businesses, almost all banks require all types of licenses.

• For micro businesses, most collateral is movable (e.g. cars and motorcycles). Small and medium businesses mainly use land and buildings.

BI Survey Results: The Main Constraints

• Human resources development, including in finance, marketing and production. • Legal development, especially in the area of easier licensing. • The credit guarantor institution (e.g., offices, premiums, and claim procedures)

In reviewing these results, three main points stand out from the rest. First, problems of access to finance become more difficult as the scale of the business decreases, despite efforts at increased flexibility on the part of the banks. Second, existing government programs are not working well. And third, manpower weaknesses (on the sides of the MSMEs, banks and possibly government institutions) look to be at the root of many of the most basic problems.

V.1.2 MSME Lending & Policies in Indonesia

Notwithstanding the many problems mentioned in the previous Section, lending to MSMEs in Indonesia appears to have a relatively strong, established position in banks’ portfolios. Its share in total lending has been climbing in recent years, but may have stabilized around 50% (see Figure 49). This is just a bit lower than the 55% of GDP which they produce, according to the Minister of Cooperatives and Bank Indonesia. In addition, the quality of loans to MSMEs compares favorably with conventional loans, using the metric of non-performing loans (Figure 49, and a presentation by Bank Indonesia staff during field work for this report). This is consistent with the experience of BRI, which has 86% of their loan portfolio extended to MSMEs, while maintaining an NPL ratio that is a bit below the industry average (2.9% vs. industry’s 3.3%, as of September 2008).

Conference Edition

130

Figure 49: MSME Lending

Source: Bank Indonesia Source: Bank Indonesia; data for end 2008

For Indonesia’s policymakers, SME lending has long been on the policy agenda. For example, in 1992 banks were required to lend at least 20% of their credit portfolio to SMEs, although this policy was abandoned in 2001. Instead, banks were asked to establish self-determined targets, and to report against them (see Chapter IV; for other government programs, see Chapter II). In late 2007, the Government launched its People’s Business Credit program (KUR), which is guaranteed by PT Asuransi Kredit Indonesia (Askrindo) and Perum Sarana Pengembangan Usaha (SPU) up to a ceiling of Rp500 million per small business unit. In parallel, Askrindo and SPU were recapitalized (at a cost of Rp850 billion and Rp600 billion, respectively) to allow them to offer more loan insurance for banks in case their SME loans turn sour.

Despite its short lifetime, anecdotal evidence indicates that the KUR program is already encountering challenges.100 By way of examples, as of early 2009 only six banks101 were eligible for the guarantee program, which gives these six banks a considerable competitive advantage over others, most notably the BPRs. It is unclear how the guarantee program’s terms (e.g., coverage for first time, unbanked borrowers starting up new businesses) are being enforced. It is also unclear how the authorities can ensure that banks aren’t channeling bankable projects into the KUR program in order to benefit from the guarantee program. In responding to these points, BI notes that the Government is considering expanding the KUR program to include the Regional Development Banks (Bank Pembangunan Daerah; BPDs). Regarding enforcement of KUR terms,

100 The Government of Indonesia is currently undertaking a review of the KUR program. The KUR program has been doubled in size in 2009 from that in 2008. 101 These were Mandiri, BRI, BNI, BTN, a Sharia arm of a state bank and Bukopin. Notably, all are state banks except for Bukopin, which historically has been very close to Bulog, the state logistics company. It’s also notable that the BPDs (under consideration for expansion of the KUR program, as noted in the main text) are also state banks, because they are owned by the provincial governments.

Conference Edition

131

BI appears to rely heavily upon its Debtor Information System (SID) to check the borrower’s credit history and to see if the project has already received credit. In the course of its review of the KUR program, the government might consider allowing more banks to qualify for the program, including a good mixture of private banks and BPRs, not just state-owned banks. Also, the pricing structure of the guarantee could provide an incentive to increase the quantity and quality of small loans, and the terms should be designed ultimately to convince bank management that the guarantee is not needed.102 An exit strategy would also be useful.

As next steps, the Government could look at the other government programs run out of various ministries that attempt to channel financing into preferred sectors (see Chapter II.6). These programs are heavily dependent upon government financing and they crowd-out other micro-finance operations that are often struggling to be self-sustaining. Also, these programs’ subsidized nature invites corrupt administrative practices and encourages a culture of non-repayment. Like the KUR program, these would benefit from an independent, external assessment of their effectiveness. Participation could include local think tanks, industry associations and NGOs with special interests in the area.

Reflecting the ineffectiveness of these many programs, there are continuing controversies as to whether MSMEs have adequate access to finance. Proponents of this view argue ‘under-lending’ to MSMEs because they account for 99% of all business units and 96% of the employed labor force (Figure 50), but receive only around half of bank credits. Comparing output to lending by sector (Figure 50), SMEs operating in agriculture, construction, trade, transportation and business (financial) services are said to be particularly under-serviced. Proponents conclude that there are many untapped, potential investments that could be undertaken by MSMEs, if the financing were available. In this regard, they point to foreign-owned and joint venture banks as the main culprits;103 those banks account for 14% of total lending, but only 4% of lending to MSMEs. Proponents believe that an active SME lending policy would boost job creation and real sector growth in Indonesia.

This issue is partly explained by definitions; Bank Indonesia’s definitions do not match those of the Government.104 For practical reasons,105 Bank Indonesia currently defines small-scale credits by size of credit, not by size of the borrower. Consequently, roughly half of MSME credits are for

102 For example, the price of the guarantee could be proportional to the size of the MSME portfolio. And, the guarantee could only kick-in after a time threshold (say 3 months) when NPLs exceeded some pre-determined ratio (say 3%). 103 State-owned banks (SoB) are sometimes cited as neutral is this regard (see Figure 50), but this is only true for them as a group. The great majority of SoB lending to MSMEs is accounted for by BRI. 104 BI defines a micro credit as up to Rp50 million; a small credit is between Rp50 million and Rp500 million; and a medium-sized credit is between Rp500 million and Rp5 billion. (p. 79 of Bank Indonesia (2005)). According to BI staff, this definition is under review, with the intention of bringing it into line with the government’s definition. For GoI’s definition of an MSME, see footnote 116. 105 For example, it is very difficult to tell the difference between a motorcycle purchased for personal use and one that is purchased for small business purposes, like an ojek (a motor cycle taxi).

Conference Edition

132

consumption (that is, for consumer loans), not for investment.106 Effectively then, lending to MSMEs is only about half the amounts normally cited from official sources.

106 See Table 9.5 in Bank Indonesia (2007).

Figure 50: MSME Characteristics

MSME & Large Enterprise Business Units

MSME & Large Enterprise Employees

Breakdown by Loans Breakdown by Output

Breakdown of SME Lending by Type of Bank

Source: Bank Indonesia. Data for June 2007

Electricity, Gas, and Water

0%

Construction0%

Trade, Restaurants, and

Hotel27%

Transportation and

Communication6% Agriculture

53%

Services6% LEs

0.01%

Financial, Rental, and Business

Services0%

Manufacture Industry

7%

Mining1%

Electricity, Gas, and Water

0%

Construction1%

Trade, Restaurants, and

Hotel24%

Transportation and

Communication4%

Agriculture43%

Services11%

LEs4%

Financial, Rental, and Business

Services1%

Manufacture Industry

11%

Mining1%

By Lending

31%9% 20% 21% 31%

64%25% 30%

52%90%

69%91% 80% 79% 69%

36%75% 70%

48%10%

0%20%40%60%

80%100%

Agr

icul

ture

Min

ing

Man

ufac

turin

g

Elec

trici

ty,

Gas

, and

Wat

er

Cons

truct

ion

Trad

e

Tran

spor

tatio

n

Busin

ess

Serv

ices

Sosia

l Ser

vice

s

Oth

ers

(%)

SME LEs By GDP

87%

8% 13% 1%

44%75%

30% 17%40%

9%

3% 12%8%

22%

21%

24% 47% 8%

4%

89% 75% 92%

34%4%

46% 36% 52%

0%20%40%60%80%

100%A

gric

ultu

re

Min

ing

Man

ufac

ture

Indu

stry

Elec

trici

ty, G

as,

and

Wat

er

Build

ings

Trad

e,Re

staur

ants,

and

Tran

spor

tatio

nan

dFi

nanc

ial,

Rent

al, a

nd

Serv

ices

(%)

SEs MEs LEs

35% 34%

8% 14%

40%48%

14%4%3%

0%

20%

40%

60%

80%

100%

Total Lending SME Lending

Non-ForexCommercial Banks

JV and ForeignOwned Banks

Foreign ExchangeCommercial Banks

RegionalDevelopment Banks

State-owned Banks

Conference Edition

133

Another factor concerns the high overheads associated with conventional commercial bank lending, in part because they are subject to relatively strict regulatory oversight. More to the point, MSME lending is still risky unless there is strong loan supervision, which entails high administrative costs that can only be borne by institutions with a strong local presence. Even BRI, with one of the largest branching networks in the world, leaves the lower end of this segment to other players. For most conventional commercial banks—especially foreign and joint venture banks - this business is difficult, because they lack the requisite neighbourhood knowledge, relationships and trust. As technology brings costs down (see Box 56), the more aggressive commercial banks can be counted upon to move into these segments, because they will be attracted by the opportunities for profit. In the meantime, the real issue is how to bring down artificial barriers to faster progress; how to rationalize existing ineffective government programs; and how to find better ways to deliver financial services by institutions better-equipped to do the business.

V.2 Migrant Workers and Access to Finance

As mentioned in Chapter II, large numbers of Indonesians work abroad. In 2006, approximately 680,000 Indonesian migrant workers left Indonesia to work overseas, bringing the estimated total to about 4.3 million (World Bank, 2008). There are many issues associated with these migrant workers and this Section looks at two that fit well with the subject of this report. These are: some socio-economic characteristics of these migrant workers, using results of the household survey of Chapter III; and their difficulties in accessing financial services at different stages of their migration.

One striking aspect of this work is the disconnect between existing TKI financial products (e.g., insurance and savings) and the needs as expressed by the migrant workers themselves. In this context, it should be noted that the World Bank conducted a more extensive survey devoted expressly to migrant workers in late 2008. It will expand considerably upon the survey material of Chapter III.

V.2.1 A2F Findings on Indonesian Migrant Workers

This Section expands upon previous work (e.g., World Bank (2008a)), using the survey of Chapter II. This provides information on households that have a migrant worker living abroad, with a sample size of about 10% of total respondents (see the upper left panel in Figure 51). It is not a survey of migrant workers, per se, and for some sub-categories the number of observations is small, which reduces statistical reliability. With these qualifications in mind, the main patterns of these workers are as follows:

Conference Edition

134

Figure 51: Characteristics of Migrant Workers

How Many  T imes  Did  Y ou  Work  Abroad?(in %)

0.00 20.00 40.00 60.00 80.00 100.00

Malays ia

B runei

S ingapore

Hongkong

Taiwan

S audi Arabia

K uwait

Other

One time Two times Three times More than three times

T otal Res pondentsby Mig rant Worker Abroad

10.1%

89.9%

Migrant Non Migrant

Mig rant Workers by Gender

29.6%

70.4%

Male F emale

Mig rant Workersby L evel of Education  & Gender

(in% )

100.0

32.919.0

32.5 35.3

76.0

29.6

67.181.0

67.5 64.7

24.0

70.4

0.00

20.00

40.00

60.00

80.00

100.00

Never go tos chool

Did notcompleteP rimarys chool

P rimaryS chool

S econdaryS chool

S enior HighS chool

Univers ity Total

Male F emale

Conference Edition

135

• About 80% of the (legal) migrant workers are female, and more than 85% of them work in informal sector in relatively unskilled jobs, such as household domestics (e.g., maids and nannies; Figures 51 and 52).

• By education level, just over half (53%) have a primary school education or below, and males dominate the extremes, that is the well and very poorly educated (see the middle panel of Figure 51).

• 45% are repeat migrants, with Kuwait, Saudi Arabia, Hong Kong and Taiwan popular repeat destinations. Singapore is the least popular repeat destination (the lower panel in Figure 51).

Figure 52: Migrant Workers, by Type of Work and Legal Status

(in per cent)

The limited, available data suggest that many Indonesians work abroad illegally, which greatly complicates policy in this area.107 As illegal workers, they simply cross the border on foot or by boat, usually to a neighboring country, such as Malaysia. Results of this survey indicate that 10% are illegal workers, but that looks low based upon other available information.108 Most seem to be

107 Legal migrant workers are defined as those who register as a migrant worker to Ministry of Manpower and Transmigration, and hold a proper migrant workers’ passport and working permit in the destination country. 108 For example, see http://www.reliefweb.int where 600,000 were estimated to be in the state of Sabah alone in 2002. In 2005, the Government of Malaysia estimated that there were more than 1 million illegal migrant workers in the country, mainly from Indonesia and other countries in South and SE Asia (reported on http//burmalibrary.org). Joseph Liow (‘Malaysia’s Illegal Indonesian Migrant Labour Problem: In Search of

63.9

2.7

8.2

11.0

3.9

9.1

30.3

3.6

1.1

21.7

22.3

13.4

0.90.6

0.0 14.0 28.0 42.0 56.0 70.0

Domestic  Worker

Industry Worker

F ac tory Worker

P lantation  Worker

C onstruc tion  Worker

S ervice  Worker

Other

L egal Illegal

Conference Edition

136

men in Asia-Pacific destinations, dominantly Malaysia. It appears that they tend to work on construction projects, in plantations and service industries, including as domestic workers (Figure 52).

V.2.2 Migrants & Access to Financial Services

For purposes of this report, issues of migrant workers’ access to financial services are reviewed according to the different stages of the migration process, namely: i) pre-departure (pre-migration); ii) during migration; and iii) post-migration. This approach is useful because migrants’ needs are different at different stages of the process (see Figure 53).

Figure 53: Indonesian Migrant Workers and Access to Finance

Solutions’, on http://www.questia.com) writes that Indonesians into Malaysia are arguably the second largest flow of illegal immigrants in the world (after Mexicans into the United States).

Post-Migration Period

Migration Period

Pre-departure Preparation

Period

Recruited & Registered  

Destination Country with Employer 

Returning to Indonesia & the village PJTKIs 

NEED  CREDIT for: Placement Fees, Insurance, Transportation 

NEED  SAVINGS  

NEED SAVINGS  NEED INVESTMENT/CREDIT 

NEED CHANNEL FOR        REMITTANCES NEED SAVINGS for both TKIs overseas and for families in Indonesia 

Conference Edition

137

V.2.3 The Pre-Departure Stage

Registration with PJTKI. Legally registered migrant workers are normally recruited either by a PJTKI109 (a licensed Indonesian Overseas Employment Agency) or by agents of PJTKI, usually referred to as ‘sponsors’. PJTKI’s functions are:

• To receive the placement request from overseas clients; • To recruit migrant workers in Indonesia; • To prepare all necessary documentation for migrant workers; and • To provide training, especially for those who plan to work as household domestics.

Depending on the type of job and the timing of the job order from overseas, migrant workers usually stay at PJTKI’s training center (with accommodations for migrant workers) from a few weeks (for repeat migrants) up to a year, before departing for the destination country.

Credit Needs during the Pre-departure Stage. As soon as they decide to work abroad, either formally or informally, most migrant workers face an immediate need for credit. The credit is needed to cover the cost of PJTKIs, agents of PJTKIs and sponsors.

In the case of PJTKIs, there are a series of administration services and training programs from PJTKI. These include: the administrative cost of a passport; government levy and working visa for their destination country; medical examination fees; insurance; transportation; and the fee for PJTKI. The cost varies by the destination country, but generally ranges from US$335 to US$950 for legally registered migrant workers.110

109 Perusahaan Jasa Tenaga Kerja Indondesia, popularly known as “PT” Indonesia (PJTKI), now is Perusahaan Penempatan Tenaga Kerja Indonesia Swasta (PPTKIS). 110 Fees for securing plantation and construction job are between $335 and $560. Clearly, the cost for illegal migrant workers is cheaper than for legals, which is part of the incentive for migrants to choose the illegal route (World Bank, 2008a).

Conference Edition

138

Figure 54: Migrant Workers in the Formal Sector, Financed by Employers

For some PJTKIs, formal financial service providers (e.g., commercial banks, BPRs and overseas employers) do provide loan schemes to cover pre-departure costs. However, such opportunities are usually limited to those who work in the formal sector, for instance at a plantation or a factory. In those cases, the employer in the destination country usually makes monthly deductions from the migrant worker’s salary to repay the loan. Banks indicate that credit risks are too high for them to consider financing migrant workers’ expenses at this stage. They would consider financing through a co-signer, but normal commercial considerations would apply to the co-signer (that is, they would lend against the co-signer’s collateral, not the prospective income of the migrant worker).

The survey results of Chapter II indicate that the most popular way (20%) of financing these expenses is from personal savings, especially for repeat migrants (Figure 55). The use of savings for this purpose by repeat migrants is evidence that financial benefits do accrue to migrant workers, despite the many problems they encounter along the way. Other popular ways, especially among first time migrants, are: borrowing from the employer (17%); borrowing from family members (16%); and borrowing from the work sponsor (15%). Only 5% borrow from a bank, which highlights the banks’ limitations in servicing this group. Products more suitable to migrant workers’ needs would be very helpful.

