23250 vol. 2 May 2002 , ina cr Marguerite S. Robinson The Microfinance Revolution Volume 2: Lessons from Indonesia © 2002 by the International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, NW, Washington, D.C. 20433 USA All rights reserved Manufactured in the United States of America First printing May 2002 The findings, interpretations, and conclusions expressed in this book are entirely those of the author and should not be attributed in any manner to Open Society Institute or to the World Bank, its affiliated organizations, or members of its Board of Executive Directors or the countries they represent Library of Congress Cataloging-in-Publication Data has been requested Edited, designed, and laid out by Communications Development Incorporated, Washington, D C. and San Francisco, California The Microfinance Revolution Volume 1: Sustainable Finance for the Poor Volume 2: Lessons from Indonesia Volume 3: The Emerging Industry Marguerite S. Robinson The World Bank, Washington, D.C. Open Society Institute, New York Praise for The Microfinance Revolution "This is volume one of three that Dr. Robinson is preparing for World Bank pub- lication. I especially liked three aspects of this tome: it makes a strong case for em- phasizing deposit mobilization in microfinance efforts; it does a thorough job of covering recent literature on this topic; and the discussion is enriched with anthro- pological perspectives that remind us that microfinance clients are complex people. Proponents of the new paradigm of development finance will warm to Robinson's treatment of interest rate policies, deposit mobilization, stress on sustainability, em- phasis on qualhty of services, and the profitability of providing mucrofinance if done correctly... I enjoyed reading volume 1 and look forward to chewing on volumes 2 and 3." -Dale WAdams, Professor ofAgricultural Economics, Emeritus, The Ohio State University; con- sultant in ruralfinance. From a review of volume I posted on Ohio State's online Development Fi- nance Network, 9 September 2001 "If you are looking for a book that can provide a global overview of the current mi- crofinance sector with a historic flavor from both a theoretical and practical perspec- tive, you have one place to go: Marguerite S. Robinson's new book, The Microfinance Revolution. Robinson's vast experiences in Asia and skllls as a financial anthropologist are well combined to provide a gourmet treat to a wide variety of readers including academics, nucrofinance practitioners, consultants, policy makers, and researchers. In short, it is a must read for all stakeholders in mucrofinance." -Geetha Nagarajan, Associate Editor, The MicroBanking Bulletin. From a review of vol- ume 1 in theAsian Development Bank's quarterly newsletter, Focal Point for Microfinance, September 2001, p. 8 "The book examines how the diverse demand for financial services.. .for the poor and low-mcome households and their enterprises .., can be met on a large scale and on a sus- tainable basis. [The author] does this in a stimulatmg manner by drawing from the basics of commercial microfinance. She combines the basics with her insights gained from years of field experience in microfinance and policy advice to polhcymakers in a number of developing countries.The result is a unique study.. .The book is a must for those who need to keep abreast of the changing landscape of microfinance.The readers of this book will, in my view, gain both insight and foresight." -NimalA. Fernando, Lead Rural Finance Specialist,Asian Development Bank " The Microfinance Revolution is a magnificent contribution to the theory and practice of international development. It is a much-needed wake-up call for economists who have long pooh-poohed the potential of microfinance institutions for promoting sav- ings and investment and alleviating poverty. Likewise, it will alert advocates of subsi- dized microfinance that the financial needs of the vast majority of the poor can be met by commercially based microlending." -David E. Bloom, ClarenceJames Gamble Professor of Economics and Demography, Harvard University "Dr. Robinson has written a magnificent work that provides a jolt of energy as well as wise guidance to the fledgling mucrofinance industry. This book will quickly be- come required reading for students and professionals in and around the mucrofinance industry, for donors and government agencies, and for investors.This is also the first book that, through thoughtful analysis, vivid images, and extensive research, wllE beck- on commercial bankers and the rest of the 'real world' to sit up and take interest in mucrofinance. It will thus be a potent force in fusing the small scale, donor-driven mi- crofinance of today with the formal financial systems of tomorrow-systems that will provide high-quality financial services on a permanent and ever increasing scale to mil- lions of poor households around the world." -Elizabeth Littlefield, Chief Executive Officer, Consultative Group to Assist the Poorest; Director, World Bank; andformer Managing Director,JP Morgan "Marguerite Robinson has produced a major work that will unquestionably lie at the very center of microfinance literature for many years to come. Dr. Robinson is umquely qualified, having spent many years living in tiny villages as an anthropolo- gist, seeing informal finance as it happens, and having spent many years advising top policymakers on how to design effective financial services for the poor, most notably in Indonesia with Bank Rakyat Indonesia projects. Her account of the paradigm shift in microfinance is both exhaustively researched and provocative. She has a wonderful ear for stories; her book is full of marvelous phrases, excerpts, and anecdotes from the world of poor people's finance, in addition to being a wellspring of quantitative doc- umentation for the trends about which she writes. Highly recommended!" -Robert Peck Christen, Senior Adviser, Consultative Group to Assist the Poorest;Academic Director, Microfinance Training Program, Naropa University, Boulder, Colorado "The Microfinance Revolution is an ambitious achievement that will be the definitive work on microfinance now and for some time to come. In clear, convincing, and often el- egant language, Marguerite Robinson gives us the fruits of her deep experience and painstaking research.This book provides the most complete statement existing on how mucrofinance arose, how it works, and why it matters. The Microfinance Revolution views mucrofinance from the commercial or financial systems perspective. Robinson sets mi- crofinance in its correct place as one important tool in the 'poverty alleviation tool- box: In so doing she dispels the fuzzy myths surrounding the image of microfinance as a panacea for poverty. Every microfinance professional wlll want a copy of this work as a comprehensive reference for the field. Every polhcymaker or donor will be remuss if he or she makes decisions about microfinance without first internalizing Dr. Robin- son's messages." -Elisabeth Rhyne, Senior Vice President,ACCI6N International;former Director, Office of Mi- aoenterprise Development, US.Agencyfor International Development; author, Mainstreamung Microfinance: How Lending to the Poor Began, Grew and Came of Age in Bolivia Praise for The Microfinance Revolution (continued) "The Microfinance Revolution is a tour de force remarkable both for the breadth of its vision and for the wealth of experience it captures. Dr. Robinson folds page after page of tellng information about real people and their financial behavior, and about real institutions and their achievements, into a vigorously argued-and sometimes con- troversial-synthesis. Anyone interested in financial services for poor people should read it." -Richard Rosenberg, SeniorAdviser, Consultative Group to Assist the Poorest "Marguerite Robinson's book succeeds admirably in presenting and analyzing the fun- damentals of microlending and mobilizing savings among the poor. In distilling the essence of microfinance, Dr. Robinson demonstrates with extraordinary clarity that the application of commercial principles to microfinance ensures the long-lasting ca- pacity of institutions to reach those previously excluded from financial services.This book combines the detailed, painstaking research of a noted scholar with the experi- ences of successful microfinance institutions around the globe, and provides a view of remarkable scope and exceptional weight. Dr. Robinson's work is not only an essen- tial contribution to our current understanding of microfinance, but also a key resource for laying out the future of this field." -Marna Otero, President and Chief Executive Officer, ACCI6N International "If you read Finance at the Frontier, published in 1991,you should read The Microfinance Revolution, published in 2001. If you did not read Finance at the Frontier and you seek an authoritative source about nucrofinance, you should still read The Microfinance Revolution. " -J.D. Von Pischke, President, Frontier Finance International; author, Finance at the Frontier "For more than 20 years Marguerite Robinson has been at the forefront of the 'mucro- finance revolution' she documents so lucidly and persuasively in thds book. She was deeply involved in the transformation and development of Bank Rakyat Indonesia's mi- crobanking (umt desa) system, now the largest mucrofinance mstitution in the world with more than 20 nullion cients.This book brings together the author's wealth of practice- based wisdom and draws on her experience of working with institutions all over the world. It is a valuable, important, and necessary addition to the library of anyone seri- ously interested in microfinance." -Graham A. N Wright, Programme Director, MicroSave-Africa; author, Microfinance Systems "Marguerite Robinson has written a wonderful book. Its declared aim is to make the case for large-scale commercial microfinance, a cause that Dr. Robinson champions with passion, logic, and plentiful examples from her years of experience. But in the process she sheds light on a host of important and contentious issues in microfinance, and the outcome is a work that will enormously enrich the debates it is bound to en- gender." -Stuart Rutheford, Chairman, SafeSave; author, The Poor and Their Money "Hardheaded yet warmhearted, Marguerite Robinson's compendium points a way to- ward including greater numbers in the world's wealth. Based on her sound and wide- ranging field experience, her investigations up and down official ladders, and her wide knowledge of theory in several disciplines, her work is provocative yet immensely prac- tical. Robinson's contribution shows how incentives and efficiency might finally take account of local power structures, human dignity, and reciprocity.The result is just the sort of guide that could help people who deal financially across continents, and across classes, to avoid the ideological excesses of the 20th century in the 21s". It is a remark- able bridging act." -Parker Shipton,Associate Professor ofAnthropology, Boston University; author, Bitter Money "Marguerite Robinson has spent 20 years at the cutting edge of microfinance. In this book Marguerite gives us a history lesson and a guide on how to build commercial finance that fits the needs of the world's poor majority. Policymakers, finance leaders, and anyone who wants to join this revolution in banking must read this book." -Nancy M. Barry, President, Women's World Banking "This book tells a long overdue story-that of commercial microfinance institutions. It highlights the world's most efficient rural microfinance institution, Bank Rakyat In- donesia's microbanking system. Marguerite Robinson provides extensive analysis of the remarkable traits that have made microbanking at BRI an unprecedented success. This program has achleved massive outreach to millions of low-income savers and bor- rowers. All this has been accomplished in the past decade without subsidies; in fact, it is a highly profitable operation. BRI's path-breaking achievements have often been overshadowed by other, overpublicized programs. The Microfinance Revolution is a timely publication that clearly demonstrates the tremendous potential embedded in well-designed microfinance programs." -Jacob Yaron, Senior Rural Finance Adviser, World Bank; author, Successful Rural Finan- cial Institutions "In the past five years the enormous promise of access to capital as an effective tool for the world's poor has erupted into the world's consciousness. But the facts have often come intertwined with myth and legend, until oft-repeated misinformation threat- ens today to debase the accomplishments of truth. In this fog Marguerite Robinson's book, The Microfinance Revolution, arrives as a beacon. In it she combines her exten- sive first-hand experience, gained initially in Asia and then around the world, with the intellectual rigor of the first-rate scholar she also is. The result is a rare, comprehen- sive look at mucrofinance that is long on analysis and short on sound bites. By asking the right questions and seeking the tough answers around the globe, she expands our understanding even though we in the field might from time to time squirm in our seats. In the process she has presented all of us who are seriously committed to the field-practitioners, policymakers, academics, public servants, and most of all, the poor of the world-a wonderful gift of intellect and expertise." -Michael Chu, Chair, Capital Markets; former President and Chief Executive Officer, ACCI6N International;former Chairman of the Board, BancoSol iBRD 31766 THAI ~~~~~~~~INDONESIA ' SELECTED CITIES AND TOWNS OWBO \ > L S IA X > 19 @0 PROVINCE HEADQUARTERS &D , 8 > R X _, t S J L v 2 c , NATIONAL CAPITAL Andaman j ; ex 9tc < < S PHI - - PROVINCE BOUNDARIES --INTERNATIONAL BOUNDARIES ISDI ACEB II JAWA BARAT 22 KAUMAANTANTIMUR That'land ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~2 SUMATERA UTARA 12 JAWA TENGAH 23 SULAWESI UTARA 9 2 D~~~~~~~RAU 13 0I YOGYAKEARTA 24 GORONTALO Sulu RIAUARCHIPEIAGOC 14 JAWA TIMUR 2 SULAWENA Sea ~~~~~~~~~~~5JIAMBI __ - - 1 6 BNALSA TENGGARA IIARAT 2276 SSULAWES TENGIGARA- - 6 3ENGKUW 17 NUSA TENGGARA TIMUR 28 MALUKU UTARA 7 SUMATERA SELATAN is TIMOR 29 MAIUKUJ 8L AMPUNG 19 KAUMANTAN BARAT 30 PAPUJA (IRLAN JAYA) Bando. 9BANGKA BEMNG 20 KAUWAMAN TENGAVi~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~9RAG8,RETTJG 8 ASMIBA TNGS AcehX 10 i'BANT5 | t A Y 5 BRUNEI jf j > /IAEN 21 KAUMANTAN SEIATAN Celebes I.15 8 K/;L1~r'~ - o ,>Ai ^14 Se a . I S PACIFIC OCEAN --mw A} SiB = B23 1 ~~~~~~~~~~~~~~q(r-d I Mono1 ] / 2 ~ ~ ~ APORE ~~ ~22 Mndo / Ni> T. (D AN OCEAN PrnfidA N 4 t 2 '-' - 20, , , , ,, I oRr20 'PaIe '.' s7ebIundoneg cotom opmrPlnkraayo Java Sea Ujung P27anndan,m < D s l @ -7 [ on th^s mop o'o not imply. on /be parf of 7ho WorS Bem p uBanda Ky -3 ocsopM.cc oa a rr 1 6 / FU - ESTIMOR Arazfura Sea "~Gulf of __ _ _ _ _ _ _ __ _ _ _ _ _ _ _ __ _ _ _ _ _ _ _Carpentaria 0 100 200 300 ADDOKm,, flBi -, - prdocod by,1 Il9Rp O8IQ .. Ol g ! The 1 W-rhi &kA S T A I T). b-1n6-o, colo-, dorooooDo, on8d o-y6 mi,,ifonoco 81-,0AU0,,R A L .noIh,, ,oop do -01 !.ply, 80th. pot of Th. Wold Book c,oo, my' jodg-do 00 th. IOQOII01t .O-y wYrwMy,o oo.y .odonn,ro ____________________________________________ 1~~~~~~~~~~~~~~~~~~20' FEBRUARY 2002 This book is dedicated to all those who have led the microfinance revolution around the world. I add a special, personal dedication to those in Indonesia who developed large, financially self-sufficient microfinance institutions. For the first time ir history, they made commercial microbanking available on a large scale to low-income people. Ah Wardhana Sugianto, in memoriam Kamardy Arief I Gusti Made Oka Sri Adnyani Oka About the author Marguerite S. Robinson is a social anthropologist and internationally recognized expert on microfinance. She received her B.A. and Ph.D. from Harvard University and served as professor of anthropology and dean of the College of Arts and Sciences at Brandeis University before jolmng the Harvard Institute for International Development, where she worked from 1978-2000. She has worked extensively in rural areas and among the urban poor in India, Sri Lanka, and Indonesia-where she served for many years as an adviser to the Mimistry of Finance and to Bank Rakyat Indonesia. She has also worked in other Asian countries and in Latin America and Africa, advising governrnents, banks, and donors, and is the author of many papers and books on development and microfmance. Contents Foreword Ira W Lieberman xxi Introduction Ali Wardhana xxv Preface xxxiii Acknowledgments xlv PART 3 COMMERCIAL MICROFINANCE IN INDONESIA 1 Overview 2 Chapter 8 An Introduction to Indonesia 18 The Javanese Wayang Shadow Theater 24 Indonesia's Environment, Demographics, Early History, and Struggle for Independence 25 The Sukarno Era, 1949-67 28 The Soeharto Era, 1967-98 30 The Javanese concept of power 33 Pancasila, the guiding principles of Indonesian government 33 The economy, 1966-96 34 The government, the military, and politics during the Soeharto period 40 Corruption and the growth of the conglomerates 45 Indonesia on its 50th birthday, 1995 46 The Indonesian Crisis and the Resignation of President Soeharto 47 The nature and extent of the crisis 48 xi The roots of the crisis 50 Indonesia in Transition 56 The economy and the financial system, 1998-99 57 Governance and politics, 1998-99 58 President AbdurrahmanWahid, 1999-2001 60 President Megawati Soekarnoputri 71 Democracy and the Messy State 76 Notes 78 Chapter 9 Rural Development and Rural Financial Institutions in Indonesia 82 Geographic and Demographic Diversity 86 Rural Development 88 Varieties of agriculture 88 Cultivation on Java in the early 20th century 89 The green revolution 90 Rural development during the Indonesian crisis 93 Developing Rural Financial Institutions 93 European background 95 The early Indonesian people's banks 96 Bank Perkreditan Rakyat (People's Credit Banks, or BPRs) 98 Rural Finance in the 1980s 99 Government Microfinance Initiatives in the 1990s 102 The 1992 Banking Law and BPRs 102 KUK and commercial banks 104 Levying corporations and wealthy individuals for funds for subsidized credit programs 105 "Left-behind" villages: INPRES Desa Tertinggal 105 Bank Indonesia, government ministries, and rural credit programs 106 Rural Finance in the 1990s 107 Developing the BPRs 109 Six examples of rural financial institutions and programs 115 Rural Financial Institutions in Indonesia in 2000: What Have We Learned? 136 Policies and politics 136 Financial intermediation 137 Rural financial institutions: public and private 138 Learning from experience 138 Stability 139 Notes 139 Chapter 10 Where the Microfinance Revolution Began: Bank Dagang Bali 144 Bank Dagang Bali: Formative Elements 147 Oka family history 147 The Okas as informal commercial lenders 148 Bank Pasar Umum 149 Bank Dagang Bali: Development and Performance, 1970-96 150 xii The Microfinance Revolution: Lessons from Indonesia Customer Relations, Organization, Products, and Services 155 Organization and management 156 Loan products 157 Savings products 157 Lotteries 158 Mobile savings teams 158 Public relations 159 Bank Dagang Bali's Savers 160 Low-income savers 161 Lower-middle- and middle-income savers 162 High-income savers 163 Notes 163 Chapter 11 How to Fail in Financing the Poor: Bank Rakyat Indonesia's Unit Desa System, 1970-83 166 Bank Rakyat Indonesia's History, 1895-1970 172 The Development of the Unit Desa System, 1970-83 175 BIMAS loans to rice farmers 180 Kredit Mini and Kredit Midi loans 184 Deposit mobilization in the unit desa system, 1970-83 186 Losses in the unit desa system, 1970-83 187 Why Did BIMAS Fail? 187 The program's goals were incompatible 188 BIMAS credit was tied to input packets 188 Other agencies selected BIMAS borrowers, but BRI had to collect the loans 189 BIMAS did not reach many poor rice farmers 189 Policies for loan forgiveness and rescheduling during crop failures were badly planned and encouraged corruption 190 BRI did not have the organization, human resources, or motivation to manage unit desas effectively 190 Successful agricultural credit programs require successful agriculture-but insecticides supplied in the BIMAS input packet caused severe crop destruction 191 Examples of BIMAS in Four Rice-growing Environments 194 Village G (East Java): BIMAS results in an ideal rice intensification environment 195 Village C (West Java): Most households benefit from rice intensification, but not necessarily from BIMAS 198 Village R (South Sulawesi): A village with large potential for rice intensification and low participation in BIMAS 200 Village P (South Sumatra):A village where nearly everything that could have been done wrong in BIMAS implementation was done wrong 202 Implications of BIMAS: Results in four villages 205 The End of the BIMAS Era 208 Notes 210 Contents xiii Chapter 12 Success in Microlending: The KUPEDES Credit Program, 1984-96 216 Indonesia's Enabling Conditions 219 Bank Rakyat Indonesia's New Approach to Microlending at the Unit Desas 221 Implementing KUPEDES 221 Microcredit: subsidized and commercial 222 Perception Gaps among Policymakers, Bureaucrats, andVillagers 224 Policymakers 225 Bureaucrats 226 Villagers 228 The Transition from BIMAS to KUPEDES 229 Policy issues 231 Three crucial decisions 232 Cutting back on credit subsidies 233 Starting the KUPEDES credit program 234 Basic Principles of KUPEDES 235 KUPEDES and unit desa reorganization 237 Selecting KUPEDES borrowers 238 Characteristics of the KUPEDES Loan Product 239 Loan purposes 239 Loan sizes 240 Interest rates 240 Loan terms 241 Implementation of the KUPEDES Credit Program 244 From application to collection 244 Measuring and managing delinquency 245 Outreach 246 KUPEDES Performance, 1984-96 252 Outreach and repayment history 254 Geographic coverage 255 Unit desa profitability 256 Notes 258 Chapter 13 Mobilizing Massive Savings: Bank Rakyat Indonesia's Unit Desas, 1984-96 262 Why Did BRI Emphasize Savings in Its Unit Desas? 265 To finance the demand for credit and enable institutional sustainability 265 To limit government risk 267 To encourage rural savings mobilization 267 Developing and Testing the New Savings Program 267 Early field research 268 The planning stage 269 The first SIMPEDES pilot project 270 The second-stage SIMPEDES pilot project 276 Expansion and Market Penetration of the New Savings Program 279 Preparing to expand the savings program,January-March 1986 280 xiv The Microfinance Revolution: Lessons from Indonesia Expanding the savings program to all unit desas, April-September 1986 291 Learning market penetration, 1987-89 293 Unit Desa Savings Products, 1996 296 SIMPEDES and SIMASKOT 297 TABANAS 297 Deposito Berjangka and giro accounts 298 The interest rate structure 298 Cost of funds 299 Lotteries 299 The transfer price 300 Performance in Savings Mobilization, 1984-96 301 Performance by instrument 302 Notes 304 Chapter 14 Institutional Development for Large-Scale Sustainable Microfinance: The Transformation of the Unit Desas, 1984-96 306 Old and New Cultures 309 Restructuring the Unit Desas: The Transition Process 311 From branch windows to profit centers 312 Changes in organization and management 313 Unit desa location 316 Marketing and public relations 319 Learning from mistakes 321 Organization and Management of the Unit Desas, 1996 322 At the head office 323 At the regional offices 327 At the branches 327 At the unit desas 328 At the village posts 331 The unit desa system and other bank divisions 331 Reporting Unit Desa Performance at the Unit, Branch, and Regional Levels 332 The unit development reports of two unit desas 333 The aggregated unit development reports of two branches 335 Unit performance in two regional offices 335 Staff Incentives and Training 337 Incentives 337 Training 339 Retention 342 Bank Rakyat Indonesia: Microfinance Outside the Unit Desa System 343 The Badan Kredit Desa 343 Pembinaan Peningkatan Pendapatan Petani-Nelayan Kecil 347 Microfinance at BRI's branches: comparing the BKDs and P4K 350 Microbanking in a Division of a Multipurpose Commercial Bank: Structural Issues 352 External dangers to sustainable microfinance 354 Contents xv Internal dangers to sustainable microfinance 355 Why Do Structural Issues Matter? 359 Notes 362 Chapter 15 Commercial Microfinance in Indonesia: Stability in Crisis, 1997-98 364 Indonesia's Financial Crisis 369 The collapse of the banking system 373 Bank reforms and corporate debt restructuring 375 Challenges ahead 380 The Stability of Microbanking during the Crisis 382 The Badan Kredit Desa, 1997-98 382 Bank Dagang Bali, 1996-2000 384 BRI's unit desa system, 1996-2000 388 The unit desas at the start of the 21st century 396 Commercial Microfinance in Indonesia:What Has Been Learned? 402 Commercial microfinance can be both economically and socially profitable 403 Stability in crisis 403 Notes 407 Appendix Rupiah-U.S. Dollar Exchange Rates, Consumer Prices, and Performance of Bank Rakyat Indonesia's Unit Desa System, December 1996-December 2000 410 Table of Contents for The Microfinance Revolution, Volume 1: Sustainable Finance for the Poor 414 Glossary and Acronyms 421 Bibliography 433 Index 451 Tables 8.1 Economic indicators for Indonesia, 1996-2000 35 9.1 Outstanding loans in Indonesia's financial system, 1985 100 9.2 Loans and savings in Indonesia's financial system, 1995 108 9.3 BPRs and BRI unit desas: number of units and clients, September 1998 110 9.4 BPRs and BRI unit desas: average loans and deposits, 1997 111 9.5 Loans and savings in five rural financial institutions and the PHBK program 116 9.6 Estimated earnings adjusted for subsidies and bad debt for 34 original BKKs/SK and 3 LKPs of NusaTenggara Barat, 1995 121 9.7 Viability indicators for Bank Shinta Daya and Mitra Karya, 1995 129 xvi The Microfinance Revolution: Lessons from Indonesia 9.8 Estimated subsidy dependence index and required interest rates to cover the program costs of the PHBK program, fiscal 1991-96 135 10.1 Bank Dagang Bali profit and loss statement, 1970-96 151 10.2 Bank Dagang Bali balance sheet, 1970-96 152 10.3 Bank Dagang Bali outstanding loans, 1970-96 152 10.4 Bank Dagang Bali savings, 1970-96 153 10.5 Size distribution of Bank Dagang Bali loans and savings, 1996 154 11.1 Features of BIMAS and Kredit Mini 185 11.2 Household borrowing in three Indonesian villages by source of the largest loan, 1980-81 195 11.3 BIMAS participation in four Indonesian villages, 1980-81 196 11.4 Evaluation of BIMAS in four Indonesian villages 206 12.1 A comparison of BIMAS and KUPEDES 220 12.2 Size distribution of KUPEDES loans at disbursement, 1995 241 12.3 Excerpts from responses to KUPEDES borrower survey, 1996 249 12.4 Number and value of KUPEDES loans, 1984-96 253 12.5 Loss ratios for KUPEDES loans, 1984-96 253 12.6 Geographic distribution of KUPEDES outstanding loans and unit desa deposits, 1996 255 12.7 Unit desa profits and losses, 1984-96 256 13.1 Unit desa savings in the Sukabumi pilot project, 31 October- 31 December 1984 274 13.2 The most important characteristic of SIMPEDES: responses of 144 SIMPEDES savers in the Sukabumi pilot project,January- February 1985 275 13.3 Savings mobilization in the second stage of the SIMPEDES pilot project,June-December 1985 277 13.4 The most important characteristic of SIMPEDES: responses of 76 SIMPEDES savers in the 12 new branches of the second- stage pilot project, September 1985 278 13.5 "Would you still save in SIMPEDES if the annual interest rate were 6 percent?": responses of 74 SIMPEDES savers in the 12 new branches of the second-stage pilot project, September 1985 279 13.6 Annual interest rates for unit desa savings instruments, September 1996 297 13.7 Estimated cost of savings mobilization in the unit desa system, 1996 299 13.8 Value of savings accounts by account type in the unit desas, 1984-96 302 13.9 Number of savings accounts by account type in the unit desas, 1984-96 303 14.1 Employment in the unit desa system by organizational level, 1996 325 14.2 Unit development reports of two unit desas, December 1996 334 14.3 Aggregated unit development reports of two branches, 1995-96 336 Contents xviI 14.4 Performance indicators for the unit desas of two regional offices, 1995-96 337 14.5 Curriculum of training centers for unit desa staff, 1996 340 14.6 Curriculum of training centers for unit desa branch supervisors, 1996 341 14.7 Performance of the Badan Kredit Desa, 1992-96 345 14.8 Indicators of financial self-sustainability for the unit desas, 1985,1990, and 1995 360 15.1 Performance indicators for the Badan Kredit Desa (BKD), 1996 to August 1998 383 15.2 Bank Dagang Bali profit and loss statement, 1996-2000 385 15.3 Bank Dagang Bali balance sheet, 1996-2000 386 15.4 Bank Dagang Bali savings, 1996, 1998, and 2000 387 15.5 Bank Dagang Bali outstanding loans, 1996,1998, and 2000 387 15.6 Unit desa savings and lending, 1996-2000 388 15.7 Unit desa pretax profits and returns on assets, 1996-2000 388 15.8 Value of unit desa savings by account type, 1996-2000 392 15.9 Number of unit desa savings accounts by account type, 1996-2000 392 Figures 11.1 Main reasons for the failure of the BIMAS credit program for rice farmers 170 11.2 Default rates for BIMAS rice cultivation loans, 1970-84 182 11.3 Amount of outstanding BIMAS rice cultivation loans, 1970-84 182 11.4 Number of outstanding BIMAS rice cultivation loans, 1970-84 183 11.5 Area covered under BIMAS rice cultivation loans, 1970-84 183 12.1 The evolution of microcredit at Bank Rakyat Indonesia, 1970-96 223 13.1 The evolution of microsavings in the unit desas, 1984-96 264 14.1 Restructuring unit desa organization and management 310 14.2 Number of unit desas, 1983-96 318 14.3 Bank Rakyat Indonesia's organizational structure, 1996 323 14.4 Government administrative levels and Bank Rakyat Indonesia's organizational levels, 1996 324 14.5 Organizational structure of the Business Unit Desa Division at Bank Rakyat Indonesia's head office, 1996 325 14.6 Orgamzational structure of a unit desa and its supervising branch, 1996 329 15.1 Exchange rates in Indonesia, December 1996-December 2000 371 15.2 Consumer price index in Indonesia, December 1996- December 2000 371 15.3 Value of unit desa loans and savings, December 1996- December 2000 389 15.4 Number of unit desa oustanding loans and savings accounts, December 1996-December 2000 389 15.5 Real value of unit desa loans and savings, 1984-2000 391 xviii The Microfinance Revolution: Lessons from Indonesia Boxes 1 Old and new cultures at Bank Rakyat Indonesia's unit desas 15 8.1 Excerpts from David C. Cole and Betty E Slade's Building a Modern Financial System: The Indonesian Experience 31 8.2 Excerpts from Ali Wardhana's "Overcoming the Current Economic Downturn" 51 8.3 Excerpts from Lloyd R. Kenward's "AssessingVulnerability to Financial Crisis: Evidence from Indonesia" 54 8.4 Excerpts from Thomas L. Friedman's "What a Mess!" 77 9.1 Excerpts from David C. Cole and Betty F. Slade's Building a Modern Financial System: The Indonesian Experience 103 9.2 Excerpts from Camilla Nestor's contribution to Ohio State University's online Development Finance Network, 2 February 1999 107 11.1 Agricultural loans in Java in the 1920s and 1930s: Excerpts from Thomas A. Fruin's The Provisional Manualfor the Credit Business of the General Popular Bank 176 12.1 Basic principles of KUPEDES 236 12.2 The KUPEDES prompt payment incentive 243 12.3 Measures of KUPEDES portfolio quality 247 13.1 Basic principles of unit desa savings mobilization 266 13.2 Excerpts from Bank Rakyat Indonesia's unit desa savings mobilization casebook 281 13.3 Excerpts from Klaus Maurer's "Bank Rakyat Indonesia (BRI) (Case Study)" 301 15.1 Excerps from Mark Baird's "Corporate Restructuring in Indonesia" 370 15.2 Excerpts from Klaus Maurer and Hans Dieter Seibel's "Agricultural Development Bank Reform: The Case of the Unit Banking System of Bank Rakyat Indonesia (BRI)" 376 Map 1 Indonesia viii Contents xix Foreword Occasionally one meets someone with deep expertise in her chosen field. But This seminal work rarely does one meet such a person who can also explain her views with equal ease to both other experts in the field and to other interested parties without prior offers readers a knowledge of the field, such as government policymakers, central bank gover- nors, or even members of the general public. Marguerite Robinson is such a per- richness and depth son, having acquired deep knowledge of microfinance over some 20 years. She has worked primarily in Indonesia, advising the government and helping to cre- on microfinance ate Bank Rakyat Indonesia's unit desa system, one of the world's most successful microfinance programs. But Dr. Robinson has also provided her expertise to pol- that have long icymakers and directors of microfinance institutions in many other countries, in- cluding Bolivia, China, India, Kenya,Tanzania, andVietnam-to name just a few. been needed Dr. Robinson came to microfinance with a rich academic and profession- al background as an anthropologist, having spent many years in villages in India, Indonesia, and Sri Lanka. She describes herself as a financial anthropol- ogist, given her umque credentials to understand both people-particularly poor people in remote villages or urban slums not normally served by financial in- stitutions-and financial markets, and how the two interact. Few people have come to microfinance with such tools of the trade, and Dr. Robinson has honed those tools with long stints in Indonesia and other countries studying, observ- ing, researching, teaching, writing, and practicing microfinance. Now Dr. Robinson has bundled all that knowledge, and the result is a sem- inal work on microfinance that offers readers a richness and depth about the field that have long been needed.This long overdue book, The Microfinance Rev- olution, consists of three volumes.The first focuses on the paradigm shift in mi- xxi crofinance, the second concentrates on microfinance in Indonesia, and the third (written with PeterJ. Fidler) looks at the global experience with microfinance and documents the move to commercially viable microfinance. The niicrofinance field is not short on information.There are scores of case studies on microfinance institutions; technical, financial, and practical guides to the field; and wonderful reports on savings, interest rates, client desertion and delinquency, supervision, audit, appraisal, planning, and management informa- tion systems for microfinance institutions. There are also works on the impact of microfinance on poverty and some selective works on theory. We have all the bits and pieces, but no one has really seamed it all together. No one has provided an overview of how the industry has developed and where it is head- This book contains ed.And no one has provided an overarching theory that supports these views- until now. Marguerite Robinson does all that and more.The third volume, The wonderful anecdotal Emerging Industry, provides a global view on microfinance in developing coun- tries (excluding the transition economies of Central and Eastern Europe and richness supported the former Soviet Union, which Dr. Robinson decided not to cover due to her lack of experience in the region).That volume also explores theory and by a wealth offacts, the evolution of thinking on this subject. This book also contains wonderful anecdotal richness on a variety of mi- figures, tables, notes, crofinance themes: on microfinance institutions, on the voices of microfinance clients, on savings, and on moneylenders, as well as a unique assessment of In- and citations donesia that makes up the second volume, Lessonsfrom Indonesia. This rich anec- dotal material is supported by a wealth of facts, figures, tables, notes, and citations reflecting Dr. Robinson's academic rigor, a rigor that has rarely been brought to this field. The book's detail and richness are spun into a fine web supporting the author's basic thesis-that a fundamental shift is occurring in microfinance, inexorably pushing the industry to focus on commercially viable microfinance.This thesis and a detailed explanation supporting it are the main subject of the first volume, Sus- tainable Financefor the Poor. Only by making this shift, says Dr. Robinson, can mi- crofinance fill the "absurd gap" between the demand for and supply of microfinance services.That gap is huge: at least 80 percent of the 900 million households in low- and lower-middle-income countries do not have access to formal financial services. Most microfinance institutions are nongovernmental organizations (NGOs), often providing an array of social services.They focus on microfinance as a so- cial intervention or a poverty alleviation tool.They see a dilemma between achiev- ing commercial viability and serving the poor. For the most part they are not viable financial institutions and do not mobilize domestic savings or raise com- mercial funds.And they are largely dependent on donors to subsidize their op- erations.Yet the microfinance industry barely scratches the surface of its market potential, and the industry as currently structured cannot meet this need. But increasingly, as spelled out in this book, commercially viable microfi- nance institutions are being established as banks or nonbank financial institu- tions.They operate from a financial systems perspective, and they see microfinance as filling an important niche in the financial system by providing financial xxii The Microfinance Revolution: Lessons from Indonesia services-for profit-to the working poor. The only way to close the absurd gap between demand and supply in microfinance is for microfinance institu- tions to mobilize savings, to raise capital commercially, and to service clients through extensive branch networks.This is increasingly the case in Latin Amer- ica, as illustrated by the book's analysis of Bolivia's BancoSol. It is also true for a few large microfinance institutions in Asia, such as Bangladesh's Association for Social Advancement. Bank Rakyat Indonesia's unit desa system best illus- trates the benefits of long-term adherence to commercial principles of micro- finance, which is why thls case is an important contribution of this book. Let me try to sum up what this work offers to readers: * A detailed overview of the development of microfinance over the past 20 The author's basic years. * A global view of microfinance in the developing world. thesis is that a * A thesis on the future path of microfinance. * A coherent theory about microfinance-why it works when so many other fundamental shift development interventions fail. * Exquisite detail on a number of important microfinance topics-such as in- is occurring in formal moneylending and savings. * An important study of Indonesia, with detailed analysis of Bank Rakyat In- microfinance, donesia. * Brief studies of many other microfinance institutions inAfrica,Asia, and Latin inexorably pushing America. the industry tofocus This book reflects Marguerite Robinson's longstanding experience in mi- crofinance. Readers will quickly understand that Dr. Robinson is one of the few on commercially people with deep knowledge in her chosen field-as well as the ability to con- vey that knowledge simply and clearly to a broad range of interested readers. viable microfinance Ira W Lieberman Former Chief Executive Officer, Consultative Group to Assist the Poorest, 1995-99 Senior Manager,World Bank Foreword xxiii Introduction We in Indonesia have a special, longstanding interest in the emerging microfi- We in Indonesia nance revolution, which has made it possible for large numbers of low-income people to access institutional financial services-often for the first time. Financial have a special, services that are widely available in rural areas and in low-income urban neigh- borhoods help the poor improve their financial security, allow them to take ad- longstanding vantage of business opportunities, and facilitate the growth of their enterprises. In Indonesia sustainable microfinance in the formal sector began in 1970 with interest in the the opening of Bank Dagang Bali (BDB), a private bank in Bali, and attained nationwide coverage with the 1984 restructuring of the unit desa, or local bank- emerging ing, system of the state-owned Bank Rakyat Indonesia (BRI). BRI's umt desa system is now the largest financially self-sufficient provider microfinance of sustainable microfinance in the developing world. Indonesia's approach to mi- crofinance-making it profitable, and so widely available-helped the country revolution reduce the incidence of poverty from about 40 percent of the population in the mid-1970s to about 11 percent in 1996. In 1997, when the East Asian economic crisis began and poverty in Indonesia started to rise, BRI's microfinance system helped poor people who had lost their jobs finance informal sector enterpris- es. It also gave them secure and convenient deposit facilities-especially impor- tant to poor people in times of crisis. Hindsight is, as we all know, a powerful analytical tool. In reviewing the re- structuring of BRI's microbanking system, one can identify a number of com- ponents that might better have been done differently. In the 1970s, for example, BRI opened more than 3,500 village units to channel subsidized governrnent cred- it to rice farmers through BIMAS, the credit component of Indonesia's massive xxv rice intensification program.As it turned out, the rice intensification program was highly successful, but its credit component was not.The long-term results ofBIMAS were similar to those found in many developing countries.The subsidized loans, being at below-market interest rates and so in demand by wealthier farmers, often did not reach poor farmers. Moreover, arrears and losses were high.The program was phased out in 1985. Meanwhile, BRI's unit desa system also tried to mobi- lize savings. However, since the government required that banks lend at 12 per- cent and pay 15 percent on most deposits, there was a negative incentive for the banks to mobilize savings-and the incentive structure worked well! During the 1970s and 1980s rural borrowers who qualified for loans larg- er than those available at the unit desas also had the option of obtaining sub- By the early 1 980s sidized government credit through bank branches in district capitals.The Small Investment Loan Program, known as KIK, and the Small Permanent Working we began to realize Capital Loan Program, known as KMKP, provided loans of up to 15 million rupiah ($36,145 in 1975 and $13,333 in 1985).But these programs also resulted that not only were in high arrears and large losses to both the banks and the government, and were eventually phased out. our subsidized credit By the early 1980s we began to realize that year after year, the subsidies and arrears of BIMAS, KIK, and KMKP were large, the programs were inefficient, programs not driving and the loans generally did not reach the intended borrowers. In brief, our ap- proach to local finance was ineffective and unsustainable. Not only were our rural development, subsidized credit programs not driving rural development, they were actually slowing it down! Having recognized the severe deficiencies of these programs, they were actually we decided in 1983 to begin a new program for rural finance based on prin- ciples of commercial finance. slowing it down! But in 1983, when the Indonesian governmnent began to implement a vari- ety of financial reforms, we did not have good models or examples-or even ap- proximate ones-from other, similarly positioned countries. In many ways Indonesia was a pioneer in implementing financial reforms, and the reform of the unit desa system is a prime example.When we decided to transform it into a commercial microbanking system, we could find no example of a financial institution in any developing country that provided microfinance profitably on a large scale. The development of commercial microbanking in BRI's unit desas can best be understood in the context of the broad set of economic reforms implemented by the Indonesian government. On the whole these reforms reflected a con- sistent intent to achieve three basic objectives: * To move toward a predominantly market-based financial system. * To provide effective protection, as needed, so that the general public could benefit from the services offered by the financial system. * To build a financial system that would support the stable, healthy growth of the national economy. To move effectively toward achieving these aims, in 1983 we began to intro- duce a series of far-reaching finance, tax, trade, and investment reforms. xxvi The Microfinance Revolution: Lessons from Indonesia The oil boom of the mid-1970s through the mid-1980s had been a mixed blessing for Indonesia. One economist, writing about oil-exporting countries, concluded that the boom left most economies no better off than they would have been if oil prices had stayed at 1972 levels in real terms. But unlike most oil exporters, Indonesia capitalized on its windfall oil revenues. Even when we had ample oil revenues, we looked ahead to days when we might not be so for- tunate. A critical element of Indonesia's development strategy has been to stimu- late rural development, rural incomes, and rural employment. Thus in the 1970s a large share of our oil wealth was invested in agriculture-especially ir- rigation and new rice technologies-and in infrastructure, education, and health. Much of this investment was in rural areas, where about 80 percent of A 1983 deregulation the population lived in the mid-1970s.This investment helped ensure that agri- culture and other rural industries would continue to support rural income growth made possible the and create employment-an essential part of the foundation for our econom- ic growth since the mid-1980s. transformation of It is important to understand that this was not "trickle down" growth. Our approach to economic growth incorporates some of the poorest groups in the BRl's unit desa economy. Our food supply, especially rice, depends on the increasing produc- tivity of small farmers-supported by the government's massive rural invest- systemfrom a ment. Our e4ort drive is based on the growth of firms that create jobs for low-skilled workers. Some of the country's largest industries-including con- channeling agentfor struction, transportation, retail trade, and other services-employ large num- bers of unskilled workers, especially in the informal sector.These service sectors targeted, subsidized are quick to respond to rapid growth in other sectors of the economy. As the incomes of poor people rose, their demand for banking services in- government loans to creased.The reform of BRI's microbanking system was undertaken in order to bring about a major increase in the availability of financial services-initially a profitablefinancial for the rural population and later for low-income urban residents as well. De- cisions to provide microbanking services dehvered at the subdistrict level intermediary throughout the country, to pay positive real interest rates for savings, and to charge loan rates sufficient to cover all costs and to earn a reasonable profit for the bank were consistent with our overall reform agenda. Financial reforms were extended to rural areas with the government's first major financial deregulation package, issued in 1983.That deregulation abol- ished credit ceilngs and permitted banks to set their own interest rates on most loans and deposits. This made possible the transformation of BRI's unit desa system from a channeling agent for targeted, subsidized government loans to a profitable financial intermediary providing small loans and deposit services to clients in rural areas throughout Indonesia. In 1989 BRI extended its mi- crobanking services to urban areas as well. When making the decision to reform the unit desas, we asked ourselves three questions. First, would there be local demand for credit at the interest rates need- ed for BRI to cover all its costs and earn a profit? We studied the demand for small loans in different areas and found it to be very large. Poor borrowers were Introduction xxvii paying much higher interest rates to local moneylenders, and it seemed that they would generally welcome the rates that BRI would charge. Second, would people place their savings in BRI's village units? We con- ducted studies in many parts of the country and found huge demand for sav- ings services if the deposit instruments and services were designed to meet the needs of poor savers. Third, with an eye on the government budget, we asked how long it would take for the restructured microbanking system to break even and begin to make a profit. Under the assumptions we used, we predicted that the sys- tem, which began in 1984, would break even in two years-which it did in just under two years. And it has been profitable every year since. The BRI reforms Our approach to reforming BRI was market-based: in BRI's thousands of local mnicrobanking units, performance-based cash awards and other incentives have enjoyed motivate staff to act as bankers. Unit personnel also required tralmng to change their behavior. Most important, unit staff had to learn about the markets they remarkable success. served. Responsibility for loan decisions had to be delegated from branch of- fices to village units, while regional offices had to de-emphasize their control- The system has minded approach and become more oriented toward promotion. In a large, complex institution like BRI, these changes took careful planning and imple- been profitable since mentation.The restructuring of BRI's unit desa system was a major institutional reform-and it succeeded. As a result savers have a secure outlet for their 1986 and without funds, on which they generally earn positive real returns, while borrowers with productive uses for small loans have access to credit on commercial terms. subsidy since 1987 The BRI reforms have enjoyed remarkable success.The umt desa system has a single loan product, KUPEDES, that offers loans of 25,000-25,000,000 rupiah ($3-$3,406 in 1999) for any productive purpose. Most KUPEDES loans carry an effective annual interest rate of about 32 percent if payments are made on time. Savings instruments offer a choice between different combina- tions of liquidity and returns-enabling depositors to combine the products in ways that best meet their needs. Unit desa deposits, a highly stable source of funds, finance all KUPEDES loans.The system has been profitable since 1986 and without subsidy since 1987. Contrary to much international experience with rural finance, KUPEDES has had very high repayment rates. In Indonesia we have found that a less regulat- ed economy, with widespread access to institutional finance at the local level, can open new opportunities to people previously excluded from full partici- pation in the country's economic growth. But in 1997 a severe financial and economic crisis developed that affected all EastAsian economies, fromThailand toJapan and the Republic of Korea. Indonesia's currency fell from 2,450 rupiah per U.S. dollar in June 1997,just before the cri- sis began, to about 17,000 at its weakest point in 1998. The rupiah then recov- ered to levels of 7,000-8,000 in the fourth quarter of 1998.At the end of 1999 there were 7,430 rupiah to the U.S. dollar. Indonesia's average annual inflation for 1998 was 57.6 percent-a sharp contrast to the 1980s, when annual inflation had stayed below 10 percent.,GDP, which had grown by more than 7 percent a year xxviii The Microfinance Revolution: Lessons from Indonesia for over a decade, grewjust 4.9 percent in 1997 and fell 13.7 percent in 1998. But in 1999 inflation dropped to 20.5 percent, while GDP rose to 0.2 percent. The economic meltdown that hit Indonesia-and one can hardly describe it differently-had multiple causes. Some were self-inflicted; others were ex- ternal. Among the external events I would list the sharp decline in the world oil price, a decline in prices for other primary product exports, and a serious drought in 1997. But to explain the severity of Indonesia's economic crisis rel- ative to that of our neighbors, we have to look at internal weaknesses. Let me highlight two. First, our financial institutions were encouraged to fund risky, unprofitable ventures. Government officials could and did direct loans to favored firms and activities. Loans were rarely subjected to even the most rudimentary econom- ic and financial analysis. Second, the involvement of well-connected parties in The economic many economuc activities led to a problem of moral hazard: in the presence of a perceived guarantee, implicit or explicit, there is little incentive to avoid risky meltdown that hit behavior. In addition, actions by the government and the central bank suggested that Indonesian banks would be protected from failure. Our foreign exchange Indonesia in regime also encouraged risky behavior that, after the depreciation of the ru- piah, resulted in unmanageable debt that effectively bankrupted a substantial 1997-and one can portion of our corporations. At this writing more than two years have passed since Indonesia's crisis began. hardly describe it While it may be too optimistic to say that the crisis has passed, much has been accomplished, and there is general consensus on what needs to be done to get dfferently-had our economy back on track.With assistance from the International Monetary Fund,World Bank,Asian Development Bank, and others, an economic reform multiple causes. Some program was introduced in 1998. Structural reforms are under way. Safety net policies to protect the poor have been given high priority.The weaknesses in were self-inflicted; the financial system have been clearly identified, a bank restructuring program is in process, and the legal and supervisory framework for the banking sector others were external is being strengthened. Emphasis is being placed on making our capital markets more transparent and better regulated. Many other reforms are also in process. Numerous policy measures must still be implemented, but my prediction is that the crisis will pass and growth will resume. In the 1980s and 1990s the rapidly growing Asian economies created a base of human and physical infra- structure, and that base remains intact. It is on this base that we will eventual- ly be able to resume rapid growth. While it has been important to identify our weaknesses in order to rebuild the Indonesian economy, it is also important to identify the institutions that remained strong throughout the crisis and to understand the reasons for their strength and stability. One of those institutions is related to the subject of this book: commercial, sustainable microfinance. In sharp contrast to the Indonesian banking sector generally, commercial microbanking at BRI's unit desa system continued its wide outreach, high re- payment rate, and profitability throughout 1997-99.The system remained sta- ble and profitable throughout the crisis. Deposits in the umt desas more than Introduction XXix doubled in rupiah terms, from 7.7 trillion rupiah ($3.2 billion) in June 1997 (the month before the crisis began) to 17.1 trillion rupiah ($2.3 billion) at the end of 1999. The number of savings accounts increased from 17.0 million in June 1997 to 24.2 million at the end of 1999. KUPEDES lending has remained stable. In June 1997 there was 4.3 tril- lion rupiah outstanding ($1.75 billion) in 2.5 million KUPEDES loans. By the end of 1999 the outstanding loan balance was 6.0 trillion rupiah ($802 mil- lion) in 2.5 million loans. The repayment rate, 98 percent in June 1997, was also 98 percent in December 1999. In 1998, the worst year for the Indonesian economy in the past three decades, pretax unit desa profits were 714 billion rupiah ($89 mlllion), while the pretax return on assets for the unit desa system In sharp contrast to was 4.9 percent. In 1999 pretax profits were 1.2 trillion rupiah ($160 million) and the pretax return on assets was 6.1 percent. the Indonesian BRI's microbanking system emphasizes understanding local markets and meeting the demand for financial services from low-income households and banking sector enterprises. It provides products and services designed to be appropriate for this market segment.We now know that the unit desas are so robust that they have generally, commercial withstood an extraordinary national economic and financial crisis.This strength in microbanking has helped to mitigate the effects of the crisis on the poor and microbanking at to improve the foundations for future economic development. The creation of BRI's unit desa system cannot be separated from Dr. Mar- BRI continued its guerite S. Robinson. She actively participated in developing the unit desas into what is now a strong, viable microbanking system that provides finan- wide outreach, high cial services to low-income people in rural and urban areas throughout In- donesia. To ensure that the system would function effectively for local repayment rate, and communities-consisting largely of small farmers and microentrepreneurs- Dr. Robinson visited many unit desas and the villages they served. She co- profitability ordinated research teams that surveyed the income flows and savings habits of local people, studied their need for capital and their demand for financial throughout the crisis services, and assessed opportunities for investment in the community.The stud- ies covered villages in Java, Sumatra, Kalimantan, Sulawesi, and other Indonesian islands, and resulted in ongoing recommendations to the Ministry of Finance and the BRI about unit desa instruments and services that would be appro- priate for local demand. When decisions were made to add new savings and loan products and ser- vices in the umt desa system, to open new units, or to expand unit desa oper- ations to urban neighborhoods, Dr. Robinson advised BRI, assisting with staff training and advising on the management and supervision of unit operations. She has often returned to the units to learn whether they function properly and to advise BRI on the development of its microbanking system. BRI's unit desa system has made great progress since 1984, rapidly becoming a financial institution capable of contributing to rural development and rural employment. It has also expanded to serve low-income urban areas.At the same time, the unit desas have a considerable way to go-and like many newly de- veloped financial institutions, they face problems of institutional development. xxX The Microfinance Revolution: Lessons from Indonesia Dr. Robinson deserves credit for her active role in creating the unit desas, not only at the initial stages of their restructuring but also during the entire period of their subsequent development. This book reflects her deep insight and thorough knowledge of BRI's unit desas. During the 1990s BRI's unit desa system received nearly 1,000 interna- tional visitors from more than 30 countries.The bank has had to create a sep- arate office to serve the many international visitors to the unit desas.A number of the visitors have also visited Bank Dagang Bali, well-known as the earliest bank to institute commercial microfinance, as well as some of Indonesia's other financial institutions that provide services at the village level on a commercial basis, such as the Badan Kredit Desas (Village Credit Organizations) ofJava and Madura. We are especially Many developing countries in Asia, Africa, and Latin America are at dif- ferent stages of learning about and implementing institutional commercial mi- pleased that our crofinance. This book documents Indonesia's experience with sustainable microfinance, explores the spread of commercial financial services to low-in- approach to sustainable come people in other countries, and analyzes the ideas that underlie both. Indonesia, which has played a leading role in the Non-Aligned Movement, microbanking is usefui is active in transferring technology and sharing experiences that lead to eco- nomic growth, equity, and stability in the developing world.We are especially for other developing pleased that our approach to sustainable microbanking-which has provided poor people in Indonesia with new opportunities for economic growth and countries financial security-is useful for the development of microfinance in other de- veloping countries. Ali Wardhana Minister of Finance, Government of Indonesia, 1968-83 Coordinating Minister for Economics, Finance, and Industry, Government of Indonesia, 1983-88 Economic Adviser, Government of Indonesia, 1988- Introduction xxxi Preface When I began life as a social anthropologist in the 1960s, carrying out field work Over the years I in remote areas of developing countries in Asia, outsiders rarely visited the vil- lages where I lived. Those who did come, other than the occasional scientist, noticed that the were missionaries of various religions. Over the years I noticed that the few out- siders I encountered in the field were increasingly less likely to be missionar- few outsiders I ies-and more likely to be bankers.This gradual change was my first introduction to the then-embryonic microfinance revolution. encountered in The bankers who began showing up in small villages on their bicycles, mo- torcycles, or jeeps in the 1960s and 1970s were usually employees of local villages were less branches of state-owned banks.They came along with the green revolution.Their mission, as assigned by their governments and assorted international donors, was likely to be to find trustworthy villagers to whom they could provide credit. This, it was thought, would both help feed the population and increase rural economic growth. missionaries-and The bankers and the missionaries, who shared much of the same client pool, were curiously alike in some ways. Usually outsiders to the local community, both more likely to be tended to discover m the vlllages their own preconceptions, rather than the local realities and dynamics (a problem to which, of course, anthropologists are not im- bankers mune!). But many cared about helpmg poor people increase their incomes and improve thelr lives, and some were quite successful.They came with powerful ideas, found others already present, and often became catalysts for the cross-fertilization of thought and sometimes for the introduction of social and economic reforms. I watched from villages in different countries and continents as, over the decades, the balance switched from outsiders bringmg religion to outsiders bring- ing finance. Of course, those who lived in the villages already had both. As an xxxiii world began to develop commercial microfinance programs.Though the pro- grams differed, the underlying principles were similar. Gradually, a paradigm shift took place-from the delivery of government- or donor-subsidized credit to the development of sustainable financial intermediaries that capture local sav- ings, access commercial finance, and lend these funds to low-income borrow- ers at interest rates that enable full cost recovery and institutional self-sufficiency. The microfinance revolution developed in the 1980s (before it had a name) and came of age in the 1990s. It occurred when the many advances of previ- ous decades in market knowledge, lending methods, and savings mobilization were combined with a commercial approach to financial intermediation for low- income people, making financially sustainable formal sector microfinance pos- Gradually, a paradigm sible.This breakthrough-which also required the development of organizational structures and management resources capable of delivering microfinance ser- shift took place- vices profitably throughout an entire country-first occurred in Indonesia in the 1980s and then in Bolivia in the mid-1990s. from the delivery of Commercial microfinance is now found in many countries, where it is at different stages of development. In its most advanced form, in banks and other government- or donor- formal financial institutions, all microloans are fully financed by savings, com- mercial debt, for-profit investment, and retained earnings (in a variety of forms subsidized credit to the and combinations).As a result all savers and all creditworthy borrowers can be served, repeat borrowers can be accommodated as they expand their enterprises development of and qualify for larger loans, and many economically active poor people can be helped out of poverty. Industry standards for commercial microfinance began sustainable financial to develop in the 1990s.And in some countries intense competition has erupt- ed among commercial financial mstitutions aiming to attract the business of poor intermediaries clients. Nevertheless, in most developing countries the formal financial sector still does not serve microfinance cients.The traditional view-that it is neither im- portant nor profitable for institutions to provide commercial financial services to low-income people is still widely held.The microfinance revolution is still emerging. But it is probably irreversible: because there is massive unmet de- mand for microfinance, because it has been proven that this demand can be met profitably on a large scale, and because information about the profitability of financing the economically active poor has begun to be widely disseminated. Microfinance in the developed world is beyond the scope of this book. But many low-income people in industrial countries lack access to financial ser- vices, also with pervasive negative effects on society and the economy. Rich countries could learn many lessons on sustainable microfinance from developing countries. A number of people have asked whether, because this book is in three vol- umes, it is intended to be a reference book.While to some extent it can be used for reference, the book was not written primarily for that purpose. Rather, it is an analytical narrative on why and how capital is becoming democratized on a global scale for the first time in history. The reason the book is in three volumes is that it concerns a major revolution of our times. XXXVi The Microfinance Revolution: Lessons from Indonesia A reader who wants to learn about a particular microfinance institution- such as Bangladesh's Association for Social Advancement, Bolivia's BancoSol, or Mexico's Compartamos-can find these institutions in the indexes of these volumes and read about them. But the approach is not encyclopedic. The aim is not a comprehensive sunmmary of the institution, but an emphasis on its con- tributions (and in some cases lack thereof) to the microfinance revolution. Finance for the poor is a topic on which many opinions are held, usually passionately. This book wllU undoubtedly be controversial, as is intended. But microfinance is unusual.As in any emerging industry, debates are endemic. But in mucrofinance these debates are among people who work every day to in- crease the employment opportunities, incomes, and self-confidence of the poor.These are debates among good people. In presenting new data and analy- Financefor the poor ses and in reexamining long-held assumptions and conclusions, this book aims at stimulating constructive dialogue-in ways that will help financial institu- is a topic on which tions meet the demand for microfinance sustainably and soon. many opinions are What Is an Anthropologist Doing in Banking? held, usually During my first decade as an anthropologist, I conducted the kinds of research passionately. This I was taught at Harvard and Cambridge universities: studying the people of dif- ferent societies and recording, comparing, analyzing, writing, and teaching book will about their cultures and social structures.While the education I received was well suited for its multiple purposes, there was little in it to prepare me for the undoubtedly be fact that most of the people I would study would be poor-and in some cases starving, abused, and in bonded servitude. controversial, as is As part of extended field work in a very underdeveloped rural area in India, I had long conversations with many bonded laborers, members of "un- intended touchable" castes, and others among the desperately poor and disenfranchised. Once after a long discussion, I rose to leave a small group from whom I had been learning about their social and economic activities and their political environment.We had been sitting on the mud floor of a small, crowded, win- dowless house that provided only minimal protection from the driving rain of the monsoon. One of the men with whom I had been talking said to me,"We are pleased that you are interested in us, that you visit our houses, and that you sit and talk with us.We try to tell you whatever you want to know. But we would like to ask you a question.There is something that we cannot understand.We are sit- ting here in the mud because this is all that we have. Can you not see that we are cold and wet, and that we are poor and have nothing? But you are edu- cated and wealthy.Why do you want only to sit here and learn about our cus- toms? Why do you not also use your knowledge and resources to help us to have better lives and to improve our customs?" He was right. Since then, while continuing my anthropological research, I have worked on the social and economic development of the poor people in Preface XXXvii the societies I try to understand. Since 1979 I have worked simultaneously as an anthropologist and as a policy adviser to governments and financial institutions. As an anthropologist working on microfinance, I analyze local markets and their wider networks, the economic activities of their participants, and the na- ture and extent of the demand for financial services. My knowledge of local fi- nancial markets comes largely from those who participate in them: people of varying ages and both genders employed in a variety of occupations, from dif- ferent social, economic, religious, and political backgrounds. My anthropolog- ical training stands me in good stead here. I try to learn from whom, and at what cost, they obtain credit; how their credit options are linked with transactions in other markets; in what forms and for what purposes they save; and what they This bookfocuses like and dislike about their current methods of borrowing and saving. In the process it becomes possible to learn how informal credit markets, on how the demand government interventions, and bank programs work at the local level.The vest- ed interests that might oppose the development of institutional commercial nui- for microfinance crofinance in particular regions can be identified, and attention can be given to how such interests can be challenged, circumvented, or co-opted. It then be- can be met on comes feasible to design financial instruments and services appropriate for the social, political, and economic environment in general, and for the varied types a global scale of local demand in particular. As a policy adviser on microfinance, my role has been to learn the country's policy goals and its constraints, to provide infor- mation to decisionmakers about their country's microfinance demand and its relevance to development more broadly, and to suggest strategies to achieve pol- icy objectives, drawing on lessons from the country's financial markets and from international best practices in commercial microfinance. Plan of the Book This book focuses on how the demand for mucrofinance can be met on a glob- al scale. It documents the contributions of institutions and of people who have led the development of commercial finance for the poor, and it analyzes the principles on which the microfinance revolution is based.The book's intend- ed audience is diverse, including those with interests directly related to micro- finance, such as policymakers and other government officials, microfinance practitioners, social scientists, economists, bankers, and donors; those with more general interests in social and economic development and in the fundamen- tals of poverty reduction; and those drawn to difficult problems that can be solved only through an interdisciplnary approach. But this book is limited in a number of ways. Among these, no attempt is made to cover all the many types of financial institutions that provide some form of finance to the poor; emphasis is placed instead on the lessons from leading commercial nucrofinance providers. Second, it was not possible to cover all re- gions. For example, Eastern Europe, which has seen important growth in mi- crofinance since the breakup of the Soviet Union, is largely omitted from the xxxviii The Microfinance Revolution: Lessons from Indonesia discussion here. My impression is that microfinance in former centrally planned econouies is somewhat different from mucrofinance in most developing coun- tries, but I am not knowledgeable enough about transition economies to in- clude them in the discussion.Third, this is not a "how to" book for microfinance institutions on specifics such as operations, business planning, or financial analy- sis-though aspects of these topics are discussed, and references are provided to excellent sources on these subjects. Fourth, important as the topic is for poverty reduction and human rights, this book does not focus on gender issues. Many microcredit institutions have tar- geted poor women as clients and, as demonstrated in chapter 3 and elsewhere in the book, there is litde doubt that this approach has helped women and their fam- ilies increase their incomes and self-confidence. But this book is about large-scale Thefirst volume sustainable microfinance for all economically active poor people, women and men. Fifth, except for what clients of microfinance institutions tell us in their own considers the shift words, this book does not focus on the impact of microfinance on clients' house- holds or enterprises. Money is fungible, and the use of small loans and savings from subsidized is difficult to track accurately. Most impact studies on microfinance have deep methodological flaws, although breakthroughs are beginning and better knowl- microcredit to edge of the impact of financial services on the lives of the poor can be expected in coming decades. commercial Finally, except in the second volume-which provides extensive discussion on the development of microbanking in Indonesia-it has not been possible microfinance to provide the historical, macroeconomic, political, legal, and regulatory back- grounds for the development of microfinance institutions in the many coun- tries discussed. Because of these and other areas not discussed or covered in only a limit- ed way, I have called attention throughout the book to relevant works by mi- crofinance practitioners, bankers, financial analysts, economists, and others that will be helpful to readers pursumg in more depth specific components and analy- ses of the growing microfinance industry. Despite the book's omissions, I be- lieve it tells a critical story-one that is little known outside the microfinance industry.Writing this book brought to mind Charles Kindleberger's statement, "My thesis does not rest on small differences in quantities-or so I believe" (Kindleberger 1996 [1978], p. 5). There are difficulties in writing about a revolution in process. New ideas and practices spawn others. Realities change.Thus the emphasis here is on the principles and processes of the microfinance revolution. This book will soon be outdated as a current description of the state of microfinance-but not, it is hoped, as an analysis of the development and meaning of the microfinance revolution. The first volume: sustainable finance for the poor The book's first volume, which considers the shift from subsidized microcre- dit to commercial microfinance, has two parts. Part 1 (chapters 1-3),"The Par- adigm Shift in Microfinance," explores the reasons for the massive gap between Preface xxxix the low level of commercial microfinance generally available from financial in- stitutions and the extensive worldwide demand for such services among the poor. Chapter 1 explores the differences between the poverty lending approach to finance for the poor and the financial systems approach.The poverty lend- ing approach emphasizes lending to the poorest of the poor, while the finan- cial systems approach focuses on lending to the creditworthy among the economically active poor-people with the ability to use small loans and the willingness to repay them-and on voluntary savings mobilization. In this context a poverty alleviation toolbox is introduced. The tools in- clude food, employment, financial services, education, health care, infrastruc- Large-scale outreach ture, and the like. Credit is a powerful tool that is used effectively when it is made available to the creditworthy among the economically active poor. But is shown to depend other tools are required for the extremely poor, who have prior needs such as food, shelter, medicine, training, and employment. on institutional The focus then turns to the recent shift in microfinance from government- and donor-subsidized credit delivery programs to financially self-sufficient in- seif-sufficiencyfor stitutions providing commercial nucrofinance. The link between institutional self-sufficiency and large-scale outreach to low-income clients is examined; large- long-term viability scale outreach is shown to depend on institutional self-sufficiency for long-term viability. Chapter 2 introduces the emerging paradigm shift, considers how and why it is occurring, and discusses the implications of sustainable microfinance for social and economic development. In chapter 3 the focus shifts from insti- tutions to clients. Clients of microfinance programs in different countries pro- vide their views on the role that financial services have played in their economic activities, income growth, and household development.The voices of these clients show that microfinance helps them expand and diversify their enterprises, in- crease their incomes, raise their living standards and those of their families, and gain self-confidence. Their statements indicate strong underlying commonal- ities in microfinance demand across countries, economies, and institutions. Part 2 (chapters 4-7), "Theories of Local Finance: A Critique," reviews the theoretical background of microfinance. Four main streams of literature are analyzed. Chapter 4 considers supply-leading finance theory, its resulting subsidized credit programs, and the criticisms of this approach that have filled the literature for more than 20 years. Chapter 5 examines the imper- fect information paradigm and considers asymmetric information and moral hazard as these concepts have been applied to rural credit markets. The lit- erature on informal commercial credit markets and market interlinkages is reviewed in chapter 6, while that on the role of savings in microfinance is explored in chapter 7. These chapters share a common thread. They examine a variety of theo- ries and models that, when applied to microfinance markets, have impeded the development of formal sector commercial microfinance. The theorists' inten- tions were not to create obstacles to financing the poor, but that was often the xl The Microfinance Revolution: Lessons from Indonesia result.The theories are contrasted with the ways real microfinance markets work, and suggestions are offered for improving the theoretical framework for microfinance. The second volume: lessons from Indonesia Indonesia's exceptional accomplishments in microfinance are documented and analyzed in volume 2, which forms part 3 of the book ("The Indonesian Ex- perience," chapters 8-15). In one sense the Indonesian experience takes up a disproportionate amount of space in this book, partly because it is the exam- ple that I know best. But the choice of Indonesia for detailed examination, and particularly the long case study of Bank Rakyat Indonesia's (BRI's) mi- crobanking system, can be justified on other grounds. The second volume Indonesia is home to what is, to the best of my knowledge, the world's old- est commercial microfinance institution-the Badan Kredit Desas (BKDs), which documents and began in 1896. It is also home to Bank Dagang Bali (BDB), a private bank that opened in 1970 and is thought to be the world's oldest licensed, full-service analyzes Indonesias commercial bank providing continuous, profitable microfinance services on a substantial scale. And it is home to the world's largest fully self-sufficient mi- exceptional crofinance system: the microbanking division of BRI, which has operated profitably on a nationwide scale, without subsidy, since 1987. accomplishments In addition, it was possible to discuss here only one institution at consid- erable length and detail, and BRI's microbanking system is much less well known in microfinance internationally than some microfinance institutions in other countries that have been written about extensively. Emphasis is placed on the reasons the microfinance revolution emerged on a large scale in Indonesia, on the ways this occurred, and on the lessons for other countries. Chapters 8 and 9 present material on Indonesia's history, economy, and society (chapter 8) and on its rural development and rural financial mar- kets (chapter 9).These chapters provide the background for understanding why commercial microfinance developed in Indonesia nationwide, turning on its head the conventional wisdom of the time. Chapter 10 examines the history and performance of BDB. Chapters 11-15 document and analyze the remarkable restructuring of BRI's nation- wide local banking system from a government-subsidized credit program with high arrears and substantial losses during 1970-83 to a profitable, unsubsidized microbanking system beginning in 1984. The Indonesian section of the book, which was first written in early 1997, provides detailed material through 1996. It documents and analyzes the histo- ry of Indonesia's commercial microbanking over more than 25 years, a period when the country achieved unprecedented economic growth and massive so- cial and economic development. But in mid-1997 Indonesia was hit by its biggest economic, financial, and political crisis in three decades. The crisis that began in mid-1997 affected Southeast Asia and some East Asian countries, causing steep currency devaluations, plunging stocks, widespread bank failures and corpo- rate bankruptcies, loss of foreign investment, rising inflation, growing unem- Preface xli ployment, and increasing poverty throughout much of the region. For reasons discussed in chapter 8, Indonesia was hit hardest by the crisis. Given both the deadlines for this book and the importance of the post-1996 Indonesian material, certain compromises were adopted in writing part 3. Chapter 8 was revised to provide an introduction to Indonesia through 2000, and chapter 9 was updated with post-1996 material on rural finance. Chapters 10-14 analyze microbanking in Indonesia through 1996. But chapter 15, which concludes part 3, updates the microbanking material through 2000 and compares the 2000 results with those of 1996. Nearly everyone in Indonesia was affected by the crisis. Despite massive ef- forts by the government-aided by international agencies-to provide food, em- The Indonesian crisis ployment, and social safety nets, many low-income households faced very difficult times. Their purchasing power shrank substantially, many workers were laid off offers some basic as businesses closed or were retrenched, bank savings declined sharply in value, and some who had emerged from poverty slipped back under the poverty line. lessons about the Of crucial importance for this book is that as all this occurred and while the country's financial system collapsed, Indonesia's commercial microbanks re- extraordinary stability mained stable.They continued to serve mnllions of low-income households with- out any major interruption. In general these institutions saw the amount of rupiah that sustainable savings and the number of savings accounts increase considerably, loans held steady, repayments continued to be high, and the microbanks remained profitable. microfinance Thus the Indonesian crisis offers some basic lessons about the importance of microfinance to low-income households, and about the extraordinary sta- institutions can bility that sustainable microfinance institutions can maintain in a highly un- stable environment.Thus part 3 demonstrates how BRI's microbanking system maintain in a highly was transformed from a loss-making rural credit program to the world's largest sustainable microfinance system-and how it has continued to attain profitability unstable environment and wide outreach through good times and bad. The third volume: the emerging industry The book's third volume, in two parts, analyzes the emerging microfinance in- dustry, suggests a microfinance model for 2025, and discusses policy issues likely to be important for the microfinance industry over the next 25 years. Part 4,"Microfinance in Developing Countries:A GlobalView" (chapters 16-20), written with PeterJ. Fidler, analyzes the history and performance of selected institutions that have played key roles in the microfinance revolution- village banks, credit unions and cooperatives, NGOs, banks created by NGOs, commercial banks, central banks and bank superintendencies, microfinance net- works, international organizations and donors, and others. Its focus is on the creation and rapid spread of underlying principles and best practices of the new paradigm in varied institutional and country contexts in Asia, Latin America, and Africa, and on the further dissemination of these principles and practices. The microfinance revolution can be said to have reached a region when competitive institutions in the formal financial sector provide appropriately de- signed small loans and savings services (and in some cases other products as well), xiii The Microfinance Revolution: Lessons from Indonesia serve low-income clients efficiently, and price their products to cover all costs and risks-and when together these institutions provide financial services to a large share of the country's low-income households and enterprises. Chapter 16 offers a brief introduction to the history of microfinance as it developed in multiple regions.The contributions to miicrofinance made by non- bank financial mstitutions such as village banks, credit unions, and cooperatives- as well as the limitations of most of these institutions-are the focus of chapter 17. NGOs, along with regulated financial funds and companies (some of which are recent creations of NGOs that decided to expand microfinance outreach), are considered in chapter 18.The role of banks in microfinance is discussed in chapter 19, which highlights both the historical reluctance of most banks to enter microfinance and their growing interest in the market today. A few banks The third volume are selected for detailed discussion because of their special relevance for the de- velopment of the microfinance industry. analyzes the Chapter 20 explores the roles played in the development of commercial nii- crofinance by governments and international organizations, including inter- emerging microfinance national NGOs, foundations, networks, and donors. Emphasis is placed on three kinds of microfinance activities: information dissemination, banking laws and industry and suggests regulation and supervision of institutions providing microfinance, and capaci- ty-building initiatives that concentrate on tools, training of managers and staff, a microfinance model and institutional development.This chapter also focuses on the crucial partnerships being forged between governments and many kinds of organizations.Thus the for 2025 discussion concerns the roles played in microfinance by central banks, a donor consortium, multilateral and bilateral donors, an equity fund, an NGO, a non- profit charitable organization, a private microfinance rating company, a prac- titioner network, a training program, and an Internet list. Part 5, "The Twenty-first Century: Democratizing Capital" (chapters 21-22), analyzes the status of the microfinance revolution at the turn of the century and projects advances in the democratization of capital by 2025. A new model of institutional commercial microfinance is developed in chap- ter 21. Unlike earlier models also analyzed there, the commercial microfinance model assumes an arena in which competing formal sector institutions act as intermediaries, providing commercial loans and savings services to the eco- nomically active poor. In this model profitable microfinance institutions that are publicly regulated and supervised hold a sizable share of the microcredit mar- ket and a large share of the microsavings market. Organizational structures are streamlined for efficiency. Loan sizes are limited but savings, in any amount over a tiny minimum, are collected from the public-providing ample funding for loans and making savings mobilization cost-effective. Depending on the insti- tution, loan portfolios are also financed by commercial debt and investment. The model emphasizes horizontal links between formal and informal sectors in the same locality, as well as vertical links between local financial markets and actors in regional, national, and international arenas. The chapter ends with some thoughts on the microfinance industry in 2025. It projects a rapid advance in the market share of microcredit provided by reg- Preface xiiii ulated formal institutions, along with a substantial declne in the market share of informal moneylenders.This shift implies a much larger number of formal sector borrowers in 2025 relative to 2001, but not necessarily a major decrease in the number of moneylenders or their clients. As commercial miicrofinance develops into a competitive industry with funds to finance loans coming from capital markets, investments, and savings, the formal sector will lend more funds to far more microfinance clients.A substantial increase in the market share of nmcrosavings is also envisaged for formal sector mncrofinance institutions. The book concludes, in chapter 22, with a dliscussion of policy issues that are likely to be crucial for microfinance over the next 25 years.The focus is on the kinds of policy decisions that will probably engage governments, banks, non- "The bank is not bank financial institutions, donors, and others.The policy choices for the var- ious players are explored. a king, the bank There are many routes to large-scale, sustainable microfinance. Banks may enter the market. NGOs may become regulated, for-profit institutions.Village is a servant." banks may become linked with formal sector financial institutions. And some credit unions and cooperatives may decide to focus on niicrofinance. But the -A customer of focus here is on the basics that underlie the microfinance revolution and are common to all large-scale, profitable microfinance mstitutions.A macroeconomic Indonesia 's Bank and policy environment that permuts commercial financing and pricing enables institutional profitability and self-sufficiency. Institutional sustainability allows Dagang Bali financial services to be made widely available to the economically active poor over the long term. Profitablity engenders competition, which increases effi- ciency-improving the services available to low-income clients and lowering the costs they pay for them. What does all this mean for the poor people who become clients of these institutions? This is best explained by the clients themselves.A customer of In- donesia's Bank Dagang Bali for more than 20 years put it this way: I grew up poor and without education. I learned, though, that I could improve myself, and that the bank would help me. The president of Bank Dagang Bali is a great man.Why do I say that? Not because he is a bank president; there are many bank presidents. Because he knew that poor people fear banks, and he taught us not to be afraid. BDB taught us something important that we never knew before. BDB taught us that the bank is not a king, the bank is a servant. xliv The Microfinance Revolution: Lessons from Indonesia Acknowledgments During the years that I have been learning about microfinance, I have become greatly indebted to many people.This book draws on the help, insights, and guid- ance of many people in many parts of the world. It is not possible to mention all of them here, but I want to thank those who have helped in especially im- portant ways. First, I must record my debt to the thousands of men and women in vil- lages and low-income urban neighborhoods of developing countries who have answered my questions and taught me about their enterprises, their finances, and their lives. Most were poor in economic terms but rich in terms of wis- dom and social responsibility. My knowledge of mncrofinance is largely derived from them. Microfinance made an impression on me at an early age. My late father, Philip Van Doren Stern, wrote a story called The Greatest Gift that was made into a movie titled It's a Wonderful Life. The film, which I saw many times while grow- ing up, is about the owner (played by Jimmy Stewart) of a small-town build- ing and loan institution who fights the local establishment to provide financial services to the town's working poor. The movie's message seems to have sunk deep into my subconscious to emerge many years later. I first learned directly about the power of formal sector finance for the so- cial and economic development of low-income people from the late Burra Venkatappiah, who served at different times as deputy governor of the Reserve Bank of India (India's central bank) and chairman of the State Bank of India. As the driving force behind the famous 1954 All India Rural Credit Survey, he played a major role in the first nationwide study of rural credit and its relation xlv to economic and social development. The study found that government and cooperative credit together reached only about 6 percent of rural borrowers- mostly large farmers.As a result major changes were recommended in rural cred- it policy, laying the foundations for the country's long-term interest in providing finance to low-income borrowers. Much later the novelist Aubrey Menon, while being interviewed by the press, was asked what he considered the most important book written in India since independence. He rephed: "The 1954 All India Rural Credit Survey." I am much indebted to BurraVenkatappiah and his family for sharing with me their wisdom and their hospitality for a very long time. Among his many far-reaching contributions to the Indonesian economy, Professor AliWardhana-minister of finance from 1968-83, coordinating min- ister for economics, finance, and industry from 1983-88, and economic advis- er to the government since 1988-was primarily responsible for creating Indonesia's commercial microbanking system, the world's first large-scale sys- tem of sustainable microfinance. Ali Wardhana first played a crucial role in the economic reforms that resulted in extensive rural development in the 1970s and 1980s-and in the consequent emergence of millions of potential bank clients. He led the country's widespread financial reforms that began in 1983; among their results was a policy and regulatory environment in which com- mercial microbanking could be born and sustained. He then arranged the es- tablishment of Bank Rakyat Indonesia's (BRI's) nationwide commercial microbanking system, and he has watched over and guided its development ever since. Thus his introduction to this book is especially relevant. As Stephen Grenville, now deputy director of the Reserve Bank of Australia, said in 1994: "Not only was Pak Ali present at the creation of the Indonesian financial sec- tor as we know it, but he was midwife at its birth and its guardian as it grew up." I am deeply indebted to Pak Ali both for the privilege of having worked for him for so many years and for all that I have learned from him. I am also much indebted to Ibu Nani Gandabrata for her many years of assistance and kindness; through her example she has taught me much. I would like to express my gratitude to Drs. Radius Prawiro, for whom I worked when he was Indonesia's minister of finance (1983-88) and coordi- nating minister for the economy (1988-93), and who supported the develop- ment of microfinance in Indonesia throughout his long career in the cabinet. For many years I have been fortunate to have served as an adviser to BRI and to have had the opportunity to observe closely the development of its mi- crobanking system.To the thousands of BRI managers and staff-from the board of commissioners, president-director, and board of directors to the staff of the local bank units with whom I have interacted since 1979-I am indebted in a special way.They set the example for the development of sustainable microfi- nance on a nationwide scale. I have learned much from their achievements, as well as from the goals, strategies, tactics, and methods that lie behind BRI's suc- cess in microfinance. Kamardy Arief, former president-director of BRI (1983-92), provided long-term, active, committed leadership for the transformation of BRI's approach to banking for the poor. Djokosantoso Moeljono has been president- xlvi The Microfinance Revolution: Lessons from Indonesia director since 1994, and it was during his term that BRI's microbanking sys- tem emerged as a model of technology transfer among developing countries. A special acknowledgment must be made to the late Sugianto, BRI's man- aging director who was responsible for its unit desa (microbanking) system from the inception of its commercial approach to rmicrofinance in 1984 until his sud- den death in 1998. After the financial deregulation of 1983 made it possible, Sugianto (who hke many Indonesians used only one name) managed the tran- sition of the unit desa system from a network of banking units with high ar- rears, high losses, low savings, and low staff morale to the world's largest self-sufficient microbanking system. Sugianto once said to me, "You can suc- ceed in microfinance only if you love it." He did both.The unit desas now pro- vide financial services profitably to millions of poor people throughout the country, continuing even throughout the financial and economic crisis that began in 1997. I am much indebted to Sugianto for all that I learned from him for nearly two decades, and for his careful reading and helpful comments on chap- ters 1-14 of this book. His successor, Rustam Dachlan, who has managed the unit desas successfully at a time of great difficulty in Indonesia, has continued BRI's interest in the effort made in this book to document the development of BRI's microbanking system. From 1979-83 I coordinated with Donald R. Snodgrass the Development Program Implementation Studies (DPIS), an interdisciplinary study of In- donesian development programs conducted by the Harvard Institute for In- ternational Development (HIID) for the Indonesian Ministry of Finance. I had the privilege of coordinating the four-year DPIS study on Indonesia's rice in- tensification program, which resulted in recommendations to change the net- work of rural banks created to channel subsidized credit to rice farmers into a system of commercial financial intermediation. Many of the DPIS recom- mendations were accepted by the government and implemented by BRI. I am much indebted to the many people with whom I worked on that study. I am also grateful to the Center for Policy and Implementation Studies (CPIS) inJakarta, the Indonesian foundation that grew out of the DPIS project, and to the many people with whom I worked there.As coordinator of the CPIS local banking group, which provided assistance to BRI from 1983-90, I learned much from Ismah Afivan, Human Akil, Kwan Hwie Liong, R.J. Moermanto, Ilyas Saad, L. Hudi Sartono, Bambang Soelaksono, Sudarno Sumarto, and others. The CPIS also studied urban informal sector labor in Indonesia, and those studies helped me understand the demand for microfinance among the urban poor. From 1986-92 I served as coordinator of the CPIS informal sector group and worked closely with a number of CPIS research staff. Those whose work has been especially relevant to the issues considered here are Akhmadi, Sri Budiy- ati, Reno Dewina, Leni Dharmawan, Inca Juanita, the late Imanjuwono, Dewi Meiyani, Isono Sadoko, Kamala Chandrakirana (Nana) Soedjatmoko, and Dar- winaWismoyo, in addition to others from the CPIS mentioned above. I would also like to express my appreciation to Reitje Koentjoro for her long, contin- uing, kind assistance. Acknowledgments xlvii The research on which the Indonesian sections of the book are based was supported primarily by Indonesia's Ministry of Finance. The Ministry of Fi- nance and the coordinating minister for economy, finance, and industry pro- vided long-term support and provided me with extensive information on rural development, banking, and government policy. My work with BRI was also supported by BRI, the U.S. Agency for International Development (USAID), and the World Bank.All this generous support is acknowledged with gratitude. Much of what I have learned about successful mucrofinance has been gar- nered from years of discussions with I Gusti Made Oka, president-director of Bank Dagang Bali (BDB), and his wife Sri Adnyam Oka, and through obser- vation of their bank's operations. Putu Indra Suryatmaja provided extensive help explaining and helping me to document BDB's activities and performance. Kadek Edy Setiawan provided help with BDB's financial data. Chapter 10, on BDB's development, is based primarily on research that I carried out in 1994, sup- ported by USAID through its GEMINI Project (administered by Development Alternatives Inc. of Bethesda, Maryland) and by Calmeadow of Toronto. I am much indebted to all who helped me understand BDB's remarkable and pio- neering role in microfinance. From 1996-98 I served as coordinator of HIID's advisory project for BRI's International Visitors Program, funded by an agreement between BRI and USAID. I am indebted to BRI, USAID's Office of Microenterprise Develop- ment, the USAID mission in Jakarta, and those at BRI with whom I worked on this project: Soeseno As,Andi Ascarya,Widjojo Koesoemo, Siti Sundarl Na- sution,Tii Purwaningsih,Andrina Rivai, Iman Sarosa, andJarot Eko Winarno. This work broadened my perspective on the aspects of commercial mi- crobanking developed at BRI that are applicable directly or indirectly-to institutions in other countries. I learned much about microfinance from my work in the mud-1990s with the Reserve Bank of India, especially from R.V. Gupta, then deputy governor, andY S. P.Thorat, then additional chief general manager.This work was fund- ed by USAID through its India mission, its Office of Microenterprise Devel- opment, and the GEMINI Project; this support is gratefully acknowledged. On numerous trips to Bolivia I gained important information about mi- crofinance institutions and their clients from Francisco (Pancho) Otero, Her- mann Krutzfeldt, Maria-Elena Querejazu, and the many others with whom I have worked at BancoSol; from Eduardo Bazoberry and others at PRODEM; from Sergio Prudencio, with whom I worked at both institutions; and from man- agers and staff of other Bolivian microfinance institutions. My Bolivian work was supported partly by BancoSol and partly by USAID through its Bolivian mission and the GEMINI Project. Albert Kimanthi Mutua, C.Aleke-Dondo, and others with whom I worked at the Kenya Rural Enterprise Programme (K-REP) shared with me their ex- tensive experience on nmcrofinance and related institutional and policy issues. My work there was supported by the Ford Foundation.Jaime Aristotle Alhp, Do- xiviii The Microfinance Revolution: Lessons from Indonesia loresTorres, and others with whom I am now working at the CARD Bank (the Philippines) have helped me understand the opportunities, constraints, and in- ternal dynamics of NGO-created banks comnitted to financing microloans with voluntary savings. This work is being supported by Women's World Banking. Work with the Bank for the Poor inVietnam, sponsored by the Consulta- tive Group to Assist the Poorest, World Bank, and United Nations Develop- ment Programme; with the Bangladesh Mimstry of Finance and Bangladesh Bank, supported by the United Nations Educational, Scientific, and Cultural Organization; and with the Bank of Tanzania, supported by the bank and the World Bank, has provided important insights on how governments view the introduction of commercial microfinance.Work with financial institutions and government bodies in the Philippines and South Africa, funded by USAID; in East Africa, funded by the U.K. Department for International Development; and in China, funded by the Consultative Group to Assist the Poorest and the United Nations Development Programme, have broadened my perspective in this area.The support of these organizations is gratefully acknowledged. As a consultant to the U.S. Comptroller of the Currency on its effort to encourage bank use by people in the United States who are not participating in the formal financial sector, I had an opportunity to learn about the role of microfinance in industrial countries. I am grateful to Constance E. Dunham for making this possible, and for encouraging an ongoing comparison of lessons about financing the poor in both environments. Calmeadow, Ohio State University's online Development Finance Network, the Microfinance Training Program (first at the Economncs Institute and now at Naropa University in Boulder, Colorado), the Private Sector Initiatives Cor- poration (known as MicroRate), USAID's Office of Microenterprise Devel- opment, andWomen'sWorld Banking have provided assistance in various ways during the writing of this book, broadening my understanding of microfinance issues. I am especially indebted to Irene Arias-Hofman, Carlos Castello, Gregory C. Chen, Michael Chu, Michael Goldberg,Jennifer Isern, Elisabeth H. Rhyne, Laura 0. Robinson, Richard Rosenberg, Stuart Rutherford, Donald R. Snod- grass, and Jacob Yaron. Each read all or substantial parts of the manuscript and provided comments, corrections, criticisms, and improvements. I am also much indebted to Christopher P. A. Bennett, David E. Bloom, Robert Peck Chris- ten, James J. Fox, Mohini Malhotra,Joyita Mukherjee, and Maria Otero, who contributed to this book in many ways. My understanding of microfinance has been informed by the knowledge of these friends and colleagues, and enriched by their help. I would like to express my appreciation to many others who have contributed to my thinking about the role of microfinance in social and economic devel- opment. They include Dale W Adams, Nancy M. Barry, Lynn Bennett, Laksh- mi Reddy Bloom, John R. Bowen, Ernst A. Brugger, Barbara Calvin, Anita Campion, Catherine Mansell Carstens, Martha Alter Chen, Md. Shafiqual Haque Choudhury, Craig Churchill, Monique Cohen, David C. Cole, Martin Acknowledgments xlix Connell, Hernando de Soto, Deborah Drake, Cheng Enjieng, Todd Farring- ton, S. Malcolm Gillis, Claudio Gonzalez-Vega,Turabul Panjatan Hassan, Brig- it Helms, Richard M. Hook, Don E.JohnstonJr.,James R. Kern, Nathan Keyfitz, Klaas Kuiper, Dharma Kumar,Johan Leestemaker, Patricia Markovitch,Jonathan Morduch,Julia Paxton, Sayeeda Rahman,Ashok Rai, N.V Raja Reddy, the late Michael Roemer, Leo Schmit, Chiranjib Sen, Parker M. Shipton, Michael Simpson, Stephen Smith,Joseph J. Stern, the late Ann Dunham Sutoro, C. Peter Timmer, Norman T. Uphoff, Marzuki Usman, R. C. G.Varley, Robert C.Vogel, J. D.Von Pischke, and Damian von Stauffenberg. ACCION International, the MicroBanking Bulletin, and the MicroFinance Network provided information on countless occasions and steered me in many right directions. Scores of people contributed information and insights that were used in writing chapter 3 ("Voices of the Clients") and chapters 17-20 (which analyze the development of microfinance in institutions around the world).The assistance of all is most gratefully acknowledged. Carol Grotrian, Mary Ruggiero, and Flora Segundo provided extensive assistance in the preparation of the manuscript, and Jonathan Ramljak pro- vided exceptional help in overseeing its production.Jessica Roberts helped in multiple ways and prepared the bibliography and the glossary with much care and uncommon dedication. I am much indebted to all. Peter J. Fidler provided superb research assistance on all aspects of this book and co-wrote part 4. He has been invaluable in the effort to under- stand, untangle, clarify, document, and analyze the many issues discussed here, and I am much indebted to him. This book and its author owe an enormous debt to Ira W Lieberman. He supported the book from the start and played an extraordinary and continu- ing role in arranging for its financing, edcting, and publication-even as the book grew from one to three volumes.The help, encouragement, and support I have received from him is exceptional, and it is acknowledged with much gratitude. The book has received funding and support from the Consultative Group to Assist the Poorest, Ford Foundation, HIID, Open Society Institute, and World Bank. All this support is acknowledged with much appreciation. I am particularly grateful to have had Communications Development In- corporated as my editors. Paul Holtz and Bruce Ross-Larson provided superb editing at every stage, and I have learned a lot from both of them. I am also in- debted to Wendy Guyette, who laid out the chapters, and to Stephanie Ros- tron, who coordinated production, both with Communications Development. Deborah E. Patton compiled the index. I would also like to thank the publishers of the book, the World Bank and Open Society Institute. Paola Scalabrin of the World Bank's Office of the Pub- lisher served as project manager for the publication of the three volumes and played a crucial role throughout. I am indebted to her for help on a wide va- riety of matters. The three volumes of this book, especially the second one, are based primarily on work carried out for many years at HIID, and they draw on a number of my The Microfinance Revolution: Lessons from Indonesia earlier papers written at HIID. I am indebted to HIID for its support of my work generally, and for its support for the writing of these volumes. Members of my family have contributed to this book in a variety of ways, both direct and indirect, and I am most grateful for their help:Allan R. Robin- son, Sarah P. Robinson, Perrine Robinson-Geller, Laura 0. Robinson, Salva- tore A. d'Agostino, Eric B. Geller, and Peter Rosendorff. Once when I was at the Jakarta airport, about to leave for Boston, the im- migration officer challenged my embarkation card."It says here you are an an- thropologist," he said. "It also says you work for the Ministry of Finance.Which is it? It cannot be both!" It took some time to explain that both were correct. As an anthropologist who has been given the opportunity to serve as a policy adviser to finance ministries, central banks, and a variety of banks and other fi- nancial institutions in many parts of the world, I have been especially fortu- nate, and am most grateful. Acknowledgments Part -s Commercial Microfinance in Indonesia Indonesia was the first developing country to establish large-scale commercial microfinance systems. More than a hundred years ago the Badan Kredit Desa-village-owned banks now supervised by Bank Rakyat Indonesia (BRI)- began offering microloans commercially, and they contin- ue to do so today. Bank Dagang Bali, which opened in 1970, Indonesia was the is a private bank believed to be the oldest licensed bank in first developing a developing country specializing in commercial microfi- country to establish nance. Founded by a couple who had long experience in the informal sector and in informal moneylending, the large-scale bank provides savings and credit services and has been prof- commercial itable since the first year it opened. microfinance In 1983 the Indonesian government issued a major fi- systems. These nancial reform deregulating interest rates on most bank began in the late loans and savings accounts. At that time the Ministry of 19th century Finance decided that BRI, a state-owned commercial bank with a mandate to provide financial services in rural areas, would transform its 3,600 unit desas-bank outlets at the subdistrict level that served as channeling agents for the government's subsidized rural credit programs-into commercial microfinance intermediaries. And in 1984 BRI began transforming its loss-making unit desa system into what quickly became a profitable nationwide com- mercial microbanking system-now the largest in the world. 2 The Microfinance Revolution: Lessons from Indonesia These and other commercial microfinance institutions continue to operate profitably in Indonesia, reaching mil- lions of clients with small loans and savings services. And with decades of experience, such institutions offer many lessons that can be adapted by microfinance organizations in other parts of the world. The Indonesian institutions discussed here are important contributors to the microfi- In Indonesia it has nance revolution for another reason as well: in recent years been demonstrated they have demonstrated unmistakably that commercial li- unmistakably that crofinance institutions can be remarkably stable even in commercial times of extreme national crisis. The East Asian crisis hit Indonesia in 1997.The value of the rupiah plummeted, inflation mounted, and uneniploy- institutions can be ment, food shortages, and high prices led to social unrest remarkably stable in and violent outbreaks. By 1998 the foreign debt of com- times of extreme mercial banks reached about three times their equity, and national crisis loan losses increased dramatically because borrowers were unable (or in some cases unwilling) to make payments. Domestic loan losses increased as well, as corporations went bankrupt and the banking system melted down. President Soeharto was forced to resign after 32 years in office, and the country began a long and difficult effort to restore the economy, build a viable, transparent financial system, and un- dertake major political reforms. The process continues today-with mixed results and many obstacles remaining. Part 3 Commercial Microfinance in Indonesia-Overview 3 Yet an outsider with no knowledge of Indonesia who ex- amined the performance of BRI's unit desas during 1984-2001 would have no way of knowing that, beginning in 1997, the country had suddenly faced severe political in- stability, extensive economic hardship, and financial chaos.With giant banks and corporations collapsing all around them, the A century-old unit desas continued to increase their outreach, collect loans, tension-between mobilize savings, and remain profitable. Other commercial mi- rural credit subsidies crobanking institutions withstood the crisis as well. Part 3 (which contains all the chapters of volume 2) fo- and commercial cuses on how and why Indonesia developed large-scale microfinance--is sustainable banking systems, and is organized as follows. explored from its Chapter 8 provides an introduction to Indonesia's history, 19th century roots to economy, government, and politics. It then analyzes the its 21st century reasons for Indonesia's recent crisis, the crisis itself, and the issues attempts (through mid-2001) to restore broadly based eco- nomic growth and build a functioning democracy. Chapter 9 reviews the history of rural finance in Indonesia (and its antecedents in Europe) and examines the country's rural development and rural financial institutions.A centu- ry-old tension in Indonesian microfinance-between those who promote rural credit subsidies and those who foster in- dependent commercial financial institutions providing vol- untary savings and credit-is explored from its 19th century roots to its 21st century issues. In Indonesia large outreach 4 The Microfinance Revolution: Lessons from Indonesia and institutional profitability have been associated through- out only with a commercial approach to financial inter- mediation. Chapters 8 and 9 provide the background needed to understand the reasons that large-scale commercial mi- crofinance developed in Indonesia.The rest of volume 2 con- siders how and why this happened. Bank Dagang Bali (BDB), discussed in chapter 10, is the Bank Dagang Bali, oldest licensed general bank specializing in commercial a private bank, is microfinance that currently operates in a developing the oldest licensed country and without ever having received a subsidy. full-service bank Chapter 10 explores the history of BDB and its performance from 1970-96 (its 1997-2000 performance is discussed in specializing in chapter 15). The bank is best known for its knowledge of commercial microfinance clients and for its savings services, which are microfinance in a state of the art in microfinance. BDB is now a full-service developing country bank providing finance to large customers and corporate clients (many of whom were once the bank's poor clients) in addition to its microfinance customers. The development and growth of BRI's microbanking (unit desa) system are explored in chapters 1 1-15.The unit desa system was selected for detailed analysis here for three reasons. It is the world's largest financially self-sufficient microbanking system. It is the Indonesian microfinance in- stitution I know best, having advised on its development for more than 15 years starting in 1979.And it is much less well- Part 3 Commercial Microfinance in Indonesia-Overview 5 known internationally than some microfinance institutions in other countries that promote the poverty lending approach to microfinance. The BRI story should be understood by every government, donor, and financial institution making decisions about microfinance.The unit desas cannot-and should not-be cloned, but they offer many important Bank Rakyat lessons for the microfinance industry. Indonesia's unit Chapter 11 examines the unit desa system from its creation in 1970 to 1983, when the government decided to transform desa system IS the the units from channeling agents for subsidized rural credit world's largest programs to commercial financial intermediaries.The 1970-83 financially se!f- period was characterized by what BRI's President-Director sufficient Kamardy Arief (1983-92) called the unit desas' "old cul- microbanking ture" in sharp contrast to the "new culture" that resulted from system. This volume the fundamental institutional transformation initiated in 1984 shows how it got (see box 1 at the end of this section) .This "new culture" is de- there picted in chapters 12-15. Today's unit desas can be fully un- derstood only in the context of this transformation. The unit desa system was originally created to admin- ister the government's subsidized BIMAS credit program for rice cultivation; later many other smaller subsidized credit programs were added. BIMAS generally resulted in nega- tive outcomes for rice farmers, BRI, and the government. Chapter 11 analyzes the many reasons this program did not- and could not-work as planned.The lessons of this era are 6 The Microfinance Revolution: Lessons from Indonesia important because the transformation of the unit desa sys- tem occurred only after basic lessons about agricultural fi- nance had been learned the hard way. The reasons for Indonesia's failure with BIMAS may help other countries and financial institutions currently implementing similar pro- grams to understand why subsidized agricultural programs aimed at increasing production and raising the incomes of With giant banks poor farmers cannot achieve those goals.The failures of such and corporations programs are caused by intrinsic structural defects. Micro- collapsing all around finance outreach efforts must be based on fundamentally dif- them, BRl's unit ferent assumptions. Using the Indonesian example, this desas increased chapter shows why. Chapters 12-14 discuss, respectively, lending, savings, outreach, collected and organizational reforms in the unit desas from 1984-96. loans, mobilized Chapter 15 documents and analyzes the remarkable stabil- savings, and ity of the unit desas and other commercial microfinance or- remained profitable ganizations between 1997 and mid-2001, during the Indonesian crisis. Chapter 12 focuses on the financial sector reforms, pol- icy decisions, and other enabling conditions that made pos- sible the 1984 shift of BRI's unit desa system to a commercial microbanking intermediary. It analyzes BRI's achievements (and mistakes) in developing and implementing the unit desa credit program. And it examines the unit desas' credit per- formance and profitability from 1984-96. Part 3 Commercial Microfinance in Indonesia-Overview 7 Kredit Umum Pedesaan (KUPEDES), the unit desas' gen- eral purpose credit program introduced in 1984, provides loans to economically active poor and lower-middle-income individuals throughout Indonesia.The program is guided by the principle that institutional sustainability is required for large-scale outreach to low-income borrowers. KUPEDES Thefailures of provides loans to all creditworthy applicants for all productive subsidized rural purposes. Its interest rate is set to cover all unit desa costs and risks, and to return a profit to the system. credit programs are Loan decisions are based on evaluations by well-trained, caused by intrinsic experienced unit desa staff of borrowers' character (will- structural defects. ingness to repay) and of the viability and cash flows of their Using the enterprises (ability to repay). KUPEDES was designed Indonesian example specifically to meet the needs of low-income borrowers for (1970-83), chapter convenient bank locations, simple loan procedures, and 11 shows why flexible terms.Thus within the overall KUPEDES regula- tions, the loan purposes, maturities, and payment plans are customized for each borrower's needs. Repayment rates have been high for a number of reasons. Borrowers are selected based on their ability to use loans pro- ductively and on their ability and wilingness to repay them. Loans are provided in gradually increasing amounts, based on the borrower's repayment record and the creditworthiness of the enterprise for which the loan was taken. Borrowers are motivated to repay in order to retain their option to rebor- 8 The Microfinance Revolution: Lessons from Indonesia row on what are considered attractive ternis at reasonable cost. And unit staff treat clients with respect and courtesy. The unit desa system made a profit in 1986 and has been independent of subsidy since 1987. In 1996 the KUPEDES credit program had $1.7 billion in outstanding loans to 2.5 million borrowers and a long-term loss ratio of 2.2 percent. Ninety-five percent of the units were profitable, and the pre- Chapter 12focuses tax return on assets was 5.7 percent-substantially higher on thefinancial than typical banking industry averages in Indonesia and else- reforms and policies where.The crucial issue about profitability, however, has been that made possible BRI's use of unit desa profits to cover losses in the bank's the 1984 less successful divisions that serve wealthier borrowers with generally low repayment rates-a decision outside the con- transformation of the trol of the unit desa system. unit desas to a Chapter 13 documents the dramatic rise in savings mo- commercial bilization in BRI's unit desas from 1984-96. The units had microbanking only $18 million in savings at the time of the 1983 bank dereg- intermediary ulation. Low deposits were widely but wrongly attributed to an assumed lack of demand for financial savings instruments: it was believed that rural people were unable to save, unwilling to save in financial form, and did not trust banks. In fact, the problem was with the banks-which did not understand the nature or the extent of the demand for rural savings services- and with the government, whose regulations made it im- possible for banks to meet this demand profitably. Part 3 Commercial Microfinance in Indonesia-Overview 9 The unit desas' new savings program was developed in three phases. In 1984-85 extensive research was conduct- ed on the potential demand for unit desa savings and on products and services that would meet this demand; dur- ing this phase the new savings program was designed and tested in a two-stage pilot project.Then in 1986 the pro- The KUPEDES gram was expanded to all unit desas throughout Indone- loan product was sia. Finally, in 1987-89 methods were developed for market designed to meet the penetration. By 1989 all KUPEDES loans were funded by unit savings, as they have been ever since.The units subse- needs of low-income quently achieved increasing penetration of the Indonesian borrowers for market for rural savings and, with the opening of urban units convenient bank in 1989, urban savings in low-income neighborhoods. locations, simple Three main savings products offering different ratios of procedures, and liquidity and returns are available at all units, with some dif- flexible terms ferences between rural and urban units. Savers are permit- ted to have as many accounts as they want, enabling them to customize their use of the accounts to meet specific needs. In 1996 the unit desa system held locally mobilized savings of $3 billion in 16.1 million savings accounts. The principles of sustainable institutional microfinance carry with them a set of basic operational requirements. Chapter 14 shows that effective operations depend on an efficient organization and a microfinance-specific manage- ment structure. Such an organization does not arise in the 10 The Microfinanco Revolution: Lessons from Indonesia absence of the underpinning principles.Yet the principles can be put into operation on a large-scale, long-term basis only through appropriate institutional organization. How can the sequencing work in such a situation? This is a critical problem in microfinance, and one that is little understood.When BRI's unit desa system changed in 1984 from a loss-making channeling agent for govern- Chapter 13 ment credit subsidies into a commercial financial interme- documents the rise diary, it did not have an organizational structure that could in unit desa savings provide the prerequisites for institutional sustainability. mobilization-from Although microfinance can operate viably under a num- $18 million in ber of organizational structures, certain features of opera- 1983 (after more tions and management must be incorporated into the design of any financially self-sufficient microfinance in- than 10 years of termediary. For example, multiple components must be in collecting savings) to place to achieve the repayment rates of more than 95 per- $3 billion in 1996 cent needed for institutional sustainability. Among these components are effective, committed management; an ef- ficient, accountable, labor-intensive organization; simple, transparent accounting and reporting systems; appropriate staff training and performance-based incentives; decen- tralized authority for loan decisions; regular, meticulous in- ternal supervision; simple, suitable management information systems; and user-friendly products and services priced for institutional viability. Whatever its organizational struc- Part 3 Commercial Microfinance in Indonesia-Overview 11 ture, a sustainable microfinance institution must provide these elements. Chapter 14 focuses on the dynamics of organization- al development at BRI, specifically on the ways that the basic products of commercial microfinance-savings and credit-and their operating requirements drove the The ways in which changes in unit desa organization and management.The new savings and chapter also explores the organizational difficulties faced credit products drove by the unit desas in their role as the microbanking divi- sion of a full-service commercial bank. In addition, the major changes in chapter analyzes the performance of two large microfi- unit desa nance organizations administered (in different ways) by organization and BRI: the commercial Badan Kredit Desas (BKDs) and a management are heavily subsidized microcredit program known as P4K. explored in As would be expected, the two organizations have strik- chapter 14 ing differences in performance. Chapter 15 first examines the financial aspects of In- donesia's crisis, the collapse of the banking sector, and the government's subsequent bank restructuring efforts. It then turns to the stunning contrast that emerged during 1997-2001 between the banking sector generally and com- mercial microfinance, represented here by the BKDs, BDB, and the unit desas. During the crisis all three of these in- stitutions continued their wide outreach and profitability. The chapter also provides a more extensive analysis of unit 12 The Microfinance Revolution: Lessons from Indonesia desa performance from 1997-2000, followed by a mid- 2001 snapshot of their strengths and challenges. This final chapter of volume 2 addresses the reasons that commercial microbanking has remained stable during a fi- nancial crisis so severe that, as The Economist (18 July 1998) commented, Indonesia "would probably walk away with the prize for Asia's most desperate banking system."While there The stunning are differences among the three microfinance institutions contrast that discussed, the main reasons for their stability under crisis are emerged during the the same.These include: crisis-between the * Macroeconomic stabilization measures undertaken by the banking sector government with the support of the International Mon- etary Fund and other donors. generally and * The fact that most microbanking clients operate in the commercial domestic economy and were not directly affected by the microfinance-is the currency crisis. focus of chapter 15 * The experienced management, strong performance, and high liquidity of these institutions before the crisis. * Their well trained, motivated, and friendly staff. * The high value placed by borrowers on the option to re- borrow during the crisis. * The impetus for clients to save more and consume less in times of rising inflation and growing unemployment. * Savings moved from failing banks to trusted microfinance institutions. Part 3 Commercial Microfinance in Indonesia-Overview 13 * The decisions of these institutions to continue to make financial services available throughout the crisis. Indonesia offers four important lessons. Microfinance in- stitutions can be fully sustainable with large, nationwide out- reach over long periods. Commercial microfinance, following basic principles (see chapters 12-14), can be carried out by There are differences different types of financially self-sufficient institutions. Mi- among commercial crofinance can be profitable and stable even during severe microfinance crisis. And as a result, millions of low- and lower-middle- income savers and borrowers can improve their enterpris- institutions. But the es, increase their incomes, and gain self-confidence. Some main reasons for of these lessons (and others) are of course available from other their stability during countries. But these four are probably found in combina- crisis are the same tion only in Indonesia. 14 The Microfinance Revolution: Lessons from Indonesia Box 1 Old and new cultures at Bank Rakyat Indonesia's unit desas Old culture, 1970-83 New culture, 1984-present Overview Unit desas acted as channeling agents for the In June 1983 the government deregulated interest rates government's rural subsidized credit programs. on most loans and savings and encouraged restructuring The government set terms and ceilings for of the unit desa system. The government provided initial loans and interest rates for savings and loans funding for the KUPEDES loan program All decisions about unit desa products, services, pricing, and the like are made by BRI without government interference But the government has remained a strong supporter of the unit desa system and, when formulating financial policy, has consistently kept in mind the units' special contributions and requirements BRI managers and staff had little understanding of Since 1984 the units have provided commercial financial local markets or of the potential for microfinance. services for the economically active poor, the units' BRI saw the units as a loss-making activity that impedes profitability makes the system viable for the long term. "our real banking Hallmarks of the new culture include simplicity, transparency, professionalism, accountability, service, and knowledge of local markets Human resources Salaries of unit staff were much lower than those of Unit staff are on the same salary scale as the rest of the BRI branch staff. bank Unit staff were considered outside the BRI "family," were Unit staff are part of the BRI organization, included in corporate not treated as part of the corporate culture, events, and encouraged to act as, and regard themselves and were not viewed as bankers as, bankers There was no career track at BRI for unit staff Unit staff are on a BRI career track, based on performance, they can be promoted to BRI branches In addition, low-level branch staff can be promoted to serve at the units There was no specialized microfinance training for unit Extensive microfinance training is provided for unit staff and staff and their branch-level supervisors their supervisors, who are based at the bank's branches There were no performance-based incentives Unit staff receive significant cash incentives and formal recognition for good unit performance Unit managers had very limited authority Unit managers have considerable authority in loan approval and unit management Unit managers and staff had limited accountability There is strict accountability; unit managers and staff are held responsible for their financial decisions and for their unit's performance Part 3 Commercial Microfinance in Indonesia-Overview 15 Box 1 (continued) Organization and management Units were branch windows, and their transactions were Units are treated as individual profit centers, with each unit's posted as branch transactions. performance recorded and reported separately Unit performance reports were aggregated and incorporated A balance sheet and profit and loss statement are required into the balance sheet and profit and loss statement of each unit every month The unit desa system has its own of the supervising branch reporting system at the unit, branch, regional, and head office levels There was no unit desa division at the head office The unit desa division at the head office reports to a BRI and no unit desa section at regional offices managing director; regional offices have unit desa sections Supervision of the units was poor and irregular. Regular, intensive supervision of the units is conducted by supervisors from the branch and regional levels Thirty-two monthly reports were required from each Five monthly reports are required from each unit, providing unit, consuming a large amount of staff time but resulting more relevant information than was contained in the in little useful information 32 reports under the old system Products and pricing About 350 subsidized loan products were offered at One loan product and three basic savings/deposit products below-market interest rates One savings product was with generally positive rates of return are offered, with an provided, with an interest rate that was set below interest rate spread that enables unit profitability inflation and below the lending interest rate Interest payments on loans were transferred to KUPEDES interest is posted under unit assets and unit the supervisory branch Unit savings were also savings are posted under unit liabilities Units can save transferred to the branch, but interest to depositors was at or borrow from their supervising branch at the transfer paid by the unit price set by the head office Borrowers were selected primarily by government officials Borrowers are selected by well-trained unit staff, unit and committees. managers have the authority to approve most loans Profitability and outreach Arrears and defaults were high, as were government The units have been profitable every year since 1986 subsidies and unit desa losses. In 1983 there were and free of subsidy (except for technical assistance) since fewer than 500,000 outstanding loans, including many 1987. In 2000 the units had 2.7 million outstanding loans in default from previous years The number of savings in a portfolio of $816 million and $2 billion in savings accounts is not available, but at the time of the June in 26 million savings accounts 1983 financial deregulation the amount of unit savings was $18 million 16 The Microfinance Revolution: Lessons from Indonesia An Introduction to Indonesia The development of large-scale sustainable microfinance in Indonesia did not occur gradually, accidentally, or as a grassroots initiative; it was planned and implemented as an integral part of the country's economic develop- ment. Because Indonesia's achievements in microfinance can be understood only in a wider context, this chapter summarizes its history and economic development as background to the rest of volume 2. Only a brief account of Indonesia's rich and complex history can be given here, but references are provided throughout the chap- ter for readers who want to learn more about the coun- try's history, economy, politics, and culture. In preparation for the discussion of microbanking, the chapter that fol- lows completes the general background summary by 18 reviewing the history of Indonesia's rural development and rural fi- nancial institutions. This chapter begins with a discussion of early Indonesian histo- ry, the struggle for independence, and the post-independence era under Sukarno, who served as Indonesia's first president from 1949-67. By the mid-1960s, however, the Indonesian economy was in severe decline and Sukarno's political base had eroded. An attempted coup occurred on 30 September 1965; within hours Major-General Soeharto, head of the army's Strategic Reserve Command, took control of the army.The attempted coup-for which From 1970-96 responsibility is still in dispute-was followed by an army massacre of hundreds of thousands of communists and others whom the army Indonesia' economic blamed for the coup attempt. Soeharto emerged as the new national leader: executive power was transferred to him in March 1966, and growth averaged 7 he was appointed acting president in 1967 and president in 1968. percent a year, and The primary achievement of Soeharto's 32-year presidency was Indonesia's social and economic development under what was Soeharto' known as the New Order government. An extraordinary team of technocrats restored the economy and, as economics ministers, led government made it through three decades of remarkable growth with a strong em- phasis on equity. From 1970 to 1996 Indonesia's economic growth ry averaged 7 percent a year. During 1986-96 the country saw aver- achievements in age annual growth of nearly 8 percent, while average annual infla- tion stayed below 10 percent. poverty reduction Soeharto is widely credited with placing a high priority on poverty reduction, and his technocrats were able to help the poor increase their incomes and improve the quality of their lives.As a re- sult Soeharto won prestigious international prizes for Indonesia's achievements in food and agriculture, in family planning, and in pover- ty alleviation. But under the authoritarian New Order government the country remained politically underdeveloped, with Soeharto personally dominating all branches of the government, controlling the political opposition and the media, and ensuring that no healthy opposition-and no potentially powerful leaders-could emerge. By the 1990s the Soeharto family and a small group of their friends had become corrupt on a massive-and escalating-scale. In the process Soeharto had virtually stopped listening to his team of eco- nomics ministers. Many Indonesians became increasingly outspo- An Introduction to Indonesia 19 ken against Soeharto's authoritarian rule and the family's corrup- tion. Then in July 1997 a financial crisis erupted in Southeast Asia. In Indonesia the value of the rupiah plummeted, inflation mount- ed, corporations went bankrupt, banks closed, the financial system collapsed, and unemployment, food shortages, and high prices led to social unrest and violent outbreaks.The primary cause of the cri- sis was not the underlying economy, however, but the deep and grow- ing weaknesses in governance and in the financial system. Soeharto was forced to resign in May 1998; he was succeeded byVice Pres- Because of ident B.J. Habibie, his lifelong protege. In 1998 the International Monetary Fund (IMF) and other in- Soehartos ternational donors provided an emergency $42.3 billion soft loan authoritarian rule, financial aid package for Indonesia.The economy, which was at its lowest point of the crisis in 1998, improved gradually. Stabilization the country measures brought down inflation from 53.4 percent in 1998 to 20.5 percent in 1999 and 3.7 percent in 2000. remained politically Responding to mass demand after Soeharto's resignation, Presi- dent Habibie initiated a number of political reforms, including the abolition of restrictions on formmng political parties and the referen- wvith corruption dum that allowed the province ofEastTimor to vote overwhelmingly for independence from Indonesia. (After the vote, however, massive flourishing at the burning, looting, and killing of East Timorese were carried out by In- donesian militias with support from some of the military.) Immediate challenges were to improve governance, implement political reforms, strengthen the economy, rebuild the financial sys- tem, and maintain a safety net for the poor, whose numbers had in- creased during the crisis. But results were mixed. Inflation decreased substantially and economic growth improved. Some basic reforms were accomplished (such as freeing the media, a process begun by Habibie and completed by his successor, President Abdurrahman Wahid). But many of the country's old problems did not disappear, and new ones emerged. Indonesian general elections were held in June 1999-the first time in more than 40 years that the results of an election were not known beforehand. No party obtained a majority.The presidential election was held in October 1999, with Habibie and Megawati Soekarnoputri (Sukarno's daughter) as the front runners. Howev- er, the newly reformed People's Consultative Assembly elected Ab- 20 The Microfinance Revolution: Lessons from Indonesia durrahmanWahid as president and Megawati Soekarnoputri as vice president. Wahid was selected largely because he was a widely re- spected Muslim religious leader, known for his honesty, tolerance, and support for human rights. ButWahid is blind and had been weak- ened by strokes, and there was concern about whether he would be able to run the government. The challenges were formidable. CouldWahid build a coalition government from multiple newly formed parties with different views and priorities? Could the president control the army? Could the government resolve secessionist movements, restore the econ- Since independence omy and rebuild the financial system, and build the institutions needed for a viable democracy? Indonesia has had a The Soeharto family and cronies and their conglomerates, though weakened, remain wealthy and powerful. And many Soeharto sup- strong leader and a porters throughout the country remain active, in some cases work- weak state. JVen ing to create and maintain political instabilities. PresidentWahid failed in his effort to bring Soeharto to trial for corruption (the court dis- Soeharto resigned, missed the case, finding Soeharto too ill to stand trial). Wahid made some crucial reforms early in his administration: mak- the strong leader was ing the media fully independent, placing the military under civil- ian control, and holding some military officers accountable for g human rights abuses. And the economy showed improvement. In- state remained flation remained single digit and GDP growth, which was -13.0 per- cent in 1998 and 0.3 percent in 1999, reached 5.2 percent in 2000.' ButWahid was unable to consolidate power. He had neither the political backing nor the adrministrative skills needed to achieve sus- tained reforms. Political infighting, violence, bombings, secession- ist movements, ethnic and religious conflicts, and instabilities promoted by Soeharto supporters, Muslim fundamentalists, and others escalated. By early 2001 the rupiah was falling, the country's slow economic growth was faltering, and inflation was rising. Still constraining the economy were difficulties with the restructuring of massive corporate foreign and domestic debt, the crippled bank- ing system, and continuing large-scale corruption. Efforts to re- structure the corporate debt and to build an effective and accountable banking system, both needed for any significant growth in private investment, remained mired in political problems. Many of these prob- lems involve the Soeharto family and their friends. An Introduction to Indonesia 21 Meanwhile, the IMF and Indonesia's other donors had become dissatisfied with the glacial pace of agreed reforms and with grow- ing political instabilities. In December 2000 the IMF suspended dis- bursement of a $400 million tranche of its aid program, which remained suspended throughout the rest ofWahid's presidency. By April 2001,18 months after his election,Wahid had twice been censured by parliament. He was criticized for his government's lack of focus, coordination, and implementation; for the deterioration of the economy; for insufficient progress in legal, judicial, legislative, Indonesia is now banking, and other reforms; and for failing to curb corruption, re- solve secession issues, and bring former president Soeharto to jus- attempting to recover tice. Three months later, on 23 July 2001, President Wahid was impeached by the unanimous vote of the 591 members of the 700- from a major member People's Consultative Assembly who attended the session. economic,financial, He was replaced for the remainder of his five-year presidential term by Vice President Megawati Soekarnoputri. and political crisis The transition from the Soeharto era could not have been easy under any circumstances, but the economic, financial, and political and to build a crisis that began in 1997 made it especially difficult. And there will modern political undoubtedly be difficult times ahead. But Indonesia has many strengths. It has abundant resources: a large and educated labor state force, ample natural resources, a strategic location, an extensive in- frastructure, a large domestic market, and long experience with sustained economic growth and equity. As AliWardhana (1998b, p. 5) put it, "No one should doubt the severity of the current economic crisis.Yet neither should anyone underestimate our capacity to set our institutions right." 22 The Microfinance Revolution: Lessons from Indonesia Indonesia is perhaps the least understood of the world's major nations. Only an overview of selected aspects of this large, complex, and multiethnic coun- try is provided here. But even for a summary of Indonesia's social, economic, and political history, three underlying concepts are crucial: the role of the Ja- vanese wayang puppet theater and its teachings, the Javanese concept of power, and Pancasila-the five guiding principles of the Indonesian government. These concepts are considered here in the contexts where they are most di- rectly relevant to the discussion. In different ways, all are fundamental for an understanding of the events and processes explored here: the freedom struggle that culminated in a declaration of independence in 1945 and the formation of an independent nation in 1949, the country's economic development over the next half-century, its political dynamics, the economic and political crisis TheJavanese of the late 1990s, and Indonesia's current attempts to recover from that crisis and to build a modern political state. wayang shadow Wayang, orJavanese shadow theater, teaches central lessons about the con- duct of lhfe and about making choices-messages embodied in stories that peo- theater teaches ple learn from early childhood. Much ofJavanese philosophy is based in Hindu thought-not necessarily compatible with Western ideas-in which the cre- central lessons about ative and the destructive are the same principle, goodness cannot exist with- out evil, and opposition and complementarity are aspects of the same whole. conducting lfe and While most Javanese are Muslim,2 a strong underlay of earlier Hindu culture remains on Java, where most of Indonesia's population lives (and on Bali, making choices. where most of the population is Hindu). Three of the many lessons from the wayang are discussed below. One is that Current events are life is complex, duties conflict, and even ethical, honorable choices may lead to undesirable outcomes. Another is that whlle humans may accumulate ma- often interpreted terial goods, greed-especially among rulers-is unacceptable to the gods.The third is that a sovereign is legitimate only so long as he maintains the enlight- based on wayang ened characteristics required of a ruler and so long as he governs in the inter- est of his people. teachings Closely related is the Javanese concept of power-which is seen as an in- dependent, fixed entity that exists apart from its users and passes from one hold- er to another (Anderson 1990 [1972]).Thus the only way to gain power is to take it from someone else. Commoners can become kings if they acquire power. But to keep power, a ruler must demonstrate continuously that he still holds it. "The ruler must behave properly or his Power will ebb and vanish, and with it the good ordering and smoothness of the social system" (Ander- son 1990 [1972], p. 63). Pancasila, or the five guiding principles of the government-belief in one supreme being, Indonesian unity, humanitarianism, democracy by representative consensus, and social justice-was first proclaimed by Sukarno in a speech on 1 June 1945, two months before Indonesia declared its independence.As a national ideology, Pancasila is so all-encompassing that it provides an inclusiveness that can bind together many highly diverse groups. It also enables a powerful president to justify virtually any decision, and it can serve as an instrument of repression. An Introduction to Indonesia 23 Readers should keep these views in mind as the discussion below describes the rise and fall of both Sukarno and Soeharto, who between them ruled In- donesia for 50 years, as well as events of the post-Soeharto era and the rise of Sukarno's daughter, Megawati Soekarnoputri, who became Indonesia's presi- dent in 2001.3 The Javanese Wayang Shadow Theater Wayang, a traditional Indonesian puppet theater of multiple dimensions, in- corporates entertainment, art, spiritual teaching, and information and commentary "The ruler must on current events-all intricately interwoven.4 Extremely popular, especially on Java and Bali, the wayang both reflects and provides inputs into the wider light the minds and society. souls of his The trunks of banana trees that the puppets are inserted in subjects ... and are symbols of the world of man Man's spirit is the puppeteer, provide direction, symbol of truth and meaning. The screen's the unseen world above, responsibility, man's characters the wayangs, and the lamp's rays the Almighty example, and The audience is the all-perceiving wise man. -PakJaya, cited in Mulyono 1981, pp. 18-19 impartiality " Many wayang performances depict episodes from the Ramayana and the Mahabharata, ancient epics that came to Indonesia from India and were grad- ually adapted to local culture and society. The Ramayana, composed original- ly byValmiki in Sanskrit, tells the story of Rama, the prince of Ayodhya. One famous episode concerns the kidnapping of Sita, the wife of Rama-who is an avatar (incarnation) of Vishnu, part of the Hindu holy trinity of Brahma the Creator,Vishnu the Sustainer, and Siva the Destroyer. Sita is kidnapped by Ra- vana, the King of Alengka (Sri Lanka), and after many episodes is rescued by Rama's forces. One Javanese wayang performance5 highlights the role of Wibisana,6 the younger brother of King Rahwana.' Wibisana entreats Rahwana not to hold Sinta8 captive and not to make war on Rama,"who is a powerfuil, honest, noble- minded man of supernatural virtue" (Mulyono 1981, p. 72). Rahwana refuses to listen and banishes his younger brother.Wibisana then joins Rama's followers. The wayang performance centers on Wlbisana's choice. Serving one's country may mean betraying the truth, while serving the truth may mean betraying one's country. In the wayang the puppetmaster recites the following asWibisana's left foot touches the soil of Mangliawan, Rama's king- dom, while his right foot is still in contact with the soil of his native land: 24 The Microfinance Revolution: Lessons from Indonesia Stopping, he stands erect like a statue dressed with clothing. His heart is filled with confusion. The world grows dark in an instant.The flying birds fall to the ground, their feathers ruffled, their wings thrashing helplessly. The fish in the water float to the top and beat helplessly against the shore. In the words of the saying, not a leaf is stirring, for the wind itself is standing still. Wibisana is caught on the horns of a dilemma. "Eat the fruit of the simalaka- ma and your mother dies. Don't eat it, and your father dies.Walk forward and be wounded. Retreat and be torn to pieces. ..This is life." (Mulyono 1981, p. 74) Another episode from the Ramayana shown in wayang performances fo- Indonesia, with its cuses on the balance between the material and the spiritual, and on the prop- er way for a ruler to live.After many years as ruler, Rama set his life and kingdom strategic location in order. astride major Sri Rama did not leave worldly things behind until his children had replaced him on the throne. In the philosophy of the sealanes, has a rich wayang, man is advised to deal fully with the material side of life-only he is cautioned not to be greedy, and not to worship resource base possessions and rank, for worship belongs only to God. -Mulyono 1981, p. 149 including oil, natural In a third episode depicted in wayang performances, Rama teaches the prin- gas, minerals, timber, ciples of rule toWlblsana, who is eventually crowned King ofAlengka (Sri Lanka). The principles emphasize the duty of the king to give his subjects spirit, in- and primary energy spiration, and the means of earning their livelihood, to light their minds and souls, and to provide direction, responsibility, example, and impartialityThe ruler resources must be free of hatred and pettiness and must be solid, honest, and willing to reward those who serve the land and its people. "Every ruler who does not pos- sess these characteristics is a king without a crown, while every commoner who does is a crowned head in reality" (Mulyono 1981, pp. 136-37). Most Indonesians, including many non-Javanese, are familiar from child- hood with the teachings of the wayang shadow plays. This fact should be re- membered when considering Indonesian history (and current events)-because in Indonesia events are often interpreted and choices made in light of wayang teachings. Indonesia's Environment, Demographics, Early History, and Struggle for Independence The world's fourth most populous country, Indonesia lies between Asia and Aus- tralia on the world's largest archipelago, a strategic location on, and alongside, major sealanes from the Indian Ocean to the Pacific Ocean (see map of In- An Introduction to Indonesia 25 donesia). Divided into 23 provinces, 2 special regions, and 1 special capital city district, and covering 1.8 million square kilometers, the country comprises some 13,700 islands spread over a variety of ecological zones.A pluralistic country, in 2000 Indonesia had a population of 210 mnllion people from many cultur- al, linguistic, and ethnic backgrounds (World Bank 2002, World Development Re- port 2002). In 1999 about 60 percent of the population was rural, down from 78 percent in 1980 (World Bank 2001, World Development Report 2000/2001). But in densely populated Java and Bali, where most of the population lives, the distinction between rural and urban is not always clear. Indonesia is nearly 90 percent Muslim but is also home to Protestants, Roman Catholics, Hindus, Bud- dhists, and other religious minorities. Educating Although hundreds of languages are spoken, the nation's official language is Bahasa Indonesia (commonly called Indonesian in English), which is known Indonesians was not by nearly all Indonesians.The Malayan language has served as a lingua franca throughout the Indonesian-Malayan archipelago since the 15th century. Both a Dutch colonial present-day Bahasa Malaysia and Bahasa Indonesia, which are mutually intel- ligible, are modernized versions of the older Bahasa Melayu. Bahasa Indone- priority. At sia was proclaimed the official language of Indonesia in 1928, as part of the early independence struggle. The widespread acceptance of a national language in independence there this large, multilingual country was undoubtedly helped by the fact that the of- ficial language chosen was not thejavanese language of the dominant majority- were only 337 a lesson not learned in some other countries (India, for example).Among people 15 and older in 1998, 91 percent of men and 80 percent of women were re- higher education ported to be literate (World Bank 2001, World Development Report 2000/2001).9 The country's rich and diversified resource base includes oil, natural gas, graduates in a minerals, timber, and substantial primary energy resources. Indonesia is the world's largest exporter of liquefied natural gas. In addition to exploitation of natural population of 70 resources, the economy is based on manufacturing of consumer goods (tex- tiles, processed foods, garments, shoes, furniture) and intermediate goods (ply- million wood, cement, fertilizer), on agriculture and fishing (rice, cassava, maize, fish, and poultry and other animals for domestic consumption, and rubber, coffee, palm oil, spices, tea, cocoa, food, live animals, and fish for export), on trade, and on hotel and restaurant services.10 Early Indonesian history and culture (see Koentjaraningrat 1975b, 1985, and Fox 1980) were formed from a complex mix of influences from Chuna, India, Southeast Asia, and the Arabian peninsula. During the 15th century the Por- tuguese, Spanish, British, and Dutch fought for control of the Indonesian spice trade. By the 17th century the Dutch dominated the trade routes, and in the 18th and 19th centuries they occupied most of what is now Indonesia. Indonesia became administratively unified under Dutch colonial rule, which lasted until the Japanese occupation during World War II. The Dutch saw no need to bring significant numbers of Indonesians into government or to start preparing them to manage their own affairs. The 1930 census, the last before 26 The Microfinance Revolution: Lessons from Indonesia Indonesia's independence, showed there were 208,269 Dutch living in Indonesia.They ran virtually everything.. .M.S. Ricklefs in hls study A History of Modern Indonesia quotes Dutch governor- general B.C. de Jonge (1931-36) as saying, "We have ruled here for 300 years with the whip and the club, and we shall still be doing it for another 300 years." Education for the local people was also not high on the Dutch agenda, with the result that there were only a few hundred Indonesian college graduates out of a total population estimated at 70 million at the time of independence."1 -Masters 1999, p. 2 Independence was During the 1920s the Indonesian National Party (Partai Nationalis Indone- sia, or PNI) was organized by Sukarno-popularly known as Bung (brother) Karno, declared in 1945, later the country's first president-and six other founders. The PNI urged im- mediate independence. But the Netherlands strongly resisted the idea of Indonesian but the struggle to independence, and in 1929 Sukarno and many of his associates were imprisoned. Released in late 1931, in part because of protests by humanitarians in the Nether- gain it continued. In lands, Sukarno joined Mohammad Hatta and Soetan Sjahrir in forming a new party, the Indonesian National Education Party (Pendidikan Nasional Indone- 1949 Indonesia sia); the party's initials were intentionally kept the same as the previous independence party. Sukarno was imprisoned again in 1932 and then exiled to a remote fish- finally became a ing village in eastern Indonesia, where he remained until 1942. In 1940 Germany invaded Holland, and in 1942 Japan invaded Sumatra and unified republic with then Java. Sukarno became the nominal head of the Indonesian government under Japanese rule. He helped Indonesians when and as he could, but little Sukarno as could be done to mitigate the terrible hardship and suffering that the Japan- ese occupation brought to Indonesia. president InAugust 1945Japan surrendered to theAllies.Japan had promised to grant independence to Indonesia, but the promise was not kept and Japan handed Indonesia over to the Allies. However, during the short period between the Japanese surrender and the arrival of the Allied forces, Indonesia declared independence. A short procla- mation, written by Sukarno in Indonesian, stated: "We the people of Indone- sia hereby declare Indonesia's independence. Matters concerning the transfer of powers and other matters will be executed in an orderly manner and in the shortest time possible." Sukarno and Hatta signed the proclamation on behalf of the Indonesian people on 17 August 1945. But independence was not yet won. The Dutch returned almost immedi- ately, killing some 8,000 civilians in Jakarta by December 1945 (Neill 1973, p. 325). The Dutch tried several times to assassinate Sukarno, while the British (in Indonesia in the aftermath of the Japanese surrender) tried to arrest him. Over the next four years guerrilla warfare ensued in many parts ofJava as well as in Sumatra and Bali. Finally, on 27 December 1949 Indonesia became a fully independent state with Sukarno as its president and Hatta as its vice president. An Introduction to Indonesia 27 The Sukarno Era, 1949-67 Under President Sukarno Indonesia became a unified republic.'2 A constitu- tion was written and announced on 18 August 1945, the day after Sukarno and Hatta had proclaimed Indonesia's independence. Meant as a provisional con- stitution, it was short and general, but provided for a powerful presidency. Although the large majority of Indonesians are Muslim, the country was not established as an Islamic state. Sukarno's Pancasila was offered as a com- promise among groups with strongly different views on this and other issues. "The primary objective of this fuzzy doctrine is rooted in its first principle [be- lief in one supreme being] which aimed to undercut demands from the Mus- Sukarno' lim community for an Islamuc state" (Schwarz 1994, p. 10). Sukarno's presidency (1949-67) began with a period of parliamentary presidency began democracy. In 1950, when Indonesia became a fully independent state, a new constitution was adopted that mandated a parliamentary government, provid- with a period of ed detailed guarantees for human rights, instituted a system of checks and bal- ances for political institutions, and subordinated the military to civilian parliamentary leadership."3 In 1953 the government announced plans for general elections, which were held in 1955. More than 90 percent of the registered voters cast democracy. In 1959 ballots-in what turned out to be Indonesia's last democratic election until 1999. No party contesting the 1955 elections won a majority of the votes, and he switched to Sukarno continued as president with a coalition government. An election was also held in 1955 to elect delegates to the Constitutional 'guided Assembly (Konstituante), which was to draw up a permanent constitution. De- spite its diverse membership, the Constitutional Assembly was able to reach broad democracy "-in consensus on issues concerning human rights and safeguards against arbitrary use of power. But the delegates were polarized on some issues, most notably practice personal rule the debate over whether Indonesia should become an Islamic state. By then the economy had begun a sharp declne, a rebellion had occurred "reminiscent of inWest Sumatra,14 regional discontent withJakarta's leadership had grown, the military had become increasingly unhappy with parliamentary democracy, and Javanesefeudalism" a sense of crisis was forming. On 5 July 1959 Sukarno dissolved the Constitutional Assembly, abrogated the 1950 constitution with its division of powers and human rights protections, and decreed a return to the 1945 constitution (which remains in effect today). This ended the experiment in parliamentary democracy.What followed next was Sukarno's "Guided Democracy." Guided democracy, its title notwithstanding, meant in practice a return to a system of personal rule more reminiscent of Javanese feudalism than the chaotic democratic experiment of the 1950s.'In Guided Democracy,' Sukarno once said, with typical flair,'the key ingredient is leadership.The Guider. . . incorporates a spoonful of so-and so's opinions with a dash of such-and such, always taking care to incorporate a soupfon of the opposition. 28 The Microfinance Revolution: Lessons from Indonesia Then he cooks it and serves his final summation with "OK, now my dear brothers, it is like this and I hope you agree..." It's still democratic because everybody has given his comrnent.' -Schwarz 1994, pp. 16-17 Under Guided Democracy the influence of the Indonesian Communist Party (Partai Komunis Indonesia, or PKI) rose and became a source of serious con- cern to Islamic parties and to the military. Meanwhile, the military also became increasingly alarmed about political Islamic groups and their goal of Indone- sia as an Islamic state.The military expanded its operations under what had be- come known as the "the middle way," a doctrine formulated by Major General Abdul Haris Nasution in 1958. 