Employer (Factory/Plantation) 

 

PJTKI 

Bank 

MWs 

Indonesia  Destination Country 

Deducted from the salary of migrant workers 

Sending the total deducted amount to Bank to repay the loan 

Guarantee to the Bank 

Loans covering placement fees for migrant workers 

MWs 

Conference Edition

139

Figure 55: How Migrant Workers Finance Pre-departure Expenses

As mentioned, the majority of Indonesian migrant workers are working in the informal sector (such as housemaids, baby-sitters or drivers) and there are very few loan products offered to these migrant workers by formal financial institutions. Despite the good income prospects from working abroad, the migrants are usually forced to obtain credit from informal providers who charge relatively high interest rates. From the banks’ perspective, these potential clients are very risky, because the loan recipient is physically abroad, which risks default without recourse by the lender. Also, the loan amounts are small, probably only Rp3-5 million, so the interest payments may not cover the bank’s administrative costs. This said, a few banks view migrant workers as a potential market and do have a special department to deal with remittances and migrant worker.

To assist with the shortage of credit for migrant workers, government of Indonesia through BP2TKI (The National Commission for Placement and Protection of Indonesian Migrant Workers), has initiated a credit scheme for migrant worker headed for Taiwan, in partnership with banks from Taiwan and Indonesia (Box 19). Interest rates and charges appear to be quite expensive.

0.00 20.00 40.00 60.00 80.00 100.00

S ale of as s ets

Us e of pers onal s aving

Borrowing from other family  members

Borrowing from friends

Borrowing from a work  s pons or

Borrowing from bank

Borrowing from employer

Other

Don't K now

F irs t Time Repeater

Conference Edition

140

Box 16: Case Study: Migrant Workers Bound for Taiwan

Indonesian migrant workers to Taiwan have access to finance through a mandatory credit scheme. It covers their placement fees, thanks to an initiative of the Indonesian and Taiwanese governments. PJTKI covers the placement fee for migrant workers going to Taiwan up front, in principle allowing Taiwan-bound migrant workers to register with PJTKI without having any initial credit. Prior to departure, migrant workers sign a loan document with one of several banks appointed by BP2TKI. (As of November 2008, these were: Hua Nan Commercial Bank; China Trust Overseas Bank; Sani Bank; First Bank; Bank Mandiri; BPR Tatakarya; and BPR Peranragmada). The loan document and job contract are submitted as supporting documents for visa application to the Taiwanese Embassy prior to their departure. Upon arrival in Taiwan, the migrant worker obtains the requisite local identification, assisted by the agent in Taiwan, who is usually contracted by banks. The migrant worker opens the bank account, with the signed loan agreement and other necessary documents. With proof that migrant workers are physically in Taiwan and have a legal status in Taiwan, the bank releases the loan proceeds to PJTKI, initially covering the cost of migration. Once migrant workers start working, the employer deposit the salary of migrant workers into the bank account, and the bank will deduct the agreed amount every month, usually for one year, to repay the loan.

It would be useful to review this credit scheme more closely, especially the prospects for transferability to other destination countries.

Conference Edition

141

Box 17: Perum Pegadaian (The State-owned Pawnshop) & Access to Credit Perum Pegadaian, the state-owned and only legal pawn shop in Indonesia, has a long history of providing lower income Indonesians with access to credit. It has a substantial presence in the Indonesian microfinance sector with around 900 branches nationwide, serving 16.7 million clients and providing Rp 22.8 trillion in loans in 2007. Migrant workers with certain physical assets are among those who use Pegadaian’s credit services. Once migrant workers are abroad, family members of the migrant workers pay the interest rate on the credit received from Pegadaian until the migrant worker remits earnings. Once family members receive the remittance, Perum Pegadaian is re-paid and the asset is returned. Example: The credit to cover migration costs: Rp 4 million The collateral: Gold jewellery Appraised value of the collateral: Rp 5 million Loan to be provided: 80% of appraisal (the percentage depends on the item): Rp 4 million Administration fee: 1% of Loan: Rp 40,000 (to be paid up-front) Interest Rate: 1.3% per 2 weeks: Rp 52,000 Maturity: 4 months Repayment by the end of 4th months: Rp4million + (Rp 4 million*1.3%*8) = Rp 4,416,000 Perum Pegadaian has recently established a corporate partnership with Western Union to provide money transfer services at their branches. This service started in 60 branches from early 2008. Perum Pegadaian also finances investment activities, such as farming, fisheries, and small scale industries and merchandising. Currently, these account for 25% of their customers. Pegadaian issues bonds which hold a AA+ credit rating by PT Pefindo, indicating a sound financial condition. Pegadaian is a government monopoly and a very profitable business. SourcPerum Pegadaian Annual Report, the Staff Field Interview at Perum Pegadaian Cibadak office, West Java, Indonesia.

Needs for Saving Products during the Pre-departure Stage. By government regulation, migrant workers are required to open a savings account at a commercial bank and to buy an insurance policy prior to their departure (see the continued discussion of insurance products a few paragraphs below). Some commercial banks, such as Bank Negara Indonesia (BNI), have special saving accounts for migrant workers. These require a very low initial deposit and minimum balance, with low monthly admin fees (Table 24).

Conference Edition

142

Table 23: Sample Savings Product Offered by Banks and the Post Office

Savings Product Initial Deposit

Minimum Balance

Monthly Cost Notes

e-Batara Pos Rp20,000 Rp20,000 Rp1,500 - Product of BTN Bank jointly operationalized with PT Pos Indonesia

- ATM card available Simpedes Rp100,000 Rp50,000 - Product of BRI Bank

- ATM card available (with an additional monthly cost of Rp 3,000)

Tabungan Mandiri Rp50,000 Rp50,000 Rp9,000 - Product of Mandiri Bank- ATM card mandatory - Minimum deposit Rp 10,000

BNI TKI Rp10,000 Rp250 - Product of BNI - Should be requested

collectively (minimum 10 persons) and coordinated or recommended by the PJTKI

BNI Taplus Rp200,000 Rp6,000 - Product of BNI - ATM card available

Source: Interview with banks & post office, Cirebon, September 2007, www.posindonesia.co.id, www.btn.co.id, www.bri.co.id, www.bankmandiri.co.id, www.bni.co.id.

PJTKIs normally have a partnership with a bank and an official of that bank usually provides a short (1-hour) pre-departure training program, including details on opening a special savings account and providing a bank book to migrant workers before departure.111 This saving account constitutes proof of migrant worker status at the airport, where the migrant worker is exempted from paying the exit tax (of Rp1 million, commonly known as ‘fiskal’).

There are limitations on the use of this saving account by migrant workers. For example, the account is created only under the migrant worker’s name, so that his/her family does not have an access to this account while s/he is abroad.112 Effectively, this prohibits use of the account for remittance purposes. Also, a minimum balance is still required in the account, and monthly admin fees can result in the account being closed by the bank while the migrant worker is abroad.

Insurance Products for Migrant Workers. As mentioned, legal migrant workers going abroad through PJTKIs are required to have an insurance policy for the duration of their time

111 Such savings accounts for migrant workers are not available at the branch level. The support given by bank officials is mainly administrative; there is no financial literacy training or explanation of bank products. 112 BRI recently started a pilot program for a migrant worker’s savings account, which can be accessed by his/her family members with little administration process. There are also special migrant workers’ loan products that can be extended to a migrant worker’s family, once migrant worker repay a certain percentage of loan. This is a promising step that needs to be investigated further, and possibly adopted by more banks.

Conference Edition

143

overseas, as well as during the pre- and post-migration periods. As of early 2009, five consortiums of insurance companies provided insurance products to migrant workers.113

By regulation, every departing Indonesian migrant worker must pay Rp400,000 for insurance, which is included in their placement fee.114 However, there are serious issues as to whether these insurance policies provide reasonable service to the migrant worker. For instance, migrant workers are normally unfamiliar with insurance products, and are often unaware of their coverage; they tend to treat these expenses as just an overheard cost of working abroad. Consequently, in case of accident, illness or death during the insurance coverage period, claims are rarely submitted by the migrant worker and his/her family. Even if migrant workers do submit a claim, the process is often too long or too complicated for migrant workers.115 As a result, the claim rate by migrant workers and their families is low, according to discussions with representatives of the insurance consortium.

V.2.4 Financial Service Needs During Migration

The typical duration of migration for Indonesian migrants is two years. However, if both the employer and the migrant worker agree, there is the possibility of extending the contract for another one or two years. During the migration period, the workers often send some of their earnings back to their family in Indonesia, probably having saved up for 2-3 months or more (Box 21).

113 The biggest is Consortium Jasindo with 7 member of insurance companies lead by the Asuransi Jasindo. Another is the Konsorsium Asurasi TKI “Mitra Dhana Atmharakhsa”, currently a consortium of six insurance companies (PT. Asuransi Ramayana Tbk., Jasa Tania, General Insurance, BNI life, PT Asuransi Asoka Mas, PT. Fadent Mahkota Sahid General Insurance and Megalife Asuransi Jiwa). It has 19 offices throughout Indonesia and 6 international offices, in Kuwait, Bahrain, Malaysia, Singapore, Hong Kong and Taiwan. 114 This covers a total of 30 months of insurance, including the pre-departure, placement and post-placement stages. In the pre-departure stage, the insurance program covers death due to accidents or illness during training period; permanent partial disability; and medical treatment due to illness or accident. In the placement stage, insurance will cover accidents during and outside working hours; medical treatment expenses; death due to accident or illness (including expenses for burial and corpse repatriation); unpaid wages; and premature contract termination by the employer. During the post-placement stage, the insurance covers death due to accident or illness; permanent partial and total disabilities; and medical expenses due to accident. (Manpower and Transmigration Ministerial Decree no 157/MEN/2003, dated 9 June 2003). 115 There are instances where the PJTKI assists the migrant workers with the claim, but this is rare because the PJTKI’s responsibilities end when the migrant worker departs Indonesia for the destination country.  

Conference Edition

144

Transaction Channels. Indonesian migrant workers use both formal and informal channels for sending money home (Box 22). The choice depends upon the status of migrant workers (illegal or legal); the employment sector (informal or formal); and the destination country. Other factors include: accessibility (physical accessibility, ease of remittance process, opening hours, etc.); affordability; trust; and the migrant worker’s financial literacy. Notwithstanding the noted weaknesses of the banking system (Figure 56), formal channels dominate.

Box 18: Remittance Method and Frequency from a Middle East Country

Source: Bank staff interviews and focus group discussions, Cirebon, September 2007, FMW is female migrant worker

- US$ 500 – 2000, mostly in US$ 

- Gold jewellery, carpet, blanket , electronics 

- US$ 200‐400 - Mostly in US$ 

- US$ 200 – 1500 - Mostly in US$ 

- Account transfer or draft cheque (via registered mail) 

- Sent once in 3 – 6 months or when there is request from family at home 

- Employers usually kept FMW’s monthly wages. FMWs usually asked for their earnings when they want to remit money home 

 

Remitting Method 

- Account transfer or draft cheque (via registered mail)  

- Sent on the 4th or 5th month 

- Transport from airport to village 

- Buy furniture - Oleh‐oleh for families & 

friends 

Mostly to pay back loan (to cover the migration cost) &  daily needs. 

- Pay parents’ or husband’s debt 

- Daily needs - Build/renovate house - Medical emergency - Schooling (brother/sister or 

children)  

24th  

21st  

20th  

4th  

3rd  

1st  

- Hand carry home in cash & in kind - Some of the cash were hidden or 

sewn to underwear for safekeeping 

 

Usage Month  Amount 

Conference Edition

145

Box 19 : The Methods of Remittance: Formal vs. Informal

With regard to gender issues in this area, female migrant workers in the Middle-East rarely remit money to Indonesia for social/cultural reasons. In that region, females are not allowed to go on the street unaccompanied by a male, and language is often a problem. As a result, they rely heavily on their employers to remit their salary to Indonesia.

Formal Remittances. The great majority of formal remittance inflows are through banks, with non-bank financial institutions accounting for the remainder (Figure 56). The overseas transaction does not require the sender to hold a bank account in the destination countries; the sender simply requests the transaction with the submission of certain documents, such as an ID card, to comply with Anti-Money Laundering regulations. The cost per transaction varies by institutions, by the amount remitted and the destination country (see Table 25). Further investigation into comparative costs by country would be very useful information, especially if the information could be made available to workers before going abroad.

Further issues related to formal remittances and mobile banking are discussed in Chapter V-3.

Formal Channel 

Informal Channel 

• Bank • Nonbank Financial Institutions 

- Western Union - Money Gram - Licensed Foreign Exchange Shop. - Licensed Money Exchange Dealers 

• Friends / neighbors / family members • Informal money exchange / foreign 

exchange dealers 

The methods of Rem

ittance (O

verseas money transfer) 

Conference Edition

146

Table 24: Indicative Cost of Remitting Money from a Middle-East Country

Sending Method

Cost Borne by Duration Notes

Remitter Recipient Hand Carry None None 1-2 days - In the form of cash, gold, goods

- Policy on maximum amount of currencies leaving a destination country applies

Draft Cheque 20 Riyal 6 Riyal for mail

- Revenue stamp - ID photocopy - Indicative ‘gratitude

tip’ for post office

2-4 weeks - Cheque is sent via registered mail - Need valid ID to cash out - Post office called or sms migrant

family for incoming registered mail Account Transfer

25 Riyal None 2-3 days Need bank account & valid ID

Western Union

Rp 100 000 to transfer home Rp 1 million

- Formally none - Indicative ‘gratitude

tip’ for post office

1 hour - Need valid ID & PIN code - Can be cashed out in any WU outlet

(banks, post offices) - For post offices that are not yet

online, PIN code confirmation is done through phone calls with the Cirebon City Central Post Office

Source: Interviews and focus group discussions, Cirebon, September 2007

Figure 56: Method of Receiving Remittances, by Host Country

(in percent)

3.4

9.8

10.8

7.7 13.6

9.8

87.0

77.0

95.9

72.3

84.8

61.6

74.6

78.2

2.51

8.1

8.1

25.2

14.9

4.1

8.9

15.2

7.1

1.2

1.9

2.7

5.8

1.7

3.8

2.1

0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0

Malaysia

Brune i

S ingapore

Hongkong

Ta iwan

S audi Arabia

Kuwa it

Other

Through a  Pos t Offic e Bank  c hec k  s ent via  a Wes tern unionBank  wire  trans fer Informal money travel Through relatives /friendCarries  own money on Other

Conference Edition

147

Informal Remittances. The use of informal channels for overseas transactions is common for Indonesian migrant workers in Malaysia, where they are often illegal. The informal channels include hand delivery, informal agents, informal money changers and shops. The process tends to be simpler and quicker although the financial costs are usually higher. Illegal migrant workers are likely to use such informal channels due to the lack of proper ID, which is required by the formal channel.

There are also risks in using the informal channel. For example, migrant workers returning to Indonesia may carry money (or gold or jewellery) on behalf of others from the destination country. Before the couriers return home, they may be robbed; they may be charged higher fees for transportation; or they may be the target for corrupt immigration officials who know that returning migrant workers usually carry the cash on their person. Informal money changers and agents may also charge higher commissions of the recipient and the family of migrant workers.

Box 20: Voices of Hard Experience “I worked in a kedai minum in Kuching. I lived in a 3 X 4 m room that I rented for 150 Ringgit a month. I kept all my valuables and money in my room. It is safe, but I was once robbed by my own friend. I lost 2000 Ringgit (Rp 5,667,020),a necklace and a ring. I keep some of my money with my tauke now. I must be careful not to keep too much money there. When it amounted to 400 Ringgit, I have to cash it out because if he holds too much of my money, it would be more difficult for me to cash it out. He would say that he doesn’t have enough cash to give me my money.”

Rosalia, Sahan Village

“I worked in a kedai minum in Kuching. I send money home through a friend or by going home. I frequently go home. I always exchange my Ringgit at Ahie’s store. When sending money through a friend, I usually send around 100 Ringgit (Rp 283,350). Sometimes I send Rp by exchanging my Ringgit in Kuching. The two rates are very similar. We could also exchange our Ringgit in shops at the border.”

Dedek, Sahan Village

“I used to work in a kedai kopi in Kuching. I was paid monthly in cash. I always keep my own money; I never opened an account in Malaysia. I married a Malaysian in 2007 and now live in Malaysia. I send 200 – 300 Ringgit monthly to my parents through friends. Sometimes I asked my brother to go to the border in Sirikin. I will meet him there and hand over the money. It doesn’t cost us a thing.”

Mega, Sahan Village

“I worked in a plywood factory in Miri. I just completed my contract and will shortly go back to Malaysia to work again.What concerns me is what’s happening on the border: people are being cheated by the preman who forces returning TKI to exchange their Ringgit. They ran away before giving you the full amount Rp that was promised based on an exchange rate that you both agreed on. And this only happens on the Indonesian side of the border. The Malaysian side of the border is OK.”

Tadi, Sei Pangkalan I Village The World Bank staff interviews in Kalimantan, December 2007.

Conference Edition

148

In December 2006, Bank Indonesia issued a regulation that allows informal transfers by remittance agents. The regulation aims to prevent money laundering by allowing a gradual transition from registration to licensing of these agents. According to the regulation, existing agents are required to register by the end of 2008, and new agents will need to apply for a license from the beginning of 2009. Agents who are registered and/or licensed are obliged to record money transfer transaction, submit periodical and incidental reports to Bank Indonesia and to report any suspicious transactions to Bank Indonesia (World Bank, 2008b).

Saving During Migration. Migrant workers working in the formal sector are likely to have a bank account in the destination country, because their employer may transfer their salary to the bank account instead of paying the salary in cash. These migrant workers accumulate their savings in the bank account until the amount is large enough to send back to Indonesia. In the case of migrant workers in the informal sector (e.g., a housemaid or babysitter), the employer may keep the earnings on behalf of the migrant workers until they are ready to send their earnings home. Or the migrant workers simply keep their money themselves, for instance hiding it in their room. Typically, the migrant worker has limited free time; hours of banks’ operations may be a problem; and they are usually unfamiliar with the process of opening a bank account in the destination country. Consequently, migrant workers usually do not open a bank account unless their employer does it for them, and transfers their salary to the account.

V.2.5 The Post-Migration Period

Based on the evidence of this study, remittances are generally not used for investment, with the notable exception of building or renovating a house. Rather they go towards meeting the daily needs of family life in the village, re-paying loans, school and medical expenses and so forth. After returning home, many migrant workers find that their remitted earnings have been fully used up. Repeat migrant workers are likely to accumulate some assets (such as some savings or a better home) but there is limited evidence that their post-migration livelihood has improved much on a self-sustaining basis. Indeed, less than half (41%) of respondents reported that they had saved money from abroad.

In part, this reflects the dire on-going needs at home as well as the capacity of dependent family members to consume transfers from abroad, virtually as fast as the migrant worker can remit them. But it is also due to a lack of financial planning and financial management. As mentioned, the great majority of Indonesian migrant workers are from poor rural areas, with a lower education background. These migrant workers and their families tend to have little knowledge of financial services and products. This problem of low financial literacy is one of key constraints recognized by the Indonesian Government to improve migrant workers’ livelihood. Still, there are very few financial literacy education programs for migrant workers. Such information as migrant workers

Conference Edition

149

obtain comes from their own experience, friends, family members, the employer, neighbors or other migrant workers in the same destination country. Rarely does it arise from any form of proper, formal training, which would assist with making the best choices among the available financial services and products. Uninformed decisions early in the process often appear to trap the migrant into unfavorable loan arrangements, spending the bulk of their earnings to repay loans, rather than spending to improve their livelihood.

V.3 Mobile Banking

V.3.1 The International Context

Branchless banking, 116 whether it is mobile banking 117 or card-based transactions, 118 has considerable potential to provide access to financial services to the ‘unbanked’. For purposes of this study, mobile phone technology is particularly interesting because it provides a method for the unbanked to access financial services easily, conveniently and securely through a popular, existing delivery channel. Customers no longer have to travel a distance or wait for a long time to conduct a financial transaction; they can do it with their mobile phone, perhaps followed by a visit a nearby cash-in and cash-out agent.

There are two basic models of branchless banking, bank-based and non-bank based; both deliver financial services outside traditional bank branches. In the bank-based model, every customer has a direct contractual relationship with a prudentially licensed and supervised financial institution. In the non-bank based model, the customer buys an electronic store of value, the current value of which is tracked on the server of a non-bank, such a mobile phone operator or telecommunications company. If the system is mobile-phone based, customers only need to visit a retail agent to top-up the stored value or to convert the stored value back into cash (i.e., ‘cash-out’). These stored value instruments are referred to as e-money; effectively, they substitute for cash.

Recent years have seen a boom in mobile banking operations in some developing countries. In Kenya (a country of roughly 39 million people, where fewer than 4 million have bank accounts), the M-PESA mobile wallet service offered by Safaricom attracted 1 million registered users in 10 months.119 In the Philippines, the country’s two mobile network operators offer the functional

116 The term ‘branchless banking’ refers to financial transactions conducted by means other than formal bank branches or their field offices. For example, the transaction may be done through post offices or common retail outlets, like grocery stores or gas stations. 117 There is some disagreement over the term ‘mobile banking’. Some experts define mobile banking as any financial transactions conducted with the use of mobile phones, regardless of who provides the services. Others argue that banks must provide the underlying service; they believe that, if telecommunication companies provide the service, it should be called an ‘m-payment’ or something similar. 118 Card-based transactions include debit card or prepaid card transactions using a device that can transmit data over a network such as electronic data capturer (EDC) or by using a mobile phone. 119 See http://www.safaricom.co.ke/index.php?id=745.

Conference Edition

150

equivalent of small-scale transaction banking to an estimated 5.5 million customers.120 For the unbanked poor in countries like these, conducting transactions using a mobile phone is far cheaper than going to a bank branch, in terms of both time and out-of-pocket expenses.

These Kenyan and Filipino examples illustrate telco-led mobile banking operations that specifically target the unbanked; they are known as “transformational mobile banking”. Variations on this model include a joint venture between a telco and a bank, or a technology company with a bank, such as WIZZIT in South Africa (Box 24).

Box 21: Promising Technologial Advances – South Africa’s WIZZIT

There are good economic reasons to believe that the formal banking sector will extend its services to remote regions, if their cost of providing those servicing is sufficiently low. One promising technological advance in this regard is cell phone banking, introduced into South Africa under the branding of ‘WIZZIT’. The service is premised on the simple observation that even the poor in remote regions of the country have access to a cell phone these days. (As indicated in Chapter III, almost 95% of Indonesian households have a cellphone.) In South Africa, branchless banking through retail agents is permitted only for licensed financial institutions. (Non-banks are prohibited from accepting public deposits.) Consequently, mobile operators interested in branchless banking have created joint ventures with licensed banks to offer cell phone-based banking. WIZZIT, a third party technology company, became a division of the South African Bank of Athens in order to be able to offer cell phone- and card-based bank accounts for the unbanked. WIZZIT’s resulting product is a cellphone-based banking facility whose immediate target market is the estimated 16 million unbanked or underbanked South Africans (about 60 percent of the country's population). The initiative was conceived in early 2002 and commercially launched in March 2005. Once successful, the company plans to expand into the rest of Africa, and the company reports enquiries from Latin America and Asia. Because it is bank-based, WIZZIT can provide a wide range of regular bank services (e.g., deposit, withdrawal, payment and airtime reload services) through a mobile phone interface, ATMs, bank branches and post offices.

WIZZIT is compatible with early generation cell phones (which are popular in low-income communities), and it uses pay-as-you-go cellphones. In addition, WIZZIT account holders are issued debit cards that can be used at any ATM or in the formal retail sector. WIZZIT charges per-transaction fees that range from 99c (US$0.15) to R4.99 (US$0.78); it does not charge a monthly fee; there are no transaction limitations; and it does not require a minimum balance. WIZZIT employs over 800 "Wizz Kids" - typically unemployed university graduates from low-income communities to promote the product and help unbanked customers open accounts.

Sources: The World Resources Institute, http://www.nextbillion.net/archive/activitycapsule/wizzit and http://www.wizzit.co.za/

120 See Lyman, Porteous and Pickens (2008).

Conference Edition

151

Box 22: International Experience in Reducing the Cost of Mobile Banking In developed countries, commercial banks use ‘direct banking’ technology (e.g. ATMs and InterNet banking) to process transactions at 20% of the cost of a traditional bank teller. In other countries like Brazil, the Philippines and Kenya, banks are experimenting very successfully with technology to reduce transaction costs further. In Brazil, banks use point-of-sale (POS) terminals to deliver financial services in nearly every province at 0.5% of the cost of a bank branch. POS terminals are electronic devices connected to a telephone or other telecommunications network and placed at retail outlets for payments and disbursements. The device may read debit or credit cards or barcodes, or the device itself may be a mobile phone that can accept information transmitted by another mobile phone through short messaging service (SMS) or another protocol. Will POS staff of a retail outlet, or a postal clerk, build sufficient trust with a client on behalf of the bank to sell a wide range of banking services to customers? Recent information from Brazil suggests that this may be difficult. Thirty percent of the accounts opened at banking correspondents of Banco Popular do Brasil (a division of Banco do Brasil) never become active. After opening for business in June 2004 and attracting 1.05 million customers within six months, the division now maintains about 771,000 active accounts and is closing unprofitable banking correspondents. In the Philippines, a ‘G-Cash’ product is proving popular. It links the user’s cell phone to a cash account. The customer can deposit or withdraw cash from this wallet as needed, and credit can be transferred between mobile users. Registration and transactions (including inward remittances from Filipinos working overseas) are done by SMS; initially cash is deposited at a designated location and withdrawals require an acceptable ID. The cost per transaction is the standard SMS fee of about US¢2. However, cash deposits are a bit more expensive; the fee is 1% of the transaction value with a minimum of US¢19. It is too early to know whether the use of technology channels will be profitable enough to encourage banks to target low-income customers. No thorough profitability analysis of replacing bank branches with mobile phones or POS devices at retail outlets is available. Although using POS systems to move transactions outside the branch environment for existing customers reduces costs, this approach may take some tome before it helps banks acquire customers who live far from bank branches. Also, regulatory issues must be addressed (see Box 27) Ultimately, transaction costs still need to come down significantly for lower income groups to manage their own daily financial lives. To this end, current thinking is that the system should work for transactions of as little as $2—the daily income for many of the world’s truly poor—on agent commissions of not more than 2 percent. That means that customers’ transaction fees should be in the range of US¢2–4 (1–2 percent on a $2 transaction). Only then will the size of electronic transactions parallel the daily cash flow of the poor. In the specific case of Indonesia, costs look like they could be a small multiple of this amount, considering Indonesia’s level of per capita income, say very roughly US¢5-10 per transaction. Sources: “Using Technology to Build Inclusive Financial Systems”, CGAP Focus Note #32, January 2006; “Realizing the Potential of Branchless Banking: Challenges Ahead”, CGAP Focus Note #50, October 2008; and World Bank ‘Report on Access to Finance in Pakistan” (draft).

Transformational branchless banking—which is mostly operated by telcos with a limited banking license—usually offer only limited financial services, such as cash-in and cash-out, payment and person-to-person fund transfers. They cannot offer savings. Some countries have gone further, only allowing telco-led branchless banking to offer a cash-in payment, which limits the growth of

Conference Edition

152

branchless banking. Nonetheless, this is sufficient for many unbanked customers, because branchless banking is used mainly for payments, not for savings or credit, according to a CGAP report referenced above. For instance, in Brazil almost 78 percent of the 1.53 billion transactions conducted at more than 95,000 retail agents/merchants are for bill payments and for the payment of government benefits to individuals. Similarly, and even though WIZZIT offers a broader range of services, most of its customers use the service to buy airtime twice as often as they withdrew funds from a branch or ATM, and five times as often as they conduct money transfers.

With the number of people connected to mobile networks greatly exceeding the number of people with bank accounts, there is large potential for branchless banking, especially mobile banking, to reach unbanked customers. The main questions are: will the cost/unit of this service be reduced low enough to attract the unbanked poor (see Box 25)? And will the unbanked poor actually use these services once they are readily available on a cost-effective basis? Public education (by industry players) and transfer of knowledge (by early adopters of branchless banking) have a significant role to play in this regard.

V.3.2 The State of Mobile Banking in Indonesia

Indonesia is moving ahead quickly with mobile banking services (see Box 26). For instance, Bank Indonesia is relatively advanced in its thinking about e-money, and Indonesia is among the few countries with regulations allowing non-banks to issue e-money. As this report was being drafted, BI had just amended its current regulations, and the main issues are discussed in Box 26.

As of end-2007, twenty-three banks offered different kinds of mobile banking services to their customers.121 On the telco side, two mobile operators have been licensed by Bank Indonesia to allow them to operate mobile payment services to their customers, albeit with some limitations.

Mobile phone penetration is quite high in Indonesia, at 37 percent compared to 8.4 percent of fixed-line teledensity.122 Twelve mobile network operators are now competing for market share in Indonesia, with the big three (Telkomsel, Indosat and Excelcom) accounting for about 85% of the market with a total of 75 million customers at end-2007.

121 Information gathered in www.informasi-media.blogspot.com. 122 Paul Budde Communication. 2008. “Indonesia – Key Statistics, Telecommunications Market and Regulatory Overview.”

Conference Edition

153

Box 23: Mobile Banking Product Development in Indonesia

Among the commercial banks in Indonesia , Bank BCA and Bank Mandiri are the two biggest mobile banking operators (in term of number of users, number of transactions and volume of transactions). In reaching out to the unbanked, Bank Permata is probably the most innovative. Its customers can send money to an unbanked person with a mobile phone and that person can retrieve cash from a number of Bank Permata ATMs by pushing a button and keying-in a few data received via mobile phone. BCA has also launched Flazz, a prepaid electronic wallet targeted at young customers to offer an easy and convenient way to conduct transactions. Using radio frequency identification (RFID) technology and taking advantage of its vast ATM network, Flazz users only need to tap the card into an electronic reader (without PIN numbers, as would do done with a debit card). Flazz carries a limit of Rp1 million, or US$100, the same as other prepaid limit set by Bank Indonesia. Flazz can be reloaded in BCA branches or from any ATM. Telkomsel, the largest cellular operator in Indonesia, launched its mobile payment services in late 2007 called T-cash. Telkomsel customers can register to be a T-Cash user over the network or by coming into one of Telkomsel’s many centers or retail agents. The maximum wallet size is US$100, which looks small for purchasing and fund transfer. T-cash can only be used to cash in or convert cash to electronic money and to make payments or purchasing. Cashing out and person-to-person fund transfers are not allowed. Telkomsel signed up Fuji Image Plaza (a chain of photo processing kiosks) to be their first retail cash-in and retail agents. Subsequently, Telkomsel has been working with Indomaret, a chain of hundreds of mini markets, as retail agents. One year after its introduction, T-cash has approximately 80,000 users with some 500 retail agents. However, compared to the number of Telkomsel’s subscribers (55 million at end-2007), the number of T-cash adopters is quite small. Reasons include: a low limit of wallet size; restrictions on m-payment activities; and slow roll-out of retail agents that can handle cash-in and purchasing.

Indosat, the second biggest cellular provider in the country, also received a license to operate m-payments, but has yet to find a good strategy to reach potential customers. They only recently launched their m-payment service (called Dompetku, or My Wallet), and progress looks slow.

Bank Indonesia regulates all banking-related activity including mobile banking, whether it is banking- or telco-led, and banks have greater flexibility in offering services because they carry a full banking license. Under Bank Indonesia regulations,123 commercial banks in Indonesia can offer almost as many services over the internet and mobile banking platforms as they would over the bank branches, except opening a bank account.

123 Bank Indonesia Regulation (PBI) No. 9/15/2007 emphasized the importance of risk management in the use of ICT by commercial banks. This regulation governs, among others, the use of ATM, phone banking, electronic fund transfer, internet banking and mobile phone banking.

Conference Edition

154

Box 24: Weaknesses in the Regulatory Environment for Branchless Banking in Indonesia Despite Bank Indonesia’s relatively advanced thinking about e-money (described in the main text), there are few examples to date of the successful provision of financial services to low-income Indonesians through branchless banking. Some reasons are related to existing regulations, as summarized below. Policy recommendations are provided in Chapter VI. First, until very recently the regulatory regime limited the use of e-money to making retail payments. This happened because BI’s regulations treated e-money the same as prepaid card payment instrument, like credit, debit and ATM cards. BI changed its regulations in April 2009 (BI Regulation 11/12/2009), and the result appears to be a considerable improvement. However, the regulations do not secure the preferential claim of agents and customers over issuer creditors (for instance, in the case of bankruptcy of the e-money issuer). Also, non-banks are allowed to issue e-money, but the use is restricted to retail purchases. Second, any non-bank e-money provider that wants to extend its services to person-to-person (P2P) transfers (domestically or internationally) needs a remittance license from BI, and eligibility requirements are very tight. (For example, potential providers may need to change their articles of association, which may be difficult.) Effectively, the eligibility requirements are restricting new entry. Third, neither banks nor non-banks are currently allowed to use agents to provide financial services, like cash-out transactions. A network of agents is vital to operate on a sufficiently large scale to make a low-value transactions business commercially viable. Fourth, current Know Your Customer (KYC) regulations require banks that provide electronic banking services to meet the prospective customer, at least when the account is opened. This poses a barrier because sign-up of new customers can only be done in bank branch offices (or mobile branch offices). Effectively, banks cannot out-source their KYC procedures, which precludes customer acquisition beyond the reach of bank branches. Likewise, mobile banking cannot leverage off agents’ networks (or telecos) to sign up unbanked customers who are unable or unwilling to visit a traditional bank branch.