15 According to the middle way the armed forces By the early 1960s were not only a military force, but also a socio-political force with a role in preventing national instability. Under Guided Democracy the military received the economy was in representation in the cabinet, civil service, and parliament. Later, under the mid- dle way, the military greatly expanded its role throughout the country. steep decline. And The Sukarno administration was based on an uneasy coalition of nation- alists, religious groups, and communists known as NASAKOM (Nasionalisme Sukarno-listening [nationalism], Agama [religion], Komunisme [communism]). An inherently unstable grouping, it began to rupture in the late 1950s and early 1960s. only to what he Sukarno, a political revolutionary, was little interested in economics. By the early 1960s the economy was in steep decline. Production and investment had wanted to hear- fallen sharply. Debt soared, real per capita income declined, the country was wracked by hyperinflation, and the agricultural sector could not produce lost touch with the enough food to feed the population. Export earnings declined sharply, the coun- try's net foreign exchange reserves were negative, and the budget deficit reached Indonesian people about half of government spending. Banks, choked with regulations and un- able to attract funds because of inflation that reached more than 600 percent, ceased most commercial lending. Most of the population was poor and get- ting poorer.16 Meanwhile, Sukarno had lost touch with the Indonesian people, listening only to those who said what he wanted to hear. "Dialogue died. The flow of opimon was now only from the top down.The sounding board for public opin- ion became an echo of the leader" Jones 1971, p. 248). Increasing antagonism between the PKI and the army, along with the growing economic problems and a narrowing of Sukarno's political base, led to a violent coup attempt in late 1965. Responsibility for the attempted coup is still in some dispute. Among the possible candidates are the PKI, some army officers, the U.S. Central Intelligence Agency, President Sukarno, then-Major General (later President) Soeharto, and others.'7 The coup attempt occurred the night of 30 September 1965, when leftist army troops murdered six army generals-members of the army's General Staff- as well as an aide, then buried them in an unused well injakarta. Soeharto, who led the army's Strategic Reserve Command (Kommando Strategis Angkatan Darat, or KOSTRAD), assumed control of the army by early morning 1 Oc- An Introduction to Indonesia 29 tober. The army blamed the PKI for the attempted coup. Along with youth, nationalist, and religious organizations that were encouraged to take part, the army massacred PKI members, followers, suspects, and others who were dis- liked by the attacking mobs.The victims included many ethnic Chinese, who were widely resented by indigenous Indonesians (pribum() because of their promi- nence in economic activities (box 8.1). The bloodbath that erupted on Java and Bali in October 1965 continued into 1966; an estimated 300,000-400,000 people were killed.8 The complex- ities of the attempted coup, and of the massacres that followed, have yet to be fully unraveled. Soeharto emerged from the havoc as the new national leader. Sukarno A violent coup was permitted to retain the title of president, but he was relieved of his pow- ers and prohibited from engaging in political activity until after the next elec- attempt occurred in tion. Executive power was transferred to Soeharto in March 1966; he was appointed acting president by the Provisional People's Consultative Assem- 1965,followed by a bly (Majelis Permusyawaratan Rakyat Sementara, or MPRS)'9 in 1967, and president in 1968. Sukarno, who never regained power, died in 1970. At the massacre that time of his death, about 60 percent of Indonesians lived below the poverty line. continued into The rest of this chapter considers Indonesian history since Soeharto assumed the presidency in 1967. It focuses first on the reasons for the country's impressive 1966. Soeharto and sustained economic development over three decades, then considers the reasons for the devastating economic, financial, and political crisis that began emerged as the new in1997; progress toward recovery; and early steps toward a workable political democracy. leader and soon became president The Soeharto Era, 1967-98 When Soeharto's New Order government took de facto control in March 1966, it inherited a fractured society and an economy in chaos.Thirty years later, in 1996, the IMF ranked Indonesia seventh in its list of emerging economies. Sim- ilarly, a 1997 World Bank report on Indonesia was titled "Indonesia: Sustain- ing High Growth and Equity."According to estimates for Indonesia made by the Bank in 1996: Economic growth is expected to remain at about 7.5 percent over the next few years, despite a declining contribution from the oil sector. Ongoing trade and industrial sector reforms should continue to open profitable areas for relatively labor- intensive export industries. -World Bank News, 4 April 1996, p. 6 In stark contrast, a July 1998 World Bank report on Indonesia, called "In- donesia in Crisis: A Macroeconomic Update," stated: 30 The Microfinance Revolution: Lessons from Indonesia Box 8.W Excerpts from David C. Cole and Betty F. Slade's Building a Modem Financial System: The Indonesian Experience A brief description of the evolution of the economic role of the ethnic Chinese business- men in Indonesia [is provided]. because many financial policies resulted from the over- whelming and very obvious economic success of many ethnic Chinese in Indonesia and their seeming domination of many aspects of economic activity In the early years of the New Order, Bruce Glassburner wrote It is the Chinese who pose the most difficult problem because of their vastly disproportionate share of economic power. The indigenous majority regard the power of the Chinese with suspicion, fear and hostility The indigenous The World Bank peoples typically regard the Chinese with derision for their devotion to trade, moneylending, banking, and manufacturing, while both resenting and envying the wealth and power which they enjoy Unfortunately, for the economic poli- observed in 1998 cymakers in the government, this dominance of Chinese entrepreneurship presents a major complication Policies which attempt to revitalize the private during the sector benefit the Chinese ethnic group relatively more than the indigenous majority, so far as direct effects are concerned Attempts to build an indige- nous entrepreneur group through subsidy and protection have not been nota- Indonesian crisis, bly successful (Glassburner, 1971, pp 181-82) "No country in In 1994 Schwarz...could affirm that the problem was influencing policymaking Indonesia's small, ethnic-Chinese business community dominates the private recent history has sector Many leading pnbumi [indigenous Indonesian] businessmen conclud- ed that the technocrats' deregulation measures [of the 1 980s and early 1 990s] ever suffered such a were specifically aimed at helping the ethnic-Chinese increase their dominance of the business world The fact that many of these same pnbumi business- men also benefited from the deregulations didn't blunt their resentment [pp dramatic reversal of 80-81] How did this important economic role of the Chinese come about? Glassburner fortune" (1971, p 9) attributes it to the legacy of the Dutch colonial regime which seems to have encouraged a "special economic and social position of the Chinese" [in addition,j the extensive system of Chinese schools was much more developed than those for Indonesi- an pnbumi children When the Dutch left Indonesia, the Chinese were the strongest economic force left However, they did not have political or military power in the new independent Indonesian state and were faced with building up new relationships. The Chinese have been suc- cessful since that time in joining with pnbuml businessmen in economic activities, includ- ing links with politicians, bureaucrats, and the military In the last two decades, a number of Chinese businessmen have joined forces with President Soeharto's emerging family interests This collaboration allowed the Chinese to get around the discriminatory practi- ces against them but, on the other hand, also provided the business skills and experience of the ethnic Chinese to a venture Foreign investors often preferred working with Chi- nese businessmen in Indonesia "because they were better equipped in terms of busi- ness experience, capital and technical know-how." Chinese firms have maintained close connections with Singapore and Hong Kong which have opened up opportunities for credit and trade and given them a distinct comparative advantage The ethnic Chinese, originally mainly in commercial, service, financial, and small in- dustrial sectors began to become more involved in larger manufacturing and industrial An Introduction to Indonesia 31 Box &1 (continued) activities However, many Chinese have remained in small-scale operations Soeharto remained close to powerful ethnic Chinese businessmen and involved them in most ma- jor investments with his family, friends and close colleagues Note Cole and Slade's account was published in 1996 During 1997 and 1998 many of the protests and riots that led to Soeharto's downfall targeted ethnic Chinese, who were widely seen as control- ling the nation's economy under Soeharto's protection and at the expense of indigenous Indonesians (pribuml) Source Cole and Slade 1996, pp 322-27 Injavanese culture, power is believed to Indonesia is in deep crisis. A country that achieved decades of exist as an entity rapid growth, stability, and poverty reduction, is now near economic collapse.Within the space of one year Indonesia has independent of its seen its currency fall in value by 80 percent, inflation soar to over 50 percent, the economy swing from rapid growth to users. The ruler has even more rapid contraction, unemployment climb rapidly, and the stock exchange lose much of its value. Foreign creditors have all the power-but withdrawn, investors have retreated. Capital and entrepreneurs have fled. Long standing defects in governance, earlier it can move to camouflaged by rapid growth, have now been unmasked as fatal flaws.. .No country in recent history, let alone one the size another holder of Indonesia, has ever suffered such a dramatic reversal of fortune. -World Bank 1998b, p.1 The World Bank was not alone in not anticipating the crisis. [A] study by Kaufmann, Mehrez, and Schmukler (1998) looked at various indicators of market expectations based on the Global Competitiveness Survey (GCS), McGraw-Hill Global Risk Service, and Moody's and Standard and Poor's (S&P) ratings. Only businessmen and local investors seem to have foreseen the crisis in Thailand and Korea, and to a lesser degree in Malaysia. No one predicted the crisis in Indonesia. -World Bank 1998a, p. 43 The reasons that the crisis was not anticipated, and the multiple warning signals that were not sufficiently heeded, are discussed later in the chapter. The exceptional achievements-and extraordinary regressions-that occurred during Soeharto's 32-year presidency are best viewed in the complex context in which they developed. Along with lessons from the wayang, two concepts 32 The Microfinance Revolution: Lessons from Indonesia are important for understanding Indonesia generally and the Soeharto era in particular: the Javanese concept of power, and Pancasila, the guiding principles of the Indonesian government. The Javanese concept of power Soeharto is Javanese, as was Sukarno and as are about 45 percent of Indone- sians. The Javanese concept of power is central to understanding Soeharto's presidency. BenedictAnderson's 1990 [1972] seminal work,"The Idea of Power in Javanese Culture," analyzes the differences between Western and Javanese concepts of power. InJavanese culture power is seen as a concrete entity that exists mdependent of its users.The total amount of power remains fixed, though its distribution may vary. "Power simply exists, and is not the product of or- Pancasila, the state ganization, wealth, weapons or anything else-indeed [it] precedes all of these and makes them what they are... Since all power derives from a single ideology, is so broad homogeneous source, power itself antecedes questions of good and evil... .Power is neither legitimate or illegitimate. Power is" (Anderson 1990 [1972], pp. that it can appeal to 22-23). As Schwarz (1994, p. 45) says, "The Javanese ruler does not have some of highly diverse the power, he has all of it. Power is a zero-sum game: to get it, you have to take it from someone else. There is no sense of broadening your scope of power by groups. It also seeking a mandate from your subjects." Soeharto saw the power he took from Sukarno as emanating from God, not enables a powerful from the Indonesian people.20 In this concept of power there is no scope for power sharing or for independence in the judiciary, legislature, political parties, or president tojustify media.Yet the traditional Javanese concept of power is not necessarily believed in by the many Indonesians outsideJava or, in the modern era, by manyjavanese. virtually any Pancasila, the guiding principles of Indonesian government decision Pancasila-belief in one supreme being, Indonesian unity, humanitarianism, democracy by representative consensus, and social justice-is a state ideology of great flexibility that can be (and has been) pulled, pushed, and stretched to cover just about any situation. Accordingly, and in contrast to the Javanese concept of power, there has been a wide gap between form and practice in Pancasila. In reality Indonesia is far from the ideal of Pancasila democracy. Not only is it not democratic in the Western liberal sense, it is not democratic in the Pancasila democracy sense either. The lmperative of"consensus at all costs" leaves Indonesians with little scope to disagree with official policy...The interesting thing about Indonesia's Pancasila democracy is that it includes many of the features of democratisation-secret balloting, universal adult suffrage, regular elections-but relatively few of the individual and group freedoms on the liberalisation agenda. It has, in other words, the form of (Western) democracy but not the content. -Schwarz 1994, pp. 292, 294 An Introduction to Indonesia 33 However, Soeharto strongly promoted acceptance of Pancasila principles, although not necessarily their implementation.The general principles of Pan- casila incorporate three important characteristics.They are so universal that they are widely agreed on and cannot easily be openly opposed.As the national ide- ology, they provide protection against specific ideologies or religious values that various groups advocate for adoption by the state. And their existence has al- lowed the government to control those it declared to be following anti-Pan- casila practices. An example of the second characteristic is that Soeharto, invoking the Pan- casila state ideology, was able to hold off militant Muslim munorities advocat- ing that Indonesia become an Islamic state, as well as those who argued for Real annual income exceptions for Muslims under Pancasila ideology. An example of the third is that under Soeharto's rule and the Pancasila doctrine of tolerance, ethnic and for the average religious conflicts were prevented where possible and controlled when they oc- curred. Negative examples of the third characteristic were common. Interfer- Indonesian was ence with political parties, "mysterious killngs" of alleged criminals without trial,21 control of the press, banning of books and plays, and the like were typ- nearlyfour times ically justified as protecting the society from anti-Pancasila elements. higher in 1996 The economy, 1966-96 The New Order government is known for its rapid economic growth and de- than it was in 1970 velopment. During three decades of high growth the economy became diver- sified, as the country moved away from dependency on oil and developed an export-oriented manufacturing base. Major achievements in poverty reduction, agriculture, population control, and education brought international recogni- tion. Indonesia's population grew from 118 million in 1970 to 197 million in 1996. However, "real annual income for the average Indonesian was nearlyfour times higher in 1996 than it was in 1970" (Radelet 1999, p. 1). Between 1986 and 1996 real GDP grew by an average of 7.9 percent a year, increasing per capita incomes by nearly 80 percent. In 1996 per capita income reached $1,073 (in constant 1995 U.S. dollars; table 8.1).The focus here is on the growth of the economy from 1966-96; the crisis that began in 1997 and current reha- bilitation efforts are discussed later in the chapter and in chapter 15. Macroeconomic management and the technocrats. In 1966, as Soeharto's New Order government took power, the economy was paralyzed.The first priorities were to form a team of economic managers and to begin a process of economic restructuring.The new government introduced stabilization and rehabilitation policies, paying immediate attention to reducing inflation. Emphasis on devel- opment was next: increasing food production (especially rice), developing in- frastructure, strengthening the mining and manufacturing sectors, and encouraging foreign investment and export expansion (see Booth and McCawley 1981b). These policies were the work of a team of economics ministers that imi- tially had five members, though it later expanded.22 The team is known infor- mally as the "Berkeley mafia" (because many of its members were educated at 34 The Microfinance Revolution: Lessons from Indonesia Table 8.1 Economic indicators for Indonesia, 1996-2000 Indicator 1996 1997 1998 1999 2000 Exchange rate (rupiah/U.S. dollar, end of penod)a 2,383 4,650 8,025 7,085b 9,595 GDP growth (percent, at constant 1993 growth rates) 7 8 4 7 -13 0 0 3C 5 2C Annual inflation (percent, measured using consumer prices)a 8 0 6.2 53 4 20 5 3 7 GNP per capita Macroeconomic (constant 1995 U S dollars) 1,073 1,099 903 906 n a n a Not available policy in the 1980s a See appendix 1 for more details b The small discrepancy with tables 2 1 and 2 3 in volume 1 for the 1999 exchange rate occurs because more recent IMF data were used in this volume emphasized nonoil c Preliminary data from Biro Pusat Statistik (Indonesian Bureau of Statistics) Source IMF, International Financial Statistics, March 2001, World Bank data, Biro Pusat Statistik data exports. the University of California at Berkeley).Throughout, ProfessorWidjojo Ni- Manufactured goods tisastro has been the leader of the technocrats. ProfessorAliWardhana, Indonesia's longest-serving mmnster of finance, has served (among many other roles) as the rose from 2 percent leader in the development of Indonesia's financial system.This group of tech- nocrats has guided Indonesia's economy for more than 30 years. But as discussed of export value in later in this chapter, their influence had waned by the mid-1990s, when Pres- ident Soeharto had virtually stopped listemng to them. In the post-Soeharto 1980 to 53 percent era the senior technocrats have played important, though generally less public, roles. But despite their efforts, recent governments have allowed political pres- in 1993 sures to interfere with badly needed economic and financial reforms. The economic team first won the confidence of the country with its sta- bilization of the economy after the havoc of the mid-1960s.They won further respect, both at home and abroad, by effectively managing the economy through a series of oil booms and busts during the 1970s and 1980s, and by implementing far-reaching reforms in fiscal policy, foreign trade, and the financial sector beginning in the early 1980s. During the 1970s Indonesia's economy was heavily based on its rich nat- ural resources, especially oil and gas, copper, gold, tin, rubber, and palm oil. Ex- port revenues financed widespread infrastructure-roads, irrigation systems, ports, and schools-as well as massive health, family planning and education programs. "While there was clearly extensive waste and abuse, Indonesia managed its re- sources far better than most resource-abundant developing economies in the 1970s and 1980s" (Radelet 1999, p. 2). Agricultural output grew steadily be- ginning in the 1970s, supported by the government's investments in rural de- velopment, green revolution technology, and the relatively stable prices offered to rice farmers (see chapters 9 and 11). An Introduction to Indonesia 35 Three elements of macroeconomic management instituted by the New Order government in its early days guided the economy for decades.23 First, the gov- ernment required a balanced budget every year. This permitted the financial system to channel funds primarily to state and private enterprises that emphasized direct productive activities. It also prevented banks and the public from being forced to fund government deficits.While foreign borrowing was used to bal- ance resources and expenditures, the balanced budget rule enforced a budgetary discipline that protected the economy from deficit spending and inflationary excess.Although there were some off-budget expenditures, the balanced bud- get rule worked well for years. But by the latter part of the Soeharto era, the rule may have helped hide growing off-budget expenditures and increasing lack In the early 1 980s of transparency. Second, in what was considered a radical move at the time, the rupiah was the government made fully convertible in 1970.The open capital account placed a second ex- ternal constraint on monetary policy; it compelled the government to take rapid initiated a series of corrective action when external or internal developments threatened the bal- ance of payments. "The open capital account.. .ensures that any monetary mis- majorfinancial management will show up almost immediately in an outflow of foreign exchange. Thus convertibility imposed the discipline needed to deal with reforms to deregulate monetary pressures whenever these arose" (Wardhana 1998a, p. 128). Over time the government adopted an exchange rate mechanism that aimed at maintaimhg thefinancial sector the real international value of the rupiah by adjusting the rate to reflect move- ments in the domestic consumer price index relative to the international prices and promote a of its main trading partners. But by the mid-1990s a highly leveraged, risk-taking corporate sector and competitive economy foreign investors eager for Indonesian equity and debt could make use of the open capital account in ways that led to a substantial increase in private for- eign debt (Kenward 1999, p. 80). As discussed later, much of this debt was un- hedged and contributed significantly to the crisis in 1997. Third, the government instituted a relatively conservative borrowing pol- icy that lasted through the 1980s. For example, unlike many other oil-export- ing countries, Indonesia had a debt service ratio of only 15 percent in 1984, the year before an international collapse in oil prices (Wardhana 1998a).Thus effects on the Indonesian economy of the falling oil prices of 1985-86 could be cushioned by expanding international debt. But in the 1990s borrowing increased.A major source of the problem was the Soeharto family and friends, who used state banks to leverage funds from foreign banks for large, often nonviable, investment projects. By 1993 Indone- sia was ranked by the World Bank as the fourth largest debtor among devel- oping countries. Another important aspect of macroeconomic policy in the 1980s was the emphasis placed on the rapid increase of nonoil exports. In 1984 oil and gas revenues accounted for nearly three-quarters of Indonesia's dollar earnings from exports, but by 1989 manufactured exports accounted for nearly a third of ex- port earnings, while oil and gas contributed less than 40 percent of the total 36 The Microfinance Revolution: Lessons from Indonesia (Wardhana 1998a). Major exports include plywood, textiles, garments, shoes, toys, and furniture.24 Its large labor supply and relatively low wages give In- donesia a comparative advantage in labor-intensive products. In addition, most export industries are labor-intensive by government policy (although regula- tions and corruption have sometimes hindered implementation of this poli- cy). Manufactured exports grew by nearly 30 percent a year between 1986 and 1996.As a share of the value of exports, manufactured goods rose from 2 per- cent in 1980 to 53 percent in 1993 (World Bank 1997, World Development Re- port 1997). And the government increased efforts to deregulate the trade sector, sharply reducing multiple trade restrictions. It reduced and eventually removed numerous restrictions on foreign direct investment.And it adopted "a modern, simplified The newly tax system that removed low-income wage earners from the tax net, eliminat- ed, at least in principle, nearly all exemptions, and introduced the value added deregulatedfinancial tax" (Stern 2000, p. 2). system was to Financial reforms. In the early 1980s the government initiated a series of financial reforms to deregulate the financial sector and promote a competitive operate within the economy (see Wardhana 1994b, 1998a; Hanna 1994; McLeod 1994; Nasution 1994; Binhadi 1995; Cole and Slade 1996; and Stern 2000, forthcoming). prudential Some of the most important of the reforms are summarized below. The first reform, announced in June 1983, focused on deregulating cred- regulations. But this it. Credit ceilings were abolished. Liquidity credits from Bank Indonesia, the central bank, were reduced except for priority sectors. And banks were per- proved difficult to mitted to set their own interest rates on most loans and deposits.Among its other purposes, the June 1983 reform was crucial for the development of BRI's mi- implement, often crobanking system. In December 1987 a package of deregulation measures, known as PAKDES impossible I,25 was issued to activate and reform the stock market.Then in October 1988 a broad package of financial reforms, known as PAKTO 88,26 was announced; one of its most important aspects was its emphasis on competition, especially within the banking sector. PAKTO 88 opened new opportumties for finan- cial activity in rural areas. Restrictions on the operations of foreign banks were reduced. Requirements for opening new domestic banks were substantially eased. Banks wholly owned by Indonesian nationals were now permitted to open branches anywhere in Indonesia. And, in general, banks were given more au- tonomy and encouraged to expand their activities. Included among the provisions were a relaxation of requirements for open- ing branch offices and encouragements for banks to introduce new products. Reserve requirements were lowered from an average of 11 percent to a uni- form 2 percent on all deposits, reducing bank costs of intermediation. PAKTO 88 had a deep effect on the banking system. "The short-run impact of the PAKTO reforms was basically favorable.They had probably contributed to ac- celerating overall economic growth, without adversely affecting relative price stability and balance of payments equilibrium. In addition the banking system An Introduction to Indonesia 37 became more competitive and customer-oriented and was rapidly expanding both its facilities and services" (Cole and Slade 1996, p. 115). But in the longer term, the proliferation of banks that ensued, combined with the growing cor- ruption of wealthy and politically influential families and with inherent weak- nesses in the financial system, contributed to the financial crisis of the late 1990s (see chapter 15). Additional reforms followed, focusing on private securities markets, insur- ance, and promotion of instruments such as venture capital, leasing, and cred- it cards. PAKJAN,27 a January 1990 reform package, ended or cut back many of the subsidized credit programs that were administered by Bank Indonesia and channeled through state banks. Most of these liquidcty credit arrangements In 1970 about 60 were abolished except for those supporting food self-sufficiency and cooper- ative and small-scale enterprise development. In a political tradeoff, however, percent of PAKJAN also directed all domestic banks to allocate at least 20 percent of their loan portfolios to loans below 200 million rupiah (about $111,000 at that time).2s Indonesians lived But the growth of the financial system caused increasing concern about the need for better prudential regulation. In December 1990 the government pro- below the poverty vided for the privatization of the stock exchange, announced new regulations on loan loss provisioning, and required banks to adopt (over several years) the line. In 1996 the capital adequacy standards of the Basel Agreement. In the years that followed, the government's main priority for the financial system was to ensure that the figure was 11 newly deregulated system operated within appropriate prudential regulations. This effort included new measures of bank soundness and new laws on bank- percent ing, msurance, and pensions that were approved by parliament m the 1992 Bank- ing Law. Emphasis was also placed on restructuring bank management and operations so that banks could comply with the new regulations. Underlying the multiple, multiyear financial reforms was the overall goal of moving from the previous pattern of direct government control of the fi- nancial system to oversight of the system through prudential regulation.As dis- cussed later in the chapter and in chapter 15, however, effective implementation of the prudential regulations proved extremely difficult, often impossible, in the Indonesian political context of the 1990s. Poverty reduction and human resource development. Indonesia's rapid and sustained growth in these years made possible massive reduction in poverty, a remarkable achievement. In 1970 about 60 percent of Indonesians lived below the poverty line. By 1976 the portion of the population below the poverty line had declined to 40 percent (or nearly 54.2 million people)-and by 1996, to 11 percent (or 22.5 million people; Biro Pusat Statistik [Indonesian Bureau of Statistics] 1996, p. 570). Even during periods of financial difficulties, fiscal allocations and develop- ment expenditures for programs directly benefiting the poor were usually pro- tected. In addition, industrial development was carried out in a policy environment that emphasized the creation and expansion of labor-intensive in- dustries.The government was strongly committed to antipoverty programs and 38 The Microfinance Revolution: Lessons from Indonesia made special efforts to identify and help the poorest villages and the poorest people (see Haryono Suyono,"Indonesian Family Empowerment to Alleviate Poverty," Indonesia Times, 6 September 1997, p.7).As noted, average income in the poorest quintile of the population grew faster between 1976 and 1990 than average income for the entire population. And real annual income for the av- erage Indonesian was four times as high in 1996 than it had been in 1970. Indonesia also invested heavily in rural development. As discussed, a sub- stantial share of the oil wealth of the 1970s was invested in economic and so- cial infrastructure-much of it in rural areas, where more than 80 percent of the population lived at that time. Improvements in agriculture enabled Indonesia to become self-sufficient in rice and, despite a growing population, allowed per capita food consumption to grow, resulting in widespread improvements in nu- A substantial share trition and health. Family planning and primary education were made high priorities, and na- of the 1970s oil tionwide programs were effectively implemented beginning in the 1970s.Av- erage annual population growth fell from 2.3 percent a year during 1970-80 wealth was invested to 1.7 percent during 1990-97. The total fertility rate decreased from 4.3 in 1980 to 2.6 in 1996,29 while infant mortality (per 1,000 live births) plummeted in rural from 90 to 49 during the same period. Life expectancy at birth rose from 41 years in 1960 to 53 in 1980; by 1996, lfe expectancy was 63 years for men and infrastructure, 67 for women (World Bank 1982,1999). Education also expanded rapidly. Ninety-seven percent of the relevant age agriculture, group was in primary school in 1996 (compared with 71 percent in 1960 and 86 percent in 1977). And in 1996, 88 percent of both males and females education, and reached grade 5. Forty-two percent of the secondary school age cohort was enrolled in school that year (compared with 6 percent in 1960 and 21 percent health andfamily in 1977). By 1997 both females and males were expected to receive an aver- age of 10 years of formal schooling (World Bank, 1980,2000,2001). In the 1990s planning increasing priority was placed on tertiary education as well. Overall, the New Order government's macroeconomic policies and wide- ranging reforms achieved much over three decades, including substantial growth in GDP and per capita GNP, single-digit annual inflation from 1985-97, large increases in food supplies, dramatic nationwide successes in family planning and primary education, and a sharp reduction in poverty. Indonesia in relation to other emerging economies. Among the 15 lowest of the world's lower-middle-income economies in 1995, Indonesia (with $980) ranked eighth in GNP per capita.30 By far the most populated of the coun- tries, in 1985-95 Indonesia ranked first among the 15 in average per capita GNP growth (6.0 percent a year, compared with the median of 1.0 percent). In 1995 Indonesia had one of the highest average GNP growth rates (7.2 percent, com- pared with the median of 2.3 percent).The share of income held by the poor- est quintile of the population was highest in Indonesia and Egypt, both with 8.7 percent in 1995. Indonesia was at or near the median for all other economic and social indicators (World Bank 1997f, 1997g). An Introduction to Indonesia 39 Because of Indonesia's large population and its location in Asia, it is also useful to compare indicators for the eight Asian countries with low-income and lower-middle-income economies that in 1995 had populations of more than 50 million-in descending order of size of population: China, India, In- donesia, Pakistan, Bangladesh,Vietnam, the Philippines, and Thailand. Indone- sia, the third most populous but least densely populated of the eight countries, was above or near the median on all indicators. For the period 1985-95, In- donesia ranked third in average annual GNP per capita growth (afterThalland and China). It was third in 1995 GNP per capita (afterThailand and the Philip- pines). Among the eight countries, Indonesia was second highest in the share of income held by the poorest quintile in 1995 (8.7 percent), after Bangladesh The economics team (9.4 percent;World Bank 1997f, 1997g). was knownfor both Economic growth with equity. Indonesia made extraordinary progress in eco- nomic and social development between 1967 and 1996. Under the New Order vision and government, emphasis was placed simultaneously on raising economic growth and on reducing poverty; on improving agriculture and on increasing indus- caution-andfor try; on encouraging manufactured exports that are highly labor-intensive; on stimulating both public and private sectors; and on developing the country's decades of stable natural and human resources. The economics ministers worked as a team, drawing on a combination macroeconomic of caution and vision and a remarkable ability to learn from their mistakes, to implement tough policies, and to adapt to changing circumstances. Their management that stable and foresighted economic management was crucial for the country's development. enabled economic Since independence, however, Indonesia has been a country with a strong ruler and a weak state. As the economy grew it was also being undermned by growth with equity political forces discussed below.This intricate disconnect could not continue mdefinitely.The seeds of the recent crisis had long been germinating in Indonesian politics, even as the economy expanded year after year. The government, the military, and politics during the Soeharto period The Indonesian government operates under the original 1945 constitution which established a strong presidential form of government and a centralized admin- istrative structure, and mandated the principle of government by consensus.The cornerstone of the preamble to the constitution is the Pancasila. The discus- sion below refers to the Soeharto period. Events that occurred after his resig- nation in May 1998 are discussed later in this chapter and in chapter 15. The Indonesian military operates as both a mnlitary and a sociopolitical force under the doctrine of dwifungsi, or dual function-an outgrowth of the "mid- dle way" that developed under Sukarno's "Guided Democracy."31 An under- lying principle of the dual function concept is that national unity and political stability are preconditions for economic growth. The role of the military, broadly conceived, has been to protect political stability and actively to pre- vent its deterioration-an umbrella that has covered a wide range of activities. 40 The Microfinance Revolution: Lessons from Indonesia Under Soeharto, politics and the military were closely linked.A crucial as- pect of Soeharto's rule was his role as general, commander-in-chief, and pa- tron of the military.Another was his role in Golkar (golongan karya, or functional groups), the political arm of his regime. Golkar has been, in practice, controlled by the mlihtary-but both were controlled by Soeharto. The government. Under the constitution, the president, who serves as head of state and government, and the vice president are elected to five-year terms by the People's Consultative Assembly (Majelis Permusyawaratan Rakyat, or MPR). The MPR, which must meet at least once every five years, is also re- sponsible for reviewing and approving the government's broad policy goals.32 During the Soeharto era, MPR decisions were consistently made by (Soeharto- The military has a orchestrated) "'consensus.." Until the post-Soeharto reforms discussed later in this chapter, the 1,000- dualfunction: to member MPR consisted of the 500 members of the legislature and 500 rep- resentatives from the regions, from political organizations, and members of Golkar protect political from the armed forces.The 500-member legislative assembly, the Dewan Per- wakilan Rakyat (DPR), consisted of 425 elected members and 75 members ap- stability and to pointed by the president. The Indonesian judicial system administers three kinds of civil law: Shari'a, prevent its or Islamic law, for Muslims; a civil law based on Roman law for Europeans; and a collection of statutes, not yet codified, for other Asians such as ethnic Chi- deterioration-an nese and Indians.The same criminal law applies to all. As discussed below, however, a striking feature of Indonesian governance umbrella that has has been the wide gap between form and practice. Illustrating theJavanese con- cept of power, both the legislature and the judiciary were, in practice, under covered a wide range the direct control of President Soeharto, as was the MPR. An important, but often overlooked, aspect of governance in the Soehar- of activities to period was the substantially increased role Indonesia came to play in regional and international peacekeeping and cooperation. Under Soeharto, Indonesia became a strongly stabilizing force in Southeast Asia. During the more than 30 years of New Order government the country fought no international wars33- more than can be said for the others among the world's four most populous nations (China, India, and the United States). Moreover, Indonesia has played an important role in the Association of Southeast Asian Nations (ASEAN), in the development of the worldwide Non-Aligned Movement (NAM), and more recently in the Asia-Pacific Economic Cooperation (APEC) forum. The military. Under the doctrine of dwifungsi, or dual function, the armed forces (Angkatan Bersenjata Republik Indonesia, or Abri) play a major role at every level of society. From 1983 to 1998 between one-quarter and one-third of the seats in Soeharto's cabinets came from the military. Many village, sub- district (kecamatan), and district (kabupaten) heads were also from the military, as were many of the provincial governors who are appointed by the central government. An Introduction to Indonesia 41 Soeharto came from the army. He was born in 1921, the son of a village official in a poor village in Central Java. His parents, who divorced soon after he was born, both remarried.While growing up Soeharto moved among var- ious extended family households. He attended junior high school but had to leave because his family could not afford the expense. But he later attended a school inYogyakarta run by Muhammadiyah, a large Islamic organization, fin- ishing his studies when he was 18. Soeharto enlhsted in the Royal Netherlands Indies Army, but shortly afterwards the Dutch surrendered to the Japanese. He then enlisted in the Japanese-sponsored army of Indonesia; when the Japanese left Indonesia, he became an officer in the Indonesian revolutionary army. After independence Soeharto made his way up in the army and was eventually ap- Illustrating the pointed head of the army's Strategic Reserve Command-the base from which he assumed control of the army on 1 October 1965, following the attempted Javanese concept of coup. As president, Soeharto strongly supported the armed forces. power, Soeharto Soeharto came to power on the army's coattails and the repressive might of the army has been the single most important factor controlled the in undermining potential opponents throughout his tenure.The army has forcefully suppressed any number of legislature,judiciary, demonstrations ... But open shows of force are not really the New Order's style. Soeharto's-and the armed force's-objective is People's social control, not military control ... The military has tried to prevent unwanted political activity rather than rely on repression Consultative once that activity has appeared. Assembly, military, A second important feature of Soeharto's rule is his frequent and shrewd use of patronage to buy off critics particularly from within media, and Abri. . .As the New Order progressed so did the art of patronage. Revenues collected from Soeharto's close business associates in opposition parties sectors such as oil, construction, and agro-business-often washed through non-profit foundations-have enabled Soeharto to expand the distribution of patronage to potential critics in political, religious, and social circles. -Schwarz 1994, pp. 39-41 During the first half of the New Order government the army and Soeharto were practically synonymous (Schwarz 1994, p. 282). From the early 1980s on, however, the relationship grew more distant-especially in the 1990s with the escalating corruption of the Soeharto family and the rise of huge private wealth in corporate conglomerates. As the conglomerates grew in the 1990s, the influence of the armed forces (and the patronage it received) declined some- what. Nevertheless, Soeharto maintained control over the military, both through patronage and by continually shifting the appointments of its officers. The Indonesian military is not a monolith; its members hold a range of views. Some supported the preventive approach to political disruption as imple- 42 The Microfinance Revolution: Lessons from Indonesia mented, while others thought Soeharto took prevention too far. By the mnd- 1990s many members of the armed forces said they thought it was time for Soeharto to leave office (Schwarz 1994, ch. 10). Some, especially retired gen- erals, played an important role during the 1990s in criticizing Soeharto's ex- cesses and his growing unwllhngness to listen to others. As retired general Sayidiman Suryohadiproyo put it, "Soeharto no longer listens to anyone, not Abri nor anyone else.This is the danger we are facing" (quoted in Schwarz 1994, p. 284). The politics. During most of the Soeharto era Indonesia had a two-party political system-but both were opposition parties.What they were opposing was technically not a party but a government-sponsored coalition of groups, Indonesia was Golkar.34 Starting with an army-organized association of anticommunist func- tional groups known as Golkar, Soeharto announced in 1967 that Golkar micromanaged would become his parliamentary vehicle. The original groups soon lost their influence within the organization as Golkar came to be controlled by the mil- through a mixture of itary, with active support from the civil service and its civilian members. As Schwarz (1994, p. 31) comments, "The mass-based groups present at Golkar's top-down control, birth faded into insignificance,just as the masses disappeared from the politi- cal scene more generally." largesse, religious The military-backed Golkar quickly became the dominant political or- ganization in the country, also controlling the opposition parties.35 Seman- and ethnic tolerance, tics aside, Indonesia became essentially a one-party state. But the political power of Golkar was not held by the organization or by the army-it was held by military repression, Soeharto.36 From 1973 until after Soeharto's resignation 25 years later, only two other political parties were permitted. The United Development Party and economic (Partai Persatuan Pembangunan, or PPP) was formed in 1973 by the merg- er of four Islamic parties, and the Indonesian Democratic Party (Partai development Demokrasi Indonesia, or PDI) was formed the same year from five Christ- ian and nationalist minority parties. Several political groups (not parties) were also officially recognized. However, the political opposition was carefully controlled by Soeharto and Golkar. Until Soeharto's resignation the two opposition parties continued to serve largely as actors in the stage-managed general elections and MPR's pres- idential elections. Government officers and the military were required to vote for Golkar. Civilian voters were coerced, threatened, and paid to vote for Golkar-an approach that routinely produced the large majorities Golkar pre- dicted beforehand. The MPR, most of whose members were appointed or approved by Soe- harto, met every five years. Soeharto was elected seven times by unanumous ac- clamation, and his policies were routinely endorsed. As Juwono Sudarsono, minister of education in President Habibie's cabinet and minister of defense in President Abdurrahman Wahid's cabinet,37 put it, "Every five years 1,000 peo- ple met in a congress and there was not one who was brave enough to stand up and say, 'Enough"' (quoted in The New York Times, 13 June 1999, p. 3). An Introduction to Indonesia 43 Under Soeharto a large, complex, and pluralistic society was micromanaged through a carefully orchestrated mixture of top-down control, largesse, religious and ethmc tolerance, military repression, and economic development. Soeharto stood at the apex of the pyramid; his appointees sat in each of the key executive, legislative, and judicial branches of government. He dominated the cabinet and the state bureaucracy. He dominated the armed forces.. .He dominated the People's Consultative Assembly... He was the central figure in Golkar, the army-backed political movement, and had crippled the effectiveness of the two "opposition" political parties. He The Soehartos were dominated the judicial branch, weak as it was, and had stacked its key positions with long-time associates, all of them generals. involved in He appointed the men who sat in the Supreme Advisory Council... His writ extended into every department and into corruption "in every every state-run corporation; it reached down, if he chose, to every village. He wielded enormous power both because the 1945 form of state action Constitution confers enormous power on the president and because he was the dominant influence over the army, itself the that had an dominant force in society. In short, he had established himself as the paramount figure in a society in which deference to economic value." authority is deeply rooted. -Jenkins 1984, pp. 13-14 Indonesia was It has also been argued, however, that the authoritarian nature of Soehar- known to's political regime made possible many of the policles that promoted the coun- try's economic growth and development. internationally as a The [authoritarian] nature of the regime made possible kleptocracy macroeconomic policies marked by moderation and continuity, encouraging savings and investment. It also permitted policies that favored long-term development in regard to oil, agriculture, and industry, which, over time, and on balance, were beneficial economically and socially. -Bresnan 1993, pp. 292-93 But the strong leader-weak state syndrome required that the country's un- derdeveloped judicial, legislative, political, and financial systems stay that way. And no implementable process for succession to the presidency could be de- veloped while Soeharto remained president. "Mr. Suharto systematically neutered the most promising of his subordinates" (Seth Mydans in The New York Times, 6 May 2001, p.3).38 As Soeharto intensified his control, and as corruption mounted, cracks began to appear and then widen in the country's support for its president.This took place increasingly in the mulitary, in religious and ethnic groups, and among 44 The Microfinance Revolution: Lessons from Indonesia members of the media, students, intellectual leaders, civilian elites, and others. When the East Asian crisis hit Indonesia in 1997-sharply raising unemploy- ment, prices, and poverty-the growing dissatisfaction spread to ordinary peo- ple as well. Corruption and the growth of the conglomerates Corruption was an intrinsic part of Soeharto's rule. An outgrowth of the au- thoritarian nature of the regime and its lack of transparency and accountabil- ity, corruption took both public and private forms-though the distinction was often blurred.There was the vast private wealth amassed by the Soeharto fam- ily and their friends-through monopolies, subsidies, tax and credit benefits, and the like-at the expense of the state, of other firms, and of the Indonesian The 1 990s saw the people.The first family had major interests in virtually every sector of the econ- omy: oil and gas, shipping, banks, mmning, lumber, agroindustry, automobiles, rapid growth of huge toll roads, communications, and others.There was also massive funding for Golkar and its activities.These funds provided the means for the extensive patronage corporate that kept Soeharto in power. conglomerates owned The rise of the conglomerates. The 1980s and 1990s saw rapid growth of huge corporate conglomerates owned by the Soehartos and their friends, and by the Soehartos protected by the Soeharto regime. Closely related was the issue of the ethnic Chinese in the conglomerates. In return for their business skills, financial ex- and theirfriends perience, and investments in the first family's ventures, some ethnic Chinese were given the monopolies, subsidies, licenses, and other favors that enabled andprotected by the them to compete unfairly with most indigenous firms (see box 8.1). The scale of corruption was massive, pervasive, and continually escalating Soeharto regime (Bresnan 1993; Schwarz 1994; Cole and Slade 1996; The Economist, 30 March 1996,pp.41-42; The NewYork Times, 31 October 1997, p. C-7)."The [Soehar- to] regime was immersed in corrupt practices in the granting of licenses, lend- ing of funds, letting of contracts, and every other form of state action that had an economic value" (Bresnan 1993, p. 292). By the 1990s the soaring growth in the private wealth of the few favored families-combined with decades of inaction in domestic political develop- ment-raised, both domestically and internationally, increasing concerns about the country's economic and political stability. Because the domestic press and other media were controlled by the government, they were not permitted to report or comment directly on high-level corruption or on many aspects of domestic politics. Some members of the media, however, became masters at writing or speaking between the lines and expressing opinions through analo- gies. Some wayang puppetmasters used the foibles and heroics of ancient kings, queens, courtiers, warriors, and commoners to provide indirect com- mentary on the current scene. It was difficult for the government to object to a performance of Rama teaching the principles of rule toWibisana or other traditional wayang stories-and the messages being sent reached widespread popular audiences. An Introduction to Indonesia 45 The international media, however, increased their forthright criticisms of the Soeharto family and the conglomerates during the late 1980s and the 1990s. Articles critical of the Soeharto family and rule were banned in Indonesia dur- ing the 1970s and 1980s, but the ban became difficult to enforce during the 1990s (because there was so much criticism), and censorship relaxed somewhat. By 1997 there was hardly an international newspaper or political or econom- ic periodical that did not report high-level corruption in Indonesia. The in- ternational press referred to Soeharto's Indonesia as a kleptocracy (The Boston Globe, 2 June 1999, p. 26). The technocrats and the conglomerates. Struggles between the technocrats Soeharto was elected and the conglomerates occurred on numerous occasions, with Soeharto bal- ancing one against the other.The role of the technocrats in attempting to con- seven times by tain the conglomerates was widely recognized. As Bresnan (1993, p. 282) put it, "They [the technocrats] provided Soeharto with two essential ingredients acclamation. "Every to econonuc policy: the confidence of the international financial commumty, and policies and programs that produced demonstrable results." five years 1, 000 But powerful political and economic interests repeatedly sought to gain con- trol of financial institutions and resources. One of the most important contri- people met in a butions of the technocrats in the Indonesian government, often with support from international financial institutions, was to contain these special interests. congress and not one The technocrats were particularly skillful in maintaining their basic principles and using these to frustrate repeated attempts to circumvent financial controls was brave enough to and manipulate the financial system (Cole and Slade 1996, p. 355). While the technocrats tended to be able to get Soeharto's approval for their stand up and say policies in times of economic difficulty, ironically their influence diminished during periods when the economy was strong and financial resources were avail- 'Enough"' able for distribution.The technocrats managed to contain the conglomerates for many years. But in the mid-1990s the conglomerates' influence on Soe- harto surged, while that of the technocrats declined. Indonesia on its 50th birthday, 1995 At the time of its 50th birthday in 1995, Indonesia had accomplished much in a relatively short time and in a large and geographically diverse country with multiple languages, ethmc groups, religions, and cultural traditions. Sukarno had led the struggle for independence, unified the country, and put Indonesia on the world map. Until mid-1997 Soeharto presidency's produced strong eco- nomic growth, widespread development (including increased food production, education, health, and family planning), and a substantial decrease in poverty. The country's regional and international responsibilities had increased substantially, and Indonesia provided a stabilizing force in Southeast Asia. Yet at the 50th birthday celebration there were severe underlying problems and deep fissures in the political economy.The corruption of the Soeharto fam- ily and their friends, the president's control over the legislative and judicial branch- es, the political process, and the media, as well as the concerns-both at home 46 The Microfinance Revolution: Lessons from Indonesia and abroad-about presidential succession had led to increasing unease and mounting opposition to Soeharto. Indonesia had never seen a peaceful transi- tion in leadership. Power-sharing was not its strength, and there was little preparation for a transition to a new president. Whatever Indonesians as individuals might have thought about the Javanese concept of power, there was a growing view that, in any case, the Soeharto fam- ily was not behaving in a manner befitting a holder of power in Javanese cul- ture.A 1990 play about a fictional king and his four children who accumulated enormous wealth was closed by police order on its 1th day of playing to stand- ing-room audiences in Jakarta-because the government found it to be anti- Pancasila and a threat to security (Schwarz 1994, p. 232).The play, Suksesi, by Nano Riantlarno, ends with the following chorus: Some puppetmasters Don't show what can be obtained used the wayang to so that the people won't want it. The king always tries to keep his subjects provide indirect empty in heart, full in stomach, commentary on weak in desire, but strong in bones. current events. It was The king always tries to make sure that the people don't know dfficult to object to and those who know don't dare, Rama teaching the let alone act. And in consequence principles of rule all will be orderly and stable, orderly and stable...39 By the nation's 50th birthday in 1995, many in Indonesia thought that Soe- harto's continuing rule would not be in the interests of the Indonesian people. The Indonesian Crisis and the Resignation of President Soeharto It was in this context that the Southeast Asian crisis that began in July 1997 hit Indonesia.40 The Thai baht collapsed, the currencies of neighboring countries came under pressure, and a regional crisis with global ramifications developed rapidly. It soon extended well beyond Southeast Asia and became known as the East Asian crisis. Banks and corporations.. .borrow[ed] large amounts of international capital, much of it short-term, denominated in foreign currency, and unhedged. As time went on, this inflow of foreign capital tended to be used to finance poorer-quality An Introduction to Indonesia 47 investments. Although private sector expenditure and financing decisions led to the crisis, it was made worse by governance issues, notably government involvement in the private sector and lack of transparency in corporate and fiscal accounting and the provision of financial and economic data. Developments in the advanced economies, such as weak growth in Europe andJapan that left a shortage of attractive investment opportunities and kept interest rates low in those economies, also contributed to the buildup of the crisis. After the crisis erupted in Thailand with a series of speculative attacks on the The Thai baht baht, contagion spread rapidly to other economies in the region that seemed vulnerable to an erosion of competitiveness after collapsed, the the devaluation of the baht or were perceived by investors to have similar financial or macroeconomic problems. currencies of -IMF 1999e,p.1 neighboring The Indonesian crisis-which turned out to be the most severe in the region-erupted shortly after the collapse of the Thai baht.The primary cause countries came under of the crisis in Indonesia was not the underlying economy. It was the closely linked weaknesses in governance and in the financial system. pressure, and a The nature and extent of the crisis regional crisis with The value of the Indonesian rupiah plunged from 2,450 rupiah to the U.S. dol- lar at the end ofJune 1997 to 14,900 rupiah at the end ofJune 1998 (see ap- global ramifications pendix 1 and table 8.1). Most of Indonesia's corporations became technically bankrupt. Banks failed.And the financial system collapsed. Unemployment in- developed creased, and inflation was 53.4 percent for 1998. Even with its strong economnc base and long record of high economic growth, Indonesia could no longer at- tract foreign capital. GDP, which had grown on average by nearly 8 percent a year for a decade, grew by just 4.7 percent in 1997 and contracted by 13.0 per- cent in 1998. In addition to the financial crisis, in 1997 Indonesia faced its worst drought in 50 years.The resulting food shortages, hoarding of food by merchants, and soaring food prices occurred at a time when many farmers' incomes had fall- en sharply.A large portion of dwindling foreign reserves had to be used to im- port food, and food shortages continued into 1998. Riots over the high prices of food and other basic commodities broke out in various parts of the coun- try, often targeting ethnic Chinese. At the same time, there was a sharp decline in international oil prices. In addition, a major new environmental problem developed-one that was directly related to corruption issues. Huge forest fires, occurring during the drought in Sumatra and Kalimantan, resulted in massive haze that caused widespread air pollution, respiratory illness, ecological destruction, and economic hardship in Indonesia. These fires affected other parts of Southeast Asia as well, including 48 The Microfinance Revolution: Lessons from Indonesia Brunei, Malaysia, Papua New Guinea, the Philippines, Singapore, and Thailand (see Seth Mydans, The NewYork Times, 26 October 1997, international section, p. 3). Many of these fires had been set to clear land for large palm oil and in- dustrial pulpwood plantations. Indonesia's coffee, palm oil, cocoa, and rubber crops were harmed by the fires and the smoke, while the reduced sunlight that resulted slowed the growth of fruits and vegetables and decreased yields of maize and rice in the affected areas.These fires also "burrowed deep into vast peat bogs and seams of coal where experts say they may continue to smoulder for years" (The NewYork Times, 26 October 1997, mternational section, p. 3). In 1998 and 1999 new fires were set in Sumatra and Kalhmantan for the same purposes, again causing widespread devastation. The value of the Well-connected palm oil plantation owners and pulp-and- paper companies in Indonesia have continued clearing land by rupiah plungedfrom burning off vast tracts of jungle, seemingly immune to laws or punishment ... Indeed, the palm-oil producers, who have set 2,450 to the US. most of the fires, may be one of the few beneficiaries.They have cleared huge new areas for planting, and as the disaster has dollar inJune 1997 spread, palm-oil prices have risen sharply on the world market. -New York Times International Section, 28 August 1999, p. B2 to 14,900 inJune Shortly after the crisis began in mid-1997, unemployment and prices rose 1998 sharply. During 1997 and 1998 the government took steps to head off antici- pated reverses in poverty reduction. Public employment schemes were insti- tuted. The government guaranteed deposits in Indonesian banks. Government-subsidized credit schemes for small loans were expanded (with predictable results). An emergency $42.3 billion financial aid soft loan package for Indonesia was developed by the IMF,World Bank, Asian Development Bank, and sever- al bilateral contributors. The January 1998 IMF-supported economic reform program contained measures designed to eliminate cartels and monopolies, to remove subsidies and special privileges from some of the gigantic government- backed projects controlled by the country's wealthiest familles, and to increase transparency generally. But when the agreement was signed, there were wide- spread doubts throughout Indonesia about whether these measures could be implemented. Serious implementation problems and delays did occur, result- ing in postponed IMF payments on a number of occasions. Soeharto, becoming increasingly confrontational, strongly resisted reforms that would harm the business interests of his family and friends. But the IMF know little about the Indonesian economy and its institutions.This unfortu- nate combination exacerbated the already full-blown crisis in Indonesia. Soeharto had served six consecutive five-year terms as president when, in March 1998, he was "elected" to a seventh term. But the crisis continued to worsen, and it was increasingly believed that power, which Soeharto had held for so long, had deserted him.When the IMF agreement was signed, a picture An Introduction to Indonesia 49 of Soeharto and Michel Camdessus, managing director of the IMF, was wide- ly distributed through the Indonesian media.The picture showed Soeharto seat- ed, signing the agreement. Camdessus stood by his side, arms folded in front of him. This picture was widely interpreted in Indonesia as a sign that power had left Soeharto. (Some thought that Soeharto's power had come to him through his wife, IbuTien Soeharto, and that power had left him at her death in 1996). But by early 1998 it was widely agreed that Soeharto, although still president, no longer held power. In May 1998 corruption, the collapse of the financial system, soaring in- flation, sharply contracting GDP, capital flight, business closures, withdrawal of international investors, escalating poverty, the unresolved succession issue, and The IMF increasing violence and social disorder brought a crisis of confidence in the gov- ernment and forced Soeharto's resignation after 32 years in office. He resigned established a $42 on 21 May 1998 and was immediately succeeded by his longtime friend and protege,Vice President B. J. Habibie, a German-trained aeronautical engineer billion soft loan who had served as minister for research and technology from 1978-98. package that The roots of the crisis The economic, financial, and political components of Indonesia's economic mandated economic downturn were mutually reinforcing.The roots of the crisis ran deep. In an August 1998 speech Ah Wardhana, economic adviser to the govern- reforms. But serious ment and former long-term finance minister, put it this way:"Let me be quite blunt. Much of the economic crisis had its roots in the serious and fundamental implementation weaknesses of our financial system. Only after these have been tackled in a forth- right manner can we look forward to a serious and sustained recovery" (Ward- problems in reforms hana 1998b, p. 1). Wardhana's analysis of the basic weaknesses of the Indonesian financial sys- delayed tem is provided in box 8.2. "Our financial systems were encouraged to fund risky and unprofitable ventures. Government officials could, and did, direct loans disbursements to favored firms or activities" (Wardhana 1998b, p.2).As he shows, bureaucrat- ic interference, lending to favored firms, close links between banks and con- glomerates, and political pressure on bank regulators led to loans that were "rarely given even rudimentary economic and financial analysis." The result was a pat- tern of high-risk lending that was exacerbated by a widespread perception that banks were implicitly guaranteed by the government and by a rmispricing of foreign credits that contributed to the large capital inflows. Firms with sub- stantial foreign exchange exposure were vulnerable;"When we were forced to abandon our managed float, the depreciation of the rupiah created unmanageable debt burdens that effectively bankrupted a substantial portion of our corpora- tions" (Wardhana 1998b, p. 3). As the crisis unfolded, it became apparent that Indonesia had far more ex- ternal debt than had been publicly known-much of it owed by companies owned by the few wealthy, well-connected families discussed earlier. The for- eign debt of commercial banks had reached about three times the banks' eq- uity. Private corporations, banks among them, owed an estimated $68 billion 50 The Microfinance Revolution: Lessons from Indonesia Excerpts from Ali Wardhana's Box 8w2 "Overcoming the Current Economic Downturn" The economic meltdown, and one can hardly describe it differently, that has hit Indone- sia, had multiple causes. Some were self-inflicted, others were external...Let me be quite blunt much of the economic crisis had its roots in the serious and fundamental weak- nesses of our financial system Let me highlight only two First, our financial systems were encouraged to fund risky and unprofitable ven- tures In a banking system where state-owned banks play a significant role, bureaucratic interference is likely to be a serious problem Government officials could, and did, direct loans to favored firms or activities In addition the close links between banks and some of the conglomerates further reduced the likelihood that loans would be objectively evaluat- Iuch of the ed Even when government pressure was absent and when banks did not engage in in- tra-group lending, loans were rarely subjected to even the most rudimentary economic and financial analysis. In part, such analysis was handicapped by the absence of disclo- economic crisis had sure requirements and accounting standards that would allow analysts to make a reason- able estimate of risk And in part political pressure was exerted on bank regulators so that its roots in the they would not report some of the most flagrant violations of our banking laws In such a situation financial institutions fail in their most basic function to serve as an efficient intermediary, channeling savings to their most productive use. It is of course serious and true that all investments have associated risks But when savers, whether domestic or foreign, have no real capacity to evaluate the risks, the real cost of capital will be under- fundamental valued and the returns on investment overstated. As a result, scarce funds will be allocat- ed to low-return, high-risk activities Second, when neither investors nor lenders expect to bear the full cost of any fail- weaknesses of our ure, they will lower their guard against risky investments This is what is meant by the term "moral hazard " Moral hazard describes a situation where, in the presence of a per- financial system ceived implicit or explicit guarantee, there is little incentive to avoid risky behavior. It is true that the Government of Indonesia never extended any explicit guarantees against bank or corporate failures But it is also true that involvement of well-connected parties in many of our economic activities generated a feeling that, to quote a line from a popu- lar American movie, "failure was not an option here " Unfortunately, in the end, failure was very much an option here Actions by government and the central bank further en- couraged the belief that Indonesian banks would not be allowed to fail. Thus, when the government supported a recapitalization of Bank Duta in 1990, provided support to Bank Danamon to stem a bank run in 1991, and in 1994 made good the losses suffered by BAPINDO, it inadvertently suggested to all that banks would be protected from failure. As Paul Krugman recently noted, such implicit guarantees can trigger asset price inflation, reduce economic welfare, and ultimately make the financial system vulnerable to collapse 1 In a similar vein, our foreign exchange regime also encouraged risky behavior. Al- though Indonesia did not peg its exchange rate, as did some other Asian countries, we did maintain a managed float within a relatively narrow band Borrowers judged that the expected loss from currency depreciation was less than the cost of hedging their foreign borrowings For many years this proved to be correct The consequent mispricing of for- eign credits contributed to the very large capital inflows and created vulnerability for firms with substantial foreign exchange exposure. When we were forced to abandon our man- aged float, the depreciation of the rupiah created unmanageable debt burdens that effec- tively bankrupted a substantial portion of our corporations No one should doubt the severity of the current economic crisis Yet neither should anyone underestimate our capacity to set our institutions right The "Asian economic miracle" was not a mirage, it was real And many of the factors that allowed Indonesia An Introduction to Indonesia 51 Box 8.2 (continued) to grow by over 7% per annum for over a decade are still here. strong infrastructure, a well disciplined labor force, and ample natural resources These are the elements that will power our growth in the future But before that can happen we must rebuild our financial institutions so that they are capable of performing the functions that a modern economic state requires to mobilize capital and effectively channel it to its most productive use 1 Paul Krugman, 1998, "What Happened to Asia?" Unpublished research paper Source Wardhana 1998b. pp 1-3, 5-6 By early 1998 it was widely believed that Soeharto had in foreign debt in 1997, with more than half of it thought not to have been hedged. Under these circumstances the fall of the rupiah in 1997-98 led to mas- lost his power. In sive increases in foreign debt and debt service payments in rupiah terms, af- fecting the entire financial sector. By 1998 nonperforming loans (loans in May 1998 he default) accounted for about 70 percent of bank credit. Another factor contributing to the crisis was related to the sudden spurt resigned and was in the number of banks and bank branches that had followed the 1988 dereg- ulation liberalizing the creation of banks and bank branches. Excluding Peo- succeeded by Vice ple's Credit Banks (Bank Perkreditan Rakyat, or BPRs), the number of banks increased rapidly-from 129 in 1988 to 233 in 1994 (Cole and Slade, 1996, p. President Habibie 114).But there was a shortage of trained staff to supervise these banks.And even staff who had been trained were, de facto, often unable to supervise effective- ly (or even to obtain accurate information) because of the lack of transparen- cy at the banks, as well as the high-level political pressures on bank regulators to look the other way-as increasing numbers of banks provided loans to fa- vored firms and to their own conglomerates. This moral hazard problem was well known, but it was off lirmits for reform.Transparency requires accountability, high accounting standards, and accurate information about the financial con- ditions of banks and their debtors and creditors. Typically none of these was available at the banks. The problem was not that the weaknesses of the banking system were unknown. Rather the problem was the lack of political support for those who wanted to improve bank supervision and strengthen the economy's financial foundations.. .Bank licenses were often handed out to politically favored persons, further weakening supervision, and bank owners who were often industrialists, used their banks to lend to their own enterprises, often in violation of bank regulations. In the end, the economic/political interests controlling the banks were more 52 The Microfinance Revolution: Lessons from Indonesia powerful and better connected than the policy makers who were trying to shape policy in this area. -Stern 2000, p. 5 As noted, the Indonesian economic and financial crisis had been largely un- foreseen (although many had anticipated a possible political crisis arising from the growing concerns about corruption, civil rights, and succession issues). Lloyd R. Kenward, senior economist at the World Bank's Jakarta office (1994-98), analyzed the reasons for the Indonesian economic and financial crisis and "the warnings [that] were there,just below the surface" (1999, p. 71). He demon- strates that although the broadest macroeconomic indicators did not predict the crisis, extensive warning indicators of other kinds were recognized before Most of the huge the crisis (box 8.3). But as he shows, these warning signs tended to be viewed individually. Had they been analyzed collectively, they might have signaled the foreign debt (much likellhood of crisis before it occurred. Thus in 1996 and early 1997 the broad macroeconomic indicators for In- of it owed by the donesia were positive-robust economic growth, low and stable inflation, fis- cal accounts in surplus, adequate official foreign exchange reserves, and others. conglomerates) was But multiple warning signs had been identified, ranging from restraints on in- flation that were only temporary to weaknesses in the current account caused unhedged. With the by high international oil prices; from diluted deregulation packages to the coun- try's weak microeconomic foundations (banks and other corporations known collapse of the for structural weaknesses and risky behavior). And then there were Indonesia's euphemistically termed "issues of governance"-which commonly meant rupiah, the debt "corruption, cronyism, and nepotism." Had these and other trends been care- fully analyzed together, the possibility of an emerging crisis might have been became identified. Kenward also analyses the main indicators of the crisis that became clear unmanageable only with hindsight: the exchange rate risk, the extent of unhedged foreign ex- change positions, and personnel changes among high-level Indonesian officials that were "warnings to any government official who tried to enforce princi- ples of transparency and even-handedness in vital areas such as banking super- vision and privatisation. No matter how badly in need of reform, some areas were virtually untouchable" (p. 86).41 Readers who want to pursue in depth the reasons that the crisis was not predicted are referred to Furman and Stiglhtz (1998) for the Asian crisis gen- erally, and to Kenward (1999,2000) for the Indonesian crisis. Kenward also pro- vides lessons for practitioners "from the perspective of a macroeconomic practitioner close to the [Indonesian] instability as it was developing" (1999, p. 86).Among the types of lessons he presents are the Indonesian technocrats'pre- cept that "good times make for bad policies" (Kenward 1999, p. 86, quotes Ali Wardhana: "Reform is rarely if ever undertaken for its own sake"); the dangers of complacency, especially after decades of success; various ways in which mi- croeconomic weaknesses magnify macroeconomic shocks; and institutional prob- lems that can constrain prediction of a crisis. An Introduction to Indonesia 53 Box 83 Excerpts from Lloyd R. Kenward's "Assessing Vulnerability to Financial Crisis: Evidence from Indonesia" In 1967 per capita income in Indonesia was less than one-half that of India, Nigeria, or Bangladesh By mid 1997, it was five times that of Bangladesh, four times that of Nigeria and three and a half times that of India Virtually all of the country's social indicators had improved significantly, too. This was an extraordinary track record that put Indonesia in the top 10% of performers among developing countries Indicators that did not work... During the months and years before the financial cri- sis, virtually all of the broadest macro indicators were very reassuring . Economic growth was robust and driven by the private sector, inflation was low and quite stable, the fiscal "No matter how accounts were in surplus, the government was pre-paying large amounts of foreign debt, the growth of some monetary aggregates was high, but generally in line with inflation and strong, investment-led growth, foreign direct investment was strong; medium and badly in need of long-term debt service was declining in relation to exports (after allowance for debt pre- payments), official foreign exchange reserves looked adequate and were trending up- reform, some areas wards, and the deficit on current account of the balance of payments looked manageable The most controversial indicator in this context was external debt But even here, the overall situation was favorable.. .doubts arose only on examination of the breakdown be- were virtually tween public and private debt .Judging only on the broadest of criteria, the years imme- diately preceding the crisis were the best of macroeconomic times for Indonesia... [And] untouchable" if anything, the signals from sensitive, high-frequency (daily) financial market indicators were even more positive than the broadest macro indicators. Helpful Indicators. recognized wamings. [But some] analysts concluded [before the crisis began] that significant risks remained Inflation was being temporarily restrained by good luck on food prices and by delays in increasing certain administered prices the growth of monetary aggregates showed no clear tendency to decelerate Also, high inter- national oil prices (and probably data limitations) were masking weaknesses in the current account, dependence on potentially volatile short-term capital was increasing key non- tradeables (such as the property sector) were booming, and there were isolated signs of nervousness among foreign investors. The authorities even publicly acknowledged some of these indicators. [Kenward documents the recognition of each of these indica- tors before the onset of the crisis 1. While the macro data were giving off mixed signals at best, there was little doubt about the quality of structural policies. Many had been moving in the wrong direction for at least one year.. Recent deregulation packages had been weak and there was ample evidence that the deregulation drive-which had largely accounted for Indonesia's suc- cesses since 1986-had lost much of its momentum Moreover, issues of governance (the old euphemism for 'corruption, cronyism, and nepotism'. ) were drawing increasing levels of international attention Viewed individually, these slippages in structural policy seem unimportant But collec- tively, and considering the number of them, they raised the likelihood of crisis [There were] weak microeconomic foundations Indonesia's bank and non-bank corporate sec- tors have long been known for certain structural weaknesses and risky behaviour In the case of the banks, their fragile state was well-documented Indonesia's banking system would amplify any serious disturbance, spreading shockwaves throughout the economy, with destabilising feedback effects on the macroeconomy. Outside the banks, corporates were well known for their high leverage and risk-taking behavior. In many ways, there- fore, the Indonesian economy was badly positioned to absorb a major macroeconomic shock. Retrospective wamrngs. In addition to the warning signs documented above, there were a few signs of trouble that are only clear with the benefit of hindsight For example, 54 The Microfinance Revolution: Lessons from Indonesia Box 83 (continued) a standard measure of Indonesia's real effective exchange rate now suggests a modest appreciation for about two years before the onset of the crisis But the modestly appre- ciating trend (of only about 5% per annum, trough to peak) could not be spotted with any reliability until only a few months before the crisis broke, which was too late to be of much use. Another indicator that deserved more attention was the extent of unhedged foreign exchange positions, particularly among non-bank corporates that were borrowing directly off-shore and through banks on-shore With the full benefit of hindsight, there were several personnel changes among key Indonesian officials before July 1997 that constituted more subtle signs of trouble in the offing Individually, these incidents seemed to be of minor importance at the time But collectively and in retrospect they were warnings to any government official who tried to enforce principles of transparency and even-handedness in vital areas such as banking supervision and privatisation No matter how badly in need of reform, some areas were virtually untouchable None of the preceding discussion is intended to imply that the warning signs indicat- ed trouble of the order of magnitude of what eventually ensued Indeed the actual out- come exceeded the worst expectations Source Kenward 1999, pp 73-86 In the end Soeharto harmed not only the state, but also the economy he had built.Two overall commentaries on the Indonesian crisis are instructive. The Indonesian monetary crisis was triggered and exacerbated by the general collapse of financial prices in Southeast Asia. The breakdown in confidence and the disruption of normal financial flows within the region was not immediately recognised by Indonesian policy makers (nor by many others)... Indonesian banks attempted to replenish their reserves from sources offshore, but this time such funds were not available...'Normal' patterns of capital flows available to Indonesia since the early 70s, in the context of an open capital account and stable macroeconomic policies, had dried up. Domestic corporates also rushed into exchange markets to try to cover or hedge their foreign exchange obligations, and thus drove down the exchange rate. Panic by domestic individuals and businesses both drew from and fed back into foreign investors' panic, resulting in massive capital outflows from Indonesia and from the region. -Cole and Slade 1998, pp. 62-63 An Introduction to Indonesia 55 And: The macroeconomic framework constructed in the 1970s and 1980s may not have been entirely consistent but it served Indonesia well. But by the 1990s the structure of the economy had changed, new interests had emerged, and Indonesia had become more integrated with the rapidly growing world financial markets. At that time the old weaknesses of the macroeconomic framework, which had been mere annoyances early on, became serious liabilities.When the Asian crisis reached Indonesia, the macroeconomic framework proved too weak to Commercial banks' withstand the onslaught. -Stern 2000, p. 7 foreign debt reached The discussion below turns to the period from the end of the Soeharto era three times their in May 1998 to the early days of Megawati Soekarnputri's presidency, which began in July 2001. Soeharto's legacy, crucial for an understanding of Indone- equity. 70 percent of sia's current transitional period, includes three decades of extraordinary eco- nomic growth with equity and social development; a nation severely bank credit was in underdeveloped politically; a massive economic, financial, and political crisis- and the basic roots of the crisis themselves, many of them still deeply entrenched defaulted loans. The and very much alive. financial system Indonesia in Transition collapsed Many changes have taken place in Indonesia since the resignation of President Soeharto in May 1998. It is still too early to understand fully the dynamics, the complexities, and the horse trading of Indonesia in the post-Soeharto era.After Soeharto's resignation, immediate rehabilitation efforts focused mainly on im- proving governance, implementing political reforms, strengthening the econ- omy, rebuilding the financial system, and maintaining a safety net for the poor, whose numbers increased during the crisis. Results were mixed. Some reforms were accomplished that would have been unthinkable just a few years earlier. But many of the country's old problems did not disappear, some that had been buried resurfaced, and formidable new ones emerged. Discussion here focuses on some major events from mud-1998 through mid-2001, during the presidencies of B.J. Habibie (1998-99) and Abdurrah- manWahid (1999-2001). (See chapter 15 for discussion of efforts to restructure the financial system during this period).This account is followed by a brief discussion of the impeachment of President Abdurrahman Wahid by the People's Consultative Assembly (MPR), and the MPR's replacement ofWahid as president byVice President Megawati Soekarnoputri, Sukarno's daughter. These events occurred in July 2001, shortly before this volume went to press. 56 The Microfinance Revolution: Lessons from Indonesia The economy and the financial system, 1998-99 A set of economic and financial reforms was instituted as part of the econom- ic reform program financed by the IMF and other donors.These reforms in- cluded establishing a tight monetary policy to stem exchange rate depreciation, rescheduling banks' external debts, and establishing institutions to work out arrangements for the external debts of corporations. An Indonesian Bank Re- structuring Agency (IBRA) was created to restructure and recapitalize the banking industry (chapter 15). Monetary and banking reforms, along with the support from the IMF, helped the economy improve. The value of the rupiah relative to the U.S. dollar re- covered from 14,900 at the end ofJune 1998 to 8,025 in December 1998, and to 7,085 at the end of 1999 (see table 8.1 and appendix 1). Moreover, Indonesia A tight monetary (which is the only Asian member of the Organization of Petroleum Export- ing Countries, or OPEC) benefited from the soaring oil prices that began in policy, debt 1999. There was a significant turnaround in GDP growth-from -13.0 per- cent in 1998 to 0.3 percent in 1999. Inflation reached 78 percent at its high- rescheduling, and est point during 1998.Annual inflation was 53.4 percent for 1998 but dropped to 20.5 percent for 1999. IMF support helped As noted, Indonesia had an impressive record of poverty reduction during the Soeharto era-with the portion of the population below the poverty line the economy dropping from 60 percent in 1970 to 40 percent in 1976 to 11 percent in 1996 (Government of Indonesia 1996, p. 570). But during the crisis real wages and improve. But employment opportunities fell, and prices of food and other basic necessities rose. As a result some who had escaped poverty returned to it. In 1998 the share of economic recovery the population living on less than $1 a day was estimated at 20 percent. By the end of 1999, however, it had dropped to its pre-crisis, 1996 level of 11 percent has been slow, (but many "near-poor" people remain just slightly above the poverty level).And farmers who produce export crops (such as coffee, rubber, and palm oil) ben- largely for political efited from the booming export market that followed the fall of the rupiah. The government, with assistance from international donors, developed a reasons multipronged strategy for helping poor households during the crisis. Subsidized food, especially rice, was provided to many of the poor. Substantial employ- ment was created, much of it through labor-intensive public work programs. And social services, particularly for health and education, were preserved. The government instituted a "stay in school" campaign and provided scholarship pro- grams and block grants for schools. Social safety net programs were given high priority by the government and donors. But many of the country's problems remained intractable. Still constrain- ing the economy are difficulties with the restructuring of massive corporate foreign and domestic debt, the crippled banking system, and continuing large- scale corruption-exacerbated as before by the weak legal and judicial systems. Soeharto's family and cronies and their conglomerates, while weakened, are still wealthy and powerful. Many Soeharto supporters throughout the country re- main active, in some cases working to create and maintain political instabili- ties. Ethnic and religious tensions have grown, and separatist movements have An Introduction to Indonesia 57 accelerated. Investors have generally stayed away. The slow progress in In- donesia's economic recovery has been mainly a result of these and other relat- ed factors, rather than of weakness in the underlying economy. This volume went to press only a few months after President Megawati Soekarnoputri took office, and it is too early to know whether her government will be able to achieve significant economic improvement. Early indications of various kinds are dis- cussed later in the chapter. Governance and politics, 1998-99 Responding to mass demand after President Soeharto's resignation, President Habibie initiated a number of political reforms. Legislation concernung polit- Still constraining the ical parties and elections was revised, and restrictions on the formation of po- litical parties were abolished. General elections were held inJune 1999, and the economy are massive presidential election by the newly constituted MPR was held in October of the same year. Controls on the media were greatly relaxed. In another widely corporate debt, the publicized reform, Indonesia agreed to hold a referendum that would offer the province of East Timor a choice between independence and broad autonomy crippled banking within Indonesia. East Timor had been a Portuguese colony (covering half an island and surrounded by Indonesian territory) until it was annexed by Indonesia system, and large- in the mid-1970s. Issues of EastTimor's status and allegations of human rights abuses carried out there by other Indonesians and by the government had been scale corruption- in contention between Indonesia and many in the international community since East Timor first became part of Indonesia. exacerbated by the Under the Habibie reforms new political parties proliferated. Long restricted by a one-party state with just two powerless and tightly controlled opposition weak judicial system parties, Indonesians explored their new options, fielding 48 parties for the June 1999 general elections.The front runners were the incumbent, B.J. Habibie, a lifelong Soeharto prot&ge and supporter but also one of the mostWesternized of Indonesia's political leaders, and Megawati Soekarnoputri, Sukarno's daugh- ter and the most visible challenger to Soeharto while he was in office. Megawati had been able to mount an open opposition to Soeharto during the last years of his rule largely because she is the daughter of the country's founding pres- ident. Soeharto's ability to control her opposition was limited, although he ma- nipulated her removal from the officially recognized opposition party she led while he was president. She contested in the 1999 general elections as head of the Partai Demokrasi Indonesia-Perjuangan (Indonesian Democratic Party- Struggle, or PDI-P).42 PDI-P won a plurality with 34 percent of the 105 million valid ballets cast. Golkar, with President Habibie as its candidate, placed second with 22 percent of the vote. The remaining votes were divided among numerous parties, lead- ing to months of coalition-forming negotiations, payoffs, battles, and deals that paved the way to the October 1999 presidential vote. After decades of suppression, people spoke their minds during the campaigns. John Bresnan's comment in June 1999 about the forthcoming presidential elec- tions was widely repeated: "I think there is a lot of sense of irony that Golkar 58 The Microfinance Revolution: Lessons from Indonesia is seen as the status quo party but is headed by one of the most Westernized figures in Indonesian political history, whereas PDI-P, the party of reform, is led by a lady solidly rooted in the feudal past" (quoted in The New York Times, 22June 1999, p.A3). But a well-known political cormnentator,WitmarWitoe- lar, when asked about criticisms that Megawati nught not be smart or knowl- edgeable enough to become president, quipped,"We already have a rocket scientist as President and look where we are now" (quoted in The New York Times, 13 June 1999,p. 3). The 1999 presidential election was held under a reformed system in which a 700-member MPR was composed of the 462 members of the legislature elect- ed in the June 1999 general elections, 38 members of the legislature appoint- ed by the military, and 200 government-appointed MPR members from Responding to mass regional and minority groups. Megawati and Habibie remained the frontrun- ners. Habibie, while tainted by his long and close association with Soeharto, demand, President had brought about considerable reform in his short time as president. Megawati had no government experience (like nearly everyone in Indonesia outside Habibie abolished Golkar), but she is the daughter of the country's founding president, she had stood up to Soeharto, and she was (and is) popular with voters. restrictions on The newly constituted MPR convened injakarta in October 1999. It first rat- ified by consensus the landslide 30 August 1999 vote in which 87.5 percent of political parties, held the people of East Timor had voted for independence-a vote that had been fol- lowed by the worst massacre in Indonesia since 1965. Indonesian militias with ties general elections, and to the military opposed the independence of East Timor, resorting to burning, looting, and indiscriminate killings of EastTirmorese. More than 1,000 people were relaxed media killed, about 70 percent of East Timor's buildings were destroyed, and more than 200,000 East Timorese were forced into camps controlled by the militias in the controls Indonesian province ofWest Timor on the same island. Order was restored only with international intervention. East Timor became a Umted Nations protectorate temporarily, but it is in transition to become an independent nation. The MPR's next order of business was President's Habibie's formal ac- countability speech in which he reported on his presidency. After days of dis- cussion, the MPR voted by a close margin-just hours before the 20 October 1999 vote for the presidency-to reject the president's accountability speech (the first such event in the history of the republic).The rejection in effect ended Habible's candidacy for president, and he withdrew from the race. Then, in Indonesia's first democratic election since 1955 and its first trans- fer of power through a vote of democratically elected representatives, the MPR elected as president Abdurrahman Wahid-a widely respected moderate Mus- lim religious leader with a reputation for being tolerant and incorruptible.The long-time head of Indonesia's largest Muslim organization, Nahdlatul Ulama with more than 30 million members,Wahid was known for his attitudes of tol- erance and inclusion, for supporting human rights, and for championing the rights of non-Muslims, ethnic Chinese, and other rminorities. But he is nearly blind and had been weakened by strokes. Wahid (who is widely called by his nickname Gus Dur) received 373 votes to Megawati Soekarnoputri's 313. An Introduction to Indonesia 59 Wahid's Muslim-based Partai Kebangkitan Bangsa (PKB), or National Awak- ening Party, had received 13 percent of the popular vote in the June 1999 gen- eral elections. He had then supported Megawati's campaign, but broke with her in October and pushed ahead with a bid for the presidency. Few took his can- didacy seriously until the last moment when Habibie withdrew his candidacy and other parties added their support to that of the National Awakening Party to elect Wahid president as a compromise candidate. On 20 October 1999, as Wahid rose to take the oath of office, President Habibie, who had received a vote of no confidence in the same room just hours before, took the new president's arm and helped him to the podium.The next dayWahid nominated as vice pres- ident Megawati Soekarnoputri, who was elected after last-minute, closed-door The MPR held maneuvers convinced challengers from Golkar and the army to withdraw. President Habibie had served as president for 17 months.Too closely iden- presidential elections tified with the Soeharto regime to survive democratic elections, he neverthe- less brought more reform to Indonesia in his short time as president than many in 1999.It voted had thought possible. not to accept President Abdurrahman Wahid, 1999-2001 In October 1999 AbdurrahmanWahid became Indonesia's fourth president, elect- President Habibie' ed in large part because of his ethnic and religious tolerance; his long record of support for democracy, clean governance, and human rights; and his repu- accountability tation for honesty and incorruptibility. But, as president, Wahid soon became widely perceived also as mercurial, unfocused, and unorganized.Wahid him- speech, effectively selfjoked,"The Suharto era here was known as the 'New Order,' the post-Suhar- to era led by BJ. Habibie was known for being 'Out of Order,' and the Wahid ending his candidacy period is known as 'No Order," (The New York Times, 3 October 2000, p.A31). But, unlike his many widely quoted witticisms, this one was not a joke. Wahid had to build a multiparty coalition governrment-which had not been seen in Indonesia for over 40 years. His first cabinet (known informally as Gus Dur's mixmaster cabinet) was widely inclusive, although of course many of its members had little government experience. Concerns focused on whether the new coalition government could hold together the Indonesian nation, control the army, restore the economy, build a viable financial system, implement the new president's ideas of honesty and tolerance, and build the institutions need- ed for a viable democratic government. Unlike the economic crisis, which had not been predicted before its onset, by early 2000 it was widely anticipated thatWahid's new coalition government, riven by internal politcal rivalries, would not make substantial headway in meet- ing these extremely difficult and complex challenges.The predictions were cor- rect. But President Wahid made some crucial contributions to Indonesia-of which the most important was to move the country toward democracy in ways that, this time, may be irreversible. Wahid's 21 months as Indonesia's president are discussed first.Then three of the major issues that dominated that period are examined: bringing the Soe- hartos to justice, the secession issues, and the economy. 60 The Microfinance Revolution: Lessons from Indonesia Early reforms and obstacles. PresidentWahid made some crucial reforms early in his administration.The media, tightly controlled under Soeharto, had been liberalized by Habibie. But it was Wahid who made the media fully indepen- dent.Although the Indonesian press suffers from a shortage of trained journalists, it-and the other media-are now playing a major role in leading the way to new openness and transparency in the country. In one of his first acts after becoming president,Wahid placed the armed forces under civilian control (where they had been in the 1950s). He later sus- pended army chief GeneralWiranto (who had been Soeharto's last military com- mander, continuing under Habibie) from Wiranto's cabinet position as coordinating minister for political affairs and security. Along with four other high-ranking generals,Wiranto had been found at fault by an Indonesian gov- In Indonesia'first ernment panel and the United Nations for the violence and destruction in East Timor that followed the vote for independence there.Wiranto resigned his cab- transfer of power inet position in May 2000. Wahid also reshuffled military officers, promoting reformers. And he through a vote of pledged to hold the military accountable for past and present human right abuses. Indonesian investigators began questioning generals about the role democratically elected of the military in the burning, looting, and killing that took place in East Timor in 1999. In May 2000 a landmark human rights tribunal convicted representatives, the 24 Indonesian soldiers and one civilian of massacring 57 villagers during a 1999 operation mounted against rebels in the special region (daerah istime- MPR elected wa) of Aceh in northern Sumatra. But despite PresidentWahid's early reform efforts, many deeply entrenched Abdurrahman problems remained, and new ones arose rapidly.Wahid was unable to consol- idate power. Although he pushed hard for institutional reforms, he had neither Wahid as president the political backing nor the administrative skills to achieve significant progress in the desperately needed legal,judicial, legislative, financial, and other reforms. And the obstacles the new president faced were formidable. In sharp contrast to Mr. Suharto.. .Mr.Wahid's ability to enact his policies is dangerously limited. As he built his top-down structure of personal control, Mr. Suharto so eviscerated the nation's institutions that his successors have inherited a hollow and nearly useless bureaucracy."The government here is almost hydroponic," said aWestern diplomat-a floating structure with dangling, ineffectual roots." -Seth Mydans, The New York Times, 3 June 2000, pp.Al, 4 And Daniel S. Lev, an expert on Indonesia, put it this way: He [Wahid] is a president without the institution of the presidency. Every major administrative institution and political institution has essentially been destroyed, crushed.There is really nothing left to work with. It doesn't matter, in a sense, how good An Introduction to Indonesia 61 people are that you appoint because the institutions are devastated. There's corruption everywhere. -Quoted in The New York Times, 3 June 2000, p. A4 Under PresidentWahid, constitutional amendments were drafted in prepa- ration for presentation to the MPR in August 2000. Proposed amendments in- cluded direct presidential election, the creation of a two-house legislature, a system of judicial review, and decentralization of some powers to the provinces. But Wahid did not get far with these amendments. Political infighting, violence, bombings, growing secessionist movements, economic problems, and the in- stability promoted by Soeharto supporters, Muslim fundamentalists, and oth- Wahid, a respected ers took center stage.Wahid remarked,"even a hundred new presidents" won't solve Indonesia's problems (The New York Times, 30 April 2001, p.A8). moderate Muslim Constitutional amendments on governance, and institutional reforms gen- erally, were indefinitely postponed. Thus the viable institutions needed for religious leader, was long-term democratic governance were not yet under construction.Wahid was unable to gain sufficient power to bring about the democratic government of elected largely shared power in which he believes. The strong ruler had gone, but the weak state remained. because of his The president and the parliament. Wahid and Megawati were democrati- democratic views, cally elected as Indonesia's president and vice president. But Indonesia now also had a democratically elected parliament (Dewan Perwakilan Rakyat, or tolerance, and DPR)-and little experience on how the two branches of government could work together. Wahid's small National Awakening Party had little say in par- honesty. But he liament. Megawati's party, however had won a plurality in the general elec- tions and was well represented in parliament. But a rift grew between Wahid could not govern the and Megawati. As noted,Wahid's first cabinet was a broad coalition that represented com- country promises with multiple political parties. But political rivalries among cabinet members, combined with the lack of organization and focus of the president and the distance kept by the vice president, essentially paralyzed the govern- ment. In August 2000, parliament-which is dominated by the Indonesian De- mocratic Party (PDI-P; the vice president's party) and Golkar-demanded that Wahid improve his erratic and uncoordinated leadership. In response, Wahid formed a new cabinet composed mainly of people from his own Na- tional Awakening Party (which controls only about 11 percent of the seats in parliament) and from the military.This cabinet, he said, was composed of pro- fessionals who would concentrate on results, not politics. But, if anything, po- litical lines hardened. For example, despite PDI-P's power in parliament,Wahid did not appoint a single member of the PDI-P to his second cabinet. Megawati, officially in charge of cabinet affairs, did not attend the announcement of the new cabinet. By April 2001, 18 months after his election, Wahid had been twice cen- sured by parliament and had come under widespread criticism for arrogance, 62 The Microfinance Revolution: Lessons from Indonesia unpredictability and capricious leadership; for his government's lack of focus, coordination, and implementation; for legislative gridlock, the slow pace of re- forms, and insufficient emphasis on institution building. He was sharply criti- cized for failing to curb corruption, resolve secession issues, and prosecute key figures of the Soeharto regime, and for the deterioration of the economy (dis- cussed later in the chapter). At the beginning of 2001, a parliamentary investigative committee re- leased a report stating thatWahid had been aware that his personal masseur (also Wahid's spiritual adviser) had embezzled $3.7 million from a government agency. It also criticized the president for failing to declare a $2 million dona- tion from the Sultan of Brunei (which Wahid called a personal gift). But the panel found no proof that Wahid benefited personally from these transactions. "The government The president's supporters said that compared with Soeharto,Wahid's lapses are "like chickenfeed." But parliament found that Wahid had acted improperly in here is almost both cases. Parliament censuredWahid for misconduct in the two financial matters. But hydroponic-a it was widely agreed that the alleged corruption was not the real issue.The cor- ruption charge was simply the method used to express dissatisfaction withWahid's floating structure governance of Indonesia (or lack thereof). As Calvin Sims commented in The New York Times on 4 February 2001 (p.A10): with dangling, Behind the Indonesian Parliament's decision to censure President ineffectual roots" AbdurrahmanWahid for misconduct in two financial schemes last week, setting the stage for his impeachment, was widespread frustration with his erratic stewardship of the world's fourth largest population, which has been floundering.. .The corruption scandals were only the catalyst... For months there has been a growing sense of anger and despair that Mr.Wahid has failed to take the necessary steps to pull the country out of years of political and economic turmoil or end the separatist and religious violence that threaten to tear this archipelago apart. Parliament again censuredWahid in April 2001, this time for incompetence and corruption.The vote was overwhelmning: 363 to 52, with Wahid's defend- ers nearly all from his small National Awakening party. Impeachment proceedings (which can be initiated after two censures) could now legally take place. Al- though the military was now under civilian control, the army chief of staff, Gen- eral Endriartono Sutarto, warned the president not to dissolve parliament as a method of avoiding possible impeachment proceedings. In May 2001 Attorney General Marzuki Darusman, a widely respected for- mer director of the National Human Rights Commission, cleared President Wahid of any involvement in the corruption cases that had led to parliamen- tary censure. But only days later parliament voted 365-4 to convene a special session of the People's Consultative Assembly (MPR) to decide whether Pres- identWahid should be impeached. (In a cabinet shift the followmg month,Wahid An Introduction to Indonesia 63 replaced Marzuki Darusman as attorney general, citing failure in the prosecu- tion of former president Soeharto and his youngest son,Tommy Soeharto). The MPR special session was convened inJuly. Urged to resign,Wahid re- fused. He declared a state of emergency, but the military refused to carry out his order. On 23 July 2001 the MPR voted 591-0 to impeach Wahid and to replace him withVice President Megawati Soekarnoputri for the remainder of Wahid's five-year term which expires in 2004. Those in the 700-member MPR who opposed the impeachment boycotted the vote. The military sup- ported Megawati and protected the assembly building with tanks during the vote. Iromcally, Abdurrahman Wahid, the first Indonesian president to whom power was transferred democratically, also became Indonesia's first democrat- Wahid was unable ically impeached president. (Official impeachment procedures, starting with the first parliamentary censure, were carefully followed throughout). to gain sufficient Three of the most important issues that markedWahid's presidency are dis- cussed below. Each is presented simply as a snapshot of a situation at a partic- power to bring about ular point in time. But together they can provide some insight both into Indonesia during the 21 months ofWahid's presidency, and into the state of the the democratic nation that was entrusted to Megawatn Soekarnputri on 23 July 2001-the month following the celebration of the 1 00th anniversary of her father's birth. government of Taking on the Soehartos. PresidentWahid made a concerted effort to bring shared power in former president Soeharto to trial for corruption and abuse of power. But he was unable to do so. Shortly after taking office, Wahid sent a clear signal that which he believes (unlike former President Habibie), he intended broad anticorruption reforms. Charges were drawn up, and plans for Soeharto's trial were initiated.Wahid said that he would pardon the former president if he were convicted, provided that his illegally acquired wealth was returned to the nation.The pardon, however, would not extend to the Soeharto children (Soeharto's wife, Ibu Tien Soehar- to, had died in 1996). But in the process of moving toward trial, the govern- ment found that the Soehartos' cronies and loyalists remain strong, wealthy, and widespread in Indonesia. In early 2000 the government began to prepare the case against the for- mer president.Then in May, Soeharto was placed under house arrest "to smooth the investigation and in order to finish the case as scheduled," according to the attorney general, Marzuki Darusman (The New York Times, 30 May 2000, p.A3). But Soeharto's doctors said that he was too ill to stand trial. Although his trial was to begin on 31 August 2000, Soeharto did not appear in court; his doc- tors testified that he could not appear because of illness.The case was adjourned until September 14, and the attorney general and government prosecutors con- tinued with documentation and preparation.Then a spate of bombings rocked Jakarta. Bombs were set off at the attorney general's office (where work on the case was in preparation), at a crowded parking garage underneath the stock exchange (on the eve of the scheduled resumption of the trial), at Christian religious or- ganizations, and at other sites. It is widely believed that the bombings were car- 64 The Microfinance Revolution: Lessons from Indonesia ried out by Soeharto supporters in an attempt to derail the Soeharto trial and to undermine efforts to restore the stability of the government. By the end of September, the court dismissed the case against Soeharto on the grounds that he was too ll to stand trial. Meanwhile, in September 2000, in a reversal of a lower court decision, In- donesia's Supreme Court found Soeharto's youngest son, Hutomo Mandala Putra (known as Tommy), guilty of corruption in a multimnllion-dollar land deal. He was sentenced to 18 months in prison-the first conviction of any member of the family for graft. But Tommny disappeared. After two months he was found by security forces, but Tommny disappeared again. "He [Wahid] said... Tommy had slipped away while officers were checking with the palace on whether to arrest him. 'When the timing is right, he will be detained,' said Mr.Wahid" (The Wahid made a New York Times, 30 December 2000, p. A4). President Wahid publicly denounced Soeharto's youngest son as a crimi- strong effort to bring nal and also ordered his arrest in connection with the Jakarta bombings. But, the government was not successful in arrestingTommy in connection with the Soeharto to trialfor bombings, or in having him serve his prison term for corruption. The Wahid government continued its efforts to bring members of the Soeharto family to corruption and abuse court on corruption charges, but many Indonesians were highly skeptical that this would happen-and it did not. of power. But he In July 2001-just three days after Megawati Soekarnoputri took over as Indonesia's president-Syafiuddin Kartasasmita, the Supreme Court justice who was unable to do so had sentenced Tommy Soeharto to prison in the corruption case, was shot and killed in broad daylight by gunmen on motorcycles. According to the Jakar- ta police, two suspects were arrested and confessed. They said they had killed the justice under orders from Tommy Soeharto who had supplied the guns and paid them each 100 miUlion rupiah ($10,500) for the killing.Jakarta po- lice chief Sofyan Yacob accused Tommy Soeharto of masterminding the as- sassination, and the Jakarta police, assisted by National Police Headquarters, mounted a large-scale-but unsuccessful-search for the fugitive son of the former president. One of the men suspected of killing the justice was report- ed to have died in police custody. In a telling conmmentary on the state of Indonesian justice, a year after the Supreme Court's September 2000 conviction of Tommy Soeharto for corruption, the court overturned its own conviction. Seth Mydans, in an article called "If Tommy's so hard to find, justice is indeed blind," quoted Hendardi, chairman of the Indonesian Human Rights Association," Police officers are afraid that if they help catch Tommy, they could be killed." Mydans commented, "Did the message get through [to the court]?"And Anien Rais, speaker of parliament, said, "This spells disaster for the rule of law in Indonesia" (The New York Times, 4 October 2001, p. A4). But Indonesia's current president is the daughter of Sukarno-from whom Tommy's father, Soeharto, took power in 1966. Four months after Megawati Soekarnoputri became president,Tommy Soeharto was arrested (in a Jakarta mansion) and taken into custody by Jakarta police for al- legedly plotting the murder of Supreme Court Justice Syafiuddin Kartasasmi- An Introduction to Indonesia 65 ta.This arrest and the results that follow, which are far from certain, will be wide- ly viewed as a signal of power and direction in Indonesia. Can the center hold? AsWB.Yeats wrote in The Second Coming (1919),"Things fall apart; the center cannot hold."This was one of the most crucial issues that PresidentWahid faced (and that President Megawati now faces). The Indonesian nation, forged with so much struggle in the 20th century, faces multiple, escalating separatist movements at the start of the 21"t century. Recent surges of violence by secessionists in some provinces are thought to be assisted, in varying degrees, by Soeharto supporters intent on destabilizing the government. Ethnic and religious In Aceh, a special region in northern Sumatra that has rich oil and natur- al gas reserves, strongly Muslim secessionists have been fighting for an inde- rivalries, long pendent Islamic state for decades. Under the Soeharto government, the independence movement was repressed by the Indonesian military; 5,000 peo- controlled under the ple, mainly civilians, are estimated to have been killed in Aceh during the 1 990s (The New York Times, 17 September 2000, p. 9).The independence movement Soeharto regime, has become more overt, visible, and violent since Soeharto's resignation. Con- flict between the independence movement and the military has turned Aceh became openly into a "virtual war zone"(The New York Times, 11 November 2000, p.A3).Tens of thousands of Acehnese fled their villages for refugee camps. Despite an in- violent and ternationally brokered cease fire agreement between the Indonesian govern- ment and the Free Aceh Movement signed in Geneva on 2 June 2000, violence destabilizing has continued. In November 2000, PresidentWahid blamed the army and po- lice for escalating the violence onAceh and for 19 civilian deaths that occurred there. He said, "Acehnese people are my religious brothers. I want to ask [the top military and police officials involved] 'Since when are guns used in nego- tiations? If you are using guns, then please retire"' (The Boston Globe, 11 No- vember 2000, p. A15). In an attempt at compromise, the government offered Aceh increased au- tonomy and a larger share of the profits from its oil and gas supplies. President Wahid even agreed that a form of Islamic law could be imposed in Aceh, even though the Indonesian nation has never been an Islamic state. ButWahid ruled out independence for Aceh. However, many Acehnese say they will settle for nothing less than independence. A growing separatist movement exists in Papua (also known as Irian Jaya, located on the island of New Guinea)-which has giant copper and gold mines, vast timber resources, and offshore oil and gas.The indigenous Papuans of Irian Jaya have long aspired to independence from Indonesia. But Irian Jaya is re- source rich, and is the home of the Freeport mine, a subsidiary of Freeport- McMoran Copper & Gold, based in New Orleans (Louisiana, United States). This mine is Indonesia's biggest taxpayer and largest private employer. Since 1992 P.T. Freeport Indonesia has paid $1.5 billion in taxes, royalties, and divi- dends, as well as $160 million for regional development. (The Boston Sunday Globe, 10 September 2000, p. A23). 66 The Microfinance Revolution: Lessons from Indonesia Given the wealth and strategic locations of both Aceh and Irian Jaya, the Indonesian government is not prepared to consider independence for either province. Unlike the independence movement in East Timor, there is little in- ternational support for these independence movements because, if successful, they could destabilize the largest country in southeast Asia-and one that con- trols some of the world's most important shipping lanes. But it is an open question as to whether these conflicts can be resolved, es- pecially given the destabilizing aims of some Soeharto supporters and the dearth of functioning legal,judicial, and administrative institutions. In other areas ethnic and rehgious rivalries, long controlled under the Soe- harto regime, are now openly violent and highly visible. In the province of Cen- tral Kalimantan, anmmosity has existed for decades between the indigenous Dayaks Wahid accused and immigrants from other parts of Indonesia (mainly people from the island of Madura who were resettled in Kalimantan by the Indonesian government Soeharto supporters, as part of its transmigration program). But recently, the clashes between the two groups escalated into a campaign of ethnic terrorism, with the Dayaks trying some members of the to drive the Madurese from what they consider Dayak homeland. Security forces have made little attempt to stop the murders and violence (see The Boston Globe, military, and 27 February 2001, p.A9). But the future of Central Kalimantan is not the con- cern only of the Dayaks, the Madurese, and the government.Vast timber and Muslim extremists mining interests operating in Kalimantan, many connected with the Soehar- tos, are also major players in the provincial maelstrom. of using sectarian In West Timor, the defeated East Timorese militias-reorganized, rearmed, and tacitly backed by some Indonesian military units-continued to terrorize violence to create and harass the East Timorese refugees who fled or were forced into West Timor during the violence that followed East Timor's vote for independence (an es- political instability timated minimum of 50,000 refugees remained in the camps as of September 2001). But the East Timorese militias and their Indonesian military commanders were not brought to justice, and militia fighters continued to cross the border into EastTimor, attacking both EastTimorese civilians and United Nations peace- keepers stationed in East Timor. Only Eurico Guterres, the prominent militia leader widely believed to have been responsible for many of the worst crimes that took place during the East Timor massacre, was sentenced (on a charge of weapons possession)-to six months of house arrest. No one else-none of the other East Timorese militia leaders, none of the Indonesian officers who commanded them-has had to serve even a day in detention.. .Desperate to avert the creation of an international tribunal, Indonesia promised to conduct its own trials. The government formally agreed to share information with the new East Timorese administration and to provide witnesses and defendants for separate trials there. So far, both East Timorese and United Nations officials say, none of these promises have been kept... Former Defense Minister An Introduction to Indonesia 67 Juwono Sudarsono, in an interview, had little to say in defense of Indonesian justice. "The court system is in shambles," he said. In addition, President AbdurrahmanWahid is too engrossed in his own problems of political survival to push the issue and too dependent on the political support from the military to challenge it".. ."It's been a farce all along" said one United Nations official in Jakarta. -Seth Mydans, The New York Times, 16 May 2001, p.A4 In September 2000 the militias killed three United Nations aid workers who were working to repatriate East Timorese who were still in West Timor. For- Efforts to restructure eign workers then withdrew from West Timor. The Indonesian government pledged to disband the militias. But it did not (or could not) do so, and the mili- corporate debt and to tias remain active. In May 2001 a Jakarta court sentenced six militia members to prison terms, none of which exceeded 20 months, for the murders of the build an effective three United nations workers. and accountable This case apparently proceeded, unlike the others, because the victims were foreign United Nations employees. But the banking system Indonesian prosecutors did not seem to have their hearts in the case, declining to charge the men with murder or manslaughter remain mired in even though some of them admitted to stabbing the victims. "The sentences make a mockery of the international political problems community's insistence that justice be done in this horrific case" the United Nations High Commissioner for Refugees said. -Seth Mydans, The New York Times, 16 May 2001, p.A4 Meanwhile, religious pressures have been rising in Indonesia, with fights between Muslims and Christians becoming commonplace in some parts of the country.Wahid accused Soeharto supporters, some members of the mnlitary, and Muslim extremists of using sectarian violence to create political instabldity. Much of the violence between the religious groups has occurred in Maluku province in eastern Indonesia where it erupted in 1999 and then spread in 2000 to the island of Lombok, east of Bali.When military and police were sent to restore order, some of them divided along religious lines, each supporting local groups of his own religion.Jakarta has also seen numerous bombings of churches, restau- rants, and mghtclubs carried out by Muslim vigilante groups. Slow progress on economic andfinancial reform. Indonesia's progress in re- covering from the crisis has been slow. The widespread corruption and deep political uncertainties are major causes of the country's continuing economic problems.And President Wahid, who did not allocate high enough priority to the economy, made some poor choices in selecting his economic advisers.Yet, there has been encouraging economic progress on some important fronts. GDP growth reached 5.2 percent in 2000, representing an 18.5 percent turn- 68 The Microfinance Revolution: Lessons from Indonesia around in two years (see table 8.1). Inflation in 2000 was down to 3.7 percent. Oil prices stayed firm and there was strong external demand for Indonesian oil. And non-oil exports in 2000 were up substantially over 1999, reaching an all- time high (although their rate of growth was below that of their pre-1997 level). Mark Baird, the World Bank's country director for Indonesia remarked that in Indonesia, "you see a much healthier economy than you might expect read- ing reports overseas" (The New York Times, 1 November 2000, p.W1). However, many difficulties remained.At the end of December 2000 the ex- change rate was 9,595 rupiah to the U.S. dollar-compared with 7,085 at the end of 1999 (see table 8. 1).And by the end of March 2001 the rupiah had de- clined to 10,400 to the U.S. dollar.The rupiah continued to slide, with the ex- change rate at over 11,000 rupiah to the U.S. dollar from mid-April until July "The Indonesian 23 when President Wahid was impeached and President Megawati took office. By August 14 the rupiah had appreciated to 8,425 to the U.S. dollar, but by Bank Restructuring the end of October 2001 it had again dropped to 10,435. A huge foreign debt remains, estimated at over $150 billion, with debt ser- Agency is at the vice at over 20 percent despite multiple debt rescheduling agreements (Bar- clays Bank Country Report, 28 November 2000). Efforts to restructure the epicenter of a corporate debt and to build an effective and accountable banking system, both needed for any significant growth in private investment, remain mired in po- struggle over litical problems (chapter 15). Corporate governance remains a formidable chal- lenge. In Indonesia, corporations have long been accustomed to operating in Indonesia's economic a closed and collusive environment. Changing this corporate culture under pre- sent circumstances is extremely difficult. soul" The government assumed the bulk of recapitalizing the banking system, which is expected to amount to over 50 percent of GDP (see Barclays Bank Country Report, 28 November 2000); some estimates run as high as 80 per- cent of GDP (Tim Healy and Tom McCawley, in Asiaweek, 13 August 1999). But progress in restructuring the banking sector-through bank mergers, man- agement changes, loan collection efforts, and reschedulng of loans-has been slow, mainly because of continuing institutional weaknesses and political in- fighting. The Indonesian Bank RestructuringAgency (IBRA) was created in 1998 to reorganize the banking system and to manage the liquidation, mergers, re- structuring and recapitalization of faling banks. Its asset management unit was established to administer the bad loans of the banking system. It also plays a major role in restructuring corporate debt. IBRA's performance has been mixed.While some progress has been made, IBRA has often not been a match for the Soehartos and the conglomerates (chapter 15). [IBRA] has an impossible task and powerful opponents... [It] was created in 1998 to rescue a basket case of a banking system that included more than 200 poorly capitalized, opaque institutions burdened with billions of dollars in bad debt and usually owned by the debtors. IBRA today controls assets equal to 57 percent An Introduction to Indonesia 69 of GDP Meanwhile it has become an inefficient, less-than- competent, sometimes corrupt bureaucracy supported.. .by a weak central government... .IBRA is at the epicenter of a struggle over Indonesia's economic soul. On one side are the still-powerful forces of the old regime-Suharto's children and cronies. And on the other are economic reformers and activists who want retribution for the pain they see inflicted on the country by sweetheart deals and outright corruption. -Caragata 2001, p. 2 However, a new IBRA governance structure was adopted in 2000, mcluding During thefirst half appointment of a governing board composed of independent professionals. Some progress on restructuring banks and corporate debt has begun to occur. But a of 2001 the rupiah recent high-level IBRA appointment has raised some concerns because it is widely believed that the official is closely connected to the Soehartos-who fell steeply, economic have strong vested interests in IBRA's asset sales, bank privatization, corporate debt restructuring, and other actions. growthfaltered, Vast amounts of funds are idle ($45 blllion at the end of December 2000). At that time the average loan-to-deposit ratio of the banks was 36 percent (The inflation rose, and Nikkei Weekly, 5 February 2001). Some large banks have loan-to-asset ratios as low as 6 percent. Rather than lending, bankers have been keeping their funds dissatisfied donors in the central bank. In part this is because, in an effort to curb inflation, the central bank interest rate has been as high as 4 percent above the rates offered held back by commercial banks. Partly it is because bankers do not want to expand lend- ing at a time of high political risk. And partly it is because most of the larger disbursements banks have substantial funds in illiquid state bonds. There were also problems at Bank Indonesia, the central bank. In late 2000 the Supreme Audit Agency reported that during the financial crisis, nearly $9 billion ($17.2 billion at December 2000 exchange rates) provided in emergency loans to private-sector banks, many controlled by Soeharto cronies, had been diverted to unauthorized loans and other spending."The Supreme Audit Agency estimates that as much as 95 percent of the money may never be repaid. It blames Bank Indonesia for failing to track it. Much of the money went to banks con- trolled by Suharto cronies, and 59 percent was misspent, says the agency... If [the deficit] were all charged to the bank, it would go bankrupt" (Asiaweek, 22 De- cember 2000). Given the slow progress in restructuring public and private debt and in re- building the banking system, the continuing power of the Soeharto family and cronies, rampant corruption, the weak legal and judicial systems, regional un- rest, ethnic and religious tensions, and political uncertainties, foreign investors are generally staying away from Indonesia. In a now-famous interview with Dow Jones newswires in May 2000, then-coordmating minister for the economy Kwik Kian Gie said,"If I were a foreign investor, I wouldn't come to Indonesia.The law enforcement is not there, but not only that, the whole thing is so confus- ing."The minister may have been impolitic, but he was not wrong. 70 The Microfinance Revolution: Lessons from Indonesia Investors (domestic as well as foreign) were waiting to see whether Presi- dentWahid could estabhsh enough political authority to create a viable investment environment. But during the first half of 2001 the rupiah was falling steeply, the country's slow economic growth was faltering, and inflation was rising. The IMF and Indonesia's other donors had become dissatisfied with the glacial pace of agreed-upon reforms and with the growing political instabili- ties. In December 2000 the IMF suspended disbursement of a $400 million tranche of its aid program for Indonesia. And at the World Economic Forum in Davos, Switzerland in early 2001, Stanley Fischer, deputy director of the IMF, reported that the progress of structural reform in Indonesia had been very slow. In April 2001, the IMF again decided not to disburse the $400 million of sched- uled assistance which was held back throughout the rest ofWahid's presiden- Wahid declared a cy because of the views of the IMF and other donors that Indonesia had not complied with parts of the agreement. state of emergency In early 2001 Japan, Indonesia's largest donor, informed the Indonesian gov- ernment that its official development assistance to Indonesia would be reduced. but the military The World Bank and the Asian Development Bank stated similar intentions. TheWorld Bank announced that it would cut its annual loan to Indonesia from refused to respond. $1.2 billion to $400 mnllion. The Bank's vice president for East Asia and the Pacific,Jamal-ud-din-Kassum, said "Improved governance and progress on the He became reform program will be key determinants of future levels of assistance from the World Bank" (The Nikkei Weekly, 5 February 2001). Indonesia sfirst Indonesia under President Wahid. Indonesia's fourth president was a toler- democratically ant, libertarian, intellectual who encourages debate and believes strongly in democracy and its institutions. He understood that for Indonesia to become a impeached president well-functioning democratic state, transparency and powersharing would be re- quired. But Indonesia is still a neophyte in these areas.And the covert, but strong and well-financed, opposition to such reforms plays by its own rules. Because of the weak, corrupt legal system and judiciary, there are not only no accept- ed rules, but also few referees (those who attempt to provide justice, as Supreme Court Justice Syafiuddin Kartasasamita did, risk assassination). The president's low level of managerial and administrative skills, poor health, and lack of support in parliament, the country's lack of functioning in- stitutions, the strength and wealth of the Soeharto forces, the economic prob- lems, the secessionist movements, and the generally chaotic political scene combined to make it impossible forWahid to govern the country effectively. He was impeached in July 2001 in a democratic process, according the rule of law, at a time when the country needed to move on. President Megawati Soekarnoputri By vote of the People's Consultative Assembly (MPR), Megawati Soekarno- putri replaced former PresidentWahid in a peaceful transition of power on 23 July 2001. She is to serve as president until the 2004 general elections (she will be eligible to run for a five-year term as president at that time). However, An Introduction to Indonesia 71 Megawati faces the same difficult problems as Wahid did. Indonesia is a weak state, lacking basic government and civilian institutions, and it is riddled with corruption and special interests. But a strong ruler in the mold of the coun- try's first two presidents may no longer be acceptable to Indonesians. President Megawati is known as a conservative nationalist. It is widely ac- cepted in Indonesia that Megawati believes the power she now holds to be- long to her by destiny. She is the immovable object-stolid, silent, imperious-a puzzle to her countrymen even as she commands unrivaled popularity .. Her deep and dignified silences create a circle of awe Vice President around her .. Perhaps, some say this could only happen here, where Javanese culture reveres silence and where power is seen as a Megawati mysterious mantle that ... envelops a leader of its own accord. -Seth Mydans, The New York Times, 6 May 2001, p. 3 Soekarnoputri, But Megawati also commands the largest political party and the strongest Sukarno's daughter, political support from the people. And she is supported by the military. In the last few months of her vice presidency, whenWahid was concerned mainly with replacedformer his political survival, Megawati made crucial decisions to restore relations with the IMF-over the objection of the then-coordinating minister for the econ- President Wahid as omy. She adjusted and restructured the budget to bring it in line with the IMF- supported program, removed subsidies on fuel, and took other urgent actions. president in a In contrast to the MPR's decision about the presidency, the vice presidential election by the MPR was embroiled in bitter partisan politics.There were two peaceful transition leading candidates. One,AkbarTandjung, is closely associated with the Soeharto regime. At the time of the election he was the speaker of the Indonesian par- on 23July 2001 liament (DPR). He was also the leader of Golkar. The other candidate was Hamzah Haz, a conservative Muslim politician who had been a leader of the opposition to Megawati's candidacy for the presidency after the 1999 general elections, and who lost to her in the 1999 contest for vice president.The view he expressed then was that "Islam does not allow women to lead governments." In 2001 Hamzah, leader of the United Development Party (a Muslim party), won the election for vice president in the third round of a close race. The political infighting for the vice presidency prompted one political com- mentator, Chatrib Basri, to comment on a concept of power now emerging in Indonesia (quoted by Seth Mydans in The New York Times, 26 July 2001, p.A8): "What's most scary is to see how the elite aren't at all disturbed about what's going on. If there's been any profitable business in these last years [the post-Soe- harto times] it's been the business of buying and selling power." It is not that money was never before exchanged for power in Indonesia. But the image is different now, perhaps reflecting deeper changes under way. In the new image power is bought and sold. In the traditional image money flows from power. President Megawati's choices for her cabinet were widely praised both in Indonesia and internationally. "Much to the market's delight she chose people 72 The Microfinance Revolution: Lessons from Indonesia well-known and generally highly regarded in business circles to fill the key eco- nomics, finance and business posts" James Castle, Business Times, 10 August 2001). TheJakarta Post commented, "Megawati names rainbow cabinet... .Most key cab- inet posts, particularly in the economic arena, were given to professionals and bureaucrats, but there were enough seats left to placate the political factions and secure her the crucial support of the legislature" (10 August 2001). Only 4 cab- inet positions (out of 32) are held by members of the military. In Indonesia cabinets are named. Megawati named hers the Gotong Roy- ong Cabinet. Gotong royong is a phrase that is deeply imbedded in Indonesian consciousness. It refers to the ancient but continuing tradition of mutual help found in Indonesian villages-with the idea that everyone contributes to a pro- ject, and everyone benefits. It is widely accepted Economics andQfnance. With her strong cabinet appointments in economics that President and finance and her restoration of Indonesia's relations with the IMF, Presi- dent Megawati showed early indications of a serious intent to restore the econ- Megawati believes omy. Shortly after Megawati's first month in office, the government signed a new agreement preparing the way for resumption of a $5 billion IMF loan the power she now program and its long-delayed $400 million installment. The agreement in- cluded limiting the budget deficit and setting a timetable for selling or pri- holds to belong to vatizing failing corporations and banks. The Jakarta Post (24 August 2001) editorialized: her by destiny A new agreement with the IMF will.. .strengthen market confidence as it will greatly help smooth Indonesian relations with its creditors... One should not take lightly the good understanding shown on the part of foreign creditors, given the government's foreign debts of about $65 billion and corporate foreign debts of almost $70 billion...The hardest part of the job is for the government to demonstrate its real implementation capability in delivering on [its] promises.. .Learnmg from the bitter experiences and mistakes of the previous government, the Megawati administration should develop the kind of capability that reflects three fundamentals in the strategic interactions between people and government officials: accountability, transparency and predictability. The minister of finance in the Gotong Royong cabinet, Boediono, had pre- sented seven strategic points for stabilizing the Indonesian economy to an in- ternational Conference on Indonesia held in Tokyo on 30 March 2001:4 * Restoring Indonesia's battered self-confidence. * Reestablishing law and order. * Improving policy decisionnaking and implementation. * Normalizing the financial system. An Introduction to Indonesia 73 * Achieving true economic recovery, propelled by new Investments as well as a more productive use of existing resources and assets. * Maintaining macroeconomic balance. * Managing poverty and inequality. If these goals can be effectively implemented, Indonesia would be well on its way to recovery. It is too early to comment on the new government's progress toward its economic goals. But as is discussed below, it is worth not- ing that concerns are beginning to surface that little change appears to be under way and that the cabinet's economics team is running into the same difficul- ties faced by a variety of predecessors. Megawati made Other government priorities. In her Independence Day speech on 16 Au- strong cabinet gust 2001, President Megawati outlined a number of the government's prior- ities (see Seth Mydans, The New York Times, 17 August 2001, p.A8). Megawati appointments in emphasized national stability, human rights, and the need for fighting corrup- tion-about which she said, "Unlike in a feudalistic society, which doesn't con- economics and sider corruption a serious mistake, in a democratic society it is a big problem." She apologized for human rights abuses commnitted by the military in separatist finance and showed rebellions and said the military must reform itself. "We need to pay more at- tention to human rights.We need a security force which is effective, highly dis- indications of a ciplined, and under the control of the government." For the first time, Megawati acknowledged the independence of East serious intent to Timor, which she had opposed in 1999. But she ruled out independence for Aceh or Irian Jaya, encouraging them to "help build a new Indonesia." restore the economy Megawati also stated that the country's constitution (prepared in 1945 by her father, later abrogated, but reinstated by him in 1959) needs to be revised and updated. She proposed that an independent constitutional commission be established to seek the people's views on issues and then provide a compre- hensive draft of the amendment for review and enactment by the People's Con- sultative Assembly (MPR). MPR Speaker Amien Rais responded positively to this proposal, and discussion of the proposed commission is expected by the MPR. The issues now are whether President Megawati will be able to implement her government's priorities and achieve for Indonesia a balance between sta- bility and reform-while also controlling the military, reigning in the Soehar- tos, keeping the country united, maintaining the nation with the largest Muslim population in the world as a non-Islamic state (and managing Indonesia's re- sponse to issues resulting from the war in Afghanistan), building relations with the IMF and the international donor community, and holding her support in the legislature and with the people. These are difficult tasks and there are formudable obstacles. But Megawati has the support of most Indonesians, of the army, and of a majority in parlia- ment.The issue is political will.Three recent views lllustrate, in different ways, growing concerns about corruption, inertia, and lack of political will. 74 The Microfinance Revolution: Lessons from Indonesia It seems that the more things change, the more they stay the same. Corruption flourishes in new, inventive ways, there is no functioning structure to penalize wrongdoing, economic and legal reform is at a standstUll and businessmen are refusing to repay debts. Indeed, more than three years after Soeharto's downfall, there is little fundamental difference in how Indonesia's.. .people are being governed... It all adds up to a familiar gloomy picture. Megawati began her term in office with great promise by appointing well-regarded economics ministers. But the so- called "dream team" is already showing signs of paralysis- thanks in large part to a lack of political will on the part of the president ... Says former Attorney-General Marzuki But inertia is strong. Darusman... "The government is fast becoming immobilized because of inertia." "Corruption -John McBeth, Far Eastern Economic Review, 1 November 2001, pp. 17, 19 flourishes... there is Modern Indonesia is a crazy place. Incoherent, unprincipled and no functioning cymcal... It's not a good time for anyone with decency to be in government. structure to penalize -Sarwono Kusumaadmadja, a cabinet nmister in the Soeharto and Wahid governments, quoted by John McBeth, Far Eastern Eco- wrongdoing, nomic Review, 1 November 2001, pp 18-19 economic and legal Unlike Habibie and Abdurrahman Wahid], the new president [Megawati] possesses both the legitimacy and the power base reform is at a that are requisites for becoming an effective leader in the post- Soeharto era. But where is Megawati's politcal capital being used? standstill" Unfortunately, the answer is, towards no apparent end... If Megawati were to be more courageous and ask her followers to support her in making the hard decisions on economic reform and national security, her opponents would have a difficult time in maintaining the status quo of yesteryear. Megawati and her cabinet have the political capital to make a difference in Indonesia's future, but they should be mindful of the old adage, "Use it, or lose it." -JamesVan Zorge, TheJakarta Post, 19 November 2001 Although President Megawati is not known as a reformer, she has demon- strated courage and political skills under difficult circumstances. Megawati and her government have the political capital to make the hard decisions on the economy, to crack down on corruption, and to begin building the institutions Indonesia so desperately needs. But as of this writing in 2001, it is still too early to know whether the government will exercise the political will needed to meet these challenges effectively. An Introduction to Indonesia 75 Democracy and the Messy State Indonesia's future is well beyond the scope of this book, as well as beyond its author's crystal ball-reading abilities. But history may provide some perspec- tive. Sukarno led the independence struggle and unified the country. Soehar- to took over an economy in chaos and provided three decades of unprecedented economic growth and development.The vast corruption that permeated the country under Soeharto, along with the decades of arrested political develop- ment, were recognized and challenged by the same people who had grown up better fed and housed, healthier, and better educated because of the priorities and policies for economic development of Soeharto's New Order government Herfather built the and its economics ministers. President Habibie, Soeharto's protege, began the political reform process that President Wahid then broadened and deepened. Indonesian nation. The country's long-term commitment to its national goals of growth, eq- uity, and stability remain intact (although its goal of harmony seems more prob- It is much to be lematic). But Indonesia's history as a weak state with a strong ruler led, in the Soeharto era, to a system of governance characterized by pervasive corruption hoped that and a severe lack of transparency and accountability. Eventually the governance choked the development.The ensuing crisis was devastating.The transition from Megawati will build the Soeharto regime could not have been easy under any circumstances. But it has been especiaUy difficult because of the severe crisis that triggered the end the institutions that of the era. The political climate has provided a major impediment to the country's enable a strong much-needed institution building. The Soeharto forces are still able to evade justice, undermine reforms, and destabilize the country. And the dearth of ex- democratic state perienced national leaders is a direct Soeharto legacy. Confidence in the Indonesian economy has not returned, and few foreign investors are back.Viability has not been restored in the corporate sector. Se- vere difficulties stand in the way of restructuring the financial system-the out- come of which will affect the future distribution of wealth in Indonesia. In addition, the growing religious violence poses a threat to munority religions-and perhaps eventually to the choice made at the time of inde- pendence that Indonesia would not be an Islamic state. And the separatist movements provide dangers to the country's unity and to Southeast Asian stability. On the other side of the ledger, the transition from PresidentWahid to Pres- ident Megawati took place peacefully, lawfully, and democratically. Parliament is no longer a rubber stamp. International donors have begun to return. The milhtary is controlled by civilians (though to what extent the military is under control is an issue), and the role of the armed forces has been reduced.The press and other media are free to write and say what they see and what they believe. And on 30 August 2001 East Timor, a United Nations protectorate, elected its first government-in preparation for a transition to independent statehood in 2002. And the many East Timorese refugees remaiming in West Timor began to go home. 76 The Microfinanco Revolution: Lessons from Indonesia Box 8A Excerpts from Thomas L. Friedman's "What a Mess!" In the cold war we had authoritarian, democratic, and Communist states. Now we have authoritarian states (North Korea, Iraq), democratic states (America, France), democratiz- ing states (Poland, Chile, Hungary), failed states (Sierra Leone, Liberia) and messy states-namely Russia and Indonesia In messy states the old authority structure that allocated resources, enforced con- tracts, and collected taxes-President Suharto in Indonesia and the Communist Party in Russia-has broken down but has not been replaced by a new authority that can play the same role The result, in Indonesia and Russia, is rampant corruption and a fragmentation of power in which neither the army, the Parliament, the executive, nor the remnants of the old order have the strength to assert their will Messy states too big to fall, too mes- 'No one should sy to work That's why in messy states you never quite know-when arms are sold, people doubt the severity of murdered, or payoffs demanded-whether this is by design of those ostensibly in charge or because no one is in charge "In Suharto's time," said Witmar Witoelar, Indonesia's popular talk show host, "things were clear-you always knew who to pay and how long the current economic to wait Now you never know who's in charge Before people were being killed by the government Now they are killed because of no government " crisis. Yet neither Source Thomas L Friedman, "What a MessI" The New York Times, 3 October 2000, p A31 should anyone underestimate our These are among the many remarkable achievements that have occurred in the short period since Soeharto's resignation. And Indonesia continues to capacity to set our have abundant resources: human capital (including a large and educated labor force), ample natural resources, a highly strategic location, an extensive infra- institutions right" structure, a large domestic market, and long experience with sustained economic growth and development. In his strategy for the stabilization of the economy, Minister of Finance Boechono listed first the restoration of Indonesian self-confidence. Many In- donesians have suffered financially, physically, and emotionally in recent years. And some have lost pride and confidence in their country. Restoring self-con- fidence is a sine qua non for restoring the economy and addressing the na- tion's problems. Building responsible, accountable institutions that can take over the broad political and administrative functions previously performed by the military is the most difficult, the most pressing, and the most important challenge for Pres- ident Megawati Soekarnoputri. Her father built the Indonesian nation. It is much to be hoped that Megawati will build the institutions that enable a strong de- mocratic Indonesian state. Indonesia has become known internationally as a'messy state': "too big to fail, too messy to work" (box 8.4). But as Mulyono (1981, pp. 178-80) points out in his chapter entitled,"TheWayang: Messy but Loved Down the Centuries," wayang performances (with their "sometimes grotesque incongruities") have been teaching for thousands of years how truth is found beneath mess. An Introduction to Indonesia 77 There is no doubt that there are difficult and messy times still ahead. But with its extensive human and natural resources, it iS likely that Indonesia wil recover and move on.AsAliWardhana (1998b,p. 5) put it,"No one should doubt the severity of the current economic crisis.Yet neither should anyone under- estimate our capacity to set our institutions right." Notes 1.The 1999 and 2000 GDP growth data are preliminary figures from Biro Pusat Statistik (BPS), the Indonesian Bureau of Statistics. 2.Javanese refers to an ethnic and linguistic category of people who have tradi- tionally inhabited the central and eastern parts of the island of Java. Another group, the Sundanese, are found in the western part of the island. However, some Javanese live in other parts of Indonesia, while other munorities also live on Java (see Koent- jaraningrat 1985). 3. For discussion and analysis of Indonesia's history, society, and politics, see Raf- fles (1977 [1817]); Crawfurd (1993 [1820]); Soedjatmoko (1960,1967); Palmier (1960); Feith (1962);Soedjatmoko and others (1965);Shaplen (1969);Zamu'Ddmn (1970);McVey (1967); U.S. Central Intelligence Agency (1968);Anderson and McVey (1971); Neil (1973); Emmerson (1976); Anderson (1990); Jackson and Pye (1978); McDonald (1980);Jenkms (1984);Van Neil (1984);Robison (1986,1988,1990,1992, 1993);Crouch (1988 [1978]); Crouch and Hill (1992);Bresnan (1993);Mackie and Maclntyre (1993); Vatiklotis (1993); and Schwarz (1994). For discussion and biblhographic references, see Nell (1973) for Indonesia's environment and early history; Geertz (1963, 1984), Koentjaraningrat (1975a, 1975b, 1985), Fox (1980), andWhite (1983) for Indonesian culture and society; Anderson (1990), Bresnan (1993), and Schwarz (1994) for polit- ical history; and Booth and McCawley (1981a, 1981b), Cole and Slade (1996, 1998), Wardhana (1994b, 1997b, 1998a, 1998b), Prawiro (1998), Kenward (1999, 2000), and Stern (2000, forthcoming) for the economy and finance. For analysis of Indonesia in recent years, see also various articles by Seth Mydans in The New York Times (Mydans is bureau chief of the Times for Southeast Asia). 4. Not all wayang is performed with puppets, although this is its most common form.Wayang orang (human wayang) is enacted by people who play the roles of the wayang characters. 5.The names and episodes given here are those used inJavanese wayang perfor- mances of the Ramayana; they differ somewhat from those in theValniki Ramayana and from those used in India today. 6.Vlbhlshana in the Indian versions of the episode. 7. Ravana in the Indian versions. 8. Sita in the Indian versions. 9. In 1999, 14 percent of Indonesians age 15 and older were reported to be illit- erate (World Bank 2002, World Development Report 2002). But statistics based on cen- sus and survey data often use low standards to determine literacy and thus may overstate functional literacy. 10. Other agricultural products include cloves,coconuts, fruits and vegetables, peanuts, soybeans, sugarcane, sweet potatoes, tobacco, and goats, pigs, sheep, and cattle. Both fish- ing and fish farmung are common. 11. "At the time of independence in 1945 there were only 230 pribumi [indige- nous Indonesians] and 107 ethmc Chinese higher education graduates" (Cole and Slade 1996, p. 324). 78 The Microfinance Revolution: Lessons from Indonesia 12. See Schwarz (1994, chapter 1) for an excellent, succinct analysis of the ideol- ogy and pohtics of the Sukarno era.This section draws heavily on that chapter. 13. Indonesia's first (1945) constitution was superceded by its second constitution (1949), which was written following negotiations with the Dutch over a cease fire. The1949 constitution, considered tainted by Dutch influence, was replaced in 1950 by a third constitution written by Indonesians. In 1959 Sukarno abrogated the 1950 constitution and decreed a return to the 1945 constitution. 14. In 1958 a group of military officers set up a rebel government based in West Sumatra. "The Revolutionary Government of the Republic of Indonesia (PRRI) ... .did not seek to break up the Indonesian nation. Rather, its formation reflected the frus- tration of regional military commanders with the armed forces headquarters and the civilian political leadership in Jakarta, and their desire to see a new national govern- ment " (Schwarz 1994, p 13). Against the advice of the U.S. embassy in Jakarta, the rebel government received logistical and mnlitary aid from the U.S. Central Intelligence Agency (CIA), which was concerned that Sukarno was too close to the Indonesian communist party. 15.Jenkins (1984, p 2) points out that although the phrase "the middle way" is commonly attributed to Nasution, it was Professor Djokosutono who named the con- cept "the muddle way." 16. See Higgins (1968 ch. 3); Glassburner (1971); Mackie (1971); Booth and Mc- Cawley (1981b); and Arndt (1984) for analysis of the economuc situation during this period. For articles on many aspects of the Indonesian economy, see issues of Ekono- mi dan Keuangan Indonesia and Bulletin of Indonesian Economic Studies. 17. For various viewpoints about the controversy surrounding the attempted coup, its background, and its effects, see U.S. Central Intelligence Agency (1968);An- derson and McVey (1971); McDonald (1980); Crouch (1988 [1978]); Bresnan (1993); Schwarz (1994); Kahin and Kahin (1995); and Gardner (1997).A recently declassified official U.S. State Department history that describes U.S. policy on Indonesia in the mid-1960s states:"Gradually the [U.S.] embassy came to realize that Indonesia was un- dergoing a full-scale purge of PK.I. influence and that these killings were overlaid with longstanding and deep ethnic and religious conflicts." The history includes a 2 De- cember 1965 memorandum from Ambassador Marshall Green to the State Depart- ment supporting payment to a key civilian member of an organization known as Kap-Gestapu, which was involved in the campaign against the communists that was backed and coordinated by the army.The ambassador's memorandum commented,"The chances of detection or subsequent revelation of our support in this instance are as minimal as any black bag operation can be" (T7he NewYork Times, 28July 2001, p.A3). 18. Schwarz (1994, p.20) considers these the most credible estimates of the num- ber killed, noting that estimates of deaths range from less than 100,000 to more than 1 mnllhon.The Indonesian word amok (uncontrolled, berserk) is often used to charac- terize the events of these months. 19. Sementara, which means temporary, was added to the acronym MPR because the People's Consultative Assembly (MPR) was called to meet at a time when there had been long-postponed elections (McDonald 1980, p. 58).The People's Consulta- tive Assembly meets every five years to elect the president and approve basic policy. In practice, until after Soeharto's resignation in 1998, the MPR had a strong tenden- cy to elect the incumbent and rubber-stamp government policy. In this case "the MPRS was still dominated by Sukarno's appointees.. .but did Soeharto's bidding after hear- ing an unusually contrite Sukarno" (McDonald 1980, p.58). 20. Schwarz (1994, p.46) quotes a 1991 Soeharto statement:"I have always asked God to guide me in each of my tasks. And thank God, to this day...I have never felt that I have failed. And if people think I have been wrong, I think: Who is it who can An Introduction to Indonesia 79 rightfully gauge my mustakes?Who decides if something is wrong? I believe that what- ever I do, after I've asked for guidance and direction from God, that whatever the re- sults, these are the results of His Guidance." 21. "The police and other branches of the mulitary responded to a rise in the crime rate in 1983-85 by executing some 5,000 suspected criminals in various cities through- out Indonesia, all without benefit of trial. In many cases the bodies were dumped in public places to serve as a warning to the community. At the time the military vigor- ously denied responsibility for the wave of mysterious klllings, which was called petrus in Indonesian.. .But some years later, in his autobiography, Soeharto adnutted that pet rus had been a government-sponsored operation from the start" (Schwarz 1994, p. 249). 22. See Bresnan (1993, ch. 3), Cole and Slade (1996), and Stern (forthcoming) for discussion of the technocrats and their role in Indonesia. 23.This section draws fromWardhana 1994b, 1997b, 1998a; Cole and Slade 1996; Prawiro 1998; and Stern 2000, forthcoming. I am grateful to Joseph J. Stern, an expert on the Indonesian economy, for his help in providing econonuc data for this chapter. 24. By 1994 Indonesia was the world's 3'' largest exporter of footwear in the world (after Italy and China), 11 ' largest exporter of garments, 12 h largest exporter of tex- tile fibers, and 13th largest exporter of furniture. 25. PAKDES I is an acronym for Paket 23 December 1987. 26. Paket October 1988. 27. Paket January 1990. 28. See discussion of Kredit Usaha Kecil (KUK) in chapter 9. In April 1997 the requirement was changed to 22.5 percent of the loan portfolio or 25 percent of net bank credit expansion. 29. The number of children that would be born to a woman if she were to live to the end of her childbearing years and bear children in accordance with current age- specific fertility rates (World Bank 1997f). 30.The 1996 GNP per capita of the 15 countries ranged from $770 for Lesotho to $1,390 for Ecuador.The remaining countries (from lowest per capita GNP to high- est) were Egypt, Bolivia, Macedonia FYR, Moldova, Uzbekistan, Indonesia, the Philip- pines, Morocco, Syria, Papua New Guinea, Bulgaria, Kazakhstan, and Guatemala. 31. See Jenkins (1984), Crouch (1988), and Schwarz (1994) for discussion of the role of the military in Indonesia. 32.The MPR is also the only body that can amend the constitution. 33. There were, however, numerous instances of guerrilla fighting within In- donesia, often generated by regional independence movements (as in Aceh, East Timor, and Irian Jaya), with varying degrees of military involvement.There were also guerrilla engagements across neighboring country borders, as in Papua New Guinea 34. Techmcally the joint secretariat of the Golkar functional groups is a govern- ment-sponsored coalition of groups such as agricultural workers, factory workers, women, youth, and businesspeople. 35. In 1993 Soeharto arranged for Information Minister Harmoko to became the first civilian Golkar leader, but Golkar was stlll controlled by the mlitary. 36. For'discussion of Indonesia's political economy, see Bresnan 1993 and Schwarz 1994; see also Cole and Slade 1996, ch. 10. 37. Upon Soeharto's resignation in May 1998 he was succeeded by B.J. Habible, the vice president. In 1999 general elections were held andWahid was elected presi- dent by the MPR. 38. The former president spells his name Soeharto, and this is the spelling nor- mally used in Indonesia. The foreign press, however, often uses the spelling Suharto. 39. See Schwarz (1994, ch. 6) for discussion of this play, the third in a trilogy, and the context in which it was ordered closed. 80 The Microfinance Revolution: Lessons from Indonesia 40. For analysis of the crisis in Indonesia see Wardhana (1998b); Cole and Slade (1998);World Bank (1998b); Radelet (1999); Kenward (1999,2000); and Stern (2000). For discussion of the East Asian crisis more broadly, see World Bank (1998a); Furman and Stiglitz (1998); IMF (1999c, 1999d, 1999e); Lane (1999); Lane and others (1999); Radelet and Sachs (1998,1999); Stiglitz andYusuf (2001). 41. Furman and Stiglitz (1998, p. 71) argue that while countries would do well to improve transparency, this does not inoculate them against a crisis. "In the case of Indonesia, there is a plausible case that... the crisis may have been due to the expecta- tion that corruption was going to be reduced.A substantial fraction of the profits and value of many companies. ..may have been based on their political connections to the Suharto regime and the favors that followed from them. The worrisome news in the fall of 1997 was not that this corruption and nepotism existed. Rather, it was that these connections or favors might dry up, either because of the increased transparency promised by the reforms or because of the increased likelihood that Suharto's regime would end because of his poor health or political vulnerabilhty.The costs of this openness for many investors-rather than corrupt practices by the government-may have played a role in the large outflow of capital that was the central feature of the crisis." 42. Megawati's earlier party was the PDI.When Soeharto realized that she was at- tracting attention and support as head of PDI, he had her removed as party leader. She then formed PDI-P; the P for Perjuangan (struggle) symbolizes the struggle against the Soeharto regime. 