Source: ‘Diagnostic Report on the Legal and Regulatory Environment for Branchless Banking in Indonesia”, June 2009, by CGAP in cooperation with IFC and GTZ.

Looking at both the bank-led and telco-led model under current regulations, bank-led mobile banking looks more likely to provide a wide range of services to unbanked users. Nonetheless, most banks seem unable to extend these services beyond their existing customer base, partly because of regulatory issues (see Box 27). One of the solutions may lie in the Sharia banking system (see Box 6); important innovations are taking place in this segment and--because it’s Sharia-based-- some previously unbanked segments are being attracted by the service.

V.3.3 Mobile Banking to Improve Access to Financial Services

By contrast, the Telco-led model offers greater potential for reaching the unbanked poor in rural areas because of existing wide mobile network coverage. But the services don’t meet the needs of the poor, specifically no cash-out; no person-to-person fund transfer capabilities; and small wallet size.

Conference Edition

155

With Indonesia’s wide geographical spread, the main issue is how to reach the unbanked poor in remote, rural areas by cost-efficient means. Because of its reach, mobile banking offers much more promise than either internet banking or the costly brick and mortar of building additional branches. Telkomsel (which has applied to be a remittance agent) claims that its network has covered almost all districts (kacamatans) in Indonesia. The other players are currently expanding their networks rapidly, including by sharing base transceivers in the rural areas. These developments look likely to further reduce service providers’ costs significantly, especially for smaller players that want to focus their reach into rural areas.

On the side of customers, the price of handsets has dropped in recent years, making mobile phones more affordable for many low income people. These days, many operators even offer a bundling program, whereby the customer gets a low cost handset if s/he buys Rp300 thousand (US$30) worth of airtime.

These factors, of course, have to be supported by the regulatory environment (see Box 27) and by the services on offer, especially their convenience in conducting mobile banking transactions. Most people want the ability to cash-in and cash-out; send money to their families; make balance inquiries; reload airtime; and purchase goods. Considering the popularity of short messaging services (SMS), if banks or telcos were to offer mobile banking services using SMS as the transactions mechanism, there could be a huge take up, including from the unbanked poor segment of the market. Further work is needed to better access the prospects, including for helpful interventions of one form or another.

Remittances Using Mobile Banking. In some countries, person-to-person fund transfers are allowed under current regulations. For example, in the Philippines, remittances are one of the driving forces behind the success of mobile banking. Filipino migrant workers in Singapore, Hong Kong and the Middle East send millions of US dollars back home every month.

The same situation could unfold in Indonesia, if the regulatory framework permits. Currently, Telco-led mobile banking players are not allowed to conduct person-to-person fund transfer; only bank-account-to-bank-account transfers are permitted. Maxis, a cellular operator in Malaysia is able to provide remittance services to Indonesian migrant workers in Malaysia and Singapore. Indonesian migrant workers in the two countries can use Maxis services by sending money to their relatives providing their relatives have bank account in one of 5 banks that cooperate with Maxis in Indonesia.

At drafting of this report, two operators in Indonesia were working on a pilot with an Indonesia banks to allow Indonesian migrant workers in Malaysia and Hong Kong to remit money using Indonesian mobile operators. Beyond that, the mobile operators would need to build a chain of cash-out agents to allow for easy and convenient cash withdrawal. Under current circumstances, it looks quite difficult for small shops to apply to become remittance agents. To accelerate the

Conference Edition

156

process, there may be a role for a pilot program of some type, possibly in the form of a public-private partnership.

V.4 Policy Issues

MSMEs and Access to Finance. Unlike other parts of this report, the MSME issues of access to financial services are virtually one-dimensional, that is, they are almost exclusively credit-related. On the basis of available evidence, these problems of MSMEs get more difficult as the scale of the enterprise declines, with problems of access mainly arising at the micro level. At the micro-enterprise level, there are many problems, only one of which concerns their access to finance. Resolution of access-to-finance issues would certainly help micro-enterprises, but other, more basic problems need to be addressed, too, like manpower constraints.

Indonesia has been pro-active in MSME policies in this area for many years, but the past emphasis has been on subsidized credit programs. In general, Indonesia has not been well-served by most of these programs, and some have already been abandoned. In a welcome step, the Government is undertaking a review of one of these, the relatively new KUR program. As part of its review, the Government might consider allowing more banks to qualify for the program (including more private banks and BPRs); pricing the guarantee to provide an incentive to increase the quantity and quality of small loans; and to design an exit strategy. Depending upon the outcome of this assessment, the government could review similar programs.

Another useful step forward would be for drafting of the Micro-Finance law to re-gain some momentum under strong leadership from key stakeholders such as the Coordinating Ministry of Economic Affairs, Bank Indonesia, or the Ministry of Finance. As part of the process, linkage programs between commercial banks and BPRs could be expanded to include non-bank MFIs, and it would be helpful if a similar role could be defined for NGOs. At present, there is no obvious vehicle for linking NGOs with MFIs (or lower-end BPRs), even if the NGO is prepared to take an equity position.

Migrant Workers and Remittances. Many policy issues in this area concern practices in other countries, but many are a legitimate topic for Indonesia to raise in international for a. For example, to reduce foreign barriers to exports of Indonesian labor services in the course of international negotiations for trade in services, especially in bilateral and regional agreements, or in negotiating its Memoranda of Understanding on Migrant Workers with host countries. The objective should be to better balance protection of the interests of the workers themselves with interests of employers and recruitment agencies. World Bank (2008b) makes several practical suggestions in this regard, regarding for example, acceptable forms of identification that do not constrain worker access to formal sector services and exempting from formal identification requirements for small transfers (as done in the Korea-Mongolia migrant worker corridor). World Bank (2008b) also recommends: the development of customized products for Indonesian workers during their three phases of migration (for both the workers and their families back in the villages);

Conference Edition

157

and regulating the informal providers, while increasing the competition among them and maintaining workers’ accessibility to informal providers.

To make significant, sustainable progress, the main economic issue is how to convince the commercial banks that this business is sufficiently profitable. One practical possibility would be to explore possibilities for wider use of domestic guarantors (or co-signers) for loans to migrant workers. Indeed, international institutions or NGOs with particular interests in these workers might consider a pilot project to act as the guarantor, in a public-private partnership. The economic objective would be to convince private participants that this really is a commercially viable business.

Another possibility (to be explored in the Bank’s follow-up study on migrant workers) is whether Indonesian banks are missing a commercial opportunity by not expanding their overseas financial services to migrant workers beyond remittances. For example, it would be useful to know the distribution of migrant workers’ net incomes (that is, their gross overseas salaries less overseas expenses like travel and admin costs) and their number of dependents back in the village. This distribution could be compared to the income distribution of bank account holders in Indonesia as one measure to assess their individual commercial potential. Overall market potential would also need to consider other factors, like: their total numbers; their geographical concentration in the overseas market; and possibly their income distribution relative to other residents in the overseas market.

Anti-money laundering (anti-terrorism) regulations are another barrier to international remittances, because migrant workers often have difficulty meeting Know Your Customer (KYC) regulations. World Bank (2008b) advocates a well-targeted financial regulatory framework (in Malaysia and Indonesia) that ensures remittances are not used for illegal purposes, to balance with better instruments and services that attract migrant workers. As a practical first step, the government could look at negotiating windows of exemption (or minimum documentation) for small transfers, possibly beginning with countries that are not high-risk conduits for terrorist financing (e.g., Singapore, Hong Kong and Japan).

Mobile Banking. The policy issues in this area mainly concern managing prudential risk and finding low-cost ways to deliver services that lower-end consumers want. Very recently, Bank Indonesia has made regulatory progress in some areas, but more needs to be done. For example, potential non-bank e-money providers are subject to special licensing and regulatory regimes, and eligibility requirements look like they are restricting new entrants. Also, to deliver mobile banking services cheaply, the economies of scale offered by a network of retail agents looks vital in order to get unit costs down to a commercially viable level. This may entail allowing banks the discretion to out-source services to non-bank third parties, while holding the banks responsible for agent activities. For mobile banking to reach deeply into the ‘financially unserved’, there are also costly KYC issues to be addressed. For example, simplified KYC requirements for low-risk, low-value accounts (and transactions) to permit the remote opening of bank accounts in isolated areas,

Conference Edition

158

and they would allow agents to facilitating the opening of new accounts. Policy recommendations in all these areas are provided in the following Chapter.

Conference Edition

159

Chapter VI: Policy Recommendations for Improved Access to Financial Services

This Chapter pulls together key policy recommendations from the diverse material of the preceding chapters of this report, offering practical suggestions for improving access to financial services by low income households.

VI.1 General Strategic Matters

The Government of Indonesia has also placed high importance on increasing access to financial services for a greater segment of the population and MSMEs, acknowledging this as a key constraint to development. The evidence of this report indicates that it’s important for policymakers to maintain certain broad, strategic objectives when addressing issues of access to financial services. Among the most important is broad-based policies to raise economic growth and incomes – higher incomes are the single biggest determinant of access to financial services. Likewise, it’s vital to maintain financial sector stability in order to ensure confidence in the system for the population at large. Policies aimed at increasing access to financial services are likely to be much more effective if based on the foundation of policies aimed at sustainable growth and financial sector stability.

Policy makers’ focus should be on broader access to all types of financial services for the lower-income and poor as opposed to a narrow focus on access to credit. Credit is important for the poor, but savings requirements rank much higher. Much of the lower-income segments find existing financial products to be inappropriate to their needs. Designing and pilot testing appropriate products through partnerships could potentially open up more customers for the formal financial sector and vice versa, provide access to the formal financial system for a greater share of the Indonesian population. Wider access to financial services by lower-income Indonesians needs both public and private sector interventions as well as some innovative public-private partnerships.

Commercial banks dominate Indonesia’s financial system, but survey results indicate that they currently service a relatively small proportion of the population. Most Indonesians still rely on the informal sector for financial services. Expanding the formal sector’s reach should be a priority; the key will be to reduce unit costs to the point where relatively expensive formal-sector services are available at affordable prices to low-income Indonesians, who often live in remote, costly locations.

Regulatory reform, detailed below, offers a low cost solution to addressing many of these issues. In refining this strategy, commercial banks (especially those with a wide regional reach) have an important role to play because they will be the aggressive entrants into new, profitable markets. Also, they are most likely to introduce new technologies (like mobile banking) and they will put competitive pressure on other providers to improve their services and reduce prices. But that said, they are unlikely to be the primary answer to addressing the short- to medium-term issues of better access to finance, because the commercial banks do not reach deeply enough into poorer, more

Conference Edition

160

isolated parts of Indonesia. As mentioned, the informal sector currently supplies most Indonesian lower-end households with their financial service needs.

This report finds that improved access to financial services by lower-end households would be better served by working on the legal and regulatory framework for BPRs, MFIs, cooperatives and pawnshops, including as regards mobile banking which holds considerable promise for reducing costs and extending reach for all service providers. In considering targeted interventions, survey results indicate that the target for improved access to financial services should be the rural, uneducated poor who live off Java. Off Java residents are more than twice as likely to have neither a bank account nor a loan, than are on Java residents.

VI.2 Regulatory Issues

Turning to more specifics on the regulatory front, this is where sizable, transparent gains can be made without recourse to sizable amounts of public financing. Some practical suggestions are offered below, by major topic area.

VI.2.1 Mobile Banking

This area holds great promise for extending access because of the low unit costs involved and the existing widespread use of mobile phones, even among the poor. However, a necessary condition for success is a regulatory framework that permits commercial agencies to operate on a large scale. Indonesia’s thinking in this regard is relatively advanced, and very recently further improvements have been made.

But more hurdles need to be overcome. For instance, non-bank service providers can issue e-money, but only for payment purposes. If they want to offer person-to-person services, they need a remittance license and eligibility requirements are currently an (unintended) barrier to entry. BI needs to define eligibility in simpler ways. Also, the regulatory framework should allow e-money providers to use networks of agents to offer and maintain their products. Furthermore, KYC regulations should allow agents to sign-up new customers, at least those below a relatively low threshold.

VI.2.2 Commercial Banks

For the time being, one important instrument for broader access, namely new entry, is of limited applicability for commercial banks, owing to current policies that emphasize consolidation of numbers and larger size of institutions. However, there are other promising avenues, most notably mobile banking, already mentioned. Beyond this, some other steps would be helpful. For example, monthly administration fees are often a problem for small accounts and the absence of an official policy on dormant accounts seems to be contributing to banks’ monthly admin fees. Easier policies for banks to close-out inactive, non-zero accounts could be offered to banks. Basic banking services

Conference Edition

161

- ‘no frills accounts’ – with low administration fees could be encouraged as bank Indonesia is already doing and some banks are already offering this. Recent regulations concerning relocations of bank branches and ATMs look unnecessarily restrictive. More general descriptions of location should suffice, thereby giving banks more flexibility to adapt quickly to local market conditions. It would also be useful to ease official regulations on branching, at least to bring them into line with Bank Indonesia’s current, relatively relaxed approach to implementation.

VI.2.3 BPRs

Large gains look possible in this area, especially for the all-important small BPRs that operate in remote locations. BI seems to be re-thinking its policies in this area, possibly moving towards stratifying its regulation of BPRs according to size of the BPR. This would be a welcome step and looks like the ideal opportunity to move away from the current one-size-fits-all-BPRs approach to regulation.

By way of useful regulatory changes, there could be a lower tier of minimum start-up capital for small BPRs in remote locations. Foreign investors and NGOs could be allowed to take ownership positions in larger BPRs that are looking for capital. Likewise for small BPRs, especially those in locations without adequate communications services, some reporting requirements that may be costly and ineffective could be eased. Furthermore, written disclosure requirements could be waived in areas of low financial literacy, to be replaced by oral briefings of new (or potential) customers, including in the local language, if necessary. Moreover, the possibility of easing KYC regulations for opening small accounts and waiving the requirement of taxpayer numbers for small loans would make formal sector accessible for larger numbers of poor and MSMEs. Finally, for the sake of regulatory transparency, the current tight branching requirements could be brought into line with the spirit of BI’s liberal approach to implementation.

At the same time, BI’s supervision capacity for BPRs looks over-stretched, despite strong efforts to strengthen the function. BI might consider additional, temporary assistance by contracting firms that specialize in micro-finance, similar to the way it has out-sourced supervision of the former BKDs to BRI.

VI.2.4 LPS

The Indonesian Deposit Insurance Corporation, LPS, has an important role to play in providing security for depositors, because bank deposits are the single most important financial service and because ‘security’ is bank clients’ main concern. Since it was formed in 2005, LPS has adequately demonstrated its capacity to reimburse insured deposits and close BPRs, without loss of public confidence in Indonesia’s financial system. At this juncture, the main policy requirement is continued adequate financing, so that it is able to continue to play an effective role. This is particularly necessary in the event that LPS gets involved in the larger, more expensive commercial

Conference Edition

162

bank system. Also, it would be useful to find better ways of informing depositors of the limits to their coverage, especially in regions of limited financial literacy.

VI.2.5 Cooperatives

There are a large number of issues related to cooperatives, the most important of which appear to be prudential. These need to be addressed on a timely basis, as a defensive measure against memberships losing their existing access to financial services. Concurrently, there needs to be an upgrading of the Ministry’s regulatory and supervisory capacity. If necessary, this could include temporary outsourcing of the function to firms specializing in micro finance. In due course, the government could also consider moving regulation and supervision of deposit-taking cooperatives to a consolidated supervisory authority (OJK) and providing the members with deposit insurance under LPS.

VI.2.6 Pawnshops

The only legal pawnshop in Indonesia is currently a profitable, state-owned monopoly. This area should be officially opened up to the private participation. In parallel, the Ministry of Finance would need to upgrade its capacity for supervision of pawnshops.

VI.2.7 Other Financial Institutions

The most productive way forward for this diverse group would be restoring momentum to the drafting of a new Micro-Finance Law, and encouraging public debate during the process. It will be important that the new Law emphasize facilitation and access. Also, it would be very helpful if their reach could be extended further, beyond their traditional base of Java and Bali. Implementation of the recent joint decree regarding micro-finance institutions would ensure continued access to financial services on the part of those who need it most, namely, the rural poor.

VI.3 MSMEs

Indonesia has a long history of providing support to MSMEs. However, policy needs to shift further towards more effective delivery vehicles, including wider access to financing. The process has already started with GoI initiating a review of its KUR credit guarantee program. Depending upon the results of the KUR review, GoI might consider independent, external assessments of its remaining subsidized credit schemes. These could be conducted under the auspices of the Coordinating Ministry of Economic Affairs or the Ministry of Finance and might include input from local universities, think tanks, the private sector and other stakeholders.

Conference Edition

163

VI.4 Overseas Workers and Remittances

The Government could help migrant workers have greater access to financial services by taking up certain key issues in negotiations of bilateral (or regional) trade in services agreements and Memoranda of Understanding on Migrant Workers, which are negotiated with several host countries. The objective would be to better balance the interests of the workers themselves with the interests of employers and recruitment agencies. Details could focus on issues like: acceptable forms of worker ID; better facilities for workers’ remittances; and exemptions to KYC regulations for small remittances. Initially, it might be necessary to focus these negotiations on countries that are low-risk terrorist threats, for example Hong Kong and Taiwan.

To provide better access to finance during the pre-departure phase, banks could make greater use of co-signers, probably family members who remain resident in Indonesia. Indonesia’s development partners with special interests in this area might consider acting as the co-signer for pilot projects to demonstrate the commercial attractiveness of this market in some form of public-private partnership. During the pre-departure phase, financial literacy training looks like it would be helpful, as would financial counselling services in the host country. GoI should provide such services, but donors or NGOs may have to take the initiative, at least initially. Back in the village, returnees might be recruited to assist with pre-departure training.

VI.5 Matters of More Limited Concern

Based upon the evidence of this report, note should be taken of several areas because they are currently matters of lesser concern for policymakers. This is important because many of these areas are considered problematic in other countries, or because they are generally believed to be problematic in the Indonesian context. These include the following points:

• The number of banks has fallen markedly in the past decade, but the reach of financial services has still increased significantly (due to ATMs, branching and expansion of lower-end financial institutions).

• Physical access is not a generalized problem. The exception is rural off-Java, especially if travel entails water transport. On average across Indonesia, the availability of financial services is actually better than access to key public services, health and education.

• Credit is not the main financial product being demanded by low income households. A savings account is the single most important financial service sought by Indonesians. Credit from banks ranks well down the list. Other forms of credit are widely dispersed among service providers.

• Lack of collateral does not appear to be a widespread problem. It is problematic for some banks, but banks are not the main source of credit. Some banks, like the MFIs, already lend against income and not collateral.

• Gender is not a significant issue in almost all areas covered by the survey of Chapter III.

Conference Edition

164

Table 25: Key Policy Recommendations on Improving Access to Financing

(by policy area, with priorities indicated by grey highlighting) Area Policy Recommendation

General Strategy

Emphasize access to financial service broadly as opposed to a narrow focus on access to credit Focus primarily on broad-based policies to raise incomes

Focus on the legal & regulatory framework, especially mobile banking, BPRs and MFIs. For targeted interventions, concentrate on the rural, uneducated poor in off-Java regions.

Ensure financial stability and depositor protection, because bank deposits are the single most important financial service and 'security' is clients’ major concern.

For severely under-serviced areas, consider a policy of defining minimum service standards and out-sourcing provision of those services to private providers, using competitive bidding that is funded on-budget and supervised by stakeholders. Legal & Regulatory Framework for: Mobile Banking Ease licensing & eligibility requirements for non-bank service providers to issue e-money. Allow banks and non-banks to use networks of agents to offer financial services.

Ease KYC regulations which require banks to meet prospective customers, at least when an account is opened. Allow agents to sign-up new customers.

Conference Edition

165

Key Policy Recommendations on Improving Access to Financing (continued) (by policy area, with priorities indicated by grey highlighting)

Area Policy Recommendation

Commercial Banks

Formulate an official policy on dormant accounts, to give banks more flexibility to close out inactive, non-zero accounts. Encourage basic banking account or 'no frills' accounts, which would include very low monthly admin fees

Loosen regulations on relocations of branches and ATM machines to give banks more flexibility in meeting clients' needs.

Ease regulations on branching, to bring them into line with BI's relatively relaxed policy on implementation.

BPRs Implement the proposed 'tiering' system for regulation, with reduced reporting and management requirements, especially for small BPRs in remote locations

Allow partial ownership by foreign investors and by NGOs in poor remote regions. Consider lower minimum start-up capital requirements for small BPRs in remote locations. Ease reporting requirements for small BPRs in regions with poor IT infrastructure.

Ease written disclosure requirements in regions of low financial literacy, replacing them with oral briefings to new account holders in the local language.

Ease KYC regulations, especially as they apply to potential clients’ taxpayer numbers. Ease official branching requirements, to bring them into line with the current spirit of BI's implementation policy.

Conference Edition

166

LPS Ensure that adequate financing is available for prompt resolution of BPR/bank failures. Find better ways to inform depositors of limits to LPS's coverage, for example, a public education program.

Key Policy Recommendations on Improving Access to Financing (continued) (by policy area, with priorities indicated by grey highlighting)

Area Policy Recommendation ---Other MFIs Restore momentum to the drafting of a Micro-Finance Law, which needs to emphasize facilitation and access. , In the meantime, implement the joint decree on micro-finance to ensure continued access to services. ---Pawnshops Officially open this sub-sector to private ownership. ---Cooperatives Address issues of prudential regulation and institutional weaknesses to protect members’ existing access to financing. Up-grade supervisory capacity. In the short-term, the function could be out-sourced to firms specializing in micro-finance. MSMEs Shift policy towards more effective delivery vehicles than subsidized credit programs. Follow the review of KUR by an independent assessment of remaining GoI subsidized credit programs.

Conference Edition

167

Ease policies as regards identification requirements for borrowing from banks & BPRs, particularly as concerns taxpayer numbers.

Key Policy Recommendations on Improving Access to Financing (concluded)

(by policy area, with priorities indicated by grey highlighting) Area Policy Recommendation Overseas Workers & Reduce other countries' impediments to remittances during the negotiation of Memoranda of Understanding on Guest Workers or bilateral (or

regional) trade agreements on services Remittances

In these forum, seek better balance between the interests of the workers themselves with those of recruiters ands employment agencies. Details could focus on issues like: acceptable worker ID; limits on KYC reporting; and ID exemptions for small remittances.

Investigate possibilities for greater use of co-signers of pre-departure loans for migrant workers. Initiate pilot projects in partnership with banks and (as necessary) development partners

Pre-departure training needs to include briefings on access to financing; NGOs in popular destination countries might consider offering counselling services to migrant workers; returnees in home villages might be another low-cost source of assistance.

Conference Edition

168

Annexes Draft

Annex A: Some Technical Details of the Survey Annex B: Predictors of Financial Participation; Linear Probability Regressions for Savings of

Any Kind

Annex C: Predictors of Financial Participation; Linear Probability Regressions for Formal Savings

Annex D: Predictors of Financial Participation; Linear Probability Regressions for Informal

Savings Annex E: Predictors of Financial Participation; Linear Probability Regressions for Loans from

Any Source Annex F: Predictors of Financial Participation; Linear Probability Regressions for Formal

Loans Annex G: Predictors of Financial Participation; Linear Probability Regressions for Informal

Loans Annex H: Predictors of Financial Participation; Linear Probability Regressions for All Types of

Insurance Annex I: Predictors of Financial Participation; Linear Probability Regressions for Financial

Literacy Annex J: Summary of Regulatory Framework for Banks, with Attention to Issues of Access Annex K: Summary of Regulatory Framework for Cooperatives, with Attention to Issues of

Access Annex L: Summary of Regulatory Framework for Finance Companies, with Attention to Issues

of Access

Conference Edition

169

Annex ___ Access vs. use of financial services - Some Definitions, Distinctions and Concepts

It is useful to clarify ‘access to financial services’, which is a much broader concept than use of financial services, or use of specific financial services like credit. For many observers, ‘access to financial services’ is synonymous with ‘access to credit’. This report takes a much broader view of the issue, seeing credit as only one of several instruments of financial services that are important to the poor, for instance: their deposits; guarantees of their deposits; and transfers of remittances. This is illustrated in Figure 1, showing the different products of banks (including MFIs) and non-banks. Of banks’ products, poor households and SMEs make good use of the services indicated in green (especially savings accounts and remittances); they make moderate to good use of those services indicated in light blue; and they make limited use of those services highlighted in yellow. Pensions, highlighted in red were not covered by this report. It should be noted that as regards credit services, only the fully collateralized lending of pawnshops is heavily used by the poor and MSMEs.

Figure 57: Access to Financial Services Versus Access to Credit

(by financial service and service provider)

Savings Accounts (time & savings deposits)

Banks: Transactions (remittances, transfers, debit & credit cards)

Credits (roughly half or less to SMEs) Pawnshops (all are small-scale credits) Finance companies (limited amounts for SMEs, e.g. leasing) Non-banks: Insurance (limited amounts for lower-income customers) Pensions (not covered by this report) Along similar lines, many households may be genuinely excluded from access because they choose not to use formal financial services. It is therefore important to distinguish between voluntary and involuntary exclusion from use of formal financial services, as illustrated in Figure 2. In particular, voluntary exclusion should not be counted as lack of access, and research indicates that the numbers are large.124 By way of illustration, some wealthy households may choose not to borrow because they are self-sufficient in cash flow, or because they are financially very conservative. Similarly, deeply religious households may choose not to conduct business with mainstream commercial financial institutions. It is important to note that while these households may choose to 124 See, for example, Johnston and Morduch (2008). This work indicates that half of creditworthy, poor households simply do not want credit.

remain exdemand fbetween a Figure 58

Aservices fobase, possimportancwith the land InterNhigher-enconsidera

Popu

xcluded, volufor financial access and us

: Difference

Another impofor existing clsibly via newce of the formatter, essentiNet banking

nd financial seable promise

ulation

Financially

Source: Claes

Financially

untary exclusservices is

se is understo

Between Ac

ortant distinctlients; and ‘ad

w channels fomer, especialially because are example

ervice providto extend ser

Using Fiservi

Not Using Serv

Excluded

ssens (2005)

Included

sion is definefairly sensitiood, the repo

ccess to and U

tion regardindding-on newr delivery of lly to traditioit is an instru

es of add-on ders. Mobilervices to new

inancial ices

Financial vices

ed at the prevaive to their port focuses o

Usage of Serv

ng access conw clients’, byf services. Thonal financialument for poservices that banking is a

w, lower-end

Voluntari

InvoluExc

ailing price of sprice. There

on the use of

vices

ncerns the dify reaching mohis report recl intermediarioverty allevia

are essentialan example o

customers.

ly Excluded

untarily cluded

Coservices. Thifore, althougfinancial ser

fference betwore deeply incognizes the ies, but is mo

ation. Creditlly only relev

of a new deliv

price orconstrain

will be

No neeAwar

Due to hicr

Dudiscrim

Due to othor client

onference Eds report findgh the distinrvices.

ween: ‘add-onnto the populcommercial ore concernet and debit ca

vant to existinvery channel

r income nt. Assume

rejected

ed or No reness

gh risk/bad edit

ue to mination

her product t features

dition

170

ds that nction

n’ lation

ed ard ng, with

Conference Edition

171

Annex A

Survey methodology

Survey Unit of Observation The Indonesian Access to Finance Survey (A2F) is a hybrid household and individual level survey that blends individual level socio-economic characteristics with household level financial statistics. Whereas the survey recorded socio-economic data (such as, education level and employment status) for all household members above the age of 15, questions about finances were recorded for each household as a whole. The interviewee in the household was the person with financial decision-making power, referred to the ‘head of household’; 59% of the time, this person was also the figurative head of the household.125 Although individual level data are arguably more reliable and permit analysis of within-household heterogeneity, several factors led to the choice of the above survey design. First, individual level surveys are time-consuming and costly. By focusing the A2F survey only on the head of household, the survey was expanded substantially in scope and scale. Second, the overarching theme of the survey is access to finance, which is generally a household-level decision. Recent research in household survey design seems to justify a head of household survey. Specifically, Cull and Scott (2009) investigate whether asking financial access questions only to the head of household leads to systematic inaccuracies as compared to obtaining this information individually from all household members.126 The concern is that not all individuals’ use of financial services is known to the head of the household. Their preliminary results show that for basic financial access questions, there were no significant differences between answers obtained from the sum of individuals within the household and answers obtained only from the household head, although the randomly selected household member did provide less complete data. Sampling Methodology of the Indonesian Survey The sample selection methodology was designed to ensure national representation. Multi-stage random sampling was employed with population-weighted selection occurring first at the province level, next at the sub-province (Kabupaten/Kota) level, and finally at the village (Desa/Kelurahan) level. Overall, 4 provinces on Java and 6 provinces off Java were selected, namely Banten, West Java, Central Java, East Java (all provinces on Java); and Aceh, Jambi, West Kalimantan, North Sulawesi, Maluku, and Nusa Tenggara Barat (off Java provinces). On Java, 16 population-weighted sub-provinces (excluding Jakarta and Yogyakarta) were chosen at random. Within each of these 16 sub-provinces, 4 population-weighted villages were chosen at random,

125 The ‘figurative head’ is usually the eldest person in the family. Typically, Indonesian households are joint-family setups where parents live with their children and the children’s respective families. Under such cases, the figurative head would be the father, whereas the eldest son may be the financial head of the household who is responsible for all financial decision making. 126 In order to understand and quantify potential errors and biases, the authors conducted a methodological experiment in Ghana in 2007 where they randomly implemented individual and household head only data collection methods along with a third method using a randomly selected household member.

Conference Edition

172

resulting in a total sample of 64 villages. The selection was stratified by urban/rural, with the final sample containing 34 urban villages and 30 rural villages. Off Java, the country was divided into 6 sub-regions, and one (population-weighted) province was chosen at random from each sub-region. Next, within each of these provinces, 2 population-weighted sub-provinces were chosen at random. Finally, within each of these 12 sub-provinces, 4 population-weighted villages were chosen at random, resulting in a sample of 48 villages. The urban/rural stratification resulted in a final sample of 18 urban villages and 30 rural villages, owing to the overwhelming rural nature of off-Java villages. Within each of the 112 villages in the survey, 30 households were randomly selected in each village for survey interviews, resulting in our final sample of 3,360 survey household respondents (1,920 on Java and 1,440 off Java). The interviews were conducted between July and December 2007. Sampling Weights The results from the Nationwide Access to Finance survey can be extrapolated to represent the whole population in Indonesia by weighting the result for each household with its probability of being selected out of the whole household population. The sample weights are calculated separately for Java and Non-Java households since sample selection is stratified this way. The probability of a household being selected is calculated as a function of the probability of the province being selected, the probability of the desa (village) being selected, the probability of the RW (sub-village) being selected, and the probability of the household being selected within the RW. Let p1 denote the probability of the province being selected out of all the provinces in Java and Non-Java, respectively. Let p2 denote the probability of the desa being chosen out of all desa within the province. Let p3 denote the probability of the RW being chosen out of all the RWs within the desa. Finally, let p4 denote the probability of the household being chosen out of all the households in the RW. The sample weight for each household is then denoted by the expression:

1 / (p1*p2*p3*p4)

or, the inverse probability of selection. The interpretation of the sampling weight is that it characterizes the number of households in the entire population that are represented by this particular sample household. Comparison of A2F with PODES and SUSENAS This section presents some basic comparisons of survey characteristics between the A2F survey and two existing sources of financial data in Indonesia, namely the PODES and SUSENAS, supply-side census and demand-side surveys, respectively, both conducted by the National Statistics Bureau (BPS).

Conference Edition

173

Village Potential Statistics (PODES) is a nation-wide census of all villages in Indonesia and contains information on various socio-economic conditions at the village level.127 SUSENAS is a nationally representative socio-economic survey conducted annually by the BPS. Comparisons here provide evidence that the Access to Finance Survey is nationally representative, within a reasonable statistical range, Comparison of A2F with PODES 2005 The latest available PODES census was conducted in April 2005, whereas the A2F Survey was conducted two years later, in 2007. Also noteworthy is that PODES is a census and therefore covers all villages in Indonesia, whereas A2F is a sample with 30 households per village presenting our village level aggregate data. Some representative comparisons are provided in Figures 57 and 58.

127 The PODES is conducted three times every ten years. The main topic of each of these three corresponds to the periodic censuses (population, year ended with 0; agriculture, year ended with 3; economic, year ended with 6).

Conference Edition

174

``

Figure 59: Samples in PODES 2005 vs. Access to Finance 2007, at Village Level Urban Rural # of Poor Households

Do You Have a Fixed Phone Line? Do You Have Cell Phone Connection?

Is There any Commercial Bank? Is There any BPR?

17.6

46.4 42.9

82.4

53.6 57.1

0

20

40

60

80

100

National PODES

PODES ATF

Rural

Urban

32.8 27.1 33.1

67.2 72.9 66.9

0

20

40

60

80

100

National PODES

PODES ATF

Non‐poor

Poor

34.7

60.769.6

65.3

39.330.4

0

20

40

60

80

100

National PODES PODES ATF

No

Yes70.6

86.694.6

29.413.4

5.4

0

20

40

60

80

100

National PODES PODES ATF

No

Yes

6.616.1

25.9

93.483.9

74.1

0

20

40

60

80

100

National PODES PODES ATF

No

Yes

5.713.4 14.3

94.386.6 85.7

0

20

40

60

80

100

National PODES PODES ATF

No

Yes

Conference Edition

175

More detailed statistical analysis was conducted for basic village level characteristics for the 112 villages in A2F (Table 27). The results find no significant differences between A2F and PODES in terms of: urban/rural split; population figures; availability of telephones; the existence of branches of People’s Credit Banks (BPRs); and Village Credit Unions (KUDs). There were some minor, but statistically significant, differences in: the proportion of poor households (27% in PODES vs. 33% in A2F); presence of pawn shop (4% in PODES vs. 9% in A2F); presence of commercial banks (16% in PODES vs. 26% in A2F); and presence of other credit institutions, such as cooperatives (78% in PODES vs. 93% in A2F).128 Matching observations on deciles rather than point estimates, there is very little statistically significant difference between village-level characteristics reported in PODES and those in A2F. Overall, the sample selection in A2F matches fairly well with census results from PODES.