43."The International Conference on Indonesia: Strategy for the Stabilization of the Indonesian Economy and its Sustainable Development in the Future," at Mita Kaigisho, Tokyo, 30 March 2001. An Introduction to Indonesia 81 Q Rural Development and Rural Financial Institutions in Indonesia Located on the world's largest archipelago, Indonesia is home to more than 200 million people and hundreds of eth- nic groups, most of whom live in rural areas. The country encompasses a large variety of environments ranging from dense tropical rainforests to terraced rice paddies to tiny islands of coral reef. Its rural areas contain a rich variety of natural resources, including oil, natural gas, and an abun- dance of minerals. The main food crop is rice-grown in irrigated wetlands, drylands, swamps, and even jungles where small plots of land are burned and cultivated in ro- tation. Other food crops-maize, cassava, vegetables, and fruits-are cultivated on rainfed lands, while plantation agriculture, a legacy of the Dutch colonial era, produces ex- port crops such as palm oil, rubber, coffee, tea, and the spices for which the Indonesian islands are famous. With its 82 thousands of islands, the country also has a large supply of fish, although the industry is not fully developed. This chapter examines Indonesia's rural development and its cen- tury-old history of rural financial institutions.The country is home to the world's largest sustainable rmicrofinance system and many local commercial microfinance institutions. But a long-standing tension prevails between those who promote massive rural credit subsidies and those who foster independent commercial financial institutions. A similar tension exists between supporters of member-based financial cooperatives and those who advocate publicly or privately owned rural Indonesia is home to financial organizations that serve the public. Examples are provided of these and other recurrent themes in Indonesia's rural finance. the worlds largest The green revolution came to Indonesia in the 1970s, making pos- sible major gains in the production of rice and other food crops.The financially seW- government invested much of its oil wealth in rural areas, developing sufficient infrastructure, agriculture, and human capital. New methods of rice cul- tivation brought higher yields and multiple cropping that at first ben- microbanking system efited mainly rural elites with valuable irrigated ricelands. But soon the benefits reached smaller farmers as well, and the increased labor re- and many smaller quirements of multiple cropping offset labor displacements caused by commercial new cultivation techniques. Eventually the commercialization of agri- culture and growing off- farm employment opportunities brought sub- microfinance stantial production growth, rapidly increasing monetization, and rising per capita incomes to rural areas. By the early 1990s many areas of rural institutions Indonesia had gained substantially from the green revolution. But considerable rural poverty remained, as did significant regional and local disparities in income distribution and employment opportunities. Rice production in Indonesia increased from 22 million tons of dry unhulled rice in 1976 to 50 million tons in 1996.As incomes rose, so did per capita rice consumption and food consumption generally. In addition, massive investments in education, health, and family planning improved nutrition and health and increased awareness, skills, and ex- pectations. Roads and communication facilities opened up inter-island travel and migration. Middle- and upper-income rural households gen- erally prospered, especially in irrigated lowland regions. Many less de- veloped rural areas also saw economic growth and increasing employment opportunities, though to a lesser extent. But inhabitants of poorer regions, especially on some eastern islands, had fewer op- portunities for development. Rural Development and Rural Financial Institutions in Indonesia 83 Until the late 1980s rural Indonesia was served by few commercial banks other than the unit desas of Bank Rakyat Indonesia (BRI). There were, however, many commercial financial institutions in rural areas, usually owned by provincial or local governments. But these institutions typically operated on a relatively small scale. In the late 1980s banking deregulation led to a substantial increase in the number of banks and bank branches in the country. Some of these operated in rural areas, but usually only in well-developed regions. The market share (in value of outstanding loans) of rural financial In Indonesia a institutions in the financial system hardly varied between 1985 and 1995-it was about 2.5 percent in 1985 and about 2.1 percent in century-old tension 1995. In 1995 the unit desas, which are examined in chapters 11-15, accounted for 48 percent of the number of loans in microfinance exists between those institutions, 63 percent of the value of outstanding microcredit, 76 who promote rural percent of the number of savings accounts, and 81 percent of the value of savings. credit subsidies and The subsidized credit programs implemented by commercial banks and funded by the government and donors generally suffer from high those who foster costs and high arrears.They often do not reach the poor-or reach independent them in ways that do not suit borrowers' needs. And the programs are not sustainable.Worse, they deflect attention (and adaptation) from commercialfinancial Indonesia's long history of commercial, viable microfinance in rural financial institutions, some private banks, and the unit desas. institutions This chapter examines both commercial and subsidized rural fi- nancial institutions and programs. First the century-long develop- ment of Indonesia's commercially oriented People's Credit Banks (Bank Perkreditan Rakyat, or BPRs) is reviewed. This is followed by discussion of rural finance in the 1980s, government initiatives in the 1990s, and development of the BPRs after the 1992 Bank- ing Law. Finally, the chapter analyzes six very different rural finan- cial institutions and programs using data from 1995-98. These organizations-four BPRs (public and private), a Grameen Bank replicator, and a large subsidized credit program linking banks and self-help groups-illustrate well the range of approaches to micro- finance in Indonesia-and their results-as well as the current dilemmas and recurrent themes in Indonesian rural finance. The first three BPRs discussed (two public and one private) are sustainable financial intermediaries that serve both poor and non- 84 The Microfinance Revolution: Lessons from Indonesia poor clients.The fourth BPR, as it operates in one district, provides a classic example of how corruption, politicization, and lack of ac- countability can prevent profitable commercial microfinance while putting poor savers' money at risk.The highly subsidized Grameen replicator offers a rigid, targeted rmicrocredit program in a province that already has hundreds of experienced BPRs offering flexible mi- crofinance programs. As a result the replicator has low capital, low outreach, and 39 percent financial self-sufficiency in 1995 (Seibel and Parhusip 1998, p. 15).The subsidized program linking banks and self-help groups-a large, unnecessarily complicated, inefficient Commercialization program with layers of intermediaries-attained substantial outreach but at very high cost (in fiscal 1995/96 the annual effective inter- of agriculture and est rate would have had to have been 277 percent to fund full pro- gram costs). Moreover, borrowers paid high annual effective interest of-farm employment rates-typically more than 100 percent and in some areas up to 450 resulted in percent. In 1997 the start of the economic crisis and a severe drought caused production growth, substantial hardship in many rural areas.Agricultural growth fell to less in than 1 percent in 1997, a sharp decline from the 3 percent growth of creasing 1996. Marginal areas and poor people were hit particularly hard by monetization, and the drought and the crisis, as employment decreased and prices rose. Social safety net programns were crucial for the rural poor, yet in many rising rural per cases were inadequate. There was considerable variation in food se- curity by province and district, with some provinces facing severe food capita incomes shortages. But the 1998/99 and 1999/2000 rice harvests were good, and production of other food crops improved as well. In addition, es- pecially on some of the Outer Islands where the rural economy is dom- inated by export crops such as palm oil, coffee, and rubber, the plunging rupiah resulted in increased exports and substantial incomes for farmers. But as discussed in chapter 8, as of mid-2001 economic and political difficulties continue in Indonesia. Yet even as Indonesia's financial system collapsed during the cri- sis, the leading commercial microfinance institutions-a number of the BPR systems, BRI's unit desas, and Bank Dagang Bali-remained liq- uid, profitable, and stable. Together these institutions serve about half the county's households-providing a powerful lesson about the sta- bility of sustainable microfinance institutions even in times of severe economic, financial, and political crisis. Rural Development and Rural Financial Institutions in Indonesia 85 In 1976, 80 percent of Indonesia's 136 million people lived in rural areas (World Bank, World Development Report 1980).Although there has since been a substantial increase in the urban population, much of it due to rural-urban rmigration, the country remains predominantly rural. In 1999,60 percent of In- donesia's 207 million people lived in areas classified as rural (World Bank, World Development Report 2000/200 1). But there are substantial differences among the many Indonesian islands, and rural conditions vary considerably. For example, the 1995 population density onJava-where 59 percent of the country's pop- ulation lived on 7 percent of its land-was 900 per square kilometer. In Irian Jaya (now Papua)-where 1 percent of the population lived on 22 percent of the country's land-population density was 8 per square kilometer (Govern- Massive investments ment of Indonesia 1996, p.47). Infrastructure, agriculture, communications, ed- ucation, health, and the like tend to be more developed in rural Java and Balh in agriculture, and less developed in rural Papua, NusaTenggara, and Kalhmantan. Parts of Suma- tra and Sulawesi fall along different points of this continuum. education, health, andfamily planning Geographic and Demographic Diversity improved nutrition There is enormous geographic variation in the Indonesia archipelago. The country has mountains, hills, plateaus, plains, rivers, lakes, and volcanos. It has trop- and health and ical ramforests, swamplands, savanna, elevated coral reefs, areas of intensive irri- gated agriculture on rich alluvial soils, and places where little grows on dry, increased awareness, leached-out earth. Indonesia is rich in natural resources, with oil and natural gas, tin, coal, bauxite, nickel, gold, silver, manganese, copper, and other minerals. skills, and Asian land vertebrates are found in the western islands of Sumatra,Java, Bali, and Kalimantan, which had land connections to the continent when the Sunda expectations shelf, part of the Asian continental shelf, was largely exposed during the Pleis- tocene period. Similarly,Australian animals moved overland into IrianJaya and other eastern Indonesian islands when much of the Sahul shelf, part of the Aus- tralian continental shelf, was exposed. Alfred Russel Wallace, the British natu- ralist who explored Indonesia in the mid-19th century, found that a line drawn through the Makassar and Lombok Straits (east of Ball and west of Sulawesi) marked the division between Asian andAustralian fauna. It was later found that Asian and Australian fauna coexist in a transitional area in the central Indone- sian islands' around Sulawesi, the largest of a group of islands that lies between the two great shelves. Rainfall is highest in the western part of the country and decreases toward the east; the typical range is from more than 200 inches a year to less than 40 inches.The country's rural population is concentrated in areas where there is recent or continuing volcanic activity.Young volcanic soils are highly fertile and permit the growth of the dense rural populations found in Java and Bali. Stated on Indonesia's national emblem are the words binneka tunggal ika (diverse but united).The country is home to several hundred ethnic groups, most with mutually unintelligible languages.Java is populated by a Javanese 86 The Microfinance Revolution: Lessons from Indonesia majority, a large minority of Sundanese primarily in WestJava, and several small- er ethnic minorities. Moving from west to east across the archipelago, small ethnic groups and small speech communities become more pronounced until in Irian Jaya there are small bands with separate languages spoken by fewer than 100 people (Koentjaraningrat 1975b, p. 54). But except for some elderly people, nearly all Indonesians now speak the national language, Ba- hasa Indonesia. Indonesia is heavily Islamic, with Muslims making up nearly 90 percent of the population; the largest minorities are Protestant, Roman Catholic, Hindu, and Buddhist. In studying the cultures of Indonesia, Koentjaraningrat (1975b, pp. 57-60) placed the hundreds of ethnic cultures into four general categories: 6 rural microfinance * Small groups of isolated peoples whose livelihoods are based on shifting cul- tivation, found mainly in the smaller eastern islands and in Irian Jaya. institutions are used * Interior peoples whose livelihoods are based on swidden2 or irrigated agri- culture, with rice as the main crop. These groups live in village communi- to illustrate the ties and interact with people in nearby administrative towns. Examples are found in North Sumatra, Central Sulawesi, and the eastern islands. range of approaches * Coastal peoples who cultivate rice as their main crop.This is a heterogeneous mix of maritime peoples descended from Malays, Javanese, South Indians, and results in Arabs, Persians, Portuguese, English, Dutch, and Chinese. * Interior settlements of people whose subsistence is based on wet rice agri- Indonesian rural culture.These village communities, characterized by significant social strat- ification, are located most notably in Java and Bali. Elaborate agrarian-based finance royal courts developed in some of these areas.TheJavanese gentry have dom- inated Indonesia's government and development since independence. However, many changes have occurred since Koentjaraningrat's 1975 clas- sification. The opening of road travel throughout Indonesia, rural-urban mi- gration, inter-island mugration, the green revolution, the growth of overseas employment opportunities, and the explosion of infrastructure, communica- tions, and education have changed the country greatly. Koentjaraningrat's cat- egories remain relevant and important for understanding Indonesia, but the dynamics of these groupings are shifting. As James J. Fox (The Independent Monthly, 17-19 February 1990, p. 18) put it in a discussion ofJava: But, in truth,Java is a single settlement.What makes it a single settlement is its incredible flow of traffic. The Dutch left the island with a peasant population very much tied to the land in separate subsistence-oriented villages. But over the last 20 years Mitsubishi and Mercedes Benz have managed to open up the vllUages; Mitsubishi with its mini-van, the Colt, that could- and did-reach the most isolated villages, and then Mercedes with huge buses that take passengers across the island in less than a day. Rural Development and Rural Financial Institutions in Indonesia 87 Rural Development Given its geographic and cultural variety, it is not surprising that many types of agriculture are found throughout the Indonesian archipelago.The country supports a vast range of food types and other agricultural products. Varieties of agriculture The main rice supply comes from sawah (irrigated riceland) agriculture. Sawah is a field that is meant to be flooded; it is surrounded by a small dike, and if it is on the mountainside, the slopes must 1hen thefinancial be terraced. Sometimes the flooding is left to the rains, but the result of this procedure is unpredictable, for rainfall might be system collapsed too much, too Iltde, or poorly timed. Usually ditches or bamboo pipes lead water to the sawah from a nearby stream. during the crisis, -Nell 1973, p. 43 BRls unit desas, Sawah, which refers to wet-field cultivation on irrigated rice fields, regardless of the source or quality of irrigation, is the environment in which most rice is BDB, and many grown in much of Java, Madura, Bali, and Lombok, parts of Sumatra and Su- lawesi, and in scattered parts of other areas. People's Credit Estate (plantation) agriculture was introduced to Indonesia by its Dutch col- onizers. Plantation crops include palm oil, rubber, sugarcane, tobacco, coffee, Banks (BPRs) tea, and spices.These crops are grown on large estates but most are also culti- vated by smaliholders. Depending on the crop, plantations are found in both remained profitable highlands and lowlands. Timber production is concentrated in the Outer Is- lands (outside Java and Bali). and stable Another type of cultivation is known as tegalan (dryland) agriculture; this refers to land that is normally continuously cropped but is not irrigated. Rice, maize, cassava, sweet potatoes, peanuts, soybeans, vegetables, and fruits are common in rain- fed agriculture. Pekarangan (mixed garden) cultivation is usually found in small house- hold or village plots. In ladang (or swidden) agriculture, an ancient technique, a small forest area is burned and rainfed crops are grown in the ashes.The cultivat- ed area is productive for a few years but then yields declne rapidly and the cul- tivator moves on, burning a new plot on which to plant yams and other root crops, rice, and other crops. After some years the original plot is often cultivated again, and the cycle begins anew. Ladang cultivation is common in parts of Sumatra, Kali- mantan, Sulawesi, and Maluku and other eastern islands. In 1995,23.9 percent of the country's land was in estate cultivation, 19.6 percent was in ladang, 16.5 per- cent was in wooded areas, and 14.6 percent was in sawah. The rest was used for housing and other purposes (Government of Indonesia 1996, p. 151). Rice, the primary staple food of Indonesia, is grown on irrigated wetlands, on drylands, in swamps, in garden areas, and in shifting ladang cultivation. An- imal husbandry is common, especially for goats, cows, buffalo, poultry, and (in non-Muslim areas) pigs. Fish come from both salt and fresh waters and from 88 The Microfinance Revolution: Lessons from Indonesia fish cultivation in brackish water ponds, fresh water ponds, cages, and flooded sawah fields.3 Shrimp cultivation has also been developed in recent years. Cultivation on Java in the early 20th century Thomas A. Fruin, president of the AlgemeeneVolkscredietbank (AVB), a pre- cursor of BRI, wrote an extraordinary manual for the AVB in 1935. The Pro- visional Manualfor the Credit Business of the General Popular Banke was based on decades of work that had brought Fruin and others into close contact with In- donesian farmers.The AVB and its predecessor, the Volkscredietwezen (Popu- lar Credit System) early credit institutions established in Indonesia by the Dutch colornal administration to provide banking services to indigenous Indonesians- are discussed in chapter 11. What is of interest here is the analysis of indige- OnJava in 1995, nous farming in Indonesia that Fruin constructed to learn how best to provide and recover loans from different kinds of farmers. Many of the inputs into the 59 percent of manual came from articles by Fruin and others in the monthly journal Volks- credietwezen, published under several names beginmning in 1913, but from 1931 Indonesia's on known as Volkscredietwezen. Many of today's rural banks could learn from Fruin's section on types of population lived on loans in indigenous farming on Java (Fruin 1994 [1935], p. 132). He begins: 7 percent of its land. If one examines the credit business of any former popular credit bank one is often struck by the large measure of uniformity In IrianJaya 1 revealed; this is due to the fact that no more than one or two different types of loans can be found. Further study then reveals percent of the that exactly the same types were used by many similar institutions. In other words, insufficient account was taken of the different population lived on kinds of agriculture within a region or of the differences between regions. If this situation is to be improved, each local 22 percent of the office will have to investigate, district by district, what the normal types of farming are. It is not therefore sufficient to know land what the principal crops are in the region, to consult monthly figures on planting and harvesting of each crop, to have some idea about the kinds of fruit trees which are grown, about trade or about employment; one must consider farming as an organic whole, i.e. find out how crop rotation operates in particular types of farming, how large such a farm has to be to support the farmer and hls family, which crops are intended for sale, which exclusively for food and which for both, and in the latter case in what proportions they are divided, whether and to what extent such a business usually derives cash income from horticultural and orchard produce, whether cottage industries or coolie work normally provide additional income, and so on. The KUPEDES loan instrument, created much later for BRI's unit desas, reflects Fruin's lessons. KUPEDES provides loans in a wide range of amounts, Rural Development and Rural Financial Institutions in Indonesia 89 offering 36 combinations of repayment and maturity options so that borrow- ers can select loans appropriate for their particular needs. In the 1920s and 1930s Fruin painstakingly analyzed different types of farm- ing from the viewpoint of credit extension (see chapter 11). These included cycles of rice cultivation on irrigated wetlands and on higher-elevation rain- fed drylands, crop rotations of rice and secondary crops such as maize, cassava, and vegetables, and cultivation of both cash crops such as tobacco and coffee and of crops that may be for subsistence or for sale (rice, cassava, fruits). Although Fruin's purpose was to examine farming in order to establish credit procedures, much can also be learned from his manual about the various types of farming and the marketing of farm produce that were then common in different areas The main rice ofJava.While there have been considerable changes in farming in Java (and else- where in Indonesia) as a result of the green revolution, Fruin's principles of agri- supply comes from cultural credit and many of his observations aboutJavanese agriculture remain relevant today. irrigated ricelands. The green revolution But rice is also In the mid-1960s international agricultural research began to demonstrate that new high-yielding technology could significantly increase agricultural grown on rainfed productivity. Based on high-yielding seed varieties, substantial growth in the use of chemical fertilizers, insecticides, new cultivation and management tech- drylands, in swamps, niques, expanding irrigation, and in some cases new agricultural machinery, the green revolution resulted in a large-scale shift to commercial cultivation of food- and inforest areas grains in many developing countries. In Indonesia-as in most countries-the results varied by region, by crop, and by farmers' access to land, water, labor, using shifting credit, and marketing facilities. cultivation Cultivation patterns. The green revolution made possible major gains in the production of rice and other food crops. Increases in rice production have come from both intensification, primarily on Java and Bali, and extensification, pri- marily on some of the Outer Islands. Increases in rice production from inten- sification have come from higher yields and increased cropping intensity (two or three crops a year instead of one). Other areas that have achieved significant rice intensification are North, West, and South Sumatra and South Sulawesi. In 1996 average yields in sawah paddy production were 54 quintals4 of dry un- hulled paddy (gabah) per hectare on Java and 40 quintals per hectare off Java, with a range in average quintals per hectare from 28 in Irian Jaya and East Timor to 55 in EastJava. For Indonesia as a whole the average yield for sawah paddy production was 47 quintals of gabah per hectare, while the average yield for dryland paddy cultivation was 22 quintals of gabah per hectare (Government of Indonesia 1996, pp. 164-65). In contrast to intensification, rice extensification occurs by bringing new lands under cultivation. Beginning in the 1970s, far more new land was culti- vated with rice off Java than on Java. But much of the land used to extensify rice cultivation outside Java is only marginally suitable for rice growing (see 90 The Microtinance Revolution: Lessons from Indonesia chapter 11), andJava produces more than 60 percent of the country's rice. Over- all, rice production in Indonesia increased from 22 million tons of gabah in 1976 (when the population was 136 million) to 50 million tons in 1996 (when the population was 197 million). But the demand for rice increased faster than the population.As incomes rose, so did per capita rice consumption. Most of those whose main diet had been cassava or maize switched to rice when they could afford to do so. In 1996 Indonesia produced 17 mnllion tons of cassava, of which 55 percent came from Java, and 9 million tons of dry loose maize, of which 61 percent came from Java (Government of Indonesia 1996, pp. 166-67). ThusJava remains Indonesia's principal source of rice and other staple food crops.The Outer Islands contribute much of the country's export crops, such as palm oil, coffee, rubber, and timber, as well as oil, natural gas, and minerals. The green revolution Results. The effects of the green revolution in Indonesia, especially for rural significantly Java, have been much debated (see Mannmng 1987). During the 1970s many so- cial scientists reported that the commercialization of agriculture onJava was caus- increasedfood ing increasing inequalities among the rural population (Penny and Singarimbun 1973; Collier and others 1974; Budhi-santoso 1975; Hinkson 1975; Sinaga and production. But the Collier 1975; White 1976, 1979; Palmer 1977b; Gordon 1978; Hart and Sisler 1978; Sinaga 1978; Sayogyo 1982 [1973]). It was argued that: results varied by * The benefits of the new agricultural technologies reached primarily large farmers' access to farmers who had access to irrigation, credit, and high-yielding inputs ap- propriate for well-irrigated sawah. land, water, labor, * The new methods, especially for weeding, harvesting, and rice hulling, low- ered costs for cultivators but decreased employment opportunities for laborers credit, and markets (Collier and others 1974; Collier, Soentoro, Hidayat, andYuliati 1982; Bud- hisantoso 1975; Sinaga and Collier 1975; Singarimbun 1976; Sinaga and oth- ers 1977; Collier 1978; Sinaga 1978; Manning 1987, ch. 4). * Rural elites and urban residents were consolidating the fragmented lands that poor farmers could no longer afford to cultivate. * New agricultural methods and changing forms of labor contracts weakened the patron-client and communal ties believed to demonstrate Geertz's (1963) concept of a cultural norm of"shared poverty" (Collier and others 1974; Bud- hisantoso 1975; Ruttan and Binswanger 1978; Collier 1981; Hart 1986a, 1986c; Manning 1987, ch. 5).5 In addition, attention was drawn to the increasing replacement of cottage industry products by manufactured goods. There is some truth to all these statements. But the dismal picture of in- creasing income inequality, poverty, and polarization portrayed in much of the literature of the 1970s and early 1980s was oversimplified and did not provide an accurate representation of rural realities.6 Thus while the new methods for cultivating rice benefited the rural elites earlier and more extensively than was the case for small farmers, some farmers with small plots also gained from the higher yields and multiple cropping made Rural Development and Rural Financial Institutions in Indonesia 91 possible by the new technologies. In addition, the increased labor requirements of multiple crops offset much of the labor displacement and employment dis- tribution problems caused by the new cultivation techniques and labor contracts. The government's extensive investment of oil wealth in rural areas in the 1970s-in agriculture, infrastructure (irrigation, roads, markets), health, fami- ly planning, and education-had generally succeeded, by the 1980s, in improving agriculture, creating employment, and raising rural incomes. But the distribu- tion of these benefits varied considerably by region, locality, social and economic status, occupation, and gender. Increasing employment generation included both off-farm work in rural areas (trade, construction, food processing, services) and employment opportunities for rural workers in urban areas (both in informal Although large petty trade, transportation, construction, and small-scale manufacturing, and in then-emerging formal sector labor-intensive export activities). farmers often control Although large farmers often control more irrigated riceland (sawah) than they own, sawah is generally not concentrated in large holdings.This outcome more irrigated is probably due to the multiple employment activities that enabled many small farmers to retain their lands, and to the better education and increasing investment riceland (sawah) opportunities available to large farmers who were able to diversify their eco- nomic investments. than they own, Most middle- and upper-income rural households have prospered, especial- ly in the irrigated lowland regions ofJava, Bali, and Sulawesi.Two or three crops sawah is generally a year are now normally cultivated in these regions, creating additional employ- ment. In 1996 agriculture accounted for just 17 percent of GDP, yet employed not concentrated in 44 percent of the Indonesian workforce (Government of Indonesia 1996, p. 45). Yields have increased considerably, and crops are frequently sold in bulk for cash. large holdings Widespread improvements in roads and related growth in transportation- popularly called the "Colt revolution"7-m combination with the greater em- ployment opportunities generated by labor-intensive exports, expanded income sources beyond the village. This, in turn, increased remittances and transfers of funds to rural areas. Overall, irrigated lowland villages saw economic growth and increasing mon- etization, much higher crop yields, added employment opportunities for the many landless, an emerging pattern of multiple income sources (including new off-farm activities in manufacturing, services, and trade), and higher mobility (see Keyfitz 1985 and Fox 1990 for accounts of the rapid development of two East Java villages over time).8 Many less developed rural areas were similarly characterized, though to a lesser extent, by economrc growth, rising employment opportunities, multiple income sources, and increasing monetization. But the inhabitants of poorer re- gions, especially in parts of Kalimantan and some eastern islands, had fewer op- portunities for development.9 In 1996 the rural population below the poverty line varied across provinces from 3.5 percent (Bali) to 33.1 percent (East Timor) (Government of Indonesia 1996, p. 574). 10 Special programs, public and private, were instituted to reach the "left behind" (tertinggal people and villages. Even in Java some poor villages remain left behind. In the mid-1980s I talked 92 The Microfinance Revolution: Lessons from Indonesia with a man who worked in Jakarta as a laborer for a microenterprise that sort- ed and sold recyclable waste purchased from ragpickers. From this work he was able to send about $20 a month to his family in a poor Central Java village. I asked him what was the main difference between living in the city and living in his village. He answered without hesitation, "In Jakarta I can eat every day and as much as I want. I can eat until I burst. In the village I might eat today but I won't know today if I will eat tomorrow." Rural development during the Indonesian crisis In 1997 the economic crisis, along with the devastating drought, eroded and in some cases erased gains that had been made in rural development. Particu- larly hard hit were marginal areas and the poor. The social safety net programs Employment was discussed in chapter 8 were crucial for the rural poor, but many of the poor were not reached, were not adequately provided for, or were not reached at the generated by the time of greatest need.As a result the number of people below the poverty line increased from 11 percent in 1996 to an estimated 20 percent in 1998. Some labor requirements of rice eaters switched back to cassava, maize, and sweet potatoes. Prices rose, em- ployment dropped, some children left school, health care and medicine were multiple crops and not always available or affordable, and the gap between prevailing conditions and people's expectations widened. rural off-farm work, But the 1998/99 and 1999/2000 rice harvests were good, and production of other food crops-maize, cassava, soybeans, potatoes, and peanuts-improved and by opportunities as well. In addition, especially in some outer islands where the rural economy is dominated by export crops such as palm oil, rubber, and coffee, the precipitous for rural labor in fall of the rupiah resulted in increased exports and substantial incomes for farm- ers. By the end of 1999 the percentage of people below the poverty line was back urban areas to the pre-crisis 1996 level of 11 percent. Overall, the impact of the crisis on the poor was generally more severe in urban areas than in the countryside. Even with the substantial difficulties Indonesia faces as of mid-2001, there is no question that its rural poor have made strong gains over the past three decades. Although some rural inhabitants are still below the poverty line, most have benefited from more diverse employment opportunities, increased real in- comes, and better nutrition, health, education, and housing. Huge governmnent investment in rural areas during the Soeharto era-in agricultural technology, infrastructure, education, health and family planning, communications, and food processing and other off-farm rural activities-laid a firm foundation for continuing development. When Indonesia achieves the political stability and will to implement needed economic and financial reforms, these rural assets can serve as a springboard for future development. Developing Rural Financial Institutions Indonesia has a long history of rural financial institutions, dating to the late 19th century when the early Volksbank (People's Bank) and Afdeelingsbank (Dis- Rural Development and Rural Financial Institutions in Indonesia 93 trict Bank) were introduced by the Dutch colonial admimstration. After inde- pendence, different kinds of rural financial institutions continued and many new ones were developed, variously called rural banks, village banks, market banks, people's banks, and the like.The generic term for these small local banks is Bank Perkreditan Rakyat (People's Credit Bank, or BPR), but as will be discussed, this term has multiple meanings. BPRs are small independent rural financial institutions, widely scattered. Many offer both savings and credit facilities. Until recently, with the publica- tion of Steinwand (2001), relatively little had been published about most of them-with the exception of the Badan Kredit Kecamatan (BKK) of Central Java and the Badan Kredit Desa (BKD) of Java and Madura, and a few other Some villages well-known systems.This chapter and others in this book (chapters 11, 14, and 17) explore selected Indonesian BPRs. Readers interested in more in-depth remain poor. As one study of Indonesia's nearly 9,000 People's Credit Banks are referred to Stein- wand (2001), which begins with the BPRs' European antecedents in earlier cen- man said, "In turies and traces their development in Indonesia from their beginmngs during the Dutch colonial period up to the current Indonesian crisis.11 Jakarta I can eat In addition to the BPRs and BRI's unit desas, Indonesia's rural areas have long had ubiquitous rotating savings and credit associations (ROSCAs) as well every day and as as nonrotating associations (ASCAs) and other forms of savings and loan asso- ciations. But most of these informal associations have had limited outreach. In- much as I want. In formal commercial moneylenders provided most of the credit in rural areas. And most rural Indonesians kept most of their savings at home. In the second my village I won't half of the 1980s, however, things began to change. In analyzing the development of mucrofinance in rural Indonesia, certain know today if I will recurrent themes emerge in both colonial and Indonesian contexts. These themes underlie much of the discussion throughout volume 2. (Most are also eat tomorrow" relevant, in various ways, to mucrofinance in other countries, as discussed in vol- ume 3.) * Policies andpolitics. Indonesia has a century of experience with different kinds of unsubsidized, profitable microfinance institutions. Many different kinds of People's Credit Banks have been developed without assistance from donor agencies (donors have been involved only on rare occasions). Initial funding came from wealthy individuals or local governments. Equity was created from retained earnings. But Indonesia also has a history of large-scale subsidized rural credit programs. A long-standing tension has continued, generation after generation, between supporters and opponents of credit subsidies (see Schmit 1991).There is a similar dynamic between those who support mem- ber-based cooperatives and those who advocate nonmember-based financial institutions. Examples are provided here of the same political tensions in dif- ferent periods-tensions that have resulted in both policy oscilations over time and simultaneous implementation of contradictory policies. Generally the ten- dency has been for the Ministry of Finance and the National Development Planning Board (Badan Perencanaan Pembangunan Nasional, or BAPPENAS) 94 The Microfinance Revolution: Lessons from Indonesia to support commercial microfinance, while the Ministries of Cooperatives and Agriculture have frequently advocated subsidized credit programs channeled through village cooperatives. Both sides were powerful, and both approach- es reached every Indonesian village. (It should be remembered that injavanese philosophy, opposites are part of the same whole.) * Financial intermediation. Many People's Credit Banks have been (and are) fi- nancial intermediaries, not microcredit institutions. Loans are generally fi- nanced from savings and equity, and in some cases with commercial bank loans. From the beginning of the People's Credit Banks in the 19,h century, there has been a strong view in Indonesia that self-sufficient financial inter- mediaries work better than rmicrocredit programs. * Ruralfinancial institutions: public and private. Most of Indonesia's commercial The impact of the People's Credit Banks, as well as BRI's unit desas, are public enterprises, pro- viding a sharp contrast to the currently dominant international view that po- recent crisis on the tentially self-sufficient microfinance institutions are best located in the private sector. The Indonesian experience has demonstrated that commer- poor was generally cial mucrofinance can flourish in both sectors. * Learningfrom experience. There has been a long history in Indonesia that, when more severe in urban institutions do not work well, lessons are learned and new models are de- veloped-whether it be conversion from cooperative to noncooperative banks areas than in the during the colonial era or the BRI unit desas' transformation from subsi- dized credit delivery to commercial financial intermediation toward the countryside end of the 20h century. Examples are provided throughout this volume. These themes are as relevant for Indonesia today as they were a century ago. But as will be discussed, much has been accomplished in the interim. European background European banking began more than 500 years ago in Italy with the provision of credit to the aristocracy and the church, and with payment services for long- distance trade. But the Italhan banks did not provide financial intermediation. Between 1490 and 1520 German and Swiss banks developed into financial in- termediaries and took the lead in European finance.As Bergier (1979, quoted by Steinwand 2001, p. 49) put it, "The Italians had given banking its name and its instruments.The Germans gave it its place in the economy and society." By the 181' century the development of microbanking had begun. The structural changes from an agricultural based society to an industrial one and the rapid demographic growth after the end of the extended European wars (1618-1648) led to a tremendous increase in poverty in Europe. One response was the establishment of various forms of microfinance institutions during the 18th and 19th centuries...Their common goals were to fight poverty and to release the poor from the grasp of the moneylenders.They can be grouped into four categories: credit- Rural Development and Rural Financial Institutions in Indonesia 95 focused charity funds.. .community based funds which started with donated funds but developed over time into fully fledged commercial intermediaries; savings banks...and the cooperative banks. -Steinwand 2001, p. 51 Indonesia's early people's banks and district banks were introduced by the Dutch at the end of the 19th century.The founders of these banks were direct- ly influenced by the growth of riicrofinance institutions in 19th century Europe. A few examples of the European antecedents will be useful (for further discus- sion see Steinwand 2001, chapter 3). One was the Irish loan funds started in 1720 In Indonesia there byJonathan Swift, author of Gulliver' Travels, which provided small loans to poor traders (see Hollis and Sweetman 1997,1998 and Steinwand 2001).This model has been a strong was widely copied by other wealthy individuals, and by 1840 the loan funds reached about 20 percent of Irish households. 12 Another model was the savings viewfor over 100 banks that began in Germany in 1778 and spread to many European countries in the early 19th century. As Tilly (1994, p. 305) comments: "The chief motive years that se!f- [for the development of the savings banks] was to encourage thrift and economic self-dependence among the poorer segments of the population. ..and not the mo- sufficient financial bilization of financial resources to finance investment." A third important influence was the cooperative banks that began in two intermediaries serve regions of Germany in about 1850. Schultze-Delitzsch started his first credit association, mainly urban-based, in the northeast of the country, and Freidrich the poor better than Raiffeissen began his first savings and loan association, mainly rural-based, in the southeast.The two models had somewhat different approaches.Among the microcredit programs differences, the Schultze-Delitzsch model incorporated limited liability for members (to attract deposits from clients who were not poor), whereas in the Raiffeisen model members' liability was not limited. Another difference is il- lustrated by Wolff (1893, p. 64; quoted in Steinwand 2001, p. 56): "Schultze- Delitzsche throughout put the lender's interest foremost, Raiffelsen the borrower's." It was Raiffeisen's approach to cooperatives that eventually became the model for most European cooperative banks.13 In the Netherlands the first savings banks were established in 1817 with the explicit aim of assisting low-income people. But these banks were not effec- tive in reaching poor clients (Steinwand 2001, p. 59). In the 1880s, however, an agricultural crisis led to the formation of a Netherlands government com- mission to advise on improvements for the agricultural sector. The commis- sion recommended establishing local cooperative banks modeled on the German Ralffeisen banks.The first such bank in the Netherlands was officially opened in 1897, and the banks, which eventually became known as Rabobanks, mul- tiplied. By 1920 there were some 1,250 Rabobanks in the Netherlands.14 The early Indonesian people's banks The Dutch colonial administrators in Indonesia introduced various forms of People's Credit Banks beginning in 1895.The establishment of a series of banks 96 The Microfinance Revolution: Lessons from Indonesia that eventually became Bank Rakyat Indonesia (BRI) began that year (see chap- ter 11). But it was not only BRI that was developed by Dutch colonial offi- cials; many district banks and lumbung desas (iterally "village granaries," but known as paddy banks) were begun at about the same time in Java and Madura. The original plan was to construct a cooperative rural banking system in Indone- sia based on the Raiffeisen model.The lumbung desas would form the village- level credit cooperatives, and the district banks would become the second tier of the cooperative structure. District banks. Seventy-one district banks disbursed an average of 755 loans in 1926. Data for 1929 show that nonperforming loans (loans in default) in the district banks were 3 percent of the amount of outstanding loans.15 More The Indonesian than half of the district banks' deposits were from (predominantly European) time deposits in 1913, but by 1926 time deposits had declined to 23 percent experience of total deposits, while 55 percent of the banks' deposits came from (mainly Indonesian) savings accounts (Steinwand 2001, pp. 73-83). demonstrates that These financial intermediaries soon became financally self-sufficient. In 1911 there were 75 district banks, of which 67 (89 percent) were subsidized. But by commercial 1926 only 6 of 90 banks (7 percent) were subsidized (Suharto 1996, quoted in Steinwand 2001, p. 80). microfinance can Lumbung desas. By 1910 there were 12,542 lumbung desas in Java and Madu- flourish in both ra; the number gradually declined to 5,451 by 1940.The original lumbung desas were supposed to be funded by members' donations paid at the end of the Mus- public and private lim fasting month. But this approach did not work well, and a modified sys- tem was introduced that did not follow the Ralffelsen model so closely. The sectors modified lumbung desas were not strictly member-based: they had external fi- nancing as well as members' contributions, and grew much faster and performed better than those developed using the earlier model. Originally the lumbung desas provided (and collected) loans in paddy.Their main purpose was to level out seasonal fluctuations in rice supply. But by the early 1900s some of the lumbung desas in the more developed regions ofJava began to lend and collect loans in cash.As a result of this development, in 1904 the colonial administration introduced bank desas (village banks) operating on a currency basis. The lumbung desas achieved wide and deep outreach in Java and Madura but did not work well in many outer islands where paddy farming was less im- portant. By 1918 there were no lumbung desas outside Java and Madura. But in 1926 the lumbung desas reached nearly 1.3 mullion borrowers, 16 percent of the approximately 8 million households inJava and Madura.The loans pro- vided by the lumbung desas were short, small, and profitable. However, outstanding loans as a share of total lumbung desa assets were 27 percent in 1913, and only 2 percent in 1926. "Thus 98 percent of the Lum- bung Desas' funds were idle and were not used for making loans to villagers! Of course, such a low rate of investment was only possible because the Lum- Rural Development and Rural Financial Institutions in Indonesia 97 bung Desas were owned by the villages and funded to 99% by equity" (Stein- wand 2001, p. 91).The source of the lumbung desas' equity was retained earn- ings. Because paddy stored for a number of years deteriorates in quality (and sometimes in quantity, as it can be eaten by rats), the lumbung desas sold most of their paddy stock and deposited the funds in the district banks. Bank desas. When the bank desas opened in 1904 and began lending in cash, there was considerable demand for these loans. The main sources of the bank desas' startup funds were the government, district banks, local lumbung desas, and repayable villagers' shares. In 1926,4,754 bank desas onJava and Madu- ra had 941,000 borrowers, and outstanding loans were equal to 56 percent of Large-scale total assets (down from 74 percent in 1913 because the banks' equity had grown faster than their loan portfolios; Steinwand 2001, p. 97). Default rates were gen- microcredit began in erally low, averaging about 1 percent, and overhead costs were minimal because the banks operated through the offices or residences of the heads of the vil- the 1 8th century with lages where they were based. These bank desas were highly profitable. During the first two decades of the 20th century they yielded an average 50-58 per- the Irish loanfunds cent annual net return on the average outstandmg loan amount (Steinwand 2001, pp. 96-97).16 started in 1720 by The Bank Desas were fully initiated by the government and Jonathan Swift, funded with external resources. Hence their establishment finally marked the end of .. [the] idea to establish a cooperative author of Gulliver's system based on Raiffeisen's self-help principles at the village level. . .During the following decades plans to transform the Bank Travels Desas and the district banks into a two tier cooperative system surged from time to time among the colonial officials, however they were never realized. -Steinwand 2001, pp. 92-93 In 1907 the lumbung desas and the bank desas were regulated under a de- cree for Badan Kredit Desas (BKDs).These regulations were modified several times until 1929; the 1929 version is still valid (see chapter 14). Bank Perkreditan Rakyat (People's Credit Banks, or BPRs) The term Bank Perkreditan Rakyat (BPR) was introduced by Bank Indone- sia in 1978 as a generic term for people's banks.The term includes the village- owned BKDs (the lumbung desas and the bank desas), the Lembaga Dana dan Kredit Pedesaan (LDKPs, rural financial institutions owned by provincial, dis- trict, or subdistrict governments, often in combination, or by villages), and other small financial institutions variously called People's Credit Banks, rural banks, village banks, market banks, cooperative banks, and others. Such institutions vary widely in ownership, activities, legal status, supervision, management, and size. After the 1988 financial reforms of PAKTO 88, new secondary banks were es- tablished; these were also called BPRs. PAKTO 88 stipulated that BPRs meet 98 The Microfinance Revolution: Lessons from Indonesia certain requirements, including paid-up capital or savings of 50 million rupi- ah ($28,885 in 1988). Existing BPRs were given two years to fulfill the new requirements. But in 1989 the PAKMAR reform package extended indefinitely the time limit for preexisting BPRs to meet the new regulations, essentially grandfathering them from the PAKTO 88 requirements. However, the Banking Law of 1992, which recognized only two types of regulated banks (commercial banks and BPRs serving the rural population), in effect reversed PAKMAR in this regard, stating that rural financial institutions are required to meet the BPR criteria. But following the crisis of the late 1990s, the requirements for rural financial organizations to meet BPR regulations were substantially changed. As a result the future status of rural financial organizations that are not licensed BPRs re- People's Credit mains unresolved (see the discussion later in this chapter). Today the term BPR has two primary meanings. It refers to rural finan- Banks were cial institutions that meet the criteria specified in the 1992 Banking Law (li- censed BPRs). But it is also used for the nearly 9,000 People's Credit Banks of introduced in all kinds that exist today in Indonesia (generic BPRs), most of which do not meet the new criteria.At independence, Indonesia had several thousand small Indonesia by rural financial institutions, but the district banks had been merged in 1934 into the AlgemeeneVolkscredietbank (General Popular Bank), a precursor of BRI Dutch colonial (see chapter 11). By 1966 there were more than 4,000 bank desas and nearly 2,700 lumbung desas, according to the central bank (in later Bank Indonesia administrators in the statistics these institutions were grouped together under the category BKDs). The LDKPs, however, were not included in central bank statistics until after late 1 9th century the 1988 financial reforms. The LDKPs developed further during the 1970s and 1980s.The BKDs (which are discussed in chapter 14, as they are supervised by BRI) continued to pro- vide financial services to many villages in Java and Madura, but they were not permitted to serve other areas of the country or to open new units.This is gen- erally attributed to the fact that the village-owned BKDs, although they do not operate as originally planned on the basis of Raiffeisen self-help principles, are relatively autonomous.The rural financial institutions favored by President Soe- harto were those under direct government control: the BRI unit desas, the LDKPs under provincial government authority, and the state-controlled (subsidized and loss-making) village cooperative system (Koperasi Unit Desa, or KUD). Some of Indonesia's People's Credit Banks perform exceptionally well. Oth- ers perform poorly, and many are in between. But BPRs that do not perform well have generally been allowed to die out. This is a lesson that still needs to be learned in many other countries. Rural Finance in the 1980s In 1985 BRI's unit desas accounted for about 1.0 percent of the value of out- standing loans in the Indonesian financial system; all other rural financial in- Rural Development and Rural Financial Institutions in Indonesia 99 Table 9.1 Outstanding loans in Indonesia's financial system, 1985 Amount Share of Type of Billions of Millions of financial system financial institution rupiah U.S. dollars (percent) Banks 22,933 0 20,384 9 96 83 Bank Indonesia (direct credits) 964 0 856 9 4 07 State banks 15,145 0 13,462 2 63 95 Other commercial banks 4,1060 3,6498 17 34 Foreign banks 1,073 0 953 8 4 53 From 1900-20, Development banks 6400 5689 2 70 Savings banks 1,005.0 893.3 4 24 early village banks Nonbank financial institutions 162 0 144 0 0 68 Rural financial institutions 589 1 523 6 2 49 onJava and Madura BRI unit desas 229 0 203 6 0 97 Secondary banks 214 3 190 5 0 90 earned an average Bank Pasar 1930 171 6 081__ Village banks 19 0 16 9 0 08 annual net return of Paddy banks 2 3 2 0 0 01 Pawnshops 64.8 57 6 0 27 50-58 percent of Other rural nonbank financial institutions 81 0 72 0 0 34 BKKs, KURKs, and so on 31 0 27 6 0 13 KUDs 500 444 021 the average Total credit outstanding 23,684 1 21,052 5 1000 outstanding loan Note BRI stands for Bank Rakyat Indonesia BKK stands for Badan Kredit Kecamatan, or Subdistrict Credit Organization (Central Java) KURK stands for Kredit Untuk Rakyat Kecil, or People's Small Enterprise Credit Institution (East Java) KUD stands for Koperasi Unit Desa, or Village Cooperative amount Unit Source Bank Indonesia data, World Bank 1987 stitutions accounted for about 1.5 percent (table 9.1).These other institutions included secondary banks, government-owned pawnshops, regional credit pro- grams under provincial government supervision, privately owned financial in- stitutions, village cooperatives, and various kinds of village credit organizations. Although there were many active rural financial institutions, their outreach was typically small. Thus only 2.5 percent ($524 mullion) of the value of outstanding loans in the country's financial system in 1985 ($21 billion) was provided through rural financial institutions-although 75 percent of the country's population lived in rural areas.Another 0.7 percent of the value of outstanding loans ($144 mil- lion) came from nonbank financial institutions.The other 97 percent of the value of outstanding loans in 1985 was in banks. In 1983 only 17 percent of agricultural households received government credit (Agricultural Census of Indonesia 1983),17 primarily through BRI's 100 The Microfinance Revolution: Lessons from Indonesia BIMAS, the government-subsidized rural credit program provided at the unit desas. Rural borrowers who were able to qualify for the larger loans available in urban bank branches could obtain subsidized credit in district (kabupaten) cap- itals.The KIK (Kredit Investasi Kecil, or Small Investment Loan) and KMKP (Kredit Modal Kerja Permenen, or Small Permanent Working Capital Loan) programs offered subsidized loans up to 15 million rupiah ($13,333 in 1985). But very few rural households were eligible for these and for the other subsi- dized programs for larger loans sponsored by Bank Indonesia."8 And minimum loan sizes at standard commercial banks were prohibitive for nearly all rural dwellers. Most rural credit to low-income households was through the informal sec- People' Credit tor. Typical monthly effective interest rates charged by informal commercial lenders-professional moneylenders, commodity suppliers, traders, landlords- Banks that do not to low-income borrowers in rural Indonesia ranged from about 10 percent to more than 60 percent.'9 Daily or weekly loans, when calculated as effective perform well have monthly rates, ranged from about 10 percent to more than 1,500 percent (see chapter 1, box 1.1).Yet in 1985 inflation was just 4.5 percent.The highest-in- generally been terest loans tended to go to the lowest-income borrowers because these bor- rowers had little bargaining power and few or no other options, and because allowed to die out- smaller loans are usually more expensive for lenders than larger ones.20 Loans were often obtamed at low or no financial cost from relatives, friends, a lesson that many or rotating savings and credit associations (ROSCAs, called arisan in Indonesian). But as discussed in chapter 6, such credit may entail nonfinancial costs and is countries still need typically provided only for emergencies, special occasions, or relatively small loans. These loans are frequently not fungible and are often inappropriate (in amount, to learn timing, or option to reborrow) for working or investment capital. Rural savings facilities were scarce except for BPRs and some credit co- operatives. Most rural people kept their cash savings in the house and in ROSCAs. Savings in gold, crops, animals, raw materials, and finished goods were common. Most commercial banks did not collect voluntary savings from rural areas, and especially not from poor people. Interest rates at state banks were reg- ulated by the government, and with annual effective interest on loans set at 12 percent and interest on most savings accounts set at 15 percent, collecting sav- ings was unprofitable. Even at BRI, after a decade of offering savings accounts nationwide through the unit desa system, savings at the units totaled only $18 million at the time of the June 1983 financial deregulation. But by the end of the 1980s the transformed, commercialized units had $471 million in savings. The defining moments for microfinance in the 1980s were, first, the 1983 financial deregulation, which made it possible for banks to set their own in- terest rates on most loans and savings accounts; and, second, PAKTO 88 (and reform packages that followed), which substantially liberalized restrictions on opening new domestic banks and bank branches and encouraged autonomy and competition in the banking sector. Some effects were immediate. For ex- ample, shortly after PAKTO 88 was issued Bank Dagang Bali opened branch- Rural Development and Rural Financial Institutions in Indonesia 101 es in many areas of Bali where the bank had previously served clients through a de facto mobile banking service (because it had not been permitted to open branches; see chapter 10).As discussed in chapter 8, PAKTO 88 had long-range effects, both positive and negative, on the banking sector. It also had important effects on the development of BPRs, considered later in this chapter. Government Microfinance Initiatives in the 1990s By the early and mid-1990s new government initiatives and presidential in- structions on microfinance in rural areas were thick on the ground.At one end The 1983financial of the continuum was the Banking Law of 1992, which defined banks, their roles, and their supervision, and which aimed at strengthening the banking in- deregulation that dustry, including banking for low-income people.At the other end were a wide variety of politically motivated credit programs-ranging from an effort to pro- allowed banks to set vide massive numbers of highly subsidized $9 loans with an additional grant of less than $1, to a directed credit program through which Indonesian banks interest rates on were required to allocate 20 percent of their loan portfolios to loans of up to 200 million rupiah ($105,000 in 1990) for small enterprises.The impetus for most loans and many of the credit programs of the mid-1990s-which emphasized with con- siderable fanfare their importance for indigenous Indonesians (pribuml)-was savings accounts was mixed. It stemmed from a government priority to provide finance to low-in- come indigenous Indonesians, from the need to maintain rural elites' active sup- a defining moment port for Golkar (it being correctly assumed that rural elites would receive a substantial share of the banks' directed credit), and from attempts to deflect at- for microbanking tention from the growing wealth and corruption of the conglomerates and to defuse rising resentment against ethnic Chinese. The 1992 Banking Law and BPRs The 1992 Banking Law had been in preparation since the early 1980s but had been repeatedly delayed because of interagency disagreements (especially be- tween the Ministry of Finance and Bank Indonesia) about basic issues.A com- promise version mediated by the leading technocrats was submitted to Parliament in 1991, and the law was passed in 1992.Among other issues, the Banking Law defined banks, set limits on bank supervisors, specified the extent of special- ization by banks, and delineated the role of and supervision for very small banks (see Cole and Slade 1996). The law recognized two types of regulated banks: general banks (bank umum) with paid-in capital of 10 billion rupiah ($4.8 million in 1992) and Peo- ple's Credit Banks (BPRs) serving the rural population as well as low-income people and small and microenterprises in urban areas, with a minimum capi- tal requirement of 50 million rupiah ($24,250 in 1992). Mariyanto Danoesputro, then managing director of Bank Indonesia, said in 1997 that "the objective of BPR is to modernize the rural population and to help free the small people from the moneylenders" (quoted in Steinwand 2001, p. 161). 102 The Microfinance Revolution: Lessons from Indonesia Box 9.1 Excerpts from David C. Cole and Betty F. Slade's Building a Modem Financial System: The Indonesian Experience With the existence of nearly 8,000 rural financial institutions at the time the Banking Law was being examined, it was necessary to give consideration to their special require- ments These financial institutions had different types of ownership, size, activities, legal status, and supervision For example, some only gave loans but did not collect deposits PAKTO, 1988 did not deal directly with these institutions, but instead authorized the li- censing of BPRs The BPR had to be either a limited liability company, a regional govern- ment enterprise or a cooperative, and it had to have paid-up capital or basic and compul- sory savings (as the case warranted) of at least Rp 50 million Existing BPRs were given two years to adjust to the new regulations Following PAKTO it became clear that the failure to consider the special aspects of The 1992 Banking the other types of rural institutions and the implicit requirement that these institutions must adhere to the BPR rules was not in the best interests of rural finance In March Law defined banks 1989 (the PAKMAR package), the time limit for existing BPRs to adjust was made indefin- ite, I e they were grandfathered-given the status of old-style (grandfathered) BPRs Thus there were the old BPRs and the new BPRs By the time the Banking Law passed and their roles. It parliament there were 848 new BPRs, including a few that transformed from other status to the new BRP status aimed at The 1992 Banking Law defined BPRs as banks which are permitted to accept de- posits only in the form of time deposits, savings, and/or other of similar types BPRs in- cluded all banks other than general banks, therefore the BPR category included both the strengthening the grandfathered rural financial institutions and all the newly licensed BPRs Article 58 recog- nized the existence of rural financial organizations, and basically gave flexibility to the gov- banking industry, ernment to set rules for the change of status of LPDs, BKDs, BKKs, BKPDs or other or- ganizations similar to BPRs But it was clear that PAKMAR had been reversed rural financial institutions would have to meet BPR requirements [But in 1999 the effects of including banking the crisis resulted in significant changes in requirements for rural financial organizations with regard to meeting BPR requirements) for low-income Source Cole and Slade 1996, p 129 people The Banking Law limited the BPRs "in terms of location, function and portfolio composition. They are precluded from taking demand deposits and participating in the payments system. Their main role is to take time and sav- ings deposits and to extend credit" (Cole and Slade 1996, pp. 128-30; see also Sukarno 2000 and Steinwand 2001).The changing status of the BPRs during the late 1980s and early 1990s is summarized in box 9.1. Although the Banking Law stated that rural financial institutions are required to meet BPR criteria, the crisis and the collapse of the Indonesian financial sys- tem in 1998 caused changes in Bank Indonesia's views and in government pol- icy. The crisis provided a strong incentive for Bank Indonesia to reduce the number of financial institutions under its supervision. In 1999 new regulations increased capital requirements for opening new BPRs or branches by 10 times, to 500 mnllion ruplah.Adjusting for inflation, the new capital requirement was more than three times the previous requirement (Steinwand 2001, p. 181; see also Timberg 2000, p. 18). Rural Development and Rural Financial Institutions in Indonesia 103 Assuming that the [1999] regulatory framework for BPR. . .does not undergo substantial changes... .it is very likely that the... two- fold banking system consisting of commercial banks and BPR will develop into a three-fold system with large commercial banks, new medium sized BPR that work on a provincial level and operate an extensive network of branches and service posts, and the small unit-bank BPR as well as the LDKP and BKD at the bottom level. -Steinwand 2001, p. 182 KUK and commercial banks "The objective of In 1990 the government introduced a new directed credit program called Kredit Usaha Kecil (Small Business Credit, or KUK; see Martokoesoemo 1993; BPR is to McGuire, Conroy, and Thapa 1998; Ravicz 1998;Timberg 2000; and Steinwand 2001). All Indonesian banks were required to lend at least 20 percent of their modernize the rural volume of credit to small enterprises. However, the maximum loan size was 200 million rupiah ($105,000 in 1990); in 1997 the maximum was raised to 350 population and to million rupiah ($75,000 at the end of 1997). Banks could meet the KUK loan quota in three ways: direct lending to en- helpfree the small terprises, lending to BPRs or banks that would onlend to small enterprises, or by issuing Surat Berhaga Pasar Uang (SBPU-KUK), a special money market peoplefrom the instrument. Most commercial banks do not have extensive branch networks, and they had little experience making loans of this size; most of their loans were moneylenders" much larger. Thus the banks often tried to channel their KUK funds through other institutions, typically BPRs.The Association of Indonesian National Pri- vate Banks and the Association of Indonesian Rural Banks jointly created a foun- dation to receive funds from commercial banks and to onlend to BPRs, funded by a 1 percent spread between the rates at which the foundation received and lent funds. But many BPRs could not absorb all the KUK funds supplied by commer- cial banks, and a substantial amount of these finds went to the interbank money market rather than to small businesses.Another problem was that the large KUK funds discouraged BPRs from mobilizing savings. In theory KUK loans had no minimum and were supposed to be made available for small enterprises. In prac- tice these loans rarely reached poor clients, though they did reach some small and medium-size enterprises. But many banks filled their KUK quotas by financing cars and housing rather than small enterprises (Martokoesoemo 1993, p. 108). During the crisis the obligation for banks to provide KUK loans was sus- pended, and KUK lending declined rapidly from 65,890 billion rupiah in fis- cal 1998 to 38,171 billion rupiah in fiscal 1999 (Steinwand 2001, p. 179). In December 1998 nonperforming KUK loans (loans in default) were 23 percent of outstanding KUK loans (Timberg 2000, p. 21)-a very high default rate by microfinance standards, but much lower than Indonesian commercial banks' nonperforming loans as of the same date (officially 59 percent of outstanding loans, but widely believed to have been considerably higher). 104 The Micrafinance Revolution: Lessons from Indonesia Levying corporations and wealthy individuals for funds for subsidized credit programs A 1994 Ministry of Finance regulation and a 1995 presidential instruction re- quired state-owned enterprises (including banks) to use 5 percent of their prof- its to support small and medium-size enterprises and to reduce poverty (the Badan Usaha Milik Negara, or state-owned enterprise, policy; see McGuire, Conroy, andThapa 1998; Meyer 1998; Ravicz 1998; and Steinwand 2001).The levy, which was outside the tax system, was collected and administered by a foun- dation created by President Soeharto. In addition, private corporations and in- dividuals with annual incomes above 100 million rupiah ($43,328 in 1995) were required to donate 2 percent of their profits or income to the Soeharto foun- dation to be used for poverty reduction. Half of these funds were to be used Indonesian banks for loans to low-income people at low interest rates, usually 6 percent a year. The other half was used for poverty-related training and research (30 percent) were required to lend and for support of the state-owned cooperative credit insurance company (20 percent). at least 20 percent The Prosperous Family Program, implemented by the National Family Plan- mng Coordinating Board (Badan Koordinasi Keluarga Berencana Nasional, or of their credit BKKBN) through its village outlets, was a major recipient of these funds. Known as Tabungan Kesejahteraan Rakyat (Savings for a Prosperous Family, or volume to small TAKESRA) and Kredit Usaha untuk Kesejahteraan Rakyat (Credit for a Pros- perous Family, or KUKESRA), the program operates through women's groups. enterprises-up to When a group is established, each woman receives a grant of less than $1. Poor women receive initial loans of $9 at a highly subsidized 6 percent annual ef- loans of $105, 000. fective interest rate; 10 percent of the payments are returned to borrowers who repay the loans.The program received about 500 million rupiah during its first But the poor were year of operation ($210 million in 1996). By mid-1997 nearly 10 million households were reported to have participated in this program. The Prosper- rarely reached ous Family Program was not expected to cover the cost of its loans (and in any case had little experience collecting loans). The National Family Planning Coordinating Board has built a vast infra- structure throughout Indonesia and enjoys a well-deserved international rep- utation for its role in substantially reducing Indonesia's population growth rate since the 1970s. But its Prosperous Family Program, a large, hlghly publicized, unsustainable microfinance program, emphasized short-term political priori- ties at the expense of viable microfinance development. In addition, its large- scale delivery of highly subsidized credit put the program in unfair competition with the BPRs and BRI's unit desas, which provide loans to much the same market at commercial interest rates. This is just one of many examples of op- posing microcredit policies and approaches being implemented simultaneous- ly in the same villages during the mid-1990s. "Left-behind" villages: INPRES Desa Tertinggal A 1993 presidential instruction created the INPRES DesaTertinggal (IDT) pro- gram for backward (literally, "left-behind") villages (see McGuire, Conroy, and Rural Development and Rural Financial Institutions in Indonesia 105 Thapa 1998, Seibel and Parhusip 1998, and Parhusip and Seibel 2000). Coor- dinated by the National Development Planning Board (Badan Perencanaan Pem- bangunan Nasional, or BAPPENAS), the program provided $600 million over three years (1994-97) for about 28,000 rural villages identified as being espe- cially poor-about one-third of the country's villages. Most are located on the Outer Islands, and the program represents an effort by the government to re- dress long-standing regional inequalities. The program provides funds for de- veloping physical infrastructure, for income-generating activities of poor people organized into self-help groups, and for personnel (teachers, social workers, and others) who help these groups. Each of the program's villages received 20 nrldhon rupiah ($9,100 at the end A political strategy of 1994) for the income-generating component. These finds were provided as grants to the self-help groups, which then make loans to group members in a developed that revolving credit scheme. Each group decides the terms on which the loans are made; the groups are not required to set interest rates that cover costs. By mid- emphasized visible 1997 about 3.3 million people had participated in the program, with an average loan of $85 (McGuire, Conroy, andThapa 1998, p. 158). Ismawan (2000, p. 4) re- (but not viable) ports that as of February 2000 Bina Swadaya (Self-Reliance Development Foun- dation), which helps groups lnk with banks in many parts of Indonesia, had creditfor poor consulted and held policy discussions with about 100,000 IDT self-help groups. A 1995 evaluation concluded, however, that the IDT program had multi- indigenous ple problems, including the role played by facilitators without backgrounds in commercial or production activities, the development of business plans that had Indonesians, little potential for profit, and-once again-an emphasis on short-term polit- ical gains derived from highly publicized disbursement of loans to the poor rather undermining than on professional program design and implementation (Seibel and Parhusip 1998, pp. 3-4). commercial Bank Indonesia, government ministries, and rural credit programs microfinance Bank Indonesia and several ministries have been involved in rural credit pro- jects and programs in different ways. Bank Indonesia implements the govern- ment's Microcredit Project as well as the Linking Banks with Self Help Groups Program, a subsidized credit program partly funded by the German Agency for Technical Cooperation (Gesellschaft ftir Technische Zusammenarbeit, or GTZ) .21 The central bank's main role in microfinance, however, is to provide subsidized liquidity credits channeled through the banking system (see chap- ter 12). Major recipients of Bank Indonesia credit subsidies include government- sponsored village cooperatives (Koperasi Unit Desas, or KUDs) which, while important for state ideology, have a long history of defaults and mismanage- ment; and Kredit Usaha Tani (Credit for Farm Enterprises, or KUT), a Min- istry of Agriculture program channeled through BRI's branches (but not through the unit desa system), also with a poor repayment record. As noted, the Indonesian government is not a monolith when it comes to cred- it subsidies. But with the rise of the conglomerates in the early and mid-1990s, the massively increasing crony capitalism, and the growing resentment against eth- 106 The Microfinance Revolution: Lessons from Indonesia Box 92 Excerpts from Camilla Nestor's contribution to Ohio State University's online Development Finance Network, 2 February 1999 I've become increasingly dismayed with the mounds of subsidized credit the Indonesian government, the World Bank, and UNDP (among others) are pumping into Indonesian vil- lages, ostensibly as a response to the economic crisis Within the past four months, dis- tortions have emerged in the rural credit markets-cheap credit has begun to squeeze lo- cal financial service providers who must charge a market interest rate to survive If this continues, the risk of failure for BPRs (rural banks) and other rural financial institutions will be compounded-subsidized competition on top of an economic crisis If RFIs [rural fin- ancial institutions] stumble as a result of the subsidized loan programs, where will people save? And where will they borrow once the subsidized projects end? Furthermore, there's the risk of devaluing the term "loan " If these government/donor subsidized credit The government projects, which charge 0%-20% interest per year, have no enforceable repayment re- quirements and no penalties for default, people will consider the loan a gift. views charging Perhaps [this occurs] because donors cave into pressure from the government, which views charging market interest rates on micro-loans "harmful" to the poor, and po- litically unwise market Interest rates We know that 1) directed credit most often ends up in the hands of those with con- nections to the distributors, rather than with the worthiest clients, 2) loan programs with on micro-loans vague repayment terms and subsidized interest rates are unlikely to see very impressive repayment rates, and 3) subsidized credit eventually dies-the donors cannot keep inject- " t h ing such vast amounts of money Why perpetuate this? harmful to the Source Camilla Nestor, 2 February 1999 poor and politically nic Chmnese, a political strategy was adopted that emphasized visible (if not viable) unwuse" benefits to poor indigenous Indonesians. During the 1990s rural credit subsidies increased to massive proportions, as illustrated in a February 1999 letter to Ohio State University's Development Finance Network from the Indonesia project man- ager of Catholic Relief Services (box 9.2). Catholic Relief Services is active in nucrocredit development in Indonesia and other developing countries. Rural Finance in the 1990s Following PAKTO 88, the 1988 deregulation package that liberalized the banking sector, and the 1992 Banking Law, new banks and bank branches pro- liferated rapidly. By 1998 there were 222 commercial banks in Indonesia: 130 private national banks with 3,976 offices and a 46 percent share of the total as- sets of commercial banks; 7 state-owned banks with 1,602 offices (excluding the unit desas) and 40 percent of conimercial bank assets; 58 foreign and joint venture banks with 121 offices and 13 percent of assets; and 27 Regional De- velopment Banks (Bank Pembangunan Daerah, or BPDs) with 555 offices and 2 percent of assets (Stemwand 2001, pp. 147-48).22 The number of BPDs and state-owned banks did not change between 1987 and 1998, but the number of private banks doubled, from 64 to 130, and the number of foreign and joint venture banks more than quintupled, from 11 to 58. Rural Development and Rural Financial Institutions in Indonesia 107 Table 9.2 Loans and savings in Indonesia's financial system, 1995 Savings Credit Number of Amount Number Amount savings Type of Number Billions of Millions of of loans Billions of Millions of accounts financial institution of units rupiah U.S. dollars (thousands) rupiah U.S. dollars (thousands) Microfinance institutions 12,843 5,077 2,203 4,712 7,423 3,220 19,084 BPRs and secondary banks 1,948 1,566 679 1,232 1,226 532 2,969 LDKPs 1,978 224 97 261 118 51 456 BKDs 5,345a 93 40 955 63 27 1,176 BRI unit desas 3,482b 3,194 1,386 2,264 6,016 2,610 14,483 Commercial banks 240 234,611 101,783 91,168 214,764 93,173 49,904 Total 13,083 239,688 103,986 95,880 222,187 96,393 68,988 Note BPR stands for Bank Perkreditan Rakyat (People's Credit Bank, excluding here the LDKP and BKD) LDKP stands for Lembaga Dana dan Keuangan Pedesaan (Rural Fund and Credit Institution) BKD stands for Badan Kredit Desa (Village Credit Organization) BRI stands for Bank Rakyat Indonesia a Only 4,806 BKDs were active in 1995 b Table 12 7, based on BRI data, shows 3,512 unit desas at the end of 1995 The figure in this table probably came from an earlier month in 1995 Source Adapted from Seibel and Parhusip 1998, table 1 3, and Parhusip and Selbel 2000, table 7 3 By the mid-1990s the rural financial sector had also expanded consider- ably. Indonesia was home to a plethora of financial institutions operating in rural areas, some old and some new (those opened after PAKTO 88).These includ- ed state-owned and private commercial bank branches; provincial government banks; credit organizations owned by provincial, district, and subdistrict gov- ernments and by vlllages; cooperatives; licensed BPRs, old and new, public and private; and thousands of generic BPRs. But many of the new BPRs had been opened inumediately after PAKTO 88, and their owners and managers often had little banking experience. The values and numbers of loans and deposit accounts in Indonesia's com- mercial banks and microfinance institutions at the end of 1995 are shown in table 9.2.Although data for the microfinance institutions are often of poor qual- ity, the general picture is clear. Comparison with table 9.1, which provides 1985 data on loans in the financial system and in rural financial institutions, is use- ful. Between 1985 and 1995 average annual inflation (GDP deflator) was 8.8 percent (World Bank 1997). In 1985 outstanding loans in the financial system totaled 23,684 billion rupiah ($21.1 billion); in 1995 the total was 239,688 bil- lion rupiah ($104.0 billion). Tables 9.1 and 9.2-which are derived from different studies and different sources-are not directly comparable.23 Still, they show that the value of out- standing loans in rural financial institutions (including BRI's unit desa system) as a percentage of total loans in the financial system hardly varied between 1985 108 The Microfinance Revolution: Lessons from Indonesia and 1995-it was about 2.5 percent in 1985 and about 2.1 percent in 1995. The other 98 percent or so was in banks.Table 9.2 also shows that in 1995 mi- crofinance institutions had about 5 percent of the number of loans in the fi- nancial system, 28 percent of the number of savings accounts, and 3 percent of the value of savings. Data are not available, however, on commercial banks' share of outstand- ing rural loans or microloans, or of small savings accounts. These shares likely increased somewhat between 1985 and 1995, but there is no way of knowing how significant that increase might be. Table 9.2 also enables comparison ofBRI's 3,482 unut desas with 9,271 other (heavily rural) nmicrofinance institutions in 1995.With only 27 percent of the aggregate number of units in these mncrofinance istitutions, the unit desas dom- The value of loans inated with 48 percent of the total number of loans, 63 percent of the value of outstanding loans, 76 percent of the number of savings accounts, and 81 per- in ruralfinancial cent of the value of savings. institutions as a Developing the BPRs After PAKTO 88 and the passage of the 1992 Banking Law, indigenous In- share of total loans donesians were encouraged to open BPRs. Most of the BPRs that opened after 1988 are privately owned and financed commercially from local resources: sav- in the financial ings, commercial debt (mainly from KUK loans during most of the 1990s), and equity. But in 1999 Bank Indonesia allocated concessional lines of credit to the system was 2.5 BPRs. In September 1998 there were 8,699 generic BPRs, reaching 5.6 million percent in 1985 and clients. Table 9.3 shows the number of units and clients for the 1,576 private licensed BPRs, for an estimated 2,317 local-government-owned LDKPs (in- 2. 1 percent in 1995 cluding those that are licensed BPRs),24 and for the 4,806 active vlllage-owned BKDs. To provide a more complete picture of Indonesian microbanking, the table also shows data for BRI's unit desas. Steinwand (2001, p. 188) notes the well-known difficulties of collecting data for thousands of small, independent BPRs and the "notorious inconsistency of BPR statistics." But he has made an important contribution by collecting and analyzing much previously unpublished data. In tables 9.3 and 9.4 licensed BPRs that are privately owned are shown sep- arately from those that are publicly owned (mainly former LDKPs). These ta- bles are based on Steinwand's tables and therefore reflect his data organization. Because he was interested in comparing public and private BPRs and most of the licensed BPRs are privately owned, Steinwand used the larger data set of generic BPRs for his comparison. In his tables he separated the private BPRs from the public LDKPs and BKDs, and he included former LDKPs that are now licensed BPRs in the LDKP category (see table 9.3). In 1998 the 2,262 licensed BPRs, private and public, had $175 million in outstanding loans, $56 million in savings, and $83 million in time deposits (Stein- wand 2001, p. 190). Among the three BPR categories (private BPRs, the LDKPs, and the BKDs), private BPRs had the fewest active units (1,576) but Rural Development and Rural Financial Institutions in Indonesia 109 Table 9.3 BPRs and BRI unit desas: number of units and clients, September 1998 Active units Clients Share of total Share of total Institution Number (percent) Numbera (percent) BPR systems 8,699 70 5,615,215 26 Private licensed BPRs 1,576 13 3,107,612 14 LDKPsb 2,31 7c 18 1,700,000c 8 BKDs 4,806 39 807,603 4 BRI unit desas 3,703 30 16,050,000 74 Among microfinance Total 12,402 100 21,665,215 100 institutions, BRIs Note BPR stands for Bank Perkreditan Rakyat (People's Credit Bank, excluding here the LDKP and BKD), LDKP stands for Lembaga Dana dan Keuangan Pedesaan (Rural Fund and Credit Institution) BKD stands for Badan Kredit Desa (Village Credit Organization) BRI stands for Bank Rakyat unit desas had 63 Indonesia. a For the BPRs Steinwand uses the number of savers for the number of clients because every borrower has a savings account For the unit desas, the author's calculations estimate the number of percent of the value clients from the number of savings accounts All borrowers in unit desas also have savings accounts, but since individuals may have more than one, the number of savings accounts is larger than the number of savers (clients) BRI does not track the number of savers, only the number of accounts In of outstanding loans September 1998 there were 21 4 million savings accounts in the unit desas Estimating that roughly a quarter of the savers had two accounts, this would leave 16,050,000 savers (clients) b Includes former LDKPs that are now licensed BPRs and 81 percent of c Estimated Source Steinwand 2001, table VI 2, BRI unit desa monthly report for September 1998 the value of savings in 1995 the most clients (3.1 million) in September 1998 (table 9.4).Private BPRs also had the most assets ($308 million, compared with $91 million for the LDKPs and $28 million for the BKDs; Steinwand 2001, p. 230).25 The BPRs and BRI's unit desas. BRI's unit desas, with 30 percent of the combined number of BPR and BRI bank units, had 74 percent of the total clients. Private BPRs, with 13 percent of the units, had 14 percent of the clients. But the public LDKPs and BKDs, with 57 percent of the units, had only 12 percent of the clients.While tables 9.2 (for 1995) and 9.3 (for 1998) are drawn from different sources and are not directly comparable, they indicate clearly that during the crisis years the unit desa system retained its dominant position in microfinance. But tables 9.3 and 9.4 also show that the generic BPRs had, in aggregate, substantial numbers of small loans and savings accounts during the crisis. As at the unit desas, BPR savings increased during the crisis, and the BPRs were able to place the funds in high-yieldmg commercial bank deposits. Non- performing loans at the BPRs and at the unit desas did not increase significant- ly. The BPR and unit desa figures are not always comparable, however. For example, the village-owned BKDs, still operating under 1929 regulations, are lim- ited in their writeoffi; thus their default rate includes substantial bad debt accu- 110 The Microfinance Revolution: Lessons from Indonesia Table 9.4 BPRs and BRI unit desas: average loans and deposits, 1997 Average Average Average outstanding loan savings balance time deposit Percentage Percentage Percentage Rupiah of GNP Rupiah of GNP Rupiah of GNP Institution (thousands) per capita (thousands) per capita (thousands) per capita Private licensed BPRs 1,338 26 141 30 8,165 158 LDKPs LPNs 582 11 77 1.5 2,182 42 LPDs 535 10 130 2 5 2,204 43 BKKs/CJ 424 8 134 3 0 4,026 78 BKDs 142 3 14 03 27 05 BRI unit desas 1,791 35 424 80 8,739 169 Note In June 1997, just before the crisis began, the exchange rate for rupiah was 2,450 to the U S dollar, but by December 1997 the rupiah had fallen to 4,650 to the U S dollar Percentage of GNP per capita ($1,110 in 1997) is calculated based on the year-end exchange rate BPR stands for Bank Perkreditan Rakyat (People's Credit Bank, excluding here the LDKP and BKD) LDKP stands for Lembaga Dana dan Keuangan Pedesaan (Rural Fund and Credit Institution) LPN stands for Lumbung Pitih Nagari (People's Credit Bank, West Sumatra) LPD stands for Lembaga Perkreditan Desa (Rural Credit Organization, Bali) BKK/CJ stands for Badan Kredit Kecamatan/Central Java (Subdistrict Credit Organization) BKD stands for Badan Kredit Desa (Village Credit Organization) BRI stands for Bank Rakyat Indonesia Source Steinwand 2001, table VI-3, BRI unit desa monthly report for December 1997, World Bank 1997g, author's calculations based on Steinwand and BRI data mulated from the past. But BRI's unit desas have an automatic system for writ- ing off bad debt (chapter 12). None of the three BPR systems (private BPRs, the LDKPs, and the BKDs) has an adequate loan loss reserve policy. Underprovisioning for loan losses by the BPRs disguises their real position and prevents timely adjustments for fu- ture lending. In contrast, BRI's unit desas provision according to Bank Indonesia regulations, follow international accounting practices, and are required to have regular audits. BPRs: public and private. Indonesia's nearly 9,000 generic BPRs can be analyzed based on their age, legal status, and ownership. First, old and new BPRs can be differentiated depending on whether they were opened before or after PAKTO 88 (October 1988). Second, BPRs are defined by their legal status. Licensed BPRs, most of which opened after 1988, meet Banking Law criteria and other subsequent regulations. The third distinction is between public and private BPRs. Most licensed BPRs are privately owned, though a number of publicly owned former LDKPs are also licensed BPRs. The public unlicensed BPRs are officially considered in transition to becoming licensed BPRs. Private BPRs are a heterogeneous group with many kinds of owners, in- cluding individual investors, limited companies of investors, commercial banks, Rural Development and Rural Financial institutions in Indonesia 11 industry groups, and NGOs.These BPRs tend to be clustered in well-developed areas with intense competition (Jakarta, Surabaya,Yogyakarta, Bali), while pub- lic BPRs operate in rural areas in many parts of the country, typically allocat- ed by subdistrict. Steinwand (2001, ch. 6) points out some important differences between pri- vate and public BPRs: * Competition. Private BPRs compete with one another. Public BPRs gener- ally do not compete because of their separate locations (although they may have other competitors, especially BRI's unit desas). * Savings. Public BPRs mobilize more savings at a lower cost than do private 8, 699 generic BPRs.This pattern is attributed to greater trust by communities, related to implicit or explicit government deposit guarantees. BPRs continued to . Loans. Public BPRs have smaller average loans and lower default rates than do private BPRs (see table 9.4). Steinwand suggests that since public serve 5.6 million BPRs have a lower cost of savings than the private BPRs, they are gen- erally able to charge lower interest rates on loans and make lower-risk in- clients in 1998-as vestments. In addition, fierce competition among private BPRs in Java and Bali has led to high defaults (see chapters 18 and 19 for discussion of how the country's intense competition among institutions providing microcredit in Bolivia and other Latin American countries in the late 1990s also resulted in high financial system defaults). * Commercial bank links. Licensed BPRs, most of which are private, are super- collapsed vised by Bank Indonesia. Private licensed BPRs have no formal links with commercial banks (except for BPRs owned by commercial banks). But many have deposit accounts at commercial banks, and some have loans.And some BPRs have formed BPR networks. In contrast, public BPRs have close links, both financial and supervisory, with commercial banks. "The public BPR are better supervised both by the formal supervisory agencies as well as by informal control mechanisms compared to their private counterparts" (Stein- wand 2001, p. 267). * Profitability. Public BPRs are more profitable than private ones, but because many BPRs do not adequately provision for loan losses and do not under- go regular audits, reported profits are often not adjusted for loan loss provi- sions and writeoffs or for inflation. This makes it difficult to obtain a true picture of BPR profitability (see Ravicz 1998).As Steinwand (2001, p.292) shows, unadjusted figures indicate that in 1997 the average returns on assets for private licensed BPRs was 1.0 percent; for the three main systems of pub- lic LDKPs, 4-6 percent; and for the village-owned BKDs, 10 percent. Un- adjusted figures also indicated a return on equity in 1998 of 5 percent for private BPRs, compared with 17-34 percent for the three public LDKP sys- tems and 11 percent for the BKDs. But Steinwand (pp. 249,251) comments: In absence of more audited data, we calculated a "reserve- adjusted ROA [return on assets]" under the assumption that 112 The Microfinance Revolution: Lessons from Indonesia collateral value is zero and that all adjustments have to be made within one year... Under this "worst case scenario" the adjusted ROA of all BPR [groups] would be negative, however those of the BPR unit banks [private licensed BPRs] would be more than twice as low as those of the other BPR systems: -15% for the BPR umt banks; -7% for the LPN; -2% for the LPD; -3% for the BKK; and -6% for the BKD. The real situation is likely to fall somewhere between the reported fig- ures and the worst case scenario. It is of course different for different insti- tutions, and is further complicated by regulations on writeoffs. Moreover, these are 1997 figures. By the end of 1997 Indonesia was in the midst of its The public BPRs, worst crisis in three decades. Steinwand (p. 293) concludes: "The public LDKP are the most successful as a group, have BPR system and ... outperform. . . the private ones, a finding quite in contrast to mainstream thinking which clearly favors private [microfinance institu- higher savings, lower tions]."These results raise some interesting issues and questions. default rates, deeper The private BPRs are new, primarily urban or semiurban, heterogeneous in ownership, generally not well supervised, without strong links to commercial outreach, and higher banks, and they often have relatively little banking experience, at least at the start. In contrast, public BPRs tend to be more experienced and more trust- profits than the ed; they are rural; owned by provincial, district, and subdistrict governments or by villages; and are supervised by commercial banks. As a group the pub- newer private BPRs lic BPRs have higher savings, lower default rates (though still high compared with the unit desas), deeper outreach, and higher profits than the private BPRs. But the three LDKP systems used as examples in many of Steinwand's comparisons with the private licensed BPRs (primarily because of the rela- tively reliable data available for the LDKP systems) are among the best LDKPs in Indonesia. The Lembaga Perkreditan Desa (LPD) of Bali is widely con- sidered the best LDKP system in Indonesia, and the Badan Kredit Kecamatan of Central Java (BKK/CJ) is generally considered the second best (see Stein- wand 2001, pp. 292-93). Had the three systems been compared with the best of the private BPRs, the results might have been somewhat different.The of- ficially licensed BPRs (both public and private) have a wide range of per- formance, and some are in the process of upgrading. As Timberg (2000, pp. 21, 25) notes: It is reported [in December 1998] that 45% of BPRs are not healthy or almost healthy... The BPRs, though generally in better shape than the commercial banks, need to go through a parallel process of closing some BPRs and restructuring others ... Further, there is a need for a tightening of procedures and lessening of costs and a more aggressive marketing Rural Development and Rural Financial Institutions in Indonesia 113 orientation. Upgrading management is already occurring as networks of BPRs put out standard management packages- just like a franchisor. I visited one bank, taken over by a private chain, which had moved from 95% NPL [non-performing loans] to under 5% in one year. It is also worth noting that many public BPRs operate in quasi-monopo- listic environments. Yet private BPRs, generally operating in areas of high competition, have higher assets and wider outreach than public ones. In 1998, with only 18 percent of the number of BPR units (8,699), private licensed BPRs served 3.1 million clients, or 55 percent of the number of BPR clients (see table Many public BPRs 9.3). Private BPRs, not born with linkages like their public counterparts, are are quasi- developing links-both with commercial banks and with other BPRs.And pri- vate BPRs are learning to operate in an intensely competitive environment- monopolies. Yet which is likely to be where the future of mucrofinance lies (see chapter 21). The public LDKPs and BKDs had smaller average loans and savings balances private BPRs, than the private BPRs (see table 9.4). But this would be expected because, un- like public BPRs-which are typically dispersed to cover as many rural subdistricts generally operating as possible-private BPRs tend to cluster in relatively well-developed areas, many of them urban and semiurban.Yet with an average outstanding loan of 1.3 rrul- in highly competitive lion rupiah in 1997 ($288 at the end of 1997, or 26 percent of per capita GNP) and an average savings balance of 141,000 rupiah ($30, or 3 percent of per capi- areas, have on ta GNP), private BPRs appear to be providing financial services to low-income segments of the developed areas in which they operate. average higher assets Private BPRs serve better-off clients as well; their time deposit accounts had an average balance of $1,756 in 1997 (158 percent of GNP per capita). and wider outreach The fixed deposits increase the size of the average savings account, making it possible for private BPRs to serve small savers profitably.The fixed deposit ac- counts also make available substantial funds for lending, enabling the wider out- reach that private BPRs have achieved relative to public BPRs.26 There are important lessons here. First, both public and private microfinance institutions can be financially self-sufficient. Second, as Steinwand (2001, pp. 250, 324) shows, the BPRs are more likely to be self-sufficient if they are fi- nancial intermediaries, funding loans with savings.Third, over time, public BPRs that performed badly died out. The same is likely to happen to private BPRs (if donors do not interfere); in another decade it is possible that private BPRs that survive will be substantially improved. The crucial point is that as a group the generic BPRs, like BRI's umt desas- and in sharp contrast to commercial banks-survived the Indonesian crisis, main- taining outreach, liquidity, and in some cases profitability. The public and private BPRs, the state-owned BRI unit desas, and the private Bank Dagang Bali together demonstrate unmistakably that commercial microfinance-in very different kinds of institutions-can be extremely stable even in times of ex- ceptional economic, financial, and political crisis. 114 The Microfinance Revolution: Lessons from Indonesia Six examples of rural financial institutions and programs Six quite different rural niicrofinance institutions, projects, and programs are dis- cussed below to lllustrate the varied trends of the 1990s discussed above and the range in performance (table 9.5). These examples were selected based on the quality of available data and on geographic and program diversity. BRI's unit desas are not considered here because they are exarmined in detail in later chapters. The oldest of the BPRs, the Badan Kredit Desas (BKDs), and the Income Gen- erating Program for Small Farmers and Fishermen (Pembinaan Peningkatan Pen- dapatan Petani-Nelayan Kecil, or P4K) are also not discussed here because they are supervised or administered through BRI's branches and are discussed in chap- ter 14. The well-known LDKP system, the Lembaga Perkreditan Desa (LPD) of Bali, is examined in chapter 17 as a part of an analysis of village banks. As a group the The organizations discussed below include three public LDKPs, a licensed private BPR, a Grameen Bank replication project that receives funding from generic BPRs, like the Grameen Trust Fund and from an Indonesian state bank, and a large gov- ernment- and donor-sponsored program linking banks and self-help groups. the BRI unit These organizations vary considerably in age and size, and also in the envi- ronments in which they operate ranging from Sumatra in the west to Nusa desas-and in Tenggara Barat in the east. For each I have used the most recently published information I could find (varying from 1995 to 1998). It should be noted, how- sharp contrast to ever, that the data are of uneven quality and depth, and that there are some- times inconsistencies among sources for the same institution. In addition, types commercial banks- of information and definitions of indicators vary by organization. Published information is not yet available on how most of the organiza- maintained outreach tions discussed in this section were affected by the Indonesian crisis. But given Steinwand's (2001) assessment that BPR systems generally survived quite well and profitability during the crisis years (as well as corroborating evidence from BRI's unit desas and Bank Dagang Bali), it seems likely that the first two of the LDKP systems during the recent and the private BPR considered here would have weathered the crisis to vary- ing degrees. However, the third LDKP and the two programs discussed had se- Indonesian crisis rious problems-conceptual, financial, structural, or political (and in some cases all four)-even before the crisis. An early microfinance pioneer: the Badan Kredit Kecamatan of CentralJava (BKK/CJ). The oldest, best-known, and best-documented of the six organi- zations is the Badan Kredit Kecamatan of Central Java (BKK/CJ), an LDKP system owned by the provincial government of Central Java and administered through rural credit organizations located at the subdistrict level (Gadway, Gaway, and Sardi 1991; Patten and Rosengard 1991; Gonzalez-Vega and Chaves 1993; Mosley 1996; Riedinger 1994; Chaves and Gonzalez-Vega 1996; Meyer 1998; Steinwand 2001). CentralJava, along with neighboring D.I.Yogyakarta, is a world-renowned center ofJavanese culture and traditions. CentralJava is a densely populated area with 30 million people; in 1995 the province had 911 people per square kilo- meter. It is a fertile, rice-producing area with well-developed irrigation and in- Rural Development and Rural Financial Institutions in Indonesia 115 Table 9.5 Loans and savings in five rural financial institutions and the PHBK program Typical Value of effective Province; Number out- monthly Value institutional of standing interest Adjusted Adjusted Number of Nare, status; out- loans rate on Type arrears default of savings year of year standing (U.S. loans of rate rate savings (U.S. data started loans dollars) (percent) loans (percent) (percent) accounts dollars)a BKK/CJ, CentralJava;LDKP - 104 million 2-11 Individual - 127 - 71 million 1998 owned by in the 204 (1997) in the 204 provincial, district, BKKs/CJ BKKs/CJ and subdistrict that had that had governments, BPR BPR 1970 licenses licenses BKK/SK, South Kalimantan, 34,518 3 4 3 9-16 0 Individual 5 gb,c 3 5b,c _ 427,357 1995 LDKP owned million and group by provincial government, 1985 LKP (four Nusa Tenggara 4,148 312,963 8 3 Individual 20 Oc.d 5 5c.d d 28,249 LKPs in Barat, LDKP one owned by district), provincial 1995 government; 1989 Bank Shinta D I Yogyakarta, 12,656 2 1 million 23 Group and 3 5e 2 oe 30,340 22 million Daya, 1995 private licensed individual BPR 1970 Mitra Karya, East Java, 255 31,899 2 5f Group 2 oe 1 oe 1,125 5,329 1995 Grameen groups compul- replicator with 1,078 sory project, loans to savings 1993 1,125 only members PHBK, 16 provinces, 6,800 5 5 million9 4-35 Group 5 7e,h Estimated - 895,345 1996 project providing group below 4 compul- technical assist- loans percente sory ance to banks, savings NGOs, and only borrower groups; 1989 116 The Microfinance Revolution: Lessons from Indonesia Table 9.5 (continued) - Not available Note BKK stands for Badan Kredit Kecamatan (Subdistrict Credit Organization) of Central Java (CJ) or South Kalimantan (SK) LDKP stands for Lembaga Dana dan Keuangan Pedesaan (a publicly owned Rural Fund and Credit Institution) LKP stands for Lumbung Kredit Pedesaan (Rural Credit Organization) PHBK stands for Program Hubungan dan Kelompok Swadaya Masyarakat (Program to Link Banks and Self-help Groups) a Compulsory and voluntary combined except where noted b Calculated only for the 34 original BKKs c Calculated by dividing the volume of loans in arrears and in default by the volume of outstanding loans The volume of loans in arrears is defined as the percentage of loan volume in arrears net of previous year's defaults All loans more than 90 days overdue are considered to be in default (Ravicz 1998, p 35) d Average for three LKP units The fourth LKP unit studied had adjusted arrears of 78 percent in 1993, but data for this unit were not available for 1994 or 1995 e The arrears rate is not defined but probably refers to loan payments more than 30 days overdue The default rate is also not defined, but in accordance with Indonesian bank regulations, it would refer to loans more than 90 days overdue f Compulsory savings were not included in the interest rate calculation, had they been included, the effective interest rate would be higher g Total loan volume for fiscal 1995/96 h Arrears are defined as the volume of payments one day late or more as a percentage of the volume of outstanding loans Source Ravicz 1998, McGuire, Conroy, and Thapa 1998, Seibel and Parhusip 1998, Parhusip and Seibel 2000, Steinwand 2001 frastructure. Central Java also has other food crops, commercial crops, and in- dustries (including tourism), and is well known for its art and handicrafts. The early years. The BKK/CJ began in 1970 as a project of the CentralJava provincial government and by 1975,486 BKKs/CJ had been established-cov- ering all but 6 of the province's 492 subdistricts.The BKK/CJ is owned by the provincial government (50 percent), district governments (35 percent), and Provincial Development Bank (Bank Pembangunan Daerah, or BPD; 15 per- cent).27 Each subdistrict unit was capitalized by an initial loan ($2,400 in 1972) from the provincial government, with a maturity of 3-5 years at an interest rate of 1 percent a month. Although the BKKs/CJ operate under the Central Java provincial gov- ernment, each is locally administered and financially autonomous. "This has en- sured consistency of overall system policies, standards, and procedures, and at the same time maintained local accountability and incentives based on local per- formance" (Patten and Rosengard 1991, p. 28). Starting in the early 1970s the BKKs/CJ provided small, short-term indi- vidual loans to rural households and collected compulsory deposits from bor- rowers. Loan maturities ranged from 22 days to six months, with several repayment options varying from daily installments to seasonal repayments. Ef- fective monthly interest rates ranged from 2.2-10.8 percent depending on the type of loan; the shorter the loan maturity and repayment intervals were, the higher the interest rates were (Patten and Rosengard 1991, p. 27; Steinwand 2001, p. 134). Loans were financed by retained earnings, local government loans, and compulsory savings. Retained earnings and current profits grew from 34 percent of funding in 1975 to 62 percent in 1985. But during the 1970s the BKKs/CJ also channeled government-subsidized loans, and BKKs/CJ with a high incidence of these loans showed high losses. Rural Development and Rural Financial Institutions in Indonesia 117 The losses, combined with corruption and mismanagement, resulted in clo- sure or severely curtailed operations in one-third of the BKKs/CJ by the early 1980s (Riedinger 1994, p. 306). In 1979 the U.S.Agency for International Development (USAID) began pro- vidmng technical and financial assistance for the BKKs/CJ (the first time that any of Indonesia's BPRs had received external support). The USAID-supported Provincial Area Development Project helped rehabilitate the system, recapitalize bankrupt BKKs/CJ, and provide training in management and bookkeeping.The project also helped the BKK/CJ system upgrade from a project to a Badan Usaha Milik Daerah (BUMD), or provincial government-owned enterprise, in 1981. Important lending principles of the BKKs/CJ-short-term working cap- The average ital loans, interest rates that cover all costs and risks and enable a profit, char- acter-based lending to individual borrowers, and incentives to management and outstanding loan at staff-were adapted by BRI in 1983 when the unit desas' new KUPEDES lend- ing program was designed (see chapter 12). CentralJava's The BKK/CJ as a provincial government enterprise. In 1981 the BKK/CJ sys- tem began a period of growth. Larger BKKs/CJ opened village posts to increase BKKs (subdistrict their outreach.The village posts, which are open once a week and operate in village halls, disburse loans and collect savings and loan installments, but they credit organizations) do not have authority to make loan decisions.There were about 4,000 BKK/CJ village posts in 1999 (Steinwand 2001, p. 202). was 8 percent of per In 1984 the ministry of finance granted the BKKs/CJ perrnission to col- lect savings from the public. But the BKKs/CJ were slow to mobilize savings capita GNP in because of the widespread view that the rural population could not or would not save in financial institutions (see chapter 13). By 1987, however, the suc- 1997 cess of the BRI unit desa voluntary savings mobilization program, especially its fully liquid SIMPEDES instrument, had become known in Central Java.That year the BKKs/CJ began offering a similar voluntary savings product called Tabungan Masyarakat Pedesaan, rural people's saving (TAMADES). But in sharp contrast to SIMPEDES,TAMADES grew slowly-even in the same vil- lages where SIMPEDES grew rapidly.There were two reasons for the slow growth ofTAMADES. Unlike BRI's unit desas, the BKKs/CJ required compulsory savings from borrowers. Mandatory savings typically discourage borrowers from placing voluntary savings in the same financial institution (see chapter 7). It is no co- incidence that early results showed that 75 percent ofTAMEDES savers were not BKK/CJ borrowers-although the savers fit the profile of potential BKK/CJ clients (Patten and Rosengard 1991, p. 39). The second reason for the slow growth ofTAMADES was that there was little incentive for the BKKs/CJ to market savings aggressively-because of the government's large directed credit program (KUK) that began in 1990.To ful- fill its KUK obligations to provide 20 percent of its loan portfolio to small busi- nesses, the state-owned Bank Ekspor-Impor Indonesia provided the BKKs/CJ with $34 million, far more than the system could safely absorb at the time.28 As a result the BKKs became overly liquid and deposited more than 60 per- 118 The Microfinance Revolution: Lessons from Indonesia cent of their funds at the CentralJava Provincial Development Bank (BPD; see Gonzalez-Vega and Chaves 1993 and Steinwand 2001). Over time, however, the BKKs/CJ invested a large part of these funds by increasing their loan portfolios.And the BKKs/CJ became more active in sav- ings mobilization. Steinwand (2001, p. 204) shows that by September 1998 an average BKK/CJ with a BPR license had $50,828 in outstanding loans and $34,653 in savings and deposits (80 percent in savings29 and 20 percent in fixed deposits).Thus 68 percent of the average loan portfolio was covered by savings and deposits; other sources of funds were retained earnings and bank loans. At the end of 1998 there were 512 BKKs/CJ, of which 204 had BPR licenses and were supervised by Bank Indonesia.The other 308 had applied for licenses and continued to be supervised by the BPD. BKK loans are As of 1997, the BKKs/CJ-both those that were licensed BPRs and those that were not-provided small loans and collected small deposits (with larger financed by savings, fixed deposit accounts raising the average account size).As percentages of 1997 per capita GNP ($1,110), the average outstanding loan at the BKKs/CJ was 8 retained earnings, percent, while the average savings balance was just 3 percent.The average time deposit balance was 78 percent (see table 9.5). and bank loans Nonperforming loans were 13 percent in 1997. The BKKs, like other BPRs, do not provision sufficiently for loan losses. Therefore, as discussed above, profitability is difficult to ascertain.30 Based on unadjusted figures for 1997, the return on assets for the BKK/CJ system was 4 percent and the return on equity was 21 percent. According to Steinwand's (2001, p. 249) worst-case re- serve-adjusted calculations, however, the 1997 return on assets would have been -3 percent. But with 512 BKKs and about 4,000 village posts in 1999, and with aver- age outstanding loans and savings balances below 10 percent of GNP per capi- ta, the BKKs/CJ-one of the best BPR systems in Indonesia-provide financial services to many low-income people throughout Central Java. Learningfrom the BKK/CJ. There is much to be learned from the BKK/CJ, which began in 1970, the same year as BRI's unit desa system. Over time, each learned from the other's successes and failures (and from its own). Not surprisingly, many of the lessons from the two systems are the same, as can be seen in the discussion of the unit desas in chapters 11-15, and many are complementary. In addition, many of these lessons reinforce the experiences of the private Bank Dagang Bali, which also began in 1970 (chapter 10). All underscore the fun- damental principles of commercial microfinance discussed in volume 1. Basic lessons from the history of the BKKs/CJ include: * Small, short-term individual loans to rural households can be provided by rural credit organizations owned by local governments at interest rates that cover costs and risks and enable profitability. * Such organizations can be managed with low operating costs, and can be suc- cessful even if open only one day a week. * Loans can be financed by retained earnings, savings, and commercial bank loans. Rural Development and Rural Financial Institutions in Indonesia 119 * Loans should have a range of maturities and repayment options suitable for the community. * Providing subsidized loans to borrowers leads to losses and can eventually result in unit closure or severe retrenchment. * Training and incentives for management and staff are essential. * Large external injections of liquidity discourage voluntary savings mobilization. * Small voluntary savings accounts can be cross-subsidized with larger time deposits. * Emphasis should be placed on appropriate loan loss provisioning and bad debt writeoff, a remaining problem at the BKKs/CJ. * Donors that support institutional capacity building for critical needs get max- With 512 BKKs imum return on their investments. * Links between microfinance institutions and local governments and banks and about 4, 000 can be advantageous for both in a variety of ways. village posts, many Adapting the BKK model to a different environment: the Badan Kredit Ke- poor people camatan of South Kalimantan (BKK/SK). The South Kalimantan BKK pro- gram, modeled on the one in Central Java, began in 1985 (Gonzalez-Vega and throughout Central Chaves 1993; Chaves and Gonzalez-Vega 1996; Ravicz 1998; Steinwand 2001). The two institutions are unrelated, however. The BKK/SK is owned by the Java have easy access provincial government of South Kalimantan and supervised by South Kali- mantan's Provincial Development Bank (BPD). The account below draws to credit and savings heavily on R. Marisol Ravicz's excellent paper, "Searching for Sustainable Mi- crofinance: A Review of Five Indonesian Initiatives" (1998). services South Kalimantan, located northeast of Central Java, has an ethnically het- erogeneous population of 2.9 million people, with only 79 people per square kilometer in 1995. Known as "the province of 1,000 rivers," much local trans- portation in South Kalimantan is by water; in many areas roads and other in- frastructure are not well developed. South Kalimantan, with its estate agriculture and forest areas, produces such export commodities as timber, rubber, rattan, and palm oil. Development of the BKK/SK. South Kalimantan's BKKs began in 1985 with modest facilities and capital endowment grants of about $5,000 each.3" The first 16 BKKs/SK were financed by USAID; the South Kalimantan provin- cial government then financed 94 BKKs/SK. Although their loan portfolios are not subsidized, the BKKs/SK are provided with indirect and in-kind subsidies. The province provides training and supervision, and provincial and district gov- ernments often provide office equipment, motorcycles, and bicycles (Ravicz 1998, p. 32). After the 1992 Banking Law was passed, the provincial government obtained BPR licenses for 20 of its 34 then-existing BKKs-which were exempt from the 1992 regulation that deposit-taking institutions opened after 1992 must ei- ther meet the minimum capital requirement or become cooperatives. But nei- ther of these two alternatives for new BKKs was acceptable to the South 120 The Microfinance Revolution: Lessons from Indonesia Table 9.6 Estimated earnings adjusted for subsidies and bad debt for 34 original BKKs/SK and 3 LKPs of Nusa Tenggara Barat, 1995 (percent) Eamings measure BKKs,SK IKPs Adjusted real return on assets 1 5 -5 2 Adjusted real return on equity 12 0 -2.8 Estimated subsidy dependence indexa 0 24 Current annual interest rate 64b 160 Annual interest rate required to cover all costs (including obtaining a market return on equity) 64 198 Note BKKI/SK stands for Badan Kredit Kecamatan (Subdistrict Credit Organization) of South Kalimantan Large external LKP stands for Lumbung Kredit Pedesaan (Rural Credit Organization) of Nusa Tenggara Barat a The percentage increase in the interest rate required for the institution to operate without subsidies (including obtaining a market return on equity) injections of b " BKK interest rates vary by loan size and repayment frequency This interest rate is for one of BKK's most common loans-a four month loan with a 2 5 percent per month flat interest rate, payable in monthly installments, with a 5 percent forced savings requirement Five percent was selected for the liquidity discourage savings requirement because some BKKs require a 10 percent savings component while others require none " lRavicz 1998, p 37) Source Ravicz 1998, pp 37, 48 voluntary savings Kalimantan BPD.Therefore the provincial government stopped creating BKKs mobilization and began instead to build Institutions for Small Enterprise Finance (Lemba- ga Pembiayaan Usaha Kecil, or LPUKs).The one difference between BKKs and LPUKs, an important one, is that the older BKKs are permitted to accept de- posits (both voluntary and compulsory), while the newer LPUKs are not. By 1995 there were 76 LPUKs in addition to the original 34 BKKs-one in each of the province's 109 subdistricts and one additional unit (table 9.6). Following Ravicz (1998), here the 110 BKKs and LPUKs are referred to col- lectively as BKKs/SK except when the two categories are compared. The BKK/SK system is supervised by the district branches of the South Kahman- tan Provincial Development Bank (BPD). The early 1990s were difficult for the BKKs/SJ. At the same time as the system was expanding with its new LPUKs, the BKKs/SJ faced political de- mands to lend in inappropriate ways."It appears that political pressures on lend- ing policy in 1992 (caused by the election in that year) led to unprecedented arrears in 1993" (Ravicz 1998, p. 36). Loans in default at the 34 original BKKs, adjusted for defaulted loans not written off in previous years, were 23 percent of total loan volume in 1993; loans in arrears fewer than 90 days were 27 percent (Ravicz 1998, pp 35-36). But as discussed below, performance improved rapidly and the 34 original BKKs "did not require subsidies to operate profitably in 1994 and 1995, and could have paid equity holders a satisfactory return" (Ravicz 1998, p. 37). Informa- tion is not available about the defaults, profitabdity, or sustainability of the LPUKs. The BKKs/SK are each staffed by three to six employees. Staff training is emphasized, and a performance-based incentive system has been developed under which managers and staff share 18.2 percent of their BKK/SK's nominal prof- Rural Development and Rural Financial Institutions in Indonesia 121 its.The incentive is provided in bonuses and through a welfare fund. In 1995 each employee at the average BKK/SK received 669,000 rupiah ($290), equiv- alent to 1.6-5.4 months of salary (Ravicz 1998, p.34). BKK/SK managers are generally active, and the BPD provides regular supervision. In theory, the BPD makes all policy decisions for the BKKs... .including the types of savings and lending products units can offer, the terms [of] those instruments, how they should provision for bad debt, when they should write off loans, what their underwriting and loan servicing procedures should be, whom they should hire, how they should train staff, etc. In The first 16 South practice, individual unit managers frequently assume at least minor levels of discretion with regard to many of these policies. The Kalimantan BKKs BPD has one full-time supervisor for every 8.5 BKK units. Supervisors visit units from one to four times per month. werefinanced by -Ravicz 1998, pp. 32-33 USAID, and the Interest rates for loans range from a low of 3.5 percent a month on a de- climnng balance basis with no savings requirement (equivalent to 51 percent a provincial year) to a high of 1 percent a week on the original loan balance with a 10 per- cent savings requirement (about 196 percent a year on a declining balance basis). government then The most frequently quoted rate is a 2.5 percent flat rate a month, with a 10 percent compulsory savings requirement at the 34 original BKKs/SK, but no financed 94 others - mandatory savings at the newer LPUKs. For a four-month loan on these terms, the effective interest rate is about 3.9 percent a month, or 59 percent a an exemplary use of year, assuming no compulsory savings requirement. For the original BKKs/SK that require compulsory savings, the effective interest rate is 4.5 percent a donorfunds month, or 70 percent a year (Ravicz 1998, p. 33). Loan maturities, which vary by unit, range from 10 weeks to 18 months; repayment schedules range from weekly to monthly. Until 1995 all loans were made to individuals. But that year a pilot project in lending to self-help groups was started in one BKK/SK as an effort to reduce transaction costs. This ex- periment in group lending was influenced by the Program to Link Banks and Self-Help Groups (Program Hubungan Bank dan Kelompok Swadaya Masyarakat, or PHBK, discussed later in this chapter). But before the initial pilot project was completed and evaluated, other BKKs/SK adopted the idea. Ravicz (1998) re- ports that 59 self-help groups received loans from BKKs/SK in 1995. The BKKs/SK provide loans to "'channeling groups that simply pass loan funds down to their members and pass up repayments.The groups do not attempt to act as financial intermediaries, and there are no NGOs involved in the process" (p. 33). Loan sizes in the BKK/SK system range from $21-428, but the ranges are different in different BKKs/SK. In 1995 the BKK/SK system had $3.4 million outstanding in more than 34,500 loans (see table 9.5).32 In 1995 the average loan was 10 percent of GDP per capita (Ravicz 1998, p. ix). Savings in the 34 original BKKs was $427,000 in 1995. 122 The Microfinance Revolution: Lessons from Indonesia The amount of loans in arrears by more than 90 days was 26 percent of the value of outstanding loans in 1995. But since the units do not write off bad loans, Ravicz (1998, p. 35) recalculated the arrears rate to simulate arrears and default rates that would have been reported if the institution had written off 100 percent of its loans in default each year. Calculated in that way, and with all loans overdue more than 90 days considered to be in default, 5.9 percent of net outstanding loans were in arrears and 3.5 percent of loans were in default. Ravicz (1998, pp. 36-37) also calculated the real return on average assets for the 34 original BKKs, adjusted for subsidies and bad debt.Although the ad- justed return on assets was -20.9 for 1993, it was 0.5 percent for 1994 and 1.5 percent for 1995. The subsidy dependence index (the percentage increase in the interest rate "The higher the required for an institution to be fully self-supporting, including a market rate of return to equity holders) for the 34 BKKs was 118 in 1993. But the index share of savings to for the 34 BKKs was 0 in both 1994 and 1995 (see table 9.6).Thus in aggre- gate the 34 BKKs were financially self-sufficient in both 1994 and 1995. Op- total assets [in erating in a province with only 79 people per square kilometer (in 1995), the BKKs/SK face little competition.Their main competitors are BRI's unit desas BPRs], thefaster and informal moneylenders. the growth, the The only other major source of financing available to BKK clients is loans provided by cooperative leaders. These individuals use higher the their position in the cooperatives to function as private moneylenders.The rate they charge (20 percent flat per month) profitability, and the is about three times as high as the highest rate BKK units charge. -Ravicz 1998, p. 42 better the loan Learningfrom the BKK/SK. Only half as old as the BKK/CJ system, the performance" BKK/SK has already demonstrated many of the same lessons as its CentralJava counterpart.The first four lessons from the BKK/CJ, above, concern basic is- sues of commercial niicrofinance in rural areas. These fundamental principles are embedded in the BKK/SK as well. In addition, the importance of training and incentive systems in both BKK systems emerges clearly, as does the con- tinuing need in both systems for appropriate loan loss provisioning and bad debt writeoff. For both BKK/CJ and BKK/SK, a donor (USAID in both cases) played an important early role in building capacity; for the BKK/SK this involved fi- nancing the first of the BKK units. But in both cases the provincial govern- ment has played the leading role in developing and supervising the BKKs. Both systems went through difficult times, largely for political reasons, and through subsequent recovery periods. But in 1995 both BKK systems were at or close to financial self sufficiency, and both had widespread outreach in their provinces. In addition, both had average loan sizes that are very small relative to per capita GNP. One specific lesson from the BKK/SK is how the South Kalimantan provincial government adapted a system developed under the very different con- Rural Development and Rural Financial Institutions in Indonesia 123 ditions of Central Java. Both systems have BKKs in nearly every subdistrict. But as noted, there are vast differences between South Kalimantan and CentralJava in population density and infrastructure development, as well as in other as- pects of their environments, that affect the provision of financial services to the rural population.33 Some differences between the two BKK systems that are likely related to their different environments are: * While the BKKs/CJ have developed a huge network of village posts, this approach has not been adopted for the BKKs/SK.Village posts would like- ly not be cost-effective for many sparsely populated areas of South Kalimantan. The challenge is in . Unlike the BKKs/CJ, the BKKs/SK began lending to self-help groups in 1995, based on the PHBK model (although the PHBK program had not yet adapting the basic begun in South Kalimantan). But this form of lending had just begun at the BKKs/SK when Ravicz was there in 1995, and information on the perfor- principles of mance of the group loans is not available. And the BKKs/SK did not wait until their pilot project was completed and evaluated before expanding their commercial self-help group lending to multiple BKKs/SK. Elsewhere in the country, the PHBK program developed into a large, complicated, and high-cost program microfinance to providing credit and training. The BKKs/SK, however, opted for a pared- down version of the PHBK model. If the BKKs/SK can extricate the lend- diferent ing methodology from the rest of the high-cost PHBK program-as was done on Java by the private BPR Bank Shinta Daya, discussed below-they might environments find the self-help group lending methodology useful in South Kalimantan's sparsely populated environment, where BKK/SK staff travel long distances on poor roads. * Perhaps the most critical problem for most BKKs/SK (the newer LPUKs) is that they are not permitted to mobilize savings. Lending growth is slow because of lack of funds, and of course these BKKs/SK cannot meet the de- mand for savings services in their subdistricts. The BPD is aware that the BKK/SKs have few staff and limited supervision, serve large areas, and har- bor substantial opportunities for theft and corruption.These factors may dis- courage the provincial government from changing the status of some LPUKs. But other provinces have moved toward solving the problem of savings in new institutions by establishing them as licensed BPRs or as what Ravicz calls "nominal cooperatives" (see the discussion below of the Lumbung Kredit Pedesaan of NusaTenggara Timur). Savings remains an unresolved prob- lem for the LPULKs of South Kalimantan. Because data are not available on the performance of the LPUKs alone, it is not possible to assess the differ- ent levels of outreach and profitability of the 34 original BKKs and the 76 LPUKs. But Steinwand's (2001, p. 324) general comment about BPRs is like- ly to be relevant: "The long term performance of the BPR is closely linked to their capacity of savings mobilization.The higher the share of savings [to] total assets the faster the growth, the higher the profitability, and the better the loan performance." 124 The Microfinance Revolution: Lessons from Indonesia Indonesia is a large country with many environments. What works at the local level in rural areas in Central Java may not work as well in South Kali- mantan, and vice versa.Thus the central government left the detailed decisions about unlicensed BPRs to their respective BPDs."The central government. . .left enough scope for the provincial governments to follow their own LDPK pol- icy, as the different approaches of the BKK of South Kalimantan, LPD of Bali, and LKP [Lumbung Kredit Pedesaan] of NTB [Nusa Tenggara Barat] have proven" (Steinwand 2001, p. 217). But there is no question that important lessons about commercial micro- finance can be, and are, passed from one BPR to another. The challenge is in adapting the basic principles to different environments. Some cooperative "hen basic principles are lacking: Lumbung Kredit Pedesaan (LKP) of Nusa Tenggara Barat. The Rural Credit Organizations (LKPs) of NusaTenggara Barat leaders are also in eastern Indonesia are owned by the provincial government of Nusa Teng- gara Barat and supervised by the Provincial Development Bank (BPD). Rav- private icz (1998, pp. 43-51) analyzes the 1995 performance of the four LKPs in Dompu District (see tables 9.5 and 9.6). Like the two BKK systems discussed above, moneylenders, these Rural Credit Organizations are public LDPKs. Nusa Tengarra Barat is a poor province with an ethnically and religious- charging aflat ly diverse population of 3.6 million; in 1995 it had 181 people per square kilo- meter. NusaTengarra Barat is generally dry (although there is variation in rainfall monthly interest rate within the province) and has few commercially valuable natural resources and little industry. In some areas, especially Lombok, irrigated wet rice is culti- of 20 percent (an vated. But much of the province's population relies on fishing or dryland sub- sistence agriculture (sorghum, mnllet, mungbeans, cassava, sweet potatoes, annual effective rate yams, and the like).Among Indonesia's provinces, NusaTengarra Barat has one of the highest percentages of population below the poverty line (18 percent of 1, 600 percent) in 1996).3 The Dompu District LKPs were subject to political pressures and were gen- erally not well supervised. With one exception, they also appear not to have been well managed. Compared with the BKK systems of CentralJava and South Kalimantan, Dompu District's LKPs had low transparency and higher defaults and arrears. Although they had higher interest rates on loans than the BKKs, the Dompu LKPs were much further from financial self-sufficiency than the other two systems. Ravicz (1998, p. 50) commented,"The units [LKPs] are heav- ily influenced by the expertise and honesty of their management. For exam- ple, one unit was driven to the brink of insolvency through corrupt management practices... Further.. .the units are owned by the NTB provincial government and controlled by the provincial development bank. It appears likely that both of these entities have objectives for the units beyond profitability, growth, and client service." Development of Nusa Tenggara Barat's LKPs. Introduced in 1989, the LKPs are located in subdistrict capitals in a relatively poor province. In 1995 Dompu District's four LKPs covered 34 of the 45 villages in their subdistricts Rural Development and Rural Financial Institutions in Indonesia 125 (76 percent), but the two other subdistricts in the district were not served by LKPs. Dompu's LKPs were all established before 1992 and so were exempt from the government's regulations on LKPs opened after that year. Nevertheless, in 1995 the BPD was planning to convert two of the four LKPs mto licensed BPRs. For the other two, the BPD plan was "to comply with the letter of the law by nominally declaring the units to be cooperatives and redefining voluntary sav- ings as 'required members savings"' (Ravicz 1998, p. 51).The BPD also planned to open two other LKPs as nominal cooperatives in the two unserved Dompu subdistricts. Dompu's LKPs offer one loan product that is common to many BPR sys- The Rural Credit tems: a 12-week loan to individual borrowers repayable in 12 equal weekly in- stallments.The first installment represents the interest due on the loan, the second Organizations is a compulsory savings payment, and the final 10 installments repay the capi- tal. Assuming that the compulsory savings and the interest earned on the sav- (LKPs) of a Nusa ings are paid to the borrower at the end of the loan period, as is supposed to happen, Ravicz (1998) calculated the monthly effective interest rate to be 8.3 Tenggara Barat percent and the annual effective rate to be 160 percent (compared with 59-70 percent at the BKKs/SK). district were subject In 1995 loan sizes ranged from $21-214. One voluntary savings instrument is offered, a demand deposit with an unlimited number of withdrawals and a to political pressures, 10 percent annual interest rate. Compulsory savings, required to obtain loans, also pay 10 percent interest a year. poorly supervised, In 1995 the outstanding portfolio in the four LKPs was $313,000 in about 4,150 loans (see table 9.5).The largest LKP had 47 percent of the value of out- and in some cases standing loans and the smallest had 4 percent. Savings (voluntary and compul- sory) were $28,000 in 1995. corrupt Ravicz (1998, p. 46) notes, however, that "income statements and balance sheets provided by the BPD contain some apparent errors, omissions, and in- consistencies."The problems were not only at the BPD. "The Dompu district branch of the BPD has one full-time supervisor to oversee the four LKPs in its district.Two of these units are within walking distance of the BPD branch. Despite the fact that the supervisor oversees only four units, one unit manag- er was able to effectively bankrupt his unit by making fraudulent loans and em- bezzling funds" (Ravicz 1998, p. 43). The LKPs receive endowment grants and in-kind gifts and subsidies, in- cluding supervision by the BPD. Like the two LDKPs discussed, the LKPs do not provision adequately for loan losses and do not write off bad debt. Adjusting for subsidies, bad debt, and inflation, Ravicz found that three of the LKPs had average arrears (payments overdue 30 days or more) of 20 per- cent of outstanding loans in 1995.The fourth LKP had had the highest adjusted arrears in 1993-78 percent. But although this LKP continued to operate, data on its performance in subsequent years could not be obtained. Excluding the fourth LKP, adjusted loans in default (in arrears 90 days or more) averaged 5.5 percent in 1995. 126 The Microfinance Revolution: Lessons from Indonesia The adjusted real return on average assets for the three LKPs was -5.2 per- cent in 1995, and the adjusted real return on average equity was -2.8 percent (see table 9.6).These returns would have been lower had the LKP with the miss- ing data been included. Based on the three LKPs, "in 1995, the program would have had to charge a 198 percent annual interest rate to completely eliminate all program subsidies and pay liability and equity holders a market return on their investments" (Ravicz 1998, p. 48; see table 9.6). There were, however, substantial differences in performance among the LKPs.The best-performing LKP maintained consistently low default rates over the three years examined and in 1995 had an adjusted real return on assets of 4.9 percent, an adjusted real return on equity of 11.9 percent, and a subsidy depen- dence index of -4 percent.Thls LKP "could have provided an adjusted real re- Good management turn to equity holders of from 3 to 12 percent since 1993" (Ravicz 1998,p. 48). The LKPs report almost no institutional competition for borrowers, but they matters. Even in a compete with the BRI unit desas for savings. Like the BKKs/SK, LKPs report that "cooperative leaders use their position in the cooperative system to function as politicized, corrupt private moneylenders, charging a 20 percent flat monthly interest rate (equiva- lent to 1,600 percent per year on a declimng balance basis)" (Ravicz 1998, p. 51). environment, a good Learningfrom the LKPs. Although the data cover only the four LKPs of one district, much can be learned: manager can * Politicization is typically incompatible with accountable supervision and with generate good results profitable commercial microfinance institutions. * When owners and managers lack expertise and honesty, the institutions they manage are likely to be unsound, and their clients poorly served. * Lack of transparency occurs when owners, managers, and staff are not held accountable for their decisions and actions. * High defaults and arrears put poor savers' money at risk. * Inefficiency and corruption can lead to high interest rates on loans, unnec- essarily penalizing borrowers. * Appropriate products are necessary but insufficient for success in mi- crolending (and in savings).The LKP loan product (except for its interest rate) is similar to loan products in BPRs all over Indonesia.Where it is implemented well, this product can serve as a basis for a sustainable loan program with wide outreach. But when the institution is not properly managed, a loan program using the same product can show unsatisfactory performance. * Good management matters. Even in a poor region, and in a BPR system marked by politicization and corruption, a good manager can generate con- sistently good results in a particular unit. A licensed private BPR: serving the poor profitably at Bank Shinta Daya inJava. Bank Shinta Daya is in D.I.Yogyakarta, a special region of Indonesia (D.I. stands for Daerah Istimewa, which means "Exceptional Region"). Border- ing the province of Central Java toward the southeast, D.I.Yogyakarta holds a Rural Development and Rural Financial Institutions in Indonesia 127 special status as a sultanate in the Republic of Indonesia.Yogyakarta is small and densely populated (916 people per square kilometer in 1995); its area and its population are one-tenth those of Central Java. ButYogyakarta is a cultural and intellectual center and former capital of Indonesia. It is world renowned for its history, art, music, dance, ancient Hindu and Buddhist temples and statues, and its kraton, the palace compound of the sultans ofYogyakarta. Like Central Java, Yogyakarta has both wet rice and dryland cultivation and has well-developed irrigation and infrastructure. Bank Shinta Daya was founded inYogyakarta in 1970 as a limited liability company. A rural bank, it was established with private investment and funded its expansion from retained earnings.The account below is based on Seibel and At Bank Shinta Parhusip (1998) and Parhusip and Selbel (2000), unusual papers that analyze types of data rarely available for microfinance institutions. Daya 71 percent of Development of Bank Shinta Daya as a private rural bank. For its first 20 years Bank Shinta Daya mobilized savings and lent to individual clients, poor and loans made to poor nonpoor. In 1989 the bank began to participate in Bank Indonesia's Program to Link Banks and Self-Help Groups (PHBK), and by the end of 1995 Bank borrowers were Shinta Daya provided financial services through 310 groups with 7,750 mem- bers.At that point the bank had a total of more than 30,000 savers (with 43,000 through se!f-help accounts) and more than 12,500 borrowers (see table 9.5). It is important to note, especially in light of the costly PHBK program discussed below, that: groups. But 71 After an expensive and ultimately abortive attempt of working percent of poor through a local NGO as a financial intermediary, Bank Shinta Daya decided to seek out its own savings and credit groups which savers were are ubiquitous in Indonesia.The bank has set up a special group lending department, hired its own field workers, some of them individuals former NGO staff, and trains its own groups of the poor in group management, bookkeeping, savings mobilization and financial management. -Seibel and Parhusip 1998, p. 9 Bank Shinta Daya classifies most of its savers and borrowers as poor, but it has far more poor savers than poor borrowers. In 1995, 25,752 savers (85 per- cent of total savers) and 8,782 borrowers (69 percent of total borrowers) were poor. That poor savers outnumbered poor borrowers nearly three to one sup- ports data from other Indonesian microfinance institutions (and from other parts of the world) that more poor people want to save than borrow at any given time (see chapters 10 and 13; see Rutherford 2000; and Wright 2001).Among the nonpoor this pattern is less pronounced: Bank Shinta Daya's 4,588 non- poor savers do not greatly outnumber it 3,874 nonpoor borrowers. An especially interesting finding is that among poor borrowers, 2,582 (29 percent) were individual borrowers and 6,200 (71 percent) were from self-help groups. But exactly the opposite pattern held for poor savers: 18,352 (71 per- cent) were individual savers and 7,400 (29 percent) were from self-help groups. 128 The Microfinance Revolution: Lessons from Indonesia Table 9.7 Viability indicators for Bank Shinta Daya and Mitra Karya, 1995 (percent) Indicator Bank Shinta Daya Mitra Karya Average annual interest rate on deposits 15 2 10 0 Annual effective interest rate to borrowers 27 4 30 0 Average cost of outside funds 16.5 3.5 Financial self-sufficiency 96 0 39.0 Source Parhusip and Seibel 2000, p 171 Individual and These data also suggest what has been shown elsewhere: savers value confi- group methodologies dentiality and prefer to save as individuals rather than through groups (see chap- ter 7). were both profitable At Bank Shinta Daya the average annual effective interest rate charged to borrowers in 1995 was 27.4 percent-a long way from the 160 percent charged at Bank Shinta at the Nusa Tenggara Barat LPKs that year (and the 198 percent the LPKs would have had to charge to become profitable). Bank Shinta Daya's default rate was Daya. But the 2 percent.The average annual interest rate paid to depositors was 15.2 percent. Savings of $2.2 million represented 105 percent of the $2.1 million in outstanding individual loans in 1995.The bank's net worth at the end of 1995 was $215,000, with 36 percent in capital and 64 percent in retained earnings. Net profit was $27,426. methodology Seibel and Parhusip (1998,p. 16) calculate that the bank was 96 percent financially self-sufficient (table 9.7). accountedfor 94 As noted, Selbel and Parhusip provide information rarely available for mi- crofinance institutions. This includes a comparison of the outreach and prof- percent of the profits itability of Bank Shinta Daya's individual and group methodologies. Their findings show that group loans and group savings were of only minor impor- tance in 1995: 97 percent of the value of the bank's deposits came from indi- vidual clients, and 89 percent of the value of outstanding loans was in loans to individual borrowers.Although Siebel and Parhusip found that the individual and group methodologies were both profitable to the bank in 1995, the indi- vidual methodology accounted for 94 percent of profits. The group technology is thus found by Bank Shinta Daya to be viable as such, but adds little to the bank's overall viability. Why then does the bank engage in business with small groups? The bank's management explains this with future expectations. By providing financial services to group members with microenterprise activities, it contributes to their growth.As the members' microenterprises grow, so will their business with the bank. -Seibel and Parhusip 1998, pp. 16-17 Rural Development and Rural Financial Institutions in Indonesia 129 Learningfrom Bank Shinta Daya. In addition to the basic principles for com- mercial mucrofinance mentioned above for other institutions, Bank Shinta Daya's performance offers important lessons: * Licensed private BPRs can reach the poor sustainably (as can public BPRs of all types). * Savers, both poor and nonpoor, will save in licensed private BPRs if the BPRs demonstrate that they are trustworthy. * The poor are far more likely to use the bank's savings facilities than its cred- it services.The nonpoor are somewhat more likely than the poor to be bor- rowers. Mitra Karya, a . Although most savers are poor (89 percent), most of the value of the savings is from the nonpoor (thus raising the average account size and making it pos- Grameen replicator sible for the bank to collect savings profitably while also serving the poor). * Bank Shinta Daya, which operates its own group lending department, pro- in EastJava, had an vides both loans and voluntary savings services to self-help groups profitably (in contrast to the high-cost PHBK program discussed below). average cost offunds . Self-help groups can deepen credit markets, increasing poor people's access to commercial loans. of 3.5 percent and * Both individual and group methodologies are profitable at Bank Shinta Daya. But 94 percent of the bank's profit is from its individual methodology. was 39 percent . Bank Shinta Daya provides support for the view that BPRs that serve as fi- nancial intermediaries are more profitable and achieve higher outreach than financially se!f- those that do not mobilize voluntary savings from the public. sufficient. Its Missing the point: Mitra Karya of East Java, a Grameen Bank replicator. outreach was low The information on Mitra Karya here is drawn from Seibel and Parhusip (1998) and Parhusip and Seibel (2000).The Mitra Karya Grameen replication project is in East Java-the largest in area of the provinces on Java. In 1995 the population of East Java was 33.8 million, with 706 people per square kilome- ter. EastJava is well known as the home of earlyJavanese kingdoms, culminatmg in the Majapahit kingom-a Hindu court culture in the 14th century with an empire that stretched from parts of mainland Southeast Asia in the west to set- tlements on Irian Jaya in the east (covering a larger area than present-day In- donesia). Today East Java combines well-developed infrastructure and highly intensified agriculture (irrigated rice with two or three harvests a year, as well as many other crops) with major industrial and commercial businesses centered around Surabaya, Indonesia's second largest city after Jakarta. Development of Mitra Karya. Mitra Karya was established as a replication pro- ject in 1993 with funding from the Grameen Trust Fund and the state-owned Bank Negara Indonesia. The project is implemented by the Research Center of Brawijaya University in East Java. Mitra Karya began with 105 participants in 21 small groups in 1993; in 1995 it had 225 groups with 1,125 members, all of them women. 130 The Microfinance Revolution: Lessons from Indonesia Mitra Karya follows closely the relatively cosdy Grameen small group de- livery system, adhering to a group size of five-which is small by the stan- dards of Javanese self-help groups. Mitra Karya serves only poor borrowers and requires compulsory savings from all members. The institution's aim is to reach what it considers an optimum size of 2,000 members in 400 groups. As Seibel and Parhusip (1998, p, 9) point out, however, Mitra Karya may run into difficulty because it is unlikely to meet the regulations for BPR status when it ends its transition period and must register as a formal rural bank (BPR). In 1995 Mitra Karya had about $32,000 in outstanding loans and $5,300 in compulsory savings (see table 9.5). Its ontime repayment rate was 98 per- cent and its default rate was 1 percent. However, Mitra Karya's highly subsi- Mitra Karya "has dized average cost of outside funds was just 3.5 percent, and it had reached only 39 percent financial self-sufficiency (see table 9.7). not been able to Grameen replicator [Mitra Karya] reports liabilities in the form demonstrate ... that of soft loans received (outstanding as of 12/1995) to the amount of US$46,507 (about half from government-owned replication of the BNI [Bank Negara Indonesia] and half from Grameen Trust Fund) compared to assets in the form of loans outstanding Grameen approach amounting to a mere US$32,104.With an interest rate of 10 per cent on savings deposits and 3.5 per cent on soft loans, a in Indonesia lending interest rate of 30 per cent and an on-time repayment rate of 98 per cent, its spread should be more than sufficient may .. improve the to cover its costs.. .However, given its dependency on donor funds, its degree of full financial self-sufficiency is only 39 per poor' access to cent.. .With total assets amounting to Rp 127 million ($55,098),loans outstanding of Rp 74 million ($32,104) and financial services" a deposit base of Rp 12.3 million ($5,336) and virtually no equity base, [Mitra Karya] would have to come a long way to grow into a formal village bank, a BPR, which would require a minimum paid-in equity capital of Rp 50 million ($21,692). So far [Mitra Karya] has not been able to demonstrate convincingly that the replication of the Grameen Bank approach in Indonesia may substantially improve the poor's access to financial services. -Parhusip and Seibel 2000, p. 169 Learningfrom Mitra Karya. Mitra Karya seeks to replicate the Grameen Bank microcredit methodology in East Java. Will doing so benefit East Java or In- donesia? It seems unlikely for several reasons. First, Mitra Karya sees itself essentially as a clone.Accordingly, it has not adapt- ed its products and services to the Indonesian context. Rural Development and Rural Financial Institutions in Indonesia 131 Second, Mitra Karya promulgates the Grameen microcredit focus. But In- donesia's generic BPRs, operating on a model three-quarters of a century older than Grameen's, have generally succeeded with a commercial financial inter- mediation model. Even with its many variations, Indonesia's BPR model has four basic characteristics: it operates according to commercial principles; it typ- ically provides financial intermediation, cross-subsidizing services to the poor with services to the nonpoor; it experiments and makes changes as needed; and it allows failing institutions to die. In contrast, Mitra Karya offers a rigid, tar- geted microcredit program-instead of a flexible microfinance system-to a province that already has hundreds of BPRs suited to EastJava's conditions and sensibilities. "In general the Third, Mitra Karya raises some important issues for governments and donors.With its dependence on donor funds at an annual interest rate of 3.5 higher the number of percent, Mitra Karya's degree of financial self-sufficiency was just 39 percent in 1995. Moreover, Mitra Karya has a small deposit base and almost no equi- financial ty. Indonesia's microfinance models are far more appropriate for the country's more than 200 million people. In this context, and considering the thrust of intermediaries the 1992 Banking Law, why should an Indonesian state bank use its scarce re- sources to provide subsidies for an unsustainable credit-driven project aimed between the at reaching 2,000 people? Donors to Indonesia and the Indonesian government can make better use of their resources for microfinance. originating Combining large program subsidies with high borrower interest rates: the Pro- institutions and the gram to Link Banks and Self-Help Groups. The Program to Link Banks and Self-Help Groups (Program Hubungan Bank dan Kelompak Swadaya Masyarakat, end user, the higher or PHBK) is a credit and training program sponsored by Bank Indonesia and the German government's Agency for Technical Cooperation (Gesellschaft fur the interest rate Technische Zusammenarbeit, or GTZ).The heavily subsidized program began in 1989 and by 1996 was operating in 16 provinces.The PHBK provides tech- paid" nical assistance to banks, NGOs, and the self-help borrower groups to which the banks and NGOs lend. This account relies heavily on Ravicz (1998) and McGuire, Conroy andThapa (1998); see also Gonzalez-Vega and Chaves (1993) and Chaves and Gonzalez-Vega (1996). Development of the PHBK. In the provinces where it operates, the PHBK is managed from Bank Indonesia's branch offices.There are three PHBK lending models. In the first, savings and credit groups (Kelompok Simpan Pinjam, or KSP) serve as financial intermediaries, receiving group loans directly from banks and onlending the funds to their members. NGOs, compensated by the program, train the groups in record keeping and financial skills, providing support to the groups for up to nine months.The program also trains the participating banks, but does not compensate them.The second model is similar except that the NGOs borrow from banks and onlend to borrower groups. In the third model, banks lend to channeling groups (Kelompok Pengusaha Mikro, or KPM), with each member usually receiving an equal portion of the loan. NGOs are not involved (this was the model informally adapted by the BKKs/SK before the PHBK began 132 The Microfinance Revolution: Lessons from Indonesia in South Kalimantan).After 1992 the PHBK did not provide liquidity support for loans made under any of the models. In areas with a high concentration of banking facilities, the first model was being phased out by the mud-1990s because it required considerable investment in training and support. The second model did not work well, primarily be- cause of NGOs' poor performance as financial intermediaries, and is general- ly discouraged except on some of the Outer Islands.Thus in Java, Bali, and other areas with developed banking facilities, the program focuses on the third model (which, however, can have higher costs for banks than the other models). But in 1996, in more sparsely populated areas with a low density of rural financial organizations, the first and second models continued. By September 1996,410 banks and bank branches, 183 NGOs, and 6,800 PHBK, a large and self-help groups were participating in the PHBK (McGuire, Conroy, and Thapa 1998, p. 156). Most of the participating banks were private rural banks (BPRs), heavily subsidized followed by state-owned commercial banks. Some private commercial banks and Regional Development Banks (BPDs) also participated. The banks and credit and training NGOs were supervised by Bank Indonesia staff and consultants. Groups that function as financial intermediaries are supervised by NGOs (the first model), program, charged while groups that simply channel credits to members are supervised by banks (the third model). Ravicz (1998, p.54) comments,"This supervision is very ex- end borrowers pensive for the more complicated lending models and in remote areas. The three types of intermediaries set their own interest rates and loan annual interest rates terms, which vary considerably. Loan terms range from 10 weeks with daily re- payments to 12 months with monthly repayments. "In general, the larger the rangingfrom 46 number of financial intermediaries between the originating mstitutions and the end user, the higher the interest rate paid by the end user" (Ravicz 1998, p. 54). percent to 450 Rates also tend to be higher in remote areas and for loans with daily repayments. percent In Sumbawa, most banks were lending at 2.5 to 3.0 flat rate per month; NGOs that onlent to groups were lending at 5.0 to 5.5 percent flat rate per month; and credit groups were lending to members at a flat rate of 5.5 to 7.5 percent per month (including a 1.5 percent flat rate required savings component). On a declimng balance basis, and including fees and forced savings requirements, rates generally ranged from about 3 to 15 percent per month or 100 to 450 percent per year for end users. However in less remote locations, banks lending to channeling groups may charge rates as low as 1.9 percent per month on a declining balance basis. This rate, including fees and forced savings requirements, can translate into interest rates as low as about 46 percent per year for end users. -Ravicz 1998, p.54 All the participating groups offer members voluntary savings accounts, with a range of interest rates. In 1995 demand deposit accounts paid 12-16 percent Rural Development and Rural Financial Institutions in Indonesia 133 a year, while the interest on time deposits ranged from about 16-22 percent for a one-year deposit. Total loan volume for fiscal 1995/96 was $5.5 million; 2,931 group loans were issued that year. Arrears were higher in the program's earlier years but by 31 March 1996, the end of the fiscal year, 5.7 percent of the volume of out- standing loans was one day or more in arrears. The program's default rate for that year was estimated at less than 4 percent. Compulsory savings were $900,000 as of the same date, and equaled 19 percent of outstanding loans (see table 9.5). Data for voluntary savings were not available (Ravicz 1998, pp. 57-61). The costs of the PHBK are very high, equal to 95 percent of annual loan volume for fiscal 1995/96 (down from 357 percent in 1990/91).The cost per Infiscal 1995/96 credit group was $1,785 that year (down from $11,654 in 1992/93).These costs do not include the costs that banks incur in making PHBK-related loans."The PHBK costs were [PHBK] costs ... effectively translate into very large subsidies for borrowers" (Rav- icz 1998, p. 59). 95 percent of What is being achieved for these high costs? Is the PHBK effectively as- sisting large numbers of the poor? There is not enough evidence for a defini- annual loan volume. tive answer. But a 1993 impact study found that less than 20 percent of respondents had monthly household spending below the poverty line and 40 But PHBK "cannot percent had household spending more than twice the poverty line. Ravicz (1998, p. 61) concludes: "If these findings are representative of borrowers under the be considered to be program today, then this program cannot be considered to be primarily reach- ing the low-income population, although some low-income households are being primarily reaching served." Calculating the subsidy dependence index for the PHBK, Ravicz found that the low-income in fiscal 1995/96 the participating banks would have had to increase their an- nual lending interest rate by 158 percent to pay for the full cost of the pro- population" gram (down from 628 percent in fiscal 1990/91); see table 9.8. Estimating the current lower-bound average annual interest rate at 107 percent, Ravicz cal- culated that in fiscal 1995/96 annual interest rates on PHBK loans would have had to be 277 percent to fund full program costs (Ravicz 1998, p. 60). However, the PHBK cites unfair competition from the even more highly subsidized KUKESRA group lending program, which is implemented by the Family Planning Coordinating Board, and from the P4K program, which is ad- ministered by BRI's branches (see chapter 14) and provides loans at about one- fifth the effective rate charged by banks participating in the PHBK (Ravicz 1998, p. 63). Learningfrom the PHBK. The PHBK is a large, expensive, and unnecessar- ily complicated program. But to its credit, the PHBK spread widely in Indonesia the idea that providing financial services to self-help groups can be a useful ap- proach to microfinance. One result is exemplified by Bank Shmnta Daya: the bank learned how to modify the PHBK approach to develop a much less expensive model. Steinwand (2001, pp. 77-78) points out that the PHBK repeated mistakes made in Indonesia nearly a century earlier. 134 The Microfinance Revolution: Lessons from Indonesia Table 9.8 Estimated subsidy dependence index and required interest rates to cover the program costs of the PHBK program, fiscal 1991-96 (percent) Indicator 1990/91 1991/92 1992/93 1993/94 1199495 1995/96 Lower-bound estimate for SDI with full program costs 628 515 574 471 298 158 Lower-bound estimate for SDI with direct program costs only 163 145 163 140 98 48 Lower-bound estimated average annual current interest ratea 107 107 107 107 107 107 Estimated average annual interest rate required to fund full program costs 781 659 723 613 427 277 Estimated average annual interest rate required to fund direct program costs 282 263 282 257 212 159 Note The subsidy dependence index (SDI) is the percentage increase in the interest rate required for the program to operate without subsidies PHBK stands for Program Hubungan dan Kelompok Swadaya Masyarakat (Program to Link Bank and Self-help Groups) a Calculated on a declining balance basis, including all fees and forced savings requirements An estimated average of the rates commonly charged by rural banks and commercial banks to channeling groups Source Ravicz 1998, p 60 DeWolff von Westerrode and his successor, Carpentier Alting, propagated the idea of group-lending among the district banks already in the 1900s. . These early joint liability groups had up to 10 members and a group manager.. .who worked as an intermediary between the group members and the bank... .After some initial success, banks soon dropped the group-lending scheme [because of] embezzlement problems... Some 90 years after these first experiments, BPR started again with group- lending on a larger scale under Bank Indonesia's Linking Banks and Self-Help Groups (PHBK) project but still faced the very same problems ... However, after 5 years of unsatisfying results, this strategy was replaced by a more pragmatic approach of promoting linkages between BPR and groups of micro- entrepreneurs. Three main lessons emerge from PHBK.The first two are that large, com- plex microfinance programs with layers of intermediaries: * Are very expensive. Although the PHBK had been in operation since 1989, the annual cost of the program in 1995/96 was 95 percent of annual loan volume (not including bank costs).Annual interest rates on PHBK loans that Rural Development and Rural Financial Institutions in Indonesia 135 year would have had to be 277 percent to fund fill program costs.The PHBK developed as an unnecessarily complicated and inefficient program that may not have reached many poor people. Governments and donors can use their resources in far more efficient and effective ways to support the growth of microfinance (see chapter 22). * Lead to high interest ratesfor borrowers. The more layers of financial interme- diaries involved, the more expensive the program-and the more expensive the loan interest rate for borrowers. Ravicz (1998, p. 60) estimates the lower- bound average annual interest rate at 107 percent for 1995/96. But on some of the poorer Outer Islands borrowers were paying much hlgher annual rates- up to 450 percent on Sumbawa (Ravicz 1998, p. 54). Donors and the The third lesson from the PHBK is that the simplest model in which banks government provide channel funds to self-help groups works best, although the model needs mod- ification to work cost-effectively. Bank Shinta Daya modified the PHBK ap- massive proach, setting up its own group lending departmnent to identify and train self-help groups and to provide them with commercial financial services. In 1995 the unsustainable credit bank's annual effective interest rate for such loans was 27.4 percent. programs. Butfor Rural Financial Institutions in Indonesia in 2000: What Have We poor borrowers, loans Learned? are often cheaper Some of the lessons on rural finance in Indonesia were known a century ago. Others have been learned only in the past few years as the country, hit by a and easier to access devastating financial, economic, and political crisis, learned the extraordinary stability of commercial rural financial institutions. But as old lessons are rein- at commercial forced and new ones learned, the basic principles endure and the basic prob- lems persist. Of the five lessons below, the first four are drawn from the recurrent institutions themes discussed at the beginning of this chapter; the fifth is drawn from the crisis. Policies and politics Since the late 19th century Indonesia has been developing commnercial insti- tutions that serve as intermediaries, providing savings and credit services to local populations, including large numbers of low-income people. Large-scale sub- sidized rural credit programs were also developed in Indonesia in the early 1970s; politicians and rural elites quickly found ways to obtain and use these rationed, below-market funds.Today these two strategies coexist in the same villages- the former works well while the latter is politically expedient.Thus the BPRs, both licensed and generic, and BRI's unit desas continue their commercial focus, many of them providing microfinance profitably. Most of these institutions are publicly owned. But the central government also provides widespread, poorly designed, and highly subsidized credit through programs such as KUKESRA and Inpres 136 The Microfinance Revolution: Lessons from Indonesia Desa Tertinggal, while Bank Indonesia provides subsidized liquidity credits to village cooperatives (KUDs) and to farmers through a Ministry ofAgriculture program (KUT).These programs did not reach the poor any better in the 1990s than subsidized credit programs did in the 1970s. Meanwhile, donors and the government cooperate in bringing to Indonesia massive, complex, high-cost programs such as the PHBK that are unsuitable, unsustainable, and often result in poor borrowers paying far higher interest rates for loans than are charged in the country's commercial BPRs and BRI's unit desas. Messy (and wasteful) as this coexistence is, it seems unlikely to change any time soon. Between the BPRs and the unit desas, commercially oriented in- stitutions continue to provide savings and credit services to about half the coun- try's households. But other political forces in Indonesia are too powerful to be Microfinance in ignored. The de facto solution has been to support all kinds of microfinance programs, from the sublime to the ridiculous.This is the political solution. But Indonesia nrns from the economic reality has been that sustainable financial institutions-large and small; public and private; commercial banks, licensed BPRs, and generic the sublime to the BPRs-have gained a large and growing share of the rural finance market. ridiculous. But it is Financial intermediation Many BPRs are commercial financial intermediaries, as are BRI's unit desas. the commercial BPR loans are generally financed from savings and equity, and in some cases with commercial bank loans. Unit desa loans are financed from savings from institutions that both poor and nonpoor people.The larger savings accounts and time deposits of the nonpoor raise the average account size, making it possible for the insti- have consistently tutions to accept large numbers of small savings accounts from poor people and still remain profitable. gained market share Long evidence (even back to 16th century Europe) indicates that financial intermediaries are more stable, profitable, and sustainable than credit-focused organizations. And as discussed in this volume and in Steinwand (2001), this finding has been well documented for Indonesia. But the role of savings is not always understood or valued. Large external injections of liquidity, such as KUK funds, discourage voluntary savings mobilization. And there are credit-driven programs that ignore poor people's extensive demand for voluntary savings fa- cilities.As Seibel and Parhusip (1998, p. 21, emphasis added) comment: In sum, given the need of the poor for savings deposit facilities, it is all the more surprising that projects focusing exclusively on the poor as in the various Grameen Bank replications (including Mitra Karya) and small farmer development projects (including P4K [chapter 14]) continue to be credit-driven and offer little in terms of innovative savings products with attractive returns and convenient collection services, with regular compulsory savings as promoted by these projects tending to restrict savings to the required minimum. Unless this is reversed, we may conclude that such projects may ultimately provide more of a disservice Rural Development and Rural Financial Institutions in Indonesia 137 to the poor than a service,promoting indebtedness more than a healthy se!f-financing capacity. As with the debates on subsidies, however, the financially sound approach is the one that gains market share.The future of rural finance (and microfinance generally) is with commercial financial intermediaries (chapter 21).With its long history of sustainable microfinance, Indonesia has all the robust home-grown models it needs to provide large-scale sustainable microfinance to its eco- nomucally active poor. Rural financial institutions: public and private Indonesia has all the Most of Indonesia's commercial financial institutions that serve low-income peo- ple are public enterprises. As noted, this provides a sharp contrast to much in- robust home-grown ternational experience and to the prevailing international view that potentially self-sufficient rural financial institutions should be located in the private sec- models it needs to tor.There is little disagreement about the potential of the private sector for pro- viding commerclal finance to low-income people, even though in many provide large-scale countries private sector interest has been slow in developing. But governments around the world that are saddled with failing state-owned banks and other pub- sustainable licly owned rural financial institutions should consider carefully whether In- donesia's experiences with public commercial financial institutions-the unit microfinance to its desas, the LDKPs, the BKDs, the licensed BPRs-might be adapted to the con- ditions of their countries. Commercial microfinance can succeed in both pub- economically active lic and private institutions.There is no a priori reason to ignore either approach. poor Learning from experience Rural financial institutions in Indonesia have a long history of learning from experience. Colonial adrmnistrators moved from the idea of a cooperative rural banking system to one of profitable local banks with external funding. BRI's unit desas were transformed from loss-making subsidized credit deliv- ery systems to sustainable commercial financial intermediaries. The BKKs of Central Java did not initially mobilize savings effectively because of the large influx of KUK funds that made them overly liquid in the early 1990s. But they learned to invest the KUK funds in expanded loan portfolios; they also learned to mobilize voluntary savings from the public.The BKK systems of CentralJava and South Kalimantan learned how to overcome political pressure to make in- appropriate loans, at least to the extent required to operate sustainably (a les- son not yet learned by most of the LKPs of Dompu District, Nusa Tenggara Barat). Bank Shinta Daya learned how to add group financial services to the savings and credit facilities already provided to individual clients. Bank Shinta Daya also learned to build and operate its own group lending department in order to finance self-help groups profitably (which had not been possible under the high-cost PHBK model that it used at first). Licensed private BPRs are learning how to build BPR networks and links with commercial banks. There are, of course, many lessons still to be learned. But more low-income 138 The Microfinance Revolution: Lessons from Indonesia people are provided with financial intermediation services in Indonesia than in any other country in the world. Stability The recent Indonesian crisis had remarkably little effect on many BPRs, BRI umit desas, and Bank Dagang Ball. Low-income people (whose enterprises are mostly m the domestic economy) continued to save m the institutions they trust- ed. Savers left faltering banks and moved to the financial institutions they be- lieved were sound (public and private). Borrowers repaid their loans because, especially in time of crisis, they valued greatly their options to reborrow. They also repaid loans because they knew that their financial institutions were not lacking liquidity; thus the clients' ability to obtain new loans depended on their Commercial repayment performance, not on whether the financial institution would have sufficient loanable funds.As discussed in chapter 15, established conmmercial mi- microfinance can crofinance institutions stayed profitable throughout a crisis so severe that the country's financial system collapsed. provide wmide This is perhaps the most important lesson of all. Commercial microfinance (implemented in different kinds of institutions) can provide not only large-scale outreach and outreach and institutional profitability, but long-term stability under seriously adverse conditions. institutional profitability. It can Notes also provide stability 1.This area is known, in this context, as the Celebesian Transition; Celebes is the former name of Sulawesi. seriously 2. Swidden agriculture refers to agriculture that is characterized by a lengthy cycle under of field rotation and by burning as a method of land cultivation. The field is burned and the crops are cultivated in the ashes; after a period of time the field is abandoned adverse conditions for a number of years, while other fields are cultivated in rotation. 3. In 1996 Indonesia had more than 2 million fishing households. 4.A quintal is a unit of mass equal to 100 kilograms. 5. Geertz's (1963) concept of agricultural involution assumes that "poverty shar- ing" in the Javanese rural economy is a reaction to population pressure on the land. Geertz's view of the income-sharing activities of what he assumes to be relatively ho- mogeneous rural communities on Java has been criticized for its lack of fit with the realities of rural Javanese income distribution and social stratification (Alexander and Alexander 1982). See White (1983) for an extensive survey of the literature criticiz- ing Geertz's views. See Geertz (1984) for his reply to these criticisms, and Koent- Jaramngrat (1985) for extensive commentary on the issues. 6. A considerable body of literature attests to the agricultural improvements, in- creased employment opportunuties available to rural households, and generally improving living standards in rural areas of Indonesia that began in the nud-1970s. See Timmer (1975); Booth and Sundrum (1981); Bose (1981); Collier, Soentoro,Wiradi, and oth- ers (1982); DPIS (1983); Kasryno (1983);Wlrach and Manning (1984); Booth (1985, 1988); Glassburner (1985); Hugo (1985); Keyfitz (1985); Mazumdar and Sawit (1986); Rietveld (1986);Manmng (1987,1988);Jayasuriya and Manning (1988);and Hetler (1989). 7. Colt is the brand name of Mitsubishi's mnumvan, which is very popular in rural areas. Rural Development and Rural Financial Institutions in Indonesia 139 8.The studies of these two villages were carried out at different periods by au- thors who consulted for the Harvard Institute for International Development (HIID) and the Center for Policy and Implementation Studies (CPIS) during the 1980s, and both studies are based on excellent data.The Fox study is of Mojosari vlllage, a pseu- donym for village G, discussed in chapter 11. 9. See Hill (1989) for discussion of regional and local disparities in income dis- tribution and employment opportumties. 10. Provinces with more than than 15 percent of the rural population below the poverty lne were all located outside Java, Bali, Sumatra, and Sulawesi. 