128 It should be noted that the A2F survey yields higher numbers for each of these important categories.

Figure 59: Samples in PODES 2005 vs. Access to Finance 2007, at Village Level (Continued)

Is There any Pawn Shop? Is There any Other Credit Institution?

Is There any KUD?

1.3 3.6 8.9

98.7 96.4 91.1

0

20

40

60

80

100

National PODES PODES ATF

No

Yes

55.1

77.792.9

44.9

22.37.1

0

20

40

60

80

100

National PODES PODES ATF

No

Yes

11.221.4

13.4

88.878.6

86.6

0

20

40

60

80

100

National PODES PODES ATF

No

Yes

Conference Edition

176

Comparison of A2F with SUSENAS 2007 SUSENAS is also a nationally representative 2007 survey, but with a much larger sample size of 227,202 households, and comparisons are possible between socio-economic characteristics between SUSENAS and A2F at the household level. When comparing households in SUSENAS with A2F (Table 28), there were significant—albeit small—differences in gender of household head, age composition, education levels, monthly non-food consumption, and usage of formal credit. These differences are not likely to indicate any systematic sample selection bias. Rather, they probably stem from A2F’s randomly chosen respondents being picked, by coincidence, from the districts that have a slightly higher probability of

Table 26: Statistical Comparison of Village-Level Variables in the A2F Survey and PODES

Variable Survey Obs. Mean Std. Dev. Sig. Diff.

Urban Village? PODES 112 0.46 0.50 A2F 112 0.43 0.50 Diff 0.04

Population Size PODES 112 5,824 4,854 A2F 112 6,049 5,200 Diff (225)

Number of Poor Households PODES 112 390.80 452.63 A2F 108 542.43 468.06 Diff -151.63 *

Village has fixed phone line? PODES 112 0.61 0.49 A2F 112 0.70 0.46 Diff -0.09

Village has cell phone coverage? PODES 112 0.87 0.34 A2F 112 0.95 0.23 Diff -0.08 **

Village has commercial bank? PODES 112 0.16 0.37 A2F 112 0.26 0.44 Diff -0.10 *

Village has BPR? PODES 112 0.13 0.34 A2F 112 0.14 0.35 Diff -0.01

Village has pawn shop? PODES 112 0.04 0.19 A2F 112 0.09 0.29 Diff -0.05 *

Village has another kind of credit institution?

PODES 112 0.78 0.42 A2F 112 0.93 0.26 Diff -0.15 ***

Village has a KUD? PODES 112 0.21 0.41 A2F 112 0.13 0.34 Diff 0.08

Note: * = significant at 10%; ** = significant at 5%; ** * = significant at 1%.

Conference Edition

177

having male and younger household heads, lower high school graduates, higher non-food consumption, and lower usage of formal credit.129

Table 27: Statistical Comparison of A2F vs. Non-A2F Villages Using SUSENAS Data

Dependent Variable: A2F District Coeff. t-stat

Female Head of HH -0.011 -1.81 * Age -0.019 -1.86 * Primary School -0.012 -1.32 High School Up -0.014 -1.67 * House has Roof -0.002 -0.1 House has Wall 0.003 0.33 Existence of Migrant Worker 0.018 0.59

Salary 0.002 0.44 Consumption: Food 0.000 0.02 Consumption: Non-food 0.034 2.51 **

Credit: from Formal Sector -0.019 -1.96 ** Credit: from Informal Sector -0.020 -1.29 Province Fixed Effects YES Observations 76,513 R-Squared 0.122

Note: * = significant at 10%; ** = significant at 5%. Some Characteristics of A2F Sample Respondents As mentioned, the sample for the A2F Survey was selected in a manner to ensure national representation. Survey questions were posed to the household member ultimately responsible for financial decisions. Looking at respondents’ broad characteristics, for 59% of the population, the respondent was the household head; they were overwhelmingly Muslim (93%), married (83%) and half are female (Figure 59).

129 This stands in contrast to the result noted in the previous footnote.

Conference Edition

178

Figure 60: A2F Sample Characteristics

Continuing with the broad picture, respondents indicate that the national population is 42% urban, with substantial variation on and off Java (Figure 60). On Java, almost 45% of the population is urban whereas, off Java only 22% is urban (Figure 60).

Figure 61: Urban/Rural Split

77.7%

22.3%

55.1%

44.9%

Off Java Respondents On Java Respondents

Rural Rural Urban

Urban

49.5%50.5%

Female Male

Gender of Respondent

40.5%

59.5%

No Yes

Respondent is HH Head

Conference Edition

179

By regional distribution, most respondents are from Java; 35% of respondents from East Java, 33% from West Java, and 17% from Central Java. Off Java provinces account for the remaining 15% (Figure 61).

Figure 62: Respondents, by Province

By age, 90% of the sample respondents are above 25 years old, with the average respondent around 40 years of age, both on and off Java (Figure 62). Splitting the sample by gender, we find that the median age for female respondents is much younger than male respondents, though both distributions look quite similar (also Figure 62).

Figure 63: Sample Characteristics, Age by Gender

05

10

20 40 60 80 100 20 40 60 80 100

Female Male

Perc

enta

ge o

f Sam

ple

Age of Respondent

Age Profile by Gender of Respondent

.0.7 1.8

3.1

2.8

1.8

35.0

17.3

33.3

1.6

2.8

N-MALUKU N-SULWS

W-KALIMAN

W-NUSATENG BANTEN

E-JAVA

C-JAVA W-JAVA

JAMBI

ACEH

Respondents by Provinces (%)

Conference Edition

180

Most respondents have attended at least primary school (89%), and only 10% have education beyond high school (Figure 63). These figures are very similar for on and off Java residents.

Figure 64: Respondents, by Employment & Education

A large proportion of respondents are employed (73%) and nearly half of them (47%) are self-employed. Around 49% are wage earners, and 36% work in agriculture and 22% in trade. As would be expected, there is substantial variation between and Java and off Java residents; whereas wage income dominates farm income on Java (51% vs. 33%), these figures are reversed off Java (34% vs. 56%). The large urban-rural divide is the likely explanation for more farm based activities off Java.

Figure 65: Respondent Wage Composition

32,1

6,1

0,823,30,75,0

27,6

3,7 0,7

Respondent Wage Composition

Daily

Weekly

Every 2 weeks

Monthly

Quarterly

One off payment only

19.6%

30.0%

17.1%

23.1%

10.2%

No formal education Primary school/equivalentSecondary school/equivalent Senior High school/equivalentUniversity

Respondents by Education Level

8.9%

14.3%

45.6%

1.8%

25.6

3.8%

Government employee Private employeeSelf employed Employer Freelance/Casual labor HH worker not being paid

Respondents by Employment

Conference Edition

181

At the national level, the wage composition of respondents presents a complicated pattern (Figure 64). Only 23% receive a regular monthly wage; 32% receive wages on a daily basis; another 8% are paid regularly, but at different frequencies; and 28% have irregular compensation schedules. Some 39% of households own and operate a household-level enterprise, whereas only 10% have a migrant worker abroad, which matches the percentage reported in PODES.130 Home ownership is quite high in the sample--around 3/4s of the population owns their residence, both on and off Java. Most houses are permanent buildings made from brick-walls (75%), tile-roofs (86%), and tile-floors (54%), with the average number of rooms per home is 5. Brick houses are much more common on Java (79%) as compared to off Java (59%). Of these homes, 99% on Java have electricity and 87% have telephone service. By comparison, 87% of homes off Java have electricity and only 43% have telephone service. Access to tap water is low both on and off Java with an average of only 23% of homes having access to tap water as their main source of drinking water. Yearly per capita consumption expenditure equals $1,068, and is expectedly higher on Java than off Java. This average national consumption expenditure figure is roughly half of Indonesia’s GDP per capita ($1,918).

130 There are regions in the Indonesia where large proportions of the local population have migrant workers abroad, and the sample was not stratified according to migrant worker status. This could explain why we find such a low proportion of the national household population as having migrant workers abroad. The World Bank Indonesia office recently conducted another household survey focusing on access to finance for migrant workers and their families in East Java, Nusa Tengga Barat and Nusa Tengga Timur in late 2008. This survey will likely provide a better picture of access to finance issues for migrant worker families. The data from this survey are currently being collected and analyzed.

Conference Edition

182

Annex B: Predictors of Financial Participation Linear Probability Regressions for Savings of Any Kind

Basic Province Fixed Village Fixed Results Effects131 Effects132 Urban household 0.032 0.032 --

(0.023) (0.023) -- Household size -0.010 -0.001 0.002

(0.006) (0.006) (0.006) Respondent is head of household -0.094 -0.070 -0.066

(0.030)*** (0.030)** (0.028)** Male 0.019 0.008 0.009

(0.029) (0.028) (0.027) Age 0.003 0.001 -0.002

(0.004) (0.003) (0.004) Age squared 0.000 0.000 0.000

(0.000) (0.000) (0.000) Married 0.010 0.006 0.004

(0.028) (0.028) (0.027) Attended school 0.045 0.073 0.106

(0.040) (0.039)* (0.039)*** Muslim -0.072 -0.037 -0.018

(0.027)** (0.028) (0.032) Log of consumption expenditure 0.100 0.103 0.091

(0.009)*** (0.009)*** (0.010)*** Employed 0.057 0.022 0.026

(0.026)** (0.026) (0.026) Own house -0.011 -0.015 -0.001

(0.024) (0.023) (0.025) Number of rooms 0.013 0.012 0.016

(0.005)** (0.006)** (0.006)*** Have electricity 0.142 0.107 0.086

(0.046)*** (0.045)** (0.053) Have telephone 0.145 0.109 0.167

(0.029)*** (0.033)*** (0.035)*** Have tap water 0.064 0.057 0.057

(0.021)*** (0.021)*** (0.026)** Own enterprise 0.055 0.061 0.068

(0.020)** (0.020)*** (0.020)*** Household has migrant worker abroad 0.103 0.125 0.080

(0.033)*** (0.034)*** (0.037)** Financial literacy score 0.155 0.114 0.087

(0.038)*** (0.038)*** (0.038)** Cognitive/Math skills score 0.224 0.196 0.188

(0.056)*** (0.056)*** (0.055)** Risk averse -0.026 -0.015 -0.007

(0.020) (0.020) (0.019) Time consistent preferences 0.013 0.012 0.008

(0.029) (0.028) (0.027) Constant -1.480 -1.438 -1.253

131 Exploits within-province variability and controls for heterogeneity across provinces. 132 Restricts analysis to within-village variation and controls for variation across villages.

Conference Edition

183

(0.141)*** (0.147)*** (0.156)*** Observation 3360 3360 3360 R-squared 0.267 0.297 0.372 Standard errors in parentheses: * significant at 10%; ** significant at 5%; ***significant at 1*

Annex C: Predictors of Financial Participation Linear Probability Regressions for Formal Savings

Basic Province Fixed Village Fixed Results Effects Effects Urban household 0.070 0.075 --

(0.024)*** (0.025)*** -- Household size -0.018 -0.013 -0.005

(0.006)*** (0.006)** (0.006) Respondent is head of household -0.025 -0.012 -0.011

(0.033) (0.032) (0.032) Male 0.007 -0.001 0.012

(0.031) (0.031) (0.031) Age 0.009 0.008 0.003

(0.004)** (0.004)** (0.004) Age squared 0.000 0.000 0.000

(0.000)** (0.000)** (0.000) Married -0.011 -0.010 -0.014

(0.029) (0.030) (0.030) Attended school 0.055 0.068 0.084

(0.036) (0.035)* (0.036)** Muslim -0.035 -0.020 0.038

(0.041) (0.050) (0.057) Log of consumption expenditure 0.131 0.134 0.115

(0.009)*** (0.010)*** (0.010)*** Employed -0.014 -0.030 -0.011

(0.028) (0.029) (0.028) Own house 0.003 0.001 0.035

(0.026) (0.026) (0.027) Number of rooms 0.022 0.022 0.025

(0.006)*** (0.006)*** (0.006)*** Have electricity -0.028 -0.031 -0.034

(0.044) (0.042) (0.050) Have telephone 0.059 0.040 0.167

(0.029)** (0.033) (0.038)*** Have tap water 0.094 0.082 0.063

(0.026)*** (0.027)*** (0.033)* Own enterprise 0.064 0.068 0.074

(0.023)*** (0.023)*** (0.022)*** Household has migrant worker abroad 0.217 0.215 0.207

(0.035)*** (0.036)*** (0.039)*** Financial literacy score 0.143 0.123 0.072

(0.042)*** (0.042)*** (0.040)* Cognitive/Math skills score 0.231 0.219 0.190

(0.054)*** (0.055)*** (0.055)*** Risk averse 0.025 0.034 0.031

(0.021) (0.021) (0.021) Time consistent preferences 0.022 0.022 0.019

Conference Edition

184

(0.031) (0.031) (0.030) Constant -2.093 -2.100 -1.869

(0.148)*** (0.160)*** (0.174)*** Observation 3360 3360 3360 R-squared 0.297 0.309 0.381 Standard errors in parentheses: * significant at 10%; ** significant at 5%; ***significant at 1*

Annex D: Predictors of Financial Participation Linear Probability Regressions for Informal Savings

Basic Province Fixed Village Fixed Results Effects Effects Urban household 0.026 0.017 --

(0.034) (0.033) -- Household size -0.002 0.007 0.007

(0.008) (0.008) (0.008) Respondent is head of household -0.167 -0.123 -0.123

(0.044)*** (0.044)*** (0.043)*** Male 0.016 0.000 -0.009

(0.043) (0.044) (0.043) Age -0.003 -0.005 -0.007

(0.005) (0.005) (0.005) Age squared 0.000 0.000 0.000

(0.000) (0.000) (0.000)* Married 0.013 0.013 0.006

(0.039) (0.037) (0.037) Attended school -0.012 0.021 0.052

(0.044) (0.044) (0.045) Muslim -0.134 -0.079 -0.048

(0.043)*** (0.053) (0.093) Log of consumption expenditure 0.063 0.063 0.058

(0.015)*** (0.015)*** (0.015)*** Employed 0.129 0.077 0.072

(0.037)*** (0.038)** (0.036)** Own house -0.036 -0.061 -0.043

(0.037) (0.036)* (0.037) Number of rooms 0.008 0.007 0.013

(0.010) (0.010) (0.010) Have electricity 0.143 0.088 0.071

(0.037)*** (0.036)** (0.039)* Have telephone 0.184 0.128 0.053

(0.030)*** (0.037)*** (0.052) Have tap water 0.050 0.046 0.067

(0.045) (0.046) (0.066) Own enterprise 0.036 0.040 0.064

(0.032) (0.031) (0.031)** Household has migrant worker abroad -0.087 -0.060 -0.116

(0.051)* (0.047) (0.048)** Financial literacy score 0.165 0.126 0.122

(0.057)*** (0.056)** (0.054)** Cognitive/Math skills score 0.098 0.080 0.097

(0.064) (0.064) (0.065) Risk averse -0.063 -0.060 -0.045

Conference Edition

185

(0.030)** (0.029)** (0.028) Time consistent preferences 0.008 0.019 0.002

(0.042) (0.041) (0.039) Constant -0.822 -0.732 -0.598

(0.216)*** (0.222)*** (0.230)** Observation 1847 1847 1847 R-squared 0.149 0.211 0.336 Standard errors in parentheses: * significant at 10%; ** significant at 5%; ***significant at 1*

Annex E: Predictors of Financial Participation Linear Probability Regressions for Loans From Any Source

Basic Province Fixed Village Fixed Results Effects Effects Urban household -0.013 -0.035 --

(0.027) (0.028) -- Household size 0.018 0.021 0.019

(0.007)*** (0.007)*** (0.007)*** Respondent is head of household 0.006 0.000 -0.004

(0.039) (0.039) (0.039) Male -0.034 -0.027 -0.029

(0.038) (0.038) (0.038) Age 0.002 0.003 0.004

(0.005) (0.005) (0.005) Age squared 0.000 0.000 0.000

(0.000) (0.000) (0.000) Married 0.088 0.074 0.060

(0.037)** (0.038)* (0.038) Attended school 0.079 0.090 0.077

(0.044)* (0.044)** (0.046)* Muslim 0.148 0.095 0.073

(0.047)*** (0.059) (0.073) Log of consumption expenditure 0.045 0.053 0.056

(0.011)*** (0.012)*** (0.012)*** Employed 0.076 0.080 0.072

(0.033)** (0.033)** (0.033)** Own house -0.047 -0.054 -0.063

(0.030) (0.030)* (0.033)* Number of rooms -0.003 -0.005 -0.001

(0.007) (0.007) (0.008) Have electricity 0.099 0.044 0.010

(0.053)* (0.053) (0.065) Have telephone 0.047 0.023 0.062

(0.030) (0.035) (0.041) Have tap water -0.042 -0.023 -0.043

(0.031) (0.032) (0.042) Own enterprise 0.011 0.005 0.014

(0.025) (0.025) (0.026) Household has migrant worker abroad 0.028 0.022 -0.004

(0.039) (0.040) (0.043) Financial literacy score 0.076 0.060 0.071

(0.049) (0.049) (0.049) Cognitive/Math skills score 0.088 0.071 0.075

Conference Edition

186

(0.063) (0.064) (0.066) Risk averse 0.003 0.000 0.002

(0.025) (0.025) (0.025) Time consistent preferences 0.001 0.005 -0.004

(0.034) (0.034) (0.035) Constant -0.669 -0.639 -0.691

(0.175)*** (0.187)*** (0.206)*** Observation 3360 3360 3360 R-squared 0.085 0.098 0.147 Standard errors in parentheses: * significant at 10%; ** significant at 5%; ***significant at 1*

Annex F: Predictors of Financial Participation Linear Probability Regressions for Formal Loans

Basic Province Fixed Village Fixed Results Effects Effects Urban household -0.001 -0.004 --

(0.023) (0.023) -- Household size 0.004 0.004 0.010

(0.006) (0.006) (0.006)* Respondent is head of household -0.040 -0.043 -0.041

(0.026) (0.026) (0.027) Male 0.030 0.033 0.033

(0.027) (0.027) (0.027) Age 0.006 0.006 0.004

(0.003)* (0.003)* (0.003) Age squared 0.000 0.000 0.000

(0.000) (0.000) (0.000) Married 0.002 -0.002 -0.024

(0.028) (0.028) (0.028) Attended school 0.011 0.010 0.011

(0.025) (0.026) (0.026) Muslim 0.030 0.040 0.036

(0.038) (0.050) (0.059) Log of consumption expenditure 0.057 0.060 0.050

(0.009)*** (0.010)*** (0.010)*** Employed -0.019 -0.015 0.004

(0.028) (0.028) (0.028) Own house 0.041 0.041 0.049

(0.024)* (0.024)* (0.026)* Number of rooms 0.012 0.011 0.017

(0.006)* (0.006)* (0.006)*** Have electricity -0.001 -0.005 -0.031

(0.022) (0.018) (0.023) Have telephone 0.041 0.031 0.096

(0.021)** (0.026) (0.035)*** Have tap water 0.033 0.039 0.074

(0.029) (0.030) (0.041)* Own enterprise 0.020 0.020 0.043

(0.023) (0.023) (0.023)* Household has migrant worker abroad 0.001 -0.004 0.009

(0.032) (0.034) (0.036) Financial literacy score 0.065 0.063 0.041

Conference Edition

187

(0.037)* (0.037)* (0.037) Cognitive/Math skills score 0.086 0.081 0.126

(0.041)** (0.042)* (0.044)*** Risk averse 0.031 0.029 0.024

(0.021) (0.021) (0.020) Time consistent preferences 0.009 0.009 0.003

(0.030) (0.031) (0.029) Constant -1.111 -1.136 -1.057

(0.146)*** (0.158)*** (0.164)*** Observation 3360 3360 3360 R-squared 0.103 0.106 0.219 Standard errors in parentheses: * significant at 10%; ** significant at 5%; ***significant at 1*

Annex G: Predictors of Financial Participation Linear Probability Regressions for Semi-formal Loans

Basic Province Fixed Village Fixed Results Effects Effects Urban household 0.020 0.022 --

(0.018) (0.019) -- Household size 0.006 0.007 0.008

(0.004) (0.005) (0.005)* Respondent is head of household 0.021 0.025 0.027

(0.022) (0.022) (0.024) Male -0.026 -0.029 -0.027

(0.022) (0.022) (0.023) Age 0.005 0.005 0.005

(0.002)** (0.002)** (0.003)* Age squared 0.000 0.000 0.000

(0.000)** (0.000)** (0.000)** Married 0.027 0.023 0.025

(0.0176) (0.018) (0.018) Attended school 0.016 0.025 0.019

(0.017) (0.018) (0.020) Muslim 0.060 0.092 0.117

(0.022)*** (0.029)*** (0.048)** Log of consumption expenditure 0.012 0.016 0.020

(0.007) (0.008)** (0.008)** Employed 0.024 0.014 0.013

(0.020) (0.020) (0.022) Own house -0.036 -0.039 -0.030

(0.022) (0.022)* (0.022) Number of rooms 0.005 0.005 0.002

(0.006) (0.006) (0.006) Have electricity -0.043 -0.038 -0.019

(0.025)* (0.023) (0.026) Have telephone 0.030 0.023 0.011

(0.015)** (0.018) (0.028) Have tap water 0.089 0.087 0.038

(0.026)*** (0.027)*** (0.040) Own enterprise -0.006 -0.003 -0.002

(0.019) (0.019) (0.019) Household has migrant worker abroad -0.011 -0.012 -0.012

Conference Edition

188

(0.021) (0.021) (0.024) Financial literacy score 0.031 0.021 0.015

(0.032) (0.033) (0.034) Cognitive/Math skills score 0.076 0.067 0.056

(0.030)** (0.030)** (0.033)* Risk averse -0.016 -0.010 -0.013

(0.017) (0.017) (0.017) Time consistent preferences -0.037 -0.036 -0.044

(0.027) (0.027) (0.027) Constant -0.370 -0.431 -0.472

(0.114)*** (0.120)*** (0.129)*** Observation 2870 2870 2870 R-squared 0.068 0.077 0.126 Standard errors in parentheses: * significant at 10%; ** significant at 5%; ***significant at 1*

Annex H: Predictors of Financial Participation Linear Probability Regressions for Informal Loans

Informal loans Basic Province Fixed Village Fixed Results Effects Effects Urban household -0.029 -0.059 --

(0.033) (0.034)* -- Household size 0.018 0.022 0.013

(0.008)** (0.008)*** (0.008) Respondent is head of household 0.019 0.011 -0.001

(0.041) (0.041) (0.041) Male -0.052 -0.043 -0.046

(0.040) (0.040) (0.039) Age -0.005 -0.005 -0.001

(0.005) (0.005) (0.006) Age squared 0.000 0.000 0.000

(0.000) (0.000) (0.000) Married 0.096 0.082 0.083

(0.038)** (0.039)** (0.037)** Attended school 0.073 0.083 0.074

(0.044) (0.045)* (0.046) Muslim 0.126 0.027 -0.022

(0.047)*** (0.060) (0.079) Log of consumption expenditure 0.016 0.021 0.028

(0.013) (0.014) (0.014)** Employed 0.102 0.108 0.088

(0.036)*** (0.037)*** (0.035)** Own house -0.060 -0.069 -0.094

(0.035)* (0.035)* (0.035)*** Number of rooms -0.016 -0.018 -0.015

(0.008)* (0.009)** (0.009)* Have electricity 0.133 0.070 0.039

(0.055)** (0.054) (0.068) Have telephone 0.018 -0.009 0.030

(0.032) (0.038) (0.048) Have tap water -0.137 -0.113 -0.159

(0.037)*** (0.038)*** (0.047)*** Own enterprise 0.008 -0.002 -0.005

Conference Edition

189

(0.030) (0.031) (0.031) Household has migrant worker abroad 0.038 0.036 0.006

(0.044) (0.046) (0.050) Financial literacy score 0.044 0.028 0.054

(0.052) (0.052) (0.051) Cognitive/Math skills score 0.008 -0.006 -0.037

(0.065) (0.066) (0.068) Risk averse -0.004 -0.011 -0.015

(0.028) (0.029) (0.028) Time consistent preferences 0.016 0.022 0.022

(0.040) (0.041) (0.040) Constant -0.021 0.084 0.008

(0.201) (0.217) (0.237) Observation 2587 2587 2587 R-squared 0.067 0.087 0.162 Standard errors in parentheses: * significant at 10%; ** significant at 5%; ***significant at 1*

Annex I: Predictors of Financial Participation Linear Probability Regressions for All Types of Insurance

Basic Province Fixed Village Fixed Results Effects Effects Urban household 0.065 0.073 --

(0.026)** (0.027)*** -- Household size 0.014 0.016 0.020

(0.006)** (0.007)** (0.007)*** Respondent is head of household -0.122 -0.108 -0.077

(0.034)*** (0.034)*** (0.033)** Male 0.122 0.115 0.111

(0.032)*** (0.032)*** (0.032)*** Age 0.010 0.009 0.004

(0.004)** (0.004)** (0.004) Age squared 0.000 0.000 0.000

(0.000)* (0.000)* (0.000) Married -0.082 -0.068 -0.053

(0.032)** (0.032)** (0.032) Attended school 0.011 0.012 0.005

(0.040) (0.040) (0.040) Muslim 0.048 0.015 0.022

(0.041) (0.049) (0.053) Log of consumption expenditure 0.073 0.064 0.065

(0.011)*** (0.011)*** (0.012)*** Employed -0.071 -0.082 -0.070

(0.030)** (0.031)*** (0.030)** Own house -0.020 -0.019 0.020

(0.028) (0.028) (0.029) Number of rooms 0.016 0.020 0.020

(0.006)*** (0.006)*** (0.006)*** Have electricity -0.171 -0.185 -0.163

(0.050)*** (0.051)*** (0.061)*** Have telephone 0.175 0.179 0.076

(0.027)*** (0.031)*** (0.041)* Have tap water 0.081 0.058 0.056

Conference Edition

190

(0.029)*** (0.029)** (0.038) Own enterprise -0.008 -0.006 0.000

(0.024) (0.024) (0.024) Household has migrant worker abroad -0.115 -0.103 -0.121

(0.036)*** (0.037)*** (0.041)*** Financial literacy score 0.093 0.091 0.049

(0.046)** (0.045)** (0.046) Cognitive/Math skills score 0.200 0.205 0.162

(0.059)*** (0.060)*** (0.059)*** Risk averse 0.012 0.014 0.000

(0.023) (0.023) (0.023) Time consistent preferences -0.021 -0.026 -0.015

(0.034) (0.034) (0.033) Constant -1.099 -0.950 -0.767

(0.167)*** (0.178)*** (0.190)*** Observation 3360 3360 3360 R-squared 0.188 0.200 0.274 Standard errors in parentheses: * significant at 10%; ** significant at 5%; ***significant at 1*

Annex J: Predictors of Financial Participation Linear Probability Regressions for Financial Literacy

Basic Province Fixed Village Fixed Results Effects Effects Urban household 0.051 0.042 --

(0.013)*** (0.013)*** -- Household size -0.007 -0.003 -0.001

(0.003)** (0.003) (0.003) Respondent is head of household 0.000 0.006 0.010

(0.019) (0.019) (0.019) Male 0.031 0.028 0.028

(0.018)* (0.018) (0.018) Age 0.005 0.004 0.003

(0.002)** (0.002) (0.002) Age squared 0.000 0.000 0.000

(0.000)*** (0.000)** (0.000)* Married -0.022 -0.026 -0.018

(0.018) (0.018) (0.018) Attended school 0.094 0.106 0.093

(0.022)*** (0.023)*** (0.023)*** Muslim -0.039 -0.054 -0.053

(0.025) (0.031)* (0.038) Log of consumption expenditure 0.019 0.021 0.019

(0.006)*** (0.006)*** (0.006)*** Employed -0.015 -0.025 -0.016

(0.016) (0.016) (0.016) Own house -0.033 -0.033 -0.015

(0.014)** (0.014)** (0.015) Number of rooms 0.001 0.000 -0.002

(0.003) (0.004) (0.004) Have electricity 0.016 -0.015 -0.012

(0.027) (0.027) (0.032) Have telephone 0.045 0.013 0.017

Conference Edition

191

(0.015)*** (0.017) (0.020) Have tap water 0.004 0.008 -0.006

(0.015) (0.015) (0.020) Own enterprise 0.025 0.026 0.024

(0.013)** (0.013)** (0.013)* Household has migrant worker abroad 0.010 0.017 0.021

(0.018) (0.018) (0.019) Cognitive/Math skills score 0.447 0.420 0.395

(0.030)*** (0.030)*** (0.032)*** Risk averse -0.017 -0.013 -0.012

(0.013) (0.013) (0.013) Time consistent preferences 0.021 0.021 0.019

(0.017) (0.017) (0.018) Constant -0.292 -0.220 -0.138

(0.090)*** (0.097)** (0.106) Observation 3360 3360 3360 R-squared 0.282 0.298 0.332 Standard errors in parentheses: * significant at 10%; ** significant at 5%; ***significant at 1*

Conference Edition

192

Annex K: Summary of Regulatory Framework for Banks, With Attention to Issues of Access

Regulations Concerning: Authority 1/ Summary of Main Content Notes on Implications for poorer households & MSMEs

1. Entry into Industry

1.1 Minimum Capital Requirements BI Regulations Vary by type of bank; high for new banks; Tier 1 for existing banks rising thru 2010 Acts as barrier for new, small banks. Less restrictive for BPRs & Sharia

1.2 Ownership Requirements BI Regulations Very open for conventional banks. No foreign ownership for BPRs No foreign ownership (incl. NGOs) hurts BPRs, which lack human resources and financing.

1.3 Management Requirements BI Regulations Size, composition & competence of Directors, Commissioners & Supervisors (for Sharia) Education, numbers & Supervisory Board (for Sharia) excessive for small BPRs.

1.4 Fit & Proper Tests BI Regulations Review of controlling shareholders, managers & executives. Not significant.

1.5 Risk Management Certification BI Regulations For management & executives; separate requirements for InterNet banking & mutual funds. Restrictive certification looks difficult for smaller banks, but not applicable to BPRs.

1.6 Compliance Director BI Regulations A member of the Board of Directors must be appointed Compliance Director. Not significant.

2. Asymmetric Information

2.1 Know Your Customer (KYC) BI Regulations Application is part of 'management' component of bank rating. Can entail serious sanctions. Dismissal of manager is one sanction; very serious for a small bank or BPR.

2.2 Transparency of Finances BI Regulations Financials published (& on BI website) regularly; less frequently for BPRs. Doubtful effectiveness for BPRs because of low financial literacy in poor, remote areas.

2.3 Transparency in Products BI Regulations Requires transparency in bank products and use of customer data in written form. Written info ineffective for BPRs because of low literacy in poor remote areas.

2.4 Debtor Information System BI Regulations Requires reporting of substantial info on loans of all sizes. Unnecessary for very small borrowers; questionable if taxpayer # & business license needed.

2.5 Deposit Insurance Act No. 24 of 2004 In principle, places a ceiling on insured deposit rates. Unclear if the ceiling is operative.

2.6 Customer Complaint Resolution BI Regulations Written policy needed for resolving complaints. Complaints must be settled < 20 days. No evidence of negative impact.

3. Prudential Supervision

3.1 Soundness Rating System BI Regulations BPRs subject to CAMEL system; commercial banks subject to augmented CAMEL. CAMEL does not incorporate 'access' criteria. Results not published.

3.2 Capital Adequacy Ratios BI Regulations 8% CAR, including lower risk for certain guaranteed small business loans (KUK). No evidence of negative impact.

3.3 Legal Lending Limits BI Regulations Limits on individual, affiliated & unaffiliated borrowers. Broad, complicated definition of 'affiliated parties' especially onerous for BPRs.

3.4 Loan Loss Provisioning BI Regulations Provisioning %'s less demanding for BRPs than other banks. If loans not backed by collateral, provisioning requirements burdensome.

3.5 Statutory Reserves BI Regulations Complicated system including primary, secondary & additional reserves; addtion'l reservesdepend on 3rd party funds & loan-deposit ratios. Special rates for Sharia banks

Additional reserve requirements soak up excess liquidity, but reduce loanable funds.

4. Pricing of Products

4.1 On Deposits Guarantee rate set by LPS LPS sets nominal ceiling on deposit rate for eligibility for deposit guarantee. Unclear if ceiling is operational; better if set as a spread relative to some benchmark.

4.2 On Lending Various Ceilings on some subsidized credit programs (KKP & KUR, latter to compensate for higher oil prices). Rates (16%) are close to market rates, but subsidized or guaranteed.

Banks have additional requirements (e.g., business licence, taxpayer number, collateral)

Subsidized programs hamper development of a resilient, integrated market.

Conference Edition

193

5. Servicing of Products

5.1 On Deposits BI Regulations Largely definitional. Minimum balances & monthly admin fees can wipe out interest on small accounts.

5.2 On Lending Banking Act No. 10/1998 Banks encouraged to disburse funds into MSME loans by inclusion in Annual Business Plan; disclose their MSME lending plans and announcing results.

Borrowers resent requirements to have a taxpayer number & business license.

5.3 Taxation of Interest on Deposits MoF No.

51/KMK.04/2001 20% withholding tax on time & savings deposits >Rp7 1/2 million. Not significant, but should be raised to allow for inflation since 2001.