11. Much of the discussion in this chapter of the early Indonesian people's banks and their European antecedents, as well as the later development of the BPRs, is drawn from Steinwand (2001). Important sources for Indonesia's rural financial institutions outside BRI and the BKDs (for which references are provided in later chapters) are Fruin (1994 [1935], 1999 [1933]); Gonzalez-Vega (1982b); Gadway, Gadway, and Sardi (1991);Patten and Rosengard (1991);Bouman and Moll (1992); Gonzalez-Vega and Chaves (1993); Martokoesoemo (1993); Pearson and Garland (1993); Hanna (1994); Mosley (1996); Riedinger (1994); Chaves and Gonzalez-Vega (1996); Holloh (1998); Meyer (1998); Lapenu (1998); McGuire, Conroy, and Thapa (1998); Ravicz (1998); Seibel and Parhusip (1998);Timberg (2000); and Parhusip and Seibel (2000). 12.The Irish loan funds gradually declined after 1840 and ceased operation in about 1940.There is no clear evidence that the Dutch knew about the Irish loan funds when they introduced rural financial institutions to Indonesia, but as Steinwand (2001, p. 53) points out, these funds have a striking similarity to Indonesia's Lembaga Dana dan Kredit Pedesaan (LDKPs, or rural fund and credit institutions, discussed later in this chapter). 13.A third group of cooperatives was begun in 1883 by Wilhelm Haas, in part to get around the problems between the two other groups. The Haas cooperatives be- came the largest group in Germany in the early 19th century, but the Raiffelsen model of cooperative banks became the most popular in Europe (see Steinwand 2001, p.56). 14. The name Rabobank comes from Raiffeisen, as these banks were known in the Protestant north of the Netherlands, and from Boerenleenbanken, as they were called in the Catholic south. 15. During the world depression, however, nonperforming loans at the district banks rose rapidly, peaking at 56 percent in 1933 and gradually declning to 14 percent in 1937. 16. There were also bank desas on the Outer Islands, especially popular in West Sumatra, but most had to be liquidated because of high defaults (related, in many cases, to managerial and supervision problems). 17. See also DPIS (1983); Sutoro and Haryanto (1990); and the reports from the 1970s and the 1980s of the Agro-Economnc Survey,Bogor. For views about the survey, see Sayo- gyo (1982 [1973]), Collier, Soentoro,Wiradi, and others (1982); and Strout (1985). 18.As noted by ProfessorAliWardhana in the introduction to this book, these pro- grams had high arrears and resulted in high losses to the implementing bank and the government. 19. Based on the findings of DPIS/CPIS/BRI research teams, 1979-90, this rep- resents the normal range of interest rates charged to low-income borrowers by informal commercial lenders in Indonesian villages, but lower and higher rates were also re- ported. See also Alexander (1986); Mai and Buchholt (1987); Sutoro and Haryanto (1990); and Germndis, Kessler, and Meghir (1991). 20. This is not always the case, however. If the loans are tied to interlinked trans- actions so that the creditor/commodity wholesaler, for example, buys from the bor- 140 The Microfinance Revolution: Lessons from Indonesia rower/producer at below-market prices, then small loans can be part of profitable trans- actions for the lender. 21. See McGuire, Conroy, and Thapa (1998, pp. 153-54) for discussion. Under the Microcredit Project, partly funded with a loan from the Asian Development Bank, Bank Indonesia provides loanable funds to the Bank Pembangunan Daerah (Regional Development Banks, or BPDs) and the BPRs. The BPRs lend directly to microen- trepreneurs, while the BPDs lend to small financial institutions. Nongovernmental organizations (NGOs) help organize self-help groups of borrowers. Funds are lent to borrowers at a 2-3 percent flat rate per month.The Linking Banks with Self Help Groups Program (Program Hubungan Bank dan KSM, or PHBK) is discussed later in this chapter. 22. Because of rounding in Steinwand's calculations, the percentages of comrner- cial banks' assets sums to 101 percent. 23.Table 9.1 (for 1985) compares rural loans and total loans in the financial sys- tem; table 9.2 (for 1995) shows loans in microfinance institutions as a share of loans in the total financial system. However, there is substantial overlap between the rural and microfinance institutions shown in these tables. 24. Bank Indonesia lists 1,807 Lembaga Dana dan Kredit Pedesaan (LDKPs), rural fund and credit institutions owned by local governments, for 1998. But this iS likely an underestimate because Bank Indonesia has updated figures only from some provinces. Steinwand (2001, p.230) estimates 2,317 LDKPs m 1998, includmg the LDKPs that had become licensed BPRs. He points out that while Bank Indonesia statistics show 264 LPDs in Ball in 1998, there were actually 910 (p. 172). 25.These data are from September 1998, when Indonesia was in deep crisis.The exchange rate that month was 10,700 rupiah to the U.S. dollar (compared with 2,450 in June 1997,just before the crisis). 26. Steinwand (2001, p.241) raises the valid point that some private hcensed BPRs have time deposits from only a few large depositors who are often connected to the owners, and that this can be dangerous 27. Provmcial Development Banks are also known as Regional Development Banks. 28. The bank transferred the KUK funds to the BPD, which provided each BKK/CJ with a deposit at the BPD. Since the interest rate for the deposit was set to be the same as the rate that the BKKs/CJ paid for the KUK loan, the interest rates offset each other. 29. Savings include both compulsory and voluntary savings, but Steinwand (2001) suggests that mandatory savings may now be a negligible share of total savings. 30. Profitable BKKs/CJ allocate shares of their profits to stakeholders. In the early stages of their development, the BKKs/CJ retained 70 percent of their profits, but in the 1990s this was reduced to 50 percent. Thirty percent of the profits were paid in dividends to the provincial government, five percent each went to two local funds (a development fund and a social fund), and ten percent was used for performance-based incentives for BKK/CJ management and staff. (Steinwand 2001, pp. 133-34).The dis- tribution of profits has changed somewhat over time (and also shows some variation among BKKs/CJ). 31. The capital endowment grant was actually a low-interest loan. But because BKKs/SK are required to pay only the interest on the loan and because the interest is used for partial coverage of the costs of supervising units and training staff, the loan essentially functions as a capital grant (Ravicz 1998, p. 32). 32. It is unclear whether the loans to self-help groups are included in this figure. 33.The South Kahmantan BPD reported that some BKKs/SK serve vlllages up to 70 kilometers away. Ravicz (l 998, pp. 39-40) mterviewed a BKK/SK manager who reported serving clients 25 kilometers away-where the trip by motorcycle took two Rural Development and Rural Financial Institutions in Indonesia 141 hours each way because of difficult road conditions. At that BKK only 15 of the sub- district's 27 villages were provided services because of the area's limited infrastructure. 34. NusaTengarra Barat was surpassed in percentage of population below the pover- ty line in 1996 only by its neighbors in eastern Indonesia,TimorTimur (no longer part of Indonesia as of 1999) and NusaTenggaraTimur, and by West Kahmantan. In both CentralJava and South Kalimantan 14 percent of the population was below the poverty line in 1996, higher than the national average of 11 percent (Government of Indonesia 1996, p. 572). 142 The Microfinance Revolution: Lessons from Indonesia Where the Microfinance Revolution Began: Bank Dagang Bali Bank Dagang Bali (BDB) was not the first financial institu- tion to provide microcredit profitably, nor was it the first in- stitution to collect savings from the poor. In an important sense, however, BDB was where the microfinance revolu- tion began. Today it is the longest-serving formal sector fi- nancial institution providing commercial microfinance (both savings services and loans) on a substantial scale in a de- veloping country, having done so continuously and profitably since 1970-and without ever having received a subsidy.1 BDB was founded in Bali, Indonesia, by I Gusti Made Oka and Sri Adnyani Oka, a husband and wife who started out poor, built up a little capital by working in the informal sector, became informal commercial moneylenders, opened a small secondary bank, and then-with full knowledge of 144 the microfinance market-opened BDB, a private bank specializ- ing in delivering financial services profitably to the low-income public. It is no accident that the first bank to emphasize microfinance was started by former moneylenders.The Okas knew the market from the inside, and they were experts in the methods and oper- ations of moneylenders. Further, they understood the aspirations, fears, and financial strategies of microfinance clients, how such clients could benefit from bank services, and how a bank could serve them profitably. In hindsight, it is hard to imagine how the Bank Dagang first bank with a microfinance focus could have been founded in any other way. Bali was not the This chapter first explores the early history of the Okas' inter- est and activities in coommercial microfinance, then discusses the his- f fn tory of BDB. The bank's performance from 1970-96 is reviewed, institution to and the focus then shifts to BDB's clients. BDB's performance from 1997-2000 is discussed in chapter 15. provide microcredit In 1968 the Okas opened a tiny secondary market bank with $300 in capital. The bank was located in a 2- by 4-meter room in Den- profitably} nor was it pasar, the capital of the province of Bali.The little bank quickly be- thefirst to collect came profitable because the Okas knew that there would be large demand for small loans, that they could finance these loans with sav- savingsfrom the ings, and that they could undercut moneylenders by a wide mar- gin-benefiting both their customers and the bank. By 1970 the small poor But it was market bank had made about $40,000 in profit, and in September where the of that year the Okas decided to open BDB, the second licensed pri- vate bank in Bali.The bank provided small loans and mobilized sav- microfinance ings, and it grew rapidly. By the end of 1970 BDB had assets of about $156,000; by the end of 1971 it had assets of about $573,500 and revolution began had made a profit of about $8,000 that year. Over time BDB came to emphasize savings over lending, but it has always offered both. The bank is best known for its knowledge of microfinance clients and for its savings services-which are state of the art in microfinance. BDB is now a full-service bank provid- ing finance to commercial and corporate clients as well as to mi- crofinance customers. By the end of 1996 BDB had $113 million in deposits, $94 mil- lion in outstanding loans, and $136 million in assets. The bank has Where the Microfinance Revolution Began Bank Dagang Bali 145 been profitable every year since it began. In 1990-96 annual pretax profits ranged from about $1.0 million to $1.6 million.As discussed in chapter 15, the bank continued to serve low-income customers and to operate profitably during 1997-2000 despite the crisis. Bank Rakyat Indonesia (BRI) and every other bank in In- donesia that provides commercial microfinance is in some way in- debted to BDB, and the bank's influence has spread to other countries as well. BDB is known for teaching the poor not to be afraid of banks. As noted in the preface, one of its clients told me, "BDB taught us something important that we never knew before. BDB taught us that the bank is not a king, the bank is a servant." 146 The Microfinance Revolution: Lessons from Indonesia BDB, now a full-service bank providing finance to coninercial and corporate clients as well as to microfinance customers, began with an idea-that the de- mand for financial services among the economically active poor could be met by institutions providing commercial microfinance profitably, and that these ser- vices could promote social and economic development. This was a new idea in 1970, when BDB opened its doors.As might be expected from a bank found- ed by former moneylenders, BDB, without any subsidy, has been profitable every year since it began. The story of BDB is crucial for understanding the history, psychology, and financial principles of institutional commercial microfinance. But BDB cannot be understood without a knowledge of I Gusti Made Oka and Sri Adnyani Oka, its founders. Mr. Oka is frequently invited to lecture at finance courses where As might be he speaks openly about BDB's products, strategies, and operations, about how to serve microfinance clients profitably, and about the lessons BDB has learned. expectedfrom a At one such lecture at Indonesia's Ministry of Finance, he was asked by a manager of another bank, "Why do you tell us your bank secrets? Do you not bank founded know that we will copy them?" Mr. Oka replied, "That is the reason I tell you. BDB has been successful. Therefore, we must give back to the wider commu- by former nity, and what we can give best is our knowledge.There are many poor people whose banking needs are still unmet. If you can serve them, it will help them moneylenders, and it will help the country. If you learn from BDB's experience, we are happy." BDB, without any Bank Dagang Bali: Formative Elements subsidy, has been To understand today's BDB, it is important to first understand the history of profitable every year the Oka family, the Okas' experience as informal moneylenders, and the for- mation of Bank Pasar Umum, the precursor to BDB. since it began Oka family history I Gusti Made Oka was born in a village near Gianjar, Bali, in 1932. His father died when he was three years old; three years later his mother remarried and moved away. Oka was then adopted as a grandson of the elderly woman for whom his father had worked as a servant. He grew up in her home, working in the house and on her fields. His adopted grandmother sometimes allowed him to go to school, but she was unable to pay his school fees. However, the teacher permitted him to follow the lessons informally. Although he could attend school only sporadically, the young Oka did well and was later admitted without fee. When Oka was eight years old, he went to the local Hindu priest and vol- unteered to clean the temple in the mornings before school. The priest said, "That is a good idea. The Gods will bless you." Oka replied, "That would be wonderful, but could I also have some coconuts and flowers from the temple trees?"The priest agreed, and Oka sold these for pocket money and school sup- plies. Oka graduated from primary school, but he could not afford further study. Instead he learned to sew and became a tailor in his vlllage. Where the Microfinance Revolution Began Bank Dagang Bali 147 After a few years Oka moved to Gianjar, the nearby market town, and became an apprentice to a tailor there. In 1951 he moved to Denpasar, the capital of Bali, where he began to work in a tailoring shop.A year after arriving in Denpasar, Oka rented a room and opened his own tailoring shop, which doubled as his livmg quar- ters. Because he did not have a sewing machine, he sewed by hand. His skill increased, and he received an order from the Secretary to the governor of Bali.That, in turn, led to a larger order to make the drivers' uniforms for the Balinese government. He was then able, in 1952, to approach the BRI branch in Denpasar to ask for an $11 loan to buy a sewing machine.2 The sewmg machine cost $30; he had saved $19 and BRI lent him the $11 he requested.With the machine, Oka's profits increased and he was able to repay the loan and to employ a few assistants. Oka deposited sav- Mrs. Oka began to mgs regularly in BRI-which, since he had a BRI loan, he considered his bank. Soon after moving to Denpasar, Oka had begun attending night school, where lend their savings to he studied accounting and Englhsh, and where he met his future wife. Sri Ad- nyani Oka was born in 1936 in a village in Bah.After graduating from the neigh- friends, and Mr. borhood primary school, she began volunteer work at the local hospital, delivering medicine to patients. She attended a school that prepared students Oka kept the for secondary studies, and she wanted to continue her education. However, her parents did not believe in further education for girls and would not give their accounts. Soon there permission for her to attend secondary school. In 1951 she hitched a ride at night on a truck headed for Denpasar-not a common action for a young woman was more demand in provincial Bali in 1951. She stayed with a childhood friend who was work- ing in Denpasar, and she found a job as a typist in the Health Department. She for loans than the then joined night school, where she met Mr. Oka; they were married in 1955. They began raising a family and eventually had 10 children, of whom 8 sur- Okas' savings could vived. By 2000 the Okas had more than 20 grandchildren. meet The Okas as informal commercial lenders Shortly after their marriage in 1955, Mrs. Oka perceived that her husband was deeply concerned about their financial security. As she says, "he was worried about our poverty, and I wanted to do what I could to make him happier." She asked him how much annual interest they were receiving on their BRI sav- ings account-5 percent. She said, "I can earn more than that by lending our money to friends. People want short-term loans, and I can lend out the same money many times during one year." Mrs. Oka began to lend their savings to friends, and Mr. Oka kept the accounts. Soon there was more demand for loans than the Okas' savings could meet. By then, however, the Okas were well known and trusted in the community. Thus when they had a request for a loan from someone they knew and trusted, they were able to serve as an intermediary between the potential borrower and one of a num- ber of people who had excess liquidity and who trusted the Okas'judgment. To protect the lender, Mrs. Oka required the borrower to provide gold as col- lateral; she held the collateral in case the borrower defaulted.To comply with In- donesian law, Mrs. Oka purchased the gold.The sales slip recorded the weight and karat of the gold, and stated: "Thls gold was sold by Person A to S.A. Oka for X 148 The Microfinance Revolution: Lessons from Indonesia amount. Person A can purchase this gold within one month fiom S.A. Oka forY amount [X plus the interest charges]. After one month, S.A. Oka has the right to sell the gold m the market."There was large demand for this service, and the prof- its were split evenly between the Okas and the person who provided the capital. Mrs. Oka also accepted deposits on whlch she paid interest, and which she lent out. In 1956 Mr. Oka borrowed $222 from Bank Perniagaan Umum, the first private bank m Bali.At that time the bank required that borrowers become share- holders. Accordingly, Mr. Oka had to pay 10 percent of the loan to become a shareholder of the bank. His first shareholders' meeting was an experience that had a crucial effect on his thinking and on his life. As he put it, "My wife and I already knew that when a person we trusted needed to borrow money, we could go to someone who had money and arrange the transaction.At the bank "I knew that we shareholders' meeting, I was surprised to learn that is exactly what banks do. I knew then that we could run a bank, that it could be profitable, and that it would could run a bank, help many people who were afraid of banks. From that time on, I was deter- mined to obtain a bank license." that it could be The Okas maintained very low operating costs in their informal financial ac- tivities. Mrs. Oka rode a bicycle to visit those who wanted to borrow and those profitable, and that who had excess funds to lend, and Mr. Oka did the accounting. She also worked as a typist, while he continued his tailoring business and graduated from night school. it would help many Bank Pasar Umum people who were The Okas did well in the informal financial market, and in 1968 they opened a small secondary market bank (bank pasar), Bank Pasar Umum (BPU).3 BPU, afraid of banks. I with $300 in capital, was located in a tiny room in a Denpasar market.The Okas hired three employees, and the bank began operations. was determined to Inflation was high at that time, and interest on deposits was normally stat- ed as a monthly rate. Government banks offered 6 percent a month interest on obtain a bank savings, while private banks offered 7 percent a month.The Okas took a full- page advertisement in a local newspaper statmg that "If you deposit Rp 100,000 license" now, you will have Rp 204,000 next year.Your profit will be more than your principal. However, if you want interest every month, BPU will pay you 8 per- cent a month. Do not move from your house.Just call us at this telephone num- ber.We will come to your house." The ad brought BPU many customers. BPU provided loans with one-month terms, with interest of 8 percent a month and a 3 percent fee for each loan.The Okas collected loan repayments in daily installments and immediately relent the capital. Because the BPU office was tiny and probably would not have inspired confidence on the part of their clients, the Okas and their three employees nor- mally went to the customers, rather than having the customers come to the bank. BPU quickly became profitable.The bank succeeded because the Okas com- bined intimate knowledge of the local financial market, well-honed entrepre- neurial abilities, and a sense of kinship with their low-income cients.The Okas knew how few options there were for microenterprises and for poor people who wanted loans and savings accounts. Where the Microfinance Revolution Began Bank Dagang Bali 149 By 1970, two years later, BPU had made $39,700 in profits, and the Okas decided to open a private bank as well.They borrowed $13,200 from Bank Pem- bangunan Daerah, which is owned by the provincial government of Bali, and with initial capital of nearly $53,000 they opened BDB.The Okas' experience in moneylending and in building a local market bank had given them exten- sive knowledge of the microfinance market-which they would now serve through a licensed, full-service bank. Bank Dagang Bali: Development and Performance, 1970-96 Creativefinancing, BDB opened in September 1970 with 30 staff members in a rented office in the center of Denpasar. Aware that they knew a lot about the local informal professional and financial market but little about bank management, the Okas hired three re- tired BRI employees. One was appointed chairman, one the head of the cred- transparent it section, and one the head of operations. Mr. Oka became BDB's president-director. Mrs. Oka continued to work with clients, emphasizing sav- accounting, and a ings moblization and customer relations. BDB grew rapidly. In 1972 the bank purchased an office building in down- knowledge of the town Denpasar; it also obtained a technical consultant from the U.S. Interna- tional Executive Corps Service who helped the bank improve its management. market have In 1974 BDB was selected by the Indonesian government to be the first pri- vate bank to channel credit to small borrowers under the supervision of Bank consistently marked Dagang Negara, a state-owned bank. Subsequently BDB continued acting as a channeling agent for government credit to small borrowers, but was permut- BDB's development ted to do so without state bank supervision. Creative financing, professional and transparent accounting, and a knowl- edge of the market have consistently marked BDB's development. For exam- ple, until PAKTO 88 deregulated the opening of bank branches, government regulation prohibited BDB from opening branches. But knowing that there was considerable demand for BDB services in other parts of Ball, the Okas bought a large used car from Bank Indonesia, the central bank.This car became, de facto, a BDB mobile bank. Announcements were made that at a particular time on a particular day, the car would be at a certain location. Clients went to that lo- cation and transactions were carried out there.This approach worked well dur- ing the 1980s because there was large demand for small loans and savings services, and because BDB was known and trusted throughout Bali. BDB carried out profitable financial intermediation from its mobile bank until 1988. Immediately after PAKTO 88, BDB opened branches that rapidly became profitable because they already had a large client base in the mobile bank cus- tomers. In 1987 BDB was the largest corporate taxpayer in Bali, and in 1988 it was the second largest corporate taxpayer in the Eastern Indonesia Region. Since then BDB has remained one of the top taxpayers in the province. In 1990 the bank opened branches off Bali in Indonesia's two largest cities,Jakarta and Surabaya. 150 The Microfinance Revolution: Lessons from Indonesia Bank Dagang Bali profit Table 10.1 and loss statement, 1970-96 (millions of rupiah) Indicator 1970 1975 1980 1985 1990 1995 1996 Income 4 136 826 5,786 36,702 44,173 52,777 Operating income 4 136 826 5,731 31,214 43,497 52,676 Interest, commissions, and fees 3 132 801 5,636 30,336 41,437 50,571 Other 1 4 25 95 878 2,060 2,105 Nonoperating incomea 0 0 0 55 5,488 676 101 Expenses 4 130 774 4,887 33,629 41,411 49,439 Operating expenses 4 130 774 4,832 28,100 40,907 49,429 Interest 2 92 525 3,232 19,346 30,910 37,459 Overhead 2 38 249 1,600 8,754 9,997 11,970 Salaries - 11 93 449 2,086 3,855 4,148 Other overhead 2 27 156 1,151 6,668 6,142 7,822 Nonoperating expensesa - - - 55 5,529 504 10 Net profits 0 6 52 628 2,003 1,937 2,345 Pretax profits 0 6 52 900 3,073 2,762 3,338 Taxes 0 0 0 272 1,070 825 993 - Not available Note The year-end rupiah exchange rate for one U S dollar was 378 in 1970, 415 in 1975, 627 in 1980, 1,125 in 1985,1,901 in 1990, 2,308 in 1995, and 2,383 in 1996 See table 2 1 for more complete data on exchange rates, inflation rates, and the consumer price index for 1970-99 a Refers to income and expenses from nonbanking activities Isuch as the rental and sale of buildings) Source Bank Dagang Bali data As noted, BDB has been profitable every year since it opened (tables 10.1 and 10.2). Pretax profits were about $8,000 in 1971, about $83,000 in 1980, and about $1.6 million in 1990.4 Profits fell in 1991 ($1.2 million) and 1992 ($1.06 million), primarily because the central government imposed a tight money pol- icy that affected all banks in Indonesia.When the policy was changed, BDB's pre- tax profits reverted to $1.2 million a year in 1993-95 and to $1.4 million in 1996. BDB's assets grew substantially as well, from about $156,000 in 1970 to $10.7 million in 1980 to $135.5 million in 1996, excluding assets held in the names of the owners. (The bank remained profitable during 1997-2000; see chapter 15.) The amounts of BDB's outstanding loans and savings and the number of loans and savings accounts grew substantially (tables 10.3 and 10.4). By 1996 BDB had $113 million in savings and $94 million in outstanding loans. That same year, though, there were 23 times as many savings accounts (363,859) as loans (15,645), reflecting both state-of-the-art local-savings mobilization meth- ods and a relatively conservative loan policy.5 Repayment was consistently high during 1970-96, and the occasional writeoffs were small. In 1996, noncurrent loans accounted for 1.2 percent of the portfolio; doubtful loans, 0.7 percent; and bad loans, 0.2 percent. Where the Microfinance Revolution Began Bank Dagang Baii 151 Table 10.2 Bank Dagang Bali balance sheet, 1970-96 (millions of rupiah) Indicator 1970 1975 1980 1985 1990 1995 1996 Assets 59 819 6,681 31,330 168,385 268,363 323,005 Current assets 46 767 6,525 29,689 158,941 231,305 267,969 Cash 6 57 506 627 2,693 4,765 3,960 Bank Indonesia 5 80 988 1,104 2,276 2,841 9,113 Other banks 5 36 4 9,446 39,174 42,381 33,831 Outstanding loans 30 594 5,027 18,512 114,798 183,285 223,393 Reserve for bad debt 0 0 0 0 0 -1,967 -2,328 Fixed assetsa 11 50 106 902 3,892 9,811 10,086 Other assetsb 2 2 50 739 5,552 27,247 44,950 Liabilities 39 752 6,345 29,318 153,060 247,005 298,106 Demand deposits 17 106 393 2,545 14,714 16,536 14,937 Savings deposits 1 293 3,912 13,409 41,704 52,685 59,084 Time deposits 21 257 1,258 10,977 67,952 155,772 195,762 Bank Indonesiac 0 76 639 603 19,072 1,985 1,614 Other loans 0 0 0 0 6,990 14,794 20,325 Other liabilities 0 20 143 1784 2628 5233 6384 Equity 20 67 336 2,012 15,325 21,358 24,899 Capital (incl. retained eamings) 20 61 284 1,384 13,322 19,421 22,554 Net profits 0 6 52 628 2,003 1,937 2,345 Total liabilities and equity 59 819 6,681 31,330 168,385 268,363 323,005 Note See table 10 1 for exchange rates a Includes land, buildings, vehicles, and office furniture but excludes fixed assets held in the names of the bank's owners b Includes marketable securities and investments c Long-term liabilities Source Bank Dagang Bali data Table 10.3 Bank Dagang Bali outstanding loans, 1970-96 Indicator 1970 1975 1980 1985 1990 1995 1996 Value (millions of U.S. dollars) 0 08 1.43 8 02 16 46 60 37 79 40 93 80 Number 29 597 4,639 5,344 8,285 12,814 15,645 Note Excludes U S. dollar loans, which were started in 1992 Source Bank Dagang Bali data In 1996, 72 percent of BDB's savings were in time deposits in just 2 percent of the bank's accounts. Nearly 98 percent of the accounts were in passbook savings, which accounted for 23 percent of the value of deposits.The number of giro checking ac- 152 The Microfinance Revolution: Lessons from Indonesia Table IOA| Bank Dagang Bali savings, 1970-96 Indicator 1970 1975 1980 1985 1990 1995 1996 Time deposits Amount (millions of U.S dollars) 006 0.62 201 976 3574 6749 82.14 Share of total (percent) 54 39 23 41 53 70 72 Number of accounts 25 435 492 2,494 10,624 7,593 8,015 Share of total (percent) 14 4 0 0 4 2 2 Passbook savings Amount (millions of U S dollars) 0 0.71 6 24 11 92 21.93 22 83 24 79 Share of total (percent) 3 45 70 50 34 23 23 Number of accounts 62 10,515 55,341 152,837 277,776 345,240 354,888 Share of total (percent) 35 94 99 98 96 98 98 Giro accounts Amount (millions of U S dollars) 0 05 0 26 0 63 2 26 7 74 7 16 6 27 Share of total (percent) 44 16 7 9 13 7 5 Number of accounts 90 211 311 683 1,225 1,005 956 Share of total (percent) 51 2 1 0 0 0 0 Total Amount (millions of U S dollars) 011 1 58 8 88 23 94 65 41 97 48 113 2 Number of accounts 177 11,161 56,144 156,014 289,625 353,838 363,859 Source Bank Dagang Bali data counts (current accounts with checking facilities) was less than 1 percent of accounts and only 5 percent of the value of BDB's savings and deposits (see table 10.4). Between 1975 and 1985 BDB's savings were worth more than its time deposits, but over the next decade, time deposits pulled far ahead of savings. Between 1975 and 1996, however, the number of time deposit accounts never surpassed 4 percent of the total number of savings and deposit accounts (see table 10.4).The rise in the value of time deposits after 1985 reflects primarily the product choices of a relatively few large savers, many of whom were previously small savers but remained loyal to BDB. In 1996, 18 percent of the number of BDB's outstanding loans were for amounts up to $420, and 61 percent were for amounts up to $2,098 (table 10.5). However, 10 percent of borrowers accounted for 66 percent of the value of outstanding loans. For time deposits, 44 percent of the number of accounts and 81 percent of the value of the deposits were in accounts above $2,098. For pass- Where the Microfinance Revolution Began Bank Dagang Bali 153 Table 10.5 Size distribution of Bank Dagang Bali loans and savings, 1996 Amount Share of Number of Share of (millions of total loans or savings total Indicator U.S. dollars) (percent) accounts (percent) Loansa 93.8 100 15,645 100 "$125 0.1 0 156 1 $126-420 09 1 2,660 17 $421-2,098 6 5 7 6,727 43 $2,099-10,491 24 4 26 4,537 29 BDB has been >$10,491 61 9 66 1,565 10 Time deposits 82 1 100 8,015 100 profitable every year "$2,098 156 19 4,448 56 >$2,098 66.5 81 3,527 44 since it opened in Passbook savings accounts 24 8 100 354,888 100 "$420 8 2 33 333,595 94 September 1970, >$420 166 67 21,293 6 Note The size distribution for giro accounts is not available W&EithoutL ever a Excludes U S dollar loans Source Bank Dagang Bali data receiving a subsidy book savings, in contrast, 94 percent of the number of accounts and 33 per- cent of the value of the deposits were in accounts below $420. Unlike BRI, BDB does not separate its niicrobanking activities from its ser- vices to larger commercial and corporate clhents.Thus its average loan balance- $5,990 in 1996-is not a useful measure of its rmicrobanking activities. Still, because most of the bank's microsavings activities are in passbook savings ac- counts, some sense of the those activities can be gained from the average bal- ance for each savings instrument. In 1996 the average time deposit account was $10,248 and the average giro account was $6,559. In contrast, the average pass- book savings account was $70. Thus BDB serves a large number of small clients. But most of the value of its loans and savings comes from its larger clients, many of them long-term BDB customers. In essence, the larger clients make possible financially viable service to large numbers of low-income clients. Bank competition in Indonesia increased rapidly after the PAKTO 88 deregulation. BDB's competition grew particularly fast, especially in Bali, be- cause many banks there copied and adapted the Okas' ideas and the bank's prac- tices.The results for BDB have been mixed. In 1996 BDB was overly liquid (see tables 10.2-10.4), which was also true of many other Indonesian banks at that time. Moreover, many BDB clients started to use several banks. Typically they do not leave BDB; they add other banks. But they continue to recommend BDB to their family and friends, citing its record of friendliness and service to all clients 154 The Microfinance Revolution: Lessons from Indonesia regardless of the size of their accounts. So, despite increased competition, both clients and staff have continued to be exceptnonally loyal to BDB. Although BDB has come to emphasize savings over lending, it has always offered both.The bank is best known for its knowledge of microfinance clients and for its savings services. BDB benefited not only from Indonesia's growth and development during 1970-96, but also from Bali's thriving economy-to which the bank contributed. The province experienced high export growth, substantial tourism, and rapid rural employment and income growth from agriculture, fishing, and garment and handicraft production. But BDB's growth has also derived from an approach that allowed a family-owned institution with a thin management structure to serve as many low-income clients as possible while also serving some larger customers and remaining profitable. In 1994 Mrs. Oka In an effort to give back to the community, BDB provides umversity schol- arships each year to poor Balinese students. More than a thousand students have received Indonesia 's received these scholarships. In addition, the bank contributes rice and other food to orphanages. Finally, BDB provides funds for the education of needy prima- Kartini award, a ry school students in the village where Mr. Oka was born. national honorgiven Customer Relations, Organization, Products, and Services to womenfor BDB has been built on a personal approach to customers, and the Okas know lifetime achievement well their customers and their markets. Mr. Oka's example and his teachings have spread widely. Similarly, Mrs. Oka's knowledge of microfinance clients- and BDB's example in treating low-income clients with the same respect and attention accorded its wealthier clients-have become renowned. In 1994 Mrs. Oka received Indonesia's Kartini award, a national honor given to women for lifetime achievement. Sustainable financial institutions that provide services to microenterprises and other low-income people frequently develop and articulate an institutional philosophy. There is considerable overlap in the cultures of such institutions; these include components of trust, service, high-quality financial management, accountability, loyalty, institutional reputation, and contribution to social and economic development. Nevertheless, each institutional philosophy has its own emphasis: BDB's emphasizes balance. BDB believes that to achieve harmony, loyalty, and profitability, five com- ponents must be balanced: investors, staff, customers, the government, and the community. As Mr. Oka put it: Business means serving.We must serve the investor and respect the capital that is invested. We must also serve the employees; we do this in a number of ways, for example by establishing pension funds and retirement savings accounts. Customers must be given the best service possible.We serve the government by paying taxes promptly; BDB is one of the largest taxpayers in Where the Microfinance Revolution Began Bank Dagang Bali 155 this region of Indonesia. BDB contributes to the community by providing scholarships to poor students and food to orphanages. BDB culture also incorporates the idea that if the bank encourages savings and provides loans for the important things in life-a house, education, health, and retirement-then, as the Okas say, "All members of the family will be happy with BDB and will stay with us." BDB expects to serve its clients not only for their lifetimes, but for generations. Its instruments and services have been de- veloped, and its staff trained, accordingly. BDB expects to Organization and management BDB's head office, constructed in 1993, is in Denpasar. In addition, BDB has serve its clients not 8 branches, 18 sub-branches, and 4 deposit collection points.All but two of the branches are in Bali; as noted, those in Jakarta and Surabaya began in 1990. In onlyfor their that year also BDB was approved as a private national foreign exchange bank, and it began ajoint venture with the Internationale Nederlanden Bank (ING) lifetimes, butfor in Jakarta (now ING Barings).6 BDB's board of directors has three members, mcluding Mr. Oka, the presi- generations. Its dent-director.The bank has six divisions: credit and marketing, operations and plan- mng, treasury, human resources, general and legal, and internal audit. At the end instruments and of 1996 BDB had 717 employees, about evenly split between women and men. As noted, the bank's management structure is thin. BDB compensates for services have been this by concentrating on savings mobilization-an activity at which it is ex- pert and that it can manage relatively easily. Loans are made to the extent that developed management resources enable high portfolio quality. This approach permits the bank to serve both low-income people and accordingly larger clients, but it limits potential profitability.Within its organizational con- straints, however, BDB has done well: it reaches many microfinance clients, it has always been profitable, and it has retained a sound rating throughout. When the Okas retire, BDB could go public, with the family retaining a ma- jority of the shares. Considerable attention is paid to BDB staff, and staff retention is high. BDB encourages long-term staff loyalty and emphasizes membership in the BDB "fam- ily." Staff are taught about local markets and about serving clients. All staff- even the cleaning staff-are encouraged to visit their extended families, neighbors, and friends to inform them about BDB services, and in this way to locate new customers. Staff receive a bonus for every client introduced to BDB; for borrowers, the bonus is paid after the loan is repaid. BDB staff are treated with respect and thoughtfulness by their employers, and are taught to treat their customers, both rich and poor, in the same way. Many clients with different backgrounds and incomes speak highly of BDB staff.The trust that has developed between customers and staff, reinforced by recognition of good staff performance by BDB management, has helped to build the kind ofjob satisfaction that enables the bank to retain its staff. 156 The Microfinance Revolution: Lessons from Indonesia Loan products BDB provides loans to individuals ranging from less than $100 to more than $100,000; there are no group loans. Many of today's large borrowers began as small savers and borrowers in the 1970s. Although BDB emphasizes microfi- nance, it considers itself a bank for the general public. The bank's philosophy is to provide continued financial services to good customers whether their en- terprises remain small or grow large. Borrowers range from proprietors of tiny fruit stands and barbershops to gold merchants and supermarket owners. Three types of loans are provided: * Retail commercial loans (52.5 percent of the value of the 1996 portfolio). * Consumer loans, classified as motor vehicle, housing, and personal loans (9.8 BDB staff are percent). * Loans to larger private or corporate clients, some of whom have lines of cred- treated with respect it (37.7 percent). and thoughtfulness In 1996 BDB's annual effective interest rates on loans varied from 20 per- cent (for prime customers) to 36 percent, with the interest on most loans set by their employers, at about 30 percent. For large loans collateral of 150-200 percent of the loan amount is normally required, but for small loans many forms of security are and are taught to accepted, and some small loans are provided without collateral. Most BDB loan terms are for less than one year, but loans can be rolled treat their customers, over. Housing loans are primarily for land purchase and generally have terms ranging from one to five years. Five years is the maximum term for any loan. both rich and poor, Loan appraisal, decisionmaking, and the release of funds are generally accom- plished within three to five days for new borrowers. Loans for previous clients in the same way with good repayment records can be processed in a day. The government requires Indonesian banks to provide 20 percent of their loan funds for Kredit Usaha Kecil (KUK), or Small Business Credit, defined in 1996 as loans up to 200 milion rupiah ($84,000 in 1996; see chapter 9). In ad- dition to meeting its quota, BDB acts as a channeling agent for other banks that prefer not to serve this market directly; there are many more requests for this service than BDB can accept. Under the KUTK channeling arrangement, BDB locates creditworthy borrowers and recommends them to the lending bank. BDB bears no risk and receives a fee of 1.5-2.0 percent of the loan. In 1996 BDB channeled about $9 million to small and medium-size borrowers in this way. Savings products Microsavings is BDB's specialty.The design and implementation of voluntary savings instruments and services for low-income people are highly developed. Services to savers include maintaining daily routes on which staff collect de- posits and pay out withdrawals at the customer's home or place of work, ad- vising savers on account types appropriate for their specific needs, and providing quarterly lotteries for which free tickets are given to savers according to the minimum monthly balance in their accounts. Where the Microfinance Revolution Began Bank Dagang Bali 157 As indicated, savings instruments are of three general types: time deposits, giro accounts (current accounts that provide checking services), and different types of passbook savings accounts. In 1996 the annual interest rate on rupiah time deposits was 16.0 percent for 1- to 3-month deposits and 16.5 percent for 6- to 12-month deposits. Interest on U.S. dollar time deposits was 7.5 to 8.0 percent.The annual interest rate for giro accounts ranged from 6 to 9 per- cent, in most cases paid on the minimum monthly balance. In 1996 general passbook savings, which are fully liquid, paid 12 percent annual interest on the minimum monthly balance. Another passbook savings product for larger accounts paid from 13 to 15 percent a year on the minimum daily balance on accounts ranging from $420 to $4,196. There are also long- BDB's lotteries are term passbook savings accounts for retirement (which can be withdrawn at age 55), for education (which can be withdrawn at age 19), for housing, and for well known ceremonies.The annual interest on these accounts was 15 percent in 1996. In 1996 the cost of loanable funds was 15.2 percent; of this, 11.3 percent was fi- throughout Bali. In nancial costs and 3.9 percent was administrative costs. the early years the Lotteries BDB holds lottery drawings four times a year for all holders of passbook savings most valuable and giro accounts.The lotteries, which have been held since 1971, were inspired by the Okas' knowledge of the psychology of smaller savers.They perceived that annual prize was a many poor people in Bali, as in other places, are attracted to lotteries because they view gambling as the only possible means of escape from poverty.The Okas sought motorcycle; later to transform this widespread interest in gambling by removing the lottery risk for the participants, while simultaneously encouraging them to save in BDB. BDB offered a car, Every month savers receive free of charge one lottery number for each $4 (in 1996) of their minimum monthly balance.The lotteries are extremely pop- and then a house ular and well known throughout Bali. In the early years the most valuable an- nual prize was a motorcycle; later BDB offered a car, and then a house. Smaller prizes are also given at each lottery drawing. BDB uses the lottery drawings as an opportunity to provide information about the bank's instruments and services, and more generally about the role of banking in social and economic development. Large prizes are delivered to the recipient's village, where the occasion is used to promote BDB and to ed- ucate local people about banking. Many other Indonesian banks, including BRI's unit desas, have adapted the BDB lottery for their savings instruments. In addition, when Bolivia's BancoSol began mobilizing voluntary savings in 1993, lotteries adapted from BRI-which had adapted them from BDB-were included in the product design. Financial institutions in other developing countries have also expressed interest in adapt- ing the BDB lotteries for their microsavings mobilization efforts. Mobile savings teams BDB has three types of mobile savings teams that serve clients whose work- places or homes are located along three roughly concentric circles around each 158 The Microfinance Revolution: Lessons from Indonesia branch.7 One savings team travels by foot, covering the areas nearest the branch. A second team travels by motorcycle and covers more distant clients. A third team travels by car and serves the branch's most distant savers.8 Each team con- sists of two uniformed staff members, either two women or two men. In each of the mobile savings teams, one member carries the money while the other is responsible for the account books.To discourage collusion and fraud, one member of each team is rotated every six months.The teams cover a spec- ified route every day; their primary purpose is to collect savings from BDB de- positors, but they also provide withdrawals and other banking services as needed. Although loans are handled at the branch office, the savings teams acts as a liaison between the branch and the client in arranging for loan informa- tion to be provided; when requested by clients, a savings team can deduct bor- BDB collects rowers' loan payments from their savings accounts. The daily routes are arranged to accommodate customer demand. Many savings from the customers make daily deposits; some save on specific days several times a week. Others save weekly, biweekly, or monthly. Customers often have several BDB employees of many savings accounts that are either held in the names of different family members or maintained by a single individual for different purposes. institutions in Depending on the route and the type of transport, a team can cover 50 to 100 customers a day. In a single stop at a home or workplace, a team some- Jakarta, including times collects deposits from up to 10 savers. Savers who are not located in a fixed place every day often ask their familly, neighbors, and friends to hold their the Ministry of passbooks and money, and to make deposits for them when the team visits.This approach is helpful for the savers and efficient for the bank. The arrangements Finance and Bank among the savers are private; BDB bears no risk from this practice. BDB savings teams also serve many employees in private enterprises and Indonesia public institutions.The team goes to the place of employment on payday and collects deposits from the employees.When BDB opened its branch inJakar- ta, the mobile savings teams used the techniques developed in Bali; BDB col- lects savings from the employees of many institutions in Jakarta, including the Ministry of Finance and Bank Indonesia. As with the lotteries, many Indone- sian banks have copied or adapted BDB's mobile savings teams. Public relations BDB excels at public relations.The Okas, who know many of their clients per- sonally, are active in community events and often have opportunities to rein- force old ties and make new acquaintances. Clients are given courteous personal attention and provided with helpful and clear information. BDB gives gifts to its clients on religious holidays (keeping careful track of who belongs to which religion) and for major family events. As a result of its customer service pro- gram, BDB's reputation as a bank that provides good service to low-income borrowers has spread widely by word of mouth. In addition, Mrs. Oka visits clients frequently, and Mr. Oka is a regular guest at events held by the provincial and local governments and by other organiza- tions.At such occasions he provides information to the organizations' employees Where the Microfinance Revolution Began Bank Dagang Bali 159 generally about how banking services can help their families, and in particu- lar about savings instruments suitable for the education of their children and for the employees' retirement funds. Lottery drawings are used to provide in- formation and publicity about the bank's services. BDB also receives consid- erable publicity from journalists who find that reporting on its services to low-income clients makes good newscopy. Bank Dagang Bali's Savers BDB's customers come from a wide variety of backgrounds and are engaged BDB gives gfts in many types of economic activities. Most have been poor at some time in their lives. Because of Balh's substantial tourist revenues, export growth, and rel- to its clients on atively high income growth, many of BDB's long-term clients have experienced considerable upward mobility since the bank opened in 1970.They frequent- religious holidays ly attribute their economic development, in part, to the financial services and information obtained from BDB. (keeping careful BDB customers include producers of many types of goods (food, garments, Jewelry, handicrafts, paintings, leather goods), owners and employees in shops track of who belongs of all sizes selling these and other products (building supplies, hardware, vehi- cles, electronics, tape cassettes), farmers, owners and employees in service en- to which religion) terprises (restaurants, hotels, tailors, barbershops, garages), government servants, and pensioners. This section is based primarily on a study of BDB savers that I conducted in 1994.9 At that time a number of niicrocredit institutions in other parts of the world were becoming interested in collecting voluntary savings; the main pur- pose of the study was to provide lessons on savings mobilization from a small private bank that at that time had been in operation for nearly 25 years. As part of this study, I accompanied BDB mobile savings teams on their daily routes from six branches on Bali, and interviewed savers of many types and income levels. As is characteristic ofJava and Bali generally, the households of most of the people I interviewed had multiple income sources. Many households also had multiple passbook savings accounts; these were usually for different family members or for different purposes. In addition, some households had time de- posit accounts, giro accounts, or both, and some had long-term passbook sav- ings accounts for retirement or children's education. Some were also BDB borrowers, most of whom chose to have their payments deducted from their savings accounts. Finally, some households were also clients of other banks. Many of the larger customers interviewed were long-term BDB clients. Near Denpasar this usually meant that they had used BDB for more than 20 years; a few were originally customers of Bank Pasar Umum. Elsewhere long- term clients were those who came to BDB when it opened branches in 1988, though a few of those interviewed had been clients of the early mobile bank. One of the customers interviewed had won a house in the BDB lottery; an- other had won a motorcycle. 160 The Microfinance Revolution: Lessons from Indonesia Many of BDB's savers are salaried employees, working in government or private offices. In one government school that I visited, all 15 teachers saved a small amount monthly in passbook savings accounts; 2 also had long-term BDB retirement savings accounts.The BDB team visits the school monthly, collect- ing the teachers' savings from the school treasurer, who deducts the amounts to be saved from the teachers' salaries. Similarly, BDB teams make monthly vis- its to a local office of the Department of Health; when I visited the office, sav- ings were collected from 49 employees. Savings are also mobilized from employees in private businesses. In some cases BDB collects deposits from owners and employees at the same workplace. In other cases the BBD savings teams visit the homes of the employees. BDB savers range from the lower levels of the economically active poor to BDB savers range the affluent; examples are provided below. Although the emphasis here is on savers, it should be remembered that, as in microfinance generally, clients often from the lower levels use savings and loans together. RT and BT, the husband and wife discussed in chapter 3, are good examples.They started out in 1980 as a young low-income of the economically couple, a driver and a waitress. During 1980-94, they used their BDB savings to purchase assets that were then used as collateral for BDB loans. The loans active poor to the were used to build additional businesses, from which they generated increased savings. By 1994 the couple owned 10 substantial enterprises. affluent Low-income savers Among BDB's low-income savers are many who operate microenterprises; such clients often sell cooked food, botded drinks, snacks, and garments. Some are producers-of food, garments, handicrafts, furniture, and the like. But some are employees or work for daily wages. Average net daily income from the main household enterprise ranges from about $2 to $6 but can reach $15 on a particularly good day; some households have other income sources as well. Typically these customers save regularly in BDB-often daily-in amounts ranging from $1 to $3 per deposit. Some of these savers also have small BDB loans. These savers reserve some cash from each day's revenue for the next day's expenses. In most cases the amount saved is the amount taken in that is not expected to be needed for the next day's household and enterprise expenses. Usually there is a one-day lag in depositing savings. Because the mobile sav- ings teams visit clients in the mornings, the money not needed on the previ- ous day for the present day is usually deposited with the BDB team. One customer I interviewed m 1994, MK,° had migrated fromJava in 1973; in Ball he became an itinerant peddler selling ice cream. By 1975 he had been able to obtain a fixed place in the market where he had worked ever since, sell- ing ice cream from his pushcart. His average net daily income in 1994 was about $4. His wife, who sold indigenous cosmetics and medicines as an itinerant ped- dler in the same market, earned about the same amount. MK had been a reg- ular BDB saver for five years, and he usually made deposits daily. He also had a savings account in a second bank. MK's household used its savings for its chil- Where the Microfinance Revolution Began Bank Dagang Bali 161 dren's education (one child was in primary school, one in junior high, and one in high school), for medical expenses, and for ceremomal occasions. In addi- tion to his savings account at BDB, MK was responsible for the savings of an association of itinerant ice cream peddlers who are members of his extended farnily.These savings were deposited monthly in the association's BDB account. Two other men interviewed, both pushcart vendors, each deposited $2 to $3 a day in their passbook accounts. PV, who sells soft drinks and cigarettes, had a balance of $63 at the time of my 1994 visit. CS, who sells snacks and soft drinks, had a balance of $293. Three women-LR, who sells fruits and vegetables; FB, who sells meats; and BL, who sells small ceremonial goods for religious offerings'1-all oper- FB, who had saved ated their businesses in tiny fixed stalls in a large market. All saved daily in BDB in amounts ranging from about $1 to $4. Like MK, they normally withdrew at BDB since only for school fees, ceremonies, and medical expenses. LR had been a BDB saver since 1973. In 1994 she saved about $2 a day. 1988, had.five At the time of a 1996 visit she had about $530 in her passbook account. She also had an account in another bank with $113. On the day of the visit she had passbook accounts held two BDB passbooks for friends who wanted to make deposits of $2 and $3 into their accounts. for diferent members FB, who had saved at BDB since 1988, had five passbook accounts for dif- ferent members of her family, with balances ranging from $22 to $373. She typ- of herfamily, with ically saved about $3 a day, although on a good day she could sometimes save $6 to $10. BL was a relatively new customer, saving in BDB since 1992. Her busi- balances ranging ness varies depending on the dates of particular ceremonies.When I visited her, it was close to a major festival and her enterprise was doing well. She saved $9 from $22 to $3 73 that day; on a typical day, she said, she saved about $3. Her balance was about $132. Lower-middle- and middle-income savers BDB also has many clients who are owner-operators of small and medium- size enterprises.These clients are involved in a wide range of economic activ- ities in trade, production, and services.A few examples can serve as lllustration. RW, who owns a small clothing shop, made daily deposits into five BDB bank accounts for different members of his family. He also had a BDB time deposit account and a BDB savings account for retirement. In addition, he had deposit accounts in several other banks. Another customer, JI, the owner of a small general store, had been saving in BDB for 20 years, making daily deposits. Her balance at the time of the 1994 visit was $3,339. Several of her customers leave savings with her to deposit for them in their passbook accounts when BDB staff visit her shop.A third client, ST, raises chickens and supplies chicken parts to hotels and shops. He had a BDB loan and made daily deposits into his pass- book savings account, from which his monthly loan installments were paid. DP and HP, husband and wife, had been BDB customers since 1987. DP is an artist and his wife, HP, runs a general store.They had had a series of BDB loans, and they also maintained savings and deposit accounts.They saved daily in a BDB passbook account from which their loan installments were paid.They 162 The Microfinance Revolution: Lessons from Indonesia had also saved for several years in a time deposit account in order to renovate their house and build a family temple; construction on both projects had re- cently been completed at the time they were interviewed. High-income savers BDB also has wealthy customers. Some clients regularly deposit more than $100 a day in BDB savings accounts; many also hold BDB fixed deposits, giro ac- counts, or both. Some hold lines of credit or other loans over $100,000. Many of these are long-time customers who started small and later became wealthy. Examples include AN, who owns a large, well-known woodcarving enterprise; SA, who owns a gold shop; and NE and KE, a husband and wife who own a big supermarket. These customers all have both loans and deposit accounts at I asked the Okas BDB, and at other banks as well. whether I could discuss with BRI To its clients, BDB is known as a secure bank with conveniently located branches, and as one that helps low-income clients develop their economic ac- BDB's products and tivities and earn income on their savings. It is known as a bank that serves every- body: multiple generations in the same family, employers and employees in services. The only businesses, peddlers and supermarkets. Internationally, BDB is known as the first bank that understood that the requirement, they demand for microfinance can be met through nonsubsidized, commercial fi- nancial intermediation-and then implemented these ideas sustainably and with said, was that poor substantial outreach. The rest of this volume concerns the development of BRI's unit desas, the people should benefit largest system of sustainable microfinance in the world. In the early 1980s, when Indonesia's Ministry of Finance was considering whether to close down the unit desa system or to transform it into a system of commercial financial in- termediation serving low-income people, there was no known example of reg- ulated, nonsubsidized, self-sufficient institutional microbanking anywhere in the world-except at BDB. As an adviser to the ministry, I made frequent informal visits to BDB at that time to learn the basics of commercial microfinance. When the decision was made to turn the unit desas into commercial units, I asked the Okas whether I could have their permission to discuss with BRI their insights into the microfinance market-and more specifically, BDB's products and services. Their response was characteristic: the only requirement, they said, was that poor people should benefit. Notes 1. It is possible that another developing country institution has carried out the same activities, continuously serving the low-income public for a longer period. But to the best of my knowledge and that of many microfinance experts I have consulted, BDB Where the Microfinance Revolution Began Bank Dagang Bali 163 is the only regulated, formal financial institution in a developing country to have en- gaged m contmuous, nonsubsidized, profitable large-scale financial mtermediation among low-income clients for more than 30 years. 2.This was long before the unut desa system was begun, but BRI's branches were active in the towns. 3. Bank Pasar Umum means Market Bank for the Public. The initials BPU were made the same as those of Bank Perniagaan Umum-the first, and much larger, pri- vate bank in Balh. 4.The year 1985 was the first that BDB was required to pay taxes on its profits. In 1990 BDB paid $0.5 millhon in taxes; its after-tax profit was $1.1 million. 5. Recognizing thus disparity, in 1996 BDB made a concerted attempt to increase small lending.This effort caused the number of loans to increase from 12,814 in 1995 to 15,645 in 1996-a 34 percent jump. The value of outstanding loan balances grew from $79.3 million on December 31, 1995 to $93.6 million on December 31, 1996, an 18 percent increase. 6. ING needed such a relationship to enter the Indonesian market. BDB wanted a transfer of knowledge about international banking and increased international links. Both wanted a joint venture with a bank operating in a different market segment. 7. Mobile teams typically work well in secure regions but are generally not suit- able for high-risk areas. However, BDB has operated these successfully not only in Bali, which is considered relatively safe, but also in Jakarta and Surabaya (where the teams travel with guards). 8. Not all branches have all three types, but all have the walking teams. 9 See Robinson (1995d). Most of the information in this chapter is based on the more than 25 visits I made to BDB and its savings teams between 1982 and 1998. How- ever, this section is based primarily on interviews conducted in 1994.At that time the transactions of 217 BDB clients were observed as I accompamed BDB savings teams on their normal routes; brief discussions were held with each of the savers that the teams visited. In addition, information was obtained from these clients about 142 of their employees in instances when the employees were also BDB customers Ten per- cent of the clients visited were then interviewed in more depth; while not a scientif- ically drawn sample, the savers interviewed represented both genders and a variety of socloeconomic groupings, occupations, and ages. 1 O.These are all real customers, but their names are not provided in order to pro- tect their privacy and bank confidentiality. 11. Unlike Indonesia generally, which is largely Muslim, Bali is predominandy Hindu. 164 The Microfinance Revolution: Lessons from Indonesia How to Fail in Financing the Poor: Bank Rakyat Indonesia's Unit Desa System, 1970-83 This chapter analyzes the problems of subsidized credit programs in Indonesia. In 1970 Bank Rakyat Indonesia's (BRI's) unit desa system was created to implement BIMAS, a massive government-subsidized agricultural credit pro- gram for rice cultivation; later many other, smaller subsidized credit programs were added as well. The term unit desa refers to the 3,600 bank outlets that BRI established at the subdistrict level to channel BIMAS credit to rice farmers. In general, BIMAS produced negative outcomes for rice farm- ers, BRI, and the government. In part the discussion here serves as background for the rest of the analysis in volume 2, because the transformation of the unit desa system in 1984 occurred only after basic lessons about agricultural finance had been learned the hard way. But the unit desas are also 166 shown here in their original context because the same mistakes made by BRI in the 1970s-but to a great extent corrected in the 1980s and 1990s- are being repeated today in many ministries of finance, planning, and agriculture around the world, with similar results. Indonesia's experience may help other countries and financial institutions understand why subsidized agricultural credit pro- grams like BIMAS, aimed at increasing production and raising the incomes of poor farmers on a large scale, cannot achieve these goals. Such programs cannot be significantly repaired or renovated mid- stream; their failures are caused by intrinsic structural deficiencies. The transformation New financial programs based on fundamentally different as- sumptions are needed to achieve those aims. of the unit desa During 1970-83 BRI's unit desa system was a classic example i 1984 of the fourth subsidized credit model discussed in chapter 7: it failed Y at both lending and savings. Analysis of BIMAS shows the com- occurred only afier plexities of the program as it was implemented in Indonesia's var- ied environments and highlights the many ways the program did basic lessons had not-and could not-work as planned.The focus here is on the been learned the interactions among subsidized credit, agricultural production, and rural development. hard way Multiple, interlinked reasons explain the failure of BIMAS: * The program's goals were incompatible. * Subsidized interest rates prevented institutional viability. * BRI was not allowed to select its borrowers. * Credit was tied to predetermined types and amounts of inputs that were often inappropriate for local environmental conditions and sometimes destructive to the rice crop. * In many areas credit subsidies went to wealthier villagers. * In some areas borrowers were selected for the program by gov- ernment officials who needed to meet targets-even though the borrowers' lands were unsuitable for the inputs provided. * In some years farmers could not repay their loans because of mas- sive rice crop failures, a direct result of the resurgence-causing insecticides distributed in BIMAS credit packets. * Government policy for loan rescheduling was badly planned and often corruptly implemented. * BRI's organizational structure was inadequate for effective su- pervision of the unit desas, and unit staff were poorly trained, How to Fail in Financing the Poor Bank Rakyat Indonesia's Unit Desa System, 1970-83 167 underpaid, unmotivated, and typically treated as pariahs by the rest of BRI. * In addition to failing at lending, the unit desas failed at mobi- lizing savings, mainly because of the negative interest rate spread required by the government. Despite the massive deficiencies of its credit component, In- donesia's rice intensification program was a success-one made pos- sible by a strong national commitment to attaining rice The BIMAS self-sufficiency and by the government's ability to learn from its mistakes and to adjust policies and implementation strategies to program's goals the realities of the country's environmental conditions and human resources.When official channels of input distribution or output were incompatible marketing proved inadequate, alternate channels were opened. When particular seed types or insecticides were found to be un- suitable or dangerous, they were removed from the system. Sirmi- larly, when the credit program proved ineffective, it was replaced. BIMAS failed in its mission. But it gave birth to the unit desa system-which survived and ultimately developed into the inter- nationally renowned commercial microbanking system discussed in the rest of volume 2. 168 The Microfinance Revolution: Lessons from Indonesia The problems of subsidized agricultural credit are well known. The subsidies encourage corruption and often reach rural elites rather than low-income farm- ers. The programs tend to have high arrears and high losses. And the low (or negative) spread on interest rates for savings and loans makes it impossible to achieve institutional sustainability-so most of the demand for microfinance remains unmet (see chapter 4).' Bank Dagang Bali (BDB), a private bank, began offering microfinance at commercial interest rates in 1970 (chapter 10). But that same year the Indonesian government moved in the opposite direction-toward large-scale subsidized credit programs. BRI, a government bank, was required to lend at a 12 per- cent annual effective interest rate and to pay 15 percent on most deposits. BRI's unit desa system was created in 1970 to implement BIMAS,2 a mas- The same mistakes sive subsidized credit program for rice cultivation; later the system was extended to provide loans through other, smaller subsidized credit programs.The unit desas' made by BRI in the current commercial orientation is a direct response to, and outgrowth of, that experience.To understand the transformation, this chapter examines the unit 1970s are being desas in their original context. BRI opened 3,600 bank outlets at the subdistrict (kecamatan) level in the repeated today in early 1970s; the units serve the villages of the subdistrict (the literal meaning of unit desa is village unit). Since urban units (unit kota) were added in 1989, numerous countries the terms unit desa and unit bank have come to be used as general terms for all BRI's local units, rural and urban. around the world, The results of the BIMAS program were negative for most rice farmers, BRI, and the government. Analysis indicates the complexities of the program with similar results as it was implemented in Indonesia's varied environments and highlights the many ways that it did not-and could not-work as planned. The focus here is on the interactions among subsidized credit, agricultur- al production, and rural development. In one sense this discussion serves as back- ground for the analysis in the rest of volume 2, because the metamorphosis of the unit desas to a commercial microbanking system occurred only after basic lessons about agricultural finance had been learned the hard way. The reasons BIMAS failed are shown in figure 11.1 and discussed throughout this chapter. In addition to its importance as background for the unit desas' subsequent development, BRI's BIMAS era is of direct relevance to many ministries of fi- nance, planning, and agriculture today.The same mistakes made by BRI in the 1970s are being repeated in numerous countries around the world, with sim- ilar results. The problems of subsidized agricultural credit programs are wide- ly known-yet many governments continue to provide these programs and millions of people still participate in them. Upon learning the history of the unit desa system and seeing it in opera- tion now, many international visitors to BRI have said, in effect, "Our coun- try is still trapped in the old paradigm. How was it possible to change the system in Indonesia? Who made the decisions? Tell us how the change was made; this wlfl help us at home." It is hoped that the rest of volume 2 will help to answer such questions. How to Fail in Financing the Poor Bank Rakyat Indonesia's Unit Desa System, 1970-83 169 Main reasons for the failure Fi ure 11l.1 of the BIMAS credit program for rice farmers Subsidized loans prevented institutional viability, made successful savings BRI staff were BRI was not permitted mobilization untrained and to select its own impossible, and unmotivated borrowers tended to reach large, influential farmers who often did not repay Loans were not repaid Farmers became ineligible to borrow from BIMAS Problem due to Problem due to Problem due to government's other factors factors internal to BIMAS program external to BRI BRI design 170 The Microfinance Revolution: Lessons from Indonesia Presidential decree for BIMAS debt relief in Insecticides In BIMAS cases of crop BIMAS input packets packets caused brown destructon was were Inappropriate planthopper attack, Implemented by BRI for many village which caused severe in ways that often environments rice crop failures in benefited wealthier many areas borrowers, bypassed poorer borrowers, and encouraged corruption Loans were repaid, but with difficulty, after crop loss Farmers did not want to borrow from BIMAS BIMAS participation declined BIMAS ended How to Fail in Financing the Poor Bank Rakyat Indonesia's Unit Desa System, 1970-83 171 Large-scale subsidized agricultural credit programs like BIMAS cannot both increase production and raise the incomes of poor farmers on a large scale. Using the Indonesian example, this chapter shows why. The failure of the BIMAS program in four very different kinds of villages in widely separated areas of the country is examined in detail to illustrate the general reasons for such failures.The focus here is on large-scale subsidized agricultural credit programs; donor-subsidized, usually smaller-scale, multipurpose subsidized credit programs are considered in volume 3. Bank Rakyat Indonesia's History, 1895-1970 BR!, Indonesia's BRI, Indonesia's oldest bank, celebrated its 100th birthday on 16 December oldest bank-the 1995. The modern BRI is the outgrowth of a series of banks dating from the Priyayi Bank of Purwokerto (Support and Savings Bank for Domestic Civil Ser- outgrowth of a series vants, or Hulp-en Spaar Bank der Indlandsche Bestuurs Ambtenaren), which opened in Java in 1895 and is regarded as the beginning of Indonesian rural of banks datingfrom banking. Priyayi means aristocracy or gentry; the bank was founded by Raden Wiriamaadya, aJavanese government official from a priyayi family who "want- 1895-is regarded ed to help his friends out of debt." Wiriamaadya "had noticed that during a selamatan [ceremonial feast] his host had taken a massive loan from a Chinese as the beginning of moneylender in order to meet his social obligations.Wiriamaadya was so dis- tressed by this incident that he decided to contribute to the festivities from his Indonesian rural own pocket and proposed setting up a saving fund for prominent Indonesian citizens"4 (Schmit 1994 p.1; see also BRI 1985,1995). Sieburgh, the Dutch colo- banking nial administrator in charge of Purwokerto District, supportedWiriamaadya in this effort. In 1897 De Wolff van Westerrode, Sieburgh's replacement and an admirer of the Raiffeisen cooperative banks founded in Germany in the 1850s by Friedrich W Raiffeisen, began to reorganize the bank into a cooperative or- ganization. The following year van Westerrode was appointed director of the bank, which was renamed Poerwokertosche Hulp, Spaar, en Landbouwcredi- et Bank.The cooperative approach, gaining favor under the "New Ethical Pol- icy" of Queen Wilhelmmna of the Netherlands oriented toward the welfare of colonial subjects, was generally modeled on the credit union movement that had begun in Europe in the 1850s (see chapter 9).5 VanWesterrode's aim was to create a central institution, a popular credit bank, for the village credit cooperatives in Purwokerto. Although his intention was to build a cooperative bank, the bank also provided services to nonmembers and did not strictly maintain cooperative principles (Steinwand 2001). As Schmit (1991) has pointed out, the moneylending practices of the Chinese in Indonesia, and to a lesser degree of the Arabs, were an important concern of the Dutch colonial administration because usury was identified as the prima- ry cause of deteriorating rural society in the colonies and, therefore, of the in- stability of colonial society. 172 The Microfinance Revolution: Lessons from Indonesia Who actually deserves the title as "the father" of [BRI] has always been a point of controversy. According to Cramer (1929:17) it is DeWolffvanWesterrode, according to De Wolff himself it is his predecessor as head of district, Sieburgh (Schmit 1991: 41), and according to Suharto (1996:13) it is R. Aria Wirjaatmadja [an alternative spelling for Wiriamaadya]. Others say the unique Indonesian financial system is large enough that it can easily live with several fathers. However, more important than the question of fatherhood is that the process of establishing decentralized commercial viable institutions started . . . over 100 years ago, giving ample scope for experiments,institutional learning and adjustments. -Steinwand 2001, p. 67 The process of Other People's Credit Banks began to emerge on Java; increasingly they re- establishing ceived government subsidy, thereby starting the process under which the pop- ular banks came under local government control (Schmit 1994).These became commercial viable known asVolksbank (People's Bank) and over time as Bank Perkreditan Raky- at (People's Credit Bank). institutions started In 1912 an autonomous welfare service, theVolkscredietwezen (Popular Cred- it System), was established in Indonesia.A Central Fund (Centrale Kas) was also more than 100 established, with working capital of 5 million guilders.The Popular Credit Sys- tem and the Central Fund were intended to control, support, and supervise the years ago 75 popular credit banks, 12,424 lumbungpadi (village rice banks), and 1,336 vil- lage banks that had emerged (Schmit 1994, p. 7). It was assumed that "the rural population has a 'chronic credit thirst' and is too poor to save" (Steinwand 2001, p. 46). But by 1926 domestic deposits accounted for 66 percent of the total de- posits of the district banks of Java and Madura (Stemnwand 2001, p. 82). The Volkscredietwezen was heavily influenced by J.H. Boeke, a specialist in colonial economics from the University of Leiden who had been strongly influenced by views developing in Germany-especially those of the econo- mist Sombart-about pre-capitalist societies. Pre-capitalist communities were thought to be divided into two strictly separate classes: the ruling class and the masses. The latter, the producers, were believed to have limited economic needs (see Fruin 1994 [1935]). Applying the ideas of pre-capitalist society to the colonial economy, Boeke developed what became known as his theory of economic dualism (see chap- ter 21).6 He believed that indigenous Indonesians were not market-oriented and that, therefore, the aim of the Popular Credit System should not be to pro- vide people with commercial banking services but rather to satisfy social needs (see Schmit 1994). Boeke advocated that the government adopt a protective role, and that the Popular Credit System be instituted as "a separate system based on cooperative principles for the native population" (Schmit 1994, p. 9).At the same time, it was thought that by centralizing the system and providing subsi- dized funds, the government could increase its control over the popular cred- it banks (Schmut 1994, p. 7). How to Fail in Financing the Poor Bank Rakyat Indonesia's Unit Desa System, 1970-83 173 Thomas A. Fruin, like Boeke a graduate of Leiden University, joined the Volkscredietwezen at Boeke's request. But Fruin, who had a legal background, disagreed with Boeke about some of the fundamental principles of the Popu- lar Credit System. His aim was not to "protect" the indigenous population but to provide banking services to the emerging middle class.Thus Fruin empha- sized developing banks as institutions, focusing on the importance of new au- diting regulations, supervision, control, and institutional reorganization. On hearing that Fruin had agreed to joinVolkscredietwezen, Boeke wrote to him in 1920:"Banks may indeed be commercial institutions, but in such cases their social principles transcend their immediate financial interests. It would besmirch their good name to act in any other way" (quoted in Schmit 1994, p. 13). In 1921 profits of TheVolkscredietwezen and the Central Fund were reorgamzed, and by 1925 nearly all the popular banks had agreed to centralized control. But Boeke and the the village banks administrators of the popular credit banks did not share a common view of the role of these banks. At a meeting in 1925 Boeke likened the collaboration of the were 4.2 percent of banks and theVolkscredietwezen to the "strain of an unhappy marriage." One of the bank administrators stood up and asked for"a divorce" (Schmit 1994, p. 16). the amount In 1929, 5,986 village banks made 3.3 muillion loans worth 45 milion guilders.7 In addition, 5,268 village rice banks lent 1 million quintals (100 kilo- disbursed grams) of rice to 1.1 million borrowers in the same year.According to the 1930 census, the population ofJava and Madura was 25 million; there were 13,266 villages.Thus these banks reached a large portion of villages and rural house- holds. Fruin (1999 [1933], p. 5) comments, "The operational costs of the vil- lage banks in the period 1925-1930 were 3.8 percent of the amount disbursed and those of the rice banks in 1930 were 11 kilograms plus 52 cents per quin- tal of rice lent.The fact that these institutions were profitable (for example, 4.2 percent of the amount disbursed in 1921) and that they kept large deposits in the popular credit banks (10 million guilders in 1931 inJava and Madura alone) makes it even more interesting." By the late 1920s Boeke's work withVolkscredietwezen had ended, and Fruin had become its director. In 1930 Fruin proposed the creation of the Algemeene Volkscredietbank (AVB) as a bank that, combining financial and social exper- tise, would be a socially responsible credit institution. The AVB was founded by royal decree in 1934, and Fruin was appointed president. Even Fruin believed that an indigenous Indonesian villager was not "an eco- nomically active person. He may be thankful to fate for bringing him greater prosperity but it is not part of his nature to take fate into his own hands and work towards such prosperity himself through rational consideration and the appropriate continuous effort" (Fruin 1994, p. 99 [1935]). Despite their differ- ences, however, Boeke and Fruin agreed that the popular banks were established to complement commercial institutions, and that the banks served segments of the population not otherwise reached by the formal commercial sector. As noted, Fruin wrote an extraordinary treatise on Indonesian rural cred- it, the Provisional Manualfor the Credit Business of the General Popular Bank, pub- lished by the AVB in 1935.8 As in an earlier article,"History, Present Situation, 174 The Microfinance Revolution: Lessons from Indonesia and Problems of theVillage Credit System (1897-1932)" (1999 [1933]), Fruin demonstrates a deep understanding of the economic activities, income flows, interlinked transactions, consumption patterns, and credit practices of Indone- sian microentrepreneurs, farmers, tenants, sharecroppers, salaried officials, and pensioners, and others-that is, of nearly all rural inhabitants.9 The ways that farmers used credit were reviewed by type of crop, type of land, and number of harvests per year.The demand for financial services by occupation and type of consumer was also analyzed, as was the flow of informal credit (box 11.1). Fruin outlined various financial instruments that he believed to be both suit- able for the bank and appropriate for local demand. In doing so he made use of his extensive knowledge of local markets-knowledge largely derived from the work he and Boeke carried out for their many contributions to the month- The popular banks ly periodical Volkscredietwezen, which was a model of transparency in information. served segments of [Volkscredietwezen] provides a wealth of information about the institutional history of the popular credit system. It included the population not reports, announcements, guidelines and policy proposals,journal entries submitted by bank staff, travel sketches, quarterly reviews otherwise reached by of the branch and village banks, articles on the funding of various kinds of agricultural, production, industry and services, theformal market updates and analyses, case studies of village and branch banks, client analyses, prescribed literature for service commercial sector examinations, and details of promotions, dismissals, transfers, anniversaries and courses for bank staff. It rallied its readers to protest against or support reorganization plans, publicised the rmnutes of meetings with bank directors, and reviewed the parliamentary debates of these subjects. -Schmut 1994, p. 20 In 1946, shortly after Indonesia's independence, Bank Perkreditan Rakyat became BRI.10 BRI then merged with the AVB in 1950, forming a state-owned commercial bank." Indonesia's five state-owned commercial banks each has special responsi- bilities: BRI's mandate is to provide banking services to rural areas, with an em- phasis on agriculture and rural enterprises. In the years that followed the 1950 merger, BRI became Indonesia's largest bank in terms of numbers of branch- es and customers, and one of its largest in terms of assets. The Development of the Unit Desa System, 1970-83 The origins of the unit desa system can be understood only in the context of Indonesia's long struggle for self-sufficiency in rice, the country's main staple food.When the nation's revolution for independence ended successfiuly in 1949, domestic rice production had fallen below 6 million tons. On Java, the coun- How to Fail in Financing the Poor Bank Rakyat Indonesia's Unit Desa System, 1970-83 175 Box 11.1 Agricultural loans in Java in the 1920s and 1930s: Excerpts from Thomas A. Fruin's The Provisional Manual for the Credit Business of the General Popular Bank Classification of types of farming [on Javal from the point of view of the extension of credit. 1 Farms with one harvest per year, which serves for purposes of subsistence but part of which may also serve as a cash crop These will be primarily in regions where a relatively large amount of land is owned and there is a rice surplus and where a significant part of the annual paddy harvest is sold If other income is of little importance loans to be repaid as a lump sum after the paddy harvest will be the appropriate form of credit for farmers As the paddy surplus is often re- The times at which moved by the middlemen buying it up immediately after it is harvested, the repayment date set by the bank for most farmers must be set precisely with the harvest in mind Indi- loans are extended vidual farmers who are known to retain their paddy with an eye to price increases may ei- ther be given more time to pay from the beginning or granted an extension of the credit The situation is different in the purely paddy-farming areas, where the harvested do not matter, paddy serves primarily for subsistence and local sale and where rice may even be im- ported from outside Except by a few large-scale farmers, the paddy in these areas is providing one loan not sold immediately after harvest as a cash crop but kept and at most disposed of in part very gradually In such areas it would be difficult to repay a loan immediately after the harvest from paddy revenue This is why in the Bandung [West Java] high plains does notfollow (which are such an area) seasonal loans only have to be repaid 12 months after they are taken out, or several months later if the repayment date should fall during patjekl,k quickly on the hees *[time of scarcity in the pre-harvest season] quickely on the heels 2. Farms with two harvests per year, one of subsistence crops and the other of cash of a previous one crops Tobacco and maize are grown alternately, for instance, in the hills of Central Java, like- wise tobacco and sawahpadl [irrigated rice] in Bondowoso [East Java] In such cases repay- ments must be required immediately after the harvest of the cash crops, attempts to have repayments made in two instalments from both . harvests, were not successful The times at which loans are extended do not matter, providing one loan does not follow quick- ly on the heels of a previous one This should not indeed be necessary if care is taken to keep the sums loaned small so that after repayments have been made, taxes paid and oth- er monetary expenditure completed, sufficient funds remain from the harvest of the cash crop to cover the next planting of paddy 3 Farms with two paddy harvests per year Attention should be paid to whether one of the harvests is especially intended to yield a cash crop, and if so which one, or whether both or neither is so intended [In areas where most farmers have only small plots of land] almost no paddy is sold In such areas only larger landowners with paddy surpluses intended for sale are eligible for harvest loans Others must make do with monthly loans, providing they have sufficient additional income If one of the paddy harvests is especially intended for sale as well as subsistence, loans and repayments can be restructured as for type 2 4 Two paddy harvests a year yielding sufficient monetary income Two types of loans are possible (a) repayment in two instalments, one after each harvest (the instalments do not need to be equal); or (b) repayment in one instalment after the first harvest Type a loans may be larger than type b loans [which will] in prac- tice easily result in two loans being extended per year. The amount loaned should be kept very small in the case of type b loans 176 The Microfinance Revolution: Lessons from Indonesia Box 11.1 (continued) 5 Farms with one paddy harvest and various second crops In Pasuruan IEast Java, inl a sawah area with two consecutive polowidjos [secon- dary crops] and all kinds of mixed cultivation, the popular credit bank there ... prefers to extend only monthly loans in the sawah area (seasonal loans being extended in the main only for vegetables, coffee, and tobacco cultivation in the hills), with the result that there are few farmers with loans in the region . De Vries established that farm- ers in the region could usually get by for the whole year because the working of the sawah land is paid for from the revenue of the polowidjo, while the polowidjo cultiva- tion is financed from the revenue from the paddy which is sold.... Should a need for seasonal loans arise, it would be obvious to require payments to be made in the mon- Thefundamental soon season, where possible in two or more instalments, or otherwise in a single in- stalment at the time the most profitable polowidjo crop is sold Repayments should in rinciple is to any event not be required from the subsistence crops and only one seasonal loan per year should be extended 6 [Mixed farms] structure repayments It can be particularly difficult to determine when harvest loans should be required if f l t there is no clear distinction between the seasons and if crops at various stages of ma- Oseasonal loans to turity are in the fields at the same time or there are three paddy harvests in two years, as in the hilly regions of West Java and southern Central Java The same difficul- fall when thefarmer ties can arise if a mixed farm is operating on both sawahs and tegalans (higher-situat- ed, non-terraced, non-irrigated farmland). [An example of these difficulties is] cassava, which is sometimes a subsistence crop receives more and sometimes a cash crop Where it is processed to make gaplek [dried cassava, tapio- cal, as in parts of Central and East Java, the period from July to September can be monetary income deemed to be a period of cash crop harvesting, where it is processed to produce flour (as is usually done in West Java) there are no pronounced fluctuations between sea- than at other times sons In general it is to be recommended that the bank should not be too quick to per- mit a loan to be repaid only after a year and that repayments should never be required during patjekl,k but at times when farmers have the most money to spare. The fundamental principle is that in its own interests and those of the farmers, the AVB Bank should structure repayments of seasonal loans so that they fall during a period when the tani [farmer] is receiving more monetary income than at other times and can meet repayments from that income without too much trouble, i e. usually after the har- vest of his most lucrative (in terms of gross monetary revenue) cash crops A cash crop is preferable even where part of the principal paddy harvest also serves for trading purpos- es, because expenses for ritual feasts are associated with the traditional, sacred paddy harvest which do not arise in respect of harvests of cash crops, which were introduced at a later date Source Fruin 1994 119351, pp 132-40 try's main rice-producing region, it was not until the cropping year 1954/55 that production recovered to the 1940/41 prewar level; total domestic rice pro- duction in 1954/55 was about 8 million tons (Mears 1981, pp. 20-21, 489). Thus a high priority was placed on increasmg rice production during the 1950s, and considerable growth in rice output was achieved during that decade: How to Fail in Financing the Poor Bank Rakyat Indonesia's unit Desa System, 1970-83 177 about 30 percent on Java and 60 percent off Java, with the national average at just over 40 percent. On Java these increases resulted from higher yields, as the beginnings of new high-yielding rice technology reached farmers there. OffJava the growth in rice production was attributable to a rapid increase in the area cul- tivated, as a growing population opened up new lands (Mears 1981, p. 20). Over- all, however, the substantial mcrease in production was insufficient to reduce import dependency because of the country's rapidly increasing rice consumption. With rice imports at about 1 million tons a year in the late 1950s and the balance of payments worsenung, President Sukarno announced a three-year plan (1959-62) to make Indonesia self-sufficient in rice. For a variety of reasons, in- cluding poor program design, logistical and technical problems, and poor re- Mounting economic payment, the program failed.12 Rice production did not rise during this period, and rice imports increased. problems led to By 1965 mounting economic problems had led to rapidly increasing debt, soaring inflation, and decreasing exports. Net foreign exchange reserves were rapidly increasing negative, and inflation rose above 600 percent. Although a number of rice in- tensification programs were established, they could not succeed under these con- debt, soaring ditions.13 It became very difficult to obtain inputs for rice cultivation. There was little capacity to produce fertilizer and insecticide domestically, while for- inflation, and eign exchange to pay for imports was almost nonexistent. Rapidly deteriorat- ing rural conditions resulted in sharp declines in yields, and the food supply decreasing exports. decreased drastically. Rice accounted for 21 percent of imports, but the coun- try could import little of anything.The situation led to anti-Sukarno demon- Rice intensification strations and contributed to the fall of Sukarno and the installation of the New Order government under President Soeharto. programs could not In 1967 the new government admitted the failure of the country's mass ex- tension program and instituted a number of new rice intensification programs. i4 succeed under these Input supplies continued to be a severe problem, however.As a result BIMAS Go- tong Royong (BIMAS Cooperation) was begun in 1968-69; under this program, conditions multimillion-dollar contracts were signed with foreign companies, each of which agreed to provide inputs for a particular region of Indonesia. Credit to purchase inputs was extended to farmer groups that were expected to lend to participat- ing farmers. Each farmer with a plot within a specified block of land was required to cultivate the land with rice according to government specifications. Informally known as Block BIMAS or Group BIMAS, this program was widely considered a failure. Farmers objected to what many considered coer- cion by government officials about how their lands were to be cultivated.Yields were well below expectations. Corruption scandals erupted involving some of the contractors.And by 1969-70 loan repayment rate was only 53 percent. By this time Indonesia had become the world's largest rice importer. In April 1970, after an incognito visit to paddy-growing areas, President Soeharto ordered a change in the country's approach to rice intensification. In May 1970 the government announced that it would stop contracting parts of the program to foreign companies, and would instead assume direct responsibility for the planning and implementation of the rice intensification 178 The Microfinance Revolution: Lessons from Indonesia program. It was under this new program that the unit desas were created. BRI's unit desa system began inYogyakarta in 1969, as part of a pilot pro- ject for the BIMAS credit program.The pilot project was strongly influenced by Fruin's ideas and the AVB approach, and key people involved in the pilot project, including the manager of BRI'sYogyakarta branch, had known the AVB well. Klaas Kuiper, posted inYogyakarta in 1968 as project leader of a Food and Agriculture Organization (FAO) project, participated as a member of a small team in the design of the pilot project. He wrote: Basically our view was that where the AVB/Village Banks could be profitable before the war with many small, usually short- term loans, it might be useful to try to copy at least some of these The core of the new AVB characteristics into the policies and procedures of the first village units [VUs] in the pilot project.... theVUs were NOT BIMAS approach created only to improve the BIMAS program. We wanted an alternative ... fully in independent BRI hands. Our overall aim was a delivery was to create, or rather re-create, a rural financial infrastructure, providing all bank services in rural areas, using pre-war data and program of systems as an example ... agricultural Although the experiment in Yogyakarta had only some preliminary, promising results, an expansion took place outside extension, credit, [the] province in the 1970 wet season ... Considerable pressure was exerted on BRI to expand with about 500 new VUs per input supply, and year ... This far exceeded BRI's capacity. BRI simply did not have enough trained staff to embark on such a vast expansion, output marketing neither was the internal administrative system sufficient to cope with this. Internal audit furthermore was very weak at that time. services The independent position which BRI had created for itself in Yogyakarta thereby got lost and it became again a tool in the hands of centralized government and presidential directives.'5 -Kuiper 1998 The new program, called Improved National BIMAS, began in the 1970/71 wet season.1" BIMAS expanded rapidly, becoming a national program in 1973 through Presidential Instruction 4/1973. BIMAS started as a credit program for rice farmers but was later extended to other forms of agriculture such as secondary crops (palawija) and poultry. However, the program for rice intensi- fication was by far the largest of the BIMAS programs, and is the only one dis- cussed here. The core of the new BIMAS approach was a four-part service delivery pro- gram composed of agricultural extension, credit, input supply, and output mar- keting services. Complementing this service delivery system were both a new price support system for rice and accelerated investment in irrigation. BIMAS was given high priority by the government because it was seen as the key to How to Fail in Financing the Poor Bank Rakyat Indonesia's Unit Desa System, 1970-83 179 increasing domestic rice production, which would raise rural incomes, improve nutritional standards, help maintain pohtical stability, and eliminate dependence on rice imports. The new program was to be implemented in blocks, each consisting of two to seven villages and 600 to 1,000 hectares of irrigated riceland (sawah). Each block was to have four basic facilities: * An agricultural extension worker. * A retail store, run either by government-sponsored village-level cooperatives (Koperasi Unit Desas, or KUDs) or by private traders, to supply fertilizer and other inputs. Subsidized . An agency to manage the procurement, storage, and processing of rice; it was assumed that this would be carried out by the village cooperatives but these government lending were not given a monopoly. * A new kind of village bank, the BRI umt desa, that would provide subsi- programs proliferated dized credit to farmers directly. at BRI until there In aggregate, these features represented a fresh approach to the longstand- ing problem of raising the country's rice production. were more than 350 By 1983 BRI had established a network of 3,626 unit banks at the sub- district level in rural areas throughout the country. The unit desas, which op- erated as branch windows, provided services to an average of 18 villages, though there was a wide range.'7 Staffed by four employees each (a unut head, a credit officer, a bookkeeper, and a teller), the units were supervised by BRI's network of branches located at the district level in most areas of the country. Liquidity credits were made available to BRI for the unit desa system by Bank Indonesia, the central bank, at 3 percent per year. BRI bore 25 percent of the default risk, while the Ministry of Finance assumed 50 percent and Bank In- donesia the remaining 25 percent. The government also subsidized the ad- ministrative costs of the unit desas. Until 1984 the unit desas functioned primarily as channeling agents for sub- sidized government lending programs. These programs proliferated at BRI until more than 350 subsidized credit programs-for food, cattle, poultry, fish- eries, tree crops, and the like-were provided according to government in- struction. But the BIMAS rice credit program remained the units' main activity until 1984.The 12 percent annual effective interest rate for all unit desa loans remained unchanged from 1970-83. In most years this was a negative interest rate in real terms (see table 2.1), and it was below the 15 percent a year that the government required be paid on most unit desa savings accounts.Through- out this period the government determined the unit desas' interest rates, cred- it ceilings, and criteria for borrowers. BIMAS loans to rice farmers The BIMAS credit component of the rice intensification program provided sub- sidized loans for inputs of fixed types and amounts, supplied mostly m kind (seeds, 180 The Microfinance Revolution: Lessons from Indonesia fertilizer, insecticide) by the village cooperatives (KUDs). Borrowers, howev- er, were required to repay their loans in cash to the unit desas. Although BRI was responsible for collecting BIMAS loans, it had little control over the se- lection of borrowers.That task was carried out primarily by officials of the De- partment ofAgriculture and various local government officers and committees charged with meeting government targets for BIMAS loans. Loan amounts were predetermined in accordance with the amount of land to be cultivated. BIMAS loans were supposed to be disbursed to each par- ticipating farmer just before the planting season. In fact, the loans often reached borrowers late; in such cases farmers had to plant without fertilizer and other inputs, or purchase these-with funds from other income sources, savings, or informal loans-before the BIMAS loans were received.Then-hke the many From 1976-84 on- farmers who cultivated fields for which the BIMAS inputs were not suitable in kind, amount, or both-they had to use, sell, trade, or store the BIMAS in- time repaymentfor puts received. Loans for paddy cultivation were provided separately for the wet and BIMAS loans dry seasons; for most of the BIMAS period the loans were due just after the harvest of the crop for which the inputs were used.Thus farmers who averaged only 57 paid their loans from rice sales had to sell their crop when the rice price was at its lowest. percent. Program In its early years BIMAS helped farmers learn new high-yielding methods of rice cultivation. Loan repayment during 1970-75 was better than that of most participants were of the country's subsidized agricultural credit programs (figure 11.2). In terms of the value of outstanding credit, the program peaked in the 1975/76 wet sea- primarily wealthier son, with about 55 billion rupiah ($133 million) provided to more than 2.5 mil- lion borrowers (figures 11.3 and 11.4).About 2.2 million hectares were covered farmers under the BIMAS program that season (figure 11.5). Soon afterward, however, widespread crop failures-combined with defi- ciencies in program design, input packets inappropriate for the land of many farmers, and misguided implementation strategies-resulted in higher defaults (loans with payments 90 days or more overdue; see figure 11.2) and in declin- ing program participation (see figure 11.4). In the 1983/84 wet season the value of BIMAS credit outstanding reached only about a quarter of the 1975/76 total (see figure 11.3), while the number of borrowers was down to about 15 percent of the 1975/76 total (see figure 11.4). Hectarage covered by the program fell as well, to about 12 percent of the earlier total (see figure 11.5).The wet season is a much more important rice- growing season than the dry season, but it is worth noting that in general the dry seasons followed the same general patterns as the wet seasons. According to BRI's head office, from 1976-84 on-time repayment for BIMAS loans averaged only 57 percent. During those years program partici- pants were primarily wealthier farmers: the average BINMAS loan was for about 1 hectare of irrigated land. Only a small segment of rice farmers owned 1 hectare or more of irrigated riceland. On Java a poor farming household that owned irrigated land (many did not) would typically own less than 0.25 hectare.18 How to Fail in Financing the Poor Bank Rakyat Indonesia's Unit Desa System, 1970-83 181 Figurel 1.2 Default rates for BIMAS rice cultivation loans, 1970-84 60 Percent 50 40 30 Dry season 20 10 _ W~~~~~~~~et season 10 0 Note Wet seasons span two calendar years (1970/71, 1971/72, and so on) Dry seasons occur within a calendar year (1971, 1972, and so on) Thus the 1971/72 entry shows both the dry season data for 1971 and the wet season data for 1971/72 BIMAS loans continued to be collected after the program ended, and by 1990, 83 5 percent of the amounts due had been collected Loans in default refer to loans with payments more than 90 days overdue Source BRI data, Patten and Rosengard 1991, p 63 Figure 1.3 Amount of outstanding BIMAS rice cultivation loans, 1970-84 60 Billions of rupiah Wet season 50 40 30 Note Wet seasons span two calendar years O1970171, 19711.72, and so on) Dry seasons occur within a calendar year (1 971, 1972, and so on) Thus the 1971172 entry shows both the dry season data for 1971 and the wet season data for 1971/72 Source BRI data 182 The Microf1inance Revolution: Lessons from Indonesia Number of outstanding Fi ure 1 1 A BIMAS rice cultivation loans, 1970-84 3.0 Millions of borrowers 2.5 2.0 1.5 _t ~~~~Dry season sao 1.0 0.5 0.0 Note Wet seasons span two calendar years 11970/71, 1971/72, and so on) Dry seasons occur within a calendar year (1971, 1972, and so on) Thus the 1971/72 entry shows both the dry season data for 1971 and the wet season data for 1971/72 Source BRI data Ficure 1.5 Area covered under BIMAS rice cultivation loans, 1970-84 2.5 Millions of hectares Wet season 2.0 1.5 1.0 Dry season 0.5 0.0 Note Wet seasons span two calendar years (1970/71, 1971/72, and so on) Dry seasons occur within a calendar year (1971, 1972, and so on) Thus the 1971/72 entry shows both the dry season data for 1971 and the wet season data for 1971/72 Source BRI data. How to Fail in Financing the Poor Bank Rakyat Indonesia's Unit Desa System, 1970-83 183 In 1981 the government began a special drive, Gerakan Serentak (GER- TAK), to collect overdue BIMAS loans. Using a quasi-military style of collec- tion, GERTAK succeeded in collecting a significant portion of the overdue BIMAS rice loans. By 1990, 83.5 percent of the loan amounts that had come due during the life of the program (1970-85) had been repaid. However, GERTAK had its costs: after 1981 people increasingly refused to participate in BIMAS. In the 1982/83 rainy season BIMAS achieved only 35 percent of its rupiah loan target and 36 percent of its hectare target; in the 1983/84 rainy season the figures were 12 percent of the rupiah target and 14 percent of the area target.When KUPEDES began in 1984, BIMAS was phased out, official- ly ending in 1985. BIMAS was unable Overall the BIMAS rice intensification program extended $566 mlllion (in 1985 dollars), covering 25.8 million hectares in 28.8 million loans.This repre- to achieve its goals: sented 44 percent of its rupiah target and 61 percent of its hectare target. Low repayment rates were a continual problem after 1976; as noted, however, loans it neither reached continued to be collected even after the program ended. But BIMAS was un- able to achieve its goals: it neither reached poor farmers effectively nor con- poorfarmers tributed significantly to the rice intensification effort. The reasons BIMAS could not achieve its aims are analyzed later in this effectively nor chapter. But the general failure of the program did not seriously hamper the success of the rice intensification program (which succeeded despite its cred- contributed it component).Although agricultural inputs were distributed to farmers through BIMAS, there was also a parallel distribution system under which farmers significantly to the could pay cash for inputs from private traders. Indonesian rice production dou- bled between 1970 and 1985, the year that the country first became self-suf- rice intensification ficient in rice-and the year that BIMAS ended. Indonesia's "rice seed-fertilizer revolution" resulted from a complex array effort of policies, implementation strategies, and interactions among government agencies and the private sector.The rapid adoption of high-yielding rice tech- nology was made possible by an appropriate and effectively implemented pric- ing policy; available and affordable inputs; rapid learning among rice farmers; improvements in infrastructure, especially irrigation and roads; the continuing development of new high-yielding, early-maturing rice varieties at the Inter- national Rice Research Institute (IRRI) in Los Banos, the Philippines, and pro- vision of these seeds to Indonesia as the varieties cultivated there became susceptible to rapidly evolving pests; and later, the development of a national integrated pest management program."9 Conspicuously absent from the list of components responsible for the success of the rice intensification program, how- ever, was its credit program. Kredit Mini and Kredit Midi loans Two of the many subsidized credit programs offered at the unit desas-Kred- it Mini and Kredit Midi-are of special interest. For these programs, which were directed at rural inhabitants engaged in nonfarm activities, BRI was permit- ted to design the products and to set the terms and conditions, except for the 184 The Microfinance Revolution: Lessons from Indonesia 12 percent annual effective interest rate set by the government. Beginning in 1974, Kredit Mim offered loans up to 200,000 rupiah ($201 at the time the program ended in 1983). 2' Kredit Midi, begun in 1981 and intended primar- ily for Kredit Mim "graduates," provided loans up to 500,000 rupiah ($503 in 1983 dollars); Kredit Midi also ended in 1983.21 There were two crucial differences between the program designs of Kred- it Mini and BIMAS. First, Kredit Mini provided loans in cash rather than in input packets, as in BIMAS. Second, BRI selected the loan recipients for Kred- it Mini, whereas BIMAS borrowers were selected by government officials.These were the primary reasons for the third main difference between the two pro- grams: by the 1983/84 wet season the default rate for BIMAS was 54.5 per- cent, while for Kredit Mini it was 2.3 percent in 1983 (table 11.1).22 The default ratefor Kredit Mini and Kredit Midi were much in demand because borrowers want- ed cash loans so that they could make their own decisions about how they would BIMAS was 54.5 use the credit.They repaid their loans primarily because they wanted to retain the option of reborrowing. In addition, because BRI was able to select Kred- percent, whilefor Table l lFeatures of Kredit Mini it was Tabe 1.1 BIMAS and Kredit Mini 2.3 percent Feature BIMAS Kredit Mini Years of operation 1970-85 1974-83 Target group Farmers, especially rice People engaged in rural farmers nonfarm activities Who selected borrowers? Committees of BRI government officials How were loans disbursed? In kind (fertilizer, seeds, In cash insecticides), with a small cash component Nominal annual interest rate 12 12 (percent) Loan size Predetermined by Up to 200,000 rupiah amount of land to be ($201 in 1983) cultivated Loans more than 90 days 55 2 overdue at end of program (percent) Source DPIS 1983, BRI data How to Fail in Financing the Poor Bank Rakyat Indonesia's Unit Desa System, 1970-83 185 it Mim and Kredit Midi borrowers without outside interference, it was able to develop effective criteria for establishing borrower creditworthiness and to main- tain a high repayment rate in both programs. But the loans, which were sub- sidized and therefore rationed, tended to reach better-off villagers, not the poor. Even though these loans were relatively small, they were attractive to wealth- ier rural borrowers because of their below-market interest rates and because they were provided in cash. At their peak in 1983, Kredit Mini had about $6.4 million and Kredit Midi about $4.7 million in outstanding loans. Between the two programs there were about 700,000 borrowers. Kredit Mini and Kredit Midi proved that there was widespread demand in The unit desas had vlllages for small cash loans for working capital, and demonstrated that if BRI staff could select their borrowers, they could admunister programs effectively an abysmal record in and maintain high repayment rates. The main drawback to the two programs was the government-imposed subsidy that de facto limited the funds available mobilizing deposits for lending and ensured that the loans reached primarily wealthier borrowers. BRI, however, incorporated its positive experience with Kredit Mini and during the pre- Kredit Midi in planning the KUPEDES credit instrument, which was not sim- ilarly hampered by government subsidies and regulations. 1984 period Deposit mobilization in the unit desa system, 1970-83 The unit desas had an abysmal record in mobilizing deposits during the pre- 1984 period. InJune 1983, at the time of the first major financial deregulation, there was only about $18 million in deposits in BRI's entire unit banking sys- tem-after a decade of offering deposit services nationwide. Low deposits in the unit desas resulted from multiple disincentives. First, BRI had no incentive to mobilize savings in the unit desas because government regulation required a negative interest rate spread.The lending in- terest rate was 12 percent, while the rate paid on most deposits was 15 per- cent. Since every deposit represented a loss to the bank, BRI made little effort to attract savers.As one depositor complained to me in 1983:"BRI gives beau- tiful calendars to its customers every year, but only to borrowers.As a saver, why should I not receive a calendar? Am I not also a customer?" Second, the unit desas had a special disincentive to collect savings.The units were required to deposit in the branches, without interest, the savings they col- lected. But the units had to pay the interest due to depositors. Third was the significant perception gap between the views about rural sav- ings held by many government and bank officials on the one hand, and by a large number of Indonesian villagers on the other. Officials tended to assume that most rural people had little or no savings and, in any case, distrusted banks. Among vlllagers, however, there was a high but largely unmet demand for appropriate deposit instruments. But government assumptions about low savings propensity in rural areas continued, despite growing evidence to the con- trary.23 Because bank officials assumed that there was little potential for the mo- bilization of rural savings, there was little incentive to make the attempt. 186 The Microfinance Revolution: Lessons from Indonesia Fourth, Bank Indonesia's national savings program (Tabungan Nasional, or TABANAS), the only savings instrument then available in the unit desas, was disliked by most vlllagers because it limited the number of withdrawals to two a month.The problem was not that they necessarily wanted to withdraw more than twice a month, but that their overriding concern was being able to ac- cess their savings at all times when the bank was open. The withdrawal limit was widely considered to represent an unacceptable level of risk. Finally, Bank Indonesia provided low-interest liquidity credits for the loans channeled through the unit desas.Thus there was no incentive for the units to undertake the effort required for successful savings mobilization. In 1982 I accompanied Bank Dagang Bali (BDB) mobile savings teams on their rounds to learn which of their methods could be adapted for BRI. I was The unit desa quite surprised when the BDB team entered a BRI unit desa. The BDB staff explained that it was payday for the unit desa employees, and that the BDB mo- system would either bile team came regularly on such days to collect BRI staff savings. I asked the unit desa staff why they saved in BDB (at 12 percent annual interest) instead have to be closed or of in BRI (at 15 percent).The unit desa manager replied,"Well, over there we can withdraw our money whenever we want. And BDB has better service!" fundamentally Losses in the unit desa system, 1970-83 restructured By 1983 the financial performance of the unit desas had become alarming.With large fixed costs and rapidly declinng BIMAS participation, the operational loss- es of the units more than quadrupled from about $3 millhon in 1976 to about $14 million in 1983.The annual administrative subsidy increased by six times, from about $4 million in 1976 to about $24 million in 1983. Cumulative loss- es were high, since between 1970 and 1983 the unit desa system reported a net loss every year except 1977.24 At the time ofthe financial deregulation inJune 1983, it was clear that the unit desa system would either have to be closed or fundamentally restructured.Arrears were high and losses for 1984 were projected at more than $30 million. At that time the government was (correctly) antici- pating a decline in the real value of oil revenues, and it was not prepared to con- tinue indefinitely its hughly subsidized, failing rural banking system. Why Did BIMAS Fail? Many reasons explain the failure of the BIMAS credit program for rice inten- sification, and these were interrelated in complex ways.25 The program, badly designed and poorly implemented, was based on insufficient knowledge of agri- culture, village social structure, and rural markets.The extensive experience avail- able from 75 years of BRI's banking predecessors, Fruin's handbook, and even the advice of the 1969Yogyakarta BIMAS pilot project were all ignored in the rush to fulfill government targets for credit to rice farmers. Such was the ef- fect of supply-leading finance theory when combined with an urgent need to increase agricultural production. BRI's organizational structure was inade- How to Fail in Financing the Poor Bank Rakyat Indonesia's Unit Desa System, 1970-83 187 quate for effective supervision of the units, lines of responsibility in BIMAS were confused and the program became politicized, accountability was lacking, and the subsidies engendered staff corruption. The specific problems discussed below represent the Indonesian variants of the widespread problems that have been associated with subsidized agricultural credit programs in many developing countries worldwide (see chapter 4). The program's goals were incompatible The BIMAS rice intensification program had two main goals.The first was to play a crucial role in the wider rice intensification program's goal of rapid in- crease in rice production.Accordingly, the BIMAS program emphasized loans Most ricefarmers for the cultivation of high-yielding rice varieties on well-irrigated lands. The input packets typically available through the credit program were those thought received BIA/AS to be suitable for such lands-which are valuable and so normally cultivated by wealthier villagers. inputs suitablefor The second BIMAS goal was to raise the incomes of poor rice farmers. It was assumed that without credit, such farmers would be unable to afford the in- well-irrigated lands, puts required for the new rice technology that made possible substantially high- er yields.The plan was that poor rice farmers would take BIMAS loans and cultivate but most ricefarmers high-yielding rice, mcreasing the country's rice stocks and improving the farm- ers' incomes. But the seeds and fertilizers provided under BIMAS tended to suit cultivated lands that the high-quality ricelands to which poorer farmers typically did not have access. were not well BIMAS credit was tied to input packets In the early 1970s BIMAS input packets helped introduce high-yielding seeds irrigated and synthetic fertilizers to Indonesian rice farmers. But the program was de- signed primarily for rice cultivation on irrigated lands, while nearly half the country's ricelands are located in other environments, such as rainfed lowlands, nonirrigated uplands, and swamplands. Aspects of the national program often proved unsuitable for these environments. Even irrigated ricelands have different micro-environments, so an input packet appropriate for some irrigated rice fields may not suit others. In theory, BIMAS recognized environmental diversity by providing a range of input packets suitable for different land conditions. In practice, however, only one or two BIMAS packets were available in most subdistricts, and these were usually the packets suitable for well-irrigated ricelands. After a few years of experimentation with BIMAS packets, farmers usual- ly learned which seed types, fertilizer amounts, and cultivation practices worked best on their lands. These often differed significantly from the packets and in- structions provided to them.26 However, the owners of most of the best-qual- ity ricelands (wealthier farmers) made use of the new rice technology, at least partly through BIMAS. Overall, most rice farmers received seeds and fertilizer packets suitable for well-irrigated lands, but most rice farmers cultivated lands that were not well irrigated. 188 The Microfinance Revolution: Lessons from Indonesia Other agencies selected BIMAS borrowers, but BRI had to collect the loans The Mimstry of Agriculture set national and regional targets for the numbers of farmers and hectares to be included in BIMAS during each cropping sea- son. Potential borrowers were identified by officers of the Department of Agriculture and other local officials. In general, this became a politicized process in which many rural elites were identified as BIMAS participants and, as such, captured much of the subsidies. In theory BRI could reject a partic- ular borrower, but in practice this proved difficult, often impossible. In some areas, however, BIMAS targets were set too high for the rice-grow- ing potential of the area. In these instances the pressure on government offi- cials to achieve the targets tended to result in borrowers being signed up for Following BIMAS regardless of whether they were creditworthy or their lands suitable for rice cultivation. Gonzalez- Vega's BIMAS did not reach many poor rice farmers "iron law of interest BIMAS was capriciously implemented. In some areas, poor rice farmers cul- tivating nonirrigated land were not allowed BIMAS credit. In other areas poor rate restriction," farmers whose lands were unsuitable, or who did not even cultivate rice, were instructed to participate in the credit program so that local BIMAS targets might most BIMAS credit be met. Such farmers often felt little need to repay unwanted loans. But as loan delnquents, they were ieligible for future credit subsidies.Thus their only op- went to larger portunity to build a good credit rating in the formal sector was lost. The government assumed that the subsidized loans would cover the first farmers two or three years' purchases of fertilizer and other inputs, and that afterward farmers would not need to borrow further because they could then buy in- puts with the extra income generated by their improved yields. According to government planning, such farmers would then become "graduates" of BIMAS. Following Gonzalez-Vega's "iron law of interest rate restriction" (see chap- ter 4), however, most BIMAS credit went to larger farmers.Wealthier villagers quickly perceived subsidized credits as desirable financing. Those who repaid tended to continue their participation in BIMAS-as did others who did not repay their loans but who were locally influential and so permitted to rebor- row without repaying their outstanding loans. Extensive field work by the De- velopment Program Implementation Studies (DPIS), Center for Policy and Implementation Studies (CPIS), and BRI from 1979 until the program ended in 1985 uncovered virtually no BIMAS graduates (Robinson and Snodgrass 1987).As one better-offJavanese farmer commented: Every year I borrow from BIMAS. BIMAS is very helpful in raising rice production and in supporting the household-for every one kilogram of fertilizer I get from BIMAS, I use half of it for my sawah [irrigated riceland].The other half I sell and use the money for my daily expenses. -DPIS 1983, p. 181 How to Fail in Financing the Poor Bank Rakyat Indonesia's Unit Desa System, 1970-83 189 By 1983 the channeling of expensive government subsidies to wealthier vil- lagers had become so apparent that the government banned from the program farmers who cultivated more than 1 hectare of irrigated riceland. But by that time nearly all BIMAS loans were in the hands of rural elites, so the govern- ment ban resulted in a program with few participants.The program had been depleted not by its graduates but by its dropouts (who found the inputs un- suitable for their lands) and by its rejects (who did not repay their loans-in many cases, as discussed later in this chapter, because of crop destruction caused by inputs in their BIMAS packets). Policies for loan forgiveness and rescheduling during crop failures Many unit staff were badly planned and encouraged corruption In cases of crop failure, Presidential Decree 2/1976 made provision for BIMAS collected bribes, loans to be rescheduled or forgiven, depending on the amount of loss. But since an official declaration of the loss suffered was required for each rice field, there curriedfavorsfrom was considerable scope for corruption.As discussed below, massive destruction of rice crops occurred, but the loan forgiveness and rescheduling provisions for elite customers, and BIMAS loans were poorly implemented. Some local governments and BRI branches instituted various regulations treated poorer that defacto allowed better-off farmers to receive loan forgiveness while pre- venting poor farmers from accessing the provisions of the presidential decree; borrowers with some of these are discussed later in this chapter. In addition, widespread farmer complaints indicated that in many cases affidavits certifying crop destruction disdain. They were given to those who paid bribes to the inspecting officials. Meanwhle poor- er farmers who had lost their crops and were unable to pay bribes were ex- rarely met a saver pected to pay their loans. As one farmer said, "The rich pay bribes, the poor pay loans." BRI did not have the organization, human resources, or motivation to manage unit desas effectively BRI unit staff were untrained, unmotivated, poorly supervised, and general- ly demoralized. As discussed in chapter 14, they were poorly paid, treated as inferior by the rest of BRI, and blamed for the losses of the unit desas-which were largely outside their control. In turn, many unit staff collected bribes from borrowers for the below-market loans they controlled, curried favors from elite customers, and treated poorer borrowers with disdain.They rarely met a saver. With the loan interest rate at 12 percent and the interest rate for most sav- ings at 15 percent, with staff responsible for collecting BIMAS loans in cash from borrowers they had not been allowed to select and who had borrowed mostly in kind, and with the unit desa system treated within BRI as a pariah, there was no institutional framework within which the umt staff could work effectively. Subsidies-for both the BIMAS portfolio and the program's ad- ministrative expenses-diminished the already low motivation of unit staff to locate and make loans to creditworthy borrowers, and to collect these loans. 190 The Microfinance Revolution: Lessons from Indonesia Typically, neither unit nor branch staff knew the areas in which they worked or understood the local markets they served. Some branches and units instituted and implemented local policies that contradicted and undermined the intent of national policy; examples are provided later in this chapter. The institutional reorganization of the unit desa system and the upgrad- ing of its staff that began in 1984 are discussed in chapter 14. These changes were essential for the development of the unit desas into a sustainable micro- finance system. Successful agricultural credit programs require successful agriculture-but insecticides supplied in the BIMAS input packet caused severe crop destruction Typically, neither Analysts sometimes evaluate agricultural credit programs as though the caus- es of the successes or failures are all in the financial design.These are, of course, unit nor branch staff important. But agricultural issues also play a crucial role in the success or fail- ure of agricultural loan programs; these are often insufficiently understood by knew the areas in those desigmng and implementing the credit programs. Continued, large-scale crop failures will invariably affect the performance of agricultural credit pro- which they worked grams. As will be demonstrated here, many of BIMAS's repayment problems were caused by preventable agricultural failures that were linked to the inputs or understood the provided in BIMAS packets. Farmers whose crops are wholly or mostly destroyed usually cannot repay local markets they the full amount of their credit on time. If this situation is not adequately pro- vided for (as it was not in Indonesia in the mid-1970s), and if crop destruction served occurs frequently (as it did in Indonesia then), credit programs structured around particular crops are certain to have low repayment. The example provided below was selected in part because it was so seri- ous for Indonesia at the time.Although the problem of insecticide-induced crop destruction has largely been overcome there, this example was also chosen be- cause it is still widely relevant to agricultural credit programs in other coun- tries today. It should go without saying that agricultural credit programs that supply inputs that destroy borrowers' crops are unhkely to succeed. But this has not been said enough. A series of rice crop failures in Indonesia's primary rice-growing areas- where the immediate cause was destruction by the brown planthopper (Nila- parvata lugens Stal-provide a striking example of the many factors that underlie successes and failures in agriculture, and in agricultural credit programs. A brief sununary of rice seed and insecticide use at that time is provided so that the strongly negative effect of certain types of insecticides on the BIMAS cred- it program can be fully understood. Insecticide-induced crop destruction. High-yielding rice seeds developed in the Phlippines by the International Rice Research Institute (IRRI) were first released to Indonesian farmers in 1967.These were followed shortly afterward by a number of other IRRI rice seeds. But the high yields obtained from these How to Fail in Financing the Poor Bank Rakyat Indonesia's Unit Desa System, 1970-83 191 rice varieties in the first few years could not be maintained. Because of the new cultivation methods that accompanied the introduction of high-yielding rice seeds, especially the use of certain insecticides, the new rice types rapidly be- came sensitive to pests and diseases previously controlled by traditional culti- vation methods no longer practiced. Scientists from a number of countries had found that the use of some types of insecticides results in the rapid emergence of new biotypes of insecticide- resistant brown planthopper, as well as in the death of the natural enemies of the brown planthopper (such as certain spiders, beetles, and dragonflies) .27 This research demonstrated that a rapid increase in brown planthopper populations, known as resurgence, is caused by the nonselective use of such insecticides. The brown In effect, insecticides actually cleared the way for the brown planthopper planthopper by destroying its natural enemies, allowmg it enormous scope to burst forth, expand, and multiply. Sublethal doses of outbreaks that insecticide appeared to stimulate female reproduction . . . Retrospective interpretation of the evidencefrom the 1970s, based on IRRI disrupted research findings, indicates quite clearly that the brown planthopper outbreaks that disrupted Indonesia' rice intensificationprogramfrom 1974 Indonesia ' rice to 1979 were insecticide-induced. -Fox 1991, p. 75, emphasis added28 intensification It is estimated that in each of the 1974/75 and 1975/76 cropping years, program were about 400,000 hectares of new and traditional rice varieties were destroyed by the brown planthopper on Java. In 1976/77 the number increased to more than induced by 450,000 hectares (Fox 1991, pp. 68-69).To put these figures in perspective, in 1974/75 and 1975/76 the number of hectares onJava destroyed by the brown insecticides provided planthopper each year represented about 13 percent of the area covered under BIMAS throughout all of Indonesia in both rainy and dry seasons. In 1976/77 by BIMAS the figure was about 18 percent. Brown planthopper attacks on rice occurred outside Java as well.The figures for hectares of rice destroyed would, of course, be higher if all of Indonesia were included. The brown planthopper, however, presents another danger, as it carries a virus that causes a disease known as "grassy stunt." Rice plants with grassy stunt do not form grains, and there is no harvest. (Indonesian farmers call this "fam- ily planning rice.") Grassy stunt often appears in the rice crop following one that has been heavily attacked by the brown planthopper. "Cumulatively for the period 1974 to 1977, it is estimated that rice yield losses due to brown plan- thopper and grassy stunt disease in Indonesia exceeded 3 million tons of paddy worth more than $500 million" (Fox 1991, p. 69). Intensive research on rice breeding at IRRI during the late 1970s and early 1980s resulted in the development of new rice varieties (IR-36 and others that were genetically similar) which were resistant to the brown planthopper blotypes of that period. During 1979-85 Indonesia's production increased 49 percent, from 17.8 million tons of milled rice to 26.5 tons. But the use of resurgence-causing 192 The Microfinance Revolution: Lessons from Indonesia insecticides continued. The more the outbreaks of brown planthopper infesta- tion increased, the more farmers were inclined to use these insecticides-espe- cially since the cost to farmers was low because the resurgence-causing insecticides were supplied in the input packets of the heavily subsidized BIMAS program. Populations of a new brown planthopper biotype that was resistant to the insecticides began to develop in the mid-1980s; this biotype induced large rice crop losses in Indonesia in 1986, especially on Java and Sumatra. As each new biotype of the brown planthopper developed, only the fewer and fewer rice varieties that were resistant to that biotype could be planted. The need to protect against crop devastation by planting seed varieties resistant to specific pests resulted in massive erosion of Indonesia's rice gene pool.29 The rapid loss of genetic diversity was not only an Indonesian prob- The need to protect lem, but one that occurred throughout tropical Asia as the IRRI high-yield- ing rice varieties displaced traditional rice types during the 1970s (Hargrove against crop and others 1979; Fox 1991). By the 1979/80 rainy season 67 percent of In- donesia's irrigated riceland was planted in IRRI varieties that were closely devastation by related genetically (Fox 1991, p. 66).This percentage increased in the 1980s because of the continued cross-breeding among a relatively few, closely re- planting seed lated rice varieties. The combination of increasing genetic uniformity and continued use of varieties resistant to resurgence-causing insecticides brought Indonesia to a critical threshold in the mid-1980s.The role of these insecticides in the crop destruction of the brown specific pests resulted planthopper and in the precipitous genetic decline of rice varieties had been known since the mid-1970s by Indonesian and international agricultural ex- in massive erosion of perts. But powerful international insecticide compames-and some Indonesian officials-had strong interests in maintaining the status quo, so BIMAS pack- Indonesia s rice gene ets continued to contain resurgence-causing insecticides. It took time for the information about the role of these insecticides in inducing rice crop destruc- pool tion to reach President Soeharto.When it did, he acted immediately. A new national pest management strategy. Faced in 1986 with a much larg- er potential outbreak than had previously been possible, the government re- sponded with a major new pest management policy, announced in Presidential Decree 3/1986.The decree: * Banned the use on rice of 57 varieties of insecticides and began a national policy of integrated pest management that replaced regular spraying of in- secticides with a variety of biological and cultural pest control mechanisms. * Implemented, with the help of the Food and Agriculture Organization (FAO), a massive training program for farmers. * Provided for limited spraying only in identifiable outbreak situations. * Reduced and then eliminated pesticide subsidies. While these steps were of fundamental importance for Indonesia's rice pro- duction, the problem has not been entirely resolved.The decree permrtted resur- How to Fail in Financing the Poor Bank Rakyat Indonesia's Unit Desa System, 1970-83 193 gence-causing insecticides to be used on secondary crops; when these are ro- tated with rice crops on the same lands, the rice crops can be affected. More- over, these insecticides are available in the market, and some farmers continue to use them on their rice crops as well as on secondary crops. The resurgence-causing insecticides supplied in BIMAS packets had a devastating effect on both the country's rice crop and the BIMAS credit pro- gram.The program suffered severe defaults, and farmers whose crops had been destroyed by the brown planthopper attacks and who had not obtained debt relief, were unwilling to continue taking BINiAS loans. This is an extreme, although not an isolated, example. Many agricultural issues-with both positive and negative effects on loan programs-must be con- In 1986 the sidered carefully in order to understand correctly, and to improve, the perfor- mance of rural credit programs in developing countries. government responded with a Examples of BIMAS in Four Rice-growing Environments major new pest This section analyzes the BIMAS credit program for rice farmers as it had de- veloped by the early 1980s in four Indonesian villages.30 Each village reflects management policy a different ecological environment, a different pattern of land and labor use, and a different potential for rice intensification.The many reasons that BIMAS failed-as well as the fundamental incompatibility between the program's dual goals of rapid increases in rice production and of poverty alleviation-are well illustrated in the dynamics and the results of program implementation in these villages. The four villages, located in the provinces of EastJava (village G),WestJava (village C), South Sulawesi (village R), and South Sumatra (village P), are not meant to be representative of the country, nor necessarily of their provinces and districts.They were selected to represent a broad range of environments, in order to examine the implementation of four different government programs.31 The discussion below refers to the four villages as they were in 1980-81, when field studies were carried out in each village and when each already had extensive experience with BIMAS. The information about BIMAS imple- mentation gathered from these villages was used to support the policy mem- orandums that recommended to the government that the unit desa system be changed from a channeling agent for subsidized credit to a commercial finan- cial intermediary. At the time of the study there was still relatively little institutional finance in the four villages; table 11.2 shows borrowing of all types in three of the vil- lages.32 There was considerable variation among the villages in terms of reported loans; 77 percent of household heads in the East Java village reported an out- standing loan, compared with 36 percent in the South Sulawesi village. In none of the three villages, however, did institutional credit account for more than 20 percent of reported loans. For all other borrowing, villagers depended on in- formal sources. 194 The Microfinance Revolution: Lessons from Indonesia Table 1 1.2 Household borrowing in three Indonesian villages by source of the largest loan, 1980-81 (percent except where otherwise noted) Village G Village C Village R Source of loan (East Java) (West Java) (South Sulawesi) Number of households in village 775 760 711 Household heads reporting outstanding loans 76.6 58.9 358 Institutional credit 17 9 20.1 14.0 BIMAS 6 6 16.5 2 7 Bank/other 11 3 36 11.3 Informal credit 82 1 79 9 86.0 Traders 16.5 26 9 2.1 Relatives 37 4 14.8 40 5 Other (nonrelatives) 28 2 38 2 43.4 Total 100 100 100 Note The table shows three of the four villages studied intensively under the Development Program Implementation Studies (DPIS) project carried out by HIID for Indonesia's Ministry of Finance, 1979-83 Comparable data on credit are not available for the fourth village, village P in South Sumatra However, tables 11 3 and 11 4 provide other data for all four villages Extensive interviews about BIMAS participation and a household survey were carried out in village C in 1979-80, and in the other three villages in 1980-81 Selected follow-up studies in these villages were carried out in the mid-1 980s by the Center for Policy and Implementation (CPIS), an Indonesian foundation that was the successor to DPIS Source DPIS 1983 As noted, BIMAS sought to play a crucial role in the wider rice intensifi- cation program's goal of rapid increase in rice production, and it sought to raise the incomes of poor farmers. The first BIMAS goal was realized in village G (East Java) and to some extent in village C (West Java). The second goal was reached only in village C, and only to a limited extent. But the rice intensifi- cation program, if not its BIMAS credit component, did help raise the incomes of the poor in village G because the multiple rice crops made possible by the new technology created additional employment. Neither BIMAS goal was met even remotely in the two vlllages offJava. Village G (East Java): BIMAS results in an ideal rice intensification environment At the time of the study village G, located in East Java, was experiencing rapid economuc growth and agricultural intensification and had well-developed in- frastructure and irrigation facilities. Thus conditions for intensified rice pro- duction were highly favorable for the cultivation of multiple rice crops using How to Fail in Financing the Poor Bank Rakyat Indonesia's Unit Desa System. 1970-83 195 high-yielding technology on irrigated lands. Consequently, the well-irrigated, high-quality sawah (irrigated riceland) in village G was so expensive that few villagers could afford to own or cultivate irrigated rice.33 With a population of 3,559 and a land area of 325 hectares, village G was the most densely settled of the villages (table 11.3). It also had the largest per- centage of vlllage land, almost two-thirds, in sawah. However, the 201 hectares of sawah in the village were owned by just 25 percent of the households. Only 34 percent of village G's households in 1980-81 farmed land at their own risk, and only 10 percent were engaged in full-time farming. Most village households were occupied in other economic activities, including small-scale trading, wage labor, and local industries such as construction, transportation, and food processing. Table 1 1.3 BIMAS participation in four Indonesian villages, 1980-81 Feature Village G Village C Village R Village P (East Java) (West Java) (South Sulawesi) (South Sumatra) Population 3,559 3,092 3,099 1,456 Number of households 775 760 711 246 Land area (hectares) 325 644 3,519 625 Sawah area (hectares)a 201 208 1,025 68 Household ownership of sawah None 577 295 450 117 (percentage of households) (75) (39) (63) (48) Less than 0 25 hectares 48 343 39 8 (percentage of households) (6) (45) (6) (3) 0 25 hectares or more 150 122 222 121 (percentage of households) (19) (16) (31) (49) Households cultivating rice in 1980/81 267 611 467 167 (percentage of households) (34) (80) (66) (68) Households selling rice (percentage of 176 190 32 21 households cultivating rice in 1980/81) (66) (31) (7) (13) Household participation in BIMAS Ever-participants (percentage of 226 349 91 63 1980/81 sawah cultivators) (85) (57) (19) (38) Ever-participants 226 349 91 63 (percentage of households) (29) (46) (13) (26) 1980-81 participants 78 59 72 3 (percentage of sawah cultivators) (29) (10) (15) (2) 1980-81 participants 78 59 72 3 (percentage of households) (10) (8) (10) (1) a Irrigated rice fields Source DPIS 1983 196 The Microfinance Revolution: Lessons from Indonesia The combination of good rainfall, excellent irrigation, available fertilizers, and high-yielding seeds resulted in substantial increases in rice production under the national rice intensification program.34 Village G represents almost precisely the environment for which the rice intensification program was designed, but most of its inhabitants were too poor to be rice farmers.Thus by 1980-81 the num- ber of village households that had participated in BIMAS in at least one season represented 85 percent of the vlflage's rice cultivators. But they represented only 29 percent of vlllage households-the wealthiest of village G's residents. Participation and repayment in the BIMAS rice program were excellent in village G from the program's inception there in 1969 through 1975. Cov- erage peaked in the 1975/76 rainy season, with 138 participants covering 180 hectares. But much of the village's rice crop was destroyed by brown planthopper Village G is an disease in the 1976/77 rainy season. A direct effect of the insecticides provid- ed to village G farmers in their BIMAS packets, this destruction greatly affected almost ideal the repayment of that season's loans and subsequent BIMAS participation.The brown planthopper had two kinds of effects. First there were the direct effects environment for rice of massive crop destruction, lost income, and farmers' inability to repay BIMAS loans. Second were severe indirect effects resulting from poor implementation intensffication, but of the presidential decree on BIMAS loan forgiveness and rescheduling in cases of crop destruction. most of its In implementing the presidential decree on BIMAS loan forgiveness and rescheduling for the 1976/77 rainy season, a district-level decision was made inhabitants were too to reduce by 55 percent the BIMAS debts of village G farmers whose crops were destroyed by the brown planthopper. However, the policy was not im- poor to be rice plemented as stated.The vlllage head (lurah) declared the fields of 32 partici- pants to have been affected by the 1976/77 brown planthopper attack. For those farmers farmers-a group that included the vlllage head and many village officials (pa- mong)-the district ruling was applied: their BIMAS debts were reduced be- cause of crop loss. However, 73 other BIMAS participants in village G whose rice crops were also destroyed by the brown planthopper attack were not covered by this de- cision. These farmers resented the implementation of the presidential decree because influential villagers and those able to pay bribes had their loan oblig- ations reduced, while other BIMAS borrowers who had lost their crops were instructed to repay in full.They refused to do so.According to BRI records four years later, no BIMAS debt incurred that season by these farmers had been re- paid-in August 1981 BRI records listed 73 borrowers from village G who still owed a total of $3,100 for the 1976/77 rainy season.35These farmers remained ineligible for new BIMAS loans and could not build their credit ratings at the unit desa, whereas the 32 farmers whose loans had been reduced remained el- igible to continue borrowing. BIMAS participation began to decline rapidly after the 1976/77 season. Among the 226 of village G's rice farmers who had ever participated in BIMAS, only 78 (35 percent) were still participating in 1980/81.After 1976/77 most BIMAS participants in village G were larger farmers.While the BIMAS How to Fail in Financing the Poor Bank Rakyat Indonesia's Unit Desa System, 1970-83 197 loan amount per hectare was set at $56, the average BIMAS loan in vlllage G in 1980/81 was $81.Thus the average loan was taken for cultivation of about 1.5 hectares of irrigated land. Only 23 percent of the participants that season borrowed for cultivation on less than 1 hectare.36 The largest BIMAS borrower was the village head, who borrowed about $280 in 1980/8 1; the second largest borrower was his father-in-law.The average BIMAS loan taken by village of- ficials was $130. Given the village's experience with the presidential decree for debt relief, many smallholders said they were afraid of being in debt to BIMAS.Thus vil- lage G's relatively few small rice farmers either never participated in BIMAS, dropped out of the program after the 1976/77 season, or became ineligible for In village G the BIMAS loans. Looking back at the rice intensification program, village G farmers say that government its primary benefit was the introduction of synthetic fertilizers. But the 200 kilo- grams of fertilizer per hectare provided in BIMAS packets was not considered succeeded in sufficient for village G conditions.Thus even BIMAS participants were regu- lar market purchasers of fertilizer.37 There was also an economnc incentive to increasing rice purchase fertilizer from private traders because it was slightly cheaper on the open market. output significantly, In village G the government succeeded in increasing rice output signifi- cantly, but it failed to reach low-income rice farmers with BIMAS subsidies. but BIMASfailed Overall, rice cultivation in village G was both highly intensified and highly re- stricted. However, some poor residents of village G benefited from the increased to reach low-income employment that became available because of the multiple cropping and high- er yields made possible by the rice intensification program. rice farmers Village C (West Java): Most households benefit from rice intensification, but not necessarily from BIMAS Located in West Java with a population of 3,092 in 1980, village C is a farm- ing community located at an altitude of 490 meters; most of its 644 hectares consists of dry land (tegalan) and forest (see table 11.3). Only one-third of the land was sawah; most of the sawah was rainfed rather than permanently irri- gated, although some was irrigated sufficiently to permit two rice crops a year. In village C the land, which was not of particularly high quality, was controlled by a large number of smallholders.Thus 91 percent of its 760 households were involved in farming activities and 80 percent cultivated rice in 1980-81.38 In village C, 61 percent of households owned sawah, compared with 25 per- cent in village G.Village C had many more small rice farmers than vlllage G: 45 percent of village C households owned less than 0.25 hectares of sawah, com- pared with 6 percent in vlllage G.While rice production was higher in village G, where high-yielding varieties of rice were cultivated under excellent con- ditions, BIMAS reached a much larger portion of the inhabitants of village C. Many village C households participated in this program in the early 1970s: 349 households (57 percent of the households of vlllage C rice farmers, and 46 percent of all households) were BIMAS ever-participants. Fertilizer use rapidly 198 The Microfinance Revolution: Lessons from Indonesia became widespread, but after one or two trials with high-yielding seeds, nearly all village C farmers reverted to cultivation of local rice varieties.Village C rice farmers said that because their land is hilly and most of their sawah rainfed, they found local rice varieties to be more suitable than high-yielding varieties. The BIMAS input packets distributed in village C, which contained 300 kilograms of urea and 100 kilograms of triple superphosphate (TSP) per hectare,39 were larger than the village G packets, which contained 200 kilo- grams of urea per hectare.Yet village G farmers cultivating high-yielding seeds on well-irrigated lands and many village C farmers cultivating local varieties on less irrigated lands both needed larger quantities of urea than were provid- ed in the BIMAS packet.40 Farmers in village C typically handled the packet problems by unpackaging the packets and customizing the inputs for their own "I dare not join fields. They used what they needed from the packets, purchased additional in- puts as required in cash (if they could afford to), and sold or stored any excess BIMAS again inputs supplied in the packets. During 1976-78, however-for four successive rice cultivation seasons- because maybe my village C was devastated by brown planthopper and rat attacks.The farmers had been instructed to plant high-yielding rice seeds that were thought to be re- sawah will again be sistant to the brown planthopper. Some farmers, especially those who cultivated fields within the contiguous block of the best sawah, did so. But the fields plant- attacked by the ed in hlgh-yielding rice varieties were destroyed by a new biotype of the brown planthopper and by rats, as were the fields of most farmers who had contin- brown planthopper" ued to plant local seeds. Under the 1976 presidential decree for BIMAS debt relief in cases of crop destruction, only sawah that had been cultivated with high-yielding seeds was eligible for loan forgiveness or rescheduling. Except for the farmers with the best sawah, all other village C farmers had planted local rice varieties.Thus the provision for debt relief did not apply to them.The few farmers with the best lands received partial loan forgiveness, while those who cultivated local rice va- rieties did not-even though in many cases their crops had been completely destroyed.Village C farmers said that given the conditions in their village, the high-yielding seeds provided to them were appropriate only for the best lands, so the effect of the presidential decree was to favor the better-off farmers and to ignore the crop destruction suffered by smaller farmers. By the 1980/81 season, only 17 percent of village C's 349 BIMAS ever- participants, and only 10 percent of households cultivating sawah that season, were still participating in the program, Massive crop destruction combined with the ineligibility of most village C rice farmers for debt relief caused BIMAS arrears to grow rapidly in this village. BRI and the government placed renewed emphasis on loan collection; the village head had to insist that loans be repaid by households that he knew would be unable to repay. Some of the farmers whose crops had been destroyed were unable to repay their BIMAS loans and were not allowed to take new loans. Others be- came frightened and sold other crops or household possessions in order to repay their BIMAS debts; many of these farmers decided never to participate in BIMAS How to Fail in Financing the Poor Bank Rakyat Indonesia's Unit Desa System, 1970-83 199 cent of households owned any sawah. Relative to villages G and C, village R had lower economic growth, fewer off-farm economuc activities, and fewer cred- itworthy enterprises. In the 1970s and early 1980s the district in which village R is located suf- fered a series of calamities because of pests and drought. There was major crop devastation from the brown planthopper in 1972-73. In 1976-77 there was se- vere drought, as there was again in 1980-1982; poor irrigation facilities mag- nified the effects of the droughts. In addition, there was far more land than could be cultivated under existing conditions; this resulted in severe crop destruction by rats that multiplied rapidly in the untended lands and created yet another risk that farmers had to assess when deciding on input application levels. Much of the sawah in village R was left uncultivated during 1980/81 be- Much of the cause of inadequate irrigation,4' droughts, pest attacks, and a labor shortage caused by the departure of villagers seeking employment elsewhere. Of the 1,025 hectares irrigated riceland in of sawah in the village, 688 hectares were cultivated that season, of which only 223 hectares were owned by village R residents.As noted, most village R farm- village R was left ers worked as sharecroppers for absentee landowners, usually under an arrange- ment in which the owner provided the inputs and the crop was shared equally. uncultivated because Problems of poor irrigation and natural calamities were further com- pounded by the fact that in this subdistrict, BRI's unit desas differentiated be- of inadequate tween the relatively few farmers who cultivated land under irrigation projects and the many more who cultivated irrigated rice on rainfed land. According irrigation, droughts, to the rules of the local BRI branch, the latter were ineligible for BIMAS loans. Instead they were provided loans that were identical to BIMAS loans in all but pest attacks, and a one respect: their loans did not contain the BIMAS loan rescheduling and for- giveness provisions for debt relief in cases of crop destruction. labor shortage For these reasons, BIMAS participation was low in village R. In sharp con- trast to the two Java villages, only 91 households (13 percent of village R house- holds and 19 percent of sawah cultivators) had ever participated in BIMAS by 1980/81.Wealthier villagers were much more likely to have participated than poorer villagers.42 All fields in village R and surrounding villages were officially declared de- stroyed in 1976 because of drought, and all BIMAS defaulters were cleared of debt repayment obligations. But this provision did not apply to the many farmers who cultivated irrigated rice on rainfed land and whose loans did not contain the provisions for debt relief in cases of crop destruction. Most were unable to repay their loans because their crops had been destroyed. For the many village R farmers cultivating irrigated rice on rainfed land, the choice was among a BRI loan without the BIMAS crop failure provision, a private loan in kind at the prevailing rate of 62 percent for six months, pur- chase in cash of whatever inputs they could afford, or rice cultivation without synthetic fertlizers and other recommended inputs. Risk aversion in village R was understandably very strong. Most village R farmers avoided BINiAS and purchased inputs in cash, and some farmers avoid- ed buying fertilizer altogether. Low BIMAS participation was also a result of How to Fail in Financing the Poor Bank Rakyat Indonesia's Unit Desa System, 1970-83 201 the widespread practice under which vllUage R residents sharecropped the lands of absentee landowners, who were generally unable to supervise input use well. Given the risks of rice cultivation in village R and their limited shares of re- turns, owners lacked incentives to provide sufficient funds for technically op- timal input use, while cultivators lacked incentives to provide the improved, labor-intensive cultivation practices advocated by extension agents. As one sharecropper said: The decision whether or not to use fertilizer rests with the owner of the fields. He is the one who provides fertilizer and insecticides, so if they are not made available, I will not use them. I do not The government have any money and do not want to risk a loss. Better to waste energy than to waste money-it could happen that I would not failed in both goals be able to repay the loan. -DPIS 1983, p. 193 in village R: rice In contrast to the two Java villages-where many BIMAS participants also output was not purchased some inputs in cash-BIMAS participation and cash purchase of in- puts were essentially viewed as alternative farming strategies in village R. increased, and Wealthier sawah owners, including absentee landowners, tended not to pur- chase inputs with cash.They took advantage of the cheap BIMAS credit, often BIAL4S credit did passing the input portion of the loan on to sharecroppers while keeping the cash portion for themselves. Some used the BIMAS inputs for their village R not reach low-income lands on lands they held outside the village. Village R farmers who cultivated their own smallholdings tended to pur- farmers chase small amounts of inputs with cash rather than to take the risks of the BRI loans that provided no debt relief in cases of crop failure. Use of this strategy increased during 1981 when an intensive government repayment drive (GER- TAK) began.The BIMAS loan collection campaign of 1981 accompanied a se- vere drought in village R; the GERTAK drive frightened and offended farmers who could not repay their BIMAS loans because their crops had been destroyed. Thus vlllage R provides an example of rice intensification in a village where knowledge and acceptance of the new technology were high, but where ecolo- gy, infrastructure, and patterns of land ownership and use impeded intensification. These problems were exacerbated by the decision of the local BRI branch not to provide farmers cultivating irrigated rice on rainfed land with the debt relief pro- visions accorded to farmers who had permanent sources of irrigation. In the end the government failed in both its goals in village R: rice output was not sigmfi- candy increased, and BIMAS credit did not reach low-income farmers. Village P (South Sumatra): A village where nearly everything that could have been done wrong in BIMAS implementation was done wrong With a population of 1,456 in 1980/81, village P had less than half the pop- ulation of each of the other three villages and a very dlfferent agricultural en- 202 The Microfinance Revolution: Lessons from Indonesia vironment (see table 11.3). Land in village P was plentiful (625 hectares), but labor was scarce.43 There was significant potential for rice intensification, but tree crops-especially coffee and rubber-predominated.Village P had only 60 hectares of locally irrigated sawah and eight hectares of rainfed sawah. In ad- dition, 30 hectares were under swidden agriculture (padi ladang) during 1980/81. The rest of the cultivated land was used for tree crops. The history of rice intensification in village P can be understood only in the wider context of its ecological environment.Although government officials tried to fulfill their BIMAS targets in village P, farmers did not want to partic- ipate.Village P farmers must balance the demands and opportumties of a com- plex agricultural environment, and BIMAS was not designed for that environment. The combination of crops that a village P farmer selects in a given year de- Village Pfarmers pends on three main factors: land ownership, the relative prices of the potential products (rice, coffee, rubber slabs), and the tradeoffs among the various demands balance the on household labor.There is not enough labor to cultivate intensified rice, main- tain coffee gardens located far from the village, and tap rubber trees daily. The demands and primary choice is between irrigated rice cultivation, with its heavy labor de- mands in certain seasons, and a cycle of upland farming involving nonirrigated opportunities of a dry rice cultivation in combinaton with coffee, rubber, and other tree crops such as durian44 and cloves. Lands are used variously in different years depending pri- complex agricultural marily on the relative prices of rice, coffee, and rubber. Thus rice production in village P was considerably influenced by large swings environment. in coffee prices. The price of robusta coffee beans in the local market varied from less than 100 rupiah to 2,000 rupiah a kilogram during the late 1970s and BIA/JS was not early 1980s.When coffee prices were high, farmers neglected and even aban- doned their sawah.When coffee prices were low, rice cultivation increased. Un- designedfor that like coffee and rubber, the main cash crops, rice was viewed primarily as a subsistence crop. Further, rice was seen as a commodity that could either be environment grown or purchased.When coffee prices were low, about a third of the rice con- sumed in village P was purchased from outside the village.When coffee prices were high, the amount of rice purchased for consumption increased. During 1980/81, 99 percent of the vlllage's households were engaged in agriculture m one way or another, and 52 percent were sawah owners. But most sawah owners were not motivated to intensify their rice cultivation because this is a comparatively labor-intensive activity, whle most of their mcome came from other crops. In addition, given the labor shortage in village P, some farmers point- ed to an explicit tradeoff between their children's education and intensified sawah cultivation.They said they preferred to send their children to school rather than to have them work on the sawah.45 BIMAS started in village P in 1970/71, but with largely unsatisfactory re- sults. Many farmers reported large percentages of empty rice husks from the first BIMAS harvest, an outcome later attributed to overapplication of urea be- cause too large an amount for the village P environment was supplied in the BIMAS packets. No one applied for BIMAS for the next two years. During 1970-80,63 farmers (38 percent of the vlllage's 167 rice cultivators and 26 per- How to Fail in Financing the Poor Bank Rakyat Indonesia's Unit Desa System. 1970-83 203 cent of its households) tried BIMAS. But only 13 of these (21 percent of the ever-participants) participated more than once. Limited participation occurred not because farmers were unaware of the benefits of rice intensification, but because BIMAS fertilizer inputs were not suited to village P soils and because the program was administered with little understanding of the vlllage's agricultural environment. In this region the gov- ernment's message that there was a need to increase awareness was correct- but it applied primarily to the implementers of the rice intensification program, not to the farmers. Labor-intensive wet rice cultivation would have been difficult to intensi- fy in village P under any circumstances. But the problems were greatly mag- The government's nified by the fact that for more than a decade the fertilizer supplied in BIMAS packets was seriously incorrect, even harmful, for rice cultivation under village message to increase P conditions.The farmers said that on their already nitrogen-rich fields, most of which do not receive sufficient water, application of the large amount of urea awareness was provided in the BIMAS packets resulted in substantially decreased yields. An agronomic team hired by the DPIS project to study the problems in correct-but it village P agreed that the BIMAS packets contained too much urea and too lit- tle TSP. The team identified four rice growing micro-environments in village applied to the P: riverain sawah, which is irrigated year-round but often floods in the rainy season; sawah irrigated by tributaries of the main river, but usually with insuf- implementers of the ficient water for a second rice crop; rainfed sawah; and nonirrigated upland paddy. The fertilizer dosages of the BIMAS packets, and in most cases also the rela- program, not to the tive proportions of fertilizer types, were incorrect for all four environments.The packets were finally changed in 1981/82 under an intensive drive to increase farmers BIMAS in the area-but the new packets doubled the amount of urea and in- creased the ratio of urea to TSP! Not surprisingly, most farmers preferred not to participate in BIMAS. BIMAS participation in village P peaked between 1974 and 1976, when 10-12 farmers signed up each year. During 1975/76 severe pest attacks dis- couraged farmers from further BIMAS participation. During 1976-78 partic- ipation was extremely low; fewer than 10 hectares were under BIMAS. By the 1980/81 wet season only three village P farmers (5 percent of ever-participants and 2 percent of all sawah cultivators that season) were participating in BIMAS.46 Although there were only three BIMAS participants in village P in the 1980/81 season, cumulative arrears from 40 loans since 1970-71 totaled 622,304 rupiah (about $966 in 1981 dollars) at that time.47 Village P farmers based their refusal to repay their loans primarily on unchanged yields result- ing from an unwanted program with inappropriate inputs.48 To address the problems of BIMAS in this area and to increase program par- ticipation, a quasi-military drive to increase BIMAS known as OperationWork for Prosperity (Operasi Karya Makmur, or OKM), was begun in the subdistrict in 1981/82. Farmers were told that even those with outstanding BIMAS debts would be permitted to participate, and that the input packet would be larger- and as it turned out, even more unsuitable. 204 The Microfinance Revolution: Lessons from Indonesia Each village head was made responsible for meeting the subdistrict OKM targets for BIMAS participants for his village.Village P's headman knew that many participants would use the fertilizer for their coffee, would not alter their sawah farming practices, and would not repay their loans.Yet he signed up his quota of participants, as he had no choice but to meet the target. Forty-eight village P farmers participated in OKM in 1981/82, compared with three BIMAS farmers in 1980/81.49The ordinary BIIMAS packet of inputs used between 1970/71 and 1980/81 contained 75 kilograms of urea and 50 kilo- grams ofTSP per hectare. The OKM orders, however, were for farmers in this subdistrict to receive packets consisting of 200 kilograms of urea and 100 kilo- grams ofTSP for the 1981/82 season.50 That year, as usual, village P farmers used much of their BIMAS urea on their coffee. In 1982/83 only two farmers signed Village Pfarmers up for the BIMAS loans, and fertlizer purchase returned to pre-OKM levels.5' Throughout the life of the program, BIMAS loans were available to any- based their refusal to one in village P, but few villagers wanted to participate. Farmers were occa- sionally reprimanded for nonrepayment of loans-but not usually to the point repay their loans on where they repaid. Farmers would sign up when the village head needed their names to reach the vlllage P BIMAS target. But since the inputs were not cal- unchanged yields ibrated to their needs and since they might not be cultivating wet rice that sea- son, BIMAS participation in village P did not carry with it perceived obligations resulting from an to intensify rice-or even to grow it!-or to repay the loans. The government failed to realize either of its goals in village P. BIMAS was unwanted program not restricted but it was unsuitable for the vlllage P environment, and few peo- ple wanted to participate. In village P, BIMAS neither intensified rice produc- with inappropriate tion nor raised the incomes of poor farmers. inputs Implications of BIMAS: Results in four villages The results of the BIMAS program for rice cultivation in these four villages provide important insights into the program and its components under differ- ent environmental conditions. Table 11.4 provides a summary evaluation of BIMAS in each village in terms of seven variables: whether rice production increased, whether poor farmers' incomes increased, the extent of BIMAS cov- erage, the appropriateness of inputs in BIMAS packets, implementation of the presidential decree for loan forgiveness and rescheduling in cases of crop de- struction, and credit repayment before and after 1976. Increase in rice production. In vlllage G excellent soil and water conditions interacted with the introduction of high-yielding rice technology to raise rice production substantially and to increase sales to the national rice stockpile. Rice production also increased in village C, though to a lesser extent and largely be- cause farmers purchased their inputs in cash. Production did not increase in the two villages offJava. Increase in incomes of poorerfarmers. The BIMAS program was, for dif- ferent reasons, restricted to the wealthier farmers in village G and village How to Fail in Financing the Poor Bank Rakyat Indonesia's Unit Desa System, 1970-83 205 Table 11 A Evaluation of BIMAS in four Indonesian villages Village G Village C Village R Village P Feature (East Java) (West Java) (South Sulawesi) (South Sumatra) Rice production increased ++ + - Poor farmers' income increased as a result of BIMAS program + (indirectly) + Coverage BIMAS widely available to villagers - + - + BIMAS ever used by more than one-third of village households - + - +a BIMAS ever used by more than half of sawah cultivators ++ + -- BIMAS used by more than 10 percent of village households in 1980-81 - - BIMAS used by more than 20 percent of sawah cultivators in 1980-81 + - Appropriateness of BIMAS input packets for village environmentb + - Implementation of regulations for loan rescheduling and forgiveness at times of crop loss -- -- -- Credit repayment. 80 percent or more before 1976 ++ + Credit repayment 80 percent or more after 1976 - - -- ++ Yes (many, much, or highly successful) + Yes (successful) - No (few, little, or indicating a problem area) -- No (very few, very little, or indicating a severe problem) a In 1981/82 because of the OKM drive, the program reached 43 households (compared with 3 in 1980/81) But most farmers used the inputs provided on coffee, not rice, and repayment was very poor The following year 2 households signed up for the program b Does not include the effect of the resurgence-causing insecticides provided in the BIMAS packets to all villages These insecticides caused massive crop damage during the 1970s in villages G, C, and R Source Adapted from DPIS 1983 R. Poor farmers in village R benefited from BIMAS not at all. In village G, however, while poor farmers did not benefit directly from BIMAS, they did benefit indirectly from the rice intensification program because of the additional employment created by double and triple cropping and in- creased rice yields. BIMAS in village C and village P was, again for differ- ent reasons, implemented in more socially and economically egalitarian ways than in the other two villages. But only in village C did poor farmers ben- efit directly from the program. In village P almost no one benefited from BIMAS because as implemented, it was inappropriate for the agricultural environment there. 206 The Microfinance Revolution: Lessons from Indonesia Extent of BIMAS coverage. While BIMAS participation was limited to the wealthier farmers in villages G and R, there was wide access to program par- ticipation in villages C and PYet only in village C was BIMAS widely used, and only until 1976. Ever-participation in BIMAS ranged from 46 percent of the households in village C to 13 percent of those in village R.The difficulty for villages C, P, and R was that BIMAS regulations were designed largely with village G conditions in mind. But in village G the good sawah was owned by relatively few households. In the other vlllages poorer farmers cultivated sawah that was primarily rainfed (as in village C), poorly irrigated (as in village P), or both (as in village R). BIMAS was not implemented in ways that suited these conditions. By the 1980/81 season, only 8-10 percent of the households in vil- lages G, C, and R participated in BIMAS, while village P had only 1 percent Smallerfarmers saw participation. their crops fail and Appropriateness of the BIMAS input packetsfor the village environment. The primary problem was that BIMAS packets contained resurgence-causing the decree insecticides that caused brown planthopper attacks on rice crops.Villages G, C, and R were all seriously affected by this problem; village P was less affected implemented to because of its much lower emphasis on rice cultivation. But in village P the difficulty was a BIMAS input packet that was inappropriate for local soil con- forgive loans for ditions, and that remained so for a decade.And when the packet was finally re- vised, it was even less appropriate for village P lands. largerfarmers, but Loan rescheduling andforgiveness at times of crop loss. Because of local BRI not for them branch decisions, corrupt implementation of the president's decree, or both, the loan rescheduling and forgiveness provisions in cases of rice crop failure were limited to the influential (village G); to those who cultivated high-yield- ing seeds, grown then only on the best-irrigated lands (village C); and to those who cultivated lands under irrigation projects (village R).These limitations caused widespread resentment among many smaller farmers who saw their crops fail and the presidential decree implemented in ways that reduced or forgave loans for larger farmers, but not for them. Credit repayment. Before 1976, BIMAS credit repayment was fairly good in the two Java villages and fair in vlllage R and village P. After 1976, howev- er, repayment was poor in all four villages.These results reflect the national trends as shown in figure 11.2. Problems with credit repayment were closely linked to the sharp decline in BIMAS participation after 1976.This occurred partly because many participants experienced severe crop destruction caused by the brown planthopper and other pests. Many of these borrowers, who could not repay their loans, were not eligible for BIMAS loan forgiveness or reschedul- ing. Being in default, they could not reborrow from BIMAS. Program partic- ipation also decreased because many borrowers who had lost their rice crops and struggled to repay their loans from other sources refused to borrow again from BIMAS. How to Fail in Financing the Poor Bank Rakyat Indonesia's Unit Desa System, 1970-83 207 Agriculture andfinance in thefour villages. The specific reasons for the fail- ure of the BIMAS credit program were different in each village, but the un- derlying reasons were similar (see figure 11.1): * Subsidized interest rates prevented institutional viability. * BRI was not allowed to select its own borrowers. * Loans were tied to predetermined packets of inputs that were often inap- propriate and sometimes destructive to the rice crop. * In many areas credit subsidies went to wealthier villagers. * In some areas borrowers were selected by government officials who need- ed to meet targets-even though the borrowers' lands were unsuitable for Cooperatives said the inputs provided, or even for rice cultivation. * In some years there were massive rice crop failures, often as a direct result farmers groups were of the insecticides that were distributed in BIMAS packets. * Government policy for loan rescheduling was badly planned and often cor- responsiblefor loan ruptly implemented. * Unit desa staff were poorly trained, poorly paid, unmotivated, and general- collection;farmers ly ignored or looked down on by the rest of BRI. said loan collection These examples demonstrate the complex ways in which agriculture and finance are interrelated in rural credit programs. They show how natural oc- was the cooperatives' currences harmful to agriculture can be greatly exacerbated by human deci- sions, and reveal how serious and costly the problems can become when the responsibility latter reinforce the former. They also underscore both the rationality of rural borrowers and the imprudence of subsidized rural credit programs. The End of the BIMAS Era In 1982 BRI began to consider closing the unit desas or transferring them to another government agency. In selected districts in Java and Ball, BRI began quietly moving BIMAS out of the unit desas to the village cooperatives (KUDs). In the 1982 dry season 352 cooperatives provided BIMAS loans for a total of 2.3 billion rupiah; in the 1982/83 rainy season 824 cooperatives ex- tended BIMAS credit of 19.4 billion rupiah.52 Although the KUDs selected to begin the transition were chosen because they were among the best-managed in the country, they generally had mini- mal facilities; a small, untrained, and poorly paid staff; low-level management; and very little banking, bookkeeping, and accounting expertise. In addition, BIMAS was financially unattractive from the cooperatives' point of view, be- cause the 3 percentage point spread between borrowing from BRI at 9 per- cent a year and relending to farmers at 12 percent was inadequate to cover the KUDs' operating costs. The cooperatives were instructed to issue BIMAS loans to farmers groups rather than to individual farmers. However, CPIS/BRI field work in five dis- 208 The Microfinance Revolution: Lessons from Indonesia tricts in three provinces53 in 1982 uncovered BIMAS farmers groups on paper, but no farmers groups actively functioning to channel BIMAS loans to farm- ers. The listed heads of the farmers groups said that they did not want to be involved because they might be held responsible for the repayment of the other group members.The cooperative officials said that the heads of the farmers groups were responsible for loan collection; the farmers said that loan collection was the cooperatives' responsibility.We did not find a single person who said that BIMAS loan collection was his or her responsibility. The West Java district where village C is located was one of the districts selected for BIMAS administration through the cooperatives. Observing this change in progress in March 1983 in village C, we found that the new approach did not work because: Most of the * The farmers groups were not real groups. "members" of the * The group leaders were given no compensation and did not want to assume responsibility for the loan repayment of group members. so-called BIMAS * The cooperatives' procedures for BIMAS loans were lengthy, expensive,54and cumbersome for the farmers. farmers groups had * No one in the village knew who was responsible for loan collection. * The cooperative was not trusted because of widespread allegations of past been deadfor more corruption and mismanagement. than 30 years! In villages in Central Java we found elaborate BIMAS farmers groups listed on paper. When we asked to meet some of the farmers, we were told that this would not be possible. On investigating further, it turned out that these lists, which had been prepared for the local cooperative (KUD) officials, had actually been copied from 60-year-old lists of village landowners. Most of the "members" of the so-called BIMAS farmers groups had been dead for more than 30 years! In late 1983 the government decided that cooperatives that collected less than 70 percent of their outstanding BIMAS credit would not be permitted to continue administering the program. Of the 74 cooperatives administering BIMAS loans in the three districts for which we collected information,55 only 5 were able to recover more than 70 percent of the BIMAS loans for the 1982/83 season.While the remaining 69 cooperatives were not permitted to adminis- ter BIMAS loans the following season, in most cases the program was not ef- fectively returned to the unit desas either. By 1983 the umt desa system, built to channel subsidized credit to rice farm- ers, was experiencing heavy fixed costs, increasing losses, and a low and decreasing level of activity. In contrast, Indonesia's highly successful rice intensification pro- gram had achieved remarkable results. Between 1970 (when Improved National BIMAS started), and 1983 (when the decision was made to end BIMAS), the production of unhusked paddy (gabah) increased 87 percent, from 17.8 million tons to 33.3 million tons. But BRI's unit desas had become a burden on the bank and on the gov- ernment. In its BIMAS phase the unit desa system was costly to the govern- How to Fail in Financing the Poor Bank Rakyat Indonesia's Unit Desa System, 1970-83 209 ment, which subsidized the overhead of the unit banks as well as the credit to borrowers.The system was unprofitable for BRI. Morale among unit staff was poor, and service to customers was negligible.And the unit desas were deemed an impediment to BRI and a hindrance to its "real" banking activities. As Ali Wardhana commented in the introduction about the pre-1984 period:"In brief, our approach to local finance was ineffective and unsustainable." Indonesia's rice intensification program was based on a strong national com- nmtment to attainmg rice self-sufficiency; its success was made possible by the government's continuing ability to identify and correct its mistakes, and to ad- just policies and implementation strategies to the realities of the country's nat- ural environments and human resources. When official channels of input But BIMAS gave distribution or output marketing proved inadequate, alternate channels were opened. When particular seed types or insecticides were found to be unsuit- birth to the unit able or dangerous, they were removed from the system. Similarly, when the cred- it program proved ineffective, it was replaced. In 1984 the BIMAS credit desa system-which program was phased out, officially ending in 1985. But BIMAS gave birth to the unit desa system-which survived and developed into an internationally developed into an renowned commercial nicrobanking system. internationally Notes renowned 1. This chapter draws heavily on the Development Program Implementation commercial Studies (DPIS) Report on Rice Intensification (DPIS 1983).The research on which that report was based was carried out primarly between 1979 and 1982 under a HIID project with Indonesia's Ministry of Finance. This was HIID's first large-scale inter- microbanking system disciplnary project; in addition to this author, its coordinators were Donald R. Snod- grass and the late Donald PWarwick. Under the DPIS, four Indonesian development projects (rice intensification, family planning, primary education, and a presidential pro- gram to provide grants for village development) were studied at the national, provin- cial, district, subdistrict, and vlllage levels in different parts of the country, from the perspectives of anthropology, economics, and sociology. (I served as coordinator of the DPIS study and report on rice intensification and the BIMAS credit program.) This chapter is based on the work of researchers from five Indonesian universities and on that of the HIID team.The five universities were: Umversitas Padjadjaran (West Java), Umversitas Kristen Satyawacana (Central Java), Universitas Airlangga (East Java), Uni- versitas Hasanuddin (South Sulawesi), and Universitas Srlwijaya (South Sumatra). Major contributions to the field work and analysis were made byJohn R. Bowen,James J. Fox, Ilyas Saad, L. Hudi Sartono, Donald R. Snodgrass, Bambang Soelaksono, Marzu- ki Usman, Saikhyu Usman, and the late Donald P.Warwick. 2. BIMAS, an acronym for Bimbingan Massal (Mass Guidance), was implement- ed through the unit desas from 1970-85 3.This account of BRI's early history is taken primarily from Schnut (1994); see also BRI (1985, 1995). See Schmit (1991) for a review of Indonesian rural banking before independence, for an historical account of BRI's development, and for an ex- cellent bibliography. See Schmit (1994) for the history of the AlgemeeneVolkscredi- etbank, Frun (1994 [1935]) for the Provisional Manualfor the Credit Business of the General Popular Bank, and Fruin (1999 [1933]) for the "History, Present Situation, and Prob- lems of the Village Credit System (1897-1932)." See Suharto (1985, 1988, 1996) and 210 The Microfinance Revolution: Lessons from Indonesia Steinwand (2001) for detailed analysis of Indonesian rural banking history. I am grate- ful to Leo Schnut, Klaas Kuiper, and Dirk Steinwand for personal communications that helped me better understand the early history of the banks that merged in 1950 to form the modern BRI. It has been possible here only to provide a brief sketch of this extremely interesting period; for detailed analysis, see Schmit (1991, 1994). 4.There are several versions of this story, differing in minor details (see Steinwand 2001). 5. See Steinwand (2001) for a history of the European credit union movement and for extensive bibliographic references. 6. Boeke's colonial theories were criticized by many, even in the 1920s. See Hig- gins (1955) and Sadli (1957). For an overview of the commentary and criticism of Boeke's theory of dualism, see Garnaut and McCawley (1980). 7. Klaas Kuiper, the translator and editor of Fruin's 1933 article, comments: "It is not easy to give the value of the Dutch guilder (NLG) for that period ... in 1932 [the monthly salary of a village bank clerk] was between 24 and 48 NLG per month . (Fruin 1999, p. 65). 8. Fruin's manual was translated into English and reissued in 1994 by the Nether- lands Ministry of Foreign Affairs in honor of the 100th birthday of BRI in 1995. 9. Fruin's 1933 article was translated and edited by Klaas Kuiper and published by the Netherlands Ministry of Foreign Affairs in 1999. 10. See Schriit (1991) for discussion of the development of BRI in 1946-70, a period that is not covered here. 11. Indonesia's leading economust, Professor Sumitro Djojohadikoesoemo, played a crucial role in the formation of BRI as a state-owned bank in 1950. 12 Known as Operation for Self-sufficiency in Rice (Operasi Pelaksanaan Swa Sem- bada Beras), the plan was based on a projected network of service centers for rice farm- ers called padi sentra. 13. An exception was a pilot project known as Action Research carried out on 105 hectares by students from the Bogor Agricultural Institute in 1963-64.The pro- ject, which involved a more active approach to farmers, was quite successful, with re- ported yields rising by an average of 50 percent. In 1964-65 this approach was extended to 11,000 hectares.The program, known as Demonstrasi Massal Swa Sem- bada Bahan Makanan (DEMAS), was also successful, with reported yields of 7.3 tons of stalk paddy per hectare compared with a national average of 2.7 tons. In 1965-66 the program, then called Proyek Bimbingan Massal and known as BIMAS (an early version), was expanded to 150,000 hectares. 14. In 1967-69, the New Order government began a number of new rice inten- sification programs in addition to BIMAS Gotong Royong. These included BIMAS Gogo Rancah, a program for rainfed lands; BIMAS Berdikari, in which provincial gov- ernments or farmers were supposed to finance input purchases; and INMAS, a pro- gram in which farmers purchased inputs for cash. But, in general, during this period of high inflation and low foreign reserves, it was difficult obtain inputs, yields were well below expectations, and credit repayment was poor. 15. I am much indebted to Klaas Kulper for providing me with important first- hand information and unpublished data about theYogyakarta pilot project and the early BIMAS program. 16.This program is referred to henceforth simply as BIMAS. 17. DPIS field observations found a range from 3 to 83 villages per unit bank, de- pending largely on the area's population density. 18. In 1988 BRI conducted a survey of KUPEDES borrowers in 16 unit desas in eight districts in four provinces.There were two surveys: a larger one with 1,404 re- spondents and a smaller, more detailed one of 192 respondents.The smaller survey in- How to Fail in Financing the Poor Bank Rakyat Indonesia's Unit Desa System. 1970-83 211 cluded riceland ownership. It was found that 48 percent of the respondents' house- holds owned no riceland and 25 percent owned plots up to 2,000 square meters.The average amount of irrigated land owned was 1,620 square meters; the average amount of dryland owned was 1,512 square meters (Sutoro and Haryanto 1990, pp. 5, 19). 19. See Mears (1981) for the classic study of the early development of Indonesia's new rice technology; see Fox (1991) for analysis of the different phases of the coun- try's rice intensification program. 20.The design of Kredit Mim was partly based on the Badan Kredit Kecamatan (BKKs), the rural credit program of the Central Java provincial government, which had begun in 1970 (see chapter 9). During the pre-1984 period, however, the BKKs were allowed to charge substantially higher interest rates than the unit desas. 21. Kredct Midi was similar in design and implementation to Kredit Mim, except that it provided larger loans. Kredit Midi lasted only a few years because, like Kredit Mini, it was ended with the introduction of KUPEDES in 1984. 22. Kredit Midi operated only for a few years; its long-term loss ratio at the end of the program in 1983 was 3.8 percent. Some Kredit Mini and Kredit Midi defaults were associated with occasional curtailment of lending by BRI at specific times and places; borrowers who thought they might not be able to reborrow had less incentive to repay. 23. For example, a 1968 study of saving behavior in 490 Yogyakarta households concluded: "For the total sample . . . the marginal propensity to save is approximate- ly 10 percent, which compares favorably with marginal savings rates derived for households in advanced nations" (Kelley and Wlliamson 1968, p. 390).Yogyakarta is not representative of Indonesia; nevertheless, some insight into the country's saving potential could have been gained had attention been paid to these and subsequent find- ings.The World Bank's 1983 study of rural credit in Indonesia found that rural sav- ings propensity was 20 percent; this result was largely ignored or disbelieved within Indonesia at the time. 24. In 1977 the unit desa system made a profit of about $0.8 million; however, the administrative subsidy for the system that year was about $3.3 mhllion, which was not adjusted for in calculating the profit.Thus the system made a loss that year, as well as all the other years of the BIMAS program. 25. Parts of this section are summarized and adapted from Robinson and Snod- grass (1987). Further details are provided there and in DPIS Report 2 (1983). 26. In many areas the standard "Packet A" was the only packet normally available. In 1980-81 Packet A consisted of 100 kilograms of urea, 50 kilograms of triple su- perphosphate (TSP), 50 kilograms of KC1/K20, 2 liters of insecticide, 100 grams of rat poison, seed, sprayer fee, and 10,000 rupiah ($16) in cash for "cost of living" ex- penses. 27.The great strength of the brown planthopper is its ablity to breed and to spread very quickly. The planthopper eggs, lodged inside rice stalks, are sheltered from in- secticide spraying. When the larvae emerge they suck the juices of the rice plant and can destroy a healthy rice crop in one or two days. IRRI researchers have identified more than 100 parasites or predators that prey on the brown planthopper. But when these natural enemues are destroyed by insecticides, the planthopper, which is extremely mobile, can spread very rapidly (see Fox 1991). See Shepard, Barrion, and Litsinger (1987) for discussion of the natural enemies of the brown planthopper. 28. See also Oka (1979); IRRI (1979, 1984); Hargrove and others (1979, 1985); Bernsten, Siwi, and Beachell (1982); Hemnrichs and Mochida (1984);Kenmore and oth- ers (1984); and Hargrove (1988). 29.The 1983 DPIS Report on Indonesia's rice intensification program called at- tention to the alarmung decrease in the Indonesian rice gene pool, and recommend- 212 The Microfinance Revolution: Lessons from Indonesia ed immedcate further study.The recommendation was accepted by the Indonesian Mi- istry of Finance.The ensuing study carried out by CPIS and successive HIID adviso- ry projects on rice seeds and insecticide use led to analysis of the use of resurgence-causing pesticides as a major cause of pest attacks in Indonesia and brought this danger to the attention of the Indonesian government Wolfgang Linser andJamesJ. Fox led the HIID work on rice seeds and insecticides and advised the government on its development of a national integrated pest management program for Indonesia. 30. Drawn from DPIS (1983), this section is based on the work of many people. The discussions of villages G, R, and P are based on the work ofJames J. Fox (village G) and John R. Bowen (villages R and P) and the DPIS field teams they coordinat- ed. I served as coordinator for village C and for the project as a whole. As part of the DPIS study, an agronomic report on each of the four vlllages was prepared by facul- ty from agricultural umversities in the four areas, under the coordination of Donald R. Snodgrass. The analysis presented here was a cooperative effort in understanding these four very different-and yet very Indonesian-villages, and it could not have been done without the skills and insights of all those who worked in these vlllages on the DPIS project. 31.At the request of the Ministry of Finance, the complete 1979-83 DPIS study covered BIMAS and rice intensification, family planning, primary school construc- tion (INPRES Sekolah Dasar), and a national village grant program (INPRES Desa). 32. In village P, the fourth village studied, field work started later than in the other three vlllages. As a result comparable information on village borrowing is not avail- able. However, there was little institutional finance used in thus village in 1980-81. 33. Sawah refers to all irrigated rice fields, whether the water source is from irri- gation facilities or from rainfall. The quality of irrigation varies considerably; village G has high-quality irrigation faclitles. 34. By 1982 average yields had increased from about 27 quintals of paddy (gabah) per hectare before the intensification program to about 40 quintals.The primary cause of increased production in village G was the double- and triple-cropping made pos- sible by the shorter growing time of the high-yielding seeds. However, the risks caused by the shrinking rice gene pool had been rapidly increasing; by 1983 more than 90 percent of the rice cultivated in village G was of one type: IR-36. 35.This was more than the combined total of all other outstandmg village G BIMAS loans from 1975-80, as of August 1981. 36. However, the extent to which loans were shared in order to meet the 0.5 hectare requirement for a BIMAS loan is not known, nor is the degree to which land hold- ings were fictitiously enlarged to raise the loan size. Nevertheless, it was widely ac- knowledged that most village G BIMAS participants in 1980/81 were larger farmers. 37.According to the DPIS sample survey in 1980-81,97 percent of BIMAS ever- participants had bought fertilizer with cash at least once, although 73 percent of this group did so only after 1975. 38. Rice intensification began very early in the West Java district where village C is located.The first program (Padi Sentra) was implemented there in 1959-62; other early programs were tried there in the mnd-1960s. Improved National BIMAS began in village C in 1970-71. 39.The agrononuc team from Fakultas Pertanian (Department ofAgriculture) at UNPAD (Bandung) examined the six classes of sawah in village C. They con- cluded that the inputs provided in BIMAS packets were suitable for lands in class- es 1-3 (75 hectares), which have semi-technical and local irrigation. However, the paddy lands in classes 4-6 (167 hectares), most of which are rainfed and less fer- tile, require more fertilizer than was contained in the BIMAS packets provided to village C farmers. How to Fail in Financing the Poor Bank Rakyat Indonesia's Unit Desa System, 1970-83 213 40.The requirement for BIMAS participation in village G was ownership of 0.5 hectare of sawah; in village C the requirement was the purchase of 25 kilograms of urea.Thus, given the BIMAS requirement of 200 kilograms per hectare for village G, a farmer had to borrow at least 100 kilograms of urea for 0.5 hectare. In contrast, a vlllage C farmer was required to borrow at least 25 kilograms of urea, which at the BIMAS rate for vlllage C of 300 kilograms per hectare, supplied only 0.08 hectare. Even the vlllage C fertilizer packets contained too little urea for the conditions of most village C sawah. 41.Village R consists of two separate and quite different neighborhoods (kampung) that were formed into one village in 1967.The sawah of kampungT is primarily rain- fed, whlile kampung M is irrigated partly by a small vlllage dam and partly by the main subdistrict system. 42.As part of the DPIS survey, village R households were classified into four eco- nomic groups. Looking at the population of rice-cultivating households by econom- ic group, it was found that of a total of 25 households in the wealthiest group, 16 (64 percent) were BIMAS ever-participants. Only 6 (11 percent) of the 54 households in the lowest economic group had participated in the program.The two highest groups combined totaled 98 households; of these, 38 (39 percent) were ever-participants, while in the lowest two groups 53 of 360 households (15 percent) were ever-participants. 43.Three of the four villages (all but village C) suffered labor shortages: in vlllage G because of the new employment opportunities created by the double and triple crop- ping made possible by the rice intensification program, and by off-farm employment; in village R because natural and humanmade disasters caused emugration; and in vil- lage P because of the traditionally low ratio of population to land. But the labor short- age was greatest in vlllage P. 44. A highly-valued fruit found in Indonesia and other parts of Southeast Asia. 45. This is a good example of how, in a particular environment, two government programs can come into conflict-in this case rice intensification and the presiden- tial instruction to increase primary schooling. 46.Those farmers who did take BIMAS loans often did so not to intensify their sawah cultivation but to obtain fertilizer for coffee gardens. This use of BIMAS had the tacit approval of BRI officials.As one village P farmer commented: "In 1979/80, I took BIMAS for the first time. I asked to be put down at the BRI as a participant, but admitted frankly that I was not going to cultivate sawah at all that year but planned to use the fertilizer for coffee plants and dry rice (padi ladang). I pronused that I would pay after the ladang harvest. This request was approved by the BRI officials and the loan has been repaid" (DPIS 1983, p. 210)." Some of the inputs received were used for coffee, some for padi ladang, and some were sold or given to others. 47. Because there were no subdistrict or vlllage records available on nondefault- ing participants by village, it was not possible to learn the percentage of BIMAS par- ticipants in vlllage P who were defaulters. Based on the DPIS sample survey, however, the number of defaulters probably represented about 50-60 percent of the total num- ber of participants. Defaults were consistently underreported by officials in this area. Thus the DPIS survey showed a considerably higher loan default rate for village P than the defaults reported for the subdistrict by the local office of the Department ofAgri- culture (25 percent during 1975-80), or for the smaller subdivision (marga) to which vlllage P belongs-which was reported to be the marga with the lowest default rate in the subdistrict (18 percent during 1975-80). 48. Another cause of BIMAS delinquency was the overdelivery of inputs, an out- come that resulted from officials encouraging farmers to report their sawah holdings in numbers that were rounded up in order to show that area targets were reached.The following example was reported from a village near village P: "In 1974-1975, there 214 The Microfinance Revolution: Lessons from Indonesia were five farmers from Village L who took BIMAS loans. Although they each culti- vated about one-half hectare of sawah, they were urged by the village headman to re- port one hectare each They were each given the one-hectare BIMAS packet. Three of the participants had lower yields than usual; two had the same yields, all because of overuse of fertilizer" (DPIS 1983, p. 227). 49. Seventy-seven vlllage P farmers (who cultivated 62 percent of the village sawah) were signed up to take BIMAS loans in late 1981 under the OKM drive. But twen- ty-nine were called so late to collect their mputs that they had already sown their seeds; all these farmers refused to participate. 50. In addition to the problems with the fertilizer supplied in the BIMAS pack- ets, there were problems with the seeds provided. The IR-36 seeds delivered in the 1982 packets were thought by farmers to be of poor quality, and in fact only half of the seeds later sprouted. According to the extension officer, this problem was known long before seed delivery because the seeds had been tested earlier at the Seed Cen- ter (Balai Bemh) and only half had sprouted there.Village P farmers replaced the nonus- able seeds they had received under BIMAS with IR-32 seeds they purchased from farmers in another village. 51.The leader of sawah activities in the village, the mantri sihing, was supposed to be the liaison between the government's agricultural extension program and the vil- lage P farmers. However, he said he was not willing to act in this role: "I have to look after my family first; only then can I think about others." He himself had not partic- ipated in BIMAS since 1973 and did not follow the advice of the extension agent "be- cause it is too much work." 52.These data were supplied by BRI. Figures from the Ministry of Cooperatives show somewhat more KUDs participating. 53. Sumedang mWestJava, Klaten and Pekalongan in CentralJava, Kediri and Malang in East Java. 54.The KUD was located further away from village C than the unit desa, so farm- ers had considerably higher transportation expenses than in the past. 55. Klaten and Pekalongan in Central Java and Malang in East Java. How to Fail in Financing the Poor Bank Rakyat Indonesia's Unit Desa System. 1970-83 215 Q Success in Microlending: The KUPEDES Credit Program, 1984-96 In 1984 Bank Rakyat Indonesia's (BRI's) unit desas began the transition that transformed them from a loss-making chan- neling agent for government-subsidized BIMAS credit for rice farmers and other subsidized rural credit programs (see chapter 11) to the world's largest, most profitable mi- crobanking system. Kredit Umum Pedesaan (KUPEDES, or general rural credit), made possible by the 1983 financial deregulation, was introduced nationwide through the unit desas in 1984. As the unit desas' commercial microcredit program that provides loans to all creditworthy applicants for all productive purposes, KUPEDES played a leading role in driving both the transformation of the unit desa sys- tem and its subsequent achievements of wide outreach and high profitability. 216 The unit desa system was profitable by 1986 and has operated independent of subsidy since 1987 (Yaron and Benjamin 1997, p. 43). By 1996 the KUPEDES credit program had 4.1 trillion rupi- ah ($1.7 billion) in outstanding loans to 2.5 million borrowers, with a long-term loss ratio of 2.2 percent. And the unit desas had a pretax return on assets of 5.7 percent-substantially higher than typical banking industry averages in Indonesia and many other countries. This chapter considers the development of KUPEDES credit and its outreach, and of unit desa profits from 1984 through 1996. Remarkably, the KUPEDES record of wide coverage and high Between 1984 and repayment continued during Indonesia's severe financial, econom- ic, and political crisis that began in 1997, as did unit desa profitability. 2000, 48 trillion The strength of KUPEDES as a credit program, and as the en- gine of unit desa profitability, derives from its design, pricing, and funding, and from unit desa organization and management. Design disbursed in and pricing are discussed in this chapter, funding in chapter 13, and unit desa organization and management in chapter 14. Chapter 15 26 million exanines the unit desa system from 1997-2000 and explores the rea- sons for its remarkable stability-even as the country's financial sys- tem collapsed around it. KUPEDES provides individual loans to economically active poor and lower-middle-class people throughout Indonesia. The basic principle on which the credit program is based is that institu- tional sustainability is required for large-scale outreach to low-in- come borrowers. Accordingly, the KUPEDES interest rate is set to cover all unit desa costs and risks and to return a profit to the sys- tem. Between 1984 and 2000, 48 trillion rupiah was disbursed in 26 million KUPEDES loans. Unlike BIMAS, KUPEDES borrowers are selected by BRI with- out any outside intervention, and the bank bears the full credit risk. Loan decisions, made by well-trained, experienced unit desa staff, are based on evaluations of each borrower's character (willingness to repay) and the viability and cash flows of his or her enterprise (ability to repay). KUPEDES is designed specifically to meet the needs of low-income borrowers for convenient bank locations, simple loan procedures, quick processes, and flexible terms.Within its overall reg- ulations, KUPEDES loan maturities and repayment plans are cus- tomized for each borrower's needs. Borrowers who repay their Success in Microlending The KUPEDES Credit Program, 1984-96 217 loans promptly and whose enterprises remain creditworthy are per- mitted to reborrow for the gradually increasing loan sizes for which they qualify, up to the KUPEDES maximum (they can then bor- row from BRI's branches, which provide larger loans).The bank does not train KUPEDES borrowers or attempt to monitor their enter- prises or the use of their loans; the premise is that borrowers know their businesses better than do bank staff. Since 1984 each unit desa has been treated as a financial unit or "profit center" with its own balance sheet and profit and loss statement- Each unit desa is a change that was essential for the effective implementation of KU- PEDES and the new savings program discussed in chapter 13. Unit treated as a 'proft desa regulations, products, reports, and job descriptions, are simple and center" with its own standardized across all units. Unit managers and staff are given substantial responsibility and held accountable for their performance.They are pro- balance sheet and vided intensive, high-quality training and financial incentives based on the performance of their unit. Simple, transparent loan accounting and profit and loss reporting enable frequent assessment of each unit's portfolio quality and immediate followup in case of problems. Finally, unit staff are reg- statement ularly supervised by well-trained supervisors from the unit's supervis- ing branch. These supervisors also receive incentives based on the performance of the units they supervise. Repayment has been high for a number of reasons, among them that: * Borrowers are carefully selected based on their ability to use loans productively and their ability and willingness to repay them. * Loans are provided in gradually increasing amounts based on the borrower's repayment record and the creditworthiness of the en- terprise. * Borrowers want to repay in order to keep their options open to borrow again on what are considered attractive terms at reason- able cost. * Unit staff treat their clients with respect and courtesy. This chapter and the two that follow are about the transforma- tion of the unit desa system from BRI's ugly duckling, wanted by no one, into the goose that lays its golden eggs. 218 The Microfinance Revolution: Lessons from Indonesia During its 15 years of operation (1970-84) BIMAS disbursed 2.7 million more loans than KUPEDES did during its first 17 years of operation (1984-2000). But BIMAS, funded by government subsidies, disbursed only 1 percent of the credit disbursed by KUPEDES, which is funded by locally mobilized savings (table 12.1). By 2000 KUPEDES had 7.8 trillion rupiah ($816 million) in 2.7 million outstanding loans. In addition, BIMAS was noted for its high defaults, while KUPEDES is characterized by high repayment rates.With BIMAS, the unit desas made high losses; with KUPEDES they have made continuing high profits. Indonesia's Enabling Conditions BIMAS,funded by Enabling conditions in the Indonesian environment were crucial for the emer- government gence of large-scale commercial microfinance at BRI, beginning m 1984. First, the country had had good, consistent macroeconomic management and polit- subsidies, disbursed ical stability for nearly 20 years. Second, there was sustained, high-quality lead- ership among economic ministers-essential for the development of the financial only 1 percent of the system generally and of the proposed large-scale system of sustainable microfi- nance. Third, the 1983 financial deregulation-the first in a series of financial credit disbursed by and other economic reforms-opened the way for profitable microfinance at BRI. Finally, large-scale commercial institutional microfinance developed in In- KUPEDES, which donesia because of the confluence in the 1980s of six ideas: isfunded by savings * It was believed that macroeconomic considerations mandated that the pri- vate sector become responsible for a larger share of national savings and in- vestment. * Substantial investment of the country's oil wealth in rural areas in the 1970s had resulted in significant rural development and higher per capita income, creating widespread demand for banking services by the 1980s. * The serious deficiencies of many government-subsidized rural credit pro- grams had become well known to some high-level government officials, and the status quo was considered unacceptable. * The interest rates of informal commercial lenders had been studied, and it was thought that it might be possible for commercial institutions to provide small loans profitably at considerably lower cost to low-income borrowers. * The major contributions that microenterprises make to the economy in gen- eral and to employment in particular, and the extensive demand for micro- finance-for both loans and savings services-were beginning to be perceived by technocrats. * Indonesia already had rural financial institutions that had consistendy pro- vided profitable commercial microfinance-both credit and savings-in some cases for many decades (see chapters 9 and 10).This meant that there were ways to learn about local markets, techniques to screen borrowers, and methods to attract savers that had been successful in Indonesia.These lessons Success In Microlending The KUPEDES Credit Program, 1984-96 219 Table 12.1 A comparison of BIMAS and KUPEDES Feature BIMAS KUPEDES Area and years of operation Nationwide, 1 970-85a Nationwide, 1 984-present Source of finance Bank Indonesia Unit desa savingsb Credit risk borne by Bank Indonesia 25 percent, Ministry of BRI: 100 percent Finance 50 percent, BRI 25 percent Program goals and borrower targets To increase rice production and provide To provide commercial microloans subsidized credit to poor rice farmers for all productive purposes to all creditworthy borrowers How well did the program achieve its goals? Not well Well Selection of borrowers Committees of government officials BRI unit desa staff and, for larger loans, branch supervisors Disbursementlrepayment of loans Primarily in kind/in cash In cash/in cash Nominal monthly effective interest rate 1.0 percent 2 8 percentc Loan maturity One to two rice harvests (changes Up to three years, depending on made over life of program) type of loan Payment schedule One payment at end of loan Choices available, monthly installments most popular Incentives for repayment Reborrowing at a below-market rate Reborrowing from a loan program suited to borrowers' needs, prompt payment incentive Number of loans (cumulative) 28 8 million (1970-85) 26 1 million (1984-2000) Value of loans (cumulative) 636 7 billion rupiah (1970-85) 47,775 8 billion rupiah (1984-2000) Number of outstanding loans at end of last year available 563,023 (1 983/84)d 2,715,609 (2000) Value of outstanding loans at end of 23 billion rupiah ($21 9 million, 7,827 billion rupiah ($816 million, last year available 1983/84 wet season) 2000) Overduese Default rate-51 percent dry season Long-term loss ratio-1 9 percent (1983); 55 percent wet season (2000) (1983/84) Share of profitable units 0 percent (1983) 98 percent (2000) Unit desa profit or loss -12 6 billion rupiah (-$12 7 million, 1983) 1,160 billion rupiah ($121 million, 2000) Note For BIMAS, the table covers only the credit program for rice cultivation a BIMAS officially ended in 1985, but it effectively ended in 1984 with the start of KUPEDES In 1984-85 BIMAS focused on collecting old loans b Startup funds for the KUPEDES loan portfolio were provided by the government for 1984-85 c For prompt payers Rates vary somewhat by loan maturity and payment plan d Includes 1983 dry season loans and 1983/84 wet season loans This was the last year that BIMAS was active in lending e. BIMAS and KUPEDES measure overdues differently, and the available data are different and not strictly comparable The default rate for BIMAS shown here is the percentage of the amount disbursed that was in default at the end of 1983 IBut BRI continued to collect BIMAS loans after the program ended, and by 1990 the default rate was 16 5 percent ) For KUPEDES the long-term loss ratio is defined as the cumulative KUPEDES amount due but unpaid since each unit opened relative to the total amount due Source BRI data, Patten and Rosengard 1991, pp 63-65 220 The Microfinance Revolution: Lessons from Indonesia would now be integrated into the planning and implementation of large- scale financial intermediation among low- and lower-middle-income peo- ple, and combined with massive organizational reform to bring about Indonesia's microfinance revolution. Bank Rakyat Indonesia's New Approach to Microlending at the Unit Desas When the KUPEDES loan program began in 1984, the unit desas were un- profitable, run-down, disorganized, and demoralized. It had taken the work of many people from 1979-83-and considerable funds from the Ministry KUPEDES of Finance to understand local microfinance demand, to prepare for the re- structuring of the unit desas, and to design the new loan product.This process implementation has sometimes been misunderstood. For example, it has been incorrectly stat- ed that in June 1983 the minister of finance asked an advisory group from required the Harvard Institute for International Development (HIID) to assess the un- profitable unit desas and that "the team came up with an experimental so- fundamental lution within a few weeks that eventually became the KUPEDES system" (Hulme and Mosley 1996, vol. 2, p. 40). In fact, the minister of finance asked changes in unit desa us for advice on the unit desas in 1979, and the recommendations for KU- PEDES were completed in 1983-so the process took a few years, not a few organization, weeks! Many institutions and more than a hundred people were involved in the work. management, Institutions and governments considering the restructuring of large rural banking systems should not be misled into thinking that a few weeks, or even operations, and a few months, will be sufficient for the careful planning of such an undertak- ing. This is a major effort, one requiring leadership, funds, and extensive work human resources at multiple levels. Of course, today the unit desas and other financial institu- tions offer many lessons that were not available in the early 1980s, and the process need not take as long as it did then. Nevertheless, the effort required to plan the restructuring of a large rural banking system should not be underestimat- ed. And planning is just the beginning. Implementing KUPEDES The introduction of KUPEDES was not simply a matter of offering a new loan product; its implementation required fundamental changes in umt desa orga- nization, management, operations, and human resources. The structure of the umts had to be changed from branch windows to individual profit centers. Unit staff had to be given more responsibility, provided the authority to carry out new tasks, trained in the skills needed to make good performance possible, of- fered the incentives to achieve high productivity, and held accountable for their performance. Analogous steps had to be taken at branch and regional offices to enable the effective supervision of unit activities, and at the head office to develop Success in Microlending The KUPEDES Credit Program, 1984-96 221 appropriate information management, coordinate planning, and maintain ef- fective control over the system. New borrower selection processes and loan collection methods had to be developed, as well as new accounting, report- ing, and information management systems. In addition, microbanking train- ing centers for unit desa staff and supervisors had to be established and the trainers trained. All this occurred; the institutional changes are discussed in chapter 14. Nei- ther KUPEDES credit nor umt desa savings should be seen simply as new prod- ucts. Crucial differences between the old and new cultures m the unit desa system made possible the extraordinary performance of KUPEDES since 1984 (fig- ure 12.1; see also the figure in the introduction to volume 2). KUPEDES dif- In 1996 the fered from BIMAS in almost every way (see table 12.1). Moreover, even though the unit desas' new institutional framework was essential for the implementa- subsidized loans had tion of KUPEDES, the BIMAS credit product would have failed even with the new unit desa organization. A good product is a necessary, though insufficient, an average balance condition for a profitable, large-scale microcredit program. 20 times that of Microcredit: subsidized and commercial The introduction of KUPEDES as a commercial nucrocredit program in 1984 KUPEDES-and and the official end of BIMAS in 1985 did not mean the end of subsidized cred- it at BRI.The old tension between social and commercial approaches to rural an arrears rate 10 banking (discussed in chapters 9 and 11) continued. KUPEDES was based on the idea that commercial microcredit is the only way to meet large-scale nm- times as high crocredit demand from low-income people on a continuing basis. But this ap- proach was not yet widely understood. In addition, because subsidized rural credit programs tend to reach local elites, it is often politically difficult to terminate such subsidies.Thus, as KUPEDES began in the unit desas, subsidized agricultural credit programs financed from the government budget were continued at BRI-but in the bank's branches, not in the unit desas. These subsidized programs were, of course, rationed. It was assumed that since KUPEDES, funded by savings, would not need to be rationed, it would gain a large market share. This is, in fact, what happened. A comparison of BRI's subsidized loans at the branches and its KUPEDES loans at the unit desas in 1996 shows dramatic differences. Only 16,670 sub- sidized loans through the branches were outstanding, less than 1 percent of the 2.5 million KUPEDES loans outstanding at the units (see figure 12.1).And at $1.7 billion, KUPEDES loans at the units had more than seven times the value of the subsidized loans at the branches. The subsidized loans do not reach the poor. At the end of 1996 the aver- age outstanding balance of the subsidized loans at the branches was $14,275- more than 20 times the average balance of the nonsubsidized KUPEDES loans ($684) at the unit desas. Measured in both cases by portfolho status (defined as aggregate overdue principal installments relative to total principal outstanding), KUPEDES arrears were 3.6 percent, while arrears for the subsidized loans were more than 10 times higher (39.3 percent). 222 The Microfinance Revolution: Lessons from Indonesia Figure 12.1 The evolution of microcredit at Bank Rakyat Indonesia, 1970-96 1970-83 1984-96 * Indonesia had a substantial oil surplus * Oil revenues declined * Extensive donor funds available * Donor funds decreased * Government provided subsidies for loan * BRI permitted to set its own interest rates External interest rates for unit desa operations on most loans and deposits * Central bank provided loan funds and set * Government-subsidized loans reduced environment interest rates and regulations for all loans substantially in scope and amount, and * Unit desas and government-sponsored channeled through BRI branches, not cooperatives introduced through the unit desas * Interest rates deregulated in 1983 Branches did not provide microcredit during Activities this period Government continued to provide some high-priority subsidized credit schemes for production of rice, secondary crops, tree crops, and sugarcane These programs were channeled through BRI's branches Microcredit Results 11996) at the * Number of outstanding loans 16,670 branches . Value of outstanding loans $0 24 billion * Average outstanding loan balance $14,275 * Arrears as a share of out- standing loans (portfolio status)a 39 3% * Net interest revenue $14.5 million Activities Activities * Large-scale, supply-leading approach to * KUPEDES introduced in 1984 rural credit * No subsidized loans * Implementation of more than 350 * KUPEDES accompanied by major unit government-subsidized credit programs desa reorganization for rice, other food crops, cattle, fisheries, * Active savings mobilization Microcredit tree crops, poultry, small business, and the like Results (1996) at the unit * Little effort to mobilize savings * Number of outstanding loans 2.5 million desas * Value of outstanding loans $1.7 billion Results * Average outstanding loan balance $684 * Loans often reached rural elites instead of * Arrears as a share of outstanding intended borrowers loans (portfolio status)a 3 6% * Severe repayment problems * Net interest revenue $269 8 million * Very low savings * Unit desa system profitable since 1986 * Continuing and increasing losses for BRI * All loans funded by locally-mobilized and government savings a Portfolio status measures aggregate overdue principal installments relative to total principal outstanding Source Adapted from an unpublished diagram by Sugianto Success in Microlending The KUPEDES Credit Program, 1984-96 223 In addition, the net interest revenue received by BRI from KUPEDES in 1996 was $270 million, more than 18 times the net interest revenue received from all the subsidized programs in the branches (see figure 12.1). Overall, the unit desas with their KUPEDES loans and savings mobilization program (see chapter 13) have been efficient and profitable since 1986, while BRI's subsi- dized programs have continued throughout to generate high arrears and large losses.' Perception Gaps among Policymakers, Bureaucrats, and Villagers Net interest revenue The bold strategy for developing rural banking nationwide through BRI, as outlined by Indonesia's economic planners in 1983, was based on previous pol- from KUPEDES icy decisions that placed high priority on the development of the country's rural areas. By the time of the financial deregulation in June 1983, rural areas were in 1996 was more generally, if not uniformly, experiencing economic growth and increasing monetization, communication, education, and demand for financial interme- than 18 times that diation (chapter 9).2 In the early 1980s the Indonesian government, anticipating a sigmficant de- receivedfrom all cline in the real value of oil revenues, began to undertake wide-ranging financial reforms and to seek increased private savings and investment. In these cir- subsidized programs cumstances long-term government funding for a nationwide rural credit pro- gram could not be obligated. Moreover, a large and continuing commitment in the branches of central bank funds for the unit desas was considered too risky given the track record of rural credit programs with high defaults in Indonesia and elsewhere. But Indonesia's economics ministers recognized the potential for a large- scale rural banking system and supported its introduction with government fund- ing. President Soeharto supported the development of BRI's unit desa system-an endorsement that was important for the success of the transition and the sys- tem's later development. As part of the wider financial reforms, BRI was per- nutted to set its own interest rates in the unit desas. But the bank was informed that after the startup period it would be on its own, with no continuing gov- ernment financial support for the unit desas. The idea behind the restructuring of the BRI umt desa system was to meet local credit demand with a loan instrument that would be suitable for all types of productive loans, charging interest rates that would enable long-term istitu- tional viability. And KUPEDES would be financed by locally mobilized savings. The views of high-level policymakers, however, were far removed from those of many bureaucrats-both inside BRI and in other government agencies- whose understanding and participation would be critical for implementing the new system. The perception was widespread among BRI and other govern- ment officials that most villagers were low-income farmers, uneducated, un- changing, and generally incapable of, or averse to, using banks. And these officials pointed to the 1970-83 record of BRI's unit banking system-high arrears, high losses, and low savings-as proof of their conclusions. 224 The Microfinance Revolution: Lessons from Indonesia But the views of many villagers were similar to those of the policymakers. Of course, the planners and bureaucrats were not monoliths, and rural inhab- itants did not hold a single view on banking. Nevertheless, opinions were ex- tensively shared within each category, while there were profound differences between bureaucrats with views of rural stagnation, and vlllagers and policy- makers who perceived the country's rural growth and increased demand for financial services.The 1983 deregulation of interest rates and the mounting em- phasis on domestic funds mobilization opened a new era in Indonesian rural banking. What made it a reality, however, was the extraordinary achievement in narrowing the perception gap between BRI officials and other government bureaucrats on the one hand and villagers and policymakers on the other. There were profound Policymakers The minister of finance and other members of the government's economics team differences between thought that a commercial microbanking approach might be able to meet what seemed to be extensive rural demand for appropriate savings facilities. Savings bureaucrats with could fund a large and growing commercial microloan portfolio, and profits from the system could build the long-term viability of the unit banking system. If ac- views of rural complished, these results would contribute to the country's economic devel- opment. It was also hoped that BRI's unit desas would serve as a model of stagnation, and commercial financial intermediation for low-income clients that would even- tually be adopted by other Indonesian banks and financial institutions. villagers and But this perspective required farsighted vision. Given the losses of the unit desas by 1983, as well as the negative attitudes of many middle- and lower-level policymakers who government officials and bankers toward rural banking, it required an excep- tionally broad view of the country's economy and society to recognize Indonesia's perceived the potential for establishing a profitable system of financial intermediation at the local level. But the government's economncs team now held this view, and BRI's country' rural newly appointed (in August 1983) president-director soon became a strong ad- vocate of commercial mncrofinance.3 However, implementing the new micro- growth finance program was difficult, and its success was far from assured when BRI's unit desa system opened its doors to commercial microfinance in 1984. The policymakers knew that a fundamental change in bureaucratic un- derstanding would be required for the shift to conunercial financial interme- diation. On the lending side the critical lessons to be learned were that subsidized credit is unavailable to most people and that informal commercial lenders typically charge very high interest rates to poor borrowers.Thus avail- able loans with interest rates that would enable institutional profitability should be in high demand-as turned out to be the case.The lessons required for suc- cessful rural savings mobilization were, first, that many people preferred to hold substantial portions of their savings in banks-if the banks could provide se- curity, convenient locations, confidentiality, a range of savings instruments of- fering liquidity and positive real returns in different combinations, and access to loans.And second, collecting savings from the general public would enable the units to serve small savers profitably. Success in Microlending The KUPEDES Credit Program, 1984-96 225 For profitable administration of a restructured unit desa system, the primary lessons that had to be learned were: * The spread between interest rates on loans and the cost of funds would have to be set to enable unit profitability, and each unit desa would have to be- come a financial unit, a profit center. * Products would have to be few in number and carefully designed to meet demand. * A management information system would be needed to provide simple, trans- parent, accurate, and timely information about the performance of each unit. * Where necessary, bank units would have to be moved from their locations Myopic vision near rice fields to local market centers. * Staff would need to be retrained. among officials was . A culture of accountability would have to be established. * Incentives would be needed to motivate unit employees to locate and serve not confined to BRI a large number of customers-who were now to be called (and treated as) clients, not beneficiaries. * Units would have to be regularly and carefully supervised by well-trained staff from each unit's supervising branch. In brief, the decision to restructure BRI's unit banking system required a complete change in the philosophy and operations of the system. Systems and operating procedures had to be designed to provide simple, transparent incen- tives to all parties-clients, staff, supervisors, and managers-for efficient op- erations. And, because of the heavy fixed costs and the losses of the old system, the fundamentals of the new system would have to be in place within months. The high-level leadership invested in the transformation of the unit desas was essential, and included the coordinating minister for economics and finance, the finance minister, and other technocrats; BRI's new President-Director Ka- mardy Arief; and Sugianto, who became responsible for the umt desa system at the time of the transition (and was appointed a BRI managing director in 1986, in charge of BRI's microbanking activities; Robinson 1998b).Without such leadership and commitment, the unit desas could not have been transformed to a commercial microbanking system. Bureaucrats As noted, however, many government officials and others held widespread, deeply rooted misconceptions about rural markets, about microfinance demand, and about villagers' attitudes. The gap in perceptions is well illustrated in a paper presented at a 1987 seminar at Bank Indonesia: The relationship between formal credit-providing institutions and their clients is based on an economic relationship ... [However, the] Indonesian village economy is underdeveloped and [such relationships] are still unsuitable in villages . .. [Our 226 The Microfinance Revolution: Lessons from Indonesia villagers] are not able to manage in a money economy ... [and they] have psychological obstacles to saving in banks. This view represents a common holdover from the theories of a dual economy-an imported capitalistic system and an indigenous, primitive,"East- ern" social system-held by J. H. Boeke and his followers (see chapter 11). Boeke's colomal theories had been criticized even in the 1920s and 1930s, and they were widely discounted in academic circles in Indonesia and internationally by the 1950s. But the concept of economic dualism stayed with many government officials for decades afterward. Many bureaucrats (and some academics) raised on these ideas were often blinded to realities in rural Indonesia, ignoring the rapid rural development of the 1970s and 1980s. Changing the Until 1984 neither BRI unit desa staff nor their supervisors in branch of- fices had been expected to act as bankers. Beginning in 1983, extensive field perceptions and the work by BRI and Center for Policy and Implementation Studies (CPIS) re- search teams in villages, unit desas, and branches found widespread ignorance ethos of BRI's staff of village activities among local BRI staff. Unit desa staff and their supervisors generally understood neither the new policies they were required to imple- was the bank's most ment nor the productive activities and financial flows of the rural areas they were supposed to serve. dfficult task-and Unit desa staff rarely visited the villages in the service areas of their units. Believing that the villagers were primitive and their needs limited, the staff typ- its most spectacular ically treated their customers with disdain. As one BRI supervisor said to me in 1985,"Why have banks for villagers? Our villagers are still uneducated.They achievement think loans are gifts from the government.They will not use banks for savings because they spend everything they earn just to live. Anyway, even if they had money, they would not keep it in a bank. They do not trust banks, and they keep any savings in animals, gold, or land."All too often, bureaucrats with per- ceptions like these were given the job of implementing the 1983 financial re- forms and of building BRI's local banking system. Myopic vision among officials was not confined to BRI. Officers of vari- ous departments (agriculture, industry, cooperatives) charged with rural devel- opment, officers of local governments, and personnel from other banks shared many of the same nusunderstandings about productive enterprises, income flows, and financial markets in rural areas. There were exceptions-some government officials knew rural areas well, and some BRI staff were dynamic and innovative in their relationships with vlllagers. But innovations were rarely appreciated by supervisors. Changing the perceptions and the ethos of BRI's staff, a sine qua non of the bank's success, was its most difficult task-and its most spectacular achievement. The generally abysmal knowledge about village conditions held by BRI and other government bureaucrats gave rise to a number of incorrect assump- tions about villagers in general, and commercial banking in particular.A num- ber of such views, common in Indonesia in the 1970s and early 1980s,5 have been reported in many other developing countries as well: Success in Microlending The KUPEDES Credit Program, 1984-96 227 * "Villager" and "farmer" are synonymous.' * Villagers-who are generally poor, ignorant, and backward-are not "bank- mmnded." * Vlllagers cannot afford commercial interest rates for loans. * Farmers need continual instruction about the inputs to be used on their fields and cannot be trusted with cash. Thus borrowers should receive all or most of their agricultural loans in input packets. * Subsidized credit is required to provide low-cost loans to low-income farm- ers who can neither save nor afford commercial credit. * Villagers do not need financial savings instruments.They rarely have excess liquidity-and when they do, they usually spend it for consumption. Those Many villagers who save prefer nonfinancial forms. perceived clearly the These views, held as they were by legions of central, provincial, and local government officials and others, presented a formidable obstacle to the re- increasing demand structuring of the unit desas. They continued to be held by some people for many years, but by 1996 the financial deregulations, BRI's experience, and the forfinancial services increased use of local banking services generally had done much to dispel such misconceptions about rural banking and microfinance-which were reduced from low-income even further by the stunning performance of Indonesian microbanking dur- ing the recent crisis. people Villagers Unlike most government and bank officials, many villagers perceived clearly the increasing demand for financial services from low-income people. Rural moneylenders, traders, commodity wholesalers, and other informal lenders recogmzed the high demand for loans.As active participants in the economic growth then occurring in the villages, they understood that the rising demand for loans, especially for working capital, reflected the new opportumties becoming available for productive rural activities. Credit might, of course, be obtained from farmly, neighbors, friends, rotating savings and credit associations (ROSCAs, or arisan), savings and loan societies (sim- pan pinjam), village rice banks (lumbungpadi), and others. But these types of lend- ing usually did not meet the growing demand for credit that would be sufficient for ongoing production purposes. Informal commercial moneylenders of various kinds were widely present, but the cost of borrowing from them was typically very high, especially for poor borrowers-who paid monthly effective interest rates rang- ing from 10 percent to well over 100 percent (see table 6.1). Government-subsi- dized credit programs were widespread but unavailable to most low-income borrowers (or inappropriate for their needs, as discussed in chapter 11). Villagers understood what most bureaucrats did not: before the "seed-fer- tilizer revolution" of the 1970s, villagers borrowed primarily to finance con- sumption. They would borrow to feed their families in the pre-harvest or "hungry" (paceklik) season, at times of crop failure, for famnily emergencies, and to cover the costs of ceremonies such as circumcisions, weddings, and funerals. 228 The Microfinance Revolution: Lessons from Indonesia Beginning on a large scale in the early 1970s, however, rural inhabitants began to borrow for production expenses (Mears 1981). Starting with loans for inputs of new agricultural technology, demand quickly expanded for loans for productive off-farm activities, local industries, trade, and services.This was the potential de- mand for institutional commercial credit perceived by villagers, moneylenders, and economics ministers, but not by many bureaucrats and bankers. The perception gap was equally large on the savings side. Key policymak- ers, informed by extensive studies of villagers, believed that if appropriate fi- nancial savings instruments were made available in conveniently located banks, village households would convert some of their excess liquidity into financial savings. But most BRI staff and government officials interviewed in the early 1980s flatly denied the existence of such rural savings and the potential for con- "We had to change version of any rural savings that might exist into financial form. Yet discussions with villagers in many parts of the country indicated huge de- the culture of the mand for insttutional savings instruments and services geared to rural needs.Vil- lagers interviewed in the early 1980s often said that they kept more cash in the unit desa system, house than they wanted because of the lack of alternative options (see chapters 7 and 13).They did not convert excess cash into institutional financial savings pri- and we did." marnly because TABANAS, the only savings instrument available at the unit desas, limited the number of withdrawals. But beyond some needs for ready cash kept -Kamardy Arief; at home, households and enterprise owners throughout the countiy said that they wanted savings accounts from which they could withdraw on demand-in a con- BRI president- vemently located, secure bank. Some made use of conveniently located People's Credit Banks (BPRs; see chapter 9), but for most people in rural areas such banks director, 1983-92 were not available, security was lacking, or products were unsuitable. These perception gaps began to close during the 1980s as bureaucrats grad- ually came to see the reality that villagers had experienced and policymak- ers had recognized. BRI's unit desa system led the way. How the closing of the gap came about is best summarized in the words of Kamardy Arief, the BRI president-director: "We had to change the culture of the unit desa sys- tem, and we did." The Transition from BIMAS to KUPEDES Ceilings on interest rates for most state bank loans and deposits were repealed and liquidity credits to banks from Bank Indonesia were sharply reduced as part of the June 1983 financial deregulation.7 Banks were given more autonomy in decisionmaking and encouraged to expand their products and activities. State banks, instead of serving primarily as channeling agents for government sub- sidies, were to be transformed into commercial banks. In this context the question of whether to close or to restructure BRI's unit desas became a critical policy issue. BIMAS was visibly moribund, and unit desa losses for 1984 were projected to be more than $30 million. In 1983 the Har- vard Institute for International Development (HIID) submitted a report to the Success in Microlending The KUPEDES Credit Program, 1984-96 229 government, Rice Intensification: Development Program Implementation Studies Re- port 2 (DPIS 1983), that reviewed the findings of the research carried out be- tween 1979 and 1983 on rice intensification and the BIMAS credit program.8 This report (referred to hereafter as the DPIS report) recommended that BIMAS be ended and that a general rural credit program with commercial interest rates be implemented through BRI's unit banking system.Various government agen- cies and agricultural institutions also issued reports on the problems of BIMAS, and the government became quite familiar with the difficulties of the unit desas. The DPIS report also proposed a substantially increased effort to mobilize rural savings.This approach was favored because it could both provide savings services to the rural population and finance the new credit program. The re- The question of port recommended that the general rural credit program be: whether to close or . Accessible to all villagers. * Flexible enough to fit village patterns of repayment, either in full after a spe- to restructure BRls cific period or in installments over the course of the loan. * General enough to minimize procedural mechanisms, eventually streamlin- unit desas became a ing the administration of vlllage borrowing. * Viable enough to operate without continuing government subsidies (DPIS critical policy issue 1983, p. 447). The proposed program was described as follows: General rural credit should be available for many kinds of economic activity carried out by villagers, such as agriculture, trade, and small scale industry. Access to credit should be based on the individual borrower's creditworthiness, not on a particular project or operation. -DPIS 1983, p. 4479 The financial deregulation that allowed banks to set their own interest rates on most loans and deposits made it possible for the government to transform the unit desas from "rice banks" to "real banks." But should this transformation be attempted? On the one hand, it would be a high-risk strategy: no bank in the world had developed sustainable commercial microfinance on the scale of In- donesia's unit desa system. On the other hand, the minister of finance had funded four years of exten- sive research that had shown clearly that the rapid economic development in many rural areas had substantially increased the demand for credit.The government also recognized the low level of rural banking activity and the high interest rates charged to rural borrowers by informal commercial lenders. In addition, com- pellng evidence had been collected showing that many villagers would save in unit desas if savings instruments and services were appropriate for their needs. Moreover, closing the unit desas would end the provision of banking services in more than 90 percent of the locations where formal banking services were 230 The Microfinance Revolution: Lessons from Indonesia available. It would also put out of work about 18,000 employees (including both unit bank staff and night guards). It was clear from the beginning that restruc- turing the unit desas would be a high-risk, high-reward strategy. Policy issues Immediately following the financial deregulation inJune 1983, Minister of Fi- nance AliWardhana requested from HIID detailed policy recommendations on what would be needed to change the unit desa system into a commercial bank- mg network, as well as estimates of what it would cost. He asked the three ques- tions mentioned in his introduction to this book: * Would rural borrowers pay interest rates for loans that enable the unit desas The Indonesian to cover all costs? * Would rural people deposit savings in units, enabling loans to be capitalized government made by savings? * How should the new system be financed, and how long would it take until three decisions that the unit desas broke even? laid thefoundations The year before the deregulation a few of the HIID consultants to the DPIS project had formed a small, informal Rural CreditWorking Group to work on for sustainable policy issues that would result in specific recommendations for restructuring the unit desa system.10 In mid-1983, at the request of the minister of finance, microfinance. The the group responded to his questions and prepared recommendations for con- verting the unit desas into a system of commercial niicrofinance. Many detailed first was that the discussions on policy and implementation were held with the Ministry of Fi- nance and BRI in 1983; the Rural CreditWorking Group worked closely with unit desas would be both the ministry and BRI in preparing the recommendations. The nationwide evidence gathered by the DPIS project was used to an- retained swer the first two questions. First, low-income rural borrowers would pay the interest rates that BRI's unit desa system would need to charge because the only other credit options for most were to borrow from family, friends, and infor- mal savings and loan groups such as ROSCAs (whose loans were often not ap- propriate for use as working capital) or from informal moneylenders at very high cost, especially for poorer borrowers. Second, rural households would save in the unit desas if they were offered security, convenient locations, confidentiality, access to loans, and a choice of instruments with different ratios of liquidity and returns. For most low-income households, liquidity-not previously available at the unit desas-was the key. In emergencies these households wanted access to their savings at all times dur- ing bank hours. In answer to the third question, the Rural Credit Working Group recom- mended that the nominal interest rate on loans for working capital be set at a 1.5 percent a month flat rate on the original loan balance (equivalent for most loans to about a 2.8 percent a month effective interest rate on the declining outstanding loan balance)."1 Borrowers who did not repay on time would pay Success in Microlending The KUPEDES Credit Program, 1984-96 231 a penalty of 0.5 percent a month flat rate. It was suggested that interest charges be stated as a flat monthly rate because rural borrowers were accustomed to the flat rates typically charged by moneylenders. In addition, it was thought that stating a flat monthly interest rate of 1.5 percent, instead of the equivalent 32 percent annual effective interest rate (for prompt payers with a 12-month loan with monthly payments),'2 might lessen potential political opposition. The Rural CreditWorking Group calculated that these rates would enable the unit desa system to break even within two years-if credit outstanding reached $201 million and if loan losses were held to 4.5 percent of amounts disbursed. It was recommended that the new loan program be funded initially from the capital of the Kredit Mini program ($67mulhion) and from Bank Indonesia liq- The second decision uidity credits at 15 percent interest a year, the same rate that the unit desas paid depositors for savings. was to end The new savings mobilization program would eventually provide additional funds; it was assumed that the interest rate for savings would be 15 percent a subsidized lending year, the rate of the nationalTABANAS savings program.The Rural CreditWork- ing Group estimated that given these assumptions, the bank would be able to at the unit desas cover all the costs of the unit desa system, to break even within two years, and then to earn a profit. and to introduce Important influences on these recommendations were the lessons on rural finance learned from BRI's Kredit Mini and Kredit Midi programs (see chap- KUPEDES at ter 11); from Bank Dagang Bali, a private bank (chapter 10); and from some of the People's Credit Banks (BPRs; chapters 9 and 14). These institutions had commercial interest learned long before how to provide and recover small loans made to low-in- come borrowers. Except for BRI's Kredit Mini and Kredit Midi, whose sub- rates sidized interest rates were controlled by the government, these institutions had also learned how to lend to low-income borrowers profitably. As discussed later in this chapter, BRI's transition from BIMAS to its new general rural credit program-which came to be called Kredit Umum Pedesaan, or KUPEDES-owed an important debt to these institutions. Basic lessons were learned from them about microcredit-and from Bank Dagang Bali and some of the People's Credit Banks, about microsavings as well. Three crucial decisions In late 1983 the Indonesian government made three decisions that laid the foun- dations for the new approach to microfinance.The first was that BRI's unit desa network would be retained, despite the fact that it was no longer required for its original purpose of financing rice farmers through BIMAS. The unit desas would be restructured into a commercial banking network that would serve rural areas. Unmts would be moved, closed, or added as necessary to provide full microfinance coverage in rural areas.The planning focus was entirely rural be- cause in 1983 all of BRI's 3,626 unit desas were located in rural areas. (In 1989, however, the unit desa system was extended to urban areas, and KUPEDES came to be known as general purpose credit rather than its literal translation as gen- eral rural credit.) 232 The Microfinance Revolution: Lessons from Indonesia The second decision was to end subsidized lending programs at the unit desas and to introduce in their place the KUPEDES program of general rural credit at commercial interest rates. Kredit Mini and Kredit Midi were ended in late 1983. Outstanding BIMAS loans were transferred to the branches, although the unit desas acted as agents for the branches in collecting BIMAS loans.AIl new BIMAS lending stopped in early 1984, and the program officially ended in 1985. The third decision was to place high priority on mobilizing rural savings through the unit desas-both to meet demnd for savings services and so that KUPEDES would not require government funding in the future. The savings side of microfinance in the restructured units is discussed in chapter 13. Loan and deposit interest rates would be set so that the spread between them would be sufficient to permit the profitability-and hence the long-term The third decision viability-of the unit desa system. KUPEDES would be introduced in tandem with the reorganization of the unit desas. Previously branch windows, the units was to place high would now become commercial financial units with their own reporting sys- tem. KUPEDES would be the units' first new product. priority on Cutting back on credit subsidies mobilizing rural One issue remained. The cutting back of the subsidized credit programs was a sensitive matter that had to be handled carefully. In Indonesia, as in most de- savings veloping countries, rural credit subsidies tend to reach local elites whom gov- ernments want to retain as political supporters. Such subsidies are notoriously difficult to withdraw. In Indonesia the solution was one of simultaneous implementation through different offices. KUPEDES would be a commercial credit program available in the subdistricts at the unit desas and financed through locally mobilized sav- ings.The subsidized agricultural credit programs financed from the government budget would be continued, but at BRI's branches in district capitals. Since the subsidized credit would be rationed-unlike the commercial credit-KUPEDES could be expected to gain a large share of the rural credit market. This was a very Indonesian solution, drawn as it was from a culture in which opposites are considered complementarities, and from a nation whose unofficial motto has been "bisa diatur" ("it can be arranged").3 As a result the government was able to retain the allegiance of many rural elites while increasing support among villagers to whom institutional credit had not previously been available. Thus when BIMAS was phased out in 1984 and officially ended in 1985, it was immediately reincarnated into Credit for Farm Enterprises (Kredit UsahaTani, or KUT), a subsidized credit program for which BRI was required to act as a chan- neling agent for the government. But KUT was offered through BRI branches, not through the unit desas. As in BIMAS, KUT loans were sectoral and tied to input packages designed for well-irrigated lands. Hence the subsidies once again tended to reach wealthier villages-in this case through loans to farmers groups. As in BIMAS, KUT arrears were high. From the program's start in 1985 through the rainy season in 1990, KUT extended $208 million in loans (in 1990 dollars).'4 By the end of 1990 the amount overdue and unpaid since the pro- Success in Microlending The KUPEDES Credit Program, 1984-96 233 gram began was $37 nullion, or 18 percent of the loans disbursed during 1985-90. In comparison, KUPEDES disbursed $2.8 billhon in credit (in 1990 dollars) from its inception in 1984 through the end of 1990. The amount overdue and unpaid since the beginmng of the program was $29.9 million (in 1990 dollars), or 1.1 percent of total loans.Thus by 1990 KUPEDES had lent more than 13 times as much as KUT, yet the absolute amount of overdues was lower in KUPEDES. KUT was then phased out and replaced by a series of other subsidized rural credit programs.As instructed by the government, BRI continues today to im- plement a small number of "high priority," highly rationed subsidized credit programs-with similar results (see figure 12.1).These programs are available The intent behind to borrowers through BRI branches, not through the unit desas. Thus KUPEDES rapidly gained a large market share and maintained a high the initial repayment rate.The losses of KUT and the other subsidized programs were es- sentially a cost to the government for allowing KUPEDES to operate with- capitalization of the out hindrance from rural elites accustomed to subsidized loans. KUPEDES loan Starting the KUPEDES credit program KUPEDES received its initial funding from the $67 million in capital of Kred- program was to it Mini (which was converted into equity for the unit desas) and from $140 million in liquidity credits from Bank Indonesia at 15 percent a year.'5 The in- provide "the subsidy terest rate was set at 15 percent (the same rate that the unit desas paid depos- itors for TABANAS savings) so that BRI would not have an incentive to rely to end subsidies. " It on Bank Indonesia as the main long-term source of funds for lending at the units, rather than to make a strong effort to mobilize local savings.16 worked The Ministry of Finance agreed to continue subsidizing BRI's operating losses on the unit desas until the system broke even, which was estimated to take two years. It also continued to support extensive technical assistance to BRI for the development of the units. The intent behind the initial capitalization of the KUPEDES loan program was to provide "the subsidy to end subsidies" (Patten 1996,p. 4).And it worked. The KUPEDES credit program began nationwide through the unit desas in February 1984. BRI's newly transformed unit banking system broke even on a monthly basis by late 1985, a little less than the two years that had been es- timated, and the units showed a profit for 1986-and for every year since. A few days before the final decision was made to proceed with the restruc- turing of the umt desas, Finance Minister Ali Wardhana called a meeting to re- view the issues involved. During the meeting he enumerated a long list of all the things that could go wrong (the Ministry of Finance represents the government as the owner of BRI). I recall becoming more and more glum as the discussion continued. At the end, however, he smiled and said, "But this is Indonesia. We wllE do it."And they did. No country attempting to transform its rural banking system should underestimate the need for sustained, high-level leadership. Unit desa products are the focus of this chapter (credit) and of the next chap- ter (savings instruments). Chapter 14, on BRI's organizational structure, might 234 The Microfinance Revolution: Lessons from Indonesia have preceded the discussion of the units' products. But a more accurate feel for the development of the unit desas can be given in this way. In fact, it was the new products-and the operational and organizational requirements need- ed to meet the large demand for these products -that drove the reorganiza- tion of BRI, step by step. Since 1984 BRI has been gradually restructured to make possible an increasingly effective nationwide implementation of the units' products. Thus the discussion in chapter 14 explores the dynamics un- derlying the development of the unit desas into individual profit centers pro- viding specialized products for the microfinance market, and analyzes the effects of the new products on BRI's institutional structure. No country Basic Principles of KUPEDES attempting to Since 1984 managers and staff of the unit desa system have worked to for- mulate and implement the key principles of profitable microcredit (box transform its rural 12.1). Of course, BRI did not invent all the principles it uses. Some were de- veloped over thousands of years by moneylenders. Some date back more than banking system a century to the early People's Credit Banks in Indonesia and their prede- cessors in Europe. Some were pioneered by NGOs and other financial insti- should tutions in other countries and continents. Some were known to Fruin in the 1 920s. Some were adapted from Bank Dagang Bali and some from the Badan underestimate the Kredit Kecamatan of Central Java.And some were a legacy of the unit desas' Kredit Mini and Kredit Midi programs (which had in turn been influenced needfor sustained, by the Badan Kredit Kecamatan). Some important common lessons came from these otherwise quite differ- high-level leadership ent sources. Among them: * Financial services can be provided profitably to low-income people. * Interest rates for loans must be set higher than those normally charged by standard banks, reflecting the fact that delivering many small loans locally is more expensive than providing a smaller number of bigger loans in large bank branches. * Loan arrears can be kept low, primarily by careful borrower selection, and because borrowers are motivated to repay in order to retain the option to reborrow when products and services suit their needs. * Institutional activities are restricted entirely to financial services. BRI also learned from Bank Dagang Bali and some of the People's Cred- it Banks (BPRs) that with appropriate products and services, substantial amounts of voluntary savings can be mobilized from large numbers of low-income savers and used to finance loans. In addition, the unit desas learned from moneylenders that there is exten- sive demand from low-income borrowers for small, short-term, general-pur- pose loans in a range of amounts, and that the loans should have simple Success in Microlending The KUPEDES Credit Program, 1984-96 235 Box 121 Basic principles of KUPEDES Outreach and profitability * Institutional sustainability is required for large-scale outreach to low-income borrowers * The KUPEDES interest rate is set to cover all costs and risks, and to return a profit to the system Borrower selection and retention, and loan terms and repayment * Borrowers are selected by BRI without any outside intervention, and BRI bears the full credit risk The amount of the . KUPEDES loans from 25,000 rupiah to 25 million rupiah ($10 50 to $10,491 in 1996) are available to all creditworthy borrowers for all productive activities, including trade and services loan is based on the . Loan decisions are based on evaluations made by well-trained, experienced unit desa staff of the borrower's character (willingness to repay) and of the viability and cash borrowers current flows of his or her enterprise (ability to repay) * The amount of the loan is based on the borrower's current income flows, not on in- come estimated to be generated by the loan income flows, not on . KUPEDES is designed to meet the needs of low- and lower-middle-income borrowers for convenient bank locations, simple loan procedures, quick processes, and flexible income estimated to terms * Within overall KUPEDES regulations, the loan maturity and payment plan are custom- ized for each borrower's needs be generated by the . Borrowers who repay their loans promptly and whose enterprises remain creditworthy are permitted to borrow again for the gradually increasing loan sizes for which they loan qualify up to the KUPEDES maximum a Borrowers who do not repay their loans are not permitted to borrow again However, payments can be rescheduled if nonrepayment is caused by catastrophic events oc- curring throughout an area * To maximize personal responsibility for loan repayment, KUPEDES lends to individual borrowers, not to groups * BRI does not train borrowers or attempt to monitor their enterprises or the use of their loans, the premise is that borrowers know their businesses better than do bank staff Administration and operations * Each unit desa is treated as a financial unit, or profit center, with its own balance sheet and profit and loss statement * All unit desa regulations, products, reports, job descriptions, and the like are simple and standardized across all units * Unit managers and staff are given substantial responsibility and held accountable for their performance. They are provided intensive, high-quality training, and financial in- centives based on unit performance * Simple, transparent loan accounting and reporting enable frequent assessment of unit portfolios and immediate followup in case of problems Accounting is on a cash, not accrual, basis, only income actually received is posted. * Automatic criteria for decisionmaking based on predetermined criteria are used when- ever feasible-for example, for loan reserve and writeoff procedures and for additions to a unit's staff * Unit staff are supervised weekly by well-trained supervisors from the unit's supervis- ing branch 236 The Microfinance Revolution: Lessons from Indonesia procedures, quick decisions, flexible terms, reborrowing facilities, and gradu- ated amounts-and that borrowers will pay the costs of such loans. Finally, the unit desa system mined its experience with Kredit Mini and Kred- it Midi. Unlike BIMAS, Kredit Mini and Kredit Midi loans had been provid- ed in cash, borrowers were selected by local unit desa staff who appraised the creditworthiness of each loan applicant, loan terms were relatively short, reg- ular installment payments on principal were required, and repayment rates had been high.While these loan programs had been constrained by the 12 percent annual effective interest rate mandated by the government for state banks, in- terest rate caps disappeared with the 1983 deregulation. BRI was thus free to build KUPEDES on the lessons of Kredit Mini and Kredit Midi, but to do so at an interest rate that would enable the units to achieve profitability. BRI does not train Central principles of KUPEDES credit include lending to individuals, pricing for profitability, methods of borrower selection and retention, loan terms borrowers; the and repayment, and a new approach to organization, administration, and op- erations at the unit desas (see box 12.1).The role of KUPEDES in the unit desas premise is that can be understood only in the context of the simultaneous reorganization of BRI's mucrobanking system. Each supported the other, and neither could have borrowers know succeeded alone. their businesses KUPEDES and unit desa reorganization To implement the new KUPEDES product and the forthcoming savings in- better than do bank struments, BRI began to reorganize the unit desas in late 1983 and early 1984 (chapter 14). As noted, by 1984 each unit desa was treated as a financial unit, staff or profit center. Separate reporting for each unit enabled unit staff and super- visors to evaluate performance monthly and to take action quickly when problems arose. Unit staff promotions and incentives were based on the per- formance of their unit. Unit desa regulations, products, reports,job descriptions, and the like were made simple and standardized across all units. This approach made it possible for the unit desas to be successfully man- aged by local high-school graduates who, after previous experience as unit desa bookkeeper, teller, and credit officer, are selected and trained by BRI and pro- moted to unit manager. Standardization and simplicity also make it possible for the branches to supervise large numbers of units effectively, and for the regional offices and the head office to oversee a large system efficiently. Unit managers and staff are given substantial responsibility, and they are held accountable for their performance.They are given intensive, high-quality training so that they will know their jobs well, and performance-based financial incentives to en- courage them to do these jobs well. In addition, unit staff are regularly and fre- quently supervised by well-trained supervisors from the unit's supervising branch. A new system of loan accounting and reporting was designed to be sim- ple and transparent.Accounting is on a cash, not accrual, basis; only income ac- tually received is posted. The staff of each unit produce five monthly reports: a balance sheet, a profit and loss statement, a credit report that details loan ar- Success in Microlending The KUPEDES Credit Program, 1984-96 237 rears over time, an achievement of targets report, and a unit development re- port (UDR).The UDR, which assesses the condition of the unit using 26 in- dicators, provides the basic monthly information used by unit and branch managers to review the performance of their units.This report, aggregated month- ly at the branches, regional offices, and head office, serves as the main indica- tor of unit desa performance over time (see chapter 14). These 5 reports replaced the 32 monthly reports required of each unit before 1984. The unit desas' management information system now provides far better information in 5 reports than was available from the 32 reports of the BIMAS era. Use of these 5 reports enables frequent and regular assess- ment of each unit's portfolio, permitting immediate action to be taken on Accounting is on a loans in arrears. To reduce opportunities for corruption and to increase staff efficiency, cash, not accrual, automatic criteria for decisions based on preset requirements are used when- ever feasible. Thus loan reserve and writeoff procedures are determined by basis; only income the length of time the overdue loan balance has gone beyond its due date, while credit officers are added to units depending on the number of borrowers, actually received is and bookkeepers and tellers are added depending on the number of daily transactions. posted The KUPEDES interest rate was set to cover the relatively high overhead costs of the units (including depreciation), and the commercial cost of funds; it covers estimated inflation and risk, and it enables profitability. Borrowers are selected by the bank without outside intervention, and BRI bears the fill credit risk. Loans are made available to all creditworthy individ- uals for all productive activities, and prompt payers whose enterprises qualify are permitted to reborrow for gradually increasing amounts.As BRI's president- director put it in 1996, "More than 80 percent of KUPEDES borrowers bor- row again, and they need to know that their ability to borrow again depends on their own performance, not on factors outside their control" (Moeljono 1996, p.4). Loan procedures are simple and quick, and 36 combinations of maturi- ties and payment terms provide borrowers with a variety of options. Selecting KUPEDES borrowers Low- and lower-middle-income Indonesian households typically have multi- ple income sources; a 1990 BRI study of KUPEDES borrowers showed be- tween 1 and 12 economic activities per household, with an average of 3.6 (Sutoro and Haryanto 1990). For a KUPEDES loan, the borrower identifies one of the household's economic activities as the loan purpose and must show that the present income of that enterprise is capable of repaying the loan.Thus the amnount of the loan is based on the current income flows of the enterprise, not on in- come estimated to be generated by the loan. Loan decisions are based on unit staff evaluations of borrowers' characters and the viability and cash flows of their enterprises.Just as the bank is fully re- sponsible for selecting borrowers and bears the full credit risk for KUPEDES, for most loans the manager and credit officer of a unit desa choose the unit's 238 The Microfinance Revolution: Lessons from Indonesia borrowers and bear the responsibility for their decisions (all unit desa managers have served previously as credit officers). Since credit officers must assess the level of business and the cash flows of microenterprises that normally do not have formal accounting records, business plans, or financial statements, they are trained to prepare simple balance sheets and income statements for many kinds of microbusinesses. For example, a credit officer evaluating the enterprise of a woman who rais- es chickens and sells eggs must know the local egg market: the prices of chick- ens of different qualities, the price of feed, the probabihty of disease, the average laying rate, the prices of eggs, transportation costs, the wholesale and retail de- mand in the area, and so on.The credit officer will then have to decide whether the potential borrower's statements about her business are essentially accurate, Since interest rates and whether the enterprise can support the projected loan payments. (The cred- it officer also evaluates other household economic activities, whlch serve as back- are not subsidized, up for loan repayment in case the borrower's egg business slumps or fails).The egg seller may be able to make the loan payments, but will she do so? The cred- capital is not it officer is also trained to evaluate the character of the potential borrower through discussions with other unit clients who know her, her neighbors, her buyers constrained, and suppliers, and her vlllage officers. Unlike BIMAS, KUPEDES credit-widely available to all creditworthy bor- KUPEDES loans rowers-has not become politicized. Since interest rates are not subsidized, cap- ital is not constrained, KUPEDES loans are not rationed, and there is little scope are not rationed, and for corruption. In addition, staff bonuses and promotions depend on high re- payment rates. For both reasons, there is little incentive to lend to unqualified there is little scope borrowers or to relend to defaulting borrowers. for corruption Characteristics of the KUPEDES Loan Product KUPEDES, a general-purpose loan instrument, is the only loan product of- fered in the umt desas (and it is available only in the units).7 The decision to offer a single loan product that would be available for all productive purposes contributed substantially to keeping the units simple and transparent-both among the key reasons for the units' success. Loan purposes In designing the KUPEDES loan product, BRI acknowledged what money- lenders and borrowers had always known: credit is fungible. As noted, a KU- PEDES loan applicant must cite one activity as the borrowing enterprise. But BRI recognizes that KUPEDES loans are often used by borrowers for other economic activities, as well as for consumption. KUPEDES loans are classified as working capital or investment capital ac- cording to the purpose stated in the loan application. Loans are available for up to two years for working capital and up to three years for investment cap- ital; most loans are repaid in monthly installments.At the end of 1995, 72 per- Success in Microlending The KUPEDES Credit Program, 1984-96 239 cent of outstanding loans were for working capital and 28 percent were for in- vestment capital (BRI 1996a, p 4).18 The most frequendy stated loan purpose is trade, which accounted for near- ly 44 percent of loans and about 45 percent of the value of outstanding credit in a 1995 survey of KUPEDES borrowers. But the share of trade loans was proba- bly overstated because it is relatively easy for small traders to demonstrate the cash flows that can justify a loan. Many traders also engage in agricultural activities and use some of their KUPEDES loans for agriculture and other purposes. Agricul- ture accounted for the stated purposes of 18 percent of the loans and 17 percent of the outstanding credit in the survey (BRI 1996a, p. 4). A large variety of other purposes made up the rest-especially small industry, transport, and other services. KUPEDES, a BRI does not permit KUPEDES loans to be taken for purposes such as cer- emonies, health care, education, housing, consumer durables, and basic house- general-purpose loan hold needs. But the units are aware that some loan proceeds are used for these purposes. For BRI the most important concerns are that the borrower can prof- instrument, is the itably use and repay the loan, and that the repayments are made as scheduled. only loan product Loan sizes KUPEDES loans were originally available for 25,000-1 million rupiah ($23-931 offered in the unit in 1984). Since then the rupiah maximum has been gradually raised: in 1996 KUPEDES loans were available from 25,000-25 million rupiah ($10.50-10,491). desas The increases in the maximum loan size have been driven primarily by the de- mands of KUPEDES borrowers who, over time, qualified for larger loans. In practice, however, few KUPEDES loans are above 5 million rupiah at disbursement ($2,098 in 1996); the 1995 KUPEDES loan survey found that only 3 percent of respondents' loans at disbursement were above that amount (table 12.2). More than three-quarters of the loans were below $900 (BRI 1996a, p.5). In 1996 the average loan at disbursement was $1,025, but most loans fell well below that amount (BRI 1997b, p. 7). Interest rates When KUPEDES began in 1984, the monthly interest rate was set at 1.5 per- cent for working capital loans and 1.0 percent for investment loans, both calcu- lated on a flat basis on the original loan balance.'9 In 1990, however, both types of loans began to be offered at the same flat rate, 1.5 percent a month.20This rate is equivalent to about a 32 percent annual effective interest rate for a 12-month loan with monthly installments if all payments are made on time. Adjusting this 32 percent rate to account for annual inflation in Indonesia in 1984-96 suggests a 22-27 percent range in real effective annual lending rates for most loans. But borrowers who make loan payments late pay a hlgher interest rate than those who pay on time (see discussion of the prompt payment incentive, below). Monthly installments are calculated so that all payments of principal and interest remain the same throughout the payment period.This is an attractive feature for KUPEDES borrowers, who like its simplicity and regularity. How- ever, there are other payment schedules, including single-payment and seasonal 240 The Microfinance Revolution: Lessons from Indonesia Table 12.2 1 Size distribution of KUPEDES loans at disbursement, 1995 Loan size (January 1995 U.S. dollars) Share of loans made (percent) <$225 14.6 $226-450 30 2 $451-900 31.9 $901-2,250 20 2 >$2,250 3.1 Source BRI 1996a, p 5 In designing KUPEDES, BRI loans.Thus there is some variation in KUPEDES annual effective interest rates acknowledged what depending on loan maturities and payment schedules; the nominal annual rate can be as low as 19 percent on some short-term, single-payment loans. moneylenders and In 1995 BRI lowered the mterest rate on larger KUPEDES loans.This change was made to reflect the fact that operating costs decrease as loan sizes increase, borrowers had and to ensure that the units would remain competitive with the many banks entering the same market niche (largely as a result of the unit desas' success). always known: credit Because of the way the KUPEDES loan product works, larger loans general- ly imply longer-term umt clients. Thus the interest rate reduction effectively isfungible rewarded good borrowers who qualified for larger loans. The flat monthly interest rate on loans of 3 million-5 million rupiah ($1,259-2,098 in 1996), for those who pay on time, was set at 1.5-1.2 percent of the original loan amount, with the larger loans carrying the lower rates. (The interest on loans above 5 million rupiah was also 1.2 percent a month). The 1.2 percent flat monthly rate is equivalent to an annual effective rate of about 26 percent for a 12-month loan with monthly installments. Thus the annual effective interest rate for most KUPEDES loans ranged from 32-26 percent if payments were made on time. The average annual effective rate was estimated at about 30 percent.2' KUPEDES loans over $2,098 had an inflation-adjusted real interest rate ranging from 17-20 percent in 1995-96. But as discussed in chapter 15, during the recent financial crisis the in- terest on larger loans was returned to the original 32 percent annual effec- tive rate, the same as the interest on smaller loans. And in September 1998 the annual effective interest rate was raised to 45 percent. It was then grad- ually lowered to the original 32 percent by late 1999. In both cases late pay- ers paid higher rates. Loan terms Maturities and repayment terms. Within its standardized package of loan terms, KUPEDES offers borrowers considerable flexibility in meeting their vary- Success in Microlending The KUPEDES Credit Program, 1984-96 241 ing credit needs. Loans are available for maturities ranging from 3 months to 24 months for working capital loans and to 36 months for investment loans. Repayment terms include monthly payments, seasonal payments, single pay- ments (for loans with maturities of one year or less), and loans with grace pe- riods up to nine months.As noted, however, most loans are repaid in monthly installments. Unit staff and most borrowers prefer the discipline imposed by reg- ular payments, and BRI has found that loans with monthly payments perform considerably better than loans with seasonal maturities, single payments, and grace periods. According to BRI's 1995 loan survey, 40 percent of KUPEDES loans were taken for 12 months, 17 percent were taken for 18 months, and 25 percent were BRI hasfound that taken for 24 months. Only 7 percent of loan maturities were longer than 24 months; 11 percent were fewer than 12 months. Borrowers are limited to one loans with monthly working capital loan and one investment loan at the same time; the combined amount can be no larger than 25 million rupiah ($10,491 in 1996). Prepay- payments perform ment of loans, with an interest rebate, is permitted for loans with maturities of nine months or more, but not within the first or last three-month period. considerably better The loan table book used to determine which terms most closely meet a borrower's needs lists 36 combinations of maturities and repayment terms. For than loans with each maturity-repayment combination, the table shows, for a number of loan amounts, the size of each payment, the principal and interest components of other payment the payment, and the amount of prompt payment incentive (see below) that can be earned in a six-month period. Since the loan term tables are printed, schedules unit staff do not calculate the terms for particular loan amounts, nor do they compute payments, thus increasing staff efficiency. Using the loan tables and reviewing the payment schedules with borrowers also helps make loan arrange- ments transparent, reducing the likelihood that staff members will attempt to collect extra fees or provide special terms to favored customers. Prompt payment incentive. A prompt payment incentive (insentif pemba- yaran tepat waktu, or IPTW) has been built into KUPEDES since the program started in 1984.A flat monthly rate of 0.5 percent of the original balance is added to each payment. If all scheduled payments are made on time and in full for six months, the incentive payments are returned to the borrower. But if any scheduled payment is not made on time, the incentive for that six-month pe- riod is treated as a late payment penalty and forfeited to the bank. For a monthly installment loan, a six-month incentive refund is equivalent (on average) to about 30 percent of the monthly payment. If a payment is late in both six-month periods of a 12-month loan, the entire incentive is forfeit- ed, and the annual effective interest rate can then exceed 40 percent. Thus the prompt payment incentive provides a substantial inducement for timely repay- ment.22 An example of the prompt payment incentive is provided in box 12.2. The prompt payment incentive is paid into the borrower's savings ac- count.All borrowers are required to have a savings account at their unit for this purpose; many use it for additional savings as well. 242 The Microfinance Revolution: Lessons from Indonesia The KUPEDES Box 12.2 prompt payment incentive The following is an example of a one-year KUPEDES loan of 1,000,000 rupiah ($420 in 1996) with 12 equal monthly payments Loan principal 1,000,000 Monthly principal payment (1,000,000/1 2) 83,333 Monthly interest payment (flat rate of 1 5 percent x 1,000,000) 15,000 Monthly prompt payment incentive (0 5 percent x 1,000,000) + 5,000 36 combinations of + 20,000 Total monthly payment 103,333 maturities and x 12 repayment terms are Total to be repaid 1,240,000 rupiah ($520 in 1996) available so that Less prompt payment incentive (if all payments made in full and on time) - 60,000 borrowers can Net amount repaid by prompt payers 1,180,000 rupiah ($495 in 1996) customize their Thus interest payments for prompt payers total 180,000 rupiah. If all payments are made in full and on time, the approximate interest rate is an 18 percent annual flat rate (based loans to meet their on the original balance), which is equivalent to about a 32 percent annual effective rate (based on the declining balance) needs Source BRI data Collateral. Until 1992 the unit desas were legally required to take collat- eral that was at least equivalent to the value of the loan. The units continued to require collateral after the law was changed, but BRI established more flex- ible rules about the kinds of security that are acceptable. Most KUPEDES loans are collateralized with land. However, a 1990 BRI study of rice-producing areas found that nearly half of unit desa bor- rowers had no riceland (Sutoro and Haryanto 1990). In addition, obtain- ing full title to land is often difficult and expensive.Therefore, if full land title is not available, lesser certificates and land tax bills and receipts are ac- ceptable. The units also accept as collateral most forms of fixed and mov- able assets for which proof of ownership can be ascertained, including furniture, machinery, motorcycles, bicycles, and household goods. Savings and payroll deductions can also be used for collateral. A 1992 BRI study of loan collateral showed that 24 percent of loans were collateralized with land or a house plot with full title, while 59 percent were secured with less- er documentation of land or house plot ownership (BRI 1996a, p. 8). In Success in Microiending The KUPEDES Credit Program, 1984-96 243 1996 a new law tightened collateral requirements for loans over 5 million rupiah ($2,098). BRI views KUPEDES collateral primarily as a psychological incentive and an indication of the borrower's intent to repay. Except in unusual cases of fraud or borrower deception, the units rarely take legal steps to collect collateral from defaulters.As Fruin's (1994 [1935], p. 149) handbook put it, "using assets as a criterion [in loan decisions] is intended primarily to maintain a reasonable re- lationship between debt and property; it is not a question of them being of much value to the bank upon enforced sale." Research on household assets carried out by BRI in the 1990s indicated that about 90 percent of Indonesian households own assets that could qualify The loan is based as collateral at a BRI unit (BRI 1998a, p. 8). Thus the loan security require- ment is not viewed as a major barrier to loan accessibility. In practice, howev- on the er, not all unit managers accept all forms of collateral, and in some cases the collateral requirement can prevent otherwise qualified borrowers from obtaining creditworthiness of a KUPEDES loan.23 one of the Implementation of the KUPEDES Credit Program households Implementation of the KUPEDES credit program includes evaluating and ap- enterprises but all proving loan applications, delivering and collecting loans, and measuring and managing delinquency; these aspects of the program are discussed in this chap- household income ter. But KUPEDES administration also involves asset-liability management; re- porting, audit and internal supervision; human resource management, logistics, streams are assessed cash flow arrangements, and other issues that are discussed in later chapters. From application to collection KUPEDES has a simple loan application and approval process.A potential bor- rower discusses his needs with his local unit desa's credit officer, manager, or both.They encourage applicants who appear to be creditworthy and discour- age those who seem unlikely to be fundable.Applicants who will probably not be found creditworthy usually go no further. Although no statistics are kept, BRI estimates that less than 5 percent of those who formally apply for KU- PEDES loans are rejected (BRI 1996a, p. 9). This is an Indonesian way of doing business.The result is that in most cases there is no direct denial of credit, the prospective borrower does not lose face by being rejected, and a potentially negative relationship between the bank and the would-be borrower is avoided.This approach is important to both the bank and the client for the long run. Perhaps the client is a saver, or may become one. Perhaps the enterprise in question will become creditworthy in time. Per- haps another family member is, or wllR become, a unit client. Future options are left open. Applicants who are encouraged to apply for credit then fill out the loan application, which requests the borrower's name, address, occupation, family sta- 244 The Microfinance Revolution: Lessons from Indonesia tus, borrowing history, loan purpose, amount and terms requested, and a brief description of how the money will be used, including the amount to be sup- plied from the applicant's own funds (BRI 1996a, p. 9). If needed, unit staff help applicants to fill out the form. After a borrower has completed the application, the credit officer visits the borrower at her workplace (in many cases this is also the applicant's home), ap- praises the activity for which the loan has been requested, and collects detailed information on household economuc activities and income flows.Although the loan is based on the creditworthiness of one of the household's enterprises, all household income streams and expenses are assessed.The credit officer prepares simple income statements and balance sheets based on information the house- hold supplies; for loans above 750,000 rupiah ($315 in 1996), these are sum- Most KUPEDES marized on a credit information form. Another form provides details on the loan collateral. loan decisions are If a loan applicant is new to the unit desa, the credit officer or unit man- ager will talk with the village head or other local government representative made by the unit about the candidate and make inquiries about the person's creditworthiness from the applicant's buyers and suppliers. Neighbors and other unit desa clients in desa manager. There the area are also questioned about the applicant. For a repeat borrower, the ap- plication process is similar but shorter, and the field visit is less detailed. are no loan Most loan decisions are made by the unit desa manager. Each unit man- ager has a loan authority limit; within overall BRI rules, the amount of the man- committees ager's authority is set by the manager of the supervising branch. The amount set depends on the unit manager's experience and performance. For loans above the unit manager's authority, the decision is made by the unit develop- ment officer posted at the branch or the branch manager; there are no loan com- mittees for KUPEDES loans. Loan applications from defaulting borrowers are rejected, although borrowers who repay in full during the loan term, but not on time, may be considered for new loans depending on the circumstances. Loans are disbursed in cash at the unit desa immediately after the loan has been approved. Loan payments are made at the units on a prearranged schedule. Measuring and managing delinquency BRI places great emphasis on measuring and managing loan delinquency. As soon as a borrower misses a payment, the credit officer is expected to visit him; many overdue payments are collected on such visits. If the loan payment has not been made after several visits, the unit manager will also visit the borrow- er. If this fails, the unit development officer from the branch will work with the credit officer and unit manager to collect the loan. Most overdue repay- ments are collected during this process. However, if a unit's collection rate falls below 95 percent in a particular month, the unit manager's loan authority can be withdrawn. All loans must then be approved by the branch until the unit's collection rate again reaches 95 percent and its manager's loan authority is re- stored. Success in Microlending The KUPEDES Credit Program, 1984-96 245 A loan installment that is unpaid is classified as overdue one week after the due date. Only the overdue installment is so classified; the balance of the loan is classified as current. Loan loss reserves are calculated monthly at each unit.A unit's reserves con- sist of the total of 3 percent of principal outstanding, 50 percent of balances on loans up to three months past their final due date, and 100 percent of bal- ances on loans three months or more past their final due date. All balances on loans 12 months past their final due date are written off, although loan collection continues. Reserve and writeoff procedures are au- tomatic; they are determined by the length of time the overdue loan balance has gone beyond its due date.This method-in which the loan officer and branch Reserve and writeoff manager do not make individual judgments about the collectibility of a loan- helps them to avoid the temptation to underestimate risks or to overstate prof- procedures are its by not reserving adequately against loss. It also provides an automatic, conservative estimate of the value of the loan portfolio. automatic; they are BRI uses four measures in assessing KUPEDES portfolio quality: portfo- lio status, long-term loss ratio, short-term loss ratio, and 12-month loss ratio determined by the (box 12.3).The first three are reported monthly by each umt and are then ag- gregated regionally and nationally.The fourth is calculated only at the region- length of time the al and national levels.Taken together, these four measurements indicate trends in portfolio quality.24 As discussed later in the chapter, the KUPEDES repay- overdue loan balance ment rate was high during 1984-96, and it continued to be high during the recent crisis (see chapter 15). is beyond its due Outreach date KUPEDES loans reach borrowers of many kinds, and the loans finance a wide variety of econonic activities distributed over thousands of miles.25 Unit desa borrowers operate small and microenterprises in trade, agriculture, livestock, poultry, dairying, fishing, and food processing; in services (transportation, restau- rants, gas stations, repair services); and in manufacturing (textiles, garments, leather goods, furniture, crafts, bricks, tiles, jewelry, herbal medicines, and others). According to a KUPEDES impact study carried out in 1988 in four provinces, the average capital investment in the borrowing enterprise was $652 (Sutoro and Haryanto 1990, p. 19).26 BRI conducted another survey in 1996 to learn more about KUPEDES borrowers' opinions on the loan product and the unit desas, and to collect data on aspects of KUPEDES (such as loan size distribution) not available from BRI's monthly reports. The 1996 survey was not designed as an impact study. Rather, its pur- pose was to learn how respondents viewed the KUPEDES product. How- ever, some of its questions concerned respondents' perceptions of the impact of their loans on their households' economic growth.While the survey re- sults should not be interpreted as definitive evidence of KUPEDES impact, they are useful as a general overview of borrowers' opinions. But because the survey was conducted by BRI staff, it is possible that-despite assurances of confidentiality (which are not given much credence in rural Indonesia)- 246 The Microfinance Revolution: Lessons from Indonesia Measures of Box 12.3 KUPEDES portfolio quality Portfolio status Portfolio status measures the aggregate amount of overdue principal installments relative to total principal outstanding On the one hand portfolio status understates the risk in the portfolio because it includes only overdue payments, not the outstanding balance of over- due loans-which are at higher than normal risk Moreover, when the amount of out- standing loans is growing rapidly, portfolio status tends to understate portfolio risk be- cause the denominator of the ratio is rising quickly But there is an offsetting tendency for portfolio status to overstate risk because most overdue installments are eventually collected In December 1996 the KUPEDES portfolio status was 3 65 percent 99 percent Of Long-term loss ratio The long-term loss ratio measures the cumulative amount that has come due and been respondents said unpaid since the unit opened relative to the total amount that has come due Because the long-term loss ratio measures KUPEDES performance since a unit opened, it is a useful historical record, but it is not sensitive to short-term changes in collection experience In that KUPEDES December 1996 the KUPEDES long-term loss ratio was 2 15 percent Short-term loss ratio loans had helped The short-term loss ratio measures monthly changes in the components of the long-term loss ratio Because it is a measure of short-term changes in several variables, it is volatile their enterprises and difficult to interpret alone A substantial change is an indication that the unit manager and branch supervisor should look closely at each component for the month to see what grow caused the sudden change in the ratio In December 1996 the KUPEDES short-term loss ratio was 1 30 percent Twelve-month loss ratio The 12-month loss ratio measures the change over the most recent 12-month period in the components of the long-term loss ratio It is a useful indicator of changes in portfolio quality because it covers a long enough period to avoid short-term volatility, but a short enough period to reflect portfolio trends In December 1996 the KUPEDES 12-month loss ratio was 1 59 percent Source BRI 1996a, pp 10-11 some clients may have responded with more positive views than they actu- ally held. The 1996 sample of 1,500 people consisted of 12 KUPEDES borrow- ers from each of 125 unit desas, located in 62 branches throughout all 15 of BRI's regions.27 Each respondent had borrowed from KUPEDES at least three times.The sample was stratified proportionately on size of loan and on eco- nomic sector, according to unit desa national data. Of the 12 borrowers sam- pled in each unit desa, 9 had outstanding loans up to and including 3 million rupiah ($1,259), 2 had loans between 3 million and 5 million rupiah, and 1 had a loan above 5 million rupiah ($2,098). In each of the unit desas sur- veyed, seven of the sample borrowers were drawn from the trading sector, three from agriculture, and one each from industry and other enterprises. Success in Microlending The KUPEDES Credit Program, 1984-96 247 From the original sample of 1,500, information was obtained from 1,341 respondents. Additional views about KUPEDES were obtained through intensive in- terviews with borrowers carried out under BRI's InternationalVisitors Pro- gram beginning in 1996; the examples of specific borrowers below are drawn from those interviews. Most borrowers (68 percent) in BRI's 1996 survey were between 30 and 50 years old, and most (68 percent) had attended primary orjunior high school (table 12.3).Twenty-nine percent of the respondents were women (in the 1988 survey 24 percent of the larger sample of 1,404 borrowers were women and 29 percent of the smaller sample of 192 borrowers were women). De facto, however, pro- 89 percent of the ceeds from KUPEDES loans to other household members are often used by fe- male household members for their enterprises. Since loans are typically used not respondents reported only for the official loan purpose but also for other household economic activ- ities, many more women use KUPEDES loans than are recorded as so doing. that the capital of Using loansfor enterprise growth. Most KUPEDES borrowers in the 1996 the borrowing survey were long-time unit desa customers; 56 percent had been clients for more than five years. Fifty-one percent reported that they had borrowed from KU- enterprise was larger PEDES between three and five times, and thirty-one percent had borrowed between six and ten times. than when they An overwhelming 99 percent of survey respondents said that KUPEDES loans had helped their enterprises grow.While the 1988 and 1996 surveys were took their first concerned with different issues, there are some areas of overlap; enterprise growth and profitability is one. A strong finding of the 1988 impact survey had been KV'PEDES loans that on average the profits of borrowers' enterprises had increased substantial- ly (by 93 percent after adjusting for inflation) after three years of KUPEDES borrowing.This increase was attributed largely to higher volumes of goods pro- duced, bought, and sold; payments made to suppliers in cash rather than on cred- it; and opportunities to work year round without work stoppages caused by lack of working capital (Sutoro and Haryanto 1990, p. 8). Eighty-nine percent of the respondents in the 1996 survey reported that the capital of the borrowing enterprise was larger at the time of their most recent KUPEDES loan than it had been at the time of their first KUPEDES loan. Longer- term borrowers typically reported enterprise growth, increasing sales, and their qualification for larger loan sizes. For example, CS, a farmer in EastJava, received his first KUPEDES loan in 1986: an 18-month working capital loan for 300,000 rupiah ($183 in 1986; BRI 1996a, p. 11). Over the next 10 years he received six additional KUPEDES loans, gradually increasing in size to 1.2 million rupiah ($504) in 1996, to finance the cultivation of rice, peanuts, chili peppers, corn, and veg- etables. CS reported that during 1992-96 his sales had nearly doubled and that in 1996 he had been able to hire six seasonal employees. Similarly, LW, the proprietor of a market stall in Central Java, took her first KUPEDES loan in 1990, an 18-month working capital loan for $316.28 She used the loan to purchase traditional Javanese snacks, candy, crackers, tea, sugar, 248 The Microfinance Revolution: Lessons from Indonesia Table 12.3 Excerpts from responses to KUPEDES borrower survey, 1996 Survey question Response Share (percent) Loan type Working capital loan 88 Investment loan 12 Loan purpose Trade 70 Agriculture 16 Industry 6 Other 8 Gender of borrower Male 71 Afarmer in East Female 29 Age of borrower at time of survey <25 1 Java reported that 26-30 5 31-35 15 during 1992-96 36-40 19 41-45 21 hsnal 46-50 13 his sales had nearly >50 26 doubled and that he Highest level of education No school 2 Primary school 40 Junior high 28 had hired six High school 26 University 4 seasonal employees Length of time as a unit 1-3 years 20 desa customer 3-5 years 24 5-7 years 26 >7 years 30 Number of KUPEDES loans taken 1 4 2 14 3 20 4 17 5 14 6-10 31 Ever late in paying a No 85 KUPEDES installment? Yes 15 If yes, reason for late payment Family needs 45 Business problems 35 Other 20 Borrowing enterprise Primary business 82 Secondary business 18 (Table continues on next page) Success in Microlending The KUPEDES Credit Program, 1984-96 249 Table 123 (continued) Survey question Response Share (percent) Is the capital of the borrowing enterprise Larger 89 larger for the most recent KUPEDES Same 8 loan than for the first? Smaller 3 Has the use of KUPEDES helped the Yes 99 borrowing enterprise grow? No 1 Has the use of KUPEDES helped the Yes 99 56 percent had well-being of the borrower's family? No 1 What effects has the growth of the Renovated or improved housing 56 renovated their borrowing enterprise had on the Able to pay more school fees, borrower's family? keeping children in school longer 52 houses and 52 Purchased household goods 33 Recreation or entertainment 31 percent had been Purchased a vehicle 26 percent had been Installed electricity 21 Purchased cattle 14 able to pay more Installed telephone 11 Other 7 schoolfees Friendliness and helpfulness of Friendly and helpful 98 unit desa staff Not so friendly or helpful 2 Unfriendly and unhelpful 0 How quickly is the borrower normally Quickly 95 served by unit desa staff? Not very quickly 4 Slowly 1 Does the borrower find it difficult to No 81 pay the KUPEDES interest rate? Yes 19 Time taken to process last KUPEDES <1 week 76 loan application 1-2 weeks 22 >2 weeks 2 Time taken to process earlier <1 week 72 KUPEDES loan applications 1-2 weeks 25 >2 weeks 3 Reasons for choosing KUPEDES Already had a long relationship with the unit desa 73 Quick loan decisions 66 Easy loan procedures 64 Good service 62 Unit located nearby 57 No other option 21 Other 7 250 The Microfinance Revolution: Lessons from Indonesia Table 123 (continued) Survey question Response Share (percent) Does the borrower plan to apply for Yes 98 another KUPEDES loan? No 2 Size of planned KUPEDES loan relative Larger 81 to current KUPEDES loan Same 17 Smaller 2 Has the borrower ever been a No 90 customer of another bank? Yes 10 76 percent said they If yes, reasons the borrower moved Lower interest rate in unit desa 55 from the other bank to the unit desa Unit desa is near house or workplace 42 received their loan in Good service 35 Forms easy to fill out 29 a week; only two Quick loan decisions 25 Many choices of loan terms 15 Payment plan not difficult 12 percentsaidittook Other 31 more than two Source BRI 1996b weeks and other items that she sells in her market stall. LW said that when she first started borrowing from KUPEDES, her monthly sales averaged 3.2 million ru- piah ($1,683 in 1990). But by 1995, when she received her seventh loan for 5 million rupiah ($2,166 m 1995), her average monthly sales had more than tripled. Improving the household's standard of living. Ninety-nine percent of the respondents in the 1996 survey said that KUPEDES loans had helped improve the well-being of their families. Respondents reported further that the growth of their enterprises had enabled them to use their income for a variety of pur- poses that benefited their families. Fifty-six percent had renovated their hous- es or made house improvements, and fifty-two percent had been able to pay more school fees, enabling their children to stay in school longer. Others re- ported installing electricity, a telephone, or both; purchasing household goods; buying a vehicle; or spending for recreation and entertainment. The 1988 survey had reported that most respondents, after an average of three years of KUPEDES loans, had said their fanulies were better off in that they ate more proteins, kept their children in school longer, renovated their hous- es, opened savings accounts, and used medical services. RH, a weaver from South Sulawesi, received a Kredit Mini loan in 1982 and then seven KUPEDES loans of progressively increasing amounts to finance her sarong weaving business (BRI 1996a, p. 13) .The Kredit Mini loan was for 150,000 ($216 in 1982). By 1996, her seventh KUPEDES loan was for 2.5 million rupi- Success in Microlending The KUPEDES Credit Program, 1984-96 251 ah ($1,049 in 1996). She employs several of her neighbors, and she reported that her average monthly sales of sarongs had increased more than 10 times during the period she had borrowed from the unit desa. In addition to helping support her family, RH was able to use profits from her weaving business to pay for her sib- lings to attend umversity. Other examples of the uses and benefits of KUPEDES loans, as described by borrowers, are provided in volume 1, box 1.1. Borrowers' views on KUPEDES loan terms and procedures. Respondents to the 1996 survey generally said that they found KUPEDES loans to be suit- able for their needs. Eighty-one percent said that they did not find it difficult to pay the KUPEDES interest rate. Eighty-five percent reported that they had Borrowers reported never been late in paying a KUPEDES installment; most of the 15 percent who said they had been late on at least one payment of one loan said the reason was unit staff to be "family needs" (45 percent) or "business problems" (35 percent). The units were reported to be efficient in making loan decisions. Seven- friendly and helpful ty-six percent of respondents said that they had received their most recent loan within a week of applying for it; only two percent said that it had taken more (98 percent) and to than two weeks for their loan to be disbursed. Borrowers reported unit staff to be friendly and helpful (98 percent) and to provide quick service (95 percent). provide quick service When asked about their reasons for borrowing from KUPEDES, as opposed to other lending sources, 73 percent of respondents mentioned that they had (95 percent) already had a long relationship with the unit desa. Other important reasons were quick loan decisions (66 percent), easy loan procedures (64 percent), good ser- vice (62 percent) and the convenient location of their unit desa (57 percent). Nearly all the respondents (98 percent) said that they planned to borrow again from KUPEDES: 81 percent of them planned to borrow a larger amount, 17 percent the same amount, and 2 percent a smaller amount. Only 10 percent of the respondents said they had ever been a customer of another bank. This corresponds closely with the 1988 study, which found that 92 percent of 192 respondents said no when asked whether they had ever re- ceived a loan from a bank or government agency.When asked why they moved to a BRI unit, the 10 percent in the 1996 survey who had banked elsewhere cited the lower KUPEDES interest rate (55 percent), the more convenient location of the unit desa (42 percent), the good service (35 percent), the simpler forms and procedures (29 percent), and the quicker loan decisions (25 percent).The many choices of loan terms and easy payment plans of KUPEDES were also mentioned as reasons for moving to the units. Overall, KUPEDES repayment and profits are high, outreach is wide, and borrowers report being generally satisfied. KUPEDES Performance, 1984-96 The number and value of KUPEDES loans outstanding from the start of the program in 1984 through 1996 are shown in table 12.4.The quality of the KU- PEDES loan portfolio from 1984-96 can be seen in table 12.5. 252 The Microfinance Revolution: Lessons from Indonesia Table 1 2A Number and value of KUPEDES loans, 1984-96 Number of outstanding Value of outstanding loans Year loans (thousands) (millions of U.S. dollars) 1984 641 103 1985 1,035 204 1986 1,232 204 1987 1,315 260 1988 1,386 313 1989 1,644 471 1990 1,893 727 1991 1,838 731 1992 1,832 799 1993 1,896 928 1994 2,054 1,117 1995 2.264 1,383 1996 2,488 1,710 Source BRI unit desa monthly reports Table 12.5 Loss ratios for KUPEDES loans, 1984-96 (percent) Year Long-term Twelve-month Short-term Portfolio status I 1984 0 98 0 98 1 40 0.54 1985 1 66 1 84 2 42 2 08 1986 2 23 2 65 2 83 4 50 1987 2 56 2 99 4.83 5 80 1988 323 455 1 11 742 1989 2 92 2.28 0 32 5 39 1990 262 201 1 56 412 1991 3 29 4 86 2 76 8 55 1992 3 32 3 40 3 27 9 10 1993 3 07 2 15 1.23 6 46 1994 2.59 0 70 0.50 4 47 1995 2 29 1 09 1.93 349 1996 2 15 1 59 1 30 3 65 Note See box 12 3 for definitions of the four loss ratios Source BRI unit desa monthly reports Success in Microlending The KUPEDES Credit Program, 1984-96 253 Outreach and repayment history In the program's first seven years the volume of KUPEDES lending grew rapid- ly and the quality of the loan portfolio remained high. Early problems were quwckly identified in part through a three-year, three-round survey of KUPEDES implementation in 1 percent of the unit desas.29The early implementation dif- ficulties included insufficient information provided by the head office to the branches and units, some overly risk-averse branch managers who permitted few KUPEDES loans to be made, rigid enforcement of collateral rules- especially insistence on full land title, and locally imposed restrictions on lend- ing to particular kinds of borrowers. Except for the reform or retirement of intransigent or incompetent branch managers-which took much longer-most KUPEDES and of these problems were overcome within a few years. By 1990 there was $727 million in 1.9 million outstanding loans and the long-term loss ratio (see box unit desa savings are 12.3) was 2.6 percent. The slow growth in the amount of KUPEDES outstanding during 1991-92 nationwide was caused primarily by credit restrictions at BRI that resulted from a tight money policy imposed by the government. This policy was imposed from June 1990 programs, reaching until April 1992 to control inflation, which had risen from 6.4 percent in 1989 to 9.4 percent in 1991 (Cole and Slade 1996, pp. 56-65). During this period many clients in the interest rates on deposits increased sharply and the government put strong pres- sure on banks to slow lending. The number of outstanding KUPEDES loans Outer Islands as decreased slightly in 1991 and 1992, and the rupiah growth of KUPEDES cred- it was limited during those years. well as onJava and Also largely as a result of the tight money policy, KUPEDES arrears began to increase in late 1991 and continued to do so until early 1993 (see table 12.5). Bali There were several reasons for the repayment problems of this period. First, some large suppliers and buyers, unable to obtain credit at urban bank branches, de- creased or ceased business activities with their local trade partners, many of whom were KUPEDES borrowers.This resulted in cash-flow problems and lower pro- duction for some unit desa borrowers. Second, some KUPEDES borrowers, concerned that they might not receive new loans, delayed payment on their current loans.Third, a prolonged drought was followed by heavy flooding and a sharp increase in pest problems in 1991-92; in many areas crop losses were high. However, BRI-because of its unit desa system-weathered the tight money period better than did most other banks.There were two important rea- sons, both of which foreshadow aspects of the crisis years beginning in 1997 (see chapter 15). Small savings accounts tend to be stable; unit savers do not take their accounts offshore and BRI had less of a liquidity problem than other banks faced. In addition, the KUPEDES interest rate had a high enough mar- gin to enable the unit desa system to remain profitable in 1991-92 without in- creasing its lending rate during this period of higher interest rates on savings and lower than usual repayment rates. By 1992 inflation was down to 7.6 percent; the monetary restraints were lifted in April 1992.The 1992/93 rice harvest was a good one, and the effects 254 The Microfinance Revolution: Lessons from Indonesia Tabie 12.6 Geographic distribution of KUPEDES outstanding loans and unit desa deposits, 1996 Indicator Java and Bali Outer Islands Total KUPEDES outstanding loans Number (millions) 1 8 0 7 2 5 Share (percent) 71 29 100 Value (billions of U S dollars) 1 17 0 54 1 71 Share (percent) 68 32 100 Long-term loss ratio 1 7 3 5 2 2 Unit desa deposits The unit desas' Number of accounts (millions) 10 5 5 6 16.1 Share (percent) 65 35 100 return on assetsfor Value (billions of U S dollars) 1 93 1.04 2 97 Share (percent) 65 35 100 1996 was 5. 7 Population (millions)a 117 6 77 2 194 8 percent-much Share (percent) 60 40 100 Note The distribution of loans and deposits shown in table is approximate, the Outer Islands are higher than the slightly underrepresented BRI aggregates reported totals by its 15 regional divisions Two of these divisions-Jakarta and Denpasar-are also responsible for loans and deposits from some regions in industry average in the Outer Islands that are included in their totals a Data are for 1995 Indonesia Source BRI data, Government of Indonesia 1996 of the tight money policy had lessened considerably by early 1993. KUPEDES arrears had also decreased by early 1993; the 12-month loss ratio was 2.15 per- cent for 1993 and 0.70 percent for 1994. Overall, during 1984-96 BRI's unit desas lent out $11.1 billion (after adjusting for inflation)30 in 18.5 million KUPEDES loans. During 1996,3,595 unit desas made 1.9 million loans, for an average of 158,000 loans per month. At the end of 1996, there was $1.7 blllion in 2.5 million outstanding KUPEDES loans.The average loan balance was $684 and the long-term loss ratio was 2.15 percent. Geographic coverage The geographic distribution of KUPEDES outstanding loans and unit desa de- posits in 1996 is shown in table 12.6.Java and Bali (with 60 percent of Indonesia's population) are somewhat overrepresented in the number of unit desa savings accounts (65 percent) and loans (71 percent), as well as in the value of savings (65 percent) and outstanding loans (68 percent). But KUPEDES and unit desa savings are clearly nationwide programs, reaching clients in the Outer Islands as well as on Java and Bali. Success In Microlending The KUPEDES Credit Program, 1984-96 255 Unit desa Table 12.7 profits and losses, 1984-96 Profit or loss Profitable units Loss-making units Retum on (millions of U.S. Share of total Share of total assets Year dollars) Number (percent) Number (percent) (percent)a 1984 -23 336 136 2,133 864 - 1985 -08 1,192 48.3 1,277 51 7 - 1986 6 1,647 72 5 626 27 5 - 1987 14 1,887 80 6 454 19 4 1988 18 2,090 80.9 495 19.1 - 1989 21 2,253 79 2 591 20 8 - 1990 34 2,708 89 1 332 10 9 30 1991 33 2,696 84.0 514 160 27 1992 41 2,744 85 9 450 14 1 26 1993 66 2,918 893 349 107 33 1994 121 3,173 93.7 215 63 5 1 1995 174 3,361 95 7 151 43 65 1996 177 3,412 949 183 5.1 5.7 - Not available a The return on assets calculated by BRI for the unit desas is a pretax ratio Although BRI pays taxes, as a bank division the units do not Source BRI unit desa monthly reports The long-term loss ratio is higher on the Outer Islands (3.5 percent) than on Java and Bali (1.7 percent), reflecting the difficulties inherent in building effective management and supervision systems covering numerous units in outlying loca- tions-especially on the world's largest archipelago. Still, with the KUJPEDES over- al long-term loss ratio of 2.2 percent in 1996, the problem is not severe. Unit desa profitability In 1983 there were no profitable unit desas, and BRI wanted to close the system because of its high losses.Table 12.7 shows the percentage of profitable units from 1984 (13.6 percent) to 1996 (94.9 percent; most of the loss-making units in 1996 were ones that had been only recently opened). The table also shows unit desa profits and losses since KUPEDES began in 1984. In 1996 the unit desa system earned $177 million in pretax profits after covering all costs, including the com- mercial cost of funds, overhead costs (including the costs of supervising the units and the costs of training unit staff and their branch-level supervisors), and loan risk. Because the units are part of BRI, all profits m the unit desa system during the previous year are transferred to BRI's general account at the end of each year.31 The unit desas' return on assets for 1996 was 5.7 percent,32 much higher than the industry average in Indonesia. Because, as a division of BRI, the unit desa system does not pay taxes, BRI calculates the unit desas' return on assets as a pretax ratio.The industry average (after tax) was 1.5 percent in 1996. 256 The Microfinance Revolution: Lessons from Indonesia The units' profits come from two main sources: revenues from KU- PEDES and interest on unit savings deposited at the branches. For the unit desa system in 1996, interest revenue was calculated at 23.1 percent of aver- age earning assets. Interest expenses (10.9 percent) and operating expenses (6.5 percent) totaled 17.4 percent, leaving a net profit of 5.7 percent (Mau- rer 1999, p. 132). In 1996 the unit desas accounted for 17 percent of BRI's rupiah loan port- folio and 42 percent of its rupiah deposits. The units' $177 million in profits also covered the losses of the rest of the bank that year ($33 million)-remarkable results for a system BRI had wanted to close just 13 years earlier. In competitive financial systems, a 1 percent return on assets is But the considered an indication of sound financial performance; figures of 2-3 percent are often recorded in the better performing extraordinarily high commercial banking systems in emerging commercial markets ... Whatever common financial indicator is used to assess their profits of the unit performance, the units have earned returns on rural financial intermediation that are well above those in the banking industry desa system cross- . . . In 1995 [the unit desas] achieved a negative subsidy dependence index; the units could have reduced the yield on subsidized wealthier their loan portfolios by 44.5 percent (from 31.6 to 17.5 percent) and still remained independent of subsidies and earned an borrowers in the adequate market rate of return on equity . . . A negative subsidy dependence index of 44.5 percent has no equal in rural or microfinance bank' other ... This indicates the "real"profitability that has resultedfrom effective ruralfinance intermediation and underscores the tremendous potential divisions for efftcient and profitable ruralfinance in other countries. -Yaron, Benjamnu, and Charitonenko 1998, pp. 165-68, emphasis added The complex issues of the extraordinarily high profits of the unit desa system-and the use of these profits to cross-subsidize BRI's wealthier borrowers in the bank's other divisions-are discussed in later chapters. But the unit desa achievements show clearly that microfinance can achieve-simultaneously and continuously-wide outreach and high profitability. However, there is still large unmet demand for small loans in Indonesia. BRI's microbanking experience is of great importance for the microfinance revolution. Considering the high costs of borrowing from moneylenders (see chap- ter 6), there is no doubt that microcredit can be profitably delivered at a fraction of the cost that many low-income borrowers pay informal lenders. In addition, the unit desas have demonstrated dramatically that mucrofinance institutions can provide savings services profitably to millions of savers, while simultaneously fi- nancmg their loan portfolios with locally mobilized savings. Since 1989 KUPEDES has been fully funded by unit desa savings. But the symbiotic relationship between savings and lending in the units extends well Success in Microlending The KUPEDES Credit Program, 1984-96 257 beyond the fact that their savings finance their lending. The relationship exists at multiple levels: * Most borrowers are also savers. * Many microenterprises are started with savings and expanded through loans. * Some small savers in the units have developed the confidence they needed to become KUPEDES borrowers. * Savings accounts provide the units with good information about the savers' creditworthiness-which is very useful when savers want to borrow. * Borrowers know that the units, with their savings, are not capital con- strained and that they can borrow again if their enterprises qualify and their The unit desas' repayment records are good-and that to a large extent they can thus con- trol their future credit options. (This point is crucial for understanding the profitability performance of the unit desa system during the recent crisis; see chapter 15.) "underscores the Unit desa clients know that their financial needs are met by a full-service bank, with efficient staff who understand that small business is business and who tremendous potential are friendly, helpful, and respectful to their customers.The development of KU- PEDES and the growth of savings in the unit desas are closely intertwined.The for efficient and ways in which the unit desas increased their savings from about $18 million at the time of the June 1983 deregulation to $3 billion in December 1996 are profitable rural discussed in the next chapter. These are two parts of the same story; one can- not be fully understood without the other. finance in other countries" Notes 1.The unit desas generated $177 mnllion in profit in 1996.The losses of the sub- sidlized programs are not provided because the branches do not report profit and loss by program. But the branches lost $33 million that year. 2. During 1979-90 extensive field work on local finance was conducted in near- ly all Indonesian provinces as part of the development of BRI's unit banking system This research was supported by the Ministry of Finance and was conducted first by Development Program Implementation Studies (DPIS) and then by its successor, the Center for Policy and Implementation Studies (CPIS), in cooperation with BRI. Both received technical assistance from the Harvard Institute for International Development (HIID), funded by the Ministry of Finance. 3.The new president-director, Kamardy Arief, came to BRI from Bank Indone- sia, where he had been responsible for rural credit, including the country's subsidized rural credit programs. His attitude toward credit subsidies changed dramatically soon after he was selected to lead BRI's reforms. Commenting that "where you sit has a profound effect on what you think about," he provided exceptional leadership in the unit desas' transition to commercial financial intermediation for low- and lower-nud- dle income people. 4.This quotation comes from a 1987 Bank Indonesia senunar on a case study of savings and credit inYogyakarta and Central Java, held in collaboration with the Pusat Penelitian Pembangunan Pedesaan (Center for Rural Development Studies) and Gad- jah Mada University. 258 The Microfinance Revolution: Lessons from Indonesia 5.This section is based on field research in nine provinces by DPIS from 1979-82 and on field research by BRI and CPIS in nearly all provinces from 1983-90 (see note 2). 6.This perception was deeply entrenched despite the 1980 census report that 47 percent of household heads in rural Java did not work, rent, or sharecrop agricultur- al lands at their own risk. Of the 53 percent who did, nearly half (43 percent) worked land of less than 0.25 hectares. 7.This section is partly based on DPIS (1983), HIID's Rural CreditWorking Group policy memorandums, and Robinson (1997c, 1998a). 8. As coordinator of HIID's DPIS Rice Intensification Project, I drafted this re- port, which was based on the work of a great many people. 9.The report also provided recommendations on debt relief policles and their im- plementation in cases of substantial crop loss. 10.The original members of the Rural CreditWorking Group wereJamesJ. Fox, Donald R. Snodgrass, and I; we worked closely with Marzuki Usman of the Ministry of Finance. In late 1982 the group was joined by Richard H. Patten, and in 1983 by David 0. Dapice and C. Peter Timmer. 11. In 1983 average annual mflation was 11.8 percent, and in 1984 it was 10.4 per- cent. In all succeeding years through 1997 inflation was below 10 percent. 12. Effective KUPEDES interest rates vary somewhat depending on loan matu- rity and payment schedule. 13.The practice of simultaneous implementation of opposite approaches through different channels is common in Indonesia. Two other examples come from BIMAS. The village cooperatives (Koperasi Umt Desas, or KUDs) were established as the sole suppliers of subsidized inputs to BIMAS participants. Severe problems caused by this approach were perceived early, however, and in 1976 the government permitted wide- spread private participation in the distribution of subsidized fertilizer, except through the BIMAS program.After the 1976 decision the share of KUDs in the rapidly grow- ing market for fertilizer plunged from 74 percent in 1975 to about 18 percent in 1982, whle fertilizer use rose rapidly. Sinilarly, although it had been planned that KUDs would be the primary agents for purchases of paddy from farmers, this marketing plan did not work well.When it was discovered that most paddy was purchased from farmers by private traders, bypassing KUDs, it was decided to continue the marketing func- tion of KUDs, which were typically run by local elites, but to allow farmers to mar- ket their paddy as they wished. 14.These figures are not adjusted for inflation because I do not have the KUT loan amounts and overdues by year. However, average annual inflation was single digit in all years between 1985 and 1990. 15. For dlscussion of KUPEDES, see Sugianto (1989, 1990a, 1990b); Sutoro and Haryanto (1990); Robinson (1991, 1992b, 1994a, 1995b, 1995c, 1997c, 1998a); Snod- grass and Patten (1991); Patten and Rosengard (1991); Gonzalez-Vega (1992); Mar- tokoesoemo (1993); BRI (1996a, 1996b, 1997a, 1997b, 1998b); Boomgard andAngell (1994); Hook (1995); Moeljono (1996); Mosley (1996);World Bank (1996);Yaron, Ben- jamin and Piprek (1997);Yaron, Benjamin, and Charitonenko (1998); Charitonenko, Patten, andYaron (1998); Institute for Development of Economics and Finance and BRI (1999); Patten, Rosengard, and Johnston (1999); and Maurer and Seibel (forth- coming). For additional references, see volume 1, chapter 2, note 16. 16.Additional fundmg for KUPEDES was later supplied by two World Bank loans- one in 1987 for $97 milhon and one in 1990 for $120 nmllion. But because these loans were made after the units had already moblized substantial savings, they were not need- ed to finance the KUPEDES portfolio. By 1989 the units had more savings than out- standing loans, and the World Bank loans were used primarily for other BRI purposes. Success in Microlending The KUPEDES Credit Program, 1984-96 259 Nevertheless, the international recognition that these loans conferred on the units was important politically in Indonesia, and the review procedures for KUPEDES perfor- mance instituted under the loan agreements helped to encourage transparency dur- ing the expansion of KUPEDES. The first loan, signed in May 1987, was for $96.7 million; the interest rate was set at the average rate of all BRI interest-bearing accounts. To encourage savings mobilization in the unit banks, a limit was placed on theWorld Bank loan funds that could be drawn: 65 percent of the increase in the amount of out- standing loans in the first year and 60 percent in the second year. The second loan, signed in August 1990, was for $120 mullion; funds could be drawn monthly at 5 per- cent of the loan or 30 percent of the increase in the amount of outstanding loans dur- ing the preceding month (whichever was lower).The interest rate for the second loan was set at the three-month rate of the central bank's Sertifikat Bank Indonesia (SBI), or the average rate of the five state-owned banks for three-month fixed deposits (whichever was lower) Both loans were made for 20-year durations, including a grace period of five years. In addition, in 1989 the Export-Import Bank ofJapan pro- vided a $28.2 mullion loan for KUPEDES. The interest is the average rate of all BRI interest-bearing accounts, and the loan was made repayable semiannually over a 15- year period starting in September 1991. 17.This section and the one that follows draw on two BRI umt desa information modules (BRI 1996a, 1997b) prepared under a BRI-U.S.Agency for International De- velopment cooperative agreement for the InternationalVisitors Program, with techni- cal assistance from HIID. In addition, Richard M. Hook, Don E.Johnston Jr., and Patricia Markovitch provided helpful information. Except where specified otherwise, the dis- cussion of KUPEDES characteristics, implementation, outreach, and performance in this chapter refers to 1996. KUPEDES in 1997-2000 is examined in chapter 15. 18.This percentage breakdown probably understates the amount of KUPEDES loans for investment capital. Because less documentation is required for working cap- ital loans, these loans tend to be preferred. Some borrowers use working capital loans at least partly for investment purposes. 19. KUPEDES working capital loans were not subsidized, but a 3 percent liquidlty credit was provided by Bank Indonesia for KUPEDES investment loans. 20. In 1990 working capital and investment loans were both provided at a 1.5 per- cent a month flat rate for the first 3 million rupiah and a 1.0 percent a month flat rate for the amount above 3 million rupiah. In 1991 the KUPEDES interest rate was changed to a 1.5 percent a month flat rate for working capital and mvestment loans of all amounts. Subsequent interest rate changes have applied to both types of loans. 21.The effective interest rate would be somewhat higher if the borrower's loss of the use of the prompt payment incentive funds during the loan period were includ- ed in the calculation. 22. In 1983, when the Rural Credit Working Group calculated the interest rate that would be needed for a two-year break-even period, we did not have a good es- timate of the portion of borrowers who would repay on time and have their incen- tive payments returned, and those who would repay late and forfeit their payments. We estimated, in a way that we thought would be conservative, that a relatively high number of borrowers would be late on at least one payment. But with KUPEDES repayment rates at 98 percent or higher during 1984-85, we had to recalculate and show that KUPEDES nught break even somewhat later than predicted since a small- er amount would be forfeited to the bank than had been estimated. In the end, however-even with a high repayment rate-the unmt desa system broke even in just under two years, as estimated. 23. In 1994 the unit desas began a pilot project, KUPEDES Skala Kecil (Small- scale KUPEDES) offering small loans without collateral. Loans for less than 500,000 260 The Microfinance Revolution: Lessons from Indonesia rupiah ($210 in 1996) were provided for maturities up to one year. In August 1997 KUPEDES Skala Kecil, operating in 432 unit desas in four BRI regional offices, had $1.1 million in 10,891 outstanding loans, for an average loan balance of about $100. Portfolio status (overdue principal installments relative to total principal outstanding) was just 0.59 percent. Noncollateralhzed small loans have since been incorporated to some extent into KUPEDES lending. But there is still considerable scope for expan- sion of this approach. 24.A fifth measure, portfolio at risk, is not used at the units although BRI has said that it might be a useful addition (BRI 1996a, p. 11). Portfolio at risk is a measure- ment of portfolio quality defined as the total outstanding balance of loans with late payments divided by the total outstanding balance of the loan portfolio. 25.This section draws in part on BRI (1996a, 1997b). 26.The 1988 survey had two components: a survey of 1,404 respondents and a smaller, more detailed study of 192 respondents drawn from the larger survey.The ref- erences made here refer to the smaller of the two surveys (which contain most of the survey questions) unless otherwise specified. 27. Because the sample consisted only of current KUPEDES borrowers, no in- formation is available about previous KUPEDES borrowers who did fiot have out- standing loans at the time of the survey. 28.This account is from an interview conducted by BRI's International Visitors Program in 1997. 29.The three rounds, carried out by BRI, CPIS, and HIID teams, surveyed 36 unit desas in 19 BRI branches in 9 provinces. See Patten and Snodgrass (1987). 30. See volume 1, table 2.3, note a for method of calculation. 31. Because the units do not retain their earnings from year to year, return on as- sets is generally used to measure the system's profitability. However, see Yaron, Ben- jamun, and Charitonenko (1998) for an analysis of the unut desas' return on equity from 1990-95, based on start-of-year equity. "The [unit desa] system has been exception- ally profitable by any banking standards.Whereas banks in low-mflation countries might earn 15-20 percent (after tax) on their average annual equity, BRI-UD [unit desa sys- tem] earned more than 60 percent on its average equity in 1990 and 1991. By 1995 this figure had more than doubled to 136 percent.When net income is measured against start-of-year equity, the return on equity increases to about 90 percent in 1990 and 1991, rising to an astounding 407 percent by 1995 (a net income of 403 billion [ru- piah] on a start-of-year capital of 99 billion [rupiah])" (p. 164). 32. The decrease in the unit desas' return on assets from 1995 to 1996 was due primarily to an increase in loans with the new lower interest rates in effect in 1996, and to an increase in the interest paid on depositors' savings. Success in Microlending The KUPEDES Credit Program, 1984-96 261 G ' Mobilizing Massive Savings: Bank Rakyat Indonesia's Unit Desas, 1984-96 In 1984, as the KUPEDES commercial credit program got under way, Bank Rakyat Indonesia's (BRI's) managers and advisers tried to answer three questions. Could the bank learn in what forms and for what purposes low-income house- holds in developing countries save?' Could it provide sav- ings instruments and services that meet the needs of low-income savers better than the savers could meet their needs themselves? And could it price those instruments and services in such a way that savings mobilization finances the loan portfolio and contributes to unit desa profitability? In hindsight, the answer to all three questions was yes. After a decade of offering savings accounts, BRI's unit desa system had only $18 million in savings when finan- cial sector deregulation began in June 1983 (see chapter 12). 262 This low level of deposits was widely-but wrongly-attributed to an assumed lack of demand for financial savings instruments: it was believed that rural people were unable to save, unwilling to save in financial form, and did not trust banks. In fact, the problem was with the banks (which did not understand the nature or the extent of the demand for rural savings services) and with the government (whose regulations made it impossible for banks to meet this de- mand profitably). The Harvard Institute for International Development's (HIID's) 1983 Development Program Implementation Studies (DPIS) report It was uwdely on rice intensification and the role of the unit desas (see chapter 11) recommended that in conjunction with the introduction of KU- believed that rural PEDES, the unit desas greatly increase their efforts to mobilize eople were rural savings. The savings program was developed in three phases: P research and design of a new savings program and testing of the pro- unwilling to save in gram in a two-stage pilot project (1984-85), the phased expansion of the new program throughout the country (1986), and develop- financialform. In ment of methods for market penetration (1987-89). Figure 13.1 shows how unit savings developed at each of these fact, theproblem was stages.The effects of this effort to build a high-quality savings pro- with the banks and gram are shown there in detail. By the end of 1989 the unit desas had designed and tested appropriate products, established an orga- the government nization able to deliver these, and developed effective techniques for penetrating the market for savings mobilization. In 1989 all KU- PEDES loans were funded by unit savings, as they have been ever since. Thus it took six years from the first stage of demand research to a successful savings program implemented nationwide. After 1989 the units achieved steadily increasing penetration of the Indonesian market for rural savings and, with the opening of urban units in 1989, urban savings in low-income neighborhoods. Between 1989 and 1996 the value of unit savings increased by more than 450 percent, from $534 million to $3 billion. (Average annual inflation was in the single digits throughout that period.) The number of sav- ings accounts increased 155 percent, from 6.3 million in 1989 to 16.1 million in 1996. Even when the country entered a period of severe crisis in 1997, the number of savers and the amount of savings in rupiah terms continued to grow steadily (chapter 15). Mobilizing Massive Savings Bank Rakyat Indonesia's Unit Desas, 1984-96 263 Figure 13.1 The evolution of microsavings in the unit desas, 1984-96 3.0 Billions of current U.S dollars Penetrating the market 2.5 All loans 2.0 funded by- KUPEDES 1.5 Market penetration methods 1.0 developed Expansion 0.5 Two-stage to all units pilot project 0. 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 Source BRI unit desa monthly reports Traditionally much of the savmgs of rural Indonesian households was held in non- financial forms such as grain, gold, and animals (see chapter 7). Many villages had rice banks where rice could be saved and borrowed. But as the economy became more monetized, it also became common to hold cash savings in the house. Al- though people continued to save both in kind and in cash, many said that they would prefer to hold some of their savings in financial institutions if secure, con- venient institutions and appropriate products and services were available. Their reasons centered around the problems associated with savmg in other forms. Gold and cash present security risks. Gram is too apt to be borrowed. Caring for more than a few animals represents an unacceptable opportunity cost to many house- holds, especially those m which the adults are workmg and the children are m school. Raw materials (wood, cloth, leather) for mucroenterprises take up space and may deteriorate in storage. And land purchase often requires a long wait before an ap- propriate match is found among land size, type, location, and sellng price. Problems of opportunity cost and liquidity were clearly perceived. For ex- ample, in densely populated areas where fodder was scarce and space limited, taking care of more than a few animals was considered too time consuming. As one Javanese farmer said, "Now we have more work opportunities than in the past, and the shepherds are all in school. If we put all our savings in ani- mals, who will take care of all of those animals?"Another commented, "When you have to pay the school fees, you cannot sell the cow's leg." 264 The Microfinance Revolution: Lessons from Indonesia Informal financial savings opportunities abound in Indonesia, in ubiquitous rotating savings and credit associations (ROSCAs, known as arisan) and other informal savings and credit associations.These were, and are, very popular. But as a savings mechanism, they too have problems. Returns are often unavailable or uncertain. In addition, the members' funds are, to varying degrees, illiquid, and the amount of their savings is publicly known.There is also a risk that mem- bers who have received their payout will not continue to make their payments (chapter 7). Another method of savings is to store funds with a member of the com- munity.Thus some farmers store excess liquidity with their commodity buyers, employees with their employers, and sharecroppers with the owners of the land they cultivate.Again, however, this approach carries some risk and generally pro- It took six years vides no returns. Savers could also put their savings in Bank Perkreditan Raky- at (People's Credit Banks, or BPRs; see chapter 9). But BPRs were not located from thefirst stage in all areas of the country. In addition, while many BPRs were financially self- sufficient, many others had high default rates, putting savers' money at risk.Thus of demand research until 1984 there was no uniformly available method of saving in rural Indone- sia, formal or informal, that was generally suitable for rural demand. to a successful How did the units move from dismal failure in mobilizing savings during 1970-83 to spectacular success since 1984? The basic principles of unit desa savings program savings mobilization are building trust; pricing for profitability; offering savers a combination of security, convenience, liquidity, confidentiality, service, and re- implemented turns; and providing staff with training and incentives-while holding them re- sponsible and accountable for performance (box 13.1).The rest of this chapter nationuide considers how these principles were developed and implemented in the unit desas. Why Did BRI Emphasize Savings in Its Unit Desas? There were suggestions in 1983 and 1984 that with the new KUPEDES com- mercial loan program, the unit desas could become self-sustaining even with- out substantial savings mobilization.There was also concern that it would be too difficult for staff to implement savings mobilization and KUPEDES si- multaneously. In addition, many people had strong doubts that there were rural savings worth mobilizing; others thought increased collection of rural savings would be too costly and couldjeopardize the viability of KUPEDES. But there were three main reasons that the government decided that the credit and sav- ings components of the newly commercialized unit desas would be planned and implemented together. To finance the demand for credit and enable institutional sustainability In the early 1980s the Indonesian government was anticipating declining oil revenues and seeking increased private savings and investment. In this context it was decided that extensive long-term funding for a nationwide rural credit Mobilizing Massive Savings Bank Rakyat Indonesia's Unit Desas, 1984-96 265 Box 13. Basic principles of unit desa savings mobilization The fundamentals of unit desa savings mobilization are building trust, offering savers se- curity, convenience, liquidity, confidentiality, service, potential access to loans, and re- turns, pricing for profitability, and providing staff with training and incentives, and holding them responsible and accountable for performance Secunty, convenience, service, and outreach * The bank must be, and must be perceived as, trustworthy * The location of the units and their working hours are arranged to be convenient for clients * Clients are treated with respect and given high-quality service "Now the shepherds * Savings are collected from the public Products and pncing are all in school. If . A choice of several savings instruments with different ratios of liquidity and returns is available, savers maintain as many accounts of whatever types they wish we put our savings * Labor-intensive, and therefore costly, savings products have interest rates that are tiered depending on the minimum monthly balance in the account The rates range from no interest for very smnall accounts to market rates for the largest accounts in animals, who will . The interest rates of savings instruments are set so that the spread between the blended cost of funds of the savings products and the KUPEDES loan interest rate is take care of the maintained at a level that permits unit desa profitability * Units with excess savings deposit them at their supervising branch and receive inter- est on these deposits, units in which savings are insufficient to finance loans borrow animals" from their branch and pay interest The transfer price-the interest rate for the units' saving and borrowing at the branches-is set monthly by BRI The transfer price serves as a signal from the head office to indicate when the relative emphasis on sav- ings and lending needs to be shifted. For example, if BRI wants the units to mobilize more savings or to slow the rate of growth of lending, the transfer price is raised * Cash management is implemented so that cash is routinely available at the units to meet withdrawals and to finance loan disbursements * Some savings products permit accounts to be held in the name of institutions or organizations * Lotteries with attractive prizes are provided for some savings instruments, with free lottery tickets given to savers in proportion to the minimum monthly balance in their accounts The drawings are used as vehicles for publicity * Savers may use their accounts to build credit ratings, savings accounts can also be used as collateral for KUPEDES loans. * Savers' accounts are kept confidential Staff training and incentives * Unit staff and their branch supervisors are well trained in learning from the local popu- lation, in systematically identifying potential savers, in helping clients use the different products offered at the units, and in maintaining good long-term relationships with savers * Unit staff and branch supervisors are provided performance-based incentives, both in cash and in institutional recognition program could not be committed from Bank Indonesia, the central bank. In- stead KUPEDES would be financed by locally mobilized savings. It was cor- rectly anticipated that meeting the country's demand for microcredit would 266 The Microfinance Revolution: Lessons from Indonesia eventually require far more funds than had been required for all previous rural credit programs combined, and that this demand could be profitably financed with locally mobilized deposits. To limit government risk The proposed rural credit program was considered too risky for long-term com- mitment of central bank funds. It was well known that many rural credit pro- grams in Indonesia and other developing countries had a history of high defaults and losses. No sustainable large-scale model of rural financial intermediation could be found in any developing country. Indonesia's economics ministers recognized the potential for a banking system that would provide commercial microfinance throughout the country and supported its introduction with government fund- "1hen you have to ing. But they also recognized the need to limit government risk. pay the schoolfees, To encourage rural savings mobilization As part of the wider policy of encouraging private savings, the government want- you cannot sell the ed to provide institutional capacity to meet what was accurately estimated as large potential demand for savings services in rural areas. It was thought that cow's leg" an approach that offered savings instruments specifically designed to meet local demand could generally provide customers with positive real returns, while also helping to build the long-term viability of BRI's unit banking system. Both results would contribute to economic development in rural areas. From the start the aim was to provide loans to low- and lower-middle-income borrowers, but to collect savings from the public-individuals ranging from the poor to the wealthy, as well as associations and institutions-that lived, worked, or operat- ed near a unit desa. Developing and Testing the New Savings Program KUPEDES began nationwide in February 1984 with the understanding that a new program of savings mobilization in the unit desas would follow.2 BRI de- cided that the savings program would be developed more gradually than KU- PEDES for two reasons. First, there was less urgency to begin quickly in savings mobilization and less information about demand. Second, given the unit desas' continuing losses and large fixed costs, the imminent termination of BIMAS, and the start of KUPEDES in all unit desas simultaneously, unit staff would have to devote their full attention in the first year to introducing the new credit program and taking on the responsiblities that accompanied it (see chapters 12 and 14). In 1984 BRI and the Center for Policy and Implementation Studies (CPIS) expanded their cooperation to include work on the new savings program.3 A small group began research and planning for the new savings program, while the rest of the unit desa management and staff worked to get KUPEDES under way. The new savings program was designed and tested in 1984-85 and im- plemented nationwide in 1986.There was considerable internal opposition dur- Mobilizing Massive Savings Bank Rakyat Indonesia's Unit Desas, 1984-96 267 ing 1983-84 to the start of a major savings initiative. But as unit desa deposits in the pilot projects began to increase dramatically in 1985, an old proverb came to be quoted in BRI: "The people who say it cannot be done should not in- terfere with the people doing it." Early field research Building on the DPIS and CPIS research on local finance that had been car- ried out for the Ministry of Finance since 1979, in 1984 BRI's unit desa sys- tem began to focus specifically on analyzing the demand for different kinds of savings instruments. CPIS, with assistance from HIID, conducted extensive re- search with BRI into formal, quasi-formal, and informal savings methods in In 1983 no rural areas. People in many parts of the country were asked what they liked and did not like about their present forms of savings. Individuals, owners and man- sustainable large- agers of small and microenterprises, and leaders of organizations were queried about their actual forms of saving and their preferred savings alternatives.4The scale model of rural aim was to design an institutional savings program that would meet the savers' demand in better ways than they could accomplish by themselves. financial These studies offered substantial evidence for two conclusions.5 First, eco- nomic growth and monetization had created a large, unmet demand for sav- intermediation could ings services. Many villagers wanted to save in banks and to convert some of their nonfinancial savings into institutional deposits-if suitable deposit prod- be found in any ucts and services could be made available. Second, there was extensive demand in rural areas for a liquid financial sav- developing country ings instrument, not then available at the unit desas. Tabungan Nasional (TA- BANAS), the national savings program administered by Bank Indonesia, was the only deposit instrument offered at the unit desas, and it limited withdrawals to two per month. When people were asked why they did not use TABANAS, which had been available through the unit desas since the early 1970s, the replies were nearly identical. From one end of the country to the other, respondents said that the restriction on withdrawals was unacceptable.As a Sulawesi woman said,"If there is a famuly emergency and we cannot withdraw our savings, then what will hap- pen to us?" Or as aJavanese farmer put it, "Suppose we had withdrawn twice already and then there was an emergency. We would be malu to ask to with- draw again." Malu, a word with a wide range of meanings in Indonesian and not easily translatable, essentially means to be ashamed.A poor saver would feel malu to request a forbidden withdrawal from a government bank. Many peo- ple said they would be afraid to request a third withdrawal, even in an emer- gency. In 1984 field work was carried out in six districts ofJava and Bali to begin identifying the needs of rural savers and analyzing their savings methods, in- come flows, and reasons for saving. The study encompassed both the demand for savings facilities and the villagers' views on saving in whatever local finan- cial organizations were locally available-ROSCAs, cooperatives, People's Credit Banks (BPRs), and the like. 268 The Microfinance Revolution: Lessons from Indonesia For the most part ROSCAs were viewed positively but peripherally. They were generally not considered suitable for long-term savings or, because of illhq- uidity and lack of confidentiality, for the main savings of a household or en- terprise. And it was recognized that they are risky. Some of the people interviewed had lost their savings in ROSCAs. Some local credit organizations required compulsory savings but were not permitted to mobilize voluntary sav- ings. Others offered voluntary savings, but many institutions were tainted by corruption and were not trusted. As a farmer in Central Java put it, "I would be afraid to put my savings in the vlllage cooperative. Maybe they would use it to cover their losses." Except for some of the People's Credit Banks and Bank Dagang Bali, there were few institutions that offered voluntary savings instru- ments for the public and were widely trusted. The aim was to BRI deposits, however, were implicitly guaranteed because BRI is a gov- ernment bank. Although there was no formal deposit guarantee, it was wide- design an ly believed that savings were safe at BRI because it is a state-owned bank. The CPIS/BRI report on the research findings stated: institutional savings Based on our impressions during the last five years of fieldwork:6 program that would (1) economic growth has far outstripped banking facilities in many rural areas; (2) considerable excess liquidity is presently found in meet the savers' the vlllages; (3) traditional forms of savings (animals, gold, grain) are often unsuitable for modern village conditions; (4) there is demand in better widespread demand for savings accounts that are secure, convenient, and liquid; (5) the BRI has a strong comparative advantage in the ways than they were rural areas . . . If attention is given to offering a wider range of financial instruments; if the unit desa system undergoes some able to accomplish reallocation of resources in order to improve the physical accessibility of banking services for customers; and finally if by themselves initiatives are taken in marketing, advertising, and training, the BRI unit desas should be able to raise sufficient savings to support the volume of lendmng required to make the system self-sufficient. -CPIS Rural Savings Report 1, 1984a, pp. 2-4. The planning stage Six main ideas emerged from the field research on savings and the early expe- riences of KUPEDES, and these were used in the design of the new savings program: * Liquidity is the primary key to mobilizing rural savings. * Developing clients' trust in the institution would be essential. * Security and convenience are crucial. * A combination of several savings instruments would be needed to meet the varied demand. * Deposits could be mobilized not only from individual households, but also from the many groups, associations, and institutions operating in rural areas. Mobilizing Massive Savings Bank Rakyat Indonesia's Unit Desas, 1984-96 269 * Deposit instruments offering different ratios of liquidity and returns could simultaneously meet demand and, in combination with KUPEDES, permit institutional profitability. The emphasis was to be on developing a microbanking system that would provide financial services nationwide and be stable, profitable, and viable for the long run. Savings and loans would be designed, revised, priced, and im- plemented together. Because in its post-1983 form the unit desa system was a large experiment with a number of unknowns, BRI decided that the spread between loan and deposit interest rates would be relatively large in the begin- ning, both to protect against risk and because it was anticipated that the sys- Few institutions tem would be somewhat inefficient at first. The design for the savings program incorporated four instruments.Village offered voluntary Savings (Simpanan Pedesaan, or SIMPEDES), a deposit instrument that would permit an unlimited number of transactions, was designated the flagship of BRI's savings instruments new rural savings program. SIMPEDES would be aimed at rural households, firms, and organizations that demanded full liquidity as well as returns. for the public and TABANAS would be continued, and Deposito Berjangka, a time deposit instrument previously available at BRI only through its branches, would now were widely trusted be offered at the unit desas as well. The fourth instrument, giro, a type of cur- rent account used primarily by institutions with special requirements, would be made available at the unit desas as well as at the branches. A transfer price mechanism was established. Because the unit desas are not independent entities, they cannot place their funds on the market. Instead units are required to deposit excess liquidlty in an interest-bearing account at their supervising branch; units in deficit borrow from their branch.The interest rate is set monthly by BRI's head office.To encourage savings mobilization, the trans- fer price was set at a rate higher than the highest time deposit interest rate; it would then be cheaper for a unit to mobilize savings than to borrow from its branch. The transfer price could stimulate savings mobilization only because the units were now treated as individual profit centers, and because staff incentives were provided for unit profitability. Previously, when there had been no pos- sibility of unit profitability, savings mobilization could not have been encour- aged through the transfer price. Over time the transfer price has served as a signal from the head office to indicate when the relative emphasis on savings and lend- ing is to be shifted, and incentives motivate staff to direct their efforts where needed. For example, during the tight money period in the early 1990s, as well as in the crisis that began in 1997, the transfer price was set high to encourage additional savings mobilization. The first SIMPEDES pilot project By mid-1984 enough progress had been made on the design of SIMPEDES to suggest that a pilot project should be the next step.The purpose of the pilot would be to test the instrument, to develop methods of estimating and locat- 270 The Microfinance Revolution: Lessons from Indonesia ing savings potential in the pilot area, and to learn how to train the staff in sav- ings mobilhzation. In order not to confuse issues in the pilot branch or over- load the unit staff, the introduction of the time deposit and giro accounts in the units was postponed until a later stage. TABANAS would be continued as usual in the units of the pilot branch; the pilot would focus entirely on SIMPEDES. The 27 unit desas of BRI's branch in Sukabunii District, West Java, were selected for the pilot because they were located in a largely rural district with- in a two-hour drive of Jakarta (important for supervision of the pilot by the head office); because they had an above average, but not exceptional, record; and because Sukabumi was a large branch with units located in different kinds of areas with multiple economic activities. Deposits could be Estimating the demandfor savings services. As a first step in preparing for mobilized not only the pilot project, CPIS/BRI teams visited Sukabunii villages and began de- veloping ways of estimating demand for unit desa savings instruments.7 Because from individual there was considerable skepticism among many BRI managers and staff about rural savings potential, a conservative approach was taken in estimating demand households, but also for BRI savings services.The research teams thought, correctly as it turned out, that even an extremely conservative approach would provide estimates so from the many much greater than current unit savings that BRI would realize that a major in- crease in savings was possible. groups, associations, Accordingly, it was assumed that: and institutions * On average, 10 percent of the income of the 40 percent of households in Sukabumi with the highest annual incomes would be available for savings operating in rural in some form, and that 20 percent of that amount (2 percent of their esti- mated annual income) would be deposited in the unit desas in the first year areas of the pilot project. * On average, 0.5 percent of the income of those households represented pre- sent savings that would be captured by the units within the first year. A more realistic but still conservative estimate of savings available to the units would have been based on perhaps 5 percent of the income of the 60 percent of the households with the highest incomes; it would also have included high- er deposits from stocks. But the aim was an estimate based on assumptions so conservative that there would be no objection to them at BRI. The subdistrict ofJampangkulon was selected for the study on rural de- mand for savings services.The study covered the 20 villages (14,398 households) that comprised the service area of the Jampangkulon unit desa. Two main data sources were used for theJampangkulon household income data; these were then cross-checked against each other.The first, the 1981 SUSE- NAS8 income figures for the 40 percent of households in rural West Java with the highest incomes, were used to estimate the income of Jampangkulon households with potential savings. As noted, it was assumed that 2 percent of Mobilizing Massive Savings Bank Rakyat Indonesia's Unit Desas, 1984-96 271 the income of those households would constitute the potential for unit desa savings. However, households with reported annual incomes below 300,000 ru- piah ($279) were excluded even if they were among the top 40 percent of in- come earners. Using this method, the estimate for potential umt desa savings from income flows in the 20 Jampangkulon villages was 59 million rupiah ($54,934) for the first year of the program. However, SUSENAS is believed to underreport income significantly, and it was realized that the actual savings po- tential was likely to be larger. The second source of data, a field study of 17 of the 20 Jampangkulon vil- lages, was undertaken by BRI/CPIS teams. In previous DPIS/CPIS field work in West Java, it had been learned that each village maintained a written list of The unit desas'first households classified into categories based on their incomes and assets.Assign- ments of households to particular categories were made by the officials of each years savings village and reviewed in a village meeting.The classifications were used to de- termine the annual contribution each household was required to make to the potentialfrom 20 village.These contributions were used to provide the matching resources from the vlllages that are required as a condition of the government's annual vlllage villages served was development grants (INPRES Desa). Different villages had developed different economic categories, but all the villages visited in West Java had such lists.9 estimated at nearly The research teams were able to obtain these lists from vlllage leaders. Both formal and informal village leaders were asked to estimate the average net in- 9 times the units' come for the households in each category on the village list.This approach typ- ically engendered discussion and debate among the leaders, and eventually a 1984 balance general consensus. Using these data, the teams estimated potential unit savings. In the 17 villages visited, 2 percent of household income flows amounted to 65.8 million rupiah ($61,266), compared with 59 million rupiah ($54,934) for all 20 villages, using the SUSENAS data. Given that the latter tend to under- report income, the results were very close. CPIS Rural Savings Report 2 (1984b, p. 10) commented, "We consider that these two methods [of estimat- ing income], one using micro data and the other using macro data, constitute a check on each other, and that the results agree." Finally, some additional conservative assumptions were made about other savings that could be mobilized by the unit desas from savings from stocks (es- timated, as noted, as 0.5 percent of annual household income) and the excess liquidity of local associations and institutions. Adding these to the estimates of potential household savings from income flows, the total savings potential for the 20 Jampangkulon villages for the first year was conservatively estimated at $78,000-nearly nine times the subdistrict's unit desa savings of $9,0321o on 31 September 1984. Starting the SIMPEDES pilot project. SIMPEDES, a fully liquid instru- ment with no limit on the number of transactions, was introduced in the unit desas in Sukabumi on 1 November 1984. It was, and remains, a unit desa in- strument; it is not offered to BRI's branch clients. The interest rate was set at 12 percent a year'l on the minimum monthly balance; this was lower than 272 The Microfinance Revolution: Lessons from Indonesia the 15 percent annual interest rate then paid onTABANAS deposits of 1 mil- lion rupiah ($931) or less.t2 But the results of the demand studies had con- vinced BRI that the liquidity of SIMPEDES would compensate for the difference in interest rates. In addition, SIMPEDES interest was compound- ed and posted monthly, whereas TABANAS interest was not compounded and was posted annually. Also in contrast to TABANAS, SIMPEDES permitted deposits to be made in the name of organizations and institutions. Field research had indicated that many members of formal and informal rural organizations would welcome the opportunity to deposit group funds in an account held in the name of the or- ganization.The members of such groups generally preferred this arrangement to the customary one under which jointly held savings were kept in the home Two months after of the group leader or treasurer, thus providing opportunities for (reportedly frequent) corruption. SIMPEDES was, therefore, designed in part to meet this the SIMPEDES latent demand from organizations and institutions. Lotteries, adapted from BDB, were held quarterly for SIMPEDES savers.The savings product was number of SIMPEDES lottery coupons a depositor held was determined by the minimum monthly balance in her account.The lottery prizes were awarded in introduced, a district-level ceremony; local dignitaries were invited and the festive setting was used as an occasion for BRI to provide information about the unit desa sys- Sukabumi' unit tem and its products and services.Although Bank Indonesia had held national- level lotteries for TABANAS for years, there was little interest in these because desa deposits had they were too remote; no one we interviewed knew anyone who had won a TABANAS lottery. No one even knew anyone who knew anyone who had won. more than doubled But a district-level lottery generated excitement because it was local. At the end of October 1984, after nearly a decade of offering TABANAS savings accounts, deposits totaled $161,060 in the unit desas of the Sukabumi branch. Shortly before the SIMPEDES pilot project pilot began, I asked a high- level Sukabumi branch officer how long he thought it would take for the Suk- aburmi unit desa deposits to double. He replied that if things went very well, deposits might double in five years. Early results of the pilot project in Sukabumi. The first pilot project in Suk- abumi quickly showed evidence of huge demand for the SIMPEDES instru- ment.13 By the end of December 1984, two months after SIMPEDES was introduced, Sukabumi's unit desa deposits had more than doubled, from $161,080 to $438,826-an increase of 172 percent (table 13.1). This rapid increase was caused, in part, by the fact that the timing of the pilot project coincided with the government's annual disbursement of village grants under INPRES Desa.'4 As part of the pilot project, these funds were, for the first time, deposited in the 162 village bank accounts that were opened in unit desas throughout Sukabumi (one account per village).The total amount deposited from this source was $150,837; the funds could be withdrawn by the villages inJanuary 1985. In addition, the government purchased land from Suk- abumi residents during this period, and $30,726 of the deposit total represents Mobilizing Massive Savings Bank Rakyat Indonesia's Unit Desas, 1984-96 273 Unit desa savings in the Table 13.1 Sukabumi pilot project, 31 October-31 December 1984 TABANAS SIMPEDES Total Amount Amount Amount (thousands of Number of (thousands of Number of (thousands of Number of Date U.S. dollars) accounts U.S. dollars) accounts U.S. dollars) accounts 31 October1984 1611 7,181 na na 1611 7,181 31 December 1984 153 0 7,030 2858 2,655 4388 9,685 Change (percent) -53 -2.1 n a n a 172 4 34 9 n a Not applicable Source CPIS Rural Savings Report 3, 1985a a few large deposits from the land sales. But even if these special sources of funds But it soon became are excluded, Sukaburmi's unit desa savings rose dramatically in two months: from $161,080 to $257,263, an increase of 60 percent. By the end of 1984, when evident that SIMPEDES was only two months old, it accounted for 65 percent of the value of savings and 27 percent of the number of savings accounts in the unit desas SIMPEDES in of the Sukabumi branch. To put the early pilot experience into perspective, it must also be noted that the pilot project was after the 1983 financial reforms, unit desa savings had increased elsewhere in the region as well, although to a lesser extent than in Sukabumi.The unit desas a highly labor- in the 30 other subdistricts of BRI's West Java regional office increased their savings from $3.3 million at the end of October 1984 to $4.3 million at the intensive- end of December 1984, an increase of 30 percent. It rapidly became apparent that the Sukabumi pilot project would be suc- and therefore cessful in mobilizing savings. Because of the popularity of the instrument, more SIMPEDES accounts were added daily. But it soon became evident that SIM- expensive- PEDES as implemented in the Sukabumi pilot project was a highly labor-in- tensive-and therefore expensive-instrument.5 There were many reasons for instrument the high labor costs including the numerous SIMPEDES transactions, the units' greatly increased bookkeeping and reporting, the work involved in the monthly compounding and posting of SIMPEDES interest, new arrangements for cash management and security, activities related to the SIMPEDES lottery, and promotional efforts. In order to close the books for the day, unit staff fre- quently found themselves working until late at night. Either additional staff would have to be employed at the units, raising costs, or the savings workload would cut into KUPEDES lending, lowering in- come and increasing risk. In either case profits would decline in the Sukabu- mi unit desas. It was decided, therefore, that SIMPEDES would have to be revised before it could be expanded to other areas. In keeping with BRI's new priorities on learning customers' views, it was decided to conduct additional research in Sukabumi villages to determine 274 The Microfinance Revolution: Lessons from Indonesia Table 13.2 The most important characteristic of SIMPEDES: responses of 144 SIMPEDES savers in the Sukabumi pilot project, January-February 1985 (percent) Account size "$47 >$47 All respondents Characteristic (N=120) (N=24) (N=144) Unlimited withdrawals 60 84 64 Maintaining at least the current interest rate 5 4 5 Lottery prizes 31 8 27 Interest posted monthly 4 4 4 In order to close the Source CPIS Rural Savings Report 4, 1985b books for the day, which aspects of the new instrument were most important to SIMPEDES savers, unit stafffrequently and which could be safely changed or eliminated. found themselves Back to the drawing board. CPIS/BRI teams surveyed 144 SIMPEDES savers in six Sukabumi unit desas inJanuary and February 1985 to determine which working until late at aspects of the instrument were important to savers and which were not. Table 13.2 shows the respondents' answers to a question asking what they thought night was the most important of four characteristics of the SIMPEDES instrument: its 12 percent annual interest rate, the lottery prizes, the unlimited number of withdrawals permitted, or its monthly posting of interest.The unlimited num- ber of withdrawals was considered most important by 64 percent of respon- dents (and by 84 percent of those with accounts over $47). Lottery prizes were next with 27 percent. Only 5 percent of the respondents said they considered the interest rate to be the most important feature of SIMPEDES. The 49 respondents who had both SIMPEDES and TABANAS accounts were asked which account they would prefer if they could have only one.Among this group, 42 (86 percent) said SIMPEDES, 2 (4 percent) said TABANAS, and 5 (10 percent) said they did not want to choose because they needed both. Additional research was undertaken in February 1985 to study the size dis- tribution of SIMPEDES accounts. 16 Because information on account size dis- tribution was not available from the monthly reports, the balances of 304 SIMPEDES accounts were obtained through a 10 percent random sampling of accounts in 13 Sukabumi unit desas. The results showed that 7 percent of the accounts contained 64 percent of the funds, while 70 percent of the ac- counts contained 9 percent of the funds. For BRI to set appropriate SIMPEDES interest rates, several factors had to be considered: * SIMPEDES costs are sensitive to average account size because of the high labor costs of the instrument and because the average account size for indi- Mobilizing Massive Savings Bank Rakyat Indonesia's Unit Desas, 1984-96 275 vidual savers in Sukabumi was low (in late 1984 and early 1985 it ranged from $47 to $57). * Most SIMPEDES account holders preferred the labor-intensive service as- pects of the instrument to the interest rates. * Since 7 percent of the accounts in the sample contained 64 percent of the funds, attention would have to be paid to possible interest rate sensitivity among the larger account holders. If a sufficient number of these were to close their SIMPEDES accounts because of a decrease in the interest rate, the average account size would decrease and SIMPEDES could become financially nonviable. SIMPEDES would Given the concerns about the cost of SIMPEDES, as well as the results of the Sukabumi studies, BRI decided to conduct a second-stage pilot project in have to be revised 12 additional districts on Java using a modified SIMPEDES instrument with a lower overall interest rate, fewer lotteries, and a number of cost-cutting ad- before it could be ministrative reforms.The second-stage pilot began inJuly 1985.The SIMPEDES pilot project in Sukabumi was also continued, making a total of 13 pilot expanded to other branches. areas The second-stage SIMPEDES pilot project The lower-cost SIMPEDES instrument was introduced in the unit desas of 12 branches in the provinces ofWest, Central, and East Java, and inYogyakarta."7 Instead of a 12 percent across-the-board interest rate, as in Sukaburni, a tiered interest rate was introduced. No interest was paid on accounts with minimum monthly balances below $22, 9 percent a year was paid on accounts between $22 and $177, and 12 percent on accounts over $177.TABANAS continued to pay 15 percent on all accounts below 1 million rupiah ($889) and 12 per- cent on amounts above that, as per Bank Indonesia requirements. Savers with accounts under $22 were informed that they could choose between the liq- uidity of SIMPEDES and the returns ofTABANAS. SIMPEDES lotteries were conducted in each pilot branch, but the frequency was reduced from four to two times a year, reducing the costs for prizes, the tax on prizes, coupons, and labor.A number of administrative changes were also made to reduce costs; for example, the lottery system was made more efficient, savers were charged a fee for the SIMPEDES bankbook, and cost-cutting changes were made in promotional activities. During the course of the second- stage pilot, further minor adjustments were made to the modified SIMPEDES instrument, and it was adapted in various ways to local conditions. In addition, Deposito Berjangka, a time deposit instrument previously available in the branches, was gradually introduced in the units of the pilot branches. Staff training in estimating and identifying potential unit desa savings was carried out in all the pilot branches. Outside West Java the villages visited did not have lists of households by economic category, so a (less accurate) method was devised to generate such lists. Formal and informal village leaders were gath- ered in what would now be called focus groups; the term was not then known 276 The Microfinance Revolution: Lessons from Indonesia 13.3 Savings mobilization in the second stage of Table the SIMPEDES pilot project, June-December 1985 (thousands of U.S. dollars) SIMPEDES as a share of unit SIMPEDES Unit savings savings in the 13 June December Increase June December Increase pilot branches, 31 1985 1985 (percent) 1985 1985 (percent) December 1985 428 4,564 966 5,479 11,405 108 40 0 Note The table presents aggregate totals from the 13 pilot branches (Sukabumi, which began as the first stage of the pilot project on 1 November 1984, Sleman, which began in the pilot on 1 February 1985, and the 11 other branches that began in the pilot on 1 July 1985) Source CPIS Rural Savings Report 7,1 986b to the researchers or BRI staff. In the discussion the village leaders were asked In the pilot 7 to estimate how many households in the vlllage "had more than enough in- come and assets to meet their needs," how many "had enough for their needs," percent of the how many "did not have enough for their needs," and how many "were very poor." In most cases this triggered discussion among the participants that led SIMPEDES to a detailed set of categories.18 The participants were then asked to estimate average income and income ranges for the different categories; savings poten- accounts held 64 tial was later estimated from this information. Of the 12 new pilot branches, one began on 1 February 1985 and the oth- percent of thefunds, ers began on 1 July 1985. Most got off to a slow start but were generally run- mng well by August or September. Deposito Berjangka (fixed deposits) and giro while 70 percent of accounts were introduced to the units, andTABANAS was continued.The total savings of the unit desas in the 13 pilot branches grew by 108 percent between the accounts held 9 30June and 31 December 1985, from $5.5 million to $11.4 million (table 13.3). By the end of 1985 SIMPEDES accounted for 40 percent of unit savings in percent of thefunds the 13 pilot branches (Sukabumu plus the 12 new ones). The tiered interest rates, fewer lotteries, and many cost-cutting adminis- trative measures of the modified SIMPEDES instrument were effective in lowering the overall cost of funds. In addition, the average SIMPEDES account ($47 at the beginning of the first-stage pilot project in Sukabumi) was con- siderably larger ($116) in the 12 new branches in the second stage of the pilot. The costs of SIMPEDES in the second-stage pilot project were similar to those of TABANAS: about 2 percentage points above the 15 percent cost of loan- able funds from Bank Indonesia in 1985. Moreover, there appeared to be con- siderable room for further efficiencies and cost decreases. In preparation for expansion of the new savings program, a CPIS/BRI study was carried out in the 12 new pilot branches in September 1985 to as- sess savers' reactions to the modified SIMPEDES product in particular, and to the unit desa savings program in general. Interviews were conducted with Mobilizing Massive Savings Bank Rakyat Indonesia's Unit Desas, 1984-96 277 Table 1 3A The most important characteristic of SIMPEDES: responses of 76 SIMPEDES savers in the 12 new branches of the second-stage pilot project, September 1985 (percent) Account size 2$90 >$90 445 >$445 All respondents Characteristic (N=14) (N=20) (N=42) (N=76) Unlimited withdrawals 93 65 79 77 Maintaining at least the current interest rate 7 5 7 7 Lottery prizes 0 25 12 13 Interest posted monthly 0 5 2 3 After 6 months Source CPIS Rural Savings Report 5, 1985c SIMPEDES accountedfor 40 169 umt desa savers in the pilot branches.The sample was structured to cover about 24 respondents in each of seven categories: large, medium-size, and small percent of unit SIMPEDES savers; large, medium-size, and small TABANAS savers; and time deposit holders. savings in the 13 As earlier in Sukabumi, questions were asked about savers' views on pos- sible changes in SIMPEDES terms.The responses were analyzed by size of ac- pilot branches count to learn whether there was a significant difference in views between savers with smaller and larger accounts. Since the latter group had to be retained if the former were to be served by a cost-effective instrument, this survey was crucial for BRI's decision about whether the modified SIMPEDES instrument was ready for nationwide implementation. Table 13.4 shows the responses of the 76 SIMPEDES savers in the sample to the question,"For your needs, what is the single most important feature of SIMPEDES?" Seventy-seven percent of respondents said that an unlimited num- ber of withdrawals was the most important aspect of the instrument. Liquidi- ty was especially important to the smallest savers, with 93 percent of those with account sizes of $90 or less selecting unlimited withdrawals as the most im- portant SIMPEDES feature. Only 7 percent of respondents cited interest rates as the most important aspect of SIMPEDES. Moreover, there was little variation in the answer across the account size distribution. However, another question asked was, "Would you still save in SIM- PEDES if the annual interest rate were reduced from 12 to 6 percent?" (table 13.5). Here 69 percent of the 74 people responding to this question said yes and 31 percent said no. The importance to the savers of the interest rate rose with the size of their savings accounts.All respondents with mminmum month- ly balances of $90 or less said they would continue saving in SIMPEDES (in this group, those with balances above $22 were receiving 9 percent interest). Among larger account holders, 76 percent of those with accounts of $90-445 said they would stay in SIMPEDES, while only 55 percent of those with ac- 278 The Microfinance Revolution: Lessons from Indonesia 'Would you still save in SIMPEDES if the annual Table 13.5 interest rate were 6 percent?": responses of 74 SIMPEDES savers in the 12 new branches of the second-stage pilot project, September 1985 (percent) Account size 2$90 >$90 445 >$445 All respondents Characteristic (N=13) (N=21) (N=40) (N=74) Yes 100 76 55 69 No 0 24 45 31 Source CPIS Rural Savings Report 5, 1985c Only 7 percent of counts of more than $445 said they would continue with SIMPEDES at a 6 respondents cited percent interest rate. In answers to other questions, 93 percent of the respondents said they would interest rates as the continue in SIMPEDES if the lottery were held annually instead of semian- nually, and 79 percent said they would stay in SIMPEDES even if there were most important no lottery. Thus SIMPEDES seemed to be popular.The costs of the instrument in the aspect of 12 new branches were considerably lower than in the first-stage Sukabumi trial. Further, it appeared that there was room for future adjustments to the instru- SIMPEDES ment if additional cost reductions proved necessary. The results of the second-stage SIMPEDES pilot project were watched care- fully not only because of the interest in the performance of the new savings in- strument, but also because the KUPEDES loan portfolio was growing rapidly. It had become clear that there was a large market for credit at the KUPEDES in- terest rate and that the units could make and collect such loans. By the end of 1985 the outstanding KUPEDES portfolio had reached $204 million and was ap- proaching full use of the funds that had been provided by the government in 1984. Since BRI would receive no additional government grants or subsidies for KU- PEDES (other than some technical assistance),the unit desas planned to mobilize substantial savings to enable the continued rapid expansion of the loan portfolio. Expansion and Market Penetration of the New Savings Program By 1985 the SIMPEDES pilot results had begun to attract the attention of In- donesia's economics ministers. In President Soeharto's January 1986 budget speech he stated that SIMPEDES was important for mobilizing rural savings- which, in turn, was important for the nation's development.The president also instructed that, from then on, all village grant funds (INPRES Desa) would be deposited directly into unit desa bank accounts held in the name of each village (as had become the practice in the districts under the SIMPEDES pilot project). Mobilizing Massive Savings Bank Rakyat Indonesia's Unit Desas, 1984-96 279 In early 1986 BRI decided that the modified SIMPEDES product was ready to be implemented throughout the unit desa system. SIMPEDES continued to be considered the flagship in a set of four savings instruments that would also include TABANAS, Deposito Berjangka, and giro accounts. Unlike the other instruments, however, SIMPEDES would not be available in the branches, but only in the unit desas. Later SIMASKOT, the urban version of SIMPEDES, was introduced in new units that were opened in cities beginning in 1989. SIMPEDES interest rates, based on savers' minimum monthly balances, were to be kept the same as in the second-stage pilot: no interest for accounts with balances below $22, 9 percent for accounts between $22 and $177, and 12 per- cent for accounts over $177. SIMPEDES would remain fully liquid, and lot- SIMPEDES would teries would continue to be held semiannually. This product would be aimed widely at households, firms, and organizations for whom liquidity was a pri- be aimed widely at ority.TABANAS would continue unchanged, with interest (also calculated on the minimum monthly balance) paid on all accounts and withdrawals limited households,firms, to two per month.TABANAS provided higher interest rates for smaller accounts than did SIMPEDES (or later, SIMASKOT). Its target was the interest-sensi- and organizations tive small saver; many of its account holders are children. As a time deposit instrument, Deposito Berjangka would provide the high- for whom liquidity est interest rates and the lowest liquidity of the four instruments. It would be offered in the unit desas for 1, 3, 6, 12, or 24 month periods, at annual interest was a priority rates ranging from 13 to 16 percent. Deposito Berjangka would be used by house- holds and enterprises that wanted higher returns and could afford to save in a nonliquid instrument, as well as by people saving for long-range goals (busi- ness expansion, construction, purchase of machinery, land purchase, education, retirement, and the like). Giro, the low-interest demand deposit account that had previously been of- fered at the branches, would now be available at the units primarily for use by institutions and for special accounts, such as those used to finance government- facilitated religious pilgrimages to Mecca. As a package, the instruments would provide savers with security, convenience, confidentiality, and a choice of instruments with different ratios of liquidity and returns. Savers could have as many accounts of whichever types they wanted. Preparing to expand the savings program, January-March 1986 In early 1986 the emphasis at BRI's head office in Jakarta turned to develop- ing a core group of trainers who would travel from region to region to ex- plain the bank's new approach to rural savings mobilization and to train regional and branch staff (who would then train the unit staff) in the operations of SIM- PEDES and the methods of finding local sources of savings. In January a train- ing session was held at the head office for the core group of trainers, with followup training in the regions during February and March. A book of case studies in savings was prepared for use by the trainers (box 13.2).'9 The cases, drawn from CPIS/BRI field work and from the experiences of the 13 pilot branches, provided examples of savings mobilized by unit desas 280 The Microfinance Revolution: Lessons from Indonesia Bolx 13.2 Excerpts from Bank Rakyat Indonesia's unit desa savings mobilization casebook * Savings potential can be found all around us, if you look you will find it * Service is the key treat potential savers with respect and consideration, listen to them and learn from their experiences and needs, and educate them about the ways in which BRI services can help them * Good relationships with local government officials and informal village leaders are es- sential Cultivate their friendship, pay attention to their concerns, and demonstrate ap- preciation for their help Attend village and subdistrict meetings regularly and widen your acquaintance with potential clients * Unit desa employees are members of the community Use your contacts with neigh- bors, friends, schools, and organizations to promote BRI services and to learn about As a package, the opportunities for mobilizing savings. * Learn to recognize events that are important as potential occasions for savings mobili- zation because large amounts of money will be transferred (land sales, rural electrifica- instruments would tion, and special opportunities such as government projects and donor disbursements in rural areas) provide savers with * The personal approach is essential Help clients (as unit staff did in providing special services to a busy midwife, in helping savers register at the Hai pilgrimage office, and security, convenience, in arranging the transfer of remittance funds from Saudi Arabia to the local unit) * Remember that depositing funds in the unit desas benefits the savers as well as BRI Households, enterprises, and institutions gain better security, receive returns, and are confidentiality, and a able to improve accounting and administration and decrease misuse of funds Explain the benefits to potential savers choice of dfferent * There is a close relationship between institutional savings and private savings These are often mobilized together (as in our examples of the high school and the health clin- ic) Sometimes the institutional savings are mobilized first, and then managers and ratios of liquidity employees of the institutions open private accounts, sometimes it is the other way around If clients are provided good service in one account, they are likely to open the and returns other * Many people and institutions need to transfer funds either occasionally or regularly Villagers need to pay their buyers or suppliers in the cities People working in Jakarta or Saudi Arabia want to send money to their families in the villages Parents in the vil- lages want to send funds to their children studying in the cities Government and pri- vate institutions need to meet their rural payrolls If your unit can help these people with their transfer needs, the unit is likely to capture their savings * This casebook has no ending yet-because the last section is to be written by youI Study the examples given here Send us your best case studies of successful savings mobilization The head office is waiting with much interest to hear from you Source CPIS Rural Savings Report 6, 1986a from a wide range of individuals, groups, private institutions, and government offices, and drew general lessons from the examples. These cases fell into two broad categories.The first represented opportunities available throughout much of rural Indonesia: mobilizing savings from rice farmers, traders, salaried em- ployees, food processing mills, schools, government offices, and the like.The sec- ond were locality-specific cases, such as industries based on local resources, farmers and traders of crops cultivated only in selected regions, tourist areas, and eth- nic-specific organizations. Mobilizing Massive Savings Bank Rakyat Indonesia's Unit Desas, 1984-96 281 In general, the training emphasized the identification and collection of sav- ings from the largest rural savers first.There were two reasons for this approach. First, the units needed to build savings quickly because of the rapidly increas- ing KUPEDES portfolio. Second, it was thought that if vlllage and subdistrict leaders and larger farmers and rural business owners became unit desa savers first, then smaller savers-many of whom were lnked in various ways with the larger ones-would follow suit. In hindsight, this generally turned out to be correct.The examples below illustrate how BRI staff were taught to mobilize savings from a wide range of potential savers. The savings casebook: mobilizing savingsfrom households. Many cases were "Service is the key: developed to demonstrate different types of household savings, to explore the rea- sons that households save, and to teach staff how to explain the unit desa savings treat potential savers instruments in ways that would make them attractive to households. The pur- pose was to teach unit staff and their branch and regional supervisors how to lo- with respect and cate potential savers and to help them match instruments to clients' savings purposes. Cases of individual savers used in the training included people who cul- consideration, listen tivated rice, vegetables, fruwt, tobacco, cloves, and coffee; who raised cattle and poul- try; who engaged in home industry production (weaving, bamboo products, to them and learn ceramics, woodcarving, tilemaking); who traded in agricultural commodities, tex- tiles, construction materials, and processed foods; who provided services (small from their restaurant and hotel owners, mechanics, midwives, motorcycle drivers providing transportation); and who worked in factories and offices. In each case attention experiences" was drawn to seasonality, income flows, purchase of raw materials, types and lo- cations of buyers and sellers, and periods of excess liquidity. Six examples illus- trate the messages contained in the training sessions on household savings. Learning the linkages, timing, and incomeflows of local industries. The training emphasized intra-industry linkages and the importance of understanding the income flows and expenses of households participating in different parts of an industry. Thus one case analyzed an example of the cattle trade in eastern In- donesia. Farmers there breed or purchase calves and fatten them; they then sell the fattened cattle to traders who live in nearby villages.These traders sell the cattle to larger traders operating in the subdistrict capital.The cattle are then sold to much larger, district-level cattle merchants. At each level these trading linkages typically involve long-term relationships. Cattle farmers tend to have excess liquidity when they sell their fattened cattle and purchases calves; this is usually a time that many households are ca- pable of saving.Vfllage and subdistrict cattle traders said they wanted a SIM- PEDES account in a unit desa so that, as one subdistrict trader put it, "I will not have to carry millions of rupiah in cash when I go to market to purchase cattle."The district-level trader in the case study already had a bank account at the district headquarters, but he wanted a SIMPEDES account in a rural unit that could be used for local transactions. This case study was used to show regional, branch, and unit staff how the cattle industry works and to encourage staff to become acquainted with cat- 282 The Microfinance Revolution: Lessons from Indonesia tie farmers and traders in their areas.These lessons were badly needed in 1986, when most unit staff did not understand their markets, did not visit potential clients, and knew little about the economic activities in the areas their units served.While visiting a number of cattle areas in eastern Indonesia that year, I asked six unit desa managers what day the weekly cattle market was held in their area. Only one knew. The traimng sessions emphasized, for example, how unit staff can develop a relationship with cattle farmers and traders, and how staff can then be intro- duced to the trading partners of such a client. Unit staff, perhaps with the help of the branch for larger traders, would then have the potential to capture de- posits from all four links in the cattle trading chain. A branch manager could show a district-level cattle trader that it would be to his advantage to have his The training main account at the BRI branch and to have additional accounts in the units. Funds could then easily be transferred (albeit slowly, at that tume) back and forth emphasized between branch and umt. Similar case studies of industry lnkages were done for coffee in NusaTeng- understanding the gara Timur (in eastern Indonesia) and for cloves in Manado, North Sulawesi. These trading linkages are of the same general type as in the cattle trade, and income flows and there is savings potential at every level: the farmers want to store lump sum in- comes, and the traders demand liquid accounts at convenient local units. But expenses of unlike the cattle trade, which continues throughout the year (although with peaks and troughs), tlming is more critical for coffee and cloves. After harvest, households most coffee and clove farmers store as much of their crop as they can afford to hold, wait for the commodity prices to rise, and then sell. In Indonesia, assuming participating in good storage conditions (which are fairly common), coffee beans can be stored for about 2 years, while cloves can be stored up to 10 years. different parts of an These case studies emphasized that knowing the local market includes know- ing the tinmng of income flows.The cases demonstrated that unit desa staff can industry mobilize significant savings if they keep abreast of the coffee and clove mar- kets and visit farmers right after their sales. But most unit managers did not yet think in this way. On one occasion I asked the manager of a unit desa in Nusa TenggaraTimur to accompany me to a coffee-growing village in a remote part of the area served by his unit. He said there was no point in going there be- cause there was no savings potential there. "That village [which I later discov- ered he had never visited] is isolated, and the people are poor and uneducated. They are not in the modern economy."When we arrived there, however, we found that a number of farmers (with considerable savings in coffee) listened regularly on their shortwave radios to London-based British Broadcasting Corporation reports of commodity prices. The training also emphasized interindustry relationships in local markets. For example, weavers in the Nusa Tenggara Timur coffee area said that their household income flows are influenced by the coffee market.When coffee prices are high, textile prices are also high.When coffee prices are low and coffee farm- ers are holding their coffee, textile prices drop.Weavers said that they saved in their finished woven products, and that they sold these when they needed money. Mobilizing Massive Savings Bank Rakyat Indonesia's Unit Desas, 1984-96 283 But if a weaver needed money when coffee prices were low, she had to sell her textiles at a low price. If unit staff could be trained to learn the timung of local financial cycles, they would know that weavers in that area have excess liquid- ity at about the same time as coffee farmers. Upon hearing about SIMPEDES, many weavers said they wanted accounts (which were not yet available in eastern Indonesia).They said that selling their textiles when coffee prices were high and putting their proceeds in the bank would provide them with security and returns, and they could still access their savings at any time. Listening to-and not making assumptions about-savers'needs. Umt staff often as- sumed that they knew their clients' needs, a holdover from the pre-1984 unit desas. I asked six unit During the pllot projects numerous examples had been collected showing how a unit staff member had assumed the opposite of what a saver actually wanted. desa managers what For example, a unit manager looking for potential savers said that he had decided not to visit a woman who ran a small restaurant because it would be day the weekly cattle "natural" for her to want to put her profits back into the restaurant so that it could be expanded."People like her are not potential savers," he explained.When market was held in he was finally persuaded to visit her and explain the features of SIMPEDES, she said,"I want to have a SIMPEDES account."When he asked about expanding their area. Only one the restaurant, she said. "I do not want to expand the restaurant. If it becomes larger, people will think I am rich, and then I will have to start giving credit. knew That will make difficulties for me as I am a widow and have no one to help me collect loans. So I want to keep the restaurant small and put my profits in the bank." She opened a SIMPEDES account. In another example, a unit manager was talking with a carpenter about the new unit desa savings instruments.The manager assumed that the potential saver wanted a SIMPEDES account, but the carpenter kept asking questions about the time deposit account.Assummng that the carpenter could not afford to main- tain a time deposit account-which requires a relatively high minimum de- posit and carries penalties for early withdrawals-the manager discouraged this idea.The carpenter, however, had done his calculations carefully. In addition to a SIMPEDES account, he wanted to put some funds into a one-month time deposit account. This account would be for his savings for purchasing wood. Each month he would review wood prices and decide whether to roll over his time deposit or close the account and purchase wood. The carpenter opened both a SIMPEDES account and a one-month time deposit account. Capturing remittances. Before 1984-and until 1986 outside the pilot pro- ject areas-BRI staff did not consider remittances a potential source of unit savings, although the opportunities were all around them. During the first stage of the pilot project the staff of a unit desa in Suk- abunii collected the names and addresses of households that had famuly mem- bers working in other parts of Indonesia who sent remittances home.The staff visited those households and told them how funds could be transferred from the person living elsewhere to the recipient's local unit desa. By February 1986 the unit had four savings accounts in which the account holders were receiv- 284 The Microfinance Revolution: Lessons from Indonesia ing regular remittances from Jakarta, Sumatra, and Irian Jaya (now Papua). The aggregate balance of the four accounts was about $1,950. In addition, this unit had three savings accounts in which the funds were derived from remittances received by households with family members work- ing in Saudi Arabia; in February 1986 the aggregate balance of these accounts was about $1,000.The unit staff had captured these savings by visiting the house- holds and offering assistance in arranging the transfer of the funds.At that time this was a complicated process that the recipients of the remittances would have been unlikely to undertake by themselves.Thus they were delighted when the unit and branch staff helped them receive their remittances at the unit. These early experiments eventually led to substantial savings in the unit desa system derived from remittances to local households. Since the units do not The case studies report remittances separately from other deposits, precise data are not available. But remittances are an important part of unit savings. emphasized that Recognizing that many SIMPEDES savers depositfrequently and rarely withdraw. The savers surveyed in the second-stage pilot project had overwhelmingly cited knowing the local an unlimited number of withdrawals as their top priority among the choices offered for the SIMPEDES product.Thus BRI tended to assume that SIMPEDES market includes savers would make frequent withdrawals, adding to the cost of the product. It was not yet understood that what many savers wanted was simply the option knowing the timing of withdrawing whenever they wanted; they did not intend to exercise the op- tion frequently, and in fact they did not. of incomeflows Thus while some savers, especially traders, used their SIMPEDES accounts to deposit and withdraw frequently, many SIMPEDES savers deposited regu- larly but withdrew rarely. Case studies were developed of people engaged in a variety of economic activities, including agriculture, petty trading, home in- dustries, and services. These savers used their accounts to build a savings bal- ance for use in emergencies, for income smoothing, or for a variety of specific purposes. Some transferred their funds from SIMPEDES to Deposito Berjangka when their account reached a level that was beyond the amount they needed available in liquid form. Some had the interest from their time deposit account automatically transferred to their SIMPEDES account. Unit staff were shown how these SIMPEDES accounts, which were very stable, were of advantage both to the savers (who would have funds available when they needed them) and to the units (which would have the use of the funds for long periods). Mobilizing large savingsfrom rural land sales. During both stages of the pilot project BRI learned that land sales in rural areas present good opportunities for mobilizing household savings, especially when large tracts are purchased from multiple small sellers by the government or by private firms.The sellers, who usually receive relatively large amounts of cash at one time, need a secure place to store some or all of their proceeds at least temporarily, and in some cases for longer periods.Whether a seller wants to use the funds from the sale to pur- chase other land or for other purposes, he or she is rarely ready to make the new purchases at exactly the time of the sale. Mobilizing Massive Savings Bank Rakyat Indonesia's Unit Desas, 1984-96 285 Three case studies of land sales were included in the casebook, providing both positive and negative lessons from unit desa experiences with this type of savings mobilization. In one case the government had purchased the land of an historical site inYogyakarta.Well ahead of the sale, the local umt desa pro- vided information about BRI products and services to the land sellers, both individually and at local meetings. BRI staff from the unit, sometimes accom- panied by staff from the branch and the regional office, attended village and subdistrict meetings to explain how the sale proceeds could be saved at the local unit desa. When the land sale took place in September 1985, BRI representatives from the branch and the unit were present when the transactions occurred.They ad- Numerous examples vised the land sellers to open unit desa savings accounts the same day "so that your money will be safe and you can have peace of mind while you make the had been collected important decisions ahead." Because there were many sellers receiving funds at once, the umt developed a special streamlined procedure for opening accounts showing how a unit and collecting money at the place of the land sale; the remaining formalities were carried out later at the unit.A total of $587,328 was deposited in the unit staff member had desa from the proceeds of the land sales.Three months later in December 1985, 81 percent of these funds was still on deposit in the umnt. assumed the A second example, a smaller one that could be more frequently replicat- ed, involved a private company that purchased land to build a shrimp hatch- opposite of what a ery in Sukabumi. The unit desa manager promoted the idea that the 13 land sellers should open savings accounts, and he discussed this with the company's saver actually representative, the village head, and the sellers.All the sellers opened SIMPEDES accounts at the local unit desa; in November 1984 these deposits totaled wanted $63,573. Nine of the thirteen sellers then used these funds for Haj pilgrimages to Mecca,20 and the unit desa manager helped them enroll at the district Haj office.The savers reported that this was of great help to them because they need- ed to go to the Haj office only once-unlike their neighbors and friends, who had to make multiple trips to accomplish the same formalities. In February 1986 (after the pilgrimages), 8 of the 13 land sellers were still SIMPEDES savers, with an average balance of about $850. The third case study was developed to show BRI staff how not to act. In this case the Department of Public Works purchased land from villagers in Cen- tralJava inJuly 1985 to construct a large dam.The provincial and district gov- ernments made an agreement with BRI that payments would be made directly into TABANAS accounts in the names of the sellers. Between July 1985 and January 1986, as the sales proceeded, $2.9 million was deposited in more than 4,000 accounts in one umt desa. But the unit's staff provided almost no infor- mation to the depositors about the savings instruments, services, and options available to them. This land sale took place in a poor area where most of the people had never seen so much money at one time. Many hastily bought inferior land, driving up local land prices. Others purchased motorcycles, televisions, and other con- sumer goods at rapidly rising prices. By January 1986, the last month of the 286 The Microfinance Revolution: Lessons from Indonesia seven-month land purchase, only 30 percent of the $2.9 million remained in the unit, many of the land sellers had become landless, and unemployment in the area had risen. In conjunction with the others, this case was a useful teach- ing device because it emphasized the important difference that unit desa staff could make to benefit both the bank and its clients. Recognizing the importance of word-of-mouth advertisingfrom satisfied clients. The cases showed many examples of this lesson.A midwife and prominent mem- ber of her community in Central Java told the manager of her local unit desa that she wanted to have a savings account but, because her clinic was always filled with patients, it was difficult for her to come to the bank during the unit's working hours. The unit manager responded by offering to provide her with special mobile service; a staff member would collect her deposits and deliver Many SIMPEDES funds as needed. She, in turn, agreed to promote savings to her customers. By February 1986 her savings balance was about $5,300, and many of her cus- savers deposited tomers, including many low-income women, had also opened accounts at the unit. regularly but Of course, this form of promotion depends on continuing good service. During one of our early research trips an elderly woman in a remote area in withdrew rarely eastern Indonesia gave a number of reasons that people in her village would be pleased if SIMPEDES accounts were made available there. She concluded by commenting, "As you can see, this area is very rural.There is nothing to do here in the evenings except to sit and talk. So if you do something good for us-if you offer us SIMPEDES accounts-we will talk for years to come about how BRI came here and helped us."Then she looked up and added,"Of course, if you do not do this, or if you do not do it right, then that is what we will talk about for years to come!" The savings casebook: mobilizing savingsfrom organizations and institu- tions. In addition to mobilizing household savings, the casebook emphasized the many opportunities for capturing rural savings from public and private or- ganizations and institutions. Examples included funds mobilized by the unit desas from 32 schools, 6 vlllage associations, 20 government offices (including vil- lage, subdistrict, and district-level government offices, as well as the local of- fices of the departments of agriculture, archaeology, cattle breeding, education and culture, family planning, fisheries, industry, and religious affairs), 11 coop- eratives, 6 agricultural estate companies, and 25 local groups (the Armed Forces Pension Association, the BusTerminal Employees Association, the Police Wives' group, the Reformed Prostitutes Group, the Tea Factory Drivers' Welfare Group, and various local women's, sports, family planning, and religious groups). Many such institutions and organizations had excess liquidity; in aggregate, they showed considerable potential for unit desa savings mobilhzation.The cases were separated into those that represented opportunities for savings mobiliza- tion throughout Indonesia and those that were locality-specific. Nationwide opportunitiesfor savings mobilization. The examples below are taken from cases of mobilizing savings from types of organizations and institutions Mobilizing Massive Savings Bank Rakyat Indonesia's Unit Desas, 1984-96 287 that are common throughout Indonesia.The training sessions emphasized that each unit should start immediately to identify such institutions in their service areas, to meet with the heads of the organizations, and to explain to them how the new unit savings program could benefit their organizations. Villagefunds. Throughout Indonesia, vlllage heads are responsible for vlllage revenues. These include the government's annual grants to all vllHages, spe- cial development funds from provincial governments, rentals of village- owned land, sales of village-owned land and of produce from village-owned lands or trees, fees and taxes, and other funds from district and provincial gov- ernments. Not all villages have all of these revenue sources, but all have some There were many of them. Traditionally such funds were held by the village head and other village opportunitiesfor officials. However, during the SIMPEDES pilot project some village heads had been convinced to deposit village revenues in the unit desas. The unit capturing rural staff spread information widely that depositing village funds into village bank accounts at the units made the funds more secure and facilhtated manage- savingsfrom public ment of the accounts, that convement and appropriate instruments were avail- able for such funds, and that bank accounts could be used as aids in meeting and private requirements of reporting and accountability. As expected, not all village heads were enthusiastic about the idea. Some, organizations and however, liked it. Others became supporters when they found it difficult to oppose publicly the growing opinion that village-owned funds should be de- institutions posited in village-owned bank accounts. Case studies were developed to show how village funds could be mobilized. For example, a Sukabumi unit desa mobilized funds that were being saved for a $3,144 multipurpose building that was to be constructed by one of the vil- lages served by the unit. A unit desa inYogyakarta convinced a village to de- posit $20,444 from its sale of village-owned land. One and a half hectares of land belonging to a Central Java village was sold for $19,556; it was later dis- covered that the village head, who had kept the funds in his house, had ris- used some of the money. The subdistrict head intervened and instructed the village head to return the funds and to deposit them in a unit desa account in the name of the subdistrict. Once this was done, the money stayed in the ac- count until it was used for approved vilage purposes. Other cases involved vil- lage accounts opened to deposit income from rentals of village land and from collection of fees and taxes. * Schools. Like land sales, savings mobilization from Indonesian schools result- ed from learning that took place in the pilot project branches and from other research carried out for the Ministry of Finance. In a 1985 trip to study sav- ings potential, I visited a government high school in Central Java.The head- master said that the funds of most schools in the area were held by their headmasters or other school officials. He said he had observed that this ap- proach led to leakage of school funds and that he did not approve of han- dling funds in this way. 288 The Microfinance Revolution: Lessons from Indonesia Therefore, he had opened six TABANAS accounts in the local unit desa and three Deposito Berjangka accounts at the BRI branch.TheTABANAS accounts were opened to overcome the limitation on withdrawals. (This was not a pilot project area and SIMPEDES was not yet available in this district.) The six TABANAS accounts were used for four purposes: the Parent- Teacher Association (with a balance of $5,155 at the end of 1985), the op- erating expenses of the school ($4,800), the Library Fund ($1,867), and the StudentActivities Fund ($888).The Deposito Berjangka accounts at the BRI branch (these accounts were not yet available at the units) were used to de- posit funds for building construction; the balance of these three accounts was $13,333.Thus by the end of 1985 this one school had $26,043 in deposits at the unit desa and its supervising branch. Many teachers, staff, and students The school had personal accounts in the unit as well.While this was a large high school, the general lesson was immediately apparent: every village in Indonesia has headmaster said the a school (some have more than one), and schools held substantial savings po- tential for the units. funds of most The cases also included examples of other kinds of schools. One was a pri- vate junior high school inYogyakarta. Since the school was private, funds could schools were held by be mobilized not only from the sources mentioned above, but also from school fees (which were paid at particular times and then used over a period their headmasters of months).The school, located in an area covered by a pilot branch, had opened a SIMPEDES account at the local unit; in November 1985 the average daily and this approach balance of the account was $1,277. Other cases covered funds mobilized from government and private primary,junior high, and high schools, universities, a led to leakage of private kindergarten, a school for retarded children, an art institute, agricul- ture schools, and a number of religious schools. In February 1986 balances ranged schoolfunds from $121 to $27,063. The school case studies were used to emphasize the point that certain kinds of savings mobilization opportunities exist everywhere in Indonesia, and to teach methods of mobilizing the savings from these sources. Religious institutions. Case studies of mosques, churches, and temples were used to demonstrate that, as with schools, religious institutions are ubiqui- tous in rural areas in Indonesia and that many were potential sources of unit desa savings.There is, of course, wide variation among such institutions. Some are large and have considerable excess liquidity; others are much smaller. Some mosques may not want to open interest-bearing accounts because of religious proscriptions on receiving interest payments. But the mosque representatives interviewed for the case studies said that since they consid- ered the payment from BRI as profit, rather than interest, this would not constitute a problem. It was also established that if a mosque did object to receiving interest payments, BRI could give the amount to the mosque as a contribution. The three main sources of funds from religious institutions for unit sav- ings were collections taken at weekly services, funds collected for special re- ligious events, and building funds. Four case studies of village or subdistrict Mobilizing Massive Savings Bank Rakyat Indonesia's Unit Desas, 1984-96 289 mosques and churches were used in the training (three mosques and one Protestant church); their combined balance was $1,772 at the time of the interviews in late 1985 and early 1986. * Health clinics. Every subdistrict in Indonesia has at least one government health clinic (Pusat Kesehatan Masyarakat, or PUSKESMAS), and there are also many private clinics.The case studies of health clhics, like those of schools and re- hgious institutions, emphasized that the methods used to identify and mo- bilize savings could be easily replicated throughout the country. One government health clinic in Sukabumi had three SIMPEDES accounts that were used for different purposes: laboratory expenses, midwifery activities, and purchase of medicines. The clinic head said that keeping the three ac- One school had counts enabled the clinic to improve its accounting and reporting, and that having the funds in the bank improved security. In addition, the clinic head $26, 043 in deposits had opened a personal account. These accounts were opened in August 1985 after visits by unit staff to at the unit desa and the clinic head and to the staff members responsible for the various funds. In December 1985 the aggregate balance of the four accounts was $6,113. its supervising Locality-specffic opportunitiesfor savings mobilization. A second set of cases was branch developed to teach staff how to identify regional, local, or one-time opportu- nities for collecting savings from groups, organizations, and institutions.A wide range of possibilities was explored, emphasizing regional specialties, develop- ment programs targeted to specific areas, and social organizations in different parts of the country with different customs (adat). * Funds held by local leaders for specific development purposes. Two Sukabumi ex- amples showed how to capture idle funds in villages that were awaiting elec- trification. In the area covered by one unit, three villages that were to be electrified had set up committees that collected funds from households that were to be electrified. Unit desa staff convinced the three committees to de- posit the funds in SIMPEDES accounts until they were needed (which turned out to be more than a year later). In December 1985 and January 1986 the electrification committee account balances in the three villages totaled $7,826. In the second example, another Sukabumi unit desa began collecting sav- ings from village electrification committees in July 1985; by January 1986 the balances in these accounts totaled $44,444. * Funds from development programs that provide assistance to households. Another lesson from the pilot districts was that regional, national, and international rural development programs that target specific areas and populations pro- vide an excellent opportunity for savings mobilization in banks with a net- work of rural outlets. One example was the case study developed to show how savings were mo- bilized from Foster Parents Plan International (FPPI), a private organization that sponsors poor children, families, and communities in many countries. 290 The Microfinance Revolution: Lessons from Indonesia The FPPI wanted to set up a mechanism in fourYogyakarta districts for the transfer of funds to 20,000 children whose foster families would receive reg- ular assistance from the organization. BRI officials approached the FPPI office to suggest that these funds be de- posited directly into unit desa savings accounts.The FPPI liked the idea, both because each foster family could have its own account and because the FPPI would be able to check whether the funds had reached each household.The FPPI office requested that arrangements be made for depositing the funds in the unit desas each month. But BRI said that the procedures would be too time consuming if the funds were deposited monthly because, on average, there would be more than 1,000 accounts per unit desa (the areas where the chil- dren lived were served by 19 units). Case studies of The FPPI's only other real option was to use a bank that operated at the district level, since no other bank had a network of subdistrict outlets. Be- health clinics, cause it was considered too far for the foster families to travel to district head- quarters, the FPPI would have had to use an intermediary to deliver the funds schools, and religious to the households, and it would have been difficult to ascertain whether the funds had reached their intended destinations. The FPPI rejected this op- institutions tion, and BRI and the FPPI agreed that the payments to the foster families would be made through the units, that payments would be made every six emphasized that months, and that each unit would not have to process more than 40 accounts a day. methodsfor Between October 1985 and February 1986 the average amount outstanding in the FPPI foster family accounts in the four unit desas examined in the mobilizing case study was $50,218; the range among the units was from $7,717 to $15,423. institutional savings could be replicated Many such case studies were developed.The training sessions emphasized that BRI's unit desa system has a strong comparative advantage in rural areas, throughout the that the units could negotiate with local governments and other institutions from a position of strength, and that in a number of cases, such as land sales country and the FPPI, the units and the branches would need to work closely togeth- er to capture the potential savings from individuals and from organizations and institutions. Expanding the savings program to all unit desas, April-September 1986 After the training of the trainers was completed, first in Jakarta and then in the regions, BRI decided that the new savings program was ready to be expand- ed to all units. A final review of the pilot projects at the end of March 1986 showed that they were doing well. In the 13 pilot branches, unit desa savings had more than doubled since July 1995 (when 11 of the 13 pilot branches began the program), and by March 1986 SIMPEDES already accounted for 46 per- cent of savings in the pilot branches. Mobilizing Massive Savings Bank Rakyat Indonesia's Unit Desas. 1984-96 291 The phased expansion of the new savings program began in April 1986. The modified SIMPEDES, Deposito Berjangka, and giro accounts would now join TABANAS in all unit desas.The centrally trained trainers traveled to each region in turn, and worked with BRI's regional, branch, and unit staff to train all staff who would be involved in the new savings program at the units. By September 1986 the four savings instruments were available through the unit banking system nationwide. CPIS/BRI teams monitored the expansion of the program, studying the training, the startup processes, and the problems that occurred as the new sav- ings instruments were introduced in different locations; assessing staff under- standing of the program and its instruments; and providing the head office with There were multiple early warning signals as necessary. In addition, a sample survey was conducted in 29 unit desas in 13 branches in 8 provinces fromJune through August 1986, implementation while the expansion was under way (CPIS Rural Savings Report 7, 1986b). Data were collected to analyze implementation problems and the cost of funds problems at the start in branches and units. Issues of BRI organization and administration as they affected the savings program were also identified (chapter 14). of the nationwide As expected, there were multiple implementation problems at the start of the nationwide expansion. Among many others, these included difficulties in expansion cash management; poor distribution of bankbooks, brochures, and posters; misunderstanding of SIMPEDES regulations; restrictions placed by branch managers on units permitted to offer SIMPEDES; and confusion about lot- teries and prizes. But one major concern-that SIMPEDES accounts would generate large numbers of transactions, increasing labor costs-was allayed.The average num- ber of monthly transactions in the sample survey was less than the two with- drawals a month permitted byTABANAS (CPIS Rural Savings Report 7,1986b). It was not that most savers chose SIMPEDES because they wanted to with- draw and deposit frequently. Rather, they chose SIMPEDES so they would have the option to make transactions whenever they wanted to do so. Costs were monitored carefully (CPIS Rural Savings Report 7,1986b; and 8, 1987). There were differences in the components of the costs of funds among the different instruments-for example, SIMPEDES had higher labor costs and lower interest costs than TABANAS or Deposito Berjangka. Never- theless, the total cost of loanable funds, including labor, in each of these three instruments was calculated to be roughly similar at about 16 percent. Giro, a low-interest current account used mamily by institutions with special require- ments, had lower costs. Despite the many implementation problems in the early stages, the nation- wide expansion of the new savings program went well.The volume of deposits in the units (in U.S. dollars) increased 42 percent between the end of 1985 and the end of 1986, from $75.4 million to $107.0 mnllion.2' Thus the new pack- age of unit desa savings instruments seemed to be both popular and affordable. A branch manager in South Sulawesi faced an unusual dilemma. As SIM- PEDES moved into the unit desas there, it became an instant success. Branch 292 The Microfinance Revolution: Lessons from Indonesia customers then demanded that SIMPEDES be made available at the branch as well, since they wanted to save using SIMPEDES and the nearest unit desa was too far away. Offering SIMPEDES at the branch was clearly against BRI reg- ulations, yet it was obvious that the demand was large. The branch manager looked through his regulation manuals, hoping to find some guidance. Finally, he found a regulation that stated that a village post, a subunit of a unit desa, could be opened at any location where the branch man- ager determined there was a strong demand for unit services there. He then opened a vlllage post of the nearest unit desa inside the branch office.This type of problem was later solved by the opening of BRI units in urban areas in 1989 and by the 1990 introduction of SIMASKOT, the urban version of SIM- PEDES, in all urban units. BRI thought that In some areas-especially rural ones-SIMPEDES became part of the In- donesian language, meaning "save" or "keep."Thus when a popular government since the new official in South Sulawesi was about to be transferred, some of the villagers protest- ed, saying, "Don't transfer him, simpedes him here." savings products Between the end of 1984 (when the first Sukabumi pilot project was two months old) and the end of 1986 (when the new savings program had become were in all units, the available nationwide), unit desa savings increased 174 percent, from $39 mil- lion to $107 million. By the end of 1986 SIMPEDES accounted for 47 per- hard work of cent of unit desa savings nationwide. In early 1987 BRI's head office breathed a collective sigh of relief, think- establishing unit ing that the new savings products, now in all units, were appropriate and cost- effective, and that the hard work of establishing unit desa savings services was desa savings services over. They were right on the former, but wrong on the latter. It turned out that designing and pricing the instruments, training the man- was over. They were agers and staff, and solving the logistical problems of expansion were necessary but insufficient conditions for mobilizing enough savings to finance the grow- wrong ing KUPEDES portfolio. On average, a unit desa covers about 18 villages. It soon became apparent that most of the savings was mobilized from the two or three villages that were located nearest the unit desa: their savings came to be known as "easy money." But if unit desa savings were to finance KUPEDES, the other villages would have to participate as well. This would require devel- oping a special methodology for market penetration. Eventually it would also lead to a new incentive program that would motivate staff to get out to those villages. Learning market penetration, 1987-89 CPIS/BRI teams continued to monitor implementation issues even after the new savings program was established in all units. Extensive field trips to many parts of the country during 1986-87 indicated that there were still vast areas of untapped savings potential and unit staff who were doing little about it.22 Many kinds of problems continued: with cash management (savers pointed out that a liquid account is only liquid if the umt keeps cash on hand); with units of high potential that were downgraded or closed because branch managers (and Mobilizing Massive Savings Bank Rakyat Indonesia's Unit Desas, 1984-96 293 sometimes unit staff as well) did not recognize the units' potential; with ad hoc, unsystematic funds mobilization; and with poor unit desa service in some areas. Typically, in early 1987, unit desa X had mobilized funds from schools but not from government offices, while unit desaY had an account from the Min- istry of Industry but not from schools. Meanwhile, unit desa Q had accounts from traders and businesses but not from organizations or institutions, while unit desa Z had mobilized savings from mosques but not from schools, and from the Ministry of Agriculture but not from local businesses. To deepen market penetration, especially in villages that were not located near unit desas, BRI decided that the second phase of expansion of BRI's new A new pilot project savings program would emphasize a coordinated and systematic approach to rural funds mobilization. Accordingly, in 1987 a new pilot project was estab- was established to lished in North and Central Sulawesi that came under BRI's Manado region- al office.This project was intended to develop what became a crucial aspect of develop a systematic BRI's unit banking system: the systematic approach to savings mobilization. approach to savings Developing the systematic approach to savings mobilization, 1987-89. Building on the methods of locating and estimating savings potential that had mobilization been developed in the pilot projects and recorded in the casebook, unit staff and their branch supervisors under the Manado regional office were given spe- cial training on how to identify and contact potential savers systematically. Under this approach, staffwere first shown how to identify the 100 largest savers in the area covered by their unit.The idea was not to mobilize savings only from those with potentially large accounts, but rather to test the systematic approach on larger savers first so that KUPEDES could be financed quickly from unit desa savings, and the hypothesis that smaller savers would follow larger ones could be tested. Once developed and tested, the systematic approach was used to locate po- tential savers of all kinds, both large and small (CPIS Rural Savings Report 9,1988b). Potential savers were systematically identified through visits to subdistrict and village officials, heads of local institutions and government offices, local or- ganizations, informal leaders, wealthy villagers, and other local contacts. From these interviews lists of potential savers were drawn up and marked on a large map of the area to be covered. In each unit a schedule was drawn up showing the staff members who would visit particular potential savers on particular dates. Staff were taught how to conduct these interviews, how to explain the uses and advantages of each type of savings instrument, and how to keep records of the visits and any followup actions required.The units used these interviews to fur- ther their knowledge about individual and group savings, as well as about wide- ly available savings sources such as schools and government offices, and sources particular to the region (such as, in this case, the clove and nutmeg trade). This training was quite successful, and many units became able to identify substantial numbers of potential savers.Through the Manado pilot project, BRI learned that, in addition to local savers, it had a comparative advantage over other banks in obtaining accounts from large urban-based corporations that conduct 294 The Microfinance Revolution: Lessons from Indonesia substantial business in rural areas. For example, some large companies distribut- ing soft drinks, cigarettes, cosmetics, and processed foods nationally demanded SIMPEDES accounts into which their local distributors could make deposits, and from which funds could be transferred to urban banks. Conversely, urban-based companies purchasing raw materials or goods wholly or partly produced in rural areas found SIMPEDES accounts useful for making payments to their suppliers. BRI decided to expand the systematic approach to all unit desas. Accord- ingly, in May and June 1987 unit and branch staff in 12 branches under 6 BRI regional offices were trained in the systematic implementation of savings mo- bilization, and the method began to be widely used. By 1988 all units were im- plementing the new method. However, it soon became apparent that during training, unit staff could lo- By 1988 the cate savers and mobilize deposits, but their follow-through tended to drop off as soon as the training was finished.The problem was no longer that staff could not systematic approach recognize savings potential.The problem was that staff found that visiting poten- tial savers in dispersed villages was hard work, and they observed that staff mem- to savings bers who visited few or no prospective savers received the same salary and benefits as those who actively followed BRI's instructions for systematac savings mobilization. mobilization was As discussed in chapter 14, a staff incentive program had been introduced in 1984 in which staff received cash incentives based on the profitability of their implemented in all unit desas. But by 1987, 81 percent of the units were profitable and many units had reached the maximum incentive payment allowed.As a result that program units-but not well. (which was later revised) no longer served as a powerful incentive for perfor- mance.Thus by 1988 the systematic approach to savings mobilization was un- Staff incentives were derstood but generally not well implemented. then developed The UnitAchievement Competition, 1988-89. Implementation of the new methods was monitored during 1988-89, leading to a number of changes.The most important of these was the development of a new staff incentive program, the Unit Achievement Competition. The program could not be based solely on performance in savings mobilization, however, because staff might then ne- glect KUPEDES and the unit desa portfolio could deteriorate. Rather, the main purpose of the competition was to reward staff on a semiannual basis for good performance in profitability, portfolio quality, and volume of lending and sav- ings. The Unit Achievement Competition was first tested in Manado, and in 1989 it was implemented throughout the unit desa system. As earlier, incentives were based on the performance of units, not individ- uals, and were paid for out of the profits of each unit. Under the Unit Achieve- ment Competition the staff members of all units that met their goals received cash awards and recognition certificates awarded by the head office in a for- mal ceremony held in each region. Units that were judged the best in their re- gion and best in the country received special awards, including visits to Jakarta and trips to neighboring countries to study other methods of microbanking. The combination of the systematic approach to savings mobilization and the Unit Achievement Competition galvanized the unit desas to mobilize sav- Mobilizing Massive Savings Bank Rakyat Indonesia's Unit Desas, 1984-96 295 ings.As shown in figure 13.1, the units began to increase their savings, though on a small scale, under the first pilot project (1984-85), and savings rose dur- ing the early expansion phase (1986) and the second pilot project in Manado (1987-89). But not until 1989 were all the components of a successful large- scale savings program in place. It was no accident that unit desa savings first financed all KUPEDES loans in 1989; this could have occurred only when the market penetration effort had become nationwide through all units. In the following years the results of the 1984-89 savings work and the continuing emphasis on market penetration be- came clear: umt savings more than quintupled, from $534 nullion at the end of 1989 to $3 billion at the end of 1996. Average annual inflation in each of Unit desa savings these years was less than 10 percent. This performance was a direct result of the combination of market research, firstfinanced all management and staff training, product design and testing, the systematic ap- proach to savings mobilization, and the Unit Achievement Competition. None KUPEDES loans could have been omitted. As will be discussed in chapter 14, it was also a re- sult of the changes in unit desa organization, management, administration, and in 1989; this could operations made during those years. At the time of the early pilot projects in unit savings mobilization, neither the have occurred only components nor the sequencing needed to build a large mucrosavings program was known. Each step led to the next and was guided by the underlying princi- when the market ples established at the time of the transformation of the units, by contmuing mar- ket research among the units' present and potential clients, and-a sine qua penetration effort had non-by the excellent management of the unit desa system at the head office. become nationwide Unit Desa Savings Products, 1996 The savings products in the unit desas in 1996 were quite similar to those in 1986.23 Interest rates fluctuated over time.TABANAS stopped restricting the number of withdrawals. SIMASKOT was added to all urban units in 1990.The minimum account size below which SIMPEDES interest was not paid was re- duced. Operating costs were lowered by increasing efficiency And other changes were made. In general, however, the 1986 instruments worked well. Households and enterprises frequently hold more than one type of deposit account in the unit desa system. Some households hold a time deposit account to save for long-term goals, a SIMPEDES or SIMASKOT account for routine income deposits and withdrawals for operating expenses, and a TABANAS ac- count for children's savings. Deposits in all unit desa savings and deposit instruments continued to be implicitly guaranteed by BRI. Although there were no explicit guarantees for savings until 1998 (see chapter 15), it was widely believed long before then that savings were safe in the unit desas because BRI is a government bank. In addi- tion, the government had protected small savers in banks that had failed; this re- inforced the belief that unit desa savings were effectively government-guaranteed. 296 The Microfinance Revolution: Lessons from Indonesia Table 13.6 |Annual interest rates for unit desa savings instruments, September 1996 Account type and size Interest rate (September 1996 U.S. dollars) (percent) SIMPEDES <$4 0 >$4-$425 10 0 >$425-$2,125 11 5 >$2,125 13 0 SIMASKOT Unit savings more $1 0 0 >$10-$425 - 11 0 than quintupled, >$425-$2,125 13.5 >$2,125 145 from $534 million TABANAS in 1989 to $3 S$4 0 >$4 130 billion in 1996 Time deposits (Deposito Berjangka) Varies with maturity 15 5-16.0 Source BRI 1996a, p 16 SIMPEDES and SIMASKOT SIMPEDES has accounted for more than half of unit desa deposits since 1987. In 1996 SIMPEDES and SIMASKOT together accounted for 76 percent of the units' $3 bilhon in deposits, and for 71 percent of the 16.1 million unit desa accounts. Both instruments permit an unlimited number of transactions, both have lotteries, and both have tiered interest rates depending on account size. SIMASKOT requires a higher minimum account balance for payment of in- terest than does SIMPEDES (see below), but SIMASKOT interest rates are some- what higher (table 13.6). In addition, SIMASKOT lotteries are held at the provincial level rather than at the district level, and lottery prizes are in cash rather than in kind. Otherwise, the instruments are basically the same. Both SIM- PEDES and SIMASKOT are available in some units where there is demand for both instruments. Neither is available in BRI's branches. TABANAS As Indonesia's national savings program,TABANAS is available in all banks, and banks are now generally permitted to set their own terms and interest rates. In Mobilizing Massive Savings Bank Rakyat Indonesia's Unit Desas, 1984-96 297 BRI and other banks, the success of SIMPEDES resulted in some changes in TABANAS.At BRI,TABANAS has permitted savers an unlhmited number of withdrawals since 1990. Because the cost of maintaining very small accounts is high relative to their earnings, at BRI TABANAS and SIMPEDES pay no interest on accounts with minimum monthly balances of $4 or below; SIMASKOT pays no inter- est on accounts with minimum monthly balances of $10 or below.TABANAS pays a higher interest rate on small accounts that are above the minimum than do SIMPEDES and SIMASKOT (see table 13.6). However, the average cost of TABANAS funds is about the same as that of SIMPEDES and below that of SIMASKOT, primarily because the cash prizes awarded in the semiannual SIMPEDES and TABANAS lotteries are much smaller than the prizes awarded in the lotteries of the other two instruments. SIMASKOT Deposito Berjangka and giro accounts accountedfor 76 The unit desas offer BRI's standard time deposit instrument (Deposito Ber- jangka), with maturities ranging from 1 to 24 months.Withdrawals are per- percent of the units' mitted before maturity, but with a penalty. Because the units' time deposit accounts compete in an interest-sensitive market, they usually carry the high- $3 billion in est interest rates of the savings products available at the units (up to 16 per- cent a year in 1996). The units also offer a giro account, a specialized deposits and 71 low-interest demand deposit instrument used primarily by institutions with special requirements. percent of the 16.1 The interest rate structure million accounts The annual interest rates of the main unit desa savings instruments are reviewed monthly by the management of the unit desa system, together with BRI's Asset and Liability Coordinating Committee. Rates are adjusted to reflect a number of factors, including changes in the money market, the liquidity needs of the units and of BRI, and the activities of the competition. Interest is calculated on the minimum monthly balance for SIMPEDES, SIMASKOT, and TABANAS. Above the minimum account size required to re- ceive interest,TABANAS has a single interest rate (13 percent in 1996). SIM- PEDES and SIMASKOT have graduated rates that are higher on larger balances (see table 13.6).This strategy is a direct result of two lessons from the pilot pro- ject experience. First, larger depositors are usually more interest sensitive than smaller ones, so the units must keep their rates for larger accounts competitive. Second, larger accounts are generally more profitable than small accounts be- cause operating costs represent a lower percentage of the average savings bal- ance in large accounts than they do in smaller accounts. Interest rates for Deposito Berjangka, which is offered throughout BRI, are set for the bank as a whole and vary with the maturity of the account. During 1984-96 the savings instruments offered in the unit desas gener- ally provided savers with positive real interest rates, with two exceptions: very small SIMPEDES, SIMASKOT, and TABANAS accounts; and giro accounts. 298 The Microfinance Revolution: Lessons from Indonesia Table 13.7 Estimated cost of savings mobilization in the unit desa system, 1996 (percent) Cost item As share of deposit balance Interest cost 12 4 Administrative cost 2 2 Cost of funds subtotal 14 6 Liquidity cost (cash holding 3 percent of deposits) 0 5 Cost of loanable funds subtotal 1 15 1 Cost of minimum reserve requirement (2 percent of deposits)a 0 3 Cost of loanable funds subtotal 11 15 4 The cost of unit Alternative, cost of borrowing from branches 15 5 desa savings a In early 1997 Bank Indonesia increased the minimum reserve requirement in two steps to 5 percent, which is required to be held as demand deposits with Bank Indonesia, carrying little or no mobilization in interest Source Maurer 1999, p 134 1996 was estimated Cost of funds at 14.6percent of Because the unit desa cost accounting system does not disaggregate between lending and savings activities, the administrative costs of savings mobilization the deposit balance cannot be precisely determined. As discussed in chapter 7, a study conducted for the Consultative Group to Assist the Poorest estimated the cost of unit desa savings mobilization in 1996 to be 14.6 percent of deposit balance (12.4 per- cent for interest and 2.2 percent for administrative costs),24 and the cost of loan- able funds (including the reserve requirement) to be 15.4 percent (table 13.7; for more details see Maurer 1999, p. 134). Between 1990 and 1996 the umut desas' cost of funds ranged from 9 percent in early 1994 to more than 19 per- cent during the tight money period of 1991 (BRI 1996a, p. 16). Lotteries The idea of usmg local lotteries as part of a financial savings product, developed in Indonesia in 1970 by Bank Dagang Bali, was adapted by the unit desas in 1984 in the Sukabumi SIMPEDES pilot project. Since 1986 BRI's unit desa system has held semiannual lotteries for SIMPEDES and SIMASKOT savers. Free lottery tickets are given to savers based on the size of their minimum monthly account balances; savers know that the higher their account balances, the higher their probability of wmning a lottery prize. In 1995 the aggregate value of the prizes was about 0.7 percent of the SIMPEDES balance and 0.8 percent of the SIMASKOT balance. As noted, SIMPEDES lotteries are held at BRI's branches at the district level. Since they occur relatively near the homes of savers, and since many partici- pants have either won a prize in the past or know someone who has won, there is considerable local enthusiasm for the lotteries. Prizes are given in kind; the first prize in each district is usually an automobile or motorcycle. Many small- Mobilizing Massive Savings Bank Rakyat Indonesia's Unit Desas, 1984-96 299 er prizes (television sets, radios, refrigerators, clocks) are also awarded. In the June 1996 lottery, held at BRI's 324 branches, there were about 16,000 win- ners.The average value of the prizes was about $260. SIMASKOT lotteries are held at BRI's 15 regional offices. Although the drawings take place at a considerable distance from the homes of many of the savers, the large cash prizes are attractive to SIMASKOT's more urban clients. In the first drawing for 1996, there were about 2,400 winners.The first prize was $65,000, and the average prize was about $435. A semiannual lottery with much smaller prizes is held for TABANAS savers, with cash prizes in 1995 totaling less than 0.15 percent of the TABANAS bal- ance. Since TABANAS pays the highest interest rates on small accounts, it has a In thefirst low-cost lottery so that the average cost of funds of SIMPEDES and TABANAS are about the same. SIMASKOT The transfer price lottery drawingfor As noted, the transfer price is the interest rate on funds that are deposited by a unit in its supervising branch or borrowed by the unit from the branch.25 This 1996, there were rate is adjusted monthly by BRI's head office according to the overall lhquidity position of the bank.The transfer price is used by BRI to indicate to the units the 2, 400 winners. The relative emphasis that should be placed on savings and loans. For example, a low transfer price sends a clear signal that the umts should concentrate on lending. But first prize was when BRI wants to encourage savings mobilization or slow down the rate of growth of lending, the transfer price can be raised substantially-as happened during the $65, 000, and the tight money period in the early 1990s and during the crisis in 1997-98. Generally the transfer price has been set slightly above the cost of mobiliz- average prize was ing savings at the units, thus creating an incentive for the units to collect savings rather than borrow from the branches. Because each umt is treated as a profit cen- $435 ter and the staff incentive programs are heavily based on each unut's profits, the transfer price has a strong, direct incentive effect on the behavior of unit employees. CGAP's study of savings in the units in 1996 (Maurer 1999, p. 131) found: 95% of the urnts are net fund providers and maintain considerable deposit balances with the branches.As a result, more than a third of the units' interest income m 1996 originated from mterest paid by the branches, whle two-thirds of interest mcome was generated from KUPEDES lending. For 1996, the average interest yield was 15.6% on [the units'] branch deposits and 30.2% on loans (net of borrower incentives), resulting in a weighted interest yield on earning assets (branch deposits and loans combined) of 22.8%. Box 13.3 shows lessons on cost reductions that emerged from examination of savings mobilization at BRI's unit desas. Emphasis is placed there on the in- stitutional viability of mobilizing small savings if cost elements and incentives are set correctly, the importance of the transfer price, and the reduction of trans- action costs for savers. 300 The Microfinance Revolution: Lessons from Indonesia Box 13.3 Excerpts from Klaus Maurer's "Bank Rakyat Indonesia (BRI) (Case Study)" * The general lesson learned is that the mobilization of small savings is costly, but it is a viable and profitable option for a financial institution if the various cost elements are seriously addressed and incentives are set correctly * Taking into account the lessons of transaction cost economics, BRI has applied a dif- ferentiated interest rate structure for SIMPEDES by tying the level of deposit rates to savings account balances Thus, lower interest costs compensate, at least in part, for the higher administrative costs of small accounts * Reducing the administrative cost associated with thousands of small accounts has been a major concern A number of innovative features have contributed to lowering these costs, e g, simplicity of design and standardization of operations, low overhead Probably the most costs, increasing staff productivity, computerization, etc * The opportunity cost of cash holding, or liquidity cost, was minimized by prudent cash essentialfactorfor management Above all, access to the BRI liquidity pool has significantly contributed to reducing liquidity costs * The framework conditions for savings mobilization have been extremely favorable as the success of small the minimum reserve requirement has been set at a low level of 2%, allowing 98% of deposits to be employed as productive assets (Only recently, the Central Bank has savings mobilization raised the minimum reserve requirements to 5% ) * Most importantly, however, and probably the most essential factor for the success of SIMPEDES and the mobilization of small savings, has been BRI's policy of the transfer has been BRIls price of funds from the branches to the units and vice versa This artificial, yet market- based, accounting price was set at a level equal or slightly above the cost of funds for transfer price policy mobilizing small savings This move has effectively eliminated the opportunity cost of small savings and has created a strong incentive for the units to engage in savings mobilization rather than borrowing from the branches * The transaction costs for savers have been limited primarily through the customers' physical proximity to the BRI units, which are located in the centers of the rural econo- my, the sub-district towns However, there are still many customers living in remote areas and villages who may find it costly to travel to the nearest BRI unit. The village service posts operated by some units help to bring down the costs but the number of these posts remains limited Source Maurer 1999, pp 135-36 Performance in Savings Mobilization, 1984-96 The unit desa system, which had moblhzed only $18 million between the early 1970s and the time of the financial deregulation in June 1983, proceeded to mobilize $3 bilion by 1996 (table 13.8).This spectacular change occurred pri- marily because after 1983, the unit desas had a strong incentive to mobilize sav- ings. Accordingly, BRI trained unit staff and their branch supervisors to learn about the markets they served, developed savings instruments designed to meet different types of local demand, provided performance-related staff incentives, and established a pricing system for savings and loans that would permit both wide outreach and unit desa profitability. Mobilizing Massive Savings Bank Rakyat Indonesia's Unit Desas, 1984-96 301 Table 13.8 Value of savings accounts by account type in the unit desas, 1984-96 (millions of current U.S. dollars) Account type 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 TABANAS 364 56 5 47 6 48 2 51 7 63 2 80 8 109 2 1464 194 9 242 9 273 0 3304 (93) (75) (45) (28) (18) (12) (9) (9) (9) 1101 1101) 11) (11) SIMPEDES 0 29 46 50 3 110 6 197 5 389 4 525 9 669.7 933 3 1,278 3 1,523 5 1,639 8 1,849 3 (1) (6) (47) (64) (69) (73) (59) (53) (57) (62) (64) (63) (62) SIMASKOT na n a n a n a n a 05 77 6 142 5 220 9 300 3 345 6 362 0 409 7 (0) (9) (11) (13) (15) (15) (14) (14) Timedeposits 07 19 24 77 257 652 169.9 2842 2501 1899 191 8 2641 3252 (Deposito Berjangka) (2) (3) (2) (4) (9) (12) (19) (22) (15) (9) (8) 110) (11) Other 19 12.4 67 77 9.8 153 37 3 696 977 86 5 743 67 5 61 2 (5) (17) (6) (4) (3) (3) (4) (5) (6) (4) (3) (3) (2) Total 393 754 107 0 1742 284 7 533 6 891 5 1,275 5 1,6484 2,049 9 2,378 1 2,606 4 2,975 8 n a Not applicable Note Numbers in parentheses are percentages of the total value of savings accounts Totals may not sum to 100 percent because of rounding Source BRI unit desa monthly reports Performance by instrument In 1996 SIMPEDES accounted for 62 percent of the unit desas' $3 billhon in sav- ings, followed by SIMASKOT (14 percent),TABANAS and time deposits (11 per- cent each), and all other deposits, including giro accounts (2 percent; see table 13.8). SIMPEDES has been the most popular account since 1990, the first year that the market penetration phase of the units' savings mobilization effort had been in full operation nationwide (table 13.9). In 1996, 63 percent of unit ac- counts were in SIMPEDES, followed by TABANAS (28 percent) and SIMASKOT (8 percent).Time deposits represented 1 percent of the units' ac- counts.All other account types, including giro, together accounted for less than 1 percent of the savings accounts in the unit desa system. In general, the package of savings instruments did what it was designed to do. SIMPEDES mobilizes most of the funds in most of the accounts. Its tiered interest rates have helped the instrument to reach millions of low-income savers and to attract and maintain the accounts of wealthier savers-thus raising the average account size and making SIMPEDES financially viable. At the end of 1996 the average SIMPEDES account was $183. The more than 4 millionTABANAS accounts are used predominantly by smaller savers who value its present-day liquidlity and want to maximize returns; 302 The Microfinance Revolution: Lessons from Indonesia Table 13.9 Number of savings accounts by account type in the unit desas, 1984-96 (millions) Account type 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 TABANAS - - - - 3 29 3 51 3 33 3 34 3 48 3 68 3 97 4.22 4 55 (66) (56) (46) (39) (35) (32) (30) (29) (28) SIMPEDES 0 004 042 097 1 64 266 350 446 550 666 7.85 8.90 10.10 - - - (23) (33) (42) (48) (52) (55) (58) (60) (61) (63) SIMASKOT na na na na n.a 0 027 054 075 0.93 110 121 1.33 (0) (4) (6) (8) (8) (8) (8) (8) Timedeposits na - - - 001 004 010 018 015 0.10 008 009 011 (Deposito Berjangka) (0) (1) (1) (2) (2) (1) (1) (1) (1) Other n a - - - 005 0.05 006 007 0 07 0.06 006 0.06 006 (1) (1) (1) (1) (1) (1) (1) (O) (O) Total - - - 420 499 626 7.26 8.59 995 11 43 1306 1448 16.15 - Not available n a Not applicable Note Numbers in parentheses are percentages of the total number of savings accounts Totals may not sum to 100 percent because of rounding Source BRI unit desa monthly reports in 1996 the average TABANAS account was $73. SIMASKOT, which effec- tively began in 1990 in BRI's urban uruts,26 had 8 percent of accounts and 14 percent of funds; in 1996 its average account size was $308.Time deposits (De- posito Berjangka) accounted for only 1 percent of accounts in 1996, but with an average account of $2,956 they were responsible for 11 percent of the units' savings. SIMPEDES, SIMASKOT, and TABANAS are stable sources of funds for the units. Deposito Berjangka is more volatile because its account holders are interest sensitive and because their accounts are relatively large. Giro, used pri- marily by institutions, is even more volatile than Deposito Berjangka, but it ac- counted for less than 2 percent of unit desa deposits in 1996. With 16.1 million savings accounts and 2.5 million outstanding loans at the end of 1996, the unit desa system provided savings services to an esti- mated 25 percent of Indonesia's approximately 46 million households. BRI does not record the number of individuals who have savings accounts. Since many households and some individuals have more than one savings account, since all borrowers have savings accounts, and since some accounts are in the names of institutions and organizations, the exact number of households served Mobilizing Massive Savings Bank Rakyat Indonesia's Unit Desas, 1984-96 303 by the units is not known. However, 25 percent of the country's households seems a reasonable estimate for the units' household outreach in 1996. As BRI has learned since the mnid-1980s-and Bank Dagang Bali and many People's Credit Banks (BPRs) learned earlier-in Indonesia there are more house- holds and small and microbusinesses with demand for secure savings facilities than with demand for small loans. Thus at any one time, nearly all unit desa clients want to save, but not all want to borrow.27 In 1996 the units had more than six times as many savings accounts as loans, and the value of savings was nearly twice as much as that of loans. At the end of 1996 the ratio of clients who saved in the umts but did not have an outstanding loan to those that borrowed and saved in the units was probably about four to one. At any one time, In addition to the ways in which the units' savings program benefits its clients (both individuals and organizations), low-income households and nearly all unit desa small and microenterprises have benefited considerably from the expanded volume of institutional credit financed by unit desa savings.However, there clients want to save, is still extensive unmet demand for small loans in Indonesia.And as BRI's ac- tivities now shift toward increased emphasis on micro and retail banking (chap- but not all want to ter 15), unit savings can be expected to finance an increasing number of KUPEDES loans. borrow Notes 1.When referring to financial savings, the terms savings and deposits are treated synonymously here. 2.This section is based m part on CPIS Rural Savmgs Reports 1 (1984a) and 2 (1984b), which concern the views of potential savers and contain early work on product design and testing. 3. In 1984 CPIS, with HIID technical assistance funded by the Ministry of Fi- nance, began a Rural Savings Group. This group, which I coordinated, worked close- ly with BRI on savings mobilization through 1990. Much of this chapter is based on the work of this group. 4. This research involved extensive, in-depth interviewing of men and women of different ages, with varied occupations and geographic locations. Since savings is a sen- sitive topic, this method was considered more useful than a formal survey in which re- spondents mught be reluctant to answer questions or might not answer them accurately. Surveys were carried out later to learn about savers' responses to particular products. 5. The discussion here concerns primarily rural Indonesia. However, field work in the late 1980s, conducted in preparation for the opening of unit desas in cities in 1989, indicated that the savings needs of people in low- and lower-middle-income urban neighborhoods are similar in many respects to those living in villages. 6.The five years referred to both HIID/DPIS and HIID/CPIS field work with BRI carried out during 1979-84. 7.This section is based in part on CPIS Rural Savings Report 2 (1984b). 8.The Survel Sosial Ekonomu Nasional (SUSENAS) is a periodic nationwide sur- vey that reports on consumption, income, and other socioeconomnc data. 9.This was an important methodological finding. It meant that in the future, unit desa staff inWestJava could obtain from these lists the names of villagers who appeared 304 The Microfinance Revolution: Lessons from Indonesia to be potential savers, and could then contact them individually to try to collect their savings. Such lists were not kept in other parts of the country, and different methods were developed elsewhere, as discussed later in this chapter. 10.This figure excludes savings accounts with balances below 10,000 rupiah ($9). 11.Annual inflation in 1984 was 10.4 percent. 12.TheTABANAS interest rate on larger accounts was 12 percent. 13. Discussion of the Sukabumu pilot project is based in part on CPIS Rural Sav- ings Reports 3 (1985a) and 4 (1985b). 14. In 1984 the grants were 1 million rupiah ($931) per village. 15. R.C.G.Varley was the first to call attention to this important problem. 16. Richard T. Monteverde provided the technical assistance to BRI on the size distribution study of SIMPEDES accounts in Sukabumi. 17. Discussion of the second-stage pilot project is based in part on CPIS Rural Savings Report 5 (1985c).The first pilot project continued in Sukabumii, but with the original SIMPEDES instrument. In 1986, when the revised savings program was ex- panded nationwide, it was introduced in Sukabumu as well. 18.This method requires a skilled and locally knowledgeable person to organize and run the discussion.While the method did not always work well, in most cases it was quite effective for estimating mnummum savings potential in the village. Followup was then required to identify the households in the various categories. 19. CPIS Rural Savings Report 6 (1986a).The cases were prepared by the CPIS Rural Savings Group working with BRI;James R. Kern and I served as the coordi- nators for CPIS work on the casebook and as the authors of the book. 20. A pilgrimage to Mecca (Saudi Arabia) during Dhu'l Hijja, made as part of the religious life of a Moslem. Haji is often used as an honorific title for one who has made such a pilgrimage. 21.The increase in rupiah terms was 107 percent (from 84.7 billhon rupiah to 175.8 billion rupiah). However, the end-of-year exchange rate for $1 was 1,125 rupiah in 1985 and 1,641 rupiah in 1986. 22.Although earlier research had been conducted outside Java, all the SIMPEDES pilot branches were selected onJava to facilitate supervision and monitoring.Therefore, the 1986-87 research was concentrated on the Outer Islands to ensure that BRI would receive input on the umts' new products and services from all parts of the country. 23.This section draws substantially from BRI (1996a and 1997b) and from dis- cussions with Richard M. Hook and Patricia Markovitch. For discussion of unut desa savings, see Robinson (1992b, 1994a, 1994b, 1995a, 1995b, 1997c, 1998a);BRI (1996a, 1997b); Charitonenko, Patten, andYaron (1998); Institute for Development of Eco- nomncs and Finance and BRI (1999); Patten, Rosengard, andJohnston (1999); and Mau- rer (1999, forthcomung). See additional references in volume 1, chapter 2, note 16. 24. This study assumed, based on a previous transaction cost analysis at BRI urnts, that one-third of admmnistrative costs is related to deposit mobilhzation and two-thilrds to lending. 25. A change in transfer price normally does not directly affect the profitability of BRI as a whole.The transfer price can, however, affect bank profitablity indirect- ly by influencmg the behavior of unit staff.Thus if the transfer price is raised substantially during periods of liquidity shortage, unit staff will be motivated to raise additional sav- ings; this provides the bank with a lower-cost alternative to other sources of funds. 26. Urban units were first opened in 1989 and SIMASKOT was tried on a very small scale that year It was offered in all urban units in 1990. 27.All unit desa borrowers are required to have savings accounts because the KU- PEDES prompt payment incentive is paid into a savings account. It appears that a few borrowers use their savings accounts only for that purpose but that most use them for other purposes as well; the numbers are not known. Mobilizing Massive Savings Bank Rakyat Indonesia's Unit Desas, 1984-96 305 Institutional Development for Large-Scale Sustainable z Microfinance: The Transformation of the Unit Desas, 1984-96 The principles of sustainable institutional microfinance carry with them a set of basic operational requirements.1 Effective operations, in turn, depend on an efficient orga- nization and management structure that are focused on mi- crofinance. Such an organization will not arise in the absence of the underpinning principles. Yet putting the principles into operation on a large-scale, long-term basis can be done only through appropriate institutional organization. How can the sequencing work in such a situation? This is a critical problem in microfinance, and one that is little understood. Therefore this chapter explores the dy- namics of organizational development at Bank Rakyat In- donesia (BRI); its purpose is to provide an example of the sequencing process that occurred there. But the process at BRI 306 was long and difficult. It is hoped that other institutions will find lessons in the process, rather than replicate it. Microfinance can operate viably in a number of quite different or- ganizational structures, as is discussed further in volume 3. But essential features of microfinance operations and management must be incor- porated into the organizational design of any self-sufficient institu- tion. For example, a number of components must be in place to achieve the loan recovery rates above 95 percent that are needed for institu- tional sustainability. Included among these are effective, committed man- agement; an efficient, accountable, labor-intensive orgamzaton; simple, Thefocus here is on transparent accounting and reporting systems; careful asset-liability man- agement; appropriate staff training and performance-based incen- how basic operating tives; decentralized authority for loan decisions; effective measurement requirements Of and management of delinquency, including appropriate mechanisms for loan provisioning and write-off; regular and meticulous internal commercial supervision; simple, suitable management information systems; and user- friendly instruments and services priced for institutional viability. microfinance drove Whatever its structure, a sustainable microfinance institution must be organized to provide these requirements. the transformation of When BRI's unit desa system changed in 1984 from a loss-mak- unit desa ing channeling agent for government credit subsidies to a commercial financial intermediary, it did not have in place an organizational struc- organization and ture that could provide the prerequisites of institutional sustainability. The focus here, therefore, is on how basic operating requirements management at BRI of commercial microfinance drove the transformation of unit desa organization and management. Complex organizational changes have taken place within BRI since the mid-1980s, and they are still occurring.The key elements of the unit desa organization that have emerged thus far include strong, effective leadership; the transformation of the units from branch windows to individual profit centers; the growth of the old unit desa system into a full bank division, with internal restructuring and rep- resentation at all regional and branch offices; an entirely new ap- proach to human resources; the development of large, state-of -the-art training centers with a curriculum designed specifically for microfinance; an intensive system of supervision of the units by spe- cially trained staff at the branch and regional offices; and-most im- portant-in the words of BRI's President-Director Kamardy Arief, Institutional Development for Large-Scale Sustainable Microfinance The Transformation of Bank Rakyat Indonesia's Unit Desas, 1984-96 307 who was appointed in 1983 to lead the reforms and served until 1992: "a new BRI culture." These and other aspects of the restructuring of the units will be analyzed. But the transformation of the unit desa system is not com- prehensible without an understanding of the role of its owner, the Ministry of Finance,2 and of other government agencies.While no longer financing the units, the government-especially the Ministry of Finance and the coordinating minister for the economy-provided them with continuing support of other kinds, both formal and in- formal.This strong comnitment to sustainable banking for the eco- The government's nomically active poor was a sine qua non for the transformation of strong commitment the system. Over the years, as financial policy was made and bank reg- ulations set, the units consistently had powerful patrons who under- to sustainable stood their needs and safeguarded their ability to provide commercial microfinance. In addition, the Ministry of Finance funded substan- microfinance was a tial foreign technical assistance for the unit desa system from 1979-90; sine qua nonfor the the consultants worked directly for the minister. Beginning in the mid-1980s funding and technical assistance for transformation of the unit desas were also supplied by donors (after the units were al- ready profitable). The donors and their consultants contributed to the system global dissemination of the best practices of Indonesian microfinance. This effort helped significantly increase awareness of sustainable microfinance abroad and buttress the position of the units at home. But as part of a larger bank, the organization of the new umt desa division was not without its problems. Directors of other divisions of the bank did not necessarily understand the units' needs. Unit prof- its supported loss-making bank divisions that provided loans at lower interest rates to larger borrowers, with low repayment rates. And the managing director responsible for the unit desa system did not have the authority to make some of the basic decisions required for mu- crobanking, nor did he control the support to his division provided by other bank divisions, such as personnel and logistics.The success of the units depended partly on an informal system of personal rela- tionships among the directors-a system that was vulnerable to change as directors came and went at BRI.After the Indonesian cri- sis began in 1997, the government addressed these problems. BRI now concentrates primarily on providing commercial finance to micro, small, and medium-size retail clients (see chapter 15).This chapter consid- ers the organizational development of the unit desas during 1984-96. 308 The Microfinance Revolution: Lessons from Indonesia The main differences between the old and new cultures in BRI's unit desa sys- tem are shown in the box that appears in the introduction to part 3.3 As indi- cated there, the emergence of the new culture was driven by the reorganization and new management of the unit system, by fundamental changes in the role of the units, by new views of staff responsibility and accountability, by a mas- sive upgrading of human resources, and by basic changes in accounting, reporting, and information management. Old and New Cultures In the old culture, discussed in chapter 11, the units were branch windows for The new culture of channeling government funds and their performance reports were aggregated with those of the branch. Unit staff were paid low salaries and had no career professionalism, track within BRI. Staff authority and accountability were exceedingly limit- ed.Training was poor and performance-based incentives nonexistent.And su- transparency, pervision of the units was weak and irregular. In 1983 the newly appointed managers at the head office inJakarta real- efficiency, initiative, ized that the system was steeped in bureaucracy and seniority, suffered from in- efficiency and lack of initiative, and offered many opportunities for corruption and accountability through the disbursement of credit subsidies.The unit desa system would have to be fundamentally restructured if the new culture of professionalism, trans- required that the parency, efficiency, initiative, and accountability were to take root. When the restructuring of the unit desa system began in 1984, it soon be- unit desa system be came apparent that building the institutional capacity needed for commercial mi- crofinance operations would require extensive reorganization within BRI. Figure fundamentally 14.1 illustrates major aspects of the processes that occurred in building an orga- nization that would meet the operating requirements of sustainable mucrofinance. restructured A loan instrument had to be developed and implemented that would en- able the units to provide large-scale, profitable outreach to low-income clients. To achieve that goal, the unit desa system would have to take on major new functions, including developing a new type of borrower selection process, new methods of loan collection and delinquency measurement and management that would produce a very high on-time repayment rate, and new systems of accounting, reporting, and information management. Similarly, a new savings program would have to be designed and implemented to meet the widespread demand for savings services and to finance the KU- PEDES loan program (see chapter 13). This required generating the capacity to develop appropriately priced savings instruments with the characteristics that savers demanded, to identify potential savers in far-flung vlllages and to col- lect their savings, and to provide savers with the security, cash-flow manage- ment, and services that would induce them to open and maintain accounts at the units. The new system had to develop the ability to design, price, and implement together these new lending and savings instruments, to create an appropriate Institutional Development for Large-Scale Sustainable Microfinance The Transformation of the Unit Desas, 1984-96 309 Figure 14,1 Restructuring unit desa organization and management Element Goal Institutional capacity Organizational changes required Element Goal needed to reach goal to develop capacity * Capacity to design and implement a commercial loan product that meets microcredit demand To provide and profitably recover small * Specialized staff training in loans from low- borrower identification, income clients selection, and loan KUPEDES nationwide, collection enabling unit * Development of systems desa profitability for tracking loan status, * Units changed from branch and large-scale reporting performance, and Untchgefrmbah outreach meaurting performanc,aging windows to individual profit measuing and managing centers delinquency * Job descriptions of unit staff * Maintenance of changed unit manager given loan consistently high on-time authority and overall responsibility, repayment rate manager and credit officer held * Capacity to design, test, responsible for their financial To mobilize and implement savings decisions, all staff held public savings products that meet varied accountable for performance on a large local demand and are priced * Substantial upgrading and scale in order for unit profitability augmentation of unit supervisory Savings to meet * Specialized staff training in staff at branch and regional levels demand from identifying potential savers * A new unit desa division small savers and capturing their savings established at the head office, and to finance * High-quality security, cash- reporting to a BRI managing KUPEDES flow management, and director. This division develops loans customer services and oversees the unit desas and coordinates their activity with * A spread between loan and other bank divisions svnsinterest rates that *Establishment of training centers savings Institu t hal for unit desa staff and their enables Institutional profitability supervisors, development of a * Effective asset-liability training curriculum appropriate for management microfinance * A simple, transparent, cash- A profitable, based accounting system nationwide * A transfer price, reviewed system of on a regular basis, that Unit desa financial signals to the units the system intermediation relative emphasis to be delivered placed on lending and locally to low- savings mobilization income clients * A career path, salary scales, incentives, and training program that encourage good staff performance and high productivity * Efficient operations and management information systems 310 The Microfinance Revolution: Lessons from Indonesia transfer price mechanism, to establish effective and efficient operating proce- dures, to develop a new staff training program designed for commercial mi- crofinance, and to provide the supervisory process and performance-based incentives necessary for high staff productivity. Fundamental organizational changes would have to be made to implement, supervise, and control such a system. At the lowest level, the structure of the units would have to be changed from branch windows to individual profit cen- ters. Unit staff would have to be given more responsibilhty, provided the au- thority to carry out their new tasks, trained in the skills needed to achieve good performance, offered the incentives to achieve high productivity, and held ac- countable for their performance. Analogous steps would have to be taken at branch and regional offices to The cultural enable the effective supervision of unit activities, and at the head office to de- velop appropriate information management, coordinate planning, and main- transformation of the tain effective control over the system. In addition, training centers designed to train unit desa staff and their supervisors in microbanking would have to be unit desa system did established and the trainers trained. Underlying and accompanying the process of institutional capacity build- not happen ing and organizational restructuring was a newly emerging unit desa philosophy-a philosophy that began to be captured in such key words and overnight. It was a phrases as trust, simplicity, transparency, accountability, commitment, service, prof- itability, security, standardization with flexibility, customer convenience, mar- long, hard process ket knowledge, incentives to staff and clients, specialized staff training, supervision, and the new BRI culture. Of these the most important was the BRI culture, which integrated the others.As Kamardy Arief, BRI's president-director, said: "We are building a new culture at BRI." It was that new culture, with all its components, that converted the unit desas that no one wanted into today's preeminent international model for large-scale sustainable microfinance. Restructuring the Unit Desas: The Transition Process The cultural transformation of the unit desa system did not happen overnight. It was a long, hard process that experienced spurts and setbacks. The changes occurred over many years and involved interlinked innovations and reforms of multiple kinds and in different timeframes.These included the transformation in operations, human resources, organization, and management that are discussed in this chapter, as well as the transition to the new ideas and products that were analyzed in earlier chapters.While the organizational structure evolved piece- meal over time, the continuity required to change the culture was provided by its high-level management.The late Sugianto, BRI's managing director responsible for the unit desa system from 1984 until his untimely death in 1998, was es- pecially important in this context, as he provided the long-term leadership need- ed for the development of the system.4 Institutional Development for Large-Scale Sustainable Microfinance The Transformation of the Unit Desas, 1 984-96 311 From branch windows to profit centers At the beginmng of 1984 all unit desas were changed from branch windows to individual profit centers. This had to be the first step of the restructuring process because it laid the foundation on which the entire system would then be built. Once the performance of individual units could be measured simply and accurately and reported regularly, staff could be held accountable, super- vision could be concentrated where needed, incentives and promotions could be based on indicators of unit performance, and each unit desa could be helped to attain outreach and profitability. To start, each unit was provided with 19 million rupiah in equity (about $19,000). Bookkeeping, reporting, and audit systems were completely trans- Until 1984 each formed. Each unit was required to maintain its own balance sheet and profit and loss statement and to report these monthly to its higher-level supervisors. unit produced 32 Units with outstanding loans above 19 million rupiah were now consid- ered as having borrowed the remainder from their supervising branches. How- monthly reports. ever, the units served as agents for the branches in collecting old loans. Unit liabilities were transferred to the supervising branches. Most were After the new system began, interest paid on a unit's KUPEDES loans was retained by the unit. Similarly, savings mobilized by a unit were no longer trans- redundant, useless, ferred to the branch but were considered a unit liability.The transfer price mech- anism was introduced: a unit with excess liquidity deposited funds at its or both. The reports supervising branch and earned interest on the deposits, while a unit with out- standing loans higher than total deposits borrowed from the branch. The in- were eventually terest rate for liquidity credit from branches to units was set higher than the highest deposit interest rate to encourage units to mobilize savings by making reduced to 5 it less expensive to do so than to borrow from the branch. Accounting standards and procedures were fundamentally changed.The pre- vious system of accrual of income earned but not yet paid was ended; only paid interest was now posted as income. Reserves for bad debt were instituted based on unit arrears, and reserves for payment of mterest on unit savings were required. Until 1984 each unit produced 32 monthly reports. Most were redundant, useless, or both, and they were often submitted too late to be of use for man- agement purposes.Yet preparing the reports occupied a large amount of staff time.The monthly reports required from units were eventually reduced to 5- which provide far better information than the 32 earlier reports. Since the profit and loss position of each unit was now available monthly, unit staff were able to take timely actions to help the unit widen outreach and achieve, or increase, profitability. In addition, the supervisors of the units at the branch and regional offices could review the performance of the units each month, and could take immediate action as necessary. Head office managers could review overall monthly progress by unit, branch, and region. This meant that poorly performmng units and their supervisors could be visited quickly to iden- tify problems and to institute corrections, while unusually well-performing units could be visited to learn successful techniques that might be replicated or adapt- ed by other units. 312 The Microfinance Revolution: Lessons from Indonesia These components of the new management information system were es- sential for the development of the new culture. In contrast, while computeri- zation ultimately played an important role in the development of the unit desas, it was not an integral part of the cultural transformation.The units were com- puterized gradually; most became profitable under manual accounting and re- porting systems. Only by 1997-long after the success of the unit desa system-did every unit have a computer. Changes in organization and management As the basic changes that would be required to transform the units into prof- itable microfinance intermediaries began to become clear, and as the implica- tions of these changes for organizational restructuring began to be understood, Most units became the process of reorganizing the system began.5 It took place at all levels-units, branches, regional offices, and the head office. profitable under In 1984 BRI's head office accorded high priority to effective management of the unit desa system. This was, of course, the first year that the system was manual accounting designed so that it could be managed effectively. Until then BRI's management of unit staff was deeply flawed. Kept out- and reporting side the "BRI family," unit staff were essentially treated as outcastes. On a sub- stantially lower salary and benefit scale than other BRI staff, they were not trained systems. It was long in the skills needed for local banking.They were responsible for operating cred- it and savings programs that were impossible to implement profitably. They had after their success no career track and could not be promoted to positions in the rest of the bank. And they were excluded from participating in BRI corporate events. that every unit had As a first order of priority, therefore, unit desa management at the head of- fice turned its attention to bringing unit salary scales in line with those of the a computer rest of the bank. After that the head office established a system of promotion that incorporated the unit staff into the wider BRI personnel system and de- veloped performance-based promotion criteria. Skill development was carried out concurrently, with unit staff beginning to be trained in basic knowledge of local financial markets, in the implementation of the new credit and savings instruments, and in the operations required for sustainable microfinance. The head office. BRI's president-director and its managing director responsible for the unit desas changed-gradually but radically-the internal supervision of the unit desa system. Unlike the reorganization of the system at the unit and branch levels, which happened quickly, head office restructuring took consid- erable time. Misunderstanding of the new role of the unit desas within BRI, combined with internal vested interests, made it difficult and time-consuming to make changes at the head office. As a result implementation of head office reorgamzation took place in stages, driven at each stage by the needs of the unit desa system-but more to the point, by its growing influence within BRI, a result of the units' ncreasing contribution to bank profits and liquidity. Until 1984 BRI's Cooperative and Program Credit Division, which was responsible for subsidized rural credit at BRI, also provided national and re- Institutional Development for Large-Scale Sustainable Microfinance The Transformation of the Unit Desas, 1984-96 313 gional supervision of the units.This arrangement was appropriate because the units had been established to administer the division's loans. But the units' new credit and savings products were based on commercial principles that were fun- damentally different from those underlying the bank's ongoing government- subsidized loans. Thus it was no longer appropriate for the unit desas to be supervised by the Cooperative and Program Credit Division. Where did the reformed unit desas belong organizationally? Nowhere. In 1984 three BRI divisions (urusan) were responsible for the bank's loans: the Co- operative and Program Credit Division, the Commercial Credit Division (for credit much larger than KUPEDES loans), and the Corporate Credit Division. Each division reported to a different managing director. It was clear, however, The unit desa that the KUPEDES credit program did not fit into any of these divisions. Therefore, when KUPEDES began, a team headed by BRI's Division for system was a success Organization was assigned responsibility for the units. But since KUPEDES grew rapidly, it was soon decided that the units needed to be placed in a division re- in search of a home sponsible for supervising credit. Once again, the Cooperative and Program Cred- it Division assumed responsibility for the units.This tune, however, a new section in the BRI (bagian) called Business Unit Desa (BUD) was created within that division. Primarily because of the personal commitments and leadership of BRI's organizational president-director and the managing director responsible for the unit desas, BUD was well managed-even during the years when an appropriate place for the structure system within BRI's organizational structure had not yet been established. Dur- ing much of the second half of the 1980s, the urnt desa system was a success in search of a home in the overall BRI structure. Although hardly a textbook example of good management, it worked as a temporary measure. Then, as the business of the units continued to expand rapidly, BUD out- grew its status as a section. In addition, it clearly no longer belonged in the di- vision to which it was assigned. But BUD could not reasonably be put into any other division.The solution eventually reached was to upgrade the BUD section to the BUD Division; the head of the new division reported directly to the managing director responsible for the unit desa system. This process of reorganization-driven dlrectly by the performance of the new unit desa products and services, and indlrectly by the broader strategy that guided the units' extraordinary performance-created a major shift in organi- zational structure at the bank.The change engendered considerable controversy. But as the units continued to earn more and more of the bank's profits, it was not possible to reverse direction. Instead, the BUD Division continued to in- crease its role within BRI. The regional and branch offices. At the regional offices new BUD sections (bagian) were created to oversee and support unit activities.These sections were charged with reviewing and correcting branch supervision of unit desa activities. At the branch level much more rigorous supervision was instituted than had been present before 1984. Previously there had been one branch supervi- sor for every 6 unit desas and one higher-level unit development officer for 24 314 The Microfinance Revolution: Lessons from Indonesia or more units.6 But neither the number of supervisors nor the usual quality of the BIMAS-era incumbents was adequate for the new system.Therefore, as dis- cussed later in this chapter, high priority was given to developing an intensive supervision system with fewer urnt desas per supervisor; these supervisors, called unit business managers, were given special training specifically related to their new job descriptions. Each branch supervisor is assigned a maximum of four unit desas; branches with nine or more units also have a coordinating and su- pervising unit development officer, and branches with many units may have an additional unit development officer. Supervisors of the unit desa system at the branch and regional offices were now held directly accountable for the financial performance of the units they supervised. Unit staff were now The unit desas. Incorporating unit staff into the BRI corporate structure given responsibility, was an essential first step in building the human resources of the unit desa sys- tem.As BRI's salary structure, promotion criteria, and career track opportuni- authority, training, ties were revised to include the units, the responsibilities of unit staff were changed dramatically. Until 1984 unit staff had httle authority and even less account- and performance- ability. They were now given responsibility, authority, training, and perfor- mance-based incentives-along with accountability and supervision. The based incentives- minimum staff for a unit continued at four: manager, credit officer, bookkeeper, and teller. But new staffing standards were established so that staff could be added along with based on a unit's workload. When the accounting mechanisms needed to evaluate each unit as an in- accountability and dependent profit center were instituted in 1984, it became possible to require and enforce staff accountability. Until then, except for legally proven fraud cases, supervision staff could not be held accountable for their performance because in most cases performance could not be evaluated. At best, unproductive or dishonest staff could be transferred. I was told of an instance during the 1970s when a unit manager was found to have stolen money from his unit. Since the case was not taken to court, he could not be dismissed. However, the unit manager was de- moted-to teller! The restructuring of the unit desa system incorporated basic changes in the allocation of staff responsibilities.Within BRI's overall rules, each unit manag- er was given authority to approve loans up to a maximum amount that was based on the manager's experience and performance.The unit manager and the credit officer were accountable for the quality of the loans they made, and unit staff were evaluated on the basis of the financial performance of their unit. Staff promotions and incentives were allocated accordingly. For the first time, women were recruited as unit staff. Staff training and incentives. In the BIMAS era unit staff had been given little training, and certainly none that would enable them to implement a sus- tainable system of microfinance. However, the goals of the new system required the development of a new kind of training. Institutional Development for Large-Scale Sustainable Microfinance The Transformation of the Unit Desas, 1984-96 315 The focus was on teaching unit staff the elements of the new banking sys- tem, including accounting, bookkeeping, operations, and financial analysis; on study- ing local financial markets, business networks, and household income flows; and on developing staffcapability in customer relations and publicity.The trainers were an ad hoc assortment of BRI directors, trainers, regional and branch staff, CPIS staff, and HIID consultants. One of the most important shifts from old to new cultures occurred when head office managers began visiting units, not only to give instructions but also to listen to unit staff. And umt staff had begun to lis- ten to their customers. Unheard of in BIMAS times, substantial information flowed from unit staff to the head office and was incorporated into training materials. In 1986-87 a systematic retraining of all unit desa staff began; the purpose An important was both to upgrade the staff and to ensure that new unit procedures became standardized throughout the system.As discussed later in this chapter, five large cultural shift training centers for unit desa staff and their branch supervisors were built in different parts of the country. An excellent example of how donor funds can occurred when head be used well in developing sustainable microfinance, the training centers were constructed in large part with funding from the U.S. Agency for Internation- office managers al Development. Under the old culture subsidized loans were provided at a below-market visited units not effective interest rate of 12 percent a year and were rationed.As a result staff- poorly paid and poorly supervised-had an incentive to allocate loans to ap- only to instruct, but plicants wealthy enough to pay bribes or influential enough to provide staff with other benefits.The new reforms addressed the incentive problems directly. First, to listen widely available KUPEDES credit at a 32 percent annual effective interest rate provided far less opportunity for staff corruption. Second, a special incentive system based on unit profitability was institut- ed for unit staff.Ten percent of each unit's annual profits was given as a bonus to all the unit's employees, up to 1.5 percent of an employee's monthly salary. However, no incentives were provided to the units' supervisory staff at branch level; it was thought that it could prove counterproductive to offer incentives to some branch staff (those supervising the units) and not to others (those work- ing on retail, commercial, or corporate banking).The lack of incentives for su- pervising staff became a long-standing source of difficulty. But for unit staff the profitability bonus provided a strong incentive for good performance. Other performance-based incentives for unit staff and their branch super- visors were developed in stages; these are discussed later in this chapter. In the early years of the transition, however, the most powerful incentive for produc- tivity was that the new system had only two years to prove itself, that the gov- ernment would not bail it out if it did not do so, and that unit staff would lose their jobs if the unit desa system failed to attain outreach and profitability. Unit desa location Rural units. Because the unit desas had been developed primarily to chan- nel BIMAS credit to rice farmers, they were originally located near irrigated ricefields (sawah). Units were opened in proportion to the extent of the sawah 316 The Microfinance Revolution: Lessons from Indonesia to be covered: one unit desa per 600-1,000 hectares of sawah on Java, and one unit per 1,200-2,000 hectares offJava. With the advent of KUPEDES and the new savings instruments, howev- er, the situation changed. Since these products were designed to appeal to a broad base of small borrowers and savers, the units would now have to be located near local market centers where they would be easily accessible to all kinds of po- tential clients. In addition, the widely dliffering productive activities in differ- ent parts of the country made a uniform approach to the location of unit desas inappropriate.Thus a more flexible method of relocating units and opening new ones was initiated. At that time nearly all unit buildings were rented. Thus, in theory, it would not be difficult to move those that were unsuitably located to sites with more 10 percent of each potential for microfinance. But the realities were more complex. In 1984 the head office instructed branch managers to divide all unit desas unit's annual profits into two categories:"potential'" and "nonpotential." Potential units were defined as those estimated to have the capacity to cover the cost of a four-person unit. was given as a Nonpotential units were those believed not to have the possibility of covering these costs. Branch managers were instructed to relocate nonpotential units or bonus to all the to combine them with nearby potential units. Most nonpotential units would then be designated as village service posts (pos pelayanan desa, or PPDs); these unit' employees were to be kept open for hlmited transactions during restricted periods. In most cases, however, branches were instructed to classify the potential of the units under their supervision before their managers and staff had been trained to assess rural banking potential. Many branch and unit managers knew little about local productive activities, financial markets, business networks, and income flows-or how to identify creditworthy borrowers. Virtually none knew anything about savings potential. As a result branch managers steeped in the old culture assumed that a unit desa that had not been active in the past would not be active in the future, or that one with high loan delinquency in the past would have high delinquen- cy in the future. Many branch managers interpreted the instructions to relo- cate units as a license to close them-since there were, they thought, no locations with banking potential in their service areas. The potential of units that served areas with significant nonrice economic activities, or those with sub- stantial savings possibilities, went unrecognized. In addition, areas considered to be economically below average or poor were ruled out as unit sites by near- ly all branch managers. Later it was discovered that even in quite poor areas there can be considerable banking potential-primarily for savings mobiliza- tion. The reason is that many people in poor regions leave to find work else- where; often they then send remittances home to their families. At the end of 1983 there were 3,626 unit banks; the number dropped by nearly one-third (to 2,469) by the end of 1984 and to 2,273 by 1986 (figure 14.2). On paper there were nearly 1,200 village service posts, a re- sult of the relocation of unit desas. But many of these posts existed only on paper.Worse, some that were announced as being open on certain days were Institutional Development for Large-Scale Sustainable Microfinance The Transformation of the Unit Desas, 1984-96 317 Figure 14.21 Number of Fi ure 14.2 unit desas, 1983-96 Many branch managers i nterpreted the instructions to relocate units as a Source BRI unit desa monthly reports license to close them often closed on those days. Customers who traveled long distances to ser- since they saw no vice posts and found them closed when they were supposed to be open rapid- ly became disillusioned about the prospects of banking at their local unit locations with desa post. The closing of more than a thousand units in 1984 happened so quickly banking potential that it took until mid-1985 for the head office to realize fully the extent of the decrease in units. By 1986, however, high priority had been placed on train- ing local staff in recognizing and assessing banking potential in rural areas, and in appropriate methods of relocating unit desas.The number of umt desas in- creased steadily after 1986, reaching 3,595 by the end of 1996, with 392 addi- tional village service posts-which have been consistently open on the dates and at the hours specified. In the late 1980s BRI began committing substantial resources for the opening of new units in areas with good banking potential and for significant improvements in buildings, communications, and equipment. In some areas BRI purchased the buildings in which the units are located. Urban units. Because the unit desas were established as part of the gov- ernment's rice intensification program, at first they were located entirely in rural areas. But by the 1980s substantial rural-urban migration, extensive circular mi- gration (in which a household or some of its members moves for part of the year to work elsewhere and returns home for part of the year), and widespread urban-rural remittances7 and other financial transfers had created an environ- ment in which considerable opportunity existed for financial intermediation between rural and urban areas. 318 The Microfinance Revolution: Lessons from Indonesia A few years after the units' new products had been introduced in rural areas, BUD managers at the head office began to believe that much of what had been learned about rural nucrofinance mught also apply in low-income urban neigh- borhoods. Market research confirmed that the demand for financial instruments and services appeared to be simnlar in many respects.Therefore, in 1989 urban units (unit kotas) began to be opened in low-income urban neighborhoods, with rapid and positive results. By the end of 1990 urban units accounted for 622 (20 percent) of the 3,040 unit banks then in the system. The urban umts had $182 million in KU- PEDES loans outstanding, about a quarter of the 1990 outstanding loan bal- ance of $727 million. In addition, they had $284 million in deposits, about 32 percent of the $892 million in the unit desa system.The urban units account- In 1983 there were ed for 26 percent of umut bank profits in 1990. After the urban units had been in place for a few years, BRI stopped 3, 626 unit banks; recording unit performance by whether the units were rural or urban. In ad- dition, some units came to be treated as semiurban; for example, they offer both the number dropped SIMPEDES (rural savings) and SIMASKOT (urban savings).Thus more recent data differentiating urban and rural units are not available. However, in 1996 to 2,273 by 1986 urban units probably accounted for about one-quarter of unit desas. The urban units started with the same products as the rural ones. Over time these instruments were modified somewhat to suit the urban demand for small loans and savings and to enable the units to operate successfully in a more com- petitive environment. Overall, however, BUD learned that low-income urban clients are much like their counterparts in rural areas: they respond positively to security, convenience, liquidity, products appropriate for their needs, good service, and a friendly environment. Critics have sometimes suggested that the umt desa system draws funds from rural to urban areas, thus contributing to urban bias.8 This is a complex issue. The urban units have been net savers from the first year they were opened. In 1990 the rural units were still net borrowers, but shortly afterward they became net savers as well. Much of the excess liquidity of the unit desas is lent to the branches, which are mostly rural. In addition, the prevalence of remittances from urban workers to their rural families suggests that some deposits in rural units originate from urban sources. However, in 1996,42 percent ofBRI's funds came from the units, while only 15 percent of the value of outstanding loans were from the unuts. Throughout the 1990s more than half of the unuts' savings were deposited in the branches, a trend that accelerated during the crisis (chapter 15). For many years much of the units' excess liquidity was channeled into large corporate loans. And since large borrowers have a history of poor repayment at BRI (and at many other Indonesian banks), unit profits were used to cross-subsidize the losses of other BRI divisions. Marketing and public relations When the main business of the units was disbursing subsidized credit, there was little interest in public relations. Borrower targets were set and filled by gov- Institutional Development for Large-Scale Sustainable Microfinance The Transformation of the Unit Desas. 1984-96 319 ermnent commuttees; given the interest rate regulations of the time, the units did not encourage savers. Customers, it was thought, needed instructions, not public relations. The situation changed in 1984 with the advent of KUPEDES, and even more m 1986 when the new savings program began nationwide. First, KUPEDES had to be introduced to the public. Unsubsidized, and unlike the institutional loans with which most villagers were familiar, KUPEDES brought with it the need for umt staff to sell a product to potential borrowers. KUPEDES customers came to be called (and treated as) clients, not as beneficiaries-a milestone in the history of the unit desa system. Savings mobilization accelerated this tran- sition because, to capture savings, unit staff had to visit clients and convince them KUPEDES that the bank was trustworthy and its services helpful. Early publicity took several forms: posters, advertising in newspapers and on customers came to be the radio, and personal visits. Following the example of Bank Dagang Bali, the SIMPEDES lottery drawings were used to publicize unit products and services. called (and treated There were setbacks. At all levels in the unit system, BRI staff were un- familiar with how to advertise unit products. The head office hired a large as) clients, not as Jakarta advertising firm to publicize unit products in the villages; the results were somewhere between disastrous and ludicrous.The posters first showed beneficiaries-a urban youth in the latest fashions waxing ecstatic over KUPEDES and SIM- PEDES.When these were rejected, the next set of posters featured pictures milestone in the of the agency's perception of farmers-poor, bent, and clad in ragged clothes. No one from the advertising agency thought of visiting a village or looking history of the unit for messages that might appeal to real villagers.The bank terminated its re- lationship with the advertising firm and looked for another. It soon became desa system apparent that at that time no Jakarta agency would know how to advertise unit desa products to rural households. The head office then took over the job of advertising unit products; after an initial learning period, BRI did very well at this. Similar problems arose at all levels in the system.When selecting prizes for the SIMPEDES lotteries, branch managers had little idea what local people would consider attractive prizes. Farming tools and sewing machines were the first at- tempt. But SIMPEDES savers said they wanted consumer goods as prizes- motorcycles, televisions, and refrigerators.The prizes were changed accordingly and were later increased so that cars could be included as well. Branch staff also learned how to adapt to their local environments the advertising materials gen- erated by the head office. As noted in chapter 13, however, the best advertising remained word of mouth from satisfied clients. As the new products were introduced, the staff trained, and the incentives implemented, umt service improved dramatically. So did word- of-mouth advertising. By the late 1980s, when the units were beginning to face competition from other banks, they had become public relations experts.The units had developed advertising messages that appealed to their client base and could be spread in multiple ways: through BRI calendars (finally given to savers as well as bor- 320 The Microfinance Revolution: Lessons from Indonesia rowers), t-shirts, posters, decals, newspaper advertisements, and television dra- mas showing the role of unit products in family sagas.9 The development of public relations provides another example of how the units were driven by the nature of the new instruments.These were products that had to be sold; by trial and error, the units learned marketing. Learning from mistakes While moving from the old culture to the new, many mistakes were made. But BRI usually learned from these mistakes. As the unit desas learned from the problems that arose in relocating units, in advertising, and in other areas, some underlying issues began to emerge. One example was the critical role of se- quencing. Asking branch managers to distinguish potential from nonpotential While movingfrom units before they were trained in identifying potential was a classic example. Another example illustrates a related problem.At the time of the transition, the old culture to branch managers were evaluated annually on a set of criteria that placed con- siderable weight on low unit arrears but relatively little weight on the value of the new, many outstanding loans; the criteria did not include an increase in the number of bor- rowers.Thus managers had an incentive to give few loans.After KUPEDES was mistakes were introduced, the evaluation criteria continued unchanged for some time. Not surprisingly, many branch managers instructed unit staff to go slow on KU- made. But BRI PEDES. One branch manager refused to allow any KUPEDES loans because, he said, the surest way to have no arrears is to have no loans! When these dif- usually learned ficulties were uncovered, the evaluation criteria were changed, and the num- ber of KUPEDES borrowers increased. from these mistakes Effective branch management of the new microfinance turned out to be far more difficult than had been anticipated. Some branch managers learned quick- ly and took great interest in the new system, but at first these were in the nunor- ity. Supervising officers at branch and regional offices tended to find it more difficult to accept the far-reaching changes being made in the unit desa system than did the unit staff.The middle managers had grown up under BIMAS and had been promoted because they were experts at implementing the old BRI culture. Some were able to change quickly, but many were unable or unwilling to do so. By contrast, many unit staff serving under these managers were eager to implement the new products, both KUPEDES and savings.They were younger, less set in the old ways, and closer to the unit customers. They were also mo- tivated by the changes in their salaries, benefits, incentives, and training. This situation led to difficulties; in some cases units tried to implement head office instructions but were blocked by their branch managers. The situation was partly alleviated by frequent visits to the branches and units by represen- tatives of the head office. In particular, the president-director and the manag- ing director responsible for the units traveled constantly, visiting branches and units, often unexpectedly.Their visits had three purposes: to check whether the new instruments and procedures were being appropriately implemented; to teach the new BRI culture at the regions, branches, and units; and to learn from local staff-a hallmark of the changed culture at BRI. Unit staff were encouraged Institutional Development for Large-Scale Sustainable Microfinance The Transformation of the Unit Desas, 1984-96 321 to discuss their ideas with the president-director and the managing director in front of their branch managers, hitherto an unimaginable event. Over time, and with training, incentives, and encouragement, many nud- dle managers changed their attitudes (to varying degrees).With others the prob- lem was solved only by their eventual retirement. Organization and Management of the Unit Desas, 1996 Fully owned by the government of Indonesia, BRI is managed by a president- director and six managing directors appointed by the Ministry of Finance; the In some cases units appointment of the president-director must be approved by the president of Indonesia."' The managing directors are operating officers of the bank; each tried to implement was responsible for at least 3 of BRI's 23 operating divisions (figure 14.3). A three-person Board of Commissioners appointed by the Ministry of Finance head office holds general supervisory authority on behalf of the ministry but does not play an active role in direct bank management. instructions but were At the end of 1996 BRI's assets totaled $14.4 billion, outstanding loans were $11.2 billion, and deposits were $8.1 billion.There were more than 45,000 em- blocked by their ployees. In addition to its divisional structure, BRI was organized into strate- gic business units. In 1996 there were three such units: corporate and international branch managers banking, commercial and retail banking, and the unit microbanking system. BRI maintains offices at all the government's admimstrative levels from its head office inJakarta to its unit posts in villages (figure 14.4).The Business Unit Desa (BUD) Division was one of BRI's 23 divisions; only this division and its managing director were directly responsible for the unit desa system (see fig- ure 14.3). However, other divisions in the bank-such as Systems and Tech- nology, Logistics, Human Resources Development, Training, and Legal Affairs-provided services to the BUD Division. At the end of 1996 the BUD Division had assets of $3.4 billion (24 per- cent of BRI's assets). KUPEDES had $1.7 billion in outstanding loans (15 per- cent of BRI's loans).And unit deposits were $3.0 billion (42 percent of BRI's deposits).The BUD Division was responsible for the large majority of BRI's clients-and all its profits. Lending and deposit taking are the units' main activities. However, other services are profitably provided on a fee basis.Thus all units handle transfer pay- ments for both clients and noncients. Many units act as payment points for tele- phone and electric bills and for the payment of property taxes. Some act as paymasters for government teachers and army staff and for pension plan pay- ments. Table 14.1 shows BUD employment at the different BRI organizational levels illustrated in figure 14.3. Although it was only 1 of BRI's 23 divisions, the BUD Division employed 23,115 staff in 1996-just over half the bank's total employment-indicating just how labor-intensive nmicrofinance is com- pared with conventional commercial banking.The proportion of female em- 322 The Microfinance Revolution: Lessons from Indonesia Figure JAL3 Bank Rakyat Indonesia's organizational structure, 1996 Board of Commisioners President- director - = - Human Systemdd D s =- - - Deeomn _ XCdnatiol _~~~~~~ a Source BRI data ployees in the unit system, which had been rising gradually, approached 25 per- cent by the end of 1996. The following sections provide a detailed discussion of BUD organization and management, reporting, staff incentives, and training in 1996.The purpose is to illustrate their relationships to umt desa performance. Beginning in 1997 further bank reorganization took place; the changes that affected microbank- ing are discussed in chapter 15. At the head office At BRI's head office the managing director responsible for the BUD Divi- sion is charged with overseeing the unit desa system; he also plays the pre- dominant role in setting overall policy for the units. However, major policies are decided by BRI's board of directors (its six managing directors and its president-director). Institutional Development for Large-Scale Sustainable Microfinance The Transformation of the Unit Desas, 1984-96 323 Figure 1M Government administrative levels and Bank Rakyat Indonesia's organizational levels, 1996 Government administrative level BRI organizational level National Head office Provincial Regional offices (15) Internal audit offices (15) District Branch offices (325) Subdistrict Unit desas (3,595) Village Service posts (392) Source BRI data At the head office the BUD Division is responsible for the development, man- agement, supervision, and reporting of the unit desa system. Consisting of 69 peo- ple in 1996, it was headed by a general manager (kepala urusan) and two deputy managers; the general manager reports to the responsible managmg director (figure 14.5).The BUD general manager also coordmates unit desa activities with divisions (such as Logistics and Human Resources Development) that provide support to BUD. 324 The Microfinance Revolution: Lessons from Indonesia Table 14.1 Employment in the unit desa system by organizational level, 1996 Organizational level Number of offices Number of employees Head office 1 69 Regional offices 15 196 Branches 325 1,612 Units 3,595 21,238 Unit staff 17,156 Guards and contract staff 4,082 Village posts 392 0 Total 4,327 23,115 Source BRI 1997b Figure 145 Organizational structure of the Business Unit Desa Division at Bank Rakyat Indonesia's head office, 1996 Managing director 1 General manager Deputy Deputy general general manager manager i M- it r 1 M researc and ~ PON.,WtL ma,o tq,,,anaqgernen irprtn ystm Source BRI data BUD's four sections. The BUD Division was organized into four sections (bagian).The Product Management Section, with a staff of 10 in 1996, was re- sponsible for developing new products, marketing and promoting those unit products, and analyzing product profitability; this section also provided train- ing in these areas. The Operations and Reporting Section, with a staff of 18, was charged with developing BUD operating standards, producing operations manuals, and monitoring the use of the standards that had been set. This sec- Institutional Development for Large-Scale Sustainable Microfinance The Transformation of the Unit Desas, 1984-96 325 tion also analyzed the data reported from the units, produced regular reports on unit development, and generated reports for management on special top- ics affecting the units. Among its other responsibilities were monitoring and reporting on unit staffing and supervision. The Unit Research and Development Section, with a staff of 10, con- ducted research on products, markets, customers, and on unit staffing, op- erations, and organization.This section also evaluated unit development and provided input to BUD on its criteria for the location of units.The Man- agement Information Systems Section had a staff of 23, of whom 10 were appointed to the computer workshop. This section was responsible for evaluating and recommending equipment; installing, testing, and repairing The BUD hardware; and developing, installing, maintaining, and improving software. It was also responsible for research and development for BUD's computer (microbanking) network. BUD also maintains an InternationalVisitors Program; in 1996 it had a staff Division was of 5, plus a part-time director who reported to the BUD general manager." Program staff accompany BUD's many international visitors on field visits, pre- responsiblefor the pare instructional materials for the program, give presentations on various as- pects of unit activity, and conduct workshops and conferences on commercial large majority of mncrofinance. During 1991-96, BUD hosted nearly 500 visitors from more than 30 countries in Asia,Africa, Latin America, Europe, the Middle East, and North BRI's clients-and America. all its profits Management issues. As a large commercial bank, BRI follows a generalist management model that results in frequent personnel changes among the bank's divisions.Among divisional general managers and section heads, for ex- ample, the average appointment lasts three to four years.While this method en- ables managers to gain broad banking experience, it limits continuity and institutional memory within divisions. This management model presents particularly difficult problems for the BU1D Division since the units' clients, products, services, procedures, pricing, and staffing differ significantly from those of the rest of the bank. For BUD, as for any com- mercial microfinance institution, it is important to develop a group of long- term managers who are experienced in and committed to the profitable operation of a commercial mucrofinance system. This was not possible under BRI's management model. With the exception of the late Sugianto, the long-term incumbent of the managing director 1 position shown in figure 14.3, there has been little con- tinuity at BUD.As soon as BUD managers gained several years of microfinance experience, they were transferred to other divisions and replaced by microfi- nance neophytes.A related problem is that many of BRI's muddle managers did not consider BUD a desirable posting. The idea that "real" bankers are in cor- porate and international banking was widespread. Despite the fact that BRI's profits came from BUD, microbanking was generally considered a low-status activity within BRI-and by the banking sector in Indonesia generally.The BRI 326 The Microfinance Revolution: Lessons from Indonesia management model illustrates one of the problems that can arise when com- mercial banks operate microfinance through a division; other difficulties that can result from this structure will be discussed later in this chapter. At the regional offices BRI's 15 regional offices are located in provincial capitals. Their role is ad- ministrative and supervisory; regional offices do not make loans or accept de- posits. They coordinate and supervise the branches in their regions. They collect and analyze data about the unit desas of the region and report this in- formation to the head office.And they communicate messages and instructions from the head office, helping the branches to understand and implement these. In addition, staff recruitment for the units is centered at the regional offices. Although it was As with many of BRI's other operating divisions, BUD is represented at each regional office through a BUD Department, divided into two groups:Ad- only 1 of BRI's 23 ministration and Reporting, and Automation. In addition, some BUD De- partment staff work directly with the BUD department manager. Other BRI divisions, BUD divisions represented at the regional offices-Logistics, Human Resources De- velopment, and Systems and Technology-are also involved with BUD activ- employed over haWf ities; the regional BUD Department is not responsible for most of the support functions used by the unit desa system at the regional level. BUD's function in the banks staff this regard is to coordinate with the relevant departments at the regional of- fices to ensure that such support is adequately provided. The BUD Department in each regional office has overall responsibility for the units in its region. Department staff analyze unit performance, profitabili- ty, and portfolio quality; help estimate employment and recruitment needs; col- lect information in fraud cases; install and maintain computers; and generally supervise and promote unit development throughout the region.The number of BUD staff at a regional office depends on the number of units in the re- gion. In 1996 there were an average of 13 BUD staff per regional office, of whom 7 were responsible for computers and automation. During 1993-96 total su- pervision costs at all levels for the unit desa system averaged 1.2 percent of loans outstanding or 17.8 percent of unit staff costs. BRI management considered this substantial expenditure on supervision essential for the strong perfor- mance of the units. At the branches BRI's 325 branches, located in district capitals, are responsible for the bank's commercial and retail banking activities; some branches are also engaged in cor- porate banking. In addition, the branches supervise the unit desas that operate in their service areas. The average branch supervises about 10 units, but there is a wide range from a few units in densely populated small districts to more than 35 units in large, more sparsely populated districts. The branch manager is responsible for all the banking activities of the branch and its units. Unit business managers, most of whom are former unit desa man- Institutional Development for Large-Scale Sustainable Microfinance The Transformation of the Unit Desas, 1984-96 327 agers, are based at the branch and carry out the daily supervision and moni- toring of the units. Each unit business manager is supposed to supervise a max- imum of four units, although in practice these managers are sometimes responsible for additional units because of understaffing at the branch level. Unit business managers play a key role in the supervision process, work- ing closely with unit desa managers, monitoring all aspects of each unit's busi- ness, conducting a running audit, and educating unit staff about new policies and instructions."2 Responsibilities also include visiting borrowers and savers and cross-checking the units' loan and deposit records with clients. Unit busi- ness managers are required to talk with all borrowers whose loans are in ar- rears; this serves the dual purpose of assisting in loan collection and identifying As soon as BUD cases of collusion between unit staff and borrowers. A unit development officer directs and supervises unit business managers managers gained in branches with nine or more umts; branches with many units may have an additional unit development officer.The unit development officer, who reports several years of directly to the branch manager, also participates in unit supervision and in the ongoing evaluation of unit performance.The number of branch staff carrying microfinance out unit-related work depends on the number of units the branch supervises; in 1996 the average was about six such employees per branch."3 experience, they At the unit desas were transferred to Most unit desas are located at subdistrict capitals and report to supervising branch- es at district headquarters.The decentralized nature of BUD can be seen from other BRI divisions table 14.1:in 1996 only 1.1 percent of the division's 23,115 staff members were located above the branch level. The physical plants of the units have changed significantly since the 1970s. At that time many units were located in run-down, dilapidated buildings; the units themselves were often dark and dirty, with peeling walls and broken fur- niture. Many units did not even have hand calculators; calculations were made on paper or on an abacus. Today the units are still small, but typically they are light, clean, and attractive. The manager has a small private office that is often used for discussions and trans- actions with clients who request privacy.The use of space at the units is careful- ly designed for efficiency and for the convenience of customers and staff. Color posters showing unit clients using unit products line the walls.The staff wear at- tractive uniforms.An increasing number of the buildings that house unit desas are owned by BRI, but most are still rented. The atmosphere of today's units is one of professionalism and service in a small, familiar, comfortable place of business. In 1996 each unit had at least one stand-alone, multiuser, personal com- puter-based system. Urban units were moving toward a system in which the units within a branch are networked with one another. Although the software systems of units and branches are not directly compatible, BUD was develop- ing plans to network units with their supervising branch. Figure 14.6 shows the structure of a unit desa and its relationship to unit supervisors at the branch level. As noted, each unit has at least four staff mem- 328 The Microfinance Revolution: Lessons from Indonesia Filure 14.6 Organizational structure of a unit desa and its supervising branch, 1996 I The idea that "real" ~~~~nN ~~~~~~~~~Banh I ~ ar i Unit manager corporate and international Credit officer Bookeeper banking was widespread Village service posts Source BRI data bers: manager, credit officer, bookkeeper, and teller. The latter two are entry- level positions.The bookkeeper, who posts transactions and keeps the unit's books, also helps customers fil out forms and perform transactions. The teler is re- sponsible for all cash transactions. Cash is kept in a safe at the unit; excess cash above the specified overnight lrimit-which varies by unit-is moved to the branch by BRI staff. Credit officers, most of whom have worked as unit desa bookkeepers and tellers, are expected to spend most of their time visiting customers. They in- form potential borrowers about KUPEDES loans, assess the creditworthiness of loan applicants, and colle,ct overdue loan payments. They also visit present and potential savers, and they serve as general representatives of the unit in the communuity. When appointed, unit managers have typicaly had 7-10 years of service as bookkeeper, teler, and credit officer. The unit manager, who signs officially on behalf of the unit, is responsible for the decisions about loans that fall with- in his or her loan authority, for the unit's savings, for its admionstration and se- curity, and for its profitability. Unit managers split their time between the Institutional Development for Large-Scale Sustainable Microfinance The Transformation of the Unit Desas, 1984-96 329 office (where they supervise staff, meet with customers, prepare reports, and work closely with the unit business manager from the branch) and the field (where they visit cients,join the credit officer for some field appraisals, and fol- low up on overdue loans).The unit manager plays a key role in promoting unit business by developing and maintaining contacts within the community. Staff are added based on a unit's activity. For example, when a unit reach- es 400 borrowers, BRI regulations permit a second credit officer. Similarly, when a unit's daily cash transactions average 200 for a six-month period, an additional teller can be added.When the average daily number of transactions posted, ex- cluding posting of interest, reaches 150, another bookkeeper can be added. In practice, however, these arrangements do not always work so precisely, and some During 1993-96 understaffing is common. This occurs for two main reasons. First, recruiting and authorization may be delayed because these involve bank total supervision divisions outside BUD, and the labor-intensive nature of the units' activities is often not fully understood elsewhere in the bank. Second, a unit manager and costs at all levelsfor the unit's supervisors at the branch, fearing an immediate decrease in unit prof- its (and in their incentive payments), may wait until there are perhaps 600 or the unit desa system more borrowers before requesting an additional credit officer.When a unit reach- es 11 staff members by adding staff under this process, it must split into two averaged 1.2 percent units in order to move its services closer to customers' homes and places of work. The staff of each unit produce a monthly balance sheet, profit and loss state- of loans outstanding ment, credit report (which details loan arrears over time), achievement of tar- gets report, and unit development report.The unit development report, which uses 27 indicators to assess the condition of the unit, provides the basic infor- mation used by unit and branch managers to review the performance of their units.These reports, aggregated monthly at the branch, regional, and head of- fice levels, serve as the main indicator of unit desa performance over time. In addition, the head office conducts periodic sample surveys to obtain informa- tion that is not collected on a regular basis, but that is useful for decisionmak- ing. Examples of such studies are those on trends in loan size distribution and on delinquency by length of loan maturity. Unit staff are normally local high school graduates. Because jobs at a unit are well paid by local standards, especially in rural areas, BRI can usually select unit staff from a large pool of applicants. BRI has found that, for unit desas, it is more efficient to invest in comprehensive in-house training for recruits who are high school graduates (and who often have some additional traimng in bank- ing, accounting, or computers) than it is to hire university graduates; howev- er, some unit staff hold university degrees. In general, experience has shown that while university graduates demand higher salaries, much of their previous training is irrelevant, their expectations tend to be higher than can be met at an entry-level unit position, they typically do not want to live near the units, and they often find it difficult to treat poor clients with the respect and friend- liness that are the hallmark of the units. In sharp contrast to the staff regulations that prevailed during 1970-83, there was now a flow of staff between units and branches. A career path was estab- 330 The Microfinance Revolution: Lessons from Indonesia lished for unit staff who perform well: unit managers can be promoted to serve at the branch as the unmt business manager or unit development officer; they can then be promoted to other positions within the bank. In addition, lower- level branch staff can be promoted to serve in the units. However, given that in 1996 there were 3,595 unit managers, about 1,000 unit business managers, and only about 250 unit development officers, promotion beyond unit manager is highly competitive. In practice, promotion for umt staff beyond the level of unit development officer is very limited. In 1996 only 4 percent of the BUD staff at the head office (3 people) had had experience work- ing at the units. In addition to the unit banking staff, each unit employs a mght guard. Since rural Indonesia generally has a low crime rate, most units do not have guards Staff are added during the day. based on a unit' At the village posts Some units operate village service posts (pos pelayanan desa, or PPDs); in 1996 activity. Thus when there were 392 such posts. These posts are established when there is substan- tial demand for financial services at a location that is within the unit's service a unit reaches 400 area but too far from the unit for clients to receive convenient service. Service posts, which do not have their own staff, are typically served by two unit staff borrowers, members: a teller and a bookkeeper. Depending on demand, the posts are open from one to five days a week; they accept loan payments and deposits and pro- regulations permit a vide information about unit products. Units with a post open on most days have two tellers and bookkeepers on their staff. Loans are not made at the posts, but second credit officer credit applications are received there on behalf of the unit. These applications are then reviewed and processed at the unit, where decisions are made and over- due payments managed in the same way as with all unit loans. Each post is part of a specific unit, and its performance is aggregated with that of its unit in the reporting system. The village posts are a good example of the unit desa system's creed of stan- dardization with flexibility. All posts carry out the same activities, but their days of operation depend on client demand. Moreover, a post with a large and grow- ing volume of business may be upgraded to a unit, while a unit with insuffi- cient business may become a post. The unit desa system and other bank divisions Because BUD is a division of a larger bank, it receives services from other di- visions that are not reflected in BUD staffing.These services include liquidity management and portfolio investment for the deposits the units keep at the branches; some personnel, logistics, and automation functions; legal assistance; internal audit; and prudential reporting to the government. The branches also provide unofficial services to the units. The most im- portant is to shield the units from administering the government's subsidized credit programs (which are provided through the branches). In addition, the branches tend to act as a safety valve for unit staffing problems. Established BRI Institutional Development for Large-Scale Sustainable Microfinance The Transformation of the Unit Desas, 1984-96 331 staff can be fired only for proven fraud; this regulation creates a serious prob- lem for the units because they operate with only a few staff members, and their staff cannot be removed for poor performance or low productivity. The prob- lem is often solved by "parking the bad unit staff at the branch," as it is some- times called at BRI.A staff member thus parked carries out minor clerical tasks at the branch while another staff member is appointed to the unit. The relationship between the units and their supervising branches is com- plex, with many ambiguities.The units offer only microfinance, a different busi- ness from the retail and commercial banking offered at the branches; the branches supervise commercial microfinance but do not provide such services themselves. But the branches administer government- and donor-subsidized credit programs. The tensions and Unit supervision generally account for a fairly small portion of branch activity and involve relatively few branch staff. But in 1996 the units were very prof- complementarities itable, while the branches were loss-making. Given these circumstances, branch managers were often uncertain about how to order their priorities. between commercial The problems caused by the provision of incentives to unit staff but not to their branch supervisors were addressed in 1996.A revised Unit Achievement microfinance and Competition program provided semiannual cash bonuses for unit business managers, unit development officers, and branch managers, up to 1.6 times their conventional monthly salaries (see chapter 13 and below).The bonus is based on the per- formance of the units supervised by these branch staff.This change helped con- banking converge at siderably, but underlying structural problems remained at the branch level.The tensions and complementarities between commercial microfinance and con- the branch, where ventional banking converge at the branch, the only level at BRI where the two systems interact directly on a daily basis. Also converging at the branch level the two systems are the longstanding tensions between commercial and subsidized rural cred- it. The dynamucs of these overlapping but sometimes conflicting approaches, interact on a daily as they interact at BRI's head office, are discussed later in this chapter. basis Reporting Unit Desa Performance at the Unit, Branch, and Regional Levels As noted, the unit development report is the basic unit progress report gener- ated monthly at each unit; it is aggregated at the branch, regional, and head of- fice levels, also monthly. The first section of the report summarizes income, expenses, profits, portfolio quality, and trends in loan and savings volume. The remaining sections provide additional details on loans, savings, unit staffing, and the average daily activity levels of the unit.Time series for each of the indica- tors reported in the unit development report are available for use by unit man- agers and their supervisors.This reporting system at each level-from the unit desa to the head office and its use by unit supervisors at the branch and re- gional levels are crucial factors in the success of the system. Examples of unit development reports for two units located in different parts ofJava (called here units A and B), their supervising branches (branches A and 332 The Microfinance Revolution: Lessons from Indonesia B), and their regional offices (regional offices A and B) are provided below to show how the reporting system works and how it is used in the management of the unit desa system. Regional office A, one of BRI's largest regional offices located in one of Indonesia's largest cities, is responsible for 35 branches (including branch A) and 600 units. Branch A is responsible for 16 unit desas, including unit A. In 1996 the district in which branch A is located had a population of 1.1 mil- lion people. Regional office B, smaller and more typical in size (although still larger than average), oversees 21 branches and 345 units. In 1996 about 730,000 people lived in the district served by branch B. This branch super- vises 18 unit desas, including unit B. Unit desas A and B both serve relative- ly poor rural regions characterized by mixed agriculture, local industry, and Also converging at retail trade. Unit desa A is located in a district capital, while unit desa B is in a subdistrict capital. the branch level are The unit development reports of two unit desas the longstanding The December 1996 unit development reports of units A and B are provid- ed in table 14.2 as examples of the unit reporting system.These are simply two tensions between examples of reported unit performance, and these units cannot be considered representative of the system. However, the reports are useful as examples of commercial and the performance of individual units and as examples of what can be learned from these reports. Both units were understaffed in 1996. Unit A was autho- subsidized rural rized for six employees and unit B for seven. Both, however, had only five employees. credit The cumulative amounts of KUPEDES loans lent by the two umts were nearly identical at $3.2 million (in 1996 dollars).4 The value of their outstanding loan balances was also quite similar.Yet at the end of 1996 unit A had only 54 percent of the number of unit B's outstanding loans.Thus unit A's average out- standing loan balance ($684) was larger than that of unit B ($410). There were several other important differences between the two units. First, unit A had nearly $2 million in deposits, more than twice as much as unit B. Second, unit B had a far better repayment record than unit A.The cumulative amount written off by unit A ($49,098) was more than 10 times that of unit B ($4,565); as a result the manager of unit A had a substantially lower loan au- thority than the manager of unit B. In 1996 unit A's long-term loss ratio (2.33 percent), although not high, was more than four times that of unit B (0.50 per- cent) . Third, unit A's 1996 profits ($139,320) were 79 percent greater than those of umt B ($77,962) .The primary reason for unit A's higher profits was its greater income from interest on deposits at the branch.While both units supplement- ed their KUPEDES income with interest income from their deposits at the branch, unit A had more than four times the amount of funds deposited at its branch as did unit B. Fourth, unit B also provided other services (such as pay- ments of taxes and electricity bills) that unit A did not. Overall, unit B has a more typical rural profile-more and smaller loans, more service, more trans- Institutional Development for Large-Scale Sustainable Microfinance The Transformation of the Unit Desas, 1984-96 333 Table 14.2 Unit development reports of two unit desas, December 1996 (U.S. dollars unless otherwise noted) Indicator Unit desa A Unit desa B Summary indicators Total income 460,763 280,005 Total expense 321,443 202,051 Profit or loss 139,320 77,962 Value of total loans outstanding 508,183 561,416 Principal collected (percent) 99 51 99 41 Portfolio status (percent overdue)a 2.92 1 56 Value of total savings 1,952,581 866,858 Loans Cumulative amount lent (1996 dollars)b 3,244,300 3,241,125 Amount lent current month 52,875 47,230 Total arrears 14,687 8,785 Cumulative amount charged off 49,098 4,565 Long-term loss ratio (percent)a 2 33 0 50 Short-term loss ratio (percent)a 4 36 -1.64 Interest collected (percent) 99 57 99 90 Manager's loan authorityc 1,259 3,147 Number of loans outstanding 743 1,369 Deposits Current account (giro) 419 2,067 National savings (TABANAS) 54,973 65,216 BRI rural savings (SIMPEDES) 1,780,109 802,682 BRI urban savings (SIMASKOT) 57,910 0 Time deposits 59,169 16,471 Other deposits 0 0 Employees ___ _ Number of employees (actual) 5 5 Number of employees (authorized) 6 7 Daily activities Average number of daily cash transactions 200 260 Average number of daily service transactionsd 0 25 a See box 12 3 for definitions of loss ratios The short-term loss ratio measures changes in the past month It can be negative if, for example, substantial collection of old loans previously charged off occurred during the month being reported b Amounts lent were not available by year, so this figure is not adjusted for inflation c This row normally provides the amount of loan authority for loans to both new and old customers, the authority for loans to old borrowers being higher than the authority for loans to new borrowers However, only the authority for loans to new borrowers was reported in these unit development reports d For example, facilitation of payments for taxes and electricity bills, and disbursement of pension fund payments Source BRI data 334 The Microfinance Revolution: Lessons from Indonesia actions, and higher repayment-than unit A, whilch serves its district capital and is somewhat more urbanized than unit B. The aggregated unit development reports of two branches Table 14.3 shows the December 1995 and December 1996 unit development reports of branch A (the supervising branch of unit A) and branch B (the su- pervising branch of unit B). These two unit development reports aggregate the performance of the 16 units that report to branch A, and the 18 units that report to branch B.The unit system in both branches was profitable in both years, although the units of branch A saw profits fall from 1995 to 1996, while in branch B profits increased. In both branches all KUPEDES loans were funded from unit desa savings. Nine- The reporting ty-five percent of the value of savings in 1996 in both branches was in fully liquid savings accounts (SIMPEDES, SIMASKOT, and TABANAS). system at each level The aggregated unit development report of Branch A for both 1995 and 1996 alerted the branch manager, the unit business managers, the unit devel- and its use by unit opment officer, and the BUD section at the regional office that the units of this branch were incurring high arrears.As soon as they receive a monthly umt supervisors are development report that shows increasing arrears or continuing high arrears, the unit business managers and the unut development officer would be expected crucialfactors in the to work with the umt managers and credit officers of the units where the prob- lems were located. Branch A's improved short-term loss ratio for December 1996 success of the system suggests that such an effort was made. The branch and regional managers should also have noted from branch A's unit development reports that the units of branch A are understaffed, that un- derstaffing was a growing problem, and that this may have been contributing to the repayment problem.The reports signal that the managers should check whether the understaffed umts are the ones with repayment problems-as turned out to be the case at umt A. The aggregated unit development report of branch B shows an excellent record. Branch B units show increasing profits, low arrears, increasing value of loans out- standing, a growing number of loans outstanding, and mcreased savings. Unit performance in two regional offices The performance of the unit desas in the two regional offices is shown in table 14.4. Both regions are largely rural but also contain urban units. Regional of- fice A is responsible for branch A and unit A, while regional office B is responsible for branch B and unit B. With long-term loss ratios of 1-2 percent, outstanding loans and savings in the hundreds of mullions of dollars, and more than 2 million clients in re- gional office A and more than 1 million in regional office B, the unit system has been successful in both regions. In 1996 both regional offices had about six times as many savings accounts as loans, and all loans were financed by sav- ings. In 1996 unit profits were $28.7 million at regional office A and $14.2 mil- lion at regional office B. Institutional Development for Large-Scale Sustainable Microfinance The Transformation of the Unit Desas, 1984-96 335 Table 14.3 Aggregated unit development reports of two branches, 1995-96 (U.S. dollars unless otherwise noted) Branch A Branch B Indicator 1995 1996 1995 1996 Summary indicators Total income 3,214,471 3,617,771 2,615,090 3,155,377 Total expense 2,429,809 2,889,215 2,023,710 2,461,854 Profit or loss 785,095 725,556 591,380 693,523 Value of total loans outstanding 6,740,035 8,284,096 5,654,274 7,026,152 Principal collected (percent) 96 24 9719 9915 99 30 Portfolio status (percent overdue)a 5 86 5 95 2 39 2 80 Value of total savings 9,616,551 10,815,359 8,279,064 8,932,376 Loans Cumulative amount lent (1996 dollars)b 45,619,151 55,760,386 32,629,895 48,574,036 Amount lent current month 854,863 1,063,365 636,666 727,024 Total arrears 395,147 492,666 135,155 196,638 Cumulative amount charged off 1,396,013 1,431,809 94,686 95,462 Long-term loss ratio (percent)a 4 56 4 01 0 84 0 70 Short-term loss ratio (percent)a 6 67 2 83 0.69 1 89 Interest collected (percent) 97 52 98 18 99 37 99 87 Manager's loan authorityc - - - - Number of loans outstanding 10,905 12,204 15,431 16,684 Deposits Current account (giro) 81,889 92,740 142,125 46,069 National savings (TABANAS) 826,256 992,866 811,525 909,534 BRI rural savings (SIMPEDES) 8,151,213 9,103,651 6,890,642 7,431,685 BRI urban savings (SIMASKOT) 99,220 91,062 69,165 88,214 Time deposits 457,972 528,327 365,537 456,752 Other deposits 6,499 6,294 0 0 Employees Number of employees (actual) 91 91 82 85 Number of employees (authorized) 94 96 82 87 Daily activities Average number of daily cash transactions 1,705 1,729 2,431 2,314 Average number of daily service transactionsd 440 545 50 114 Memorandum item Number of unit desas 16 18 a See box 12 3 for definitions of loss ratios b Amounts lent were not available by year, so this figure is not adjusted for inflation c Varies by unit d For example, facilitation of payments for taxes and electricity bills, and disbursement of pension fund payments Source BRI data 336 The Microfinance Revolution: Lessons from Indonesia Table 14A| Performance indicators for the unit desas of two regional offices, 1995-96 Regional office A Regional office B Indicator 1995 1996 1995 1996 Value of KUPEDES loans outstanding (millions of U S dollars) 238 2 288 3 1066 130.7 Number of KUPEDES loans outstanding 412,000 445,000 236,235 242,392 Value of savings (millions of US dollars) 415.6 4790 2187 2500 Number of savings accounts 2,430,000 2,646,000 1,450,401 1,527,717 Portfolio status (percent)a 4 08 4 06 2 87 3.06 Long-term loss ratio (percent)a 2 15 2 03 1 20 1 18 Twelve-month loss ratio (percent)a 1 13 1 47 071 1 10 Short-term loss ratio (percent)" 1 31 2.00 0.09 - 144a Profit (millions of U S dollars) 30 7 28 7 14 1a 14 2 Memorandum items Number of branches 35 21 Number of unit desas 600 345 a See box 12 3 for definitions of loss ratios See note a in table 14 2 for explanation of negative short-term loss ratios Source BRI data The aggregate unit business for which the regional offices are responsible is of considerable magnitude. Efficient operation of the units on this scale de- pends heavily on the reporting system and on its effective use by branch and regional supervisors. Staff Incentives and Training Successfiul management of a large microfinance system depends on well-trained and highly motivated staff. The BUD Division implements carefully planned staff incentives and carries out specially designed nucrofinance training pro- grams for all unit staff and their supervisors. These investments in human re- sources-which contribute significantly to the relatively high overhead costs of the BUD Division-are crucial to the unit desa system's scale of outreach, high repayment rate, high savings, and continued profitability. Incentives Beginning in 1984, 10 percent of each unit's annual profits was distributed to all the employees of that unit as an incentive bonus, up to 1.5 times an em- Institutional Development for Large-Scale Sustainable Microfinance The Transformation of the Unit Desas. 1984-96 337 ployee's monthly salary. This system continued for more than a decade; it was revised in 1996.At that time the percentage of unit profits available for the in- centive bonus was reduced to 6 percent, but the maximum bonuses were in- creased to 2.5 times the base monthly salary of the unit desa manager, 2.3 times the salary of the credit officer, and twice that of the other unit staff.The new system was implemented in 1997. As noted, a second incentive system, the Unit Achievement Competition, was developed in 1989 to encourage staff to move to the stage of market pen- etration in savings mobilization (chapter 13). By 1989 about 80 percent of the units were profitable; most staff had reached their maximum incentive level and had come to expect the profitability bonus as a matter of course.Thus the semn- In 1996 both annual Unit Achievement Competition was added because the unit prof- itability bonus no longer provided a strong incentive for many umt staff. regional offices had Unlike the profitability bonus, which rewards only profitability, the Unit Achievement Competition scores each unit on a range of weighted criteria. about six times as The weights, set by the head office, are changed according to BUD's needs.Thus when the competition was introduced, funds mobilization was the units' high- many savings est priority; accordingly, the weighting emphasized increases in savings. But dur- ing the tight money period of the early 1990s, when the loan portfolio was accounts as loans, declimng, portfolio quality was also given a high weight in the competition. In 1996 the criteria used in the competition, and their weightings, were prof- and all loans were itability (30 percent), portfolio quality (20 percent), management effectiveness (20 percent), increase in the number of borrowers (10 percent), increase in the financed by savings value of outstanding loans (10 percent), and increase in the value of savings (10 percent). All staff members of a unit that meets its goals are given cash awards as well as certificates of appreciation presented in a formal ceremony. Typically between one-third and one-half of the units meet the criteria to win the semi- annual competition. In each competition the staff of the units that are judged to be best in the region and best in the country are given additional rewards, which have included visits to observe rural banking systems in neighboring countries. In 1996 awards for a successful unit manager under the semiannual Unit Achievement Competition reached about $130, representing 30-45 percent of the manager's monthly salary.Awards were also given to other unit staff in pro- portion to their salaries. The combmation of bonuses and awards provides a strong incentive for good performance. Including profitability incentive bonuses and competition awards, unit staff in 1996 could earn from three to six times the annual per capita GNP of $1,073.Thus an experienced unit manager with an annual salary of $4,800 (4.5 times the 1996 per capita GNP) who managed a unit that earned good profits and met the criteria for the competition could earn additional incen- tive pay of up to $1,260 a year.This would be composed of $1,000 from the profitability bonus and $260 from meeting the criteria in both semiannual com- petitions. Managers whose units were judged best in the region or nation could 338 The Microfinance Revolution: Lessons from Indonesia earn additional incentives.Another important incentive is that highly rated staff know that they have an increased probablity of promotion. As noted, problems had resulted earlier from providing incentives to unit staff but not to their branch-level supervisors.Therefore the profitability bonus and Unit Achievement Competition rules were changed in 1996 to include unit business managers, unit development officers, and branch managers. Start- ing in 1997 these staff became eligible for an annual bonus for unit profitabil- ity of up to 1.6 times their monthly salary. Training BRI operates five large training centers for unit desa staff and their supervi- sors. Located in Bandung in West Java,Yogyakarta in the central part ofjava, BUD provides Surabaya in East Java, Padang in West Sumatra, and Ujung Pandang in South Sulawesi, these centers are responsible for training the staff from all units as specially designed well as the unit business managers and unit development officers from all branches. staff incentives and The five traimng center managers report to the manager of BRI's Train- ing Division, who reports to managing director 5 (see figure 14.3). But BUD, microfinance training under managing director 1, bears the operating costs of the unit desa training centers as well as the expenses of the trainees; BUD also sets the content of the programsfor all unit curriculum. The centers, which operate throughout the year with about 80 full-time staff and their instructors, have a combined capacity at any one time of 780 trainees. Instructors, who are drawn from the unit desa system, provide courses for new staff, courses supervisors for staff being promoted to new positions, and refresher courses for all staff at least every 18 months.The refresher courses have two purposes: to provide a continuing opportunity for long-term staff traimng, and to update employees on new policies and operations. Between 13,000 and 18,000 unit and super- visory staff are trained each year. Thus between 50 and 75 percent of the staff concerned with operating and supervising units receive some form of train- ing each year. Considerable thought and care have been put into selecting instructors, de- signing the curriculum, and developing the facilities available and teaching meth- ods used at the centers. Instructors are selected in a competitive process: they must have unit desa experience, they must be good teachers, and they must be able to carry out multiple teaching methods, including lecturing, leading dis- cussions, role playing, brainstorming with trainees, observing trainees practic- ing in the center's dummy unit desa, supervising field practice, and other techniques that emphasize creative teaching and encourage input from trainees. Guest lecturers visit the centers to provide training in special topics such as law, business, and marketing. The training centers have excellent facilities. In addition to classrooms, of- fices, discussion rooms, sports facilities, dormitories for trainees, housing for the instructors and the manager of the center, and a guest house, each training cen- ter is equipped with a computer room, a well-stocked library containing ex- Institutional Deve opment for Large-Scale Sustainable Microfinance The Transformation of the Unit Desas, 1984-96 339 Table 14 51Curriculum of training centers for unit desa staff, 1996 (number of required 90-minute sessions) Bookkeeper/teller Credit officer Unit manager Recruit- Recruit- Promo- Recruit- Promo- Subjectrinstruction ment Refresher ment tion Refresher ment tion Refresher Bookkeeping 195 10 11 5 35 0 195 11 0 Logistics 4 5 0 4 5 0 0 4 5 0 0 Computer applications 11 10 11 0 0 16 11 0 Human resources 11 0 11 0 0 17 0 0 KUPEDES regulations and procedures 24 5 185 11 5 0 22 5 165 0 Credit investigation methods 0 0 7 5 7.5 7 7 5 7 5 7 Customer relations and supervision 0 0 5 5 5 5 5 5 5 5 5 0 Assessment of customers' businesses 0 0 5 5 5 5 7 7 5 7 5 7 Loan decisions 0 0 0 0 0 5.5 5 5 5 Funding and bank services 16 5 16 6 0 21 5 11 5 0 Civil law 9 0 11 55 5 13 7 5 Analysis of unit performance 0 0 0 0 0 11 14 0 Planning and budgeting 0 0 0 0 0 12 12 0 Planning KUPEDES development 0 0 0 0 0 8 7 5 7 Interview techniques 7 5 4 5 0 0 4 5 0 0 Loan collection methods 0 0 12 12 11 12 11 5 12 Personal selling 0 0 5.5 5 5 5 5 5 5 5 0 Marketing methods 0 0 55 55 0 5 5 5 5 0 Marketing strategy 0 0 0 0 0 6.5 5 5 5 Preparation of area monographs 0 0 4.5 4 5 4 4 5 4.5 4 Leadership 0 0 0 0 0 4 5 4 5 0 Presentation techniques 0 0 0 0 0 0 Supervision 6 2 7 5.5 0 16 11 8 Lectures 2 2 2 2 2 2 2 2 Opening and closing ceremonies 2 0 2 2 2 2 2 2 Total classroom sessions and examinations 112 39 145 82 48 234 168 5 64 Field trips and discussions 0 0 10 0 0 10 0 0 Total sessions 112 39 155 82 48 244 168 5 64 Source BRI data tensive training materials developed specifically for microfinance operations at the umts, and a dummy unit desa. The curriculum of the training centers at the end of 1996 is shown in tables 14.5 and 14.6. All 23 courses in the curriculum are specific to microfinance. Even such standard courses as bookkeeping and computer applications are taught differ- ently at these centers than they are at BRI's training programs for other bank staff.The material for many of the unit desa courses-such as credit investiga- tion, loan collection methods, and knowledge of customers' businesses-have 340 The Microfinance Revolution: Lessons from Indonesia Curriculum of training centers for Table 14.6 unit desa branch supervisors, 1996 (number of required 90-minute sessions) Unit business manager Unit desa officer Subjectrinstruction Recruitment Promotion Refresher Recruitment Promotion Refresher Bookkeeping 195 55 0 19 6 0 Logistics 4 5 0 0 4 5 0 0 Computer applications 11 10 5 0 1 1 11 Human resources 17 6 5 0 17 5.5 0 KUPEDES regulations and procedures 23 10 5 0 23 8 0 Credit investigation methods 7 5 7 5 0 7 5 5 5 0 Customer relations and supervision 5 5 5 5 0 5 5 5 5 0 Assessment of customers' businesses 7 5 7 5 0 7 5 8 0 Loan decisions 5.5 5.5 0 5 5 5 5 0 Funding and bank services 21 5 5 5 0 21 5 5.5 0 Civillaw 19 5 5 0 19 0 0 Analysis of unit performance 14 14 13 14 14 13 Planning and budgeting 12 12 0 12 0 0 Planning KUPEDES development 8 7 5 7 7 5 8 7 Interview techniques 4 5 0 0 4.5 0 0 Loan collection methods 12 12 5 12 11 5 11 0 Personal selling 5 5 0 0 5 5 0 0 Marketing methods 5 5 0 0 5 5 0 0 Marketing strategy 6 5 6 5 6 7 7 6 Preparation of area monographs 0 0 0 4 5 4 5 0 Leadership 4 5 4 5 0 9 9 0 Presentation techniques 8 8 0 8 8 7 Supervision 23 16 8 22 16 8 Lectures 2 2 2 2 2 2 Opening and closing ceremonies 2 2 2 2 2 2 Total classroom sessions and examinations 249 155 50 256 142 45 Field trips and discussions 15 0 0 15 0 0 Total sessions 264 155 50 271 142 45 Source BRI data been drawn directly from unit experience since 1984.Training for credit offi- cers and unit desa managers includes sessions in the field as well as in the classroom. Credit officers are taught to prepare simple balance sheets and income state- ments for micro and small businesses of many kinds in agriculture, trade, in- dustry, and services. In the sessions on credit investigation the trainees are asked to discuss and compare their experiences and to suggest methods suit- able for general use.Thus the class may learn from one of its participants about Institutional Development for Large-Scale Sustainable Microfinance The Transformation of the Unit Desas, 1984-96 341 assessing a tilemaking business, from another about shrimp farming, from a third about shoemaking. General issues are also discussed. For example, how can the inventory of a small shop be investigated efficiently so that each item need not be counted? One method taught is to ask the shopowner the price for which he or she would sell all the stock presently on hand. Both the curriculum and the teaching methods follow the BUD strategy of standardization with flexibility.While all five training centers provide the same courses and use the same training materials, the courses change according to umt desa needs. 15 Instructors and BUD staff provide regular input into the de- sign of new courses, which can cover any topic considered important to BUD performance. For example, by late 1996 it was thought that competition from Including other banks entering the microfinance market had accelerated, and that many unit managers and credit officers did not have sufficient self-confidence to man- profitability age the more competitive situation. As a result a new course on building self- confidence was designed and introduced into the training curriculum in 1997. incentive bonuses International visitors to the training centers often ask whether the traners are the same ones that train the rest ofBRI's staff, and whether the trainers come from and competition universities or banking institutes.They are usually quite surprised to learn that uni- versity or institute faculty would not be qualified to teach mucrofinance to unit awards, unit staff staff, and that all trainers for these centers are drawn from the unit desa system. Instructors are appointed for a four-year term, with possible renewal for a can earnfrom three second four-year term.After a maximum of eight years, a BUD trainer returns to the bank in another capacity.This arrangement was developed because it is to six times annual believed that rotating trainers helps maintain the quality of the training."6 per capita GNP Retention Government service in Indonesia, as in many developing countries, is seen as a sinecure. Once appointed, employees tend to stay and turnover is relatively low. BRI staff turnover figures are not available by division, but BUD staff re- tention is known to be high. BUD managers at the regional and head office levels are subject to the bank's internal transfer rules, but a large portion re- mains at BRI. In part, the continuity of unit and BUD branch staff is due to the fact that it is difficult for BRI to lay off staff and almost impossible to fire employees in the absence of proven fraud. However, there is also high retention of good BUD staff at the units and branches.This is partly due to the difficulty of being promoted beyond the unit development officer position at the branch; within BRI there are not many al- ternatives for most unit staff and their BUD supervisors. But retention of good staff is also due to a number of positive factors: the hiring of local high school graduates for entry-level positions, the heavy investment in high-quality in-house training, the provision of good salaries (by local standards), the considerable re- sponsibility given to staff, and the substantial cash incentives and institutional recognition provided for good performance. And as with staff at Bank Dagang Bali (BDB), many BUD unit and branch staff mention their relationships with clients as an important component of job satisfaction. 342 The Microfinance Revolution: Lessons from Indonesia Bank Rakyat Indonesia: Microfinance Outside the Unit Desa System In addition to its unit desa system, which is discussed further later in the chap- ter, BRI engages in various kinds of microfinance activities through its branch- es. Two contrasting examples-with different lessons for niicrofinance are discussed here. Both provide loans that are on average smaller than KUPEDES loans, and both reach borrowers who are poorer than most unit desa borrow- ers. One is commercially financed and charges higher interest rates than KU- PEDES; the other provides donor- and government-subsidized loans. In different ways, both are administered for the government at BRI's branches, but not at the unit desas. The BKDs The Badan Kredit Desa More than 100 years old, the Badan Kredit Desa (BKDs, or village credit or- (Village Credit ganizations) were among the first microcredit institutions in Indonesia.They are located on Java and the neighboring island of Madura; each BKD is owned Organizations), by its village government.'7 administered at BKD history. The BKDs were an outgrowth of the lumbung desa (village gra- naries) established by the Dutch colonial authorities in the late 19th century"8 and BRI branches, are the slighdy later bank desas, which operated on a currency-instead of a paddy- basis (see chapter 9).The lumbung desas were originally conceived as the start of more than a a Raiffeisen-type cooperative rural banking system, while the bank desas were gov- ernment-initiated and funded by external sources. Initial capital for the bank desas hundred years old. was provided by loans from the government, the district, and the local lumbung desa, as well as repayable shares belonging to villagers. Over time a growing por- From the start, they tion of the bank desas' loan portfolios came to be funded by retained earnings. As discussed in chapter 9, the first government decree regulating BKDs (bank were set up to be desas and lumbung desas) was issued in 1907. Modified several times, it was re- vised in 1929 under the Ordonasi Staatsblad 1929 No 257, which regulated sustainable BKD ownership, management, business activities, and supervision. The idea of the lumbung desas as the lowest tier of a cooperative rural bank- ing system was never implemented. BKDs are owned by village governments; the BKDs are not cooperatives or member-owned institutions. During the colo- nial period BKDs operated in all parts of the colony, but they generally did not work well in the Outer Islands. Today BKDs exist only on Java and Madura (an island off the eastern coast ofJava). From the beginning, the BKDs were set up to be sustainable. Interest rates had been set high enough not only to cover costs but to increase the institutions' equity base through retained earnings . . . Subsidies were limited to initial start-up support after which the institutions had to become viable. Maturities and repayment schedules of the loan products were designed according to the needs of the customers. Unsecured, "character- Institutional Development for Large-Scale Sustainable Microfinance The Transformation of the Unit Desas, 1984-96 343 based" lending was common and ... the BKD experienced the benefits of low information cost due to the social and spatial proximity to their clients. Innovative techniques like group- lending and joint liablity were tested (and given up again). Strict loan enforcement was common and resulted in [satisfactory] loan performance throughout most of the time; and the fee-based compensation upon loan collection worked as an incentive for BKD staff. -Steinwand 2001, p. 103 The BKDs declined duringWorldWar II and in the years after Indonesia's In 1996 the BKDs independence. Including both lumbung desas and bank desas, the number of BKDs fell from about 13,500 in 1940 to about 4,000 in 1950 (Steinwand 2001, had more than p. 124). In the early 1950s consideration was again given to transforming the BKDs into a cooperative system, but once again the cooperative model was not 775,000 implemented.The government decided to rehabilitate the BKD system in the 1950s, and the number of BKDs gradually increased. But as noted, the Soe- outstanding loans harto government preferred rural banking that was less autonomous than the village-owned BKDs-such as the state-owned cooperatives (Koperasi Umt Desa, and more than or KUDs) and BRI's unit desas. 250, 000 voluntary Effects of the 1992 Banking Law. To comply with the 1992 Banking Law (see chapter 9), the BKDs were granted a number of collective Bank Perkred- savings accounts itan Desa (BPR, or People's Credit Bank) licenses, as they were too small to receive BPR licenses on an individual basis. But opening new BKDs was not permutted. As BPRs, the BKDs are supervised by BRI through its branch net- work on behalf of Bank Indonesia, the central bank. 19 Bank Indonesia reim- burses BRI for the salaries of its branch-level BKD supervisors. BRI, however, incurs some expense for work by its regional and head office staff on BKD data management and reporting. In 1991 BRI gave the BKDs permission to mo- bilize voluntary savings from the public. And the following year BRI recapi- talized some of the inactive BKDs. The board of a BKD usually consists of the village head as chair, a mem- ber of the village government (the secretary or the head of the village finance office), and one other member. The board receives a commission of 3 percent of loan principal collected. BKD services. At the end of 1996 there were 4,806 active BKDs, each pro- viding small commercial loans to individuals residing in the vlllage. Not all vil- lages have BKDs, and in practice these organizations sometimes serve a few residents of neighboring villages as well. Although figures by gender are not available, many, probably most, BKD customers are women. In 1996 there were more than 775,000 outstanding loans and more than 250,000 voluntary savings accounts (table 14.7).Borrowers are required to de- posit compulsory savings, but voluntary savers need not borrow. The average 344 The Microfinance Revolution: Lessons from Indonesia Performance of the Badan Kredit Table 14.7 Desa, 1992-96 (millions of U.S. dollars except where otherwise indicated) Indicator 1992 1993 1994 1995 1996 Number of active BKDs 4,731 4,889 4,801 4,806 4,806 Number of outstanding loans 1,008,756 907,451 765,586 779,652 779,599 Number of voluntary savers 322,179 311,066 283,592 267,931 251,028 Value of BKD deposits at BRI 21.4 23 9 24 8 26 0 28.0 Value of outstanding loans 31 6 33 8 37 0 40.1 43 8 Value of savingsa 6 8 7 6 8 3 8 7 8.3 Income 11,9 12.7 12 8 13 8 15.4 Expense 5 5 64 6 9 7 0 8 0 Profit (adjusted for loan losses)b 3 8 6.3 5 5 5.8 59 Adjusted return on assets (percent)c 7.5 11 4 9 2 9 3 87 a The total includes both compulsory and voluntary savings Voluntary savings, typically 15-20 percent of total savings, totaled $1 2 million in 1992 and $1 4 milkon in 1996 b Profit minus change in arrears past final due date plus actual bad debt expense c Adjusted profit / (total assets less arrears past final due date) Source BRI data outstanding loan balance in 1996 was about $56 (5 percent of the country's $1,073 GNP per capita); the average compulsory savings account balance was $32 (3 percent of GNP per capita) and the average voluntary savings account balance was $6 (less than 1 percent of GNP per capita). BKD loans rarely reach above $200. Loans are made to individuals, usually for short terms (typically 12 weeks). For the month of December 1996,146,219 loans were disbursed; of these 78 percent were for 10-12 weeks; 13 percent for 35 days; and 9 per- cent for seasonal agricultural loans. Collateral is not required.The loan process is simple and fast; approval and disbursement are often received on the same day the application is made. Most BKDs are open one morning a week (al- though some are open more frequently). They are staffed by members of the BKD board and a bookkeeper. The BRI supervisor is also present when the BKD is open; at the end of the morning, cash on hand is deposited in the BKD account at BRI (the account may be held at the supervising branch or at a near- by unit desa). Interest rates. Interest rates on loans are set by local governments and may vary from one BKD to another. However, the following example illustrates a typical arrangement.A loan of $50 is taken for 12 weeks, and 12 weekly pay- ments of $5 each are made.The first payment is for the interest; the second pay- ment is for compulsory savings (which are returned without interest to the borrower the following year at the holiday following the end of the Islamic fast- ing month). Payments 3-12 repay the loan principal. Stated interest is thus 20 Instilutional Development for Large-Scale Sustainable Microfinance The Transformation of ihe Unit Desas, 1984-96 345 percent for a 12-week period, but the effective monthly interest rate is high- er because of the forgone use of the compulsory savings. Ravicz (1998, p. 15) calculates the BKDs' annual effective interest rate to be 131 percent if savings are returned when the loan is repaid. Since savings are typically returned only once a year, the actual interest rate is higher.Thus the BKDs have a much high- er annual effective interest rate than the unit desas. Repayment. It is difficult to deterrnine BKD arrears and default rates be- cause loans in default are written off only after five years-and sometimes not even then. Like most BPRs, BKDs also do not provision adequately for loan losses (chapter 9). Ravicz (1998, p. 13) estimated default and arrears rates by The average writing off all loans in default each year. According to these calculations, the BKD arrears rate (loan volume in arrears net of previous years' defaults) was outstanding loan 21 percent in 1996. Since each BKD village board is responsible for loan col- lection, results vary widely among individual BKDs.The standardized training balance in 1996 and the discipline of unit desa staff cannot be replicated at the BKDs, which are operated by village leaders. But since a BKD serves a small community where was 5 percent of everyone is known to everyone else, many BKD loans that fall into arrears are eventually collected. GNP per capita Sustainability. BRI provides a line of credit for BKD lending operations. But by 1996 most BKDs financed their loans with retained earnings (deposit- ed at BRI) and the compulsory savings of their borrowers. BKDs that did not have adequate funds to finance their loan portfolios could borrow from BRI at a 21 percent annual effective interest rate.At the end of 1996 the BKDs had nearly $44 million in outstanding credit and $8.3 million in savings (of which 83 percent was in compulsory savings). BKD deposits at BRI totaled $28 mil- lion (see table 14.7). The BKD system has been profitable for decades.Table 14.7 shows the prof- its from 1992-96, adjusted for loan losses. In 1996 the BKDs made $5.9 mil- lion in profit after adjustment for loan losses and provisioning. The return on assets, adjusted for arrears past the final due date, was 8.7 percent. Like moneylenders, BKDs offer low transaction costs for borrowers. They fall between unit desas and informal commercial moneylenders in terms of both interest rates and the total cost of credit to borrowers. Managed by village leaders, the BKD system has a lower repayment rate than the unit desa system. But the BKD system is also profitable, primarily be- cause the lower repayment rates are compensated for by the higher interest rates for loans, interest received on BKD deposits at BRI, and the much lower op- erating costs of the BKD system. And in 1996 the system reached nearly 800,000 borrowers with an average loan balance of $56 and more than 250,000 voluntary savers with an average account size of $6. The unit desas and the BKDs together serve a substantial portion of the low-income people ofJava and Madura (where 59 percent of the country's pop- ulation lived in 1995), and the unit desas serve the rest of the country as well. 346 The Microfinance Revolution: Lessons from Indonesia The BKDs have deeper outreach; the unit desas have wider outreach.With both systems in operation, BRI makes an extraordinary contribution to financing low-income people throughout the world's fourth most populous country. But there is still room for expansion-both for the unit desas and for the BKDs (if the regulation is changed to allow new BKDs to open). Pembinaan Peningkatan Pendapatan Petani-Nelayan Kecil Sponsored by the Ministry of Agriculture and subsidized by donors and the government, Pembinaan Peningkatan Pendapatan Petani-Nelayan Kecil (Income Generating Program for Marginal Farmers and Fishermen, known as the P4K Project) was introduced as part of the second phase (1990-96) of the ministry's large income-generating project. P4K is a subsidized credit and training pro- The BKD system gram for poor rural self-help groups. Participants include small and marginal farmers, farm laborers, sharecroppers, and small fishermen organized into small has been profitable groups of 8-16 households.Women account for about half of P4K participants.20 P4K provides training (in such areas as skill development, education, health, and for decades. P4K is family welfare); loans, which require compulsory savings; and voluntary sav- ings facilities. a highly subsidized The BKDs are Indonesian (originally European) in background, and they adhere to the financial systems approach to microfinance as practiced for more credit and training than a hundred years in Indonesia. In contrast, the subsidized P4K program is a recent donor- and government-driven import influenced by the poverty lend- program for poor ing view of microfinance, but with a modified approach: P4K does not pro- vide credit to the extremely poor. The BKDs (and the unit desas) provide rural seff-help groups individual loans; P4K uses a group lending methodology. Since the BKDs and P4K serve many of the same villages (as do the unit desas), comparison among them is instructive. The microfinance component of the P4K program is administered for the Ministry of Agriculture by BRI through its branch network. In 1996, 80 per- cent of loanable funds were financed by the International Fund for Agricul- tural Development (IFAD) and 20 percent came from Bank Indonesia liquidity credits. The mean cost of funds provided to the program was a highly subsi- dized 6.3 percent a year. P4K also received support from the United Nations Development Programme (UNDP) and the Dutch government. BRI is required by the government to provide loans under the P4K pro- ject and to collect the repayments. While BRI's responsibility for P4K credit is at the branch level, many of the cash transactions are carried out at the units or even at the village posts as a service provided to the branches by the units- since the units are more conveniently located for most borrowers than the branch- es. However, the last payment of a P4K loan must be made at a BRI branch. P4K's target population is a result of the program's view that access by the poor to financial resources is based on a series of steps (IFAD 1996, p. 26).Thus the extremely poor require grants, charity, and employment. However, house- holds with annual per capita income equivalents above 320 kilograms of rice (about equal to $137 in 1996 prices) can use commercial microfinance, some- Institutional Development for Large-Scale Susta nable Microfinance The Transformation of the Unit Desas, 1984-96 347 times with special conditions.The households in between-those with annu- al per capita incomes equivalent to 240-320 kilograms of rice are provided first with revolving credit financed by grants, then with subsidized loans. Ac- cording to P4K administrators, in practice the project targets primarily house- holds with annual per capita incomes of 240-320 kilograms of rice. The P4K program was designed specifically to assist self-help groups of poor borrowers make the transition from P4K credit subsidies to commercial mucrofi- nance; thus borrowers are limited to four subsidized P4K loans with maturities of 12-18 months each. Group members are joindy and severally liable for the loans to group members; this guarantee serves as the borrowers' collateral. Interest on the loans is 12 percent a year flat rate (21.15 percent annual effective rate; McGuire, In contrast to the Conroy, and Thapa 1998, p. 154). However, actual interest is higher since the groups are required to save at BRI as a condition of receiving a loan; the amount unit desas and the of savings required from each group is $21 for the group's first loan, 10 percent of its second loan, and 20 percent of its third and fourth loans.The borrower does not BKDs, only 19 have the use of his or her compulsory savings during the loan period, although the savings of the group members are deposited in a BRI SIMPEDES account that percent of P4K pays interest on the minimum monthly balance.Taking into account the compul- sory savings requirement, the annual effective mterest rate for loans ranges from 24-33 groups received credit percent (depending on the percentage of compulsory savings required), or from 22-31 percent for prompt payers, who receive a small rebate (Ravicz 1998, p. 66). after their first loans In 1995 the credit limit for a borrower in a newly eligible P4K group was $43; for the borrower's fourth loan the limit was $128. The average loan size in 1995 was $66, or 7 percent of GNP per capita. A small number of self-help groups (less than 2 percent) continue onto a fifth loan. In this case the inter- est rises to 18 percent a year flat rate (on the original loan balance).Taking into account the savings requirement of 25 percent of such loans, and the interest paid on the savings, Ravicz (1998, pp. 66-67, 73) calculates that these terms are equivalent to a 62 percent annual effective interest rate on a 12-month loan with monthly installments. Between March 1990 and March 1996 the P4K project disbursed 75 bil- lion rupiah in loans to members of 37,289 self-help groups (IFAD 1996, p.45). In fiscal 1996, $10.9 million was disbursed in loans to an estimated 153,480 individuals in 15,137 groups.As of March 1996, there was $2.8 million in sav- ings from P4K members. During the six-year period the P4K program reached nearly half a million households, most of them in the lowest income quartile of Indonesia's popu- lation. But to evaluate the program meaningfully, a number of issues need to be addressed. Interest rates. The interest rate charged on the first four loans is 1 percent a month flat rate on the initial loan balance. But taking into account the forced savings (which are inaccessible to the borrower during the loan period), the annual effective interest rate on the declining balance paid by the borrower is 24-33 percent.Within this range, the rate depends on the amount of compul- 348 The Microfinance Revolution: Lessons from Indonesia sory savings required; Ravicz (1998, p. 71) estimates an average annual effec- tive interest rate of 27 percent. For the few groups that take fifth loans, the an- nual effective interest rate for the same type of loan, including the compulsory savings requirement, is 62 percent. Some group members deposit voluntary savings with their groups to be onlent to group members. "Each group determines the interest rate it will charge on these loans.Typically, groups lend out these funds at funds at a 5 percent flat rate per month.This is equivalent to about a 154 percent annual rate on a de- climng balance" (Ravicz 1998, p. 67). Participation. Participation in the P4K program is difficult to measure be- cause between March 1990 and March 1996 only 19 percent of P4K groups Since most P4K received credit after their first loans (those that received only one loan are known as "resting" groups), while 21 percent had never received a loan ("sleeping" borrower groups groups) (Ravicz 1998, p. 73). But the program does not track the number of loans by the year in which the groups received their first loans, and it is un- receive only one clear how many groups have been "resting" for years, how many had recently repaid their first loans, and how many were unable to obtain additional cred- loan, they do not its. Nevertheless, 51 percent of the P4K groups had received credit by March 1993 and 65 percent by March 1994 (IFAD 1996, p. 40). But as noted, during have the incentive to 1990-96 P4K had not re-lent to 81 percent of its borrower groups after their first loans. This provides a sharp contrast to both the BKDs and KUPEDES, repay that BKD which help creditworthy borrowers with good repayment records move out of poverty by continuing to lend to them in gradually increasing amounts. and KUPEDES Repayment. Information on the total volume of P4K loans in arrears is not borrowers have available. Ravicz (1998) estimates arrears by comparing the number of groups with loans in arrears with the number of groups with outstanding loans. Groups with arrears were at or below 6 percent of groups with credits through 1993. But this percentage increased to 10.5 percent in 1994 and to 18.7 per- cent in 1995. In addition, the volume of payments overdue increased from 1.6 percent in 1993 to 7.4 percent in 1995. However, Ravicz notes that these fig- ures overstate arrears because BRI does not write off P4K loans in default. The arrears appear to be caused in part by the fact that the P4K program uses agricultural extension workers for the P4K credit component.The extension workers typically do not have a financial background, and they are not under the control or supervision of BRI. In addition, they have no direct incentive to investigate the creditworthiness of self-help groups or to help ensure time- ly repayment of loans (see Ravicz 1998, p. 74). And since most P4K borrower groups receive only one loan, they do not have the incentive to repay that BKD and KUPEDES borrowers have-the assurance of another loan for qualified borrowers with good repayment records. Program costs. The costs of the P4K program are high. In 1995 the annu- al cost of the program, including Ministry of Agriculture costs, was $3.3 mnil- lnstituttonal Development for Large-Scale Sustarnable Mlcrofinance The Transformation of the Unit Desas, 1984-96 349 lion, or 30 percent of the loan volume.The cost per new group loan was $383; for old group loans it was $213 (Ravicz 1998, p. 74). Program subsidies. Ravicz (1998, pp. 70-71) calculates a lower-bound es- timate for the 1995 P4K subsidy dependence index (SDI) of 262 percent, in- dicating that instead of its 27 percent average annual effective interest rate, the P4K program would have to charge 98 percent to operate sustainably (see box 2.4 in volume 1 for discussion of the SDI). BRI's role in P4K. At BRI, the P4K program is under the Small Business, Food, and Cooperative Division, which is also responsible for other subsidized Instead of its 27 rural credit programs. Ravicz (1998, p.74) comments on BRI's administration of P4K: percent average BRI operates this program from its branches rather than through annual effective its Umt Desa system. This is highly surprising given that BRI branches are located only in district capitals, and generally make interest rate, P4K quwte large loans. In contrast, Unit Desas are located in sub-district capitals, and make much smaller loans.Thus, Unit Desas are much would have to closer to the credit end-users both geographically and in terms of the types of credits they typically issue. Relocating the charge 98 percent to program to Unit Desas would facilitate collection efforts and might also increase the proportion of groups receiving more than operate sustainably one credit as it would be much easier for groups to apply for additional credits. From the point of view of P4K participants, Ravicz is right. But since 1984 BRI has carefully separated its government- and donor-subsidized credit pro- grams, which are operated through its branches, from the microbanking divi- sion's KUPEDES credit program, which operates commercially at the umt desas. Offering a large commercial credit program that has achieved continuing high repayment rates in the same small bank oudets as much smaller subsidized cred- it programs with repayment problems would hardly be recommended under best practices in microfinance. Figure 12.1 shows clearly why P4K is adminis- tered through BRI's branches, not its unit desas. Microfinance at BRI's branches: comparing the BKDs and P4K In one sense comparing BKDs and P4K is rather like comparing mangoes and raisins: they are very different. But both provide finance to low-income peo- ple in Indonesian villages, and since resources for developing microfinance are limited a brief comparison of the two can be useful. Both provide small loans, require compulsory savings to obtain a loan, and also collect voluntary savings. Supervised by BRI at the branch level, the BKD system is commercial in orientation and charges higher annual effective interest rates on loans than do both P4K and the unit desas.The BKDs, which serve individual clients, have 350 The Microfinance Revolution: Lessons from Indonesia low operating costs. P4K is a relatively new program providing training and cred- it to groups of borrowers. Its subsidized credit component, managed by BRI also at the branch level, has high operating costs. Both programs have lower av- erage loan balances and higher arrears rates than KUPEDES at BRI's unit desas. BKD loans are financed by retained earnings, savings, and commercial loans from BRI.The BKDs receive no subsidies and are profitable after adjusting for loan losses and provisioning. In contrast, the P4K portfolio is financed by donor grants and low-cost donor and government loans. P4K was far from profitability-although its lower-bound estimated 1995 SDI of 262 percent was considerably lower than the 520 per- cent estimated by Ravicz for 1992. P4K is not profitable primarily because of its subsidized interest rate and its high operating costs. P4K is not Like the BKDs, P4K reaches large numbers of poor people-but in the case of P4K most borrowers have received only one loan, which typically has only profitable primarily a limited effect on helping poor borrowers move out of poverty. In contrast, both the BKDs and the unit desas provide repeat loans to all creditworthy bor- because of its rowers who want to reborrow.The difference between commercial and subsi- dized microcredit is symbolized by the fact that the BKDs (like the unit desas), subsidized interest call their customers clients, indicating a business association.The P4K program calls its customers beneficiaries, implying a form of charity. rate and its high In 1995 the subsidized P4K program would have had to charge an estimated nmnimum 98 percent annual effective interest rate to become sustainable. The operating costs BKDs were financially self-sufficient, despite relatively high arrears, with an an- nual effective interest rate of 131 percent (including the effect of compulsory savings, assuming that the savings are returned to the borrower at the end of the loan). In comparison, the unit desas were very profitable, with an annual effec- tive interest rate of 32 percent for prompt payers on most loans.2' It should be noted, however, that, including the compulsory savings component of their loans, P4K borrowers pay annual effective interest rates that are not much different from KUPEDES rates.And P4K group members who borrow from their P4K groups at an annual effective interest rate of 154 percent pay far higher rates than do KUPEDES borrowers.22 The BKDs and the P4K program have attained outreach that is both wide and deep. The advantages of the BKDs are that they are convenient for bor- rowers, have low operating costs, are well supervised by BRI, and are expe- rienced, trusted, well-capitalized, and profitable. The disadvantages are that they are available only on Java and Madura (although the majority of Indonesia's population lives in the area they serve), have relatively high arrears rates, and charge substantially higher annual effective interest rates than P4K or KU- PEDES. Like the BKDs, P4K's advantage is its outreach. But P4K has many disadvantages. It has high subsidies, high operating costs, and relatively high arrears. It does not provide most of its borrowers with continuing access to loans, and its field staff know little about finance. And the P4K program is not sustainable. Institut ona Development for Large-Scale Sustainable Microfinance The Transformation of the Unit Desas. 1 984-96 351 If P4K were ended, its savers and creditworthy borrowers could become unit desa or BKD clients-thus saving the P4K subsidies, allowing borrowers to reborrow, and improving the service to clients. Unlike the BKDs and the P4K program, the unit desas require collateral for most KUPEDES loans. But the KUPEDES Skala Kecil pilot project in small-scale, noncollateralized cred- it (see chapter 12, note 23) was found to be successful, and, in some cases, small loans without collateral have been incorporated into KUPEDES-although there is considerable scope for expansion of such loans. In thls context, the contin- uation of large P4K subsidies-and the administration of the P4K program by BRI-should be reconsidered. The BKDs (and their villages and cients), however, benefit substantially from Microfinance BRI supervision, lines of credit, and deposit facilities. And the BKDs, like the unit desas, continued their wide outreach and their profitability during the re- divisions with cent Indonesian crisis (see chapter 15). hundreds of Microbanking in a Division of a Multipurpose Commercial Bank: thousands or Structural Issues millions of clients Many lessons have been learned from BRI and other commercial banks about structural issues that emerge when microfinance is carried out in a bank divi- need to use the sion.23 This arrangement offers some advantages for microfinance operations. BRI's head office supports the unit desa division (BUD) in areas ranging from McDonalds internal audit to research and development, from logistics to prudential reporting. At the regional offices, BUD is provided with services from other bank divi- model-afew sions, including Human Resources Development, Logistics, and Systems and Technology. And BRI branches play a major role in the units' supervision and products in high financial intermediation. On balance, however, the benefits received by a microfinance division in a demand large multipurpose commercial bank are likely to be outweighed by the dis- advantages.These stem from organizational problems inherent in the structure of an institution in which one division is engaged in a business that differs sub- stantially from that of the others. Of course, variations in business activities, op- erations, and priorities exist among other bank divisions, but not to the same extent as in microbanking.At BRI there are some crucial differences between the operations of the microbanking division and those of the rest of the bank. Much of the argument made in this section is drawn from BRI's experi- ence, some from elsewhere. Not all of BUD's experience is relevant to other commercial banks with microfinance divisions. But since BRI has by far the most fully developed history of such a relationshlp, its lessons are worth exploring. Approaches taken by other multipurpose banks providing commercial micro- finance are discussed in chapter 19. Microfinance products must be designed to meet the demand of low-in- come clients and priced for institutional profitability. They must be delivered to large numbers of clients in many different localities, and loans must be re- 352 The Microfinance Revolution: Lessons from Indonesia covered.Thus the operations of a bank's mnicrofinance division are necessarily labor intensive-far more so than those of other bank divisions. And microfi- nance products and operations must be kept simple. Divisions serving a small number of wvealthy clients can offer multiple products of different kinds. But microfinance dlvisions with hundreds of thousands or millions of clients need to use the McDonalds model for products-a few products in high demand. What is crucial for microfinance clients is not a large number of products, but a well-designed combination of a few products that each client can custormize for his or her own use.Thus the unit desas have one loan product (with 36 com- binations of maturities and payment schedules) and three basic savings accounts (with different ratios of returns and liquidity). Keeping large-scale microfinance simple in a multipurpose commnercial bank can be difficult, as the need for sim- Keeping plicity may not be understood in other divisions with different functions. The staff of the rnicrofinance division must be recruited on criteria that dif- microfinance simple fer in some important respects from those used for the staff of the rest of the bank. Microfinance management, supervision, training, and reporting are marked- in a multipurpose ly different from industry standards. The division operates on a larger scale in terms of number of offices, staff, vehicles, and the like than the bank's other di- commercial bank can visions. Microfinance operating costs must be significantly higher than those of the other dcvisions.And the division's products must be priced well above bank- be difficult, as the ing industry norms. In the case of BRI, another difference is that the microfi- nance division is substantially more profitable than other bank divisions. needfor simplicity Yet as part of a conmnercial bank, the microfinance division does not have the authority to ensure that the basic requirements needed to conduct its busi- may not be ness are met. Its ability to function effectively depends on whether the bank's chief executive officer, and often the managing directors of the bank's other understood in other divisions, understand and support microfinance. Even high microfinance prof- its do not necessarily ensure the support of other bank directors; profits can also divisions engender rivalries. Over time, it iS likely that some managers will be support- ive of the niicrofinance division and others will be less so. It is possible to succeed in the short and medium term with a dedicated microfinance manager, a supportive chief executive officer, and high-level sup- port from relevant ministries-as occurred at BRI. But depending on the leadership of a few individuals is not an effective way to achieve long-term in- stitutional viability. The structural problem of the bank division model for microfinance can be stated simply: the responsible director does not control the main decisions that affect the activities of the microbanking division, while the bank's other directors and chief executive officer are typically not qualified to do so. For long- term sustainability, the director in charge of microbanking must have full au- thority over all aspects of the microbanking business. There are two kinds of dangers to the development of commercial mni- crobanking. One, discussed here only briefly, comes from outside the bank; this danger is considered more generally in volume 3 in the context of the broad- er issue of institutional structures and country conditions that are suitable for Institutional Development for Large-Scale Sustainable Microfinance The Transformation of the Unit Desas, t984-96 353 microfinance. The other danger, specifically related to microfinance divisions of commercial banks, comes from inside the bank. Discussed below, this is an important and little-understood potential hazard for the development of mn- crofinance in commercial banks. External dangers to sustainable microfinance Governments are generally not monoliths. Some policymakers may understand and support the development of commercial microfinance, while others may not agree with this approach.Thus while a country's ministry of finance or cen- tral bank may support commercial microfinance, government agencies charged with poverty alleviation may require the microfinance divisions of commer- The staff of the cial banks to implement subsidized credit programs-not understanding (or not caring) that such programs undermine the divisions' commercial microfinance microfinance division activities. In Indonesia an ad hoc compromise was reached at BRI (see figure 12.1). must be recruited on Since 1984 subsidized microcredit programs have been channeled through the branches, while commercial microcredit is provided through the units.This has criteria that differ been a successful arrangement for the development of the microfinance division. But the agreement was not formally mcorporated mto the permanent bank struc- from those usedfor ture, and it is vulnerable to changes imposed by external or internal sources. Another danger to sustainable microfinance arises when regulatory authorities the staff of the rest apply general industry standards without realizing how these affect the devel- opment of microfinance in both publicly and privately owned banks. Exam- of the bank ples include: * Usury laws and interest rate caps. These typically prevent microfinance divisions from covering their costs and discourage banks from entering this market. * Requirementsfor new branches. If a microfinance unit with four staff members is subject to the same regulations as a regular bank branch (such as require- ments for capital, reporting, collateral, and the like), few units will be opened and demand will not be met. * Reporting requirements. Microfinance divisions with large numbers of small loans should be permitted to report on loans in aggregate, and through sampling, rather than being required to follow the industry practice of providing reg- ular reports on all individual loans. * Requiring accounting by the accrual method. The accrual method of accounting is generally too complicated for large microfinance systems and does not encourage rapid, effective followup of delinquency.A cash accounting system in which only income received is reported is usually more appropriate for microbanking. * Loan classification. Given the large number of small loans at the local level, classification of overdue loans should not be carried out on individual loans by a loan committee but rather by automatic classification based on loan aging and repayment history. * Requiring collateral to be notarized. This requirement makes it difficult to col- lect simple forms of collateral from low-income clients and increases the cost 354 The Microfinance Revolution: Lessons from Indonesia to the borrower because notary fees tend to be high at the local levels of de- veloping countries. * Supervision of the savings of the poor. When the rnicrofinance divisions of com- mercial banks coIlect savings from the public, it is important that they be ap- propriately publicly supervised. On the one hand, industry standards for supervision are often not appropriate for microsavings; on the other, too lit- tde supervision can be dangerous, especially because microfinance is new to the banking system. Internal dangers to sustainable microfinance The relationship between a bank's microfinance division and its other divi- sions is sensitive at every level of the organization. The main internal prob- The microbanking lem faced by a microfinance division is that its purpose and requirements may be musunderstood within the bank, and that the head office may not put a high director does not enough priority on building a constructive and cooperative relationship be- tween this division and the others in the bank.The following sections discuss control the banks some of the dangers to the microfinance division-and hence to the bank- that can arise from an inadequate understanding of microfinance within the important decisions bank.With the possible exception of government requirements to implement subsidized credit programs, these internal dangers can apply to both private affecting and state-owned banks. microfinance; the The role of the microbanking unit in a multiservice bank. The core of a suc- cessful microbanking division is its labor-intensive, client-centered, small-scale other directors profit centers. But these units must be closely supervised by a higher level of the bank's organization. At BRI the branches play this role, also acting as fi- typically are not nancial intermediaries for the units. BRI branch managers must manage the commercial, retail, and sometimes corporate banking activities of the branch, qualified to do so implement the government's subsidized credit programs, and supervise the unit desas that report to the branch. The managers and BUD staff at the branches are trained to understand the role of the branch in unit desa development. But it is not easy to maintain an appropriate balance among the various functions of the branch, especially since commercial microfinance does not represent a large part of the branch's responsibilities-although it typically represents a large portion of its profits (or it covers losses from other branch activities). BRI's divisions responsible for the main activities carried out at the branch- es, as well as divisions that carry out bankwide functions such as human re- sources, management information systems, operations, and logistics, sometimes perceive the units as arms of the branches. A unit can, to some extent, serve as an arm of its branch. But if the unit is to be profitable, its service to the branch must be an ancillary activity. For example, at BRI a borrower who lives in a rural area near a unit, but who wants and qualifies for a loan larger than the KUPEDES maximum, may take a larger loan from the branch. However, the borrower may prefer to pay the loan installments at the unit because of its more convenient location. In collecting the payments for this loan, the unit performs institutional Development for Large-Scare Sustainable Microfinance The Transformation of tIhe Unit Desas, 1984-96 355 a service delivery function for the branch. Similarly, computer terminals can be installed at the umnts to service branch-based checking accounts, facilitate money transfers, and house automated teller machines. But some branch managers and representatives of a bank's other divisions tend to emphasize the unit's function in delivering services to the branch to the point that the unit's primary activity-providing financial services to the economically active poor-is threatened.This emphasis is dangerous for two reasons. First, the two functions have different basic requirements. Second, the microbanking function of the units enables the division's profitability-and in the case of BRI, the bank's profitability. The relationship Labor-technology ratios. The appropriate use of labor and technology is di- rectly related to the role of microbankmg units in a multipurpose bank.Those between a bank s who see the units as extensions of the branches tend to argue for substantial increases in the computerization of the units and for major decreases in their microfinance labor costs.Those who understand microbanking realize that small savers and borrowers demand personal service from helpful staff delivered at multiple lo- division and its cations, and that they are willing and able to pay for these services.While labor may be costly, profitable microfinance programs are labor intensive. This cru- other divisions is cial point is not always understood at a bank's head office or at its regional and branch levels. sensitive at every With better technology, transaction costs can be safely lowered in the units in a variety of ways, both in the units' services to the branch and in their level of the microbanking activities (especially in backroom operations). But operating costs for microbanking are necessarily much higher than those in the bank's organization other divisions.The many reasons include higher labor costs, higher trans- action costs, expenses related to the operations of many small bank outlets, the specialized staff training required, and the need for research and devel- opment on products and services specially designed for the microfinance mar- ket.As discussed in the section on unit desa profitability in chapter 12, the operating costs of the unit desas in 1996 were 6.5 percent of average earn- ing assets-far higher than general banking industry standards in Indonesia or elsewhere. In modern banking, however, optimal ratios of labor and technology vary considerably among bank divisions. A bank with multiple divisions needs to ensure that its mucrobanking division is not outvoted by other divisions on issues vital to the sustainability of microbanking, and that its rmcrofinance outlets do not become branch arms. It must be recognized throughout the bank that the units of the microfinance division are the bank's network for providing finan- cial services to low-income clients; they are neither branch outlets nor small branches. It is not easy to achieve this perception in a multipurpose bank in which microbanking is one among many divisions. The generalist management model. In the generalist model of management, found in many commercial banks, managers above a certain level are regular- 356 The Microfinance Revolution: Lessons from Indonesia ly transferred (typically every two to four years) from one division to anoth- er. This model does not work well for mncrofinance because microbanking is a different business from standard conmmercial banking. Under the generalist management model, soon after a manager learns the microbanking business he or she is transferred to corporate, international, or commercial banking (all of which are generally considered plum assignments relative to microbanking). In the long run this model prevents the development of a strong, stable, experi- enced management team that understands the business of the rmcrofinance di- vision.This has been a serious problem at BRI. The downsizing syndrome. Bank managers often think that the instruments and services designed and priced for commercial and corporate customers can Decisions about use simply be downsized-that is, shrunk-for use by nucrofinance clients.This is a mistake, because the demand for and pricing of microfinance are fundamentally of profits and different from other bank products. Microfinance clients want small loans and small savings accounts with the characteristics discussed in earlier chapters.AI- allocation of though some urbanized, better-off niicrobanking clients can use some of the bank's standard commercial products, most standard commercial products are resources may be not suitable for most microfinance clients. For its activities to be sustainable, a bank's microfinance division must have the authority to develop and imple- made uwth ment the products and pricing that are appropriate for its market.This has not been a serious issue at BRI because the BUD Division has defended well against inadequate it. But it is a problem in some banks. understanding of the Managing bank support services. The various divisions of a commercial bank share the services provided by the bank's support divisions (such as human re- requirements of the sources, training, and logistics); each user division should pay for the services it uses. However, the managers responsible for the bank's support divisions some- microfinance division times do not understand the special needs of the microfinance division.With- out control over its own requirements for operation, the microbanking division remains vulnerable to uninformed management decisions that could cripple or destroy the microfinance effort.At BRI this issue has been handled through constant personal negotiation. While this method can work for some time, it is not structurally sound for the long term. Use of profits and allocation of resources. Corporate decisions on the use of bank profits and the allocation of resources among divisions may be made with inadequate understanding of the requirements of the microfinance divi- sion, since these are often quite different from the needs of other divisions. As a result the microfinance division may not receive sufficient funds for its main- tenance, improvement, and expansion. Since the microfinance division controls neither the use of its profits nor the bank's investment decisions, the division cannot protect itself in this regard. Once again, the microfinance division is vul- nerable because of its differences. This has been a particularly crucial issue at BRI since the bank's profits come from its microbanking division. Institutional Development for Large-Scale Sustainable Microfinance The Transicrmation of the Unit Desas, 1984-96 357 Uninformed planning. Problems in bank planning can arise because of lack of knowledge about the business of microfinance. For example, it may be con- sidered efficient to combine microfinance units and regular bank branches when both are located in the same area (usually in towns and cities). But these enti- ties serve different clients, need different products and pricing, require differ- ent kinds of staff with different training, and have different requirements for buildings, transportation, and the like. That their consolidation is neither effi- cient nor profitable is not necessarily understood by managers with little ex- perience in microfinance. This issue has been the subject of a long-term, ongoing debate at BRI. Hhat may appear Use of depositsfrom microfinance clients. Well-run microfinance divisions are able to mobilize savings that, although small in terms of individual accounts, to be goodfor the can be very large and very stable-in aggregate. Therefore a bank that faces a shortage of funds may be tempted to instruct its microfinance division to em- bank may not be phasize savings at the expense of loans, in order to finance the activities of other bank divisions. In such cases the transfer price is normally set to encourage sav- goodfor the ings mobilization and to discourage lending. This approach can reduce rural investment and raise interest rates on mi- microbanking croloans higher than is necessary (to compensate for higher interest rates on unit savings accounts).The latter did not occur at BRI during 1984-96. But division much of the units' savings were used to finance large loans at the branches- while the demand for small loans was only partly met.What may appear to be good for the bank as a whole may not be good for the microbanking di- vision or its low-income clients; in the long run it may not be good for the bank either. Managing required subsidized credit programs. Government instructions re- quiring implementation of subsidized credit programs can threaten a bank's commercial microfinance program. Unless the government can be persuaded to change its policy, the bank must find a way to implement the subsidies in ways that do not directly endanger the market approach of its microfinance division. Otherwise the division's viability is threatened.The BRI solution has been to irm- plement the subsidized programs at the branches but not at the units (which are the locus of responsibility for commercial microfinance). Each bank affected by this problem must find its own solution; the two programs can coexist at the same bank if necessary, but not in the same division or in the same retail outlet. "Modernization. " Increasing competition tends to lead banks in develop- ing countries to believe that they must develop a "modern" corporate image. Emphasis is placed on technological superiority, gleaming modern architecture, and managers with multiple degrees, preferably awarded by foreign institutions. A bank active in the microfinance market must understand that this approach should not be imposed on its microfinance division. Nor should the division be considered a pariah because of its simpler image. The economically active 358 The Microfinance Revolution: Lessons from Indonesia poor prefer small, familiar bank outlets with friendly staff who understand their needs and treat them with respect. This problem, though widespread, has barely been recognized; it needs to be understood and resolved. Given that most of the world's population do not have access to formal financial services, and that microfinance can be highly profitable, microfinance units may well become the most important outlets of the banks of the 21st century. Why Do Structural Issues Matter? Banks considering entering the microfinance market should consider carefully The current the lessons from these experiences. BRI's unit desa system faced institutional prob- lems from both outside and inside the bank.These difficulties were typically solved restructuring of BRI on a case-by-case basis; this method allowed the units to function well but made their long-term viability dependent on the skill of their managing director, the indicates that many priorities of BRI's president-director, and the goodwill of the coordinating min- ister for the economy and a few individuals at BRI, Bank Indonesia, the Min- of the bank' istry of Finance, the Ministry of State for Planning, and elsewhere. Realizing better than anyone the vulnerability of the units because of the problems are being ad hoc decisionmaking involved, Sugianto, the long-term managing director responsible for the unit desa system, pressed in the rmid-1990s for a national addressed microfinance policy framework. He wanted to ensure that decisions crucial to the development of commercial mncrofinance in Indonesia-involving such mat- ters as the administration of government-subsidized credit programs and cen- tral bank regulations concerning accounting methods, reporting requirements, and loan classification-would not depend on particular individuals, but would be made in the context of informed microfinance policy. But his untimely death in 1998 preceded the formulation of such a policy framework, and the mo- mentum for its development lapsed. The current restructuring of BRI into a bank providing commercial finance mainly to micro, small, and medium-size retail customers indicates that many of the bank's problems-both external and internal-have been understood and are being addressed (see chapter 15). If BRI is successfully restructured in this way, its role in providing increased financial services to low- and middle-in- come people throughout Indonesia could generate an opportunity for the for- mulation of national policy on microfinance issues. If only BRI and a few scattered banks were involved in commercial mi- crobanking, the structural matters discussed here might not be so compelling. But the issue is urgent because commercial banks in developing countries are the primary institutions with the scale of operations required to reach the hun- dreds of millions of economically active poor who lack access to formal financial services.And many banks are now entering microfinance (see chapter 19). An increasing number of governments and banks have begun to understand that commnercial microbanking is in their own interest. Both push and pull forces Institutional Development for Large-Scale Sustainable Microfinance The Transformation of the Unit Desas, 1984-96 359 Indicators of financial self- Table 14.8 sustainability for the unit desas, 1985, 1990, and 1995 (percent) Indicator 1985 1990 1995 Nominal average yield on loan portfolio 27 4 31 5 31 6 Nominal average interest rate on deposits 10 5 11 3 9 7 Nominal interest rate spread 16 8 20 2 21 9 Real average yield earned on loan portfolio 21 7 22 4 20 2 Real average interest rate paid on deposits 5 6 3 6 0 3 Lowest nominal lending interest rate needed for financial self-sustainability 36 2 27 2 17 5 Lowest real lending interest rate needed for financial self-sustainability 30 1 1824 7 3 Operating costs as a share of Average annual net loan portfolio 20 5 12 9 12 6 Half of the average annual net loan portfolio and deposits 31.5 11 6 8 3 Average annual total assets 15 1 8.0 5 3 Profit (millions of U S dollars)a -0 8 34 3 170 2 Share of profitable units 48 3 891 95 7 Average annual deposit volume/average annual loan portfolio volume 0 3 1 2 2 0 Subsidy dependence index 32 2 -13 7 -4 5 Note Inflation was 4 7 percent in 1985, 7 4 percent in 1990, and 9 4 percent in 1995 There are slight discrepancies between the Yaron, Benjamin, and Charitonenko (1998) inflation figures used in this table and the data in table 2 1, which reports inflation as 4 5 percent in 1985 and 7 5 percent in 1990 a There is a small discrepancy between the 1995 profits provided in this table and current BRI data According to BRI's monthly reports, unit desa profits were 402 6 billion rupiah in 1995 Thus at the year-end exchange rate of 2,308 rupiah to $1, profits were $174 million (see table 2 2) Source Yaron, Benjamin, and Charitonenko 1998, p 160 are present. Government banks are being pushed toward commercial microfi- nance by the large losses they typically incur from subsidized credit programs. This happened in Indonesia, and it is beginning to occur in many other countries-including India and China, which together are home to about half of the world's low-income people. Both state-owned and private banks in many developing countries are also being pushed toward microfinance by increasing competition from overseas banks. Local banks, which fear competition for prime customers from large interna- tional banks, have begun to realize that foreign banks entering their country may be highly competitive in the corporate market, but are unlikely to com- pete in the microfinance market (although this may change in the coming decades). Pull forces include the profits that can be earned from commercial micro- finance, and the political benefits from governments that look favorably on banks that provide financial services to low-income people. However, the question of profits raises one of the most difficult issues that can be faced by a mucrofi- nance division of a commercial bank.Table 14.8 provides indicators of finan- 360 The Microfinance Revolution: Lessons from Indonesia cial sustainability for the unit desa system in 1985, 1990, and 1995. It shows that in 1995 the subsidy dependence index (SDI) for the BUD Division was -44.5 percent, indicating that the division could have substantially lowered its KUPEDES interest rates, raised its interest rates on savings accounts, or both- while still remaining filly sustainable and self-sufficient.24 The profits of BRI's microfinance division have been used to cross-subsi- dize other BRI divisions that serve wealthier clients but have lower repayment rates and are less efficient.Thus the units' clients, the economically active poor, are subsidizing the bank's more affluent corporate and commercial customers. Yaron, Benjamin, and Piprek (1997, p. 129) pose the crucial question: What would have been the impact on the rural economy, and The bank division particularly on BRI-UD [unit desa] clients, if the value of negative subsidies had been used to decrease the spread between model is not likely the BRI-UD's on-lending and deposit interest rates, instead of subsidizing other BRI non-unit desa activities? This question to become the is of the utmost importance because the cross-subsidization withm the BRI (as reflected in the negative SDI) results in adverse organizational income distribution, with small-scale rural entrepreneurs subsidizing the more affluent clientele. standardfor Of course, this is a particular case, not a general rule. Nevertheless, there commercial are clear structural dangers when microfinance is implemented through a di- vision of a multipurpose commercial bank. Moreover, decisions made under microbanking in the this type of bank structure can result in regressive income distribution. The dangers to microfinance in commercial banks can be overcome if they future are recognized and if high priority is given to their prevention and remedia- tion. There will be considerable incentives to do so once banks better under- stand the opportunities for profitable microfinance. With many trials and errors, BRI built an efficient organization and man- agement structure for its microbanking program.This structure enabled the BUD Division to develop the world's first fully sustainable large-scale system of mm- crofinance and to help millions of the economically active poor expand their enterprises and increase their incomes. Given the context, the scale, the risks, and the many new ideas that had to be tested, Indonesia probably could not have developed large-scale commercial microbanking under another orgamn- zational model. By using the existing unit desa infrastructure, the coordinat- ing minister for the economy and the minister of finance-who represented the government as BRI's sole shareholder-and BRI were able to build this system at that time.The owners and managers of other Indonesian organiza- tions were uninformed, unwilling, uninterested, or unable to undertake large mnicrobanking efforts. Many lessons essential for building an organizational structure to provide large-scale profitable mucrofinance can be learned from the process of budld- ing BUD at BRI. One, however, is that the bank division model is not likely lnsiitutional Development for Large-Scale Sustainable Microfinance The Transformat on of the Unit Oesas, 1984-96 361 to become the organizational standard for commercial microbanking in the fu- ture. Although they too have their problems, a bank subsidiary or an indepen- dent bank is a more likely bet (see chapter 21). Indonesia and BRI have been leaders of the microfinance revolution from its inception. Over time, both the enormous potential of commercial micro- finance and the structural problems'inherent in the bank division model of mi- crobanking came to be understood in their Indonesian context. The current restructuring of BRI-with its primary focus on providing commercial financial services to low- and middle-income people is discussed in the next chapter. It represents a clear indication that the country and the bank plan to remain at the forefront of the microfinance revolution. Notes 1. Parts of this chapter are based on Robinson (1997 c), BRI (1997a), and Suglanto and Robinson (1998). 2.The Ministry of Finance represents the government as the owner of BRI. 3. See Yaron, Benjamin, and Piprek (1997, pp. 32-33) for a table that shows the characteristics of old and new pohcy approaches and government interventions in rural finance.That table and the box in the introduction to part 3 provide useful comple- mentary information and analysis. 4. Suglanto became responsible for the unit desa system in 1984; he was appointed managing director in 1986 (see Robinson 1998b) 5. See Patten (1996); Robinson (1997c); and BRI (1997a) for further discussion. 6. Until 1984 this position was referred to as unit desa officer; both titles refer to branch-level positions concerned with supervision of units. 7. Urban-rural remittances in Indonesia have been widely reported from micros- tudies. For example, a 1987 CPIS survey of Jakarta's informal sector workers found that 65 percent of 510 workers sent an average of 31,850 rupiah ($19.37) a month to their famiihes in vlllages (CPIS 1988a). 8. See Schnit (1991, pp. 198-206) for discussion of the debates about whether SIMPEDES would contribute to urban bias or "urbanize" villagers. SIMASKOT re- sponded to existing demand in urban areas for a liquid savings account. Many urban residents who had heard about SIMPEDES requested that it be made available in the cities; this led to the development of SIMASKOT, adapted from SIMPEDES. 9.Television began to reach some villages in the mud-1980s, especlally onJava and Bali. By the 1990s television became widely available in much of rural Indonesia. 10. Much of this section is based on BRI (1997a). 11.The International Visitors Program began formally in 1996 under a coopera- tive agreement between BRI and the U.S.Agency for International Development. It was developed in response to growing international demand from governments, banks, NGOs, donors, and foundations for study tours to BRI that would enable in- stitutions in other countries to learn about the unit desa system Durmg 1996-98 HIID provided technical advice to BRI under the cooperative agreement, and I served as coordinator of the HIID project. 12. See BRI (1997a, pp. 13-14) for detailed discussion of the role of the unit busi- ness manager. 13. In addition to those discussed, two other kinds of branch employees serve the units directly: staff who perform general administrative and clerical functions related 362 The Microfinance Revolution: Lessons from Indonesia to supervision of the units, and reserve staff who fill in when there are temporary va- cancies at the units because of vacations, sick leave, trairnig leave, and the like. 14. Amounts lent were not available by year, so this figure is not inflation-adjusted. 15. Trainees are required to subnmit an extensive anonymous evaluation of each course and instructor. One topic on the evaluation questionnaire is how well the the- oretical explanations relate to the trainee's experiences; others range from assessment of the quality of the traimnig center adminstration to whether the classes are disturbed by noise.The commnents of the trainees are taken quite seriously by the training cen- ter and by BUD. 16. In practice, however, it is sometimes difficult to find appropriate places in the bank for the returning trainers. 17. See Steinwand (2001) for a discussion of BKD history; parts of this section draw from that source. 18.The early lumbung desas established in Central Java by De Wolff van Wester- rode in 1897-98 were based on the indigenous Javanese lumbung (granary) used to store paddy for consumption or seed; some lumbung provided loans to members of the village (see Suharto 1996, Steinwand 2001). 19. For discussion of the BKDs in the 1990s, see Chaves and Gonzalez-Vega 1996, BRI (1998a), Ravicz (1998), and Steinwand (2001). 20. For discussion of the P4K program, see International Fund for Agricultural Development (1996); Seibel and Parhusip (1998); McGuire, Conroy, andThapa (1998); and Ravicz (1998). 21.As noted in chapter 12, the KUPEDES annual effective interest rate of 32 per- cent a month would be higher if the prompt payment incentive (0.5 percent a month for six months or for the life of the loan, whichever is shorter) were included. But the KUPEDES prompt payment incentive maximum of 3 percent of the loan in a six- month period is much smaller than the forced savings requirements of P4K (and the BKDs). 22. It is xvorth noting here that where local officials or borrowing groups that on- lend set interest rates for loans (such as the BKDs' 131 percent annual effective inter- est rate and the 154 percent annual effective interest rate set for their members by P4K groups; Ravicz 1998, p. 67), there is a tendency for the interest rates to be set much higher than those set by efficient, profitable mucrofinance institutions (such as the KU- PEDES annual effective interest rate of 32 percent for prompt payers for most loans). This pattern occurs iti other countries as well. 23. Much of this section is taken from Sugianto and Robinson (1998). Because the death of Sugianto, BRI's long-term managing director responsible for the unit desa system, unfortunately preceded the final updating and editing of this chapter, respon- sibility for the opimons expressed here is entirely mune. 24.The subsidy dependence index (SDI) measures the percentage increase in the average yield obtained on the loan portfolio needed to compensate for the elimina- tion of all subsidies in a financial institution (see box 2.4 for discussion of the SDI). Institutional Development for Large-Scale Sustainable Microfinance The Transformaton of the Unit Desas, 1984-96 363 Commercial Microfinance in Indonesia: Stability in Crisis, 1997 to Mid-2001 The Indonesian crisis that began in 1997, its development, and its complex political, economic, and financial roots were discussed in chapter 8. The information in that chapter is essential for understanding the reasons for the failure of the financial system and the subsequent bank and corporate debt restructuring efforts discussed in this chapter. The focus in this chapter is on a dramatic contrast that became evident during the crisis-between commercial microbanking institutions, represented here by Bank Rakyat Indonesia's (BRI's) unit desas, the village- owned Badan Kredit Desas (BKDs), and the private Bank Dagang Bali (BDB), all of which remained stable and profitable, and the Indonesian financial system, which col- lapsed. The analysis explores the four years between the 364 start of the crisis in July 1997 and President Megawati Soekarno- putri's assumption of office inJuly 2001. For reasons that were primarily political rather than economic, the East Asian crisis hit Indonesia the hardest of any country in the region. In June 1997, the month before the crisis began, the In- donesian rupiah was valued at 2,450 to the U.S. dollar; in June 1998 it was 14,900 to the dollar. By the end of 1998 the value of the ru- piah had recovered to 8,025 to the dollar, but it later declined again (see the appendix for monthly exchange rates). In late 1997 and in 1998 the currency devaluation was followed by sharp increases in A dramatic contrast interest rates and inflation (which reached 78 percent at its highest point in 1998), then by steep contraction in GDP. Rising social dis- became evident order and severe political instability forced the resignation of Pres- during the crisis. The ident Soeharto in May 1998, after 32 years in office. With the fall of the rupiah, massive foreign debt-most of it un- Indonesian financial hedged and much larger than had been previously recognized-re- sulted in unmanageable debt burdens. Foreign debt in commercial system collapsed, but banks reached about three times their equity. By 1998 about 70 per- cent of bank loans were in default.The private sector's foreign debt commercial was more than twice the government's. Corporations went bank- microfinance rupt, and the financial system melted down. In mid-2000 Indonesian corporate debt was estimated at about remained stable and $120 billion, 72 percent of which was denominated in foreign cur- rencies.A large portion of the debt was concentrated in loans made proftable to a small number of Soeharto family members and close associates. By early 1999 the Indonesian banking system had a negative net worth of nearly 200 trillion rupiah ($23 billion). Banks had essen- tially ceased lending, and deposits declined as depositors lost con- fidence in the banking system. The Economist (18 July 1998) commented: "Even with the fierce competition from its neighbors, Indonesia would probably walk away with the prize for Asia's most desperate banking system." The International Monetary Fund (IMF), supported by the World Bank, Asian Development Bank, and other donors, created an emergency $42.3 billion soft loan fund for Indonesia. The IMF agreements with Indonesia stipulate that disbursement of these funds is contingent on tighter fiscal and monetary policies and a va- riety of financial and other reforms. But, as widely expected in In- Commercial Microfinance in Indonesia Stability in Crisis, 1997 to Mid-2001 365 donesia, implementation problems arose with some of the reforms. As a result the IMF postponed disbursements on a number of oc- casions. Nevertheless, under the program with the IMF, the gov- ernment of Indonesia made important and timely achievements in macroeconomic stabilization and other areas. As part of the IMF program, the Indonesian Bank Restructur- ing Agency (IBRA) was created to restructure and recapitalize the banking industry; the agency was also given a major role in corpo- rate debt restructuring. IBRA has made some progress. But formi- dable obstacles-deeply imbedded in the financial system and in By early 1999 Indonesian corporate culture-continue to impede both bank re- Indonesia's banking structuring and corporate debt recovery.The problems are greatly ex- acerbated by Indonesia's weak and corrupt legal and judicial systems. system had a When President Soeharto resigned in May 1998, he was succeeded negative net worth by his vice president, BJ. Habibie. Responding to public demand, Habibie made a number of political reforms and called for general Of nearly $23 elections. But he was too closely connected to Soeharto to be re- elected. Habibie was succeeded by Abdurrahman Wahid, who was billion elected in 1999 but impeached in 2001-essentially for not being able to govern the country.Wahid was succeeded by his vice pres- ident, Megawati Soekarnoputri (the daughter of Sukarno; see chap- ter 8), who took office in July 2001. Presidents Habibie andWahid both made some long-overdue re- forms. But Habibie would not, andWahid could not, address many of Indonesia's most urgent needs-such as making hard decisions on the economy, cracking down on corruption, bringing the Soe- hartos to justice, and building legal,judicial, financial, and other in- stitutions.As of late 2001, President Megawati appeared to have the political capital required to begin to address these exceptionally dif- ficult challenges. But it is unknown whether her government wil exercise the political will required to make the needed reforms. After discussing the financial crisis and the attempts to restruc- ture the banking system and corporate debt, this chapter shifts its focus to commercial microbanking.The performance, during the cri- sis, of three major commercial microbanking institutions is exam- ined. BRI's unit desa system is the developing world's largest financially self-sufficient commercial microbanking system, the BKDs are its oldest microfinance institutions, and BDB is the old- est private commercial bank specializing in microfinance.The BKDs 366 The Microfinance Revolution: Lessons from Indonesia and BDB are considered first; the chapter then turns to unit desa performance from 1996-2000, with some additional information for 2001 (added when the chapter was in press). In stunning contrast to the banking sector generally (including BRI's other divisions), the unit desa system maintained its wide outreach and high repay- ment rate during the crisis, remaining profitable throughout.The re- view of unit desa performance during this period is followed by an overall analysis of BRI's microbanking system at the end of 2000. The three institutions discussed here are among Indonesia's-and the world's-best commercial microfinance institutions. Not all In- Formidable donesian microfinance institutions fared as well during the crisis. But the strengths of commercial microbanking are not confined to the obstacles-deeply three institutions discussed in this chapter. Steinwand (2001) shows that the commercially oriented Bank Perkreditan Rakyat (BPRs, or imbedded in the People's Credit Banks) also generally performed well during the cri- financial and sis (chapter 9). But the purpose here is not to discuss all microfi- nance in Indonesia. It is to show that mature, well-managed judicial systems and commercial microfinance institutions of quite different types can serve large numbers of savers and borrowers-remaining stable and prof- in corporate itable even as their country's financial system collapses. The chapter concludes with an analysis of the reasons for the sta- culture-impede bility of commercial microbanking during the Indonesian crisis. bank restructuring Among many others, these include: and corporate debt * Timely macroeconomic stabilization measures were undertaken by the government, with the support of the IMF and other recovery donors. * Most microbanking clients operate in the domestic economy and were not directly affected by the currency crisis. * The government, with help from the IMF and other donors, quickly instituted emergency antipoverty programs, supplying food, creating employment, and keeping children in school.These programs, along with the stabilization measures, meant that most of the economically active poor stayed economically active-and were able to make use of commercial microfinance services dur- ing the crisis. * The microfinance institutions discussed here had mobilized sub- stantial amounts of public savings (or in the case of the BKDs, had Commercial Microfinance in Indonesia Stability in Crisis, 1997 to Mid-2001 367 retained earnings deposited at BRI). All were liquid throughout the crisis, and their lending was not capital constrained. * Borrowers greatly valued the option to reborrow in times of dif- ficulty-and thus made loan repayment a high priority. And savers tried, where possible, to save more and consume less. With giant banks * Savers valued the security, convenience, liquidity, confidentiality, products, and services offered by these institutions. All these fea- (including BRI) and tures were critical during the crisis. Some savers moved their ac- counts from failing banks to the unit desas, BDB, or the BPRs. * Interest rates for savings, especlally for fixed deposits, were high all around them, in 1998 and part of 1999 (at the unit desas they peaked at 60 per- cent in May 1998).These rates helped savers counter somewhat BRI unit desas the effects of rising inflation and underemployment. made and collected * Some high-level government officials who understood the im- portance of commercial microfinance for the economy contin- small loans, ued their strong support for the institutions providing it. * The institutions discussed had built a record of trust.They had good, increased savings, committed management, friendly and motivated staff, and prod- ucts appropriate for microfinance clients. Clients remained loyal because they valued the services they received. There are important lessons from the Indonesian experience dur- ing the crisis-for Indonesia, for other countries, and for the mi- crofinance industry. 368 The Microfinance Revolution: Lessons from Indonesia Among the East Asian countries affected by the 1997 crisis, each had some- what different reasons for the steep currency devaluations, plunging stocks, loss- es of foreign investment, rising prices, and growing unemployment that hit the region that year-although to varying degrees in different countries. In Indonesia even a long record of solid economic management with impressive achieve- ments could not alleviate growing concerns about rapidly increasing, large-scale corruption at high levels-and about the country's political stability. As discussed in chapter 8, the roots of the Indonesian crisis are complex, with multilayered and intertwined components (box 15.1).What was perceived in July 1997 as a currency crisis turned out to be a major political, economic, and financial crisis-and perhaps most important, a crisis of confidence. The extent of Indonesia's economic downturn in 1997-98 was devastat- Withfew new bank ing. The Asian Development Bank (ADB news release 32/99, 19 April 1999) reported that "the economy plunged into a vicious cycle of rising inflation, falling loans, and with exchange rates and increasing interest rates." And as noted in chapter 8, the World Bank (1998b, p. 1) commented that "no country in recent history, let alone one investors wary of the size of Indonesia, has ever suffered such a dramatic reversal of fortune." Yet an outsider with no knowledge of Indonesia who examined the per- Indonesia s financial formance ofBRI's unit desas during 1984-2001 would have no way of know- ing that beginning in 1997 the country had suddenly faced severe political and legal systems, instability, major econonmc hardship, and financial chaos.With giant banks and corporations failing all around them, the unit desas continued to increase their the country has seen outreach, collect loans, mobilize savings, and earn profits. Other commercial microbanking institutions withstood the crisis as well. As a long-time manag- little new economic er at Bank Dagang Bali observed while discussing the stability of commercial ricrofinance during the collapse of the banking system: "The proverb is right. activity Big storms fell mighty trees, but the grass still grows?" This chapter first examines the financial aspects of the crisis, the collapse of the banking sector, and the government's efforts at bank and corporate debt restructuring. That discussion provides the background for the subsequent analysis of the remarkable stability of commercial microbanking throughout this period. Indonesia's Financial Crisis By the beginning of 1999 the Indonesian banking system had a negative net worth of nearly 200 trillion rupiah ($23 billion). Although banks had essen- tially ceased lending in late 1997, they continued to incur losses. Loan repay- ment remained poor, and deposits declined.With few new bank loans, and with investors wary of Indonesia's financial and legal systems, the country has seen little new economic activity-which has impeded efforts to restore sustainable, broadly based economic growth. For reasons discussed in chapter 8 (see also box 15.1), the crisis in Indonesia was the worst among all the affected East Asian countries. In late 1997 and in Commercial Microfinance in Indonesia Siability in Crisis, 1997 to Mid-2001 369 Box 15.1 Excerpts from Mark Baird's -Corporate Restructuring in Indonesia" The structural problems that led to the 1997 financial crisis existed in varying degrees in all of the affected East Asian countries But they were more severe in Indonesia, which was consequently hit harder than the other countries in the region The problems included * a high dependence on foreign borrowing by Indonesian corporates * little reliance by corporates on equity finance * the absence of an effective legal and ludicial system to settle contractual disputes and apply a viable bankruptcy regime * a heavily concentrated family corporate ownership structure which depended for its success on non-transparent relationships with government and with banks * poor corporate governance, particularly concerning disclosure and enforcement * limited competition and prevalence of anti-competitive behavior * an over-extended state-owned enterprise sector that imposed special demands and received special favors from government while crowding out private investment Source Baird 2000, p 1 1998 the rupiah plunged against the U.S. dollar (figure 15.1; see the appendix for more details). By the end of 1998 the rupiah had recovered to 8,025 to the dollar, and by the end of 1999 to 7,085. But by the end of 2000 the rupiah had depreciated again to 9,595 to the dollar. And in April 2001 the rupiah contin- ued falling as political instability increased and proceedings for President Ab- durrahmanWahid's impeachment moved forward.The rupiah, which was at 11,675 to the dollar at the end ofApril, remained above 11,000 to the dollar until Pres- ident Megawati took office in late July.Within a few weeks the rupiah appreci- ated to about 8,500 to the dollar, but by October it had fallen to more than 10,000 to the dollar, where it remained in November and December 2001. In 1998 annual inflation was 53.4 percent-a sharp contrast to the previ- ous decade, when it had averaged less than 10 percent. But inflation fell to 20.5 percent in 1999, and in 2000 it was down to 3.7 percent. Annual inflation in 2001 (through November) was 11.4 percent. The consumer price index rose throughout 1997-2000 (figure 15.2), but consumer prices rose less sharply and showed far less volatility than the exchange rate-a point discussed later be- cause of its significance for the stability of comnmercial microbanking during this period. Most Indonesian corporations were technically bankrupt by early 1998. By mid-1998 the combined effects of political instability, the rupiah devaluation, high interest rates, capital outflows, bank failures, and growing unemployment caused a sharp contraction in the economy. GDP growth, which had averaged nearly 8 percent a year for more than a decade, was just 4.7 percent in 1997. And in 1998 GDP contracted by more than 13 percent.This dramatic down- turn of 18 percentage points in one year ranks among the world's most severe economic collapses since the 1930s. But in 1999 GDP growth was 0.3 percent, and in 2000 it reached 5.2 percent.' 370 The Microfinance Revolution: Lessons from Indonesia Figure 15.1 Exchange rate in Indonesia, December 1996-December 2000 15,000 Number of rupiah to the U S dollar 12,000 9,000 6,000 3,000 December June December June December June December June December 1996 1997 1997 1998 1998 1999 1999 2000 2000 Source IMF, International Financial Statistics (March 2001) Figure 15.2 Consumer price index in Indonesia, December 1996-December 2000 250 Index 1995=100 200 150 100 December June December June December June December June December 1996 1997 1997 1998 1998 1999 1999 2000 2000 Source IMF, International Financial Statistics (May 2001) The International Monetary Fund (IMF), supported by the World Bank, Asian Development Bank, and a number of bilateral contributors, created an emergency $42.3 billion package of soft loans for Indonesia (see chapter 8). The funds were to be disbursed in installments contingent on tighter fiscal and monetary policies, financial reforms, and various sector-specific structural re- Commercial Microfinance in Indonesia Stability in Crisis, 1997 to Mid-2001 371 forms.The parties reached agreement on a first program on 31 October 1997, but multiple difficulties ensued on both sides. On 15 January 1998 the gov- ernment of Indonesia signed a second agreement with the IMF But President Soeharto resisted the program's measures to remove subsidies, monopolies, and tax and credit privileges from many of the huge government projects controlled by the Soeharto family and its friends-which resulted in delays in disburse- ments and slow implementation of important aspects of the program. Even after Soeharto's resignation in May 1998, President Habibie did not carry out many of the reforms stipulated in the agreements. And President Wahid, who suc- ceeded him, was unable to do so. The slow pace of compliance with parts of IMF-government of Indonesia agreements led to further postponements of some In 1998 about 70 IMF disbursements during the period between 1998 and the impeachment of President Wahid in July 2001. percent of bank Nevertheless, under the agreements with the IMF significant progress was made on macroeconomic stabilization-although the stability remains fragile. loans were in Despite many severe problems, the economy has grown. default. Depositors In the three and one-half years since the fall of Suharto, three successive presidents have failed to overcome parliamentary lost confidence in the bickering and push through controversial reform measures, reign in corruption or restore investor confidence ... Indonesia's banking system. currency is sliding, inflation is rising and government finances are a mess. But its economy still managed to grow faster than Deposits declined. expected in the third quarter [of 2001]... 3.5 percent compared with the third quarter of 2000... demonstrating what theWorld Bank assetsfell Bank calls the country's ability to muddle through the global slowdown and its own political confusion. -Wayne Arnold, The New York Times, 16 November 2001 However, progress on structural reforms, especially bank and corporate debt restructuring, has been slow. Lack of political leadership, weak legal and judi- cial systems, opaque corporate and banking structures, rampant corruption at high levels, and the banking system's lack of adherence to international pru- dential standards have made bank restructuring extraordinarily difficult. Although institutions have been established to restructure corporate debt, substantial obstacles continue to impede loan recovery. Many of the largest debtors, especially the Soeharto family and its friends, have not felt compelled to repay their loans-foreign or domestic. The Indonesian Bank Restructur- ing Agency (IBRA) has had some success in its extremely difficult charge. But both bank and corporate debt restructuring continue to encounter formida- ble political difficulties for reasons that are deeply embedded in the financial system and Indonesian corporate culture. Overall, the reforms made with support from the IMF and other donors helped Indonesia stabilize its economy and recover from the depths of the cri- sis. But as discussed in chapter 8, broadly based economic recovery will require 372 The Microfinance Revolution: Lessons from Indonesia implementing politically difficult financial reforms and developing indepen- dent and professional legislative, legal,judicial, administrative, and financial in- stitutions. The collapse of the banking system In late 1997 and in 1998 the free fall of the rupiah, the poor quality of the loan portfolios held by many banks, large foreign debt, and corrupt and poor bank management and supervision resulted in massive bank failures and a general col- lapse of the financial system.2 Thefirst bank closures. At the IMF's insistence, the government closed 16 banks on 1 November 1997 in an effort to signal that major reforms were under In 1997-98, Bank way. There is little doubt that many banks needed to be closed. But the way these first 16 banks were closed is widely believed to have been a mistake (see Indonesia issued Radelet and Sachs 1998; Radelet 1999; Baker 1999; and Kenward 2000).The closures were carried out suddenly and without considering that no deposit $13 billion in insurance was in place. Depositors, who were allowed to withdraw a maximum of 20 million rupiah ($5,482 in November 1997), were told that they would liquidity credits to have to wait for the banks' assets to be liquidated before learning whether ad- ditional deposits could be recovered from their accounts. banks. Many were The result was a severe crisis of confidence in the financial system and a run on banks, especially private banks. The country was thus faced simultane- owned by the ously with precipitous currency devaluation and bank runs. In addition, cred- itors of the closed banks were uncertain whether they would recover their Soeharto family and loans-and creditors of banks still open began callng in loans.To keep banks open, Bank Indonesia issued massive liquidity credits-about $13 billion be- friends, and little tween November 1997 and June 1998 (Radelet 1999, p. 9). But a substantial portion of these credits went to banks owned by Soeharto's family and friends, repayment is and many of the credits have still not been repaid. One of the banks closed in November 1997 was owned by one of Soeharto's sons. Within weeks he re- expected ceived permission to open a new bank with a different name-but in the same building and with the same staff as the bank that had been closed! Serious mistakes were made on both sides.The IMF insisted on immedi- ate bank closures as a demonstration effect rather than waiting for a well-de- signed reform strategy that would include bank closures but also ensure depositors of prompt settlement.When implemented by the government, the IMF-imposed plan generated panic. And despite the need for IMF assistance, Soeharto maintained a business-as-usual-in-Indonesia approach-which was unacceptable to the IMF, other donors, and many Indonesians. InJanuary 1998 the government guaranteed all bank deposits, which eased the runs on banks. But deposits continued to be moved from private to state- owned and foreign banks. Nonperforming loans. Nonperforming loans (loans in default) have become an exceptionally difficult problem.3 With foreign exchange obligations soar- Commercial Microfinance in Indonesia Stability in Crisis, 1997 to Mid-2001 373 ing as the value of the rupiah fell, nonperforming loans jumped from about 9.5 trillion rupiah ($4 billion) at the end of 1996 to more than 300 trillion rupi- ah ($37 billion) at the end of 1998. In 1998 about 70 percent of bank loans were nonperforming, and interest owed on another 10-15 percent had been only partly paid (International Herald Tribune, 10 April 1999). As borrowers de- faulted on loans, the value of bank assets fell. At the same time, depositors lost confidence in the banking system and deposits declined. The extent of the nonperforming loans can be understood only in the con- text of the roots of Indonesia's crisis. As would happen anywhere during such a crisis, many borrowers were unable to repay their loans because of the cur- rency devaluation, which greatly increased their debts in rupiah terms. But in Nonperforming this case the basic problems were outgrowths of the banking practices of Soe- harto's Indonesia. loans jumpedfrom In state-owned banks a large portion of the value of nonperforming loans is accounted for by a few politically powerful borrowers who were able to bor- $4 billion in 1996 row large amounts from banks that did not evaluate the loans objectively. Re- payment has not been a priority for these borrowers.As Eko Santoso Budianto, to $3 7 billion in then IBRA's deputy chairman, commented in April 1999: "It is an open secret that most of the corporate debtors, particularly at state banks, are politically well 1998. At state connected businesspeople, including siblings of former President Soeharto and the family's associates" (Tempo, 21 April 1999). IBRA's chairman at that time, banks in 1999, the Glenn S.Yusuf, said that the 20 largest debtors owed state banks more than 60 trillion rupiah (about $7 billion) in nonperforming loans (Tempo, 20 April 1999). 20 largest defaulting The difficulties in recovering nonperforming loans from politically fa- vored borrowers arise from more than just their increased debt and unwlllingness debtors owed more to repay.The problems also stem from the poor quality of the borrowers' loan collateral, which is typically of much lower value than the outstanding debts. than $7 billion This shortfall is caused partly by the decreased value of the land and buildings that constitute much of the collateral. But it is also a result of the greatly ex- aggerated initial statements of the value of the collateral (by overstating the value of their collateral, well-connected borrowers were able to obtain larger loans). And among private banks, few had observed the regulation limiting loans to bank-affiliated companies and their owners to 20 percent of the banks' loan portfolios. Most of the nonperforming loans of private banks are held by their affiliated companies and owners of these companies.As a result repayment prob- lems in conglomerate-owned banks have been compounded by the debts and bankruptcies of the conglomerates' companies. In many cases where affiliates' reserves were low, banks wrote off the debts, depleting their capital. In addi- tion, some of the businesses financed by both state and private banks had never been feasible, and some are believed to have been fictitious. But it was not only banks and their debtors who contributed to Indone- sia's corruption-ridden financial system and the gargantuan problem of non- performing loans. Both Bank Indonesia and IBRA became involved in major scandals in 1999. A state audit faulted Bank Indonesia's accounting of the bil- lions of dollars in emergency liquidity loans made in 1997 and 1998 to sup- 374 The Microfinance Revolution: Lessons from Indonesia port failing commercial banks-many of which were owned by the Soeharto family and the conglomerates. As noted in chapter 8, the Supreme Audit Agency estimated in 2000 that 95 percent of these loans may never be repaid. "If [the deficit] were charged to the [central] bank, it would go bankrupt" (Asia- week, 22 December 2000). Meanwhile, IBRA became embroiled in a highly publicized corruption scan- dal after approving large payments to Bank Bali, a private commercial bank.4 The payments ended up in a company controlled by Golkar, Indonesia's for- mer ruling political party. Reaching paralysis. The government imposed a tight monetary policy in 1998. Many banks that already had liquidity problems faced a severe shortage Most of the of liquidity. Banks offered high interest rates to attract depositors-often re- sulting in negative interest rate spreads. At one point the annual interest rate nonpeforming loans offered by some banks for one-month fixed deposit accounts reached more than 70 percent. Such rates attracted many depositors, including some business ofprivate banks are groups that deposited working capital, hoping to cut their losses in a poor mar- ket. But few companies or individuals were qualified to borrow at interest rates held by their that would allow the banks a positive interest rate spread. Thus banks were essentially paralyzed during 1998, with annual effective affiliated companies, interest rates on loans often above 40 percent and few qualified takers. A neg- ative interest rate spread led to negative interest rate revenue (meaning banks many of which are paid out more in interest than they took in) for more than 40 banks, includ- ing some of the largest state and private banks. Overall, 49 banks suffered loss- bankrupt es in 1998. Of these, 36 saw their assets drop. BRI was no exception.Although BRI's microbanking system remnained prof- itable throughout, BRI-like other state banks (and many other banks)-was technically bankrupt. In 1998 BRI's branch network, with much of the value of its loan portfolio in large corporate loans, lost $3.4 billion against the bank's capital base of only $215 million.The profits of the unit desas had been cov- ering the losses of the rest of BRI for years (box 15.2). But in 1998 the unit desas' continuing profits-remarkable as they were during such a severe cri- sis-were insignificant compared with the huge losses of BRI's branches. By 1998 it became clear that many of Indonesia's banks would have to be liquidated or merged. Many would also have to be recapitalized. Some would have to be taken over by the government and then reprivatized. But it was widely recog- nized that such efforts would be very expensive-and very difficult politically. Bank reforms and corporate debt restructuring In August 1998 the government announced reforms designed to rebuild and to restore public confidence in what was widely being called Indonesia's "wrecked banking system" (Jakarta Post, 24 August 1998). The planned reforms included restructuring banks, improving bank regulations and laws, and upgrading and en- forcing prudential regulations. In addition, corporate debt restructuring-a crit- ical part of building a viable banking system-would be given high priority. Commercial Microfinance in Indonesia Stability in Crisis, 1997 lo Mid-2001 375 Box 1 ~ Excerpts from Klaus Maurer and Hans Dieter Seibel's "Agricultural Development Bank Reform: The Case of the Unit Banking System of Bank Rakyat Indonesia (BRI)" Profitability of BRI units versus BRI branches (net income in US$ million) 1993 1994 1995 1996 1997 1998 1999 2000 BRI (consolidated before tax) 57 67 111 145 30 -3,308 -243 n a Branch network -10 -54 -64 -33 -60 -3,397 -411 n a Unit banking system 66 121 174 178 90 89 168 121 In the 1990s, the Units and the microbanking business were generating the profits to sus- tain the rest of BRI! The rest of BRI, I e the branch network, has been running losses since several years-a fact well hidden in consolidated figures It was solely due to the Units' performance that, on a consolidated basis, BRI posted a positive net income until 1997 Needless to mention that the bill was paid by the Unit customers, the micro borrow- ers and savers. The bomb exploded in 1998 when BRI was hit by the financial crisis 56% of the branches' loan portfolio-large corporate loans for the most part-turned sour and had to be written off. This resulted in a loss of US$ 3 3 billion against a capital base of only US$ 215 million In 1998, the consolidated BRI-including the Unit Banking Systeml-was technically insolvent and effectively bankrupt, like many other banks in Indonesia at the time. Being an integral part of BRI suddenly turned into a major risk for the [profitable] BRI Unit Banking System, the risk of being dragged into bankruptcy by the ailing mother organ- ization However, the government stepped in to recapitalize BRI, the non-performing loans were transferred to the Indonesian Bank Restructuring Agency (IBRA) ..When BRI was severely affected by the financial crisis, the very existence and success of the Unit Bank- ing System convinced policy makers to save and recapitalize BRI Source Maurer and Seibel forthcoming, p 15, draft paper prepared for the International Fund for Agricultural Development Bank restructuring and recapitalization. IBRA was created in 1998 to re- organize the banking system, classify all banks, and manage bank liquidations, mergers, restructurings, and recapitalizations.5 In addition, IBRA's asset man- agement unit was established to administer the banking system's nonperform- ing loans. (As noted, IBRA also plays a major role in corporate debt restructuring.) In October 1998 the government announced that the strategy for bank re- structuring would focus on recapitalizing potentially viable banks, liquidating insolvent banks, merging state banks, and recovering liquidity support. Foreign accounting firms were brought in to conduct due diligence appraisals of all In- donesian banks. The March 1999 bank rehabilitation plan placed Indonesia's 128 private domestic banks into three categories. Category A, for which 74 banks quali- fied, required a capital adequacy ratio-the ratio of equity capital to risk- weighted assets-above 4 percent.The government determined that banks in this category met Bank Indonesia's minimum capital adequacy standards and did not require public assistance. Bank Indonesia then investigated category A banks to determine whether, based on technical qualifications and moral stan- dards, their owners and managers were fit and proper to continue in their po- 376 The Microfinance Revolution: Lessons from Indonesia sitions. However, not all category A banks had good asset structures. Some had the bulk of their assets in Bank Indonesia certificates, and much of the rest in nonperforming loans. Nine category B banks, defined as those with capital adequacy ratios be- tween -25 and 4 percent, were declared eligible for recapitalhzation.To receive government funds for recapitalization, the owners of these banks had to pro- vide cash for at least 20 percent of the recapitalization. In addition, the banks' owners and managers had to pass the fit and proper tests.The government also took over seven category B banks because their large size and extensive branch networks were required for the payments system to function, and because it was thought that they would be usef:ul for the country's future banking development. Most category C banks, defined as those with capital adequacy ratios below In 1998 BRI lost -25 percent, were closed, and their assets and liabilities were transferred to IBRA's asset management unit. Some category C banks were merged into new banks, $3.3 billion against with their nonperforming loans transferred to the IBRA unit. The first stage of bank reforms-between late 1997 and mid-1999-con- a capital base of centrated on closing failing banks and nationalizing selected banks (which were taken over by IBRA).Thus 66 private banks with a 14 percent market share $215 million. The were closed. But unlike the closing of the first 16 banks in 1997, later bank clo- sures were carried out relatively smoothly as part of a wider restructuring plan. unit desas remained Another 13 private banks, with a 21 percent market share, were nationalized- including the country's largest private bank, Bank Central Asia, owned by profitable but could close associates of former President Soeharto. The second stage of bank reforms, beginning in 1999, focused on bank re- no longer cover the capitalizations and mergers, and on corporate debt restructuring. Four of the seven state banks-Ekspor-lmpor Indonesia, Bank Dagang Negara, Bank Bumi losses of the rest of Daya, and Bank Pembangunan Indonesia, the State Development Bank, with a combined 25 percent market share-were merged in 1999 to form a new BRI bank called Bank Mandiri (in Indonesian the word mandiri means "to stand alone").The other three state banks-BRI, Bank Negara Indonesia, and Bank Tabungan Negara-were restructured and recapitalized to varying degrees.The bad assets of all state banks were shifted to IBRA'a asset management unit.The main reason BRI was recapitalized and restructured, rather than merged into Bank Mandiri, was its unit desa system. But merging four moribund state banks into one, even with high-level man- agement changes, does not necessarily transform the bank's corporate culture into one of professionalism and accountability. Many in Indonesia are skepti- cal about the future of Bank Mandiri, now the country's largest bank. If the newly capitalized bank continues to operate in the old banking culture, if the banks merged to form Bank Mandcri bring with them their political entan- glements, if the new bank's middle managers (who come from the four merged banks) conduct business as usual, or if the bank is required to provide subsi- dized loans, Bank Mandiri may soon acquire a new set of nonperforming loans. Another problem in the banking industry-one underlying many of the others-is that Indonesian accounting standards generally do not conform with Commercial Microfinance in Indonesia Stability in Crisis, 1997 to Mid-2001 377 internationally accepted accounting principles.The main reason this problem persists is the lack of transparency and accountability endenlc in Indonesia's financial system and corporate culture. For example, as part of the recapitalization process public accountants sub- mitted audits of Indonesian banks that set asset values based on financial re- ports provided by the banks. But IBRA then assigned two international finance companies, Lehman Brothers and J.P. Morgan, to reaudit the banks. The two firms used a different audit method, due diligence, which evaluates assets based on their real prices and cross-checks these prices at marketplaces. Due diligence produced dramatically different results, indicating that "the real value of certain banks is only about 20 percent of the value recognized Estimates of the cost by public accountants" (Kwik Kian Gie, then coordinating minister for the economy, finance, and industry, in an interview with theJakarta Post, 20 April of Indonesian bank 1999). Estimates of the costs of Indonesian bank recapitalization are staggering. restructuring range In November 2000 Barclays Bank estimated that the cost would be more than 50 percent of GDP (Barclays Bank Country Report, 28 November 2000). Some from 50 to 80 estimates have been higher. percent of GDP The.. recapitalization program is probably the most expensive in banking history anywhere-and almost certainly the most challenging.The total cost of recapitalizing Indonesia's banks will be about $90 bilon, which is around 80 percent of the nation's gross domestic product. . .Lehman Brothers Asia estimates that that the Mandiri recapitalization will consume nearly one- quarter of the cost needed to fix all of Indonesia's banks. -Tim Healy and Tom McCawley, Asiaweek, 13 August 1999 Reasons for the high costs of recapitalizing Indonesia's banks include the lack of transparency in the banks (which originally masked their real condi- tions), negative interest margins, and poor loan recovery. By 2000 more than 70 percent of the banking system was under gov- ernment control. Although the government plans to reprivatize banks taken over by IBRA, progress has been slow. IBRA has the authority to sell assets taken over from defaulting debtors. But attempts to sell the banks have faced numerous difficulties-chief among them that the banks' powerful former own- ers still have some ownership rights and that the legal and judicial systems are generally weak, corrupt, and politicized. The first major test case will be the government's long-delayed sale of a 51 percent share of Bank Central Asia (BCA), Indonesia's largest private bank.The sale is planned for early 2002.A timely, orderly sale is considered crucial for encouraging foreign investment and for meeting the terms of the government's agreements with the IMF. But it is widely believed that BCA's former owners are trying to regain control of the bank. If IBRA and the government can finalize this sale, it wil have an important impact on Indonesia's financial system. 378 The Microfinance Revolution: Lessons from Indonesia There has been considerable progress in a relatively short period in estab- lishing and strengthening IBRA and in closing, merging, and restructuring In- donesia's banks. But there is still a long road ahead. Corporate debt restructuring. Restructuring Indonesia's massive corporate debt is essential for the country's economic recovery. It is also essential for In- donesians to regain confidence in their nation. But it is extremely difficult. Indonesian corporate debt in mid-2000 amounted to around US$120 billion: US$85 billion for large private corporations and $US35 billion for state-owned enterprises and SMEs [small and medium-size enterprises]. Of the total, 49 percent was Corporate debt in owed to foreign creditors and 72 percent was denominated in foreign currencies. Of the large corporate debt, three-quarters mid-2000 was is non-performing and in need of restructuring. -Baird 2000, p. 1 about $120 billion Corporate debt restructuring got off to a very slow start. But by 2000 some (72 percent in important gains had been made.The two main institutions implementing cor- porate debt restructuring are IBRA and theJakarta ImtiativeTask Force JITF). foreign currencies). Like many Indonesian institutions, IBRA was established without an ap- propriate governance structure. Following the Bank Bali scandal of 1999, Of this, $85 billion IBRA was audited in 2000 for the first time.The auditors found that IBRA's accounting standards were not in accordance with generally accepted ac- was owed by large counting principles and that some payment records could not be reconciled with the records of Bank Indonesia and the Ministry of Finance. private corporations A new IBRA governance structure was adopted in 2000, including ap- pointment of a governing board of independent professionals. [BRA's rules were modified to permit more flexibility in corporate restructuring, and legal pro- tection was provided for IBRA staff who implement its rules. IBRA turned over the collection of commercial loans (those of 5 billion to 50 billion rupi- ah) to banks and focused its efforts on the largest debtors. Large corporations with debts of more than 50 billion rupiah represent more than 75 percent of the value of nonperforming loans but only about 1 percent of debtors. "The new framework has started to deliver results. IBRA has now entered into re- structuring agreements or initiated legal action to resolve 70 percent of its loans to theTop 21 obligors" (Baird 2000,p. 2).6A recent high-level [BRA appointment has raised concerns, however, because it is widely believed that the official is closely connected with the Soeharto family A commercial court was established to handle debt settlements and bank- ruptcy proceedings. But IBRA has been hampered in its efforts to take legal action against uncooperative debtors with large nonperforming loans. The agency has received adverse rulings in many of its cases; such rulings are wide- ly thought to be the result of political influence on the judicial system. Only a small fraction of the debt has been recovered. Commercial Microfinance in Indonesia Stability in Crisis, 1997 to Mid-2001 379 Unlike IBRA, which is a large creditor with the authority to compel re- structuring and asset sales,JITF does not hold debt. It acts as a mediator be- tween private debtors and creditors. In 2000 JITF's mandate was strengthened in several ways, including referral of noncooperating negotiating parties for ac- tion by the attorney general.JITF is also empowered to offer tax and other ex- emptions as incentives to cooperating parties. In October 2000, 50 companies with a combined $10 billion in debt were under active JITF mediation, while 27 with $5 billion in debt had concluded binding restructuring agreements with JITF assistance. In addition, foreign banks had reached or nearly completed agree- ments with a number of large debtors (Baird 2000, p. 2). Overall, there has been progress in corporate loan restructuring at both IBRA Large corporations and JITE But noncompliance by debtors remains a major problem, and efforts to prosecute large debtors have generally failed. IBRA recently expanded its represent only 1 legal actions against noncooperating debtors, but progress will depend largely on whether recently initiated judicial reforms can be effectively implemented percent of debtors and on whether new reforms can be instituted. but accountfor more Challenges ahead Some significant advances have been made in recapitalizing and restructuring than 75 percent of banks-helped by the use of fit and proper tests, changes in bank management, new capital adequacy standards, increased transparency, the use of internation- the value of al firms to assist in transitions, and other aspects of the rehabilitation process. But severe problems remain in the financial system, especially in bank and nonperforming loans corporate debt restructuring and related areas.The chaUenges for the financial system center on its most fundamental needs, including: * A strong, professional central bank. A strong, accountable, independent central bank is vital for the rehabilitation of the financial system. Efforts are being made to strengthen Bank Indonesia but-like the judicial system-it has be- come part of the country's political battleground. Even with recent im- provements, uncertainties remain about the central bank's governance and its supervision capacity. * Basic reforms of the legal and judicial systems. Fundamental reforms are essen- tial for Indonesia's ineffective, politicized, and corrupt legal and judicial sys- tems. The scale of the problem is monumental. "Indonesia's justice system has been badly broken for a generation. Not long ago a senior lawyer inJakar- ta claimed that up to 90% of court cases are decided by bribes to judges, pros- ecutors, and other court officials" (Robert N. Hornick, The Asian Wall Street Journal, 24-30 December 2001). But as Hornick points out, there are steps that can be taken quickly and at httle cost that would begin to move the coun- try down the road ofjudicial and legal reform. Some reforms have already been instituted. Legal and judicial reform is crucial for financial sector re- form, as well as for the nation generally. * A new corporate culture. Indonesia's corporate culture needs a major overhaul. Many of the basic causes of Indonesia's recent financial crisis-and certain- 380 The Microfinance Revolution: Lessons from Indonesia ly the extent to which the country was affected-stem from aspects of its corporate culture (see box 15.1). Poor corporate governance, lack of ac- countability and transparency, prevalent anticompetitive behavior, and sweet- heart deals for favored conglomerates and state enterprises contributed to the financial collapse. The changes needed in this difficult area will require sustained, high-level pohtical will-and example. A strong, professional IBRA. IBRA has a gigantic task. It also has powerful en- emies. Can the newly restructured IBRA take on the Soehartos and the con- glomerates-in both bank and corporate debt restructuring-better than was possible under the old IBRA? An important part of the answer lies with the judiciary. But much also depends on whether IBRA can serve as a profes- sional, accountable, and effective institution-and on whether it will have a IBRA has received strong, united government supporting its efforts. IBRA's own governance will be an important factor. IBRA faces obstacles in reprivatizing banks be- adverse rulings in cause of delays and stonewalling by former bank owners. Moreover, only a small part of corporate debt has been repaid (and some will never be col- many of its legal lected).7 [BRA needs to improve and expedite loan restructuring and asset sales with the 1 percent of debtors who hold more than 75 percent of the cases against large value of nonperforming loans. In addition, debt restructuring and asset sales must be conducted in a fully transparent manner. debtors-a result Deep operational restructuring is needed not only to maxlmize widely attributed to the future viability of the enterprise, but also to minimize the future risks for the Government and the burden on Indonesia's political influence on taxpayers. IBRA will also be under intense public scrutiny in settling these large debts with its high-profile debtors and must the judicial system avoid the perception of insider deals, which represent a large reputational risk for the government. -Baird 2000, p. 2 The contest between IBRA and the Soeharto family and its friends is of the epic proportions seen in struggles in wayang performances (see chapter 8). As Warren Caragata (2001, p. 2) put it, "IBRA is at the epicenter of a strug- gle over Indonesia's economic soul." * A restructured banking system. Failing banks have been closed, selected private banks have been nationalized, some banks have been merged, and bank re- capitalization has proceeded. But much remains to be done.Though state banks vary, they generally have liquidity problems, high operating costs, little new lending, and in some cases problematic management-as well as ongoing state bank cultures. State banks are expected to restructure, raise earnings, and pri- vatize-but progress has been slow. Reprivatization of nationalized private banks has also been slow.Two problems that affect the entire banking system are In- donesian accounting standards, which require extensive reform, and bank su- pervision by Bank Indonesia, which needs considerable improvement. Progress has been made m both areas, but much more is needed. Commercial Microfinance in Indonesia Stability n Crisis, 1997 to Mid-2001 381 It is against this background of the Indonesian banking system since the crisis-with its 70 percent nonperforming loans, numerous bank closures, liq- uidity problems, high recapitalization costs, and loss of public confidence-that the analysis turns to commercial microbanking in Indonesia during the same period.There is a dramatic contrast here one of significance not only for In- donesia, but also for the microfinance industry globally. The Stability of Microbanking during the Crisis Microbanking in Indonesia provides an extraordinary record of stability in cri- Because their direct sis.This section considers how and why different kinds of commercial microfi- nance institutions maintained high outreach and continued profitability even as participation in the country's financial system collapsed around them.The discussion begins with the Badan Kredit Desa (BKDs)-the village-owned banks supervised by BRI (see imports or exports is chapter 14)-and the privately owned Bank Dagang Bali (BDB; see chapter 10). The focus then turns to a more extensive analysis of BRI's unit desas dur- limited, ing the same period. To highlight the stability of the units during the crisis, the appendix provides indicators of unit performance by month from December 1996 microbanking clients to December 2000, shown alongside exchange rates and consumer prices.The decline in the real value of the unit desas' assets, liablities, and profits can be more are affected more by meaningfully understood by adjusting current figures to changes in the consumer price index rather than the exchange rate. (This is also the case for other micro- changes in consumer finance institutions.) Because their direct participation in imports or exports is generally limited, microbanking clients are typically affected more by changes in prices than in domestic purchasing power as measured by the consumer price index than by changes in the exchange rate (farmers who cultivate export crops are an excep- exchange rates tion).The differences between the changes in these two variables have been im- portant in this context. For example, between the second quarter of 1997 and the end of 1998, during the height of the crisis, the consumer price index rose 90 percent-compared with a currency devaluation of 232 percent. Although the crisis affected the institutions in different ways, all three re- mained relatively stable, their clients generally remained loyal, none of the in- stitutions required recapitalization or subsidy, and all continued to earn profits. Still, some important differences among them are worth exploring to better understand the ways that commercial microfinance can be affected by a severe national crisis. The Badan Kredit Desa, 1997-98 The BKDs survived the crisis well.Their performance from 1996 through Au- gust 1998 (the latest data available) is shown in table 15.1.8 In August 1998 In- donesia was in deep crisis. But BKD outreach remained stable in terms of the number of outstanding loans and voluntary savings accounts, with some increase in loans and a slight decrease in savings accounts. In rupiah terms the value of outstanding loans and voluntary savings accounts increased gradually, though 382 The Microfinance Revolution: Lessons from Indonesia Performance indicators for the Table 1 5.1 Badan Kredit Desa (BKD), 1996 to August 1998 Indicator 1996 1997 August 1998 Number of active BKDs 4,806 4,806 4,806 Number of outstanding loans 779,599 813,306 807,603 Number of voluntary savers 251,028 259,732 242,761 Value of outstanding loans Millions of rupiah 104,363 115,478 120,848 Millions of U S dollars 43 8 248 10.9 Value of voluntary savings Millions of rupiah 3,359 3,573 4,080 Millions of U S dollars 1 4 0 8 0 4 BKD deposits at BRI Millions of rupiah 66,878 76,389 80,284 Millions of U S dollars 28 0 16 4 7 2 Income Millions of rupiah 36,630 40,972 28,846 Millions of US dollars 154 88 26 Expenses Millions of rupiah 19,099 21,437 15,083 Millions of U S dollars 8 0 46 1 4 Adjusted profitsa Millions of rupiah 14,106 16,487 13,532 Millions of U S. dollars 5 9 3 5 1 2 Adjusted return on assets (percent)b 8 7 9 2 7.3 Note The year-end rupiah exchange rate for one U S dollar was 2,383 in 1996, 4,650 in 1997, and 8,025 in 1998 a Profits - change in arrears past final due date + actual bad debt expenses b Adjusted profits / (total assets - arrears past final due date) Source BRI data their dollar values declined steeply. (The discussion in this section is based pri- marily on BRI data and Steinwand 2001.) In 1996 the average loan balance was 133,868 rupiah ($56)-5 percent of the country's $1,073 per capita GNP. The average voluntary savings account balance was 13,380 rupiah ($6), less than 1 percent of per capita GNP In Au- gust 1998 the average loan balance was 149,638 rupiah ($14), or 2 percent of the $680 per capita GNP. The average voluntary savings account was 16,807 rupiah ($1.52), or less than 1 percent of per capita GNP. Stated annual effective interest rates on loans typically continued at 20 per- cent for a 12-week period (although as discussed in chapter 14, the actual ef- fective interest rate is higher because of the forgone use of conipulsory savings). BKD voluntary savers, whose accounts are kept at BRI, received the bank's in- creased interest rates during this period (see below). Commercial Microfinance in Indonesia Stability in Crisis, 1997 to Mid-2001 383 As noted in chapter 14, it is difficult to determmne BKD arrears and de- faults because the BKDs write off bad debt only after five years (and not al- ways then). Moreover, like most People's Credit Banks (BPRs), the BKDs do not adequately provision for loan losses. But BKDs serve small communities, and most loans that fall into arrears are eventually collected. BRI data on BKD defaults-which are not adjusted for defaults that were not written off-show that defaults (loans more than 90 days overdue) made up 15 percent of the out- standing portfolio in 1996,19 percent in 1997, and 20 percent inAugust 1998. But for 1996 Ravicz (1998, p. 78) adjusted default and arrears rates by writing off all loans in default each year. This adjustment resulted in an estimated de- fault rate of about 5 percent for 1996. The BKDs were In rupiah terms, BKD deposits at BRI increased steadily during this period-from 66.9 billion rupiah ($28 million) in 1996 to 80.3 billion rupi- stable during the ah ($7.2 million) in August 1998. Profits, after adjusting for loan losses and pro- visioning, increased from 14.1 million rupiah in 1996 ($5.9 mnllion) to 16.5 crisis. In rupiah million rupiah ($3.5 million) in 1997.Through August 1998 adjusted profits of 13.5 million rupiah ($1.2 million) were on track to match or possibly ex- terms, deposits rose ceed the 1997 rupiah total by year's end. Returns on assets, adjusted for loan losses and provisioning, stayed high throughout-reaching 7.3 percent in Au- steadily and profits gust 1998 at the height of the financial crisis. Overall, the BKDs remained stable during 1997 and the first eight months increased. Average of 1998. In August 1998 the BKDs served more than 807,000 cients.As noted, reported loan defaults increased (but the BKDs continued their practice of not loan balance was 2 writing off bad debts for at least five years). However, the BKDs remained prof- itable because of their relatively high interest rate on loans, the higher inter- percent of GNP est rate they received on deposits at BRI in 1997-98, and their low operating costs. Bank Dagang Bali, 1996-2000 As a small private bank, BDB faced difficult times during the crisis. (This sec- tion is based on BDB data and discussions with BBB owners and managers.) The bank remained profitable each year from 1996-2001, as it had earlier (chap- ter 10). But profits fell sharply in 1998 and 1999, even in rupiah terms. In 2000, however, BDB saw a substantial increase in profits, reaching 7 billion rupiah ($725,000) before taxes (tables 15.2 and 15.3). Between 1996 and 2000 assets more than quadrupled in rupiah terms. In dollar terms, assets in 2000 ($143 million) were higher than in 1996 ($135 million). The growth in as- sets came primarily from deposits in the interbank money market and from securities and investments. BDB deposits held in other banks jumped from 33.8 billion rupiah ($14.2 million) in 1996 to 303.0 billion ($31.6 million) in 2000. The rupiah value of BDB's outstanding loans remained fairly stable from 1996 to 1999, though there was a sharp decline in their dollar value.As in other banks, BDB's annual effective interest rate rose in 1998-to more than 41 percent-then gradually decreased. In 2000 the rupiah value of outstanding loans 384 The Microfinance Revolution: Lessons from Indonesia Bank Dagang Bali profit and loss statement, 1996-2000 (millions of rupiah) Indicator 1996 1997 1998 1999 2000 Income 52,777 89,650 165,985 188,699 191,623 Operating income 52,676 89,553 161,872 181,429 191,264 Interest, commissions, and fees 50,571 80,623 154,184 178,458 186,771 Other 2,105 8,930 7,688 2,971 4,493 Nonoperating incomea 101 97 4,113 7,270 359 Expenses 49,439 86,097 163,367 187,344 184,666 Operating expenses 49,429 86,085 163,202 187,194 184,009 Interest 37,459 58,434 139,680 162,207 160,665 Overhead 11,970 27,651 23,522 24,987 23,344 Salaries 4,148 4,201 4,388 4,433 5,488 Other overhead 7,822 23,450 19,134 20,554 17,856 Nonoperating expensesa 10 12 165 150 657 Net profits 2,345 2,496 1,841 957 4,870 Pretax profits 3,338 3,553 2,618 1,355b 6,957 Taxes 993 1,057 777 398b 2,087 Note The year-end rupiah exchange rate for one U S dollarwas 2,383 in 1996, 4,650 in 1997, 8,025 in 1998, 7,085 in 1999, and 9,595 in 2000 See the appendix for more complete data on exchange rates and the consumer price index for 1996-2000 a Refers to income and expenses from nonbanking activities lsuch as the rental and sale of buildings) b These are the 1999 audited figures But a subsequent adjustment shows that in 1999 pretax profits were 539 million rupiah and taxes were 153 million rupiah Source Bank Dagang Bali data was more than twice the 1999 total. And the dollar value of the portfolio in- creased from $33 million in 1999 to $53 million in 2000. On the liabilhtles side, rupiah deposits at BDB nearly quadrupled between 1996 and 2000, from 270 billion rupiah ($113 million) in 1996 to 1.1 billion rupiah ($116 million) in 2000 (table 15.4).The biggest increase was in time deposits- in 2000 their rupiah value was more than five times that in 1996. But of BDB's 402,000 accounts of all types in 2000, 98 percent were passbook savings. At BDB annual interest on general passbook savings accounts was raised from 12.0 percent in 1996 to 13.5 percent in 1998, while interest on time de- posits jumped from 16.0-16.5 percent in 1996 to 26.9 percent in 1998. Inter- est rates declined gradually in 1999 and 2000. The number of outstanding loans decreased from 15,645 in 1996 to 10,417 in 2000 (table 15.5). Part of this decrease was a response to increasing arrears in 1997 and 1998, although repayment remained good given the circumn- stances. BDB had been accustomed since 1970 to on-time repayment rates of 98 percent or higher. But noncurrent loans with payments 30-90 days over- due rose from 1.2 percent in 1996 to 5.7 percent in 1998. Doubtful loans with payments 91-180 days overdue rose from 0.7 percent to 1.8 percent, and bad debt (loans with payments more than 180 days overdue) increased from 0.2 per- Commercial Microfinance im Indonesia Stability in Crisis, 1997 to Mid-2001 385 Table 15.31 Bank Dagang Bali balance sheet, 1998-2000 (millions of rupiah) Indicator 1996 1997 1998 1999 2000 Assets 323,005 382,958 749,449 765,462 1,376,812 Current assets 267,969 301,723 586,603 471,210 871,769 Cash 3,960 5,415 7,980 10,644 7,525 Bank Indonesia 9,113 18,567 24,228 31,551 67,220 Other banks 33,831 18,882 300,445 191,617 303,044 Outstanding loans 223,393 265,610 259,863 245,360 513,090 Reserve for bad debt -2,328 -6,751 -5,913 -7,962 -19,110 Fixed assetsa 10,086 16,490 15,839 14,774 14,292 Other assetsb 44,950 64,745 147,007 279,478 490,751 Liabilities 298,106 328,255 692,906 709,895 1,317,168 Demand deposits 14,937 18,803 14,706 20,912 27,503 Savings deposits 59,084 54,732 107,419 95,524 92,775 Time depositsc 195,762 207,800 361,499 425,070 992,026 Bank Indonesiad 1,614 9,383 115,982 117,082 105,082 Other loans 20,325 28,883 58,830 15,808 61,783 Other liabilities 6,384 8,654 34,470 35,499 37,999 Equity 24,899 54,703 56,543 55,567 59,644 Capital (including retained earnings) 22,554 52,207 54,702 54,610 54,774 Net profits 2,345 2,496 1,841 957e 4,870 Total liabilities and equity 323,005 382,958 749,449 765,462 1,376,812 Note See table 15 2 for exchange rates a Includes land, buildings, vehicles, and office furniture but excludes fixed assets held in the names of the bank's owners b Includes marketable securities and investments c Includes certificates of deposit d Long-term liabilities e This is the 1999 audited figure But a subsequent adjustment shows that in 1999 net profits were 386 million rupiah See table 15 2, note b Source Bank Dagang Bali data cent to 2.0 percent. However, by 2000 BDB had a 97 percent on-time repay- ment rate, with noncurrent loans down to 0.9 percent and doubtful loans to 0.5 percent; bad debt was 1.8 percent.While some of the repayment problems during the crisis were in the mucroloan portfolio, most of the increase in ar- rears and bad debt was concentrated in BDB's larger loans, especially a few of its largest loans. Thus BDB's strategy for remaining viable during this difficult period in- cluded reducing the number of outstanding loans (primarily large loans, as part of its efforts to lower arrears) and taking advantage of the high interest rates offered by the interbank money market. During the crisis BDB concentrated its nmicrobanking efforts mainly on sav- ings (see table 15.4).As discussed in chapter 10 for earlier years, time deposits account for a small share of BDB's accounts (2 percent in 2000) but for a large share of the value of its savings (89 percent in 2000). In 2000 the average time 386 The Microfinance Revolution: Lessons from Indonesia Table 15A| Bank Dagang Bali savings, 1996, 1998. and 2000 Indicator 1996 1998 2000 Tme deposits' Amount (millions of rupiah) 195,762 361,499 992,026 Share of total (percent) 72 75 89 Number of accounts 8,015 11,665 7,997 Share of total (percent) 2 3 2 Passbook savings Amount (millions oi rupiah) 59,084 107,419 92,775 Share of total (percent) 22b 22 8 Number of accounts 354,888 380,450 393,146 Share of total (percent) 98 97 98 Giro accounts Amount (millions of rupiah) 14,937 14,706 27,503 Share of total (percent) 6a 3 3 Number of accounts 956 748 766 Share of total {percent) 0 0 0 Total Amount (millions of rupiah) 269,783 483,624 1,112,304 Number of accounts 363,859 392,863 401,909 Note See table 15 2 for exchange rates a Includes certificates of deposit b Small discrepancies between this table and table 10 4 are due to rounding (table 10 4 Is in U S dollars) Source Bank Dagang Bali data Table 15.5 Bank Dagang Bali outstanding loans, 1996, 1998, and 2000 Indicator 1996 1998 2000 Value (millions of rupiah) 223 393 259,863 513,090 Number 15,645 12,385 10,417 Note See table 15 2 for exchange rates Source Bank Dagang Bali data deposit balance was 124 million rupiah ($12,929). In contrast, passbook sav- ings accounted for 98 percent of savings accounts in 2000 but for only 8 per- cent of the value of savings. In 2000 BDB had more than 390,000 passbook savings accounts with an average balance of 236,000 rupiah ($25). Although BDB took a somewhat different route to commercial microfi- nance than did the BKDs and BRI's unit desas, like the others it maintained its profitability and its outreach to poor clients during the crisis. In 2001 BDB's Commercial Microfinance in Indonesia Stability in Crisis. 1997 to Mid-2001 387 performance continued to improve, and the bank earned record profits. BDB may go public when the Okas retire, with the Oka family retaining a majori- ty interest so they can ensure that BDB maintains its focus on rnicrofinance. BRI's unit desa system, 1996-2000 Continuing its profitable nationwide financial intermediation for low- and lower- middle-income people throughout 1996-2001, the unit desa system proved that such efforts can remain extraordinarily stable even in times of severe national crisis.9 Table 15.6 Unit desa savings and lending, 1996-2000 Indicator 1996 1997 1998 1999 2000 Value of outstanding loans Billions of rupiah 4,076 4,685 4,697 5,957 7,827 Millions of U S dollars 1,711 1,008 585 841a 816 Number of outstanding loans (thousands) 2,488 2,616 2,458 2,474 2,716 Long-term loss ratio (percent)b 2 15 2 17 2 13 2 06 1 90 Portfolio status (percent)c 3 65 4 73 5 65 3 05 2 51 Value of savings Billions of rupiah 7,092 8,837 16,146 17,061 19,115 Millions of U S dollars 2,976 1,900 2,012 2,408a 1,992 Number of savings accounts (thousands) 16,147 18,143 21,699 24,236 25,823 a The 1999 exchange rate used to calculate this table differs slightly from the one in table 2 1 (volume 1), resulting in small discrepancies between the 1999 dollar values of outstanding loans and savings in table 2 3 and this table These discrepancies occurred because this table uses more recent data from the International Monetary Fund (IMF) than were available for table 2 1 b The long-term loss ratio measures the cumulative amount due but unpaid since the opening of the unit compared with the total amount due c Portfolio status measures the aggregate amount of overdue principal installments compared with total principal outstanding Source BRI monthly unit desa reports Table 15.7 Unit desa pretax profits and returns on assets, 1996-2000 Indicator 1996 1997 1998 1999 2000 Pretax profite Billions of rupiah 422 9 417 0 713 7 1,190 3 1,160 5 Millions of U S dollars 177.4 897 88.9 168 Oa 1209 Return on assets (percent) 5 7 47 49 61 5 7 a At the end of each year all the unit desas' profits are transferred to Bank Rakyat Indonesia's general account Source BRI monthly unit desa reports 388 The Microfinance Revolution: Lessons from Indonesia Figure 15.3 Value of unit desa loans and savings, December 1996-December 2000 20 Trillions of rupiah 15 10 Loans _ 5 _ 0 December June December June December June December June December 1996 1997 1997 1998 1998 1999 1999 2000 2000 Source BRI unit desa monthly reports Figure 1 5A Number of unit desa outstanding loans and savings accounts, December 1996-December 2000 30 Mlillions 25 20 Savings ac t 15 10 5 Outstanding loans 0 December June December June December June December June December 1996 1997 1997 1998 1998 1999 1999 2000 2000 Source BRI unit desa monthly reports Savings and lending in the unit desas from 1996-2000 are shown in table 15.6, and profits and returns on assets are in table 15.7 (see the appendix for more detail).The value of outstanding KUPEDES loans rose from 4.1 trillion rupiah ($1.7 billion) in 1996 to 7.8 trillion rupiah ($816 million) in 2000 (fig- ure 15.3).The number of loans increased slowly, from 2.5 million in 1996 to 2.7 million in 2000 (figure 15.4). Repayment remained high and steady Commercial Microfinance in Indonesia Stability in Crisis, 1997 to Mid-2001 389 throughout. In December 2001 there was 9.9 trillion rupiah ($949 million) in 2.8 million loans, portfolio status was 2.2 percent, and the long-term loss ratio was 1.65 percent (see table 15.6 for loss ratio definitions). The value of savings (in rupiah terms) and the number of savings accounts increased dramatically during the same period. In 1996,16.1 million accounts held 7.1 trillion rupiah ($3.0 billion) in savings. By 2000 there were 25.8 mil- hon accounts with 19.1 trillion rupiah ($2.0 billion; see figures 15.3 and 15.4). By December 2001,27 million accounts held 22.0 trillion rupiah ($2.1 billion). Figure 15.5 shows unit desa deposits and loans in real terms from 1984-2000 (in constant 1984 prices). During this period deposits grew steadily, though at a slightly lower rate after the crisis began. In real terms the In rupiah terms, value of outstanding loans declined sharply in 1998 but grew steadily in 1999 and 2000. assets more than Because the unit desas are a division of a larger bank, at the end of each year all their profits are transferred to BRI's general account.The unit desa sys- quadrupled at Bank tem continued to be profitable during 1996-2000, posting a pretax return on average assets of 5.7 percent in 1996,4.7 percent in 1997, 4.9 percent in 1998, Dagang Bali 6.1 percent in 1999, and 5.7 percent in 2000 (see table 15.7). In 2001 the pre- tax return on average assets was 5.8 percent.10 between 1996 and In 1997 BRI began a partial reorganization, carried out as part of the gov- ernment's plan to sell BRI shares to the public (with the government retaining 2000 (a 6 percent a majority share) .The original plan was to float BRI shares in 1999, but the pub- lic offering was postponed because of the crisis. However, reorganization pro- increase in dollar ceeded as part of the restructuring and recapitalization plans for state banks. With the reorganization, responsibilitles and reporting relationships changed terms) somewhat from those shown in figure 14.3. Six chief operating officers, reporting to the chief executive officer, each held responsibility for one of the following divisions: micro, retail, corporate, and investment banking; human resources de- velopment; and systems and technology. Divisions responsible for other activ- ities such as audit and supervision reported direcdy to the chief executive officer. The name of the division responsible for the unit desa system was changed from Business Unit Desa to Microbanking. Further reorganization occurred during the crisis.The government decid- ed that the recapitalized and reorganized BRI would focus largely on micro and retail banking. As a result the Microbanking Division now has more au- tonomy within the bank, which may mitigate some of the problems with the old BRI divisional structure discussed in chapter 14.A new chief executive of- ficer and a new group of chief operating officers have been installed.When the bank is considered ready, shares will be sold to the public. Savings. Despite the exceptionally steep decline in currency value, deposits in the unit desas increased steadily in rupiah terms between 1996 and 2000 (see table 15.6).11 Deposits in BRI's rupiah-denominated accounts outside the unit desa system also increased substantially during this period."2 The rupiah value of unit desa savings accounts from 1996-2000 is shown in table 15.8; the 390 The Microfinance Revolution: Lessons from Indonesia Figure 15.5 Real value of unit desa loans and savings, 1984-2000 4,000 Billions of rupiah In constant 1984 prices 3,500 3,000 Deposits 2,500 2,000 1,500 o ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~on 1,000 500 0 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Source Maurer and Seibel forthcoming, p 9, draft paper prepared for the tnternational Fund for Agricultural Development number of accounts is shown in table 15.9 (see chapter 13 for descriptions of the units' savings products). Time deposits (Deposito Beijangka). As in BDB, the largest gain in the value of the unit desas' deposits (which are denominated in rupiah) was in time de- posits-which jumped from 791 billion rupiah ($323 million) inJune 1997, the month before the crisis began, to a high of 7.1 trillion rupiah ($871 billion) in May 1999.Then, as interest rates gradually decreased, the value of the umts' time deposits dropped to 3.1 trillion rupiah ($319 billion) at the end of 2000. The number of time deposit accounts at the units grew by nearly nine times-from 108,911 in June 1997 to a peak of 956,877in May 1999. These accounts then decreased to 297,495 in December 2000.At the end of 1998, at the height of the crisis, time deposits made up 4 percent of the units' accounts but 39 percent of the value of savings. But by 2000 time deposits accounted for only 1 percent of accounts, with 16 percent of the value of savings.This pattern continued in 2001. The sharp increase in the use of time deposits at the units in 1998 and 1999 reflects three important trends: Commercial Microfinarce n Indonesia Stability in Crsis, 1997 to Mid-2001 391 Table 1 5.8 Value of unit desa savings by account type, 1996-2000 (billions of rupiah) Account type 1996 1997 1998 1999 2000 TABANAS 787 917 1,192 1,528 1,838 (11) (10) (7) (9) (10) SIMPEDES 4,407 5,338 7,165 9,515 11,904 (62) (60) (45) (56) (62) SIMASKOT 976 1,225 1,271 1,802 2,081 (14) (14) (8) (10) (11) Time deposits 775 1,203 6,323 4,060 3,062 (11) (14) (39) (24) (16) Giro 122 140 193 152 222 (2) (2) (1) (1) (1) Other 24 13 2 4 8 (0) (0) (0) (0) (0) Total 7,091 8,836 16,146 17,061 19,115 (100) (100) (100) (1 00) (100) Note Numbers in parentheses are the share of the total for each account type See table 15 2 for exchange rates Source BRI unit desa monthly reports Table 1 5.9 Number of unit desa savings accounts by account type, 1996-2000 (thousands) Account type 1996 1997 1998 1999 2000 TABANAS 4,544 4,797 5,192 5,706 5,821 (28) (27) (24) (24) (23) SIMPEDES 10,100 11,670 13,877 15,791 17,601 (63) (64) (64) (65) (68) SIMASKOT 1,333 1,477 1,690 2,169 2,035 (8) (8) (8) (9) (8) Time deposits 109 138 873 509 297 (1) (1) (4) (2) (1) Giro 59 61 65 60 68 (0) (0) (0) (0) (0) Other 3 1 1 1 1 (0) (0) (0) (0) (0) Total 16,148 18,144 21,698 24,236 25,823 (100) (100) (100) (100) (100) Note Numbers in parentheses are the share of the total for each account type Source BRI unit desa monthly reports 392 The Microfinance Revolution: Lessons from Indonesia * Savers with time deposits (and other accounts) in failing banks moved their funds to time deposits at the units. * Unit desa savers who also had savings in other banks shifted these funds to the units. * Unit desa savers shifted funds from other unit savings products into time de- posit accounts to obtain their higher interest rates. InJune 1997 annual interest rates for time deposits at the units ranged from 13.5 percent for 1- to 3-month deposits to 14.5 percent for 6- to 14-month deposits. By August 1997 the annual interest rate for 1-month time deposits had reached 30 percent. Interest rates peaked in May 1998, when the annual rate for 1-month time deposits reached 60 percent. By the end of 1998, how- In 1996 BRI's unit ever, the interest rate for 1-month time deposits had fallen to 35 percent. Interest rates at the units continued to fall in 1999 as the economy improved. desas had 7 trillion Annual interest rates for fixed deposits declined from the December 1998 range of 35 percent for 1-month deposits to 16 percent for 24-month deposits to De- rupiah ($3 billion) cember 1999 rates of 11 percent for 1-month deposits and 12 percent for 24- month deposits. After May 1999, as interest rates were reduced, time deposits in 16 million gradually decreased (see tables 15.8 and 15.9). The average time deposit account at the units was 7.3 million rupiah in savings accounts. By bothJune 1997 and December 1998.But that average account was worth $2,980 inJune 1997 and $910 in December 1998. By December 2001 the units had 2001 they had 22 269,865 time deposit accounts containing 3.3 trillion rupiah ($313 million). The average time deposit was 12.1 million rupiah ($1,156). trillion rupiah ($2. 1 SIMPEDES and SIMASKOT SIMPEDES and its urban counterpart SIMASKOT together accounted for most of the funds mobilized at the units billion) in 27 during 1996-2000 and for most of the accounts (see tables 15.8 and 15.9).Among unit desa savings products, the largest absolute increase in number of accounts million accounts was in SIMPEDES, which grew by nearly 7 million accounts-from 10.8 mil- lion in June 1997 to 17.6 million in December 2000 (and to 24.2 million in December 2001). Annual interest rates for SIMPEDES ranged from 0-13 percent (depend- ing on minimum monthly balance) in June 1997; the high end of the range rose to 20 percent in the fourth quarter of 1998. But SIMPEDES interest rates were lowered in 1999: from 0-20 percent in December 1998 to 0-11 percent in December 1999. For SIMASKOT, annual interest rates rose from 0-14.5 per- cent inJune 1997 to 0-22 percent in the fourth quarter of 1998, but these rates were also lowered in 1999. The value of SIMPEDES savings at the units increased in rupiah terms from 4.9 trillion rupiah ($2.0 blllion) inJune 1997 to 7.2 trillion rupiah ($893 mil- lion) in December 1998, and to 11.9 trillion rupiah ($1.2 billion) in Decem- ber 2000-a clear indication that savers who valued liquidity were actively saving at the units. By December 2001 the SIMPEDES total was up to 15.8 trillion rupiah ($1.5 billion). In contrast to the fixed deposit accounts, SIMPEDES con- tinued to serve smaller savers.The average SIMPEDES account, which had been Commercial Microfinance in Indonesia Stability in Crisis, 1997 to Mid-2001 393 436,339 rupiah ($183) in 1996, was 516,333 rupiah ($64) in December 1998; by December 2001 it was 652,831 rupiah ($63). SIMASKOT continued to be in demand in cities, accounting for 8-9 per- cent of the number of unit desa savings accounts and 8-14 percent of the value of savings during 1996-2000.The average SIMASKOT account was 732,544 rupiah ($307) in 1996 and 751,700 rupiah ($94) in 1998. By December 2001 the average SIMASKOT account was 1.1 mnllion rupiah ($105). In 1996, before the crisis, SIMPEDES and SIMASKOT accounted for 76 percent of unit savings and 71 percent of savings accounts. Time deposits ac- counted for 11 percent of savings and less than 1 percent of accounts. But by December 1998 time deposits accounted for 39 percent of savings, while SIM- In real terms, the PEDES and SIMASKOT accounted for 53 percent. However, 72 percent of accounts were in SIMPEDES and SIMASKOT, while only 4 percent were in value of outstanding time deposits. Thus during the crisis many savers who could afford to put their funds in KUPEDES loans time deposits did so-primarily because of the high interest rates-but many kept SIMPEDES or SIMASKOT accounts as well. Smaller savers continued to declined sharply in rely on SIMPEDES and SIMASKOT. Both kinds of savers appeared to con- sider these choices their best options for savings under the circumstances. 1998 but grew Overall, unit desa savings grew from 7.1 trillion rupiah ($3.0 billion) in De- cember 1996 to 19.1 trillion ($2.0 billhon) in December 2000, and the num- steadily after that ber of savings accounts rose from 16.1 million in 1996 to 25.8 million in 2000. By December 2001 the units had 22.0 trillion rupiah ($2.1 billion) in 27 mil- lion accounts, for an average account size of 813,106 rupiah ($78). SIMPEDES and SIMASKOT accounted for 84 percent of the savings and 91 percent of the accounts.The multiple reasons that savers continued to save at the unit desas throughout the crisis are explored later in the chapter. Credit. Outstanding KUPEDES loan balances (in rupiah terms) and the num- ber of outstanding loans remained steady between 1996 and 1998 (see table 15.6). But in 1999 and 2000 the value of outstanding loan balances began to grow- from 4.7 trillion rupiah ($585 million) at the end of 1998 to 7.8 trillion rupiah ($816 mlulion) at the end of 2000.The number of loans also increased, from 2.5 million at the end of 1998 to 2.7 million at the end of 2000. In December 1996 the average loan balance was 1.6 million rupiah ($687); in December 1998 it was 1.9 million rupiah ($238), and in December 2000 it was 2.9 million rupiah ($300). By December 2001 the average loan balance was 3.5 million rupiah ($340). The KUPEDES flat monthly interest rate on all loans was raised from 1.5 percent to 2.2 percent in September 1998 (excluding the incentive of 0.5 per- cent a month collected each month and returned to prompt payers; see chap- ter 12). For borrowers who paid on time, this was equivalent to an annual effective rate of about 45 percent for most loans. After two interim decreases, the flat monthly rate was lowered to its original 1.5 percent in September 1999 (equiv- alent for prompt payers to an annual effective interest rate of about 32 percent for most loans). 394 The Microfinance Revolution: Lessons from Indonesia Monthly loss ratios during 1996-2001 are especially interesting (see box 12.3 for definitions of the loss ratios used by the unit desa system).The long- term loss ratio never rose above 2.2 percent in any month between Decem- ber 1996 and December 2001. During the same period the 12-month loss ratio peaked in January 1998-but it was only 2.4 percent.The short-term loss ratio was more volatile, as would be expected, but it peaked at 5.8 percent (also in January 1998).Although the value of the rupiah to the U.S. dollar plunged from 4,650 in December 1997 to 10,375 in January 1998, unit desa borrowers still repaid their loans that month and in succeeding months. Durmg 1996-2001 portfolio status peaked at 6.1 percent in February 1999 (considerably lower than the 9.1 percent reached in 1992 during the tight money period; see table 12.5). Borrowers valued greatly their option to reborrow in time of crisis. Some Savings in the cut back on consumption in order to repay their KUPEDES loans on time. In some cases repayment of KUPEDES installments may have impinged on nu- liquid SIMPEDES trition, health, and education needs, but it is not known to what extent this occurred.Another occurrence that cannot be quantified is that some KUPEDES account increased borrowers were so concerned about maintaining their option to reborrow from their units that they borrowed from moneylenders in order to make their KU- from 4.9 trillion PEDES payments. Another problem arose when some unit staff held back on lending to new borrowers for fear of crisis-related future risks to the borrow- rupiah ($2. 0 ers' enterprises. Finally, some borrowers were reluctant to take out loans dur- ing the crisis because of uncertainties about thelr enterprises and incomes.What billion) inJune borrowvers valued was not necessarily a new loan provided inmmediately, but rather the option to borrow whenever they wanted. As the economy began to im- 1997 to 15.8 prove during 1999, these difficulties began to ease somewhat. Overall, between 1996 and 2000, and continuing in 2001, the unit desas trillion rupiah ($1.5 had an extraordinary loan repayment record. It is impossible to discern from their performance that they were operating in a country where the financial billion) in December system had collapsed. The reasons that the units' repayment rate remained so high are examined later in the chapter. 2001 Profitability. The units have been consistently profitable throughout the cri- sis, earmng pretax returns on average assets-ranging from 4.7 percent in 1997 to 6.1 percent in 1999-that would please many banks in noncrisis countries (see table 15.7). In 1996 the unit desas made a pretax profit of 423 billion ru- piah ($177 million) after covering all costs, including the cost of their super- vision at BRI and the cost of unit desa training. In 1997 unit profits were 417 billion rupiah ($90 million). Had they been calculated on the same basis as in 1996,1997 profits would have been higher in rupiah terms than 1996 profits. But a reserve requirement imposed by the central bank in 1997 resulted in a decrease in the units' profits that year, from 477 billion rupiah ($103 million) to 417 billion rupiah ($90 millon). In rupiah terms unit profits were 71 percent higher in 1998 than in 1997. In 1998, the worst year in Indonesia in more than 30 years, pretax unit desa profits were a stunning 714 billion rupiah ($89 nillion) despite a higher cost Comrnercial Microfinance in Indonesia Stability in Crisis, 1997 to Mid-2001 395 of funds that reduced the financial spread in 1997 and 1998.13 Profits were in- creasingly generated by savings deposited at the branches, but also by KUPEDES loans. And in 1999 unit desa pretax profits reached 1.2 trillion rupiah ($168 million), 67 percent higher in rupiah terms than in 1998. In 2000 unit desa prof- its were again 1.2 trlllion rupiah ($121 million), and in 2001 they were 1.3 tril- lion rupiah ($129 mlllion). The units have always operated with a comfortable financial spread, giv- ing them the flexibility to survive difficult times and to overcome and learn from mistakes.They operate with a spread of about 20 percent of average out- standing loans, while operating costs average about 11 percent (see Maurer and Seibel forthcorming for detailed analysis of the income and cost structure of the Because of income umt desas, 1994-2000). But the spread could have been reduced after the sys- tem had been profitable for some time.The crucial issue about unit desa prof- uncertainties, some itability has been the use of unit desa profits to cross-subsidize wealthier customers of BRI's less successful divisions (see chapters 12 and 14). In com- former KUPEDES mercial microfinance it is not necessary to choose between economic and so- cial profitability; the unit desas have set an extraordinary record for both. But borrowers did not given the units' position as a BRI division, the reality has been that smaller bor- rowers pay higher interest rates to cover losses on larger loans provided at lower borrow during the interest rates to wealthier borrowers-who have much lower average repay- ment rates. As discussed in chapter 12, this practice represents a substantial re- crisis. But they gressive income redistribution. If the mandate for the restructured BRI to concentrate primarily on micro and retail banking can be implemented effec- valued highly the tively, it may be possible to lower KUPEDES interest rates, raise interest rates paid to savers, or both. option to borrow The unit desas at the start of the 2 1st century when ready In contrast to the approach taken in earlier chapters of volume 2, the focus here is not on the development of the units over time. Rather, this section provides a snapshot of the unit desas at the end of 2000 and some thoughts about their future.While the units' current strengths and challenges can be assessed, their future will depend heavily on factors beyond the control of BRI's Microbanking Division.These include the future course of the Indonesian economy and the country's governance, BRI's own governance and management, and the extent to which BRI focuses on micro and retail banking. Strengths. Institutional sustainabldity and large-scale, continuing outreach to the econormically active poor are the twin criteria for success in the new mncrofinance paradigrn.The unit desa system has been able to achieve and main- tain widespread coverage because the system is profitable and fully self-suffi- cient.The increasing volume of business in the units makes possible econormies of scale that add to the system's profits.The units' stability and profitability dur- ing the crisis underscore the depths of their strengths.A few of the system's most important strengths are discussed here. Many reinforce one another, and the whole is considerably greater than the sum of the parts. 396 The Microfinance Revolution: Lessons from Indonesia * The unit desa mission. Credit is provided to all sectors and to all levels of the economically active poor, as well as to lower-middle-income borrowers. Sav- ings are mobilized from all types of savers who live or work in a unit's ser- vice area.This approach to savings mobilization makes it possible for the units to finance KUPEDES loans to all creditworthy loan applicants whose cred- it needs are within the KUPEDES loan size range.This fact is widely known, and it contributes to the high repayment rate: borrowers know that they will not be constrained by lack of available funds at the unit-and that their abil- ity to borrow again depends entirely on their performance. * Simplicity and efficiency. Simplicity is the key to large-scale outreach in com- mercial mncrofinance.The units offer one broadly based loan product and a few carefully designed savings and deposit accounts. Unit products are easy By examining the for clients to use, and they are sufficiently flexible that they can be customized by clients to meet the needs of their households and enterprises. Unit forms, unit desas' procedures, and reporting requirements are also simple. The uniformity of all umts and their limited functions make it possible to manage thousands peformance between of such units at a high level of efficiency, using many high school graduates with in-house training as unit managers.The average number of active loans 1996 and 2001, it per credit officer was 534 in 2000 (Maurer and Seibel forthcoming).There were 3,724 unit desas that year; the average unit had 6,934 savings accounts. would be impossible Efficiency and simplicity are closely related-and essential for profitable large-scale nucrofinance. to learn that the * Management. The unit desa system was built by exceptional managers at BRI's head office who enjoyed high-level government support.This was a sine qua country had been in non for developing a system with the outreach and profitability of today's unit desas. Keeping the units simple at the local level requires high-level man- deep crisis agement of complex issues at the head office. * Financial intermediation. The unit desa system operates as a profitable nation- wide financial intermediary.The success of the strategies discussed above de- pends on an institutional system that operates many secure, conveniently located units. Each unit operates as a profit center, with staff promotions and incentives based on unit performance. Funds are intermediated using the transfer price mechanism, so all creditworthy borrowers can receive loans. (The transfer price is the interest rate paid on finds deposited by a unit in its supervising branch or on funds borrowed by a unit from the branch.) * High KUPEDES repayment. Because KUPEDES loans are widely available at attractive interest rates for all productive purposes and can be obtained and repaid easily and conveniently, most borrowers want to repay in order to keep open their option to reborrow. Other reasons for high repayment include the units' methods of borrower selection and loan collection, extensive training and substantial performance-based cash incentives for unit staff and their supervisors, loan amounts that are determined by borrowers' present cash flows, restriction of loans to ongoing enterprises, refusal to relend to de- faulting borrowers,'4 the prompt payment incentive, prompt classification of overdue loans, a cash-based accounting system, and close supervision of the Commercial Microfinance in Indonesia Stability in Crisis, 1997 to Mid-2001 397 units. High repayment rates during the crisis showed that borrowers place a high value on positive credit ratings. * Meeting savings demand. The units meet the demand for savings from the poor by also collecting savings from the nonpoor-whose deposits raise the aver- age account size and thus make it possible for the units to serve all savers prof- itably.The savings mobilhzed from the public are then available to finance the KUPEDES portfolio. All unit desa loans are financed by savings, which has been a lower-cost source of funds than commercial debt.The extent of mo- bilized savings and the rapid growth of savings during the crisis show that the units have developed savings products and services that are in high demand. * Accountability. Each unit is a profit center, with its own monthly profit and For years BRI had loss statement and balance sheet.The accounting system is set up so that the monthly performance of every unit is completely transparent. Unit staff and used unit desa their supervisors are treated as professionals and held accountable for their performance. They are given authority, training, good salaries, and cash in- profits to cross- centives, and they are placed on a career track within the bank. Unit staff and branch supervisors are provided with extensive training to equip them subsidize lossesfrom to make good financial decisions. Unit managers approve all loans of amounts that fall within their authority, and they and their credit officers are held ac- larger loans to countable for their financial decisions. Promotions are given or withheld based on performance. corporate clients, resulting in Challenges. Despite the unit desas' remarkable strengths, microbanking at BRI faces many challenges.The vulnerabilities of the unit desa system are caused substantial regressive by issues arising outside BRI, issues inside BRI but outside the Microbanking Division, and issues within the Microbanking Division. income redistribution External dangers. The units are potentially vulnerable to difficulties beyond the control of BRI: * Further downturns in Indonesia's economic and political environment could under- mine unit desa performance. The units cannot be expected to maintain stabil- ity indefimtely under severe pressures. If in the next few years Indonesia were to face hyperinflation, negative growth, substantial increases in poverty and unemployment, extended social unrest, and political fragmentation and in- stability, the performance of the units would eventually reflect the problems of the wider society. But if Indonesia moves steadily toward a stable, grow- ing economy and a freely elected government that governs effectively, the unit desas are likely to continue to demonstrate that large-scale commercial ruicrobanking can be robust, resilient, stable, and profitable. * Can BRI's mandate be implemented effectively? The government's recent deci- sion that BRI will concentrate primarily on micro and retail banking should solve many of the units' problems discussed in chapter 14.This decision could also provide microbanking with important opportunities, and perhaps BRI will eventually focus entirely on micro and retail banking. But a smooth in- 398 The Microfinance Revolution: Lessons from Indonesia stitutional transition may be difficult. BRI stllH serves some large clients, and not everyone agrees with its current mandate. The success of the newly re- structured BRI will depend partly on whether conmmercial rnicrobanking continues to have high-level, powerful supporters who understand the im- portance of the units for the economy and the country. * Subsidized credit is expanding. Another danger comes from an old debate; the gov- ernment's views on microfinance have never been monolithic.The tensions be- tween social and comrnercial finance continue, in some wvays remarkably similar to those in the days of Boeke and Fruin (see chapters 9 and 11). Short-sight- ed government efforts to use extensive subsidized credit to help agribusiness and to alleviate poverty do not work any better now than they did in the 1970s. The government rightly sees the expansion of agribusiness (induding agricul- The key to large- ture, agroindustry, and agribusiness trade) as an important potential component of rebuilding the economy. But Bank Indonesia's massive low-interest liquidi- scale outreach in ty credits to government programs for agriculture and related business and trade activities-onlent to borrowers at subsidized interest rates-pose a problem for commerciat the country and a danger to the unit desas. Large low-interest credit programs, doomed to failure for reasons discussed in previous chapters, threaten the unit microfinance is desas whether administered through BRI or not-although the former would be the worst case. Even if the credit subsidies are not adnministered irough BRI, simplicit-hbut this massive cheap credit can undermine the commercially oriented unit desas (and the countrv's many other commnercial mnicrofinance institutions). The requires high-level government should be replicating its successes, not its failures. * Inappropriate regulations continue. The unit desa system is still required to management of comply with some general banking regulations that are inappropriate for mni- crofinance (see chapter 14). In most cases ad hoc compromises have been complex issues at reached on such regulations (as with accounting, reporting, and loan classi- fication methods). But these compronuses represent informal decisions made the head office by particular individuals and could be changed at any time.The basic re- quirements for the operation of commercial microfinance should not be left to the vagaries of personnel appointments and informal relationships. Ap- propriate policies and regulations for microfinance are needed. In this con- text it is encouraging that Minister of Finance Boediono has indicated support for commercial microfinance and for thle development of a viable strategy for nrcrofinance in Indonesia (Boediono 2002). Potential problems in BRI but outside the Microbanking Division. It is too early to know how well the restructured and recapitalized BRI will perform. But an old danger signal still flashes. The Microbanking Division does not yet have the authority to make many of the decisions and resource allocations on which its business depends. The extent to which this problem impedes the develop- ment of microfinance will depend partly on how well the bank fiUlfills its man- date to provide micro and retail banking. If serving low- and middle-income retail clients becomes BRI's central business, then its managers will have an in- centive to make decisions that further that effort. But if the bank's conmirt- Ccr--nerocal Micicc inacs. In Indenesia Stability in C IF S 1997 1c- Mid-2001 399 ment to its mission is not strong enough, some problems from the past may continue, perhaps in new forms. * Management issues are crucial. Because of BRI's generalist management model, under which managers are regularly transferred from one division to another (chapter 14), the unit desa system has never been able to build a core group of high-level managers with extensive experience in microfinance. And since no other large bank in Indonesia has developed a major microfinance focus, there is only a tiny pool of Indonesian commercial bank managers who are microfinance experts.This situation makes it difficult to fill BRI's top man- agement posts. Not surprisingly, BRI's recently appointed managers did not Unit desa products come with substantial microfinance experience. Perhaps, like Kamardy Arief (BRI's president-director from 1983-92; see chapters 12-14), they will learn arefew. But they are microbanking on the job. But BRI should make it a high priority to devel- op a core group of high-level, well-trained, and experienced microfinance carefully designed to experts to lead its nucrobanking activities. * The decisionmaking processfor investments in the unit desas needs to be changed. Al- be flexible so that though the Microbanking Division has had input into BRI's investment de- cisions, it has never had direct control over major investments for niicrobanking. clients can combine Thus the division does not have the authority to decide how much of its prof- its to invest in maintaining, improving, and expanding the unit desa system. and customize the In addition, there are special problems for the unit desas arising from BRI's losses in 1998. One is a severe unit desa staff shortage; another is a lack of in- products to meet vestment funds for replacing old equipment at existing units and for purchasing computers, furniture, motorcycles, and the like for new units. their own needs . The units'savings and profits should be used dfferently. For years unit savings have been used to finance large loans (with generally low repayments), while unit profits have covered losses in other bank divisions.AAmong its other ill effects, this corporate strategy has affected the unit desa interest rates charged to bor- rowers and paid to savers. In the restructured BRI, borrowing rates could be lowered or savings rates could be raised (or both) if the political will is forth- coming. Even with increased investment for maintaining and expanding the units, some change in interest rates should be possible, benefiting unit clients. * Interdivisional coordination needs an overhaul. The Microbanlung Division still depends on other divisions for such functions as personnel, recruitment, lo- gistics, training, and the like. Managers of those divisions do not always un- derstand microbanking needs. For example, units are sometimes understaffed because their labor requirements are not well enough understood and be- cause the Human Resources Development Division (as well as regional and branch authorities) controls much of the units' recruiting and staffing. In ad- dition, problems at branch or regional levels can arise if the units are given low priority relative to other functions of these offices. If the bank develops a strong commitment to fulfill its mandate for micro and retail banking, this problem should be substantially resolved. If not, microbanking may contin- ue to be hampered by problems of interdivisional coordination. 400 The Microfinance Revolution: Lessons from Indonesia Challengesfor the Microbanking Division. There are also problems and chal- lenges that are largely mternal to the Microbanking Division.The problems tend to concern nussion drift and inertia among some managers and staff.The chal- lenges are related to meeting what is still large unmet demand for financial ser- vices among low- and lower-middle-income people in Indonesia. Both the problems and the challenges have arisen partly as a result of BRI's lack of high- level managers who are rnicrofinance experts. Managing the growth of its micro and retail banking is a major challenge for the institution. * Expanding the unit desa system should be a high priority. While the unit desa sys- tem has achieved wide coverage, few of the units have saturated their mar- kets.There are many areas where new units and village service posts could The units meet be operated profitably. Some expansion occurred during the crisis: in 1996 there were 3,595 units, and by December 2001 there were 3,823. Investment demandfor savings in the units and their expansion should become easier under the restructured, and eventually perhaps partly privatized, BRI. Such expansion would be es- from the poor by pecially timely in helping Indonesia recover from its economic difficulties.15 * Lending can be substantially increased. Despite the large outreach and high qual- collecting savings ity of the KUPEDES portfolio and its contmued emphasis on low- and lower- middle-income borrowers (and the presence of many other commercial alsofrom the microfinance institutions), there is still large unmet demand for small loans in Indonesia. The KUPEDES loan product is designed to permit borrow- nonpoor. This makes ers to start with relatively small loans, improve their enterprises, and quali- fy for larger loans. Millions of borrowers have reborrowed at larger loan sizes. it possible to provide Thus the average outstanding KUPEDES loan balance would be expected to increase gradually, as it has-from 174,000 rupiah ($161) in 1984 to 1.6 savings services to rnillion rupiah ($687) in 1996, and to 3.5 million rupiah in 2001 ($340). KUPEDES has helped many Indonesians out of poverty-in some cases twice all profitably (the second time during the crisis). But the many years of KUPEDES success have led to some mission drift. After attaining substantial profitability, some units have not sufficiendy emphasized the continuing provision of small loans to new borrowers. Continuing the ear- lier trend (chapter 12), in December 2000,42 percent of the KUPEDES port- folio was in loans whose stated purpose was trade, while 30 percent was in loans to salaried employees (favored by unut staff because loan repayment is deducted from the borrower's salary).With more than 40 million households in Indonesia and 2.8 million KUPEDES loans (and relatively little competition in many areas), there is considerable scope for increasing lending-especially for small loans to new borrowers."Given the potential of Indonesia's rural economy and the likely credit demand of micro and small businesses, the number of bor- rowers could probably double or triple" (Maurer and Seibel forthcoming, p. 14).A gradual but continuous expansion in new small loans would help In- donesia boost economic growth and increase equity. * Service to women should be expanded. The gender distribution of savers at the units is not known, but many are women.About one-quarter of KUPEDES Commercial M?crofinance in Indonesia Stability in Crisis. 1997 to Mid-2001 401 loans are taken by women, but this understates the use of KUPEDES cred- it by women because households often use loan proceeds for several enter- prises, including those operated by women. Nevertheless, there is considerable scope for increasing loans to women. Other countries and institutions have found that lending to women can have important development effects. * A better mechanism for transferringfunds is a high priorityfor clients. A safe, fast, and simple mechanism for transferring funds is a high priority for many unit clients in both urban and rural areas. Significant improvement has been made since the early 1980s, when low-income people complained that to trans- fer money to suppliers, buyers, family members, and others, they usually had to get on a bus and carry the cash to the recipient. Nevertheless, there re- Unit staff are given mains considerable room for improvement in developing rapid, efficient fa- cilities that enable funds to be transferred throughout the country from a authority, extensive sender in one unit to a receiver in another. * A small number of carefully selected new products could be developed. There is de- training, good mand in both rural and urban areas for additional unit products, including chldren's education loans, mortgage loans, checking accounts, automatic trans- salaries, and cash fers, and special time deposit accounts for education, housing, and retirement. Clients' views on potential products should be carefully elicited because new incentives. And they products and services represent both opportunities and potential dangers for the units. On the one hand, serious consideration should be given to de- are held accountable veloping products for which there is significant demand at profitable prices. On the other hand, if the units are to increase outreach and maintain prof- for their performance itability, they need to stay simple-and should be sparing in the addition of new products and services. * Joint ventures could provide income while spreading the lessons of the unit desa sys- tem. BRI has received numerous inquiries from financial institutions, gov- ernments, and donors about possible joint ventures with microfinance institutions in other countries. So far none has been activated, at first pri- marily because of staff constraints in the Microbanking Division and more recently because of the crisis and the changes under way at BRI. Develop- ing joint ventures could become an important opportunity once the re- structured BRI has had time to focus on its main priorities. Such ventures could be advantageous to BRI (both in income and experience) while helping to accelerate the international spread of commercial microfinance. Commercial Microfinance in Indonesia: What Has Been Learned? Indonesia offers four main lessons about commercial microfinance: * Millions of people can be served profitably over the long term by different types of financial institutions, and loans can be fully financed by savings, com- mercial debt, private investment, and retained earnings.The BKDs, BDB, and unit desas all provide small loans and voluntary savings services, none is sub- 402 The Microfinance Revolution: Lessons from Indonesia sidized, and all are profitable.This is also true of many People's Credit Banks (BPRs), in addition to the BKDs. * Comrnmercial mncrofinance has a set of core requirements. Thus appropriate policies and regulations (or deregulations) are required. High-level govern- ment support is crucial. Banks need skilled, conmmitted, long-term senior man- agers to build and maintain large-scale microbanking systems.Transparency, accountability, and professionalism are necessary at all levels. * Conmnercial microfinance can be both economically and socially profitable. * Microbanking can be extremely stable, even during severe national crisis. The first two lessons have been discussed at length, but some comments will be useful on the third and fourth. Use of massive Commercial microfinance can be both economically and socially profitable subsidized credit to Commercial microbanking in Indonesia has had several positive effects on de- velopment and equity. First, a large volume of credit has been provided prof- reduce poverty does itably to low- and lower-middle-income borrowers at substantially lower interest rates than were previously available to them. Second, the savings in- not work any better struments provide millions of poor savers with positive real returns and legal- ly recognized assets-while also creating new sources of investment. Third, now than it did in both loans and savings products have been extraordinarily successful in reach- ing customers who had never before borrowed or saved in formal sector fi- the 1970s. nancial institutions. Unit savings accounts provide security and income flows from returns. In Indonesia should addition, they help low-income people build their creditworthiness and self- finance business expansion, education, housing, and other basic requirements. replicate its successes, These savings have helped to deepen local financial markets and to enable small loans to be offered on a large scale. not itsfailures Many KUPEDES borrowers started years ago with small loans and over time have qualified for larger loans.The low default rate indicates that borrowers typ- ically use their loans for profitable activities and that they do not become mired in the accumulating debt cycle that often accompanies informal commercial loans from moneylenders (chapter 6). Widespread loans and savings services have made possible increased pro- duction by microfinance clients, resulting in higher incomes-and often in greater self-confidence among clients. Production growth has also increased employment, since many successful clients employ others to work in their enterprises. In addition, increased household income is often used for chil- dren's education and for improving the health and nutrition of household members. Stability in crisis The experienced comrnercial microbanking institutions discussed above have shown that they can viably serve the economically active poor in difficult times as well as in good times.The stability of the unit desas, BKDs, and BDB in time Commercial Microfinance In Indonesia Stability in Crisis, 1997 to Mid-2001 403 of extraordinary economic, political, and financial crisis is significant both for Indonesia and for the microfinance industry generally. Why has commercial microbanking remained stable during a crisis so severe that the Indonesian banking industry (including BRI) failed? There are multi- ple reasons. Some are specific to the unit desas, but others also apply to the BKDs, BDB, BPRs, and other commercial niicrofinance institutions in Indonesia. Economic and political issues. Most microfinance clients operate in the do- mestic economy and were not directly affected by the fall of the rupiah.This fact, combined with macroeconomic stabilization measures and emergency antipoverty programs, meant that most economically active poor people stayed economical- An old danger ly active-and were able to make use of commercial microfinance services. signalflashes at . Although there were many difficulties on both sides of the agreements be- tween the IMF and the government of Indonesia, support from the IMF and BRI. The other donors helped the Indonesian government achieve significant progress on macroeconomic stabilization relatively quickly. Although inflation reached Microbanking 78 percent at its highest point in 1998, it was contained before it reached hyperinflation and sharply reduced by 2000. GDP, which contracted by Division lacks more than 13 percent in 1998, grew by 5 percent in 2000. * The government, with help from the IMF and other donors, quickly insti- authority over such tuted extensive anti-poverty programs, supplying food, creating employment, and keeping children in school. crucial unit desa . Most microbanking clients were not directly affected by the currency cri- sis. However, some were indirectly affected because they sell products to, buy issues as products from, or work for (or were laid off from) firms engaged in foreign exchange transactions. But some, such as farmers cultivating export crops, recruitment, saw substantial increases in income during the crisis. * Many low- and lower-middle-income people were adversely affected by ris- investment, and ing inflation and growing unemployment and underemployment, especial- ly in late 1997 and in 1998. Borrowers greatly valued the option to reborrow expansion in time of difficulty-and thus made loan repayment a high priority. And savers tried, where possible, to save more and consume less. * In the case of the unit desa system, many savers believed that because BRI is a state bank (and the country's oldest bank), the government guaranteed unit desa savings. Deposits were in fact not guaranteed, though there was some- thing of an implicit guarantee (the government probably would have cov- ered losses by unit desa savers had it become necessary).As noted, in January 1998 the government did guarantee savings and deposits in all banks. * Some high-level government officials who understood the importance of commercial microfinance for the economy continued to watch over the major microfinance institutions during the crisis. Institutional issues. These institutions had built a record of trust.They had good, committed management, friendly and motivated staff, and products ap- 404 The Microfinance Revolution: Lessons from Indonesia propriate for microfinance clients. Clients remained loyal because they valued the services they received. * Clients trust financial institutions that are trustworthy. Well-trained, moti- vated staff at the unit desas and BDB were able to use their knowledge and skills in an increasingly difficult context, and thereby to retain their clients' trust.And BKD clients generally trust the BKD they use because it is owned by their vlllage. * BRI and BDB had long-established microbanking systems that were flexi- ble and robust. And BRI's supervision of the BKDs was especially impor- tant during the crisis. * Unit desa staff and their supervisors, whose salaries and incentive payments In commercial are high by local standards, were strongly motivated to ensure that their units performed well and that they continued in their jobs. microfinance, it is * The security, convemence, and services offered by these microbanking sys- tems were highly valued by their clients, most of whom remained loyal. not necessary to Credit. All three institutions maintained substantial liquidity throughout the choose between crisis, and lending was not capital constrained. Borrowers wanted to retain their option to reborrow and made loan repayment a high priority. social and economic * The unit desas had learned in 1991-92 (during a period when a tight mon- profitability. The etary policy was in effect and KUPEDES lending slowed) that arrears in- creased when borrowers perceived the availability of future loans to be unit desas have set uncertain. During the recent crisis KUPEDES loans were continuously available to creditworthy borrowers (although some units did not actively an extraordinary seek out new borrowers during this period). * Most borrowers in the umt desas and BDB entered the crisis with good cred- recordfor both it ratings. For them, maintaining the option to borrow in time of crisis be- came a high priority.They knew that their institutions were liquid, and they trusted them to provide future credit if they repaid their loans on time. * Over time, microcredit clients in these institutions had tended to invest prof- its in their enterprises, generally maintaining the level of business achieved with each loan. Most increased their equity and lowered their loan leverage, making them better able to withstand external shocks (see Patten, Rosengard, andJohnston 2001 for discussion of this point in the unit desa context). Savings. Savers valued the security, convenience, liquidity, confidentiality, prod- ucts, and services offered by these institutions-factors that were especially im- portant during the crisis. Some savers moved their accounts from failing banks to the unit desas, BDB, or BPRs. * Over the years many savers had built up substantial balances in their savings accounts that helped cushion the effects of the crisis (including the ability to make loan repayments even when income was low). Commercial Microfinance in Indonesia Stability in Crisis, 1997 to Mid-2001 405 * Interest rates for savings, especially for fixed deposits, were high during 1998 and part of 1999.These rates helped savers counter somewhat the effects of rising inflation and underemployment. * Accounts were opened by savers who moved deposits from failing banks to the unit desas, BDB, and BPRs even though the savers had not previously been clients of these institutions. * Unit desa and BDB savers who also had accounts in other, doubtful banks shifted funds from those banks to the units or BDB. * Traditional forms of savings (such as gold, grain, and animals) had become less popular for the reasons discussed in chapter 13. Saving in more than a small amount of gold bore an unacceptable security risk. Grain was too prone The BKDs, BDB, to be borrowed, especially in times of hardship. And saving in more than a few animals held too high an opportunity cost. and unit desas are of * At BRI the transfer price from the branches to the unit desas was 15 per- cent in June 1997. It rose to 24 percent by September 1997, fluctuated be- d4ferent institutional tween 20 and 26.5 percent in early 1998, and reached 33 percent in May 1998, where it remained for the rest of the year.Thus unit desa staff, whose types, but all provide incentive payments are based on unit profitability and overall performance, had strong incentives to follow the priority signaled by the transfer price and small loans and mobilize savings. After 1998 the transfer price was gradually reduced. voluntary savings The three institutions discussed here-a division of a state bank, a private bank, and a century-old village banking system-all maintained their outreach services, none is and profitability during the crisis. But they had (and have) different strengths and weaknesses. The unit desa system is the world's largest financially self-suf- subsidized, and all ficient microbanking system.Among the three institutions discussed here, only the unit desas carried out nationwide financial intermediation among borrowers earn profits and savers. But as a division of BRI, the units generated regressive income dis- tribution: they cross-subsidized low-interest accounts of large borrowers who had low repayment rates with higher-interest microloans that had high repay- ment.These were largely politically motivated decisions and were beyond the control of BRI's Microbanking Division. BDB focused mainly on mobilizing savings.The bank used the savings of wealthier clients to raise the average account size, making it possible to collect savings from large numbers of poor clients while remaining profitable. In con- trast, lending at BDB, while profitable, has not achieved large outreach. The 4,806 active BKDs continued to serve their villages-with the highest return on assets of the three institutions. Unlike the unit desas and BDB, the BKDs had problems with arrears and defaults. But the BKD re- payment problems were not caused primarily by the crisis. They were out- growths of a poor writeoff policy and of a long-term tolerance for arrears-because most overdue loans are eventually repaid. The BKDs are profitable mainly because of their interest rates on loans (higher than those of the unit desas or BDB), their retained earnings deposited at BRI, and their low operating costs. 406 The Microfinance Revolution: Lessons from Indonesia These three institutions all hold world records.As noted, the BKDs are the oldest commercial microbanking system in operation, the unit desas are the largest financially self-sufficient microbanking system, and BDB is the oldest operat- ing licensed commercial bank that focuses on nmcrofinance. But it is not only these commercial microfinance institutions that survived the Indonesian cri- sis. As Steinwand (2001, pp. 288, 303) makes clear, many Indonesian financial institutions that provide microfinance continued to serve their clients profitably throughout the crisis. Speaking about Indonesia's People's Credit Banks (BPRs), Steinwand comments: In contrast to the comniercial banking sector, the BPR sector has been hardly hit by the crisis. Instead some of them like the During the crisis in LPD [Lembaga Perkreditan Desa] and LPN [Lumbung Pitih Negara] gained windfall profits from the crisis. The crisis Indonesia, convincingly proved that savings deposits at BPR are ... even in difficult times the most stable source of funding ... The most commercial dynamic and best managed BPR were always those with a strong savings base. microfinance The evidence from Indonesia for the stability of commercial mnicrobank- institutions served ing in experienced institutions is extraordinary. In the country "that would prob- ably walk away with the prize for Asia's most desperate banking system" (The millions of poor Economist, 18 July 1998), profitable commercial rnicrofinance institutions con- tinued to serve millions of poor people-at a critical period when they par- people profitably-a ticularly needed access to financial services. This record is a hallmark of the microfinance revolution. hallmark of the microfinance Notes revolution 1.The 1999 and 2000 GDP data are prelimnary figures from the Biro Pusat Sta- tistik, the Indonesian Bureau of Statistics. 2. For discussion of the Indonesian financial sector collapse, seeWardhana (1998b); Cole and Slade (1998);World Bank (1998b); Furman and Stiglitz (1998);Radelet and Sachs (1998); Kenward (1999,2000); IMF (1999d); Baker (1999); Lane and others (1999); Radelet (1999); and Stern (2000). 3.There are four categories of nonperforniing loans, ranging from loans fewer than 90 days overdue but with payment irregularities to loans more than nine months over- due (for which 100 percent provisioning and writeoff are required). 4. Bank Bali should not be confused with Bank Dagang Bali.The two banks are unconnected. 5. Baker (1999) provides a detailed discussion of bank restructuring in Indonesia; see also Radelet (1999). 6.As used by IBRA, obligor means an ownership-based group of debtors.The 21 largest obligors consist of 340 debtors. 7. Baird (2000, p.2) points out thatThailand's Financial Sector RLestructuring Au- thority recovered only 22 percent of unsecured business loans. Commercial Microfinance in Indonesia Stability in Crisis, 1997 to Mid-2001 407 8.As noted in chapter 14, BKD accounting does not meet the same internation- al standards as accounting at the BRI unit desas. For example, information on BKD profits and defaults is sometimes inconsistent across years, and annual arrears figures include bad debts incurred over a long period but not written off, thus overstating ar- rears (see Ravicz 1998, p. 13). 9. For discussion of the unit desas during the crisis, see Charitonenko, Patten, and Yaron (1998); Patten, Rosengard, and Johnston (1999, 2001); Seibel and Schmudt (2000); Steinwand (2001); and Maurer and Seibel (forthcoming). 10.Although BRI pays taxes, the unit desa system does not; thus unit desa returns on assets are based on pretax profits. 11. It should be remembered that the terms savings and deposits are used synony- mously in this book except where otherwise incdcated. 12. As a state-owned bank, BRI was able to mobilize substantial savings even at the height of the crisis in 1998. Thus BRI's rupiah deposits, excluding unit desa de- posits, increased from 10.2 trlllion rupiah ($4.2 blllion) in June 1997 to 20.0 trlllion rupiah ($2.5 billion) in December 1998, an increase of 97 percent.The bank had dol- lar deposits as well (but not in the units).Total rupiah deposits for both the units and the rest of BRI increased from 17.9 trillion rupiah ($7.3 billion) inJune 1997 to 36.2 trlllion rupiah ($4.5 billion) in December 1998, an increase of 102 percent. 13.There were, however, some problem areas. Between 1996 and 1997 the num- ber of unit desas increased by 96, from 3,595 to 3,691, and the share of units operat- ing at a loss increased from 5.1 to 5.7 percent.The new units could easily account for this increase. But between 1997 and 1998 the number of units increased by only 12, to 3,703, while the share of units operating at a loss increased to 12.2 percent (the previous high during the 1990s was 16.0 percent in 1991, during the period of the government's tight money policy). However, in 1999 the number of units fell by 9, to 3,694, and the share of loss-making units was just 3.9 percent. In 2000, 97.7 per- cent of the 3,724 units were operating profitably-an all-time end-of-year unit desa record. In 2001 the number of unit desas increased by nearly 100, to 3,823-of which 96.1 percent were profitable. 14.With the exception of defaults caused by force majeure throughout a partic- ular area. 15. The issue of expansion of the unit desas became complex during the crisis. While the units were profitable throughout, BRI's massive losses in 1998 caused problems for the units.Although subsequently recapitalized, BRI is not yet rated "healthy" by Bank Indonesia and so has had difficulty obtaining permission to open new units. 408 The Microfinance Revolution: Lessons from Indonesia Rupiah-U.S. dollar exchange rates, consumer prices, and C Appendix performance of Bank Rakyat Indonesia's unit desa system, December 1996-December 2000 Dec- Jan- Feb- Sept- Oct- Nov- Dec- ember uary ruary March April May June July August ember ober ember cember Indicator 1996 1997 1997 1997 1997 1997 1997 1997 1997 1997 1997 1997 1997 Exchange rate (end of period) 2,383 2,396 2,406 2,419 2,433 2,440 2,450 2,599 3,035 3,275 3,670 3,648 4,650 Consumerpriceindex(1995=100) 1095 1113 1122 1122 1126 1130 1129 1139 1149 1162 1178 1187 1207 Value of outstanding loans Billions of rupiah 4,076 4,187 4,125 4,141 4,167 4,205 4,301 4,354 4,408 4,439 4,501 4,573 4,685 Millions of U S dollars 1,711 1,747 1,714 1,712 1,713 1,723 1,756 1,675 1,452 1,355 1,226 1,254 1,008 Number of outstanding loans (thousands) 2,488 2,508 2,525 2,511 2,524 2,524 2,547 2,544 2,545 2,565 2,585 2,599 2,616 Long-term loss ratio (percent)a 215 217 2 20 219 219 219 218 2 18 217 216 2.15 217 217 Portfolio status (percent)b 3 65 3 77 4 11 4 18 4 31 4 42 4 41 4 46 4 55 4 63 4 67 4 80 4 73 Value of savings Billions of rupiah 7,092 6,972 7,101 7,263 7,419 7,654 7,741 7,929 7,984 8,079 8,293 8,704 8,837 Millions of U S dollars 2,976 2,910 2,951 3,003 3,049 3,137 3,160 3,051 2,631 2,467 2,260 2,386 1,900 Number of savings accounts (thousands) 16,147 16,232 16,349 16,508 16,665 16,852 16,993 17,190 17,370 17,630 17,856 18,048 18,143 Pretax profit or loss' Billionsofrupiah 4229 94 326 693 1061 1432 183.1 2221 2467 2851 3341 380.8 4170 3 MMillionsof U S dollars 1774 3 9 13.5 287 436 587 747 854 81 3 871 910 1044 897 o. Return on assets (percent) 5 7 1 5 2 6 3 6 41 43 4 5 46 44 4 5 4 7 47 47 a a It 0 0 3 0 I- 0 Rupiah-U.S. dollar exchange rates, consumer prices, and performance of Bank Rakyat Indonesia's unit desa system, December 1996-December 2000 (continued) Jan- Feb- Sept- Oct- Nov- Dec- uary ruary March April May June July August ember ober ember cember Indicator 1998 1998 1998 1998 1998 1998 1998 1998 1998 1998 1998 1998 Exchange rate (end of period) 10,375 8,750 8,325 7,500 10,525 14,900 13,000 11,075 10,700 7,550 7,300 8,025 Consumerpriceindex(1995=100) 1290 1455 1535 1605 1691 1769 1921 2042 2119 2113 2115 2145 Value of outstanding loans Billions of rupiah 4,751 4,642 4,614 4,560 4,555 4,594 4,590 4,608 4,597 4,592 4,612 4,697 Millions of US dollars 458 531 554 608 433 303 353 416 430 608 632 585 Number of outstanding loans (thousands) 2,629 2,624 2,613 2,585 2,579 2,571 2,536 2,508 2,510 2,491 2,473 2,458 Long-term loss ratio (percent)a 2 22 2 21 2.20 2 19 2 20 2 19 2 20 2 17 2 15 2 15 2 13 2.13 Portfolio status (percent)b 5 07 5 31 5 35 5 55 5 73 5 77 5 94 5 81 5 86 5 99 5 92 5 65 Value of savings Billions of rupiah 9,124 9,910 10,986 12,157 12,289 13,581 14,515 15,130 15,636 16,054 16,240 16,146 Millions ofUS dollars 879 1,133 1,320 1,621 1,168 912 1,117 1,366 1,461 2,126 2,225 2,012 Number of savings accounts (thousands) 18,324 18,594 19,107 19,359 19,643 20,062 20,607 20,929 21,414 21,594 21,884 21,699 Pretax profit or lossc Billions of rupiah -1 9 22.5 61 2 1393 1525 2084 280.2 351 5 419.0 4975 5800 7137 Millionsof U S. dollars -0.2 2 6 7.3 186 14.5 140 21 5 31.7 392 659 795 889 Return on assets (percent) -0 2 1 5 2 6 4 2 3 6 3 9 4 2 4 4 4 5 4 5 4 6 4 9 a The long-term loss ratio measures the cumulative amount due but unpaid since the opening of the unit compared with the total amount due b Portfolio status measures the aggregate amount of overdue principal installments compared with total principal outstanding c Measured cumulatively within each year. At the end of each year all the unit desas' profits are transferred to Bank Rakyat Indonesia's general account d. For the 1999 end-of-year exchange rate, there is a discrepancy between table 2 1 (volume 1) and this table Thus the 1999 dollar values of outstanding loans, savings, and profits are somewhat different in table 2 3 and this table. These discrepancies occurred because updated data from the International Monetary Fund (IMF) were used for this table In addition, for the consumer price index 1990 = 100 in table 2 1; here 1995 = 100 Source. BRI unit desa monthly reports; IMF, International Financial Statistics (March, May 2001) -a Rupiah-U.S. dollar exchange rates, consumer prices, and Appendix performance of Bank Rakyat Indonesia's unit desa system, December 1996-December 2000 (continued) Jan- Feb- Sept- Oct- Nov- Dec- uary ruary March April May June July August ember ober ember comber Indicator 1999 1999 1999 1999 1999 1999 1999 1999 1999 1999 1999 1999 Exchange rate (end of period) 8,950 8,730 8,685 8,260 8,105 6,726 6,875 7,565 8,386 6,900 7,425 7,085d Consumerpriceindex(1995= 100) 220.8 2236 223.1 2217 2211 220.3 2180 216.0 214.2 2143 2149 2186 Value of outstanding loans Billions of rupiah 4,627 4,590 4,587 4,624 4,736 4,867 5,040 5,218 5,392 5,575 5,765 5,957 Millions of US dollars 517 526 528 560 584 724 733 690 643 808 776 841 Number of outstanding loans (thousands) 2,457 2,442 2,423 2,422 2,419 2,426 2,433 2,447 2,509 2,466 2,480 2,474 Long-termlossratio(percent)a 2.16 215 213 211 209 207 207 205 203 2.01 202 206 Portfolio status (percent)b 5 98 6 08 5 98 5 93 5 75 5 57 5 37 5 11 4 92 4 73 3.88 3 05 Value of savings Billions of rupiah 16,174 16,394 16,836 17,288 17,500 17,676 17,612 17,579 17,630 17,857 17,669 17,061 Millions of US dollars 1,807 1,878 1,939 2,093 2,157 2,627 2,562 2,324 2,102 2,588 2,384 2,408d Number of savings accounts (thousands) 22,197 22,362 22,647 22,889 23,180 23,272 23,347 23,387 23,473 23,790 24,089 24,236 Pretax profit or loss' Billionsofrupiah 851 183.9 2925 412.0 5543 8059 886.3 9805 1,0695 1,1462 1,197.1 1,1903 MillionsofUS dollars 95 211 337 499 683 1197 1289 1296 1275 1661 1612 1680 -I * ~~~Return on assets (percent) 6 8 7 1 7 3 7 4 7 8 9 2 8 4 8 0 7.6 7 2 6 7 6 1 a n :0 0 0 3 _). 0 CD 0 3 a. Rupiah-U.S. dollar exchange rates, consumer prices, and Appendix performance of Bank Rakyat Indonesia's unit desa system, December 1996-December 2000 (continued) Jan- Feb- Sept- Oct- Nov- Dec- uary ruary March April May June July August ember ober ember cember Indicator 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 Exchange rate (end of period) 7,425 7,505 7,590 7,945 8,620 8,735 9,003 8,290 8,780 9,395 9,530 9,595 Consumer price index (1995 = 100) 221 5 221 6 220.6 221.9 223 7 224 8 227 7 228.9 228 8 231 4 234.5 239 0 Value of outstanding loans Billions of rupiah 5,889 6,013 6,141 6,238 6,478 6,713 6,869 7,019 7,183 7,396 7,733 7,827 Millions of U S dollars 793 801 809 785 752 769 763 847 818 787 812 816 Number of outstanding loans (thousands) 2,451 2,508 2,519 2,518 2,546 2,627 2,577 2,590 2,599 2,600 2,647 2,716 Long-term loss ratio (percent)2 2 07 2 04 2 04 2 02 1 99 1 96 1 95 1 92 1 90 1 86 1 85 1 90 Portfolio status (percent)b 3 20 3 18 3 15 3 20 3 07 2 99 2 97 2 94 2 95 2 87 2 82 2 51 Value of savings Billions of rupiah 17,249 17,420 17,478 17,606 17,890 18,055 18,472 18,755 18,797 18,994 19,144 19,115 Millions of U S dollars 2,323 2,321 2,303 2,216 2,075 2,067 2,052 2,262 2,141 2,022 2,009 1,992 Number of savings accounts (thousands) 24,115 24,423 24,576 24,484 24,619 24,883 25,098 25,300 25,489 25,642 25,832 25,823 Pretax profit or lossC Billionsofrupiah 567 1596 262.2 3755 4875 5803 689.7 8099 9351 1,0487 1,1630 1,1605 Millions of U S dollars 7 6 21 3 34.5 47 3 56 6 66 4 76 6 97 7 106.5 111 6 122.0 120 9 Return on assets (percent) 3 5 4.9 5 3 5 7 5 8 5.8 5 9 6 0 6.1 6 2 6 2 5 7 a The long-term loss ratio measures the cumulative amount due but unpaid since the opening of the unit compared with the total amount due b. Portfolio status measures the aggregate amount of overdue principal installments compared with total principal outstanding c Measured cumulatively within each year At the end of each year all the unit desas' profits are transferred to Bank Rakyat Indonesia's general account. d. For the 1999 end-of-year exchange rate, there is a discrepancy between table 2.1 (volume 1) and this table Thus the 1999 dollar values of outstanding loans, savings, and profits are somewhat different in table 2 3 and this table. These discrepancies occurred because updated data from the International Monetary Fund (IMF) were used for this table In addition, for the consumer price index 1990 = 100 in table 2 1, here 1995 = 100. Source. BRI unit desa monthly reports, IMF, International Financial Statistics (March, May 2001) Table of Contents for The Microfinance Revolution, Volume 1: Sustainable Finance for the Poor Foreword Ira W Lieberman Introduction Ali Wardhana Preface Acknowledgments PART 1 THE PARADIGM SHIFT IN MICROFINANCE Overview Chapter 1 Supply and Demand in Microfinance What Is Microfinance? Estimating the Demand for Microfinance Informal Commercial Moneylenders and Their Interest Rates The Economically Active Poor and the Extremely Poor A Poverty Alleviation Toolbox The Financial Systems Approach and the Poverty Lending Approach: A Fork in the Road Financially Self-sufficient Microfinance Institutions Providing credit and savings services profitably Reaching scale Why Has the Demand for Institutional Commercial Microfinance Not Been Met? Why Does Meeting the Demand for Institutional Commercial Microfinance Matter? 415 Access to financial services Building the self-confidence of the poor Notes Chapter 2 Shifting the Microfinance Paradigm: From Subsidized Credit Delivery to Commercial Financial Services Financial Markets in Developing Countries Development of the Financial Systems Approach to Microfinance Microfinance in the 1980s: Going to scale Microfinance in the 1990s: Developing the industry Institutional Sustainability Levels of sustainability Entering the formal financial sector: scale and depth of outreach Pioneers in Large-scale Commercial Microfinance Microbanking at Bank Rakyat Indonesia BancoSol The Old Paradigm: Subsidized Credit Delivery The New Paradigm: Sustainable Comnmercial Microfinance Common elements of sustainable microfinance institutions Basic operating principles Meeting the Demand for Microfinance Notes Chapter 3 Voices of the Clients Do Poor People Understand Microfinance Products and Services, and Do They Know How to Use Them? Indonesia: Understanding interest rates Kenya: Demand for savings products Bangladesh: Managing scarce resources Peru: Credit as a launch pad Bangladesh: Saving for the future Indonesia: Using credit and savings products together Can Microfinance Help the Economically Active Poor Expand and Diversify Their Enterprises and Increase Their Incomes? The Philippmes: Use of microloans at the lower levels of the economically active poor Honduras: Developing and financing an enterprise Uganda: Building income opportunities with credit Indonesia: Building income opportunities with savings Senegal: Gaining business experience Nicaragua: Leaving the moneylenders behind Kenya: Expanding income rapidly Can Access to Financial Services Enhance the Quality of Life of the Clients of Microfinance Institutions? Bangladesh: "Tell your husband that Grameen does not allow borrowers who are beaten by their spouses to remain members and take loans" Uganda: Overconmng malnutrition 416 The Microfinance Revolution: Lessons from Indonesia Indonesia: Saving for children's education Bolivia: Providing umversity educations for sons and daughters Can Access to Microfinance Help the Economically Active Poor in Times of Severe Household Difficulty? Mexico: Supporting the family after her husband left them Ethiopia: Supporting a displaced family with her husband in jail Ecuador: Rebuilding a business Indonesia: Starting over using savings Argentina: Starting over using credit Colombia: Facing an economic downturn Kyrgyz Republic: Living thmugh the collapse of the Soviet economy Can Successful Microfinance Institutions Promote the Self- confidence ofTheir Clients? Notes PART 2 THEORIES OF LOCAL FINANCE: A CRITIQUE Overview Chapter 4 Supply-leading Finance Theory Supplying Finance in Advance of Demand How Credit Subsidies Prevent Sustainable Microfinance Large-scale subsidized programs generally do not reach low- income households Subsidized credit programs, especially in state-owned institutions, often have high default rates Subsidized credit, channeled to local elites, buys political support for governments-and once offered, is difficult to dislodge Borrowers bear high transaction costs Loan products are inappropriate for borrowers' needs Bank staff time is used unproductively Subsidized credit prevents the development of sustainable financial institutions Subsidized Credit Programs for Microenterprises Notes Chapter 5 The Imperfect Information Paradigm Imperfect Information in Credit Markets Information Flows in Rural Conmunrities of Developing Countries Imperfect Information and Rural Credit Markets Informal commercial lenders Formal sector financial institutions operating commercially in rural credit markets Imperfect Information Credit Models and Profitable Microbanks Notes Chapter 6 Informal Commercial Moneylenders: Operating Under Conditions of Monopolistic Competition Informal Conmmercial Lenders in Microfinance Markets Contents for The Microfinance Revolution, Voiume 1 417 The Size of Informal Credit Markets Debates about Moneylenders and Their Interest Rates "Malicious," monopolistic moneylenders Informal moneylenders: "value for the people" Informal moneylending as monopolistic competition The three views in a development perspective Policy implications Characteristics of Informal Commercial Loans from Moneylenders Flow credit and stock credit Interlinked transactions Availability of loanable funds The "Too Many Lenders" Problem Each lender serves a small number of borrowers A borrower typically receives credit from one informal commercial lender Large lenders use credit layering Transaction Costs and Risks for Informal Commercial Moneylenders The transaction costs of lending Risks Are the transaction costs and risks of informal commercial lenders high or low? What Interest Rates Do Moneylenders Charge? The nature of the data Examples of nominal interest rates of moneylenders in developing Asia Examples of nominal interest rates of moneylenders in other parts of the developing world What Are the Transaction Costs and Other Noninterest Costs for Borrowers? The Costs of Borrowing: Comparing Moneylenders with Bank Rakyat Indonesia's Unit Desas and BancoSol Making Microfinance Competitive Notes Chapter 7 Savings and the New Microfinance Mobilizing Microsavings: Five Broad Patterns Pattern 1: Microfinance institutions that are not permitted to mobilize voluntary savings Pattern 2: Savings as the forgotten half of microfinance Pattern 3: High savings, with lending as the forgotten half of microfinance Pattern 4: Failing at both savings and lending Pattern 5: Profitable financial intermediation Policy implications Forms of Informal Savings Cash Grain and cash crops Animals 418 The Microfinance Revolution: Lessons from Indonesia Gold, silver,jewelry, and other valuables Land ROSCAs and RESCAs Raw materials and finished goods Construction materials Saving by lending Deposits with savings collectors Saving in labor obligations and expected reciprocation for past contributions Advantages and Disadvantages of Informal Savings and Financial Savings Emergencies and unexpected investment opportunities Managing irregular income streams Long-term investments Social and religious obligations Old age and disability Why Does the Formal Sector in So Many Countries Fail to Mobilize Microsavings? Different view on mobilizing small savings Can mobilizing the savings of the poor be financially viable? The Mobilization of Public Savings by Sustainable Microfinance Institutions Preconditions Designing and pricing savings instruments Sequencing a savings program Who Benefits? Benefits to individuals and enterprises Benefits to groups, organizations, and institutions Benefits to the implementing financial institutions Benefits to governments and donors Benefits to the economy, development, and equity Notes Contents for The Microfinance Revolution, Volume 1 419 Glossary and Acronyms Note: Except for international acronyms, entries refer to Indonesia unless otherwise specified. Abri Angkatan Bersenjata Repubhk Indonesia, the armed forces adat (Indonesian) cuscom, tradition afdeelingsbank (Dutch) district bank APEC Asia-Pacific Economnic Cooperation forum arisan (Indonesian) rotating savings and credit association ASCA nonrotating accumulating savings and credit association ASEAN Association of Southeast Asian Nations AVB AlgemeeneVolkscredietbank; General Popular Credit Bank, a precursor of BRI bagian (Indonesian) literally a part or a share. BRI divisions (urusan) are di- videdinto secuons (bagian). Bahasa Indonesia the national language balai desa (Indonesian) vlllage hall Bank Mandiri A state-owned bank created in 1999 by the merger of four state banks bank pasar (Indonesian) market bank, a secondary bank bank umum (Indonesian) general bank, the first of two bank categories recognized by the 1992 Banking Law 421 BAPINDO Bank Pembangunan Indonesia, State Development Bank BAPPENAS Badan Perencanaan Pembangunan Nasional, the National Development Planning Agency BCA Bank Central Asia, Indonesia's largest private bank BDB Bank Dagang Bali BIMAS Bimbingan Massal (Mass Guidance), the government's subsidized agricultural credit program operated through BRI, 1970-85 Bina Swadaya Self-Reliance Development Foundation, based in Jakarta BKD Badan Kredit Desa, village-owned credit organizations established in the late 19,h century during the Dutch colonmal era; today they are supervised by BRI BKKs/CJ Badan Kredit Kecamatan, subdistrict credit organizations (LDKPs)in Central Java BKKs/SK Badan Kredit Kecamatan, subdistrict credit organizations (LDKPs) in South Kalimantan that are modeled on but independent of the BKK/CJ BKKBN Badan Koordinasi Keluarga Berencana Nasional, the Na- tional Family Planning Coordinating Board BKPD Bank Karya Produksi, People's Credit Bank (LDKP) m WestJava BPD Bank Pembangunan Daerah, regional development banks that became general banks under the 1992 Indonesian Banking Law (also known as provincial development banks) BPR Bank Perkreditan Rakyat, People's Credit Bank, has two primary meanings. It refers to rural financial institutions that have met the criteria specified for the second of the two bank categories estabhshed by the 1992 Banking Law (the first category is general banks) and that are li- censed as BPRs.The term BPR is also used to refer to the nearly 9,000 unlicensed People's Credit Banks (generic BPRs); the earliest generic BPRs were estab- lished in the late 19th century under the Dutch colonial regime. BPU (i) Bank Perniagaan Umum, the first private bank in Balh.(ii) Bank Pasar Umum, market bank for the public, established by I Gusti Made Oka and Sri Adnyani Oka in1968; the precursor to BDB. BRI Bank Rakyat Indonesia, a state-owned commercial bank BRI unit desa The lowest-level permanent outlet of BRI's micro- banking system, unit desas are located at subdistrict level and provide financial services to the villages (desas) of their subdistricts. Brown planthopper Nilaparvata lugens Stal, a pest that attacks rice plants. It in- duced large rice crop losses in Indonesia in the 1970s and 422 The Microfinance Revolution: Lessons from Indonesia 1980s as new insecticide-resistant biotypes of brown planthopper developed. BSD Bank Shinta Daya, a private licensed BPR in D.I.Yo- gyakarta BUD Business Unit Desa, BRI's rnicrofinance division, now called Microbanking Division BUMD Badan Usaha Milik Daerah, enterprise owned by a provincial government BUMN Badan Usaha Mlihk Negara, state-owned enterprise CGAP Consultative Group to Assist the Poorest, a multidonor- international consortium formed in 1995 to assist the de- velopment of nucrofinance. In 2001 CGAP had 29 member-donors; its secretariat is located at the World Bank in Washington, D.C. compulsory savings savings required as a condition of obtaining a loan. Also called forced or mandatory savings, such savings usually cannot be withdrawn until the loan is repaid or longer, and sometimes not until the borrower leaves the institu- tion. Since borrowers do not have access to their savings during these periods, the effective interest rate of the loan is higher than the stated interest rate. CPI consumer price index CPIS Center for Policy and Implementation Studies, an In- donesian research and policy advising institution that grew out of DPIS Dayak an ethmc and linguistic category of people indlgenous to the Central Kalimantan area deposito berangka time deposit account (Indonesian) desa (Indonesian) village DPIS Development Program Implementation Studies.A HIID project (1979-83), funded by the Indonesian Ministry of Finance, that conducted field work in a number of In- donesian provinces and provided advice to the Indone- sian government on four of its development programs. DPIS advised the government on the transformation of BRI's unut desa system from a channeling agent for subsi- dized credit to a commercial mnicrofinance intermediary. CPIS was an outgrowth of the DPIS project. DPR Dewan Perwakilan Rakyat, Indonesian parliament dwjfungsi "dual function;" refers to the Indonesian military's dual role in protecting political stability and preventing its de- terioration East Timor Formerly a Portuguese colony, East Timor (Timor Timur) was annexed by Indonesia in the mnid-1970s. In G ossary and Acronyms 423 1999 East Timor voted overwhelming for independence (a vote followed by a massacre of East Timorese by In- donesian militias, supported by some elements in the In- donesian military).East Timor then became a Umted Nations protectorate, and in 2001 it was preparing for in- dependent statehood. effective interest rate interest rate calculated on a declining balance basis-that is, the real cost of the money actually in the client's hand from time to time, in contrast to a flat interest rate calcu- lated on the original loan balance.The computation of effective interest rates should incorporate all fees, com- mussions, compulsory savings, and other costs paid by the borrower to the lender for the use of the borrowed money during each period of the life of the loan. But in practice some or all of these costs to the borrower are frequently omitted in statements of effective interest rates. FAO Food and Agriculture Organization financial intermediaries organizations that provide financial products and services (such as credit, voluntary savings, payment services, and transfer facilities), intermediating between borrowers and lenders/savers flat interest rate rate of interest calculated against the original face amount of the loan (rather than, as in effective interest rate, on the declining balances owed after successive in- stallment payments). FPPI Foster Parents Plan International gabah (Indonesian) dry, unhulled paddy GDP gross domestic product GEMINI Growth and Equity through Microenterprise Investments and Institutions, a USAID-funded project in the 1990s GERTAK Gerakan Serentak, a quasi-mulitary drive to collect over- due BIMAS loans that began in 1981 giro demand deposit account GNP gross national product Golkar golongon karya, or functional groups.Technically a coali- tion of groups, Golkar was the mulitary-backed political arm of the Soeharto regime; it is now a political party. gotong royong cooperation (Indonesian) grassy stunt a virus carried by the brown planthopper; rice plants with grassy stunt disease do not form grains and have no yield GTZ Gesellschaft fur Technische Zusammenarbeit GmbH, the German government's Agency for Technical Cooperation 424 The Microfinance Revolution: Lessons from Indonesia Habibie, BJ. Indonesia's third president, 1998-99 Haj pilgrimage to Mecca (Saudi Arabia) made as part of the religious life of a Muslim HIID Harvard Institute for International Development IBRA Indonesian Bank Restructuring Agency IDT INPRES Desa Tertinggal, presidential instruction for backward villages.A program created in 1993 to provide support to backward ("left behind") vlllages. IFAD International Fund for Agricultural Development IMF International Monetary Fund informal commercial unregulated lenders whose subsidiary, primary, or sole moneylenders occupation is the provision of credit with the expectation of profiting from the loan. Some lend at their own risk; others are brokers who intermediate between savers and borrowers-in which case the savers bear all or part of the risk. ING Internationale Nederlanden Bank, now ING Barings INPRES Desa A presidential instruction that provides an annual grant for development to each Indonesian village INPRES Sekolah Dasar A presidential instruction for the construction of primary schools IRRI International Rice Research Institute, the Philippines Javanese an ethnic and linguistic category of people who have tra- ditionally inhabited the central and eastern parts of the island ofJava JITF Jakarta Initiative Task Force. Established to restructure corporate debt,JITF acts as a mediator between private debtors and creditors. kabupaten (Indonesian) district kecamatan (Indonesian) subdistrict KIK Kredit Investasi Kecil, Small Investment Loan Program KMKP Kredit Modal Kerja Permanen, Small Permanent Work- ing Capital Loan Program KOSTRAD Kommando Strategis Angkatan Darat, strategic reserve command of the Indonesian army KPM Kelompok Pengusaha Mlkro, microenterprise group Kredit Midi A BRI unit desa loan program that provided loans up to $503 (1981-83) Kredit Mmni A BRI unit desa loan program that provided loans up to $201 (1974-83); a precursor to KUPEDES KSP Kelompok Simpan Pinjam, Savings and Credit Group Glossary and Acronyms 425 KUD Koperasi Unit Desa, government-supported vlllage-level cooperatives (KUDs should not be confused with BRI's unit desas) KUK Kredit Usaha Kecil, Small Business Credlit Program. A subsidized directed credit program begun in 1990 that directs all Indonesian banks to lend at least 20 percent of their volume of credit to small enterprises.The oblhgation of banks to provide KUK loans was suspended during the recent crisis. KUKESRA Kredit Usaha untuk Kesejahteraan Rakyat, the credit component of BKKBN's Prosperous Family Program, a subsidized Indonesian government program for women's groups KUPEDES Kredit Umum Pedesaan, the loan product of BRI's unit desas. KUPEDES is an acronym for general rural credit, but after 1989-when these loans were offered in urban units as well as rural ones-KUPEDES became widely known as general purpose credit. KURK Kredit Untuk Rakyat Kecil, Small Enterprise Credit In- stitutions {LDKPs) in East Java KUT Kredit Usaha Tani, Credit for Farm Enterprise Program. KUT is a government-subsidized agricultural credit pro- gram that is channeled through BRI's branches, but not through its unit desas. ladang (Indonesian) swidden agriculture, an ancient technique in which a small forest area is burned and rainfed crops are grown in the ashes. After a few years the cultivator leaves the small plot and starts a new one. Often the cultivator returns to the original plot after a number of years and begins the cycle again. LDKP Lembaga Dana dan Kredit Pedesaan, Rural Fund and Credit Institutions that are owned by provincial or local governments (or both). Some LDKPs have become li- censed BPRs, but most are a subcategory of BPR, using the term in its generic sense. leverage use of equity as a lever to obtain additional funds by bor- rowing or taking deposits. LKP Lumbung Kredit Pedesaan, Rural Credit Organizations (LDKPs) in Nusa Tenggara Barat, a province in eastern Indonesia long-term loss ratio a measurement of loan portfolio quality used by BRI's umt desa system for KUPEDES loans: the cumulative amount due but unpaid since the opening of the unit desa relative to the total amount due LPD Lembaga Perkreditan Desa,Village-owned Financial Or- ganizations supervised by the provincial government of Bali 426 The Microfinance Revolution: Lessons from Indonesia LPN Lumbung Pitih Nagari, People's Credit Banks (LDKPs) ofWest Sumatra LPUK Lembaga Pembiayaan Usaha Kecil, Institutions for Small Enterprise Finance (LDKPs) of South Kalimantan lumbung desa vllUage granary or paddy bank (Indonesian) Madurese a linguistic and ethmc category of people who have tra- ditionally inhabited the island of Madura and the north coast of the eastern extension of the island of Java Mahabharata One of the two main ancient sacred epics of Hinduism, the Mahabharata (literally "great" and "weighty") was composed by the sage Vyasa. "What is here is found else- where; what is not here is not to be found anywhere." Through its stories and teachings the epic illustrates prin- ciples of truth and spiritual realization.The Bhagavad Gita, containing the great battlefield dialogue between Aijuna and Lord Krishna, a reincarnation of the god Vishnu, is retold in the Mahabharata. Adapted to local culture in Indonesia, the Mahabharata has been incorpo- rated into the Indonesian wayang (shadow puppet theater) repertory. MFI microfinance institution microfinance small-scale financial services provided to low-income clients MPR Majelis Permusyawaratan Rakyat, People's Consultative Assembly Nahdlatul Ulama Indonesia's largest Muslim organization NAM nonaligned movement NASAKOM Nasionalisme, Agama, Komunisme (nationalism, religion, communism), a coalition grouping under President Sukarno New Order the Soeharto government governrment NGO nongovernmental organization 1945 constitution the original Indonesian constitution under which the na- tion operates.The 1945 constitution, replaced by two successive constitutions in 1949 and 1950, was restored in 1959. OKM Operasi Karya Makmur, a quasi-military drive to increase BIMAS participation that began in 1981/82 OPEC Orgamzation of Petroleum Exporting Countries P4K Pembinaan Peningkatan Pendapatan Petani-Nelayan Kecil, Income Generating Program for Small Farmers and Fishermen. Sponsored by Indonesia's Ministry of Glossary and Acronyrns 427 Agriculture and international donors, P4K is a subsidized group-based program with a credit component that is administered through BRI's branches, but not through its urnt desas. PAKDES I Paket 23 Desember 1987, Indonesian government pack- age of deregulation measures on capital market policy PAKDES II Paket 20 Desember 1988, continuation of PAKDES I capital market reforms PAKJAN Paket 29 Januari 1990, followup to PAKTO 88 on credit PAKMAR Paket 25 Maret 1989, followup to PAKTO 88 PAKTO 88 Paket 27 Oktober 1988, Indonesian government package of financial, monetary, and banking reforms palauwija (Indonesian) secondary crop Pancasila the five guiding principles of Indonesian state ideology: belief in one supreme being, Indonesian unity, humani- tarianism, democracy by representative consensus, and so- cial justice PDI Partai Demokrasi Indonesia, Indonesian Democratic Party PDI-P Partai Demokrasi Indonesia-Perjuangan, Indonesian De- mocratic Party-Struggle. Megawati Soekarnoputri con- tested the 1999 general elections as head of PDI-P (the P for Perjuangan symbolizes the struggle against Soeharto, who had her removed as head of the earlier PDI when he was president) PDP Provincial Area Development Project, a USAID-support- ed project to assist the BKKs of Central Java in the 1980s pekarangan (Indonesian) mixed garden cultivation, usually in small household or village plots PHBK Program Hubungan Bank dan Kelompok Swadaya Masyarakat, Program to Link Banks and Self-Help Groups. A subsidized techmucal assistance and credit pro- gram sponsored by Bank Indonesia and GTZ. PKB Partai Kebangkitan Bangsa, the National Awakening Party. AbdurrahmanWahid contested the 1999 general elections as head of the PKB. PKI Partai Komunis Indonesia, Indonesian Communist Party portfolio at risk a measurement of loan portfolio quality, defined as the total outstanding balance of loans with late payments divided by the total outstanding balance of the loan portfolio portfolio status a measurement of loan portfolio quality used by BRI's unit desas, defined as the aggregate amount of overdue principal installments relative to total principal outstanding 428 The Microfinance Revolution: Lessons from Indonesia PPD pos pelayanan desa, village service post, a subunit of a unit desa PPP Partai Persatuan Pembangunan, United Development Party pribumi (Indonesian) indigenous Indonesian quintal a umt of mass equal to 100 kilograms Ramayana ancient Sanskrit epic, written originally byValmiki. One of the two great sacred epics of Hinduism, the Ramayana illustrates righteous ideals through the story of the god Vishnu's incarnation on earth as Rama, the prince of Ay- odhya. The Ramayana, adapted to local culture in In- donesia, has been extensively incorporated into the Indonesian wayang (shadow puppet theater) repertory. resurgence-causing nonselective use of certain insecticides can cause the insecticides rapid emergence of new blotypes of insecticide-resistant pests, also killing the natural enemies of the pests; this can result in a rapid increase in pest populations, known as resurgence (see brown planthopper) ROSCA rotating savings and credit association rupiah unit of Indonesian currency savings collector an individual or association that holds the savings of clients (usually poor savers) and charges a fee for the service.Typi- cally savngs collectors operate in the informal, unregulated economy, but some are registered organizations. sawah (Indonesian) irrigated rice fields (regardless of source or quality of irrigation) SBI Sertifikat Bank Indonesia, Bank Indonesia certificate; a security issued by the central bank SDI subsidy dependence index. Measurement of the percent- age increase in the average yield obtained on a loan port- folio needed to compensate for the elimination of all subsidies in a financial institution. Shari'a Islamuc law short-term loss ratio a measurement of loan portfolio quality used by BRI's unit desas that measures monthly changes in the compo- nents of the long-term loss ratio SIMASKOT simpanan kota (urban savings), a savings product offered in BRI's urban unit desas SIMPEDES simpanan pedesaan (rural savings), a savings product offered in BRI's rural umt desas Soeharto Indonesia's second president. Acting president 1967, pres- ident 1968-98 Soekarnoputri, Megawati Sukarno's daughter; Indonesia's fifth president, 2001- present Glossary and Acronyms 429 Sukarno Indonesia's first president, 1949-67 Sundanese an ethnic and linguistic category of people who have tra- ditionally inhabited the western part of the island of Java SUSENAS Survei Sosial Ekonomi Nasional, a periodic socioeco- nomic survey that reports on Indonesian consumption, income, and other variables TABANAS Tabungan nasional, a national savings product TAKESRA Tabungan Kesejahteraan Rakyat, the savings component of BKKBN's Prosperous Famuly Program, a subsidized Indonesian government program for women's groups TAMADES Tabungan masyarakat pedesaan, a savings product of the Central Java BKKs tani (Indonesian) farmer tegalan (Indonesian) dryland that is normally continuously cropped but is not irrigated transaction costs when a lender (formal or informal) provides a loan to a borrower, both parties bear transaction costs. Such costs to the lender include obtaimng information about the credltworthiness of the borrower, administering the loan, and collecting it.Transaction costs that borrowers may incur include the opportunity cost of time spent travel- ing, preparing a loan applhcation, and (in group lending programs) attending meetings; transportation costs; and bribes to officials. TSP triple superphosphate, one of the fertilizers in the BIMAS packet twelve-month loss ratio a measurement of loan portfolio quality used by BRI's unit desas that measures the change over the most recent 12-month period in the components of the long-term loss ratio UBM unit business manager, a umt desa supervisor based at the unit's supervising branch UDO unit development officer, supervisor of unit business managers at large BRI branches UDR unit development report, a monthly report assessing the condition of each BRI unit desa based on 26 indicators UNDP United Nations Development Programme urusan (Indonesian) the primary meaning of the word is business; BRI is di- vided into divisions called urusan USAID Umted States Agency for International Development Vishnu the Sustainer in the Hindu holy trinity of Brahma the Creator,Vishnu the Sustainer, and Siva the Destroyer. Rama and Krishna are among the human forms in which the godVishnu appears. 430 The Microfinance Revolution: Lessons from Indonesia volksbank (Dutch) people's bank Volkscredietwezen popular credit system: credit institutions established in In- donesia by the Dutch colonial administration to provide banking services to indigenous Indonesians. Also the name of an early 20,h century journal about the institu- tional history of the Indonesian popular credit system. voluntary savings savings that are deposited voluntarily in a financial institution. Wahid,Abdurrahman Indonesia's fourth president, 1999-2001 wayang (Indonesian) traditional Javanese shadow theater, the most common form of which is puppet theater.Wayang interweaves en- tertainment, art, philosophy, ethics, history, mythology, and commentary on current events.An ancient art, wayang is extremely popular in Indonesia, especlally on Java and Bali. WOCCU World Council of Credlt Unions Glossary and Acronyms 431 Bibliography Afvan, Ismah, 1992 P "A Review of BRI Unit Development (Background, Philoso- phy and Historical Analysis)." Center for Policy and Implementation Studies, Jakarta. Alexander,Jennifer. 1986. "Information and Price Setting in a Rural Javanese Mar- ket." Bulletin of Indonesian Economic Studies 22 (1): 88-112. Alexander, Jennifer, and Paul Alexander. 1982. "Shared Poverty as Ideology: Agrarian Relationships in Colonial Java." Man, n.s., 17 (4): 597-619, Anderson, Benedict R. 1972. "The Idea of Power inJavanese Culture." In C. Holt, ed. Culture and Politics in Indonesia. Ithaca, N.Y: Cornell University Press. Reprinted in Anderson (1990). .1990. Language and Power, Exploring Political Cultures in Indonesia. Ithaca, NY: Cornell University Press. Anderson, Benedict R. and Ruth McVey. 1971. A Preliminary Analysis of the October 1, 1965 Coup in Indonesia. Cornell Modern Indonesia Project. Ithaca, NY: Cornell University Press. Arndt, Hans W 1984. The Indonesian Economy: Collected Papers. Singapore: Chopmen Publishers. BAAC (Bank for Agriculture and Agricultural Cooperatnves). 1997. "The Bank for Agriculture and Agricultural Cooperatives: Savings Mobilization Case Study." Paper presented at a workshop on institutional comiercial niicrofinance for the working poor, sponsored by Bank Rakyat Indonesia and the U.S. Agency for In- ternational Development, 12-13 June, Jakarta. . 1998. Profit and Loss Statementfor the Fiscal Year ended March 31, 1998 and 1997. Bangkok: Bank for Agriculture and Agricultural Cooperatives. Badan Pengendali BIMAS dan Umversitas Padjadjaran. 1975. Survai Evaluasi Pelak- sanaan (Intensfikasi Padi dan Palaw9a Tahun 1971-1978). Bandung: Universitas Padjadjaran. 433 Baird, Mark. 2000. "Corporate Restructuring in Indonesia." Statement made at the Jakarta Initiative Task Force-Indonesian Bank Restructuring Agency conference on corporate debt restructuring, 2 November,Jakarta. Baker, Susan L. 1999. "Bank Restructuring in Indonesia." Draft. Partnership for Economic Growth, a joint project of the U.S. Agency for International Develop- ment and the Government of Indonesia,Washington, D.C. and Jakarta. Bangladesh Delegation to Indonesia. 1998. "Formal Sector Microfinance: Report of the Bangladesh Delegation to Indonesia." Bangladesh Bank, Dhaka. Bank Indonesia Jakarta dengan Pusat Penelitian Pembangunan Pedesaan dan Kawasan Universitas Gadjah Mada. 1987. "Mengkaitkan Bank dengan LPSM dalam Mobilisasi Dana dan Penyaluran Kredit Studi Kasus." LSM Binaan LPSM Java Tengah dan D. I.Yogyakarta,Yogyakarta. Barclays Bank. 2000. "Country Report: Indonesia." 28 November. Baydas, Mayada, Douglas H. Graham, and LizaValenzuela. 1997. Commercial Banks in Microfinance: NewActors in the Microfinance World. Washington, D.C.: U.S. Agency for International Development. Berenbach, Shari. 1997a. "Colombia." In C. Churchill, ed. Regulation and Supervision of Microfinance Institutions: Case Studies. Occasional Paper 2.Washington, D.C.: Mi- crofinance Network. .1997b. "Indonesia." In C. Churchill, ed. Regulation and Supervision of Microfi- nance Institutions: Case Studies. Occasional Paper 2.Washington, D.C.: Microfi- nance Network. Bergier,Jean-Fran,ois. 1979. "From Italy to Germany:A New Banking Concept?" In Center for Medieval and Renaissance Studies, ed., The Dawn of Modern Bank- ing. Los Angeles: University of Cahfornia Press. Bernsten, R. H., B. H. Siwi, and H. M. Beachell. 1982. The Development and Dffu- sion of Rice Varieties in Indonesia. IRRI Research Paper 71. International Rice Re- search Institute, Los Banos, Philippines. Binhadi. 1995. Financial Sector Deregulation, Banking Development and Monetary Policy: The Indonesian Experience 1983-1993. Jakarta: Institut Bankir Indonesia. Boediono. 2001. Presentation made at the International Conference on Indonesia: Strategy for the Stabilization of the Indonesian Economy and its Sustainable De- velopment in the Future, Mita Kaigisho, Tokyo, 20 March. .2002. "Welcoming Remarks."World Bank Workshop on Sustainable Mi- crofinance,Jakarta, 12 February. Boeke,J. H. 1953. Economics and Economic Policies of Dual Societies-as Exemplfied by Indonesia. Combining revised editions of The Structure of the Netherlands Indies Economy (1942) and The Evolution of the Netherlands Indies Economy (1946). New York: Institute of Pacific Relations. Boomgard,James, and Kenneth J.Angell. 1990. "Developing Financial Services for Microenterprises: An Evaluation of USAID Assistance to the BRI Unit Desa Sys- tem in Indonesia." GEMINI Technical Report 6. Development Alternatives, Bethesda, Md. .1994. "Bank Rakyat Indonesia's Unit Desa System:Achievements and Replicability." In M. Otero and E. Rhyne, eds., The New Role of Microenterprise Fi- nance: Building Healthy Financial Institutionsfor the Poor. West Hartford, Conn.: Ku- marian Press. Booth, Anne. 1985. "Accommodating a Growing Population in Javanese Agricul- ture." Bulletin of Indonesian Economic Studies 21 (2): 115-45. .1988. Agricultural Development in Indonesia. Sydney: Allen and Unwin. Booth,Anne, and Peter McCawley, eds. 1981a. The Indonesian Economy during the Soeharto Era. Kuala Lumpur: Oxford University Press. 434 The Microfinance Revolution: Lessons from Indonesia .198 lb. "The Indonesian Economy since the Mid-Sixties." In A. Booth and P. McCawley, eds., The Indonesian Economy during the Soeharto Era. Kuala Lumpur: Oxford University Press. Booth,Anne, and R. M. Sundrum. 1981. "Income Distribution." In A. Booth and P. McCawley, eds., The Indonesian Economy during the Soeharto Era. Kuala Lumpur: Oxford University Press. Bose, Swadesh R. 1981."Has Economic Growth Immiserized the Rural Poor in In- donesia? A Review of Conflicting Evidence." In G. B. Hainsworth, ed., Village-Level Modernization in Southeast Asia. Vancouver: Umversity of British Columbia Press. Bouman, EJ.A., and H.A.J. Moll. 1992. "Informal Finance in Indonesia." In D.W Adams and D.A. Fitchett, eds., Informal Finance in Low-Income Countries. Boulder, Colo.:Westview Press. Bresnan, John. 1993. Managing Indonesia: The Modern Political Economy. New York: Columbia Unuversity Press. BRI (Bank Rakyat Indonesia). 1985. Bank Rakyat Indonesia dalam Era Pembangunan. 90th Anniversary Book. Jakarta. 1994. BRI Village Units. Jakarta. 1995. Seratus Tahun Bank Rakyat Indonesia: 1895-1995.Jakarta. .1996a. BRI Unit Products. InternationalVisitors Program.Jakarta. . 1996b. Keberadaan dan Prospek KUPEDES Sebagai Sumber Pembiayaan Usaha Kecil. Jakarta. .1997a. BRI Unit Organization and Management. International Visitors Pro- gram. Jakarta. .1997b. Introduction to BRI's Unit Banking System. International Visitors Pro- gram. Jakarta. 1998a. Badan Kredit Desa. International Visitors Program. Jakarta. 1998b. Developing the BRI Unit System: Policy Issues. International Visitors Program. Jakarta. Budhisantoso. 1975. "Rice Harvesting in the Krawang Region (West Java) in Rela- tion to High-YieldingVarieties." Center of Southeast Asian Studies Working Paper 6. Monash University, Clayton,Victoria, Australia. Bulletin of Indonesian Economic Studies. Various issues. Canberra: Australian National Umversity. Caragata,Warren. 2001. "IBRA." Asiaweek, 16 February, p. 27. Charitonenko, Stephanie, Richard H. Patten, and JacobYaron. 1998. "Indonesia: Bank Rakyat Indonesia-Unit Desa 1970-1996." Case Studies in Microfinance. World Bank, Sustainable Banking with the Poor project,Washington, D.C. Chaves, Rodrigo A., and Claudio Gonzalez-Vega. 1996. "The Design of Successful Rural Financial Intermediaries: Evidence from Indonesia." World Development 24 (1): 65-78. Christen, Robert P, Elisabeth Rhyne, and RobertVogel. 1995. Maximizing the Out- reach of Microenterprise Finance: The Emerging Lessons of Successful Programs. USAID Program and Operations Assessment Report 10.Washington, D.C.: U.S. Agency for International Development. Cole, David, and Betty Slade. 1996. Building a Modern Financial System. Cambridge: Cambridge University Press. .1998. "Why Has Indonesia's Financial Crisis Been So Bad?" Bulletin of In- donesian Economic Studies 32 (2): 61-66. Collier,William L. 1978. "Food Problems, Unemployment and Green Revolution in Rural Java." Prisma 9 (1): 38-52 (English edition). . 1981. "Agricultural Evolution in Java." In G. Hansen, ed., Agricultural and Rural Development in Indonesia. Boulder, Colo.:Westview Press. Bibliography 435 Collier,William L., and others. 1974. "Agricultural Technology and Institutional Change in Java." Food Research Institute Studies 13 (2): 169-94. Collier,William L., Harjadi Hadikoesworo, and Suwardi Saropie. 1977. Income, Em- ployment and Food Systems injavanese Coastal Villages. Papers in International Stud- ies, Southeast Asia Series 44. Ohio University, Center for International Studies, Athens. Collier,William L., Soentoro, Hidayat, andYYuliate 1982."Labor Absorption inJa- vanese Rice Cultivation." InWilbert Gooneratne, ed., Labor Absorption in Rice- Based Agriculture: Case Studiesfrom Southeast Asia. Bangkok: ILO-ARTEP. Collier,Wilham L., Soentoro, Gunwan Wiradi, Effendc Pasandaran, Kabul Santoso, and J.F Stepanek. 1982. "Acceleration of Rural Development in Java." Bulletin of Indonesian Economic Studies 18 (3): 84-101. CPIS (Center for Pohcy and Implementation Studies). 1984a. Rural Savings Report No. 1. Jakarta. 1984b. Rural Savings Report No. 2. Jakarta. 1985a. Rural Savings Report No. 3.Jakarta. .1985b. Rural Savings Report No. 4. Jakarta. 1985c. Rural Savings Report No. 5. Jakarta. 1986a. Rural Savings Report No. 6. Jakarta. .1986b. Rural Savings Report No. 7. Jakarta. 1987. Rural Savings Report No. 8. Jakarta. 1988a. Report on Informal Sector Labor. Jakarta. 1988b. Rural Savings Report No. 9. Jakarta. Crawfurd,John. 1993 [1820]. History of the Indian Archipelago. vols. 1-3. Delhi: Low Price Publications [Edinburgh:Archibald Constable and Co.]. Crouch, Harold. 1988 [1978]. The Army and Politics in Indonesia. rev. ed. Ithaca, N.Y.: Cornell University Press. Crouch, Harold, and Hal Hill, eds. 1992. Indonesia Assessment, 1992: Political Perspec- tives in the 1990's. Canberra: Research School of Pacific Studies. DPIS (Development Program Implementation Studies). 1983. "Rice Intensification: Development Program Implementation Studies Report 2."Jakarta: Harvard Insi- tute for Interntional Development and Ministry of Finance. Edmondson,Wade C., and Stella A. Edmondson. 1983. "A Decade of Vlllage De- velopment in EastJava." Bulletin of Indonesian Economic Studies 19 (2): 46-59. Ekonomi dan Keuangan Indonesia. Various issues.Jakarta: Lembaga Pendidikan Ekonomi dan Masyarakat FEUI. Emmerson, Donald K. 1976. Indonesia's Elite: Political Culture and Cultural Politics. Ithaca, N.Y: Cornell University Press. Falcon,Walter P,William O.Jones, and Scott R. Pearson. 1984. The Cassava Econo- my ofJava. Stanford, Calif.: Stanford University Press. Feith, Herbert. 1962. The Decline of Constitutional Democracy in Indonesia. Ithaca, N.Y: Cornell University Press. Fox,James J. 1991. "Managing the Ecology of Rice Production in Indonesia." In Joan Hardjono, ed., Indonesia: The Ecology of Development. Singapore: Oxford Uni- versity Press. Fox,James J., Ross Garnaut, Peter McCawley, and J.A.C. Mackie, eds. 1980. Indone- sia:Australian Perspectives. Canberra: Australian National University, Research School of Pacific Studies. Friedman,Thomas L. 2000. "What a Mess!" The NewYork Times. 3 October, p. A31. Fruin,Thomas A. 1994 r1935]. Provisional Manualfor the Credit Business of the General Popular Credit Bank. The Hague, Netherlands: Ministry of Foreign Affairs, Devel- opment and Cooperation Information Department. 436 The Microfinance Revolution: Lessons from Indonesia . 1999 [1933]. "History, Present Situation and Problems of the Village Credit System (1897-1932)."Translated and edited by Klaas Kuiper. The Hague, Nether- lands: Ministry of Foreign Affairs. Furman, Jason and Joseph E. Stiglitz. 1998. "Economic Crisis: Evidence and Insights from East Asia." Brookings Papers on Economic Activity (2).Washington, D.C.: Brook- ings Institution. Gadway,John E,Tantri M. H. Gadway, andJacob Sardi. 1991. "An Evaluation of the Institutional Aspects of Financial Institutions Development Project Phase I in In- donesia." GEMINI Technical Report 15. Development Alternatives, Bethesda, Md. Gallup,John Luke, Steven Radelet, and Andrew Warner. 1998. "Economnc Growth and the Income of the Poor." CAER (Consulting Assistance on Economic Re- form) Discussion Paper 36. Harvard Institute for International Development, Cambridge, Mass. Gardner, Paul E 1997. Shared Hopes, Separate Fears: Ffty Years of US. -Indonesia Rela- tions. Boulder, Colo.: Westview Press. Garnaut, Ross, and Peter McCawley, eds. 1980. Indonesia: Dualism, Growth and Pover- ty. In J. Fox, R. Garnaut, P. McCawley, and J. A. C. Mackie, eds., Indonesia:Aus- tralian Perspectives. vol. 2. Canberra: Australian National Umversity, Research School of Pacific Studies. Geertz, Clifford. 1963. Agricultural Involution. Berkeley: Umuversity of Califorma Press. .1984. "Culture and Social Change:The Indonesian Case." Man 19 (4): 511-32. Germudis, Dinutri, Denms Kessler, and Rachel Meghir. 1991. Financial Systems and Development 14/hat Rolefor the Formal and Informal Financial Sectors? Paris: Organi- sation for Economuc Co-operation and Development, Development Centre. Glassburner, Bruce, ed. 1971. The Economy of Indonesia: Selected Readings. Ithaca, N.Y: Cornell University Press. .1985. "Macroecononucs and the Agricultural Sector." Bulletin of Indonesian Economic Studies 21 (2): 51-73. Gonzalez-Vega, Claudio. 1982a. "Arguments for Interest Rate Reform." Savings and Development 6 (3): 221-30. .1982b. "Indonesia: Financial Services for the Rural Poor."Jakarta: Re- sources Management International. .1992. "Do Financial Institutions Have a Role in Assisting the Poor?" Eco- nomics and Sociology Occasional Paper 2169. Ohio State University, Rural Fi- nance Program, Columbus. Gonzalez-Vega, Claudio, and Rodrigo Chaves. 1993. Indonesia's Rural Financial Mar- kets. A Report for the Financial Institutions Development Project Indonesia. Washington, D.C.: U.S. Agency for International Development. .1996. "The Design of Successful Rural Financial Intermediaries: Evidence from Indonesia." World Development 24 (1): 65-78. Gordon, Alec. 1978. "Some Problems of Analyzing Class Relations in Indonesia." Journal of Contemporary Asia 8 (2): 210-18. Government of Indonesia. 1983a. Agricultural Census of Indonesia. Jakarta: Depart- ment of Agriculture. 1983b. Lampiran Pidato Kenegaraan 11.Jakarta: Department of Information. 1987. Rural Credit Sector Review, Volume II: Statistical Annexes. Projects Re- port, East Asia and Pacific Regional Office. S-050.Jakarta. .1996. Statistical Yearbook of Indonesia 1996.Jakarta: Biro Pusat Statistik (Cen- tral Bureau of Statistics). Bibliography 437 Guest, Philip. 1989. LaborAllocation and Rural Development: Migration in FourJavanese Villages. Brown University Studies in Population and Development. Boulder, Colo.:Westview Press. Hanna, Donald P. 1994. Indonesian Experience with Financial Sector Reform. World Bank Discussion Paper 237.Washington, D.C. Hansen, Gary, ed. 1981. Agricultural and Rural Development in Indonesia. Boulder, Colo.:Westview Press. Hardjono,Joan, ed. 1991. Indonesia:The Ecology of Development. Singapore: Oxford University Press. Hargrove,T. R. 1988. "TwentyYears of Rice Breeding." BioScience 38 (10): 675-81. Hargrove,T. R., and others. 1979. Genetic Interrelationships of Improved Rice Varieties in Asia. IRRI Research Paper 23. International Rice Research Institute, Los Banos, Philippines. .1985. Changes in Rice Breeding in 10 Asian Countries: 1965-84. IRRI Re- search Paper 111. International Rice Research Institute, Los Banos, Philippines. Harsoadi, M. 1952. "Orgaruzation and Functions of the Agricultural Credit Institu- tions and Credit Problems in Indonesia." In E. K. Bauer, ed., Proceedings of the In- ternational Conference on Agricultural and Cooperative Credit. Berkeley: University of California Press. Hart, Gillian. 1986a. "Exclusionary Labor Arrangements: Interpreting Evidence on Employment Trends in Rural Java."Journal of Development Studies 22 (4): 681-96. .1986b. "InterlockingTransactions: Obstacles, Precursors, or Instruments of Agrarian Capitalism?"Journal of Development Economics 23 (1). 177-203. .1986c. Power, Labor and Livelihood: Processes of Change in RuralJava. Berke- ley: Unuversity of California Press. Hart, Gillian, and Daniel Sisler. 1978. "Aspects of Rural Labor Market Operation:A Javanese Case Study." AmericanJournal of Agricultural Economics 60 (5): 821-26. Hart, Keith. 1986. "The Informal Economy." Cambridge Anthropology 6 (1-2): 845-56. Heinrichs,E.A., and 0. Mochida. 1984. "From Secondary to Major Pest Status:The Case of the Insecticide-Induced Rice Brown Planthopper, Nilaparvata Lugens, Resurgence." Protection Ecology 7: 201-18. Hetler, Carol B. 1989. "The Impact of Circular Migration on aVlllage Economy." Bulletin of Indonesian Economic Studies 25 (1): 53-75. Higgins, Benjammn H. 1955. "The Dualistic Theory of Underdeveloped Areas." Ekonomi dan Keuangan Indonesia (February): 58-78. .1968 [1959]. Economic Development. 2d ed. London. Constable. Hill, Hal, ed. 1989. Unity and Diversity: Regional Economic Development in Indonesia since 1970. Singapore: Oxford University Press. Hinkson,Jim. 1975. "Rural Development and Class Contradiction on Java."Journal of Contemporary Asia 5 (3): 327-36. Hollis, Aidan, and Arthur Sweetman. 1997. "Complementarity, Competition and Industrial Development: The Irish Loan Funds through Three Centuries." Uni- versity of Calgary and University ofVlctoria. ,1998. "Microcredit:What Can We Learn from the Past?" World Development 26 (10): 1875-91. Holloh, Detlev. 1998. Microfinance in Indonesia: Between State, Market, and Se!f Organi- zation. New Brunswick, NJ.: Transaction Publishers. Holt, Sharon L. 1991a. "Village Banking: A Cross-Country Study of a Community- Based Lending Methodology." GEMINI Working Paper 25. Development Alter- natives, Bethesda, Md. 438 The Microfinance Revolution: Lessons from Indonesia 199 lb. "Women in the BPD and Unit Desa Financial Service Programs: Lessons from Two Impact Studies in Indonesia." GEMINI Technical Report 19. Development Alternatives, Bethesda, Md. Hook, RLichard. 1995. "The Experience of Bank Rakyat Indonesia." In E. Brugger and S. Rajapatirana, eds., New Perspectives on Financing Small Businesses in Develop- ing Countries. San Francisco, Calif.: International Center for Economuc Growth and Fundaci6n para el Desarrollo Sostenible (FUNDES). Hugo, Graeme. 1978. Population Mobility in WestJava. Yogyakarta: Gadjah Mada University Press. .1985. "Structural Change and Labor Mobility in Rural Java." In Guy Standing, ed., Labor Circulation and the Labor Process. London: Croom Helm. Hulme, David, and Paul Mosley. 1996. Finance against Poverty. London: Routledge. IMF (International Monetary Fund). 1998. International Financial Statistics Yearbook 1998. Washington, D.C. 1999a. International Financial Statistics. Washington, D.C. 1999b. International Financial Statistics Yearbook 1999. Washington, D.C. 1999c. "East Asia in the Aftermath:Was There A Crunch?"Working Paper 38.Washington, D.C. .1999d. "The Asia Crisis: Causes, Policy Responses, and Outcomes."Work- ing Paper 138.Washington, D.C. 1999e. "The IMF's Response to the Asian Crisis."Washington, D.C. .2001. International Financial Statistics (March, May). Washington, D.C. International Fund for Agricultural Development. 1996. Income Generating Projectfor Marginal Farmers and Landless (P4K): Indonesia's Poverty Eradication Methodology. Jakarta: Ministry ofAgriculture. Institute for Development of Economics and Finance and Bank Rakyat Indonesia. 1999. Theory and Practice of BRI KUPEDES and SIMPEDES. Jakarta. International Labour Organization-United Nations Development Programme Em- ployment Creation Strategy Project. 1989. "The Impact of Unsubsidized Rural Credit Programs on Employment Generation in Indonesia."Jakarta. International Rice Research Institute. 1979. Brown Planthopper: Threat to Rice Produc- tion in Asia. Los Banos, Philippines. .1984.Judicious and Efficient Use of Insecticides on Rice. Los Banos, Phihppines. Ismawan, Bambang. 2000. "Bina Swadaya and Microfinance Development in In- donesia." Paper presented at the Regional Workshop on Microfinance for East Asia and the Pacific, sponsored by the World Bank Institute and World Bank, East Asia Region, Financial Sector Development Unit, 20 March, Manila, Philhppines. Jackson, Karl, and Lucian Pye, eds. 1978. Political Power and Communications in In- donesia. Berkeley: University of California Press. Jayasurlya, Sisira, and Chris Manning. 1988. "Survey of Recent Developments." Bul- letin of Indonesian Economic Studies 24 (2): 3-41. Jenkins, David. 1984. Soeharto and His Generals: Indonesian Military Politics 1975-1983. Monograph Series 64. Cornell University, Modern Indonesia Pro- ject, Ithaca, N.Y Jones, GavinW 1981. "Labor Force Developments since 1961." In A. Booth and P. McCawley, eds., The Indonesian Economy during the Soeharto Era. Kuala Lumpur: Oxford University Press. Jones, Howard Palfrey. 1971. Indonesia: The Possible Dream. NewYork: Harcourt Brace Jovanovich. Kahin,Audrey R., and George McT. Kahin. 1995. Subversion as Foreign Policy. New York: New Press. Bibliography 439 Kasryno, Faisal. 1981. "Land Tenure and Labor Relations in West Java, Indonesia: A Case Study in FourVlllages." Rural Dynamics Series 17. Agro-Economic Survey, Bogor. .1983. "Prospects of Rural Development in Indonesia." Rural Dynamics Series 23. Agro-Economic Survey, Bogor. Kaufmann, Daniel, Gil Mehrez, and Sergio Schmukler. 1998. "The East Asian Cri- sis:Was It Expected?"World Bank,Washington, D.C. Kelley, Allen C., andJeffrey G.Williamson. 1968. "Household Savings Behavior in the Developing Economies.The Indonesian Case." Economic Development and Cul- tural Change 16 (3): 385-403. Kenmore, P. E., and others. 1984. "Population Regulation of the Rice Brown Plan- thopper (Nilaparvata lugens Stal within Rice Fields in the Philippines."Journal of Plant Protection in the Tropics 1 (1): 19-37. Kenward, Lloyd R. 1999. "AssessingVulnerability to Financial Crisis: Evidence from Indonesia." Bulletin of Indonesian Economic Studies 35 (3): 71-95. 2000. "Comments for 'The Indonesian Crisis of 1997-98."' Paper present- ed at the National Bureau of Economic Research, 15 February, Cambridge, Mass. Keyfitz, Nathan. 1985. "Development in an East Javanese Village, 1953 and 1985." Population and Development Review 2 (4): 695-719. Klein, Damel B., ed. 1997. Reputation: Studies in the Voluntary Elicitation of Good Con- duct. Ann Arbor: University of Michigan Press. Koentjaraningrat, R. M. 1975a. Anthropology in Indonesia:A Bibliographical Review. Bibliographical Series 8. Komnkhjk InstitutVoor Taal-, Land-, enVolkenkunde. The Hague, Netherlands: Gravenhage Martinus Nijhoff. .1975b. Introduction to The Peoples and Cultures of Indonesia and Malaysia. Menlo Park, Calif.: Cummings Publishing. .1985.Javanese Culture. Singapore: Oxford Umversity Press. Kuiper, Klaas. 1998. Personal communication. 16 September. Lane,Timothy. 1999. "The Asian Financial Crisis:What Have We Learned?" Finance and Development 36 (3): 1-8. Lane, Timothy, Atish Ghosh,Javier Hamann, Steven Phillips, Marianne Schulze- Ghattas, and Tsidi Tsikata. 1999. "IMF Supported Programs in Indonesia, Korea, and Thailand: A Preliminary Assessment." Occasional Paper 178. International Monetary Fund,Washington, D.C. Lapenu, C6cile. 1998. "Indonesia's Rural Financial System:The Role of the State and Private Institutions." Asia Series.World Bank, Sustainable Banking with the Poor project,Washington, D.C. Liong, Kwan H. 1991. "Rural Savings Mobilization through BRI Unit Banking System: History, Achievement and Prospect." Center for Policy and Implementa- tion Studies,Jakarta. Mackie,J.A. C. 1971."The Indonesian Economy, 1950-1963." In B. Glassburner, ed., The Economy of Indonesia: Selected Readings. Ithaca, N.Y: Cornell University Press. .1980. "The Concept of Dualism and Its Application to Indonesian Agri- culture." In R. Garnaut and P. McCawley, eds., Indonesia: Dualism, Growth and Poverty. Canberra: Australian National University, Research School of Pacific Studies. Mackie,Jamie, and Andrew Maclntyre. 1993. "Politics." In Hal HllE, ed., Indonesia's New Order: The Dynamics of Socio-Economic Transformation. Sydney: Allen & Unwin. Mai, Ulrich, and Helmut Buchholt. 1987. Peasant Pedlars and Professional Traders: Sub- sistence Trade in Rural Markets of Minahasa, Indonesia. Singapore: Institute of South- east Asian Studies. 440 The Microfinance Revolution: Lessons from Indonesia Manmng, Chris. 1987. "Rural Economic Change and Labor Mobility:A Case Study from West Java." Bulletin of Indonesian Economic Studies 23 (3): 52-79. .1988. "The Green Revolution, Employment and Economnc Changes in Rural Java: A Reassessment of Trends under the New Order." Occasional Paper 84. Institute of Southeast Asian Studies, Singapore. Martokoesoemo, Soeksmono Besar. 1993. Beyond the Frontiers of Indonesian Banking and Finance. Rotterdam: Labyrint Publication. Masters, Edward. 1999. "Indonesia's 1999 Elections:A Second Chance for Democ- racy,"Asia Society, NewYork. [www.asiasociety.org/publications/Indonesia] Maurer,Jean-Luc. 1984. "Agricultural Modermzation and Social Change: The Case ofJava over the Last 15 Years." Masyarakat Indonesia 11 (2): 109-19. Maurer, Klaus. 1999. "Bank Rakyat Indonesia (BRI), Indonesia." In A. Hannig and S.Wisniwski, eds., Challenges of Microsavings Mobilization: Concepts and Viewsfrom the Field. Eschborn, Germany: Deutsche Gesellschaft fir Technische Zusamme- narbeit (GTZ). Maurer, Klaus, and Hans Dieter Seibel. Forthcoming. "Agricultural Development Bank Reform: The Case of Unit Banking System of Bank Rakyat Indonesia (BRI)." Draft. International Fund for Agricultural Development, Rome. Mazumdar, Dipak, and M. Husein Sawit. 1986. "Trends in Rural Wages,West Java, 1977-83." Bulletin of Indonesian Economic Studies 22 (3): 93-105. McDonald, Hamish. 1980. Suharto's Indonesia. Blackburn: Fontana Books. McGuire, Paul B., and John D. Conroy. 2000. "The Microfinance Phenomenon." Asia Pacific Review 7 (1): 90-108. McGuire, Paul B.,John D. Conroy, and Ganesh B. Thapa. 1998. Getting the Frame- work Right. Brisbane: Foundation for Development Cooperation. McLeod, Ross H. 1991."Informal and Formal Sector Finance in Indonesia:The Fi- nancial Evolution of Small Businesses." Savings and Development 15 (2): 187-209. .1996. "Control and Competition: Banking Deregulation and Re- regulation in Indonesia." Pacific and Asian Studies Working Paper 96-7. Australian National University, School for Pacific and Asian Studies, Canberra. McLeod, Ross H., ed. 1994. Indonesia Assessment 1994: Finance as a Key Sector in In- donesia's Development. Singapore and Canberra: Institute of Southeast Asian Studies and Australian National Umversity. McVey, Ruth T, ed. 1967. Indonesia. New Haven, Conn.: Human Relations Area Files. Mears, Leon. 1981. The New Rice Economy of Indonesia. Yogyakarta: Gadjah Mada University Press. Mears, Leon, and Sidik Moeljono. 1981. "Food Policy." In A. Booth and P. McCaw- ley, eds., The Indonesian Economy during the Soeharto Era. Kuala Lumpur: Oxford University Press. Meyer, Richard L. 1998. "The Development of Rural Financial Markets." Draft Theme Paper 3. Ohio State Umversity, Columbus, Ohio. Meyer, Richard L., and Geetha Nagarajan. 2000. Rural Financial Markets in Asia: Policies, Paradigms, and Peformance. vol. 3 of A Study of Rural Asia. NewYork: Ox- ford University Press for the Asian Development Bank. Moeljono, Djokosantoso. 1996. "Growing with Our Customers: Microfinance at Bank Rakyat Indonesia."Address to the Semiannual Meeting of the Consultative Group to Assist the Poorest (CGAP), 13 September,Washington, D.C. Moermanto, R.J. 1991. "KUPEDES as a Lending Instrument for BRI Unit Banks." Center for Policy and Implementation Studies,Jakarta. Montgomery, Robert, and Toto Sugito. 1980. "Changes in the Structure of Farms and Farming in Indonesia between Censuses, 1963-1973:The Issues of Inequality and Near Landlessness."Journal of Southeast Asian Studies 11 (2): 349-65. Bibliography 441 Mosley, Paul. 1996. "Indonesia: BKK, KURK and the BRI Unit Desa Institutions." In David Hulne and Paul Mosley, eds., Finance Against Poverty, vol. 2. London: Roudedge. Mubyarto, Loekman S., and Michael R. Dove. 1983. "Problems of Rural Develop- ment in Central Java: An Ethnomethodological Perspective."Journal of Contempo- rary Asia 5 (1): 41-52. Mulyono, Sri Djojosupadmo. 1981. Human Character in the Wayang:Javanese Shadow Play. Singapore: GunungAgung. Nasution,Anwar. 1994. "Banking Sector Reforms in Indonesia." In Ross H. McLeod, ed., Indonesia Assessment 1994: Finance as a Key Sector in Indonesia's Devel- opment. Singapore and Canberra: Institute of Southeast Asian Studies and Aus- tralian National University. Naylor, Rosamond. 1990. "Wage Trends in Rice Production on Java: 1976-1988." Bulletin of Indonesian Economic Studies 26 (2): 133-53. Neill,WilfredT 1973. Twentieth Century Indonesia. NewYork: Columbia Umversity Press. Oka, I. N. 1979. "Feeding Population of People versus Population of Insects:The Example of Indonesia and the Rice Brown Planthopper." In Proceedings of the Symposia of the IX International Congress of Plant Protection, vol. 1, Plant Protection: Fundamental Aspects. Minneapolis, Minn.: Burgess Publishing. O'Rourke, Kevin. 1993. "Reaching the Rural Poor with Banking Services." Har- vard Institute for International Development and Bank Rakyat Indonesia,Jakarta. Otero, Maria, and Elisabeth Rhyne, eds. 1994 The New World of Microenterprise Finance: Building Healthy Financial Institutionsfor the Poor. West Hartford, Conn.: Kumarian Press. Palmer, Ingrid. 1977a. The New Rice in Indonesia. Geneva: United Nations Research Institute for Social Development. .1977b. "Rural Poverty in Indonesia with Special Reference to Java." In Poverty and Landlessness in Rural Asia. Geneva: International Labour Office. Palmier, L. H. 1960. Social Status and Power inJava. Monographs on Social Anthro- pology 20. London: Athlone Press. Parhusip, Uben, and Hans Dieter Seibel. 2000. "Microfinance in Indonesia: Experi- ments in Linkages and Policy Reform." In Joe Remenyi and Benjamin Quifiones Jr., eds., Microfinance and Poverty Alleviation: Case Studiesfrom Asia and the Pacfic. London and NewYork: Pinter. Patten, Richard H. 1996. "The Development of aViable Rural Banking System in Indonesia." Harvard Institute for International Development, Cambridge, Mass. Patten, Richard H., and Jay K. Rosengard. 1991. Progress with Proflts:The Develop- ment of Rural Banking in Indonesia. San Francisco, Calif.: International Center for Economic Growth and Harvard Institute for International Development. Patten, Richard H., and Donald R. Snodgrass. 1987. "Monitoring and Evaluating KUPEDES (General Rural Credit) in Indonesia." HIID Development Discussion Paper 249. Harvard Institute for International Development, Cambridge, Mass. Patten, Richard H.,Jay K. Rosengard, and Don E.Johnston Jr. 1999. "The East Asian Crisis and Microfinance: The Experience of Bank Rakyat Indonesia through mid- July 1999." Harvard Institute for International Development, Cambridge, Mass. .2001. "Microfinance Success Amidst Macroeconomic Failure:The Experi- ence of Bank Rakyat Indonesia During the East Asian Crisis." World Development 29 (6): 1057-69. Pearson, Roland, and Dalls Garland.1993."Financial Institutions Project i Indone- sia: Developing Financial Institutions to Serve Small Enterprises." GEMINI Working Paper 41. Development Alternatives, Bethesda, Md. Penny, David H., and Masri Singarimbun. 1973. "Population and Poverty in Rural Java: Some Economuc Arithmetic from Sriharjo." Cornell International Agricul- 442 The Microfinance Revolution: Lessons from Indonesia tural Development Mimeograph 41. Cornell University, Department of Agricul- tural Economics, Ithaca, N.Y Prawiro, Radius. 1998. Indonesia's Strugglefor Economic Development: Pragmatism in Ac- tion. NewYork: Oxford Umuversity Press. Radelet, Steven.1999. "Indonesia: Long Road to Recovery." Harvard Institute for International Development, Cambridge, Mass. Radelet, Steven, and Jeffxey Sachs. 1998. "The East-Asian Financial Crisis: Diagno- sis, Remedies, Prospects." Brookings Papers on Economic Activity 1. Washington, D.C.: Brookings Institution. . 1999. "What Have We Learned, So Far, From the East-Asian Financial Cri- sis?" [http://www.hiid.harvard.edu/research/newnote.html#asia] Raffles,Thomas Stanford. 1977 [1817]. The History ofJava. Two volumes. London: AMS Press [London: Black, Parbury and Allen]. Ravicz, R Marisol. 1998. "Search for Sustainable Microfinance: A Review of Five Indonesian Initiatives." Pohcy Research Working Paper 1878.World Bank,Wash- ington, D.C. Remenyi,Joe, and Benjamun Quifiones Jr., eds. 2000. Microfinance and Poverty Allevi- ation: Case Studiesfrom Asia and the Paciftc. London and NewYork: Pinter. Rhyne, Elisabeth, and Linda S. Rotblatt. 1994. "What Makes Them Tick? Exploring the Anatomy of Major Microenterprise Finance Organizations." Monograph 9. ACCION International, Somerville, Mass. Ricklefs, Merle Calvin. 1981. A History of Modern Indonesia ca. 1300 to the Present. London and Basingstoke: Macnullan; Bloomington: Indiana University Press. Riedinger,Jeffrey M. 1994. "Innovation in Rural Finance: Indonesia's Badan Kredit Kecamatan Program." World Development 22 (3): 301-13. Rietveld, Piet. 1986. "Non-Agricultural Activities and Income Distribution in RuralJava." Bulletin of Indonesian Economic Studies 22 (3): 106-17. Robinson, Marguerite S. 1989. "The Framework and Development of the Bank Rakyat Indonesia Umt Banking System, 1970-1994." Two vols. Center for Poli- cy and Implementation Studies,Jakarta. .1991. "The Bank Rakyat Indonesia, 1970-1990. Strategies for Successful Rural Banking." Bank Rakyat Indonesia and Center for Policy and Implementa- tion Studies, Jakarta. .1992a. "The Role of Savings m Local Financial Markets:The Indonesian Ex- perience." GEMINI Working Paper 33. Development Alternatives, Bethesda, Md. . 1992b. "Rural Financial Intermediation: Lessons from Indonesia." HIID Development Discussion Paper 434. Harvard Institute for International Develop- ment, Cambridge, Mass. .1994a. "Financial Intermedliation at the Local Level: Lessons from Indone- sia." HIID Development Discussion Paper 482. Harvard Institute for International Development, Cambridge, Mass. .1994b. "Savings Mobilization and Microenterprise Finance:The Indone- sian Experience." In Maria Otero and Elisabeth Rhyne, eds., The New World of Microenterprise Finance: Building Healthy Financial Institutionsfor the Poor. West Hart- ford, Conn.: Kumarian Press. .1995a. "Indonesia: The Role of Savings in Developing Sustainable Com- mercial Financing of Small and Micro-Enterprises." In E. A. Brugger and S. Ra- japatirana, eds., New Perspectives on Financing Small Businesses in Developing Countries. San Francisco, Calif.: International Center for Economuc Growth and Fundaci6n para el Desarrollo Sostenible (FUNDES). .1995b. "Leading the World in Sustainable Microfinance: The 25th Anniver- sary of BRI's Unit Desa System." Bank Rakyat Indonesia,Jakarta. Bibliography 443 .1995c. "The Paradigm Shift in Microfinance: A Perspective from HIID." HIID Development Discussion Paper 510. Harvard Institute for International Development, Cambridge, Mass. .1995d. "Where the Microfinance Revolution Began: The First 25 Years of the Bank Dagang Bali." GEMINI Working Paper 53. Development Alternatives, Bethesda, Md. .1997a. "Introducing Savings in Microcredit Institutions:When and How?" CGAP Focus Note 8.World Bank, Consultative Group to Assist the Poorest, Washington, D.C. .1997b."Microfinance in Indonesia." UNESCO Courier (anuary): 24-27. .1997c. "Sustainable Microfinance." Based in part on a contribution by Richard H. Patten. In Dwight H. Perkins, Richard Pagett, Michael Roemer, Donald R. Snodgrass, and Joseph J. Stern, eds., Assisting Development in a Changing World: The Harvard Institutefor International Development, 1980-1995. Cambridge, Mass.: Harvard Institute for International Development. . 1997d. "Sustainable Microfinance at the Bank Rakyat Indonesia: The Economic and Social Profits." Paper presented at a workshop on institutional commercial microfinance for the working poor, sponsored by Bank Rakyat In- donesia and the U.S. Agency for International Development, 12-13 June, Jakarta. .1998a. "Microfinance:The Paradigm Shift from Credit Delivery to Sustain- able Financial Intermediation.' In Carl K. Eicher and John M. Staatz, eds., Agricul- tural Development in the Third World. 3d ed. Baltimore, Md.: The Johns Hopkins University Press. Reprinted with minor additions in Mwangi S. Kimenyl, Robert C. Weiland, and J. D.Von Pischke, eds., 1998. Strategic Issues in Microfinance. Alder- shot, England: Aldershot. .1998b. "Sugianto, 1939-1998." MicroBanking Bulletin 2. 13-14. Robinson, Marguerite S., and Donald R. Snodgrass. 1987. "The Role of Institu- tional Credit in Indonesia's Rice Intensification Program." HIID Development Discussion Paper 248. Harvard Institute for International Development, Cam- bridge, Mass. Robinson, Marguerite S., and Sudarno Sumarto. 1988. "Savings and Lending in the Umt Banking System." Center for Policy and Implementation Studies,Jakarta. Robinson, Peter. 1992. "The Village Banks of Bali (Lembaga Perkreditan Desa)." Fi- nancial Institutions Development Project (USAID and Government of Indone- sia). Development Alternatives,Jakarta. .1996. "An Evaluation of Lembaga Perkreditan Desa (LPD) Bali (The Vil- lage Banks of Bali)." Denpasar, Bali: Bank Pembangunan Daerah Bali. Robison, Richard. 1986. Indonesia: The Rise of Capital. Sydney: Allen & Unwin. .1987. "After the Gold-Rush:The Politics of Economic Restructuring in Indonesia in the 1980s." In Richard Robison, Kevin Hewison, and Richard Hig- gott, eds., Southeast Asia in the 1980s:The Politics of Economic Crisis. Sydney: Allen & Unwin. .1988. "Authoritarian States, Capital Owmng Classes, and the Politics of Newly Industrialhzing Countries:The Case of Indonesia." World Politics 41 (1). .1990. Power and Economy in Suharto's Indonesia. Manila:Journal of Contem- porary Publishers. .1992. "Indonesia: An Autonomous Domain of Social Power?" Pacfic Review 5 (4): 338-49. .1993. "Indonesia,Tensions in State and Regime." In Kevin Hewison, Richard Robison, and Garry Rodan, eds., Southeast Asia in the 1990':Authoritari- anism, Democracy and Capitalism. Sydney: Allen & Unwin. 444 The Microfinance Revolution: Lessons from Indonesia Rutherford, Stuart. 2000. The Poor and Their Money. New Delhi: Oxford University Press. Ruttan,VernonW, and Hans Binswanger. 1978. "Induced Innovations and the Green Revolution." In H. Binswanger andVW Ruttan, eds., Induced Innovation: Technology, Institutions and Development. Baltimore, Md.: The Johns Hopkins University Press. Sabirin, Syahril. 1999. "Recent Developments in the Indonesian Economy." Speech presented in March at Banque de France, Paris. Sadli, Mohammad. 1957. "Some Reflections on Prof. Boeke's Theory of Dualistic Economies." Ekonomi dan Keuangan Indonesia (December): 363-84. Sartono, Hudi. 1991. "The Rural Banking Unit in Bank Rakyat Indonesia (BRI)." Center for Policy and Implementation Studies,Jakarta. Sayogyo. 1982 [1973]. "Modermzation without Development in Rural Java."Journal of Social Studies 15: 32-67; 16: 75-87. Schnut, L.Th. 1991. Rural Credit between Subsidy and Market:Adjustment of the Village Units of Bank Rakyat Indonesia in Sociological Perspective. Leiden Development Se- ries 11. Leiden: Leaden University. . 1994. "A History of the 'Volkscredietwezen' (Popular Credit System in In- donesia, 1895-1935)." In Klass Kuiper, ed., Provisional Manualfor the Credit Busi- ness of the General Popular Credit Bank, by Thomas A. Fruin [1935].The Hague, Netherlands: Ministry of Foreign Affairs. Schreiner, Mark, and Jacob Yaron. 2001. Development Finance Institutions: Measuiring Their Subsidy. Washington, D.C.:World Bank. Schwarz, Adam. 1994. A Nation in Waiting: Indonesia in the 1990s. St. Leonards, New South Wales, Australia: Allen & Unwin. Seibel, Hans Dieter, and Uben Parhusip. 1998. "Microfinance in Indonesia, An As- sessment of Microfinance Institutions Banking with the Poor." Economics and Sociology Occasional Paper 2365. Ohio State University, Department of Agricul- tural Economics, Rural Finance Program, Columbus, Ohio. Seibel, Hans Dieter, and Petra Schmidt. 2000. "How an Agricultural Development Bank Revolutionized Rural Finance:The Case of Bank Rakyat Indonesia." Rural Finance Working Paper B 5. International Fund for Agricultural Development, Rome. Shaplen, Robert. 1969. Time Out of Hand. NewYork: Harper and Row. Shepard, B. M., A. T. Barrion, and J. A. Litsinger, eds. 1987. Friends of the Rice Farmer: Helpful Insects, Spiders, and Pathogens. Los Banos, Philippines: International Rice Research Institute. Sinaga, Rudolf. 1978. "Implications of Agricultural Mechanization on Production and Employment in Rice Areas ofWest Java." Bulletin of Indonesian Economic Stud- ies 19 (1): 53-67. Sinaga, Rudolf, and Wlliam L. Colfier. 1975."Social and Regional Implications of Agricultural Development." Prisma 1 (English edition): 24-35. Sinaga, Rudolf, and others. 1977. "Rural Institutions Serving Small Farmers and Laborers:A Case Study in the Village of Sukagalih, Garut Regency,WestJava." Rural Dynamics Series 1.Agro-Econonuc Survey, Bogor. Singarimbun, Masri. 1976. "Sriharjo Revisited." Bulletin of Indonesian Economic Stud- ies 12 (2): 117-25. Snodgrass, Donald R., and Richard H. Patten. 1991. "Reform of Rural Credit in Indonesia: Inducing Bureaucracies to Behave Competitively." In D. H. Perkins and M. Roemer, eds., Reforming Economic Systems in Developing Countries. Cam- bridge, Mass.: Harvard Institute for International Development. Soebadlo, Haryati, and C. du Marchle Sarvaas, eds. 1978. Dynamics of Indonesian His- tory. NewYork: North Holland. Bibliography 445 Soedjatmoko. 1960. An Approach to Indonesian History:Toward an Open Future. Ithaca, N.Y: Cornell University Press. .1967."Indonesia: Problems and Opportunities (I) and Indonesia and the World (II)." Dyason Memorial Lectures. Australian Institute of International Af- fairs, Melbourne. Soedjatmoko, and others, eds. 1965. Introduction to Indonesian Historiography. Ithaca, N.Y.: Cornell University Press. Soeharto. 1991. My Thoughts, Words, and Deeds:AnAutobiography.Jakarta: Citra Lam- toro Gung Persada. Steinwand, Dirk. 2001. The Alchemy of Microfinance: The Evolution of the Indonesian People's Credit Banks (BPR) to 1999 and a Contemporary Analysis. Berlin:Verlag fur Wissenschaft und Forschung. Stern,Joseph J. 2000. "Liberalization and Growth before 1997." Paper presented at the National Bureau of Economic Research, 15 September, Cambridge, Mass. . Forthcoming. "The Rise and Fall of the Indonesian Economy." Harvard University, John F. Kennedy School of Government, Cambridge, Mass. Stightz,Joseph E., and ShahidYusuf, eds. 2001. Rethinking the Asian Miracle. New York: Oxford University Press. Stoler, Ann L. 1978. "Garden Use and Household Economy in Rural Java." Bulletin of Indonesian Economic Studies 14 (2): 85-101. Strout,Alan M. 1985.'"Managing the AgriculturalTransformation on Java." Bulletin of Indonesian Economic Studies 21 (1): 62-80. Sugianto. 1989. "KUPEDES and SIMPEDES." Asia Pacfic Rural Finance Uuly- September): 12-14. .1990a. "Development of Rural Agricultural Financial Policy and the Progress and Benefits of BRI's KUPEDES Program." Bank Rakyat Indonesia, Jakarta. .1990b. "Transaction Costs at Small-Scale Banks below the Branch Level." Bank Rakyat Indonesia,Jakarta. Suglanto, and Marguerite S. Robinson. 1998. "Sustainable Microfinance as Devel- oped by the Bank Rakyat Indonesia." Bank Rakyat Indonesia,Jakarta. Sugianto, Satriyo Purnomo, and Marguerite S. Robinson. 1993. Bunga Rampai Pem- biayaan Pertanian Pedesaan. Jakarta: Institute Bankir Indonesia. Suharto, Pandu. 1985. Mengenai Bank Perkreditan Rakyat di Indonesia. Jakarta: Lem- baga Pengembangan Perbankan Indonesia. .1988. Sejarah Pendirian Bank Perkreditan Rakyat. Jakarta: Lembaga Pengem- bangan Perbankan Indonesia. 1996. "100 Tahun BPR di Indonesia."Jakarta: Infobank. Sujono, Irlan. 1976. "Growth and Distribution: Changes of Incomes in Paddy Farms in Central Java, 1968-74." Bulletin of Indonesian Economic Studies 12 (2): 80-89. Sukarno, Imam. 2000. "Policy and the Regulatory Environment for Microfinance in Indonesia." Paper presented at the World Bank Institute's regional workshop on microfinance for East Asia and the Pacific, 21 March, Manila. .2001 "Commercialization of Microfinance." Bank Indonesia,Jakarta. Sundrum, R. M. 1986. "Indonesia's Rapid Economnc Growth