5.4 Establishing Branch Offices BI Regulations Must be in Business Plan; BI rules on commercial issues; some regional restrictions. Business plan requirement restricts response to market changes. BI should not make

commercial decisions. Regional restrictions on BPRs unnecessary

5.5 Linkages with Rural Banks Encouraged by BI Aims to increase commercial bank funding for MSMEs through BPRs. Seems to have benefited BPRs by increasing their funding.

6. Exercising Claims & Remedies

6.1 Debt Restructuring Unclear Defines conditions under which loans may & may not be restructured Moral hazard problem for smaller lenders. Disrupts micro-finance institutions (BKD & LDKP).

6.2 Debt Forgiveness Various, some pending Generally restricted to Farm Business Loans (KUT) As above.

6.3 Execution of Collateral Presidential Reg.

No.89/2006 State banks must use State Receivables Settlement Agency (PUPN). Slow process & high cost, in part because kickbacks must be paid by auctioning agency.

Private banks sell directly or by auction, avoiding courts where possible. Courts too expensive for BPRs & smaller banks.

7. Industry Structure

7.1 Single Presence Policy BI Regulations Party allowed to be controlling shareholder in only 1 bank (with significant exemptions). Unclear if applies to state banks. Distraction for smaller banks and BPRs.

7.2 Mergers, Acquisitions & Consolidation Indo Banking

Architecture (API) Instrument for reducing numbers of banks & increasing their size.

Designed to reduce number of small banks. Assumes big banks are better, against strong evidence. Effectively licensing is closed for commercial banks; newcomers must buy an exiting bank.

7.3 Supervisory Actions Unclear Determines type of surveillance ('Intensive' or 'Special') from BI, including liquidation. Nothing significant.

1/ Details generally available upon request from background study.

Conference Edition

194

Annex L: Summary of Regulatory Framework for Cooperatives, With Attention to Issues of Access

Regulations Concerning Authority 1/ Summary of Content Notes on Implications for poorer households & MSMEs

1. Entry into Industry

1.1 Minimum Capital Requirements Decree of Min Coops & SMEs Sets minimum paid-up capital for primary and secondary coops. Very low minimum (Rp15 million); no significant impediment.

1.2 Ownership Requirements Law 25/1992 Sets citizenship & min.size of primary (20 persons) and secondary coops (3 coops). No significant impediments.

1.3 Management Requirements Decrees & PPs. May be managed by Board or out-sourced to professional. No significant impediments.

1.4 Fit & Proper Tests N/A N/A N/A

1.5 Risk Management Certification N/A N/A N/A

1.6 Compliance Director N/A N/A N/A

2. Asymmetric Information

2.1 Know Your Customer (KYC) Decree Determines requirements for loan applicants; not specifically AML. No significant impediments.

2.2 Transparency of Financial Condition Decree Determines users, types & disclosure of financial info. Sets threshold for formal audit. Disclosure unlikely to be effective due to low financial literacy.

2.3 Transparency in Products Decree Requires coop to have policy manual covering specific areas. Unlikely to be effective due to low literacy rates in poor remote regions.

2.4 Debtor Information System N/A Information available in members' book. No significant impediments.

2.5 Deposit Insurance N/A N/A Potentially a concern.

2.6 Customer Complaint Resolution N/A N/A No significant impediments.

3. Prudential Supervision

3.1 Soundness Rating System PP & Decree Implicitly uses CAMEL & simple rating system No significant impediments.

3.2 Capital Adequacy Ratios N/A None, but should 'remain unchanged over time'. Prudentially weak, but no significant impediments for MSMEs.

3.3 Legal Lending Limits N/A Limits are decided by the Board. Prudentially weak but no significant impediments for MSMEs.

3.4 Asset Quality Decree Scoring system for Members' Loans/Ttl Loans; NPLs; and Bad debts/NPLs

Loans to members must be >60% of Ttl loans., but no real impediment to MSMEs.

3.5 Loan Loss Provisioning N/A Scoring system for Loan Loss Provisions/NPLs. Prudentially weak but no significant impediment.

3.6 Statutory Reserves N/A N/A N/A

3.7 Credit Policy PP Must uphold good lending policies by feasibility assessment & repayment capacity. No significant impediments.

3.8 Good Corporate Governance N/A Implicit in point 1.3 above. No significant impediments.

3.9 Qualifications of Auditors Unclear Covers qualifications & appointments of 'Soundness Evaluators'. No significant impediments.

Conference Edition

195

4. Pricing of Products

4.1 On Deposits PP & Decree Sets method of distributing 'Rewards' , total of which is decided by the Board. No significant impediments.

4.2 On Lending Decree Rates & fees decided by Board. Provides options for determining rates. No significant impediments.

5. Servicing of Products

5.1 On Deposits PP & Decree Main definitions; interest rates paid to members must be higher than others. No significant impediments.

5.2 On Lending PP & Decree Defines lending priorities; maturities; and collateral requirements. No significant impediments.

5.3 Taxation of Interest on Deposits Decree No withholding tax on interest income <Rp240,000/month No significant impediments.

5.4 Establishing Branch Offices Ministerial Decision Minimum time before opening, and minimum equity, working capital & members. Minimum time looks too restrictive.

5.5 Funding Law 25/1992 & PP Allows funding from wide range of sources. No significant impediments.

5.6 Placement PP & Decree Allows excess liquidity to be placed in wide range of instruments. No significant impediments.

6. Exercising Claims & Remedies

6.1 Debt Restructuring Decree Ways of dealing with loan defaults, including rescheduling & sale of collateral

Odd that write-offs are an option with no LLPs; no significant impediments for MSMEs

6.2 Debt Forgiveness See 6.1 See 6.1 See 6.1

6.3 Execution of Collateral See 6.1 See 6.1 See 6.1

7. Industry Structure

7.1 Single Presence Policy N/A N/A N/A

7.2 Mergers, Acquisitions & Consolidation PP Gives the Minister the right to intervene if a coop's operations become unsustainable.

Bears no relation to soundness rating, but no significant impediment to MSMEs

7.3 Supervisory Actions: Liquidation PP Minister or the Board may liquidate a coop. Rare that problem cooperatives are identified by a supervision. No impediment or MSMEs.

7.4 Protection of Cooperatives Law 25/1992 Provision for certain activities to be managed only by coops. Potentially problematic for MSMEs.

1/ Details generally available upon request from background study.

Conference Edition

196

Annex M: Summary of Regulatory Framework for Finance Companies, With Attention to Issues of Access

Regulations Concerning Authority 1/ Summary of Main Content Notes on Implications for poorer households and MSMEs

1. Entry into Industry

1.1 Minimum Capital Requirements MoF Decree In 2006, minimum capital increased drastically (by 10x or more) to force consolidation.

For coops, min. capital requirement are half regular. No entry from 1995-2000 and 2002-2006..

1.2 Ownership Requirements MoF Decree Indonesians or joint ventures (up to 85% foreign). Nothing significant.

1.3 Management Requirements MoF Decree Not on black lists; (limited) experience needed. Special process for Sharia. Nothing significant.

1.4 Fit & Proper Tests MoF Decree Required of Directors, Commissioners & Branch Managers Nothing significant.

1.5 Risk Management Certification N/A N/A N/A

1.6 Compliance Director N/A N/A Looks dangerous, but does not affect MSMEs.

2. Asymmetric Information

2.1 Know Your Customer (KYC) MoF Decree Tightened up recently. Some areas (e.g., leasing) require a taxpayer number. Requirement for tax number may hinder access for the poor.

2.2 Transparency of Finances MoF Decree Requires timely publication of condensed balance sheet and P&L statement. Doubtful effectiveness for poor because of low financial literacy.

2.3 Transparency in Products N/A N/A More the important than Financial Transparency for the poor.

2.4 Debtor Information System N/A N/A Not significant for MSMEs.

2.5 Deposit Insurance N/A N/A N/A

2.6 Customer Complaint Resolution N/A N/A N/A

3. Prudential Supervision

3.1 Soundness Rating System N/A Some limited oversight of borrowings by BI. No significant impediments.

3.2 Capital Adequacy Ratios MoF Decree Net Worth of 50% of paid-up capital. No significant impediments.

3.3 Legal Lending Limits N/A N/A; a matter of internal corporate risk policy. N/A

3.4 Loan Loss Provisioning MoF Decree N/A except for Leasing companies, which has a maximum deduction against income. No significant impediments.

3.5 Statutory Reserves N/A N/A N/A

3.6 No On-site Inspections N/A Direct oversight only when there is indication of non-compliance. Looks dangerous, but doesn't hinder MSMEs.

4. Pricing of Products MoF Decree Some limits on leasing; none otherwise. No significant impediments.

5. Servicing of Products

5.1 Financial Products Bapepam Regulation Mainly definitional, mainly relating to Sharia. No significant impediments.

5.2 Loan Products Bapepam Regulation Allows joint financing with commercial banks. No significant impediments.

Conference Edition

197

5.3 Tax Treatment Unclear Concerns distinctions between lessees and leasors, and venture capital companies. No significant impediments.

5.4 Establishing Branch Offices MoF Decree Equity needs to be >50% of paid-up capital, and been in previous annual business plan. Inclusion in annual business plan looks excessive.

5.5 Funding MoF Decree Defines scope of borrowings. Non-bank > Rp 1 billion and > 1 year maturity. Minimum loan size limits sources of funding.

Max, geering ratio of 10. High geering ratio => strong liquidity & risk management.

6. Exercising Claims & Remedies N/A N/A N/A

7. Industry Structure

7.1 Single Presence Policy N/A N/A N/A

7.2 Mergers, Acquisitions& MoF Decree Unclear if need to be reported to MoF or approved by MoF. No significant impediments.

Consolidation

7.3 Supervisory Actions MoF Decree Defines process for sanctions, through to revocation of licence and liquidation. No significant impediments.

1/ Details generally available upon request from background study.

Conference Edition

198

References Asia Pacific Economic Cooperation. 2002. “Profiles of SMEs and SME Issues in APEC, 1999-

2000”, Report of the APEC Small and Medium Enterprises Working Group, World Scientific Publishing.

Ashraft, N., D. Karlin and W Lin. 2006. “Tying Odysseus to the Mast: Evidence from a

Commitment Savings Product in the Philippines”. Quarterly Journal of Economics, 121(2): 673-697.

Asian Development Bank. 2007. “Low Income Households’ Access to Financial Services.” . 2000. “Rural Financial Markets in Asia: Policies, Paradigms and Performance.” Bank Indonesia and GTZ under Project ProFI. 2000. “Legislation, Regulation and Supervision of

Microfinance Institutions in Indonesia.” Jakarta, Indonesia. Bank Indonesia . 2004. Annual Report . 2005. “Results of Study of Micro, Small and Medium Businesses (MSMEs) Profile”. . 2007. Annual Report. . 2008. “Laporan Survei Nasional Pola Remitansi TKI Tahun 2008”. Bureau of Balance of

Paymen, Directorate of Statistics and Monetary, Jakarta, Indonesia. Beck, T., A. Demirguc-Kent and S. Martinez Peria. 2004. “Finance, Inequality and Poverty: Cross-

country Evidence.” Policy Research Working Paper 3338. World Bank, Washington, D.C. . 2005. “Reaching Out: Access to and Use of Banking Services across Countries.” Policy

Research Working Paper 3754. World Bank, Washington, D.C. . 2006. “Banking Services for Everyone? Barriers to Bank Access around the World.”

Policy Research Working Paper 4079. World Bank, Washington, D.C. Binhadi, 1995. “Financial Sector Deregulation, Banking Development and Monetary Policy: The

Indonesian Experience, 1983-1993.” Institut Bankir Indonesia, Jakarta. Cole, David C. and B. F. Slade, “Building a Modern Financial System; The Indonesian

Experience”Cambridge University Press CGAP, IFC, and GTZ. 2009. “Diagnostic Report on the Legal and Regulatory Environment for

Branchless Banking in Indonesia”. Notes on Regulation of Branchless Banking in Indonesia

Claessens, S. 2005. “Access to Financial Services: a Review of the Issues and Public Policy

Objectives”. Policy Research Working Paper Series, the World Bank, Washington, D.C.

Conference Edition

199

Conroy, John D. 2008. “Financial Inclusion, A new microfinance initiative for APEC”. The Foundation for Development Cooperation, presented at Microfinance Workshop, the APEC Business Advisory Council, Jakarta, Indonesia, 23 January 2008.

Demirguc-Kunt, A., T. Beck and P. Honohan. 2008. “Finance for All: Policies and Pitfalls in

Expanding Access”. Washington, D.C. GTZ. 2001. “ProFi Microfinance Institutions Study”, in cooperation with Bank Indonesia. . “Regulation, Supervision & Support of Non-Bank, Non-Cooperative Micro-Finance

Institutions”, prepared by ProFi in cooperation with Bank Indonesia and the Ministry of Finance.

. 2007. “Complementing Microfinance Services in Indonesia Through Providing

Microinsurance”, GTZ Promotion of Small Financial Institutions (ProFI) project, Jakarta, Indonesia.

Haggard, Stephan. 2000. “The Political Economy of the Asian Financial Crisis.” Institute for International Economics, Washington, D.C.

Hanson, J. 2009. “Indonesian Banks: Major Issues and Policies in the International Financial Tsunami”, World Bank mimeo.

Johnston, Don, Jr. and Jonathan Morduch. 2007. “Microcredit and Microsaving: Evidence from

Indonesia”, with support from USAID, the Ford Foundation and the Gates Foundatiom. Draft dated March 12, 2007.

. 2008. “The Unbanked: Evidence from Indonesia”. Oxford University Press. Ledgerwood, J. 1998. “Microfinance Handbook: An Institutional and Financial Perspective”. World

Bank, Washington D.C. Lyman, T., David Porteous, and Mark Pickens. 2008. “Regulating Transformational Branchless

Banking: Mobile Phones and Other Technology to Increase Access to Finance.” CGAP Focus Note 43. Washington, D.C.

Manning, C. 2000. “Labour Market Adjustment to Indonesia’s Economic Crisis: Context, Trends

and Implications”, Bulletin of Indonesian Economic Studies, 36(1), 2000, 103-36. and M. Cronin. 2008. “Indonesia’s Interests in Labour Services at International Trade

Negotiations”. A report funded by AusAID for the Ministry of Trade and the Ministry of Finance in Indonesia.

Organization for Economic Cooperation and Development. 2008. “Indonesia Economic

Assessment”. OECD Economic Surveys. Paris. Schwarz, Adam. 1994. “A Nation in Waiting; Indonesia in the 1990s.” Westview Press, Boulder and

San Francisco. Skully, M.T. 1994. “The Development of the Pawnshop Industry in East Asia.” The Mansholt

Graduate School of Social Studies, The Netherlands.

Conference Edition

200

World Bank. 1993. “The East Asian Miracle”. New York: Oxford University Press. . 2001. “Rethinking the East Asian Miracle”. New York: Oxford University Press. . 2004. “Brazil: Access to Financial Services”, World Bank Report No. 27773-BR. . 2005. “AML/CFT Regulation: Implications for Financial Services Providers That Serve

Low Income People”, CGAP, Focus Note No. 29 . 2006a. “East Asian Finance: The Road to Robust Markets”, Washington, D.C. . 2006b. “Unlocking Indonesia’s Domestic Financial Resources: The Role of Non-Bank

Financial Institutions”, Poverty Reduction and Economic Management, Finance and Private Sector Development, Indonesia Country Management Unit.

. 2006c. “Improving Access to Finance for India’s Rural Poor”, Washington, D.C. . 2006d. “Revitalizing the Rural Economy: An assessment of the investment climate faced

by non-farm enterprises at the District level” Consultative Draft, the World Bank Office Jakarta.

. 2006d. “Strengthening Pensions and Insurance Markets in Indonesia”. Proceedings of an

International Workshop, Jakarta March 21-22, 2006, the Ministry of Finance and the World Bank.

. 2006e. “Making the New Indonesia Work for the Poor”, The World Bank Office, Jakarta. . 2007a. “Access to Finance and Economic Growth in Egypt”, the World Bank, Washington,

D.C. . 2007b. “Access to Financial Services in Nepal”, the World Bank, Washington, D.C. . 2008a. “The Malaysia-Indonesia Remittance Corridor: Making Formal Transfers the Best

Option for Women and Undocumented Migrants”. World Bank Working Paper 149. Washington, D.C.

. 2008b. “Investing in Indonesia’s Health: Challenges and Opportunities for Future Public

Spending.” Health Public Expenditure Review. The World Bank Office Jakarta. . 2008c. “The Study on State-Owned Pawn Company: Perum Pegadaian, Their role of

improving access to finance”, PowerPoint presentation to the World Bank Country Office in Jakarta, dated October 22, 2008.

. 2009a. “Access to Finance in Nigeria: Microfinance, Branchless Banking and SME

Finance”, CGAP, World Bank, Washington D.C. . 2009b. “Bringing Finance to Pakistan’s Poor, Access to Finance for Small Enterprise and

the Underserved”, World Bank, Washington, D.C. . 2009c. “Banking the Poor, Measuring Access in 54 Economies”, World Bank, Washington,

D.C.

Conference Edition

201

Improving Access to Financial Services in Indonesia

CONFERENCE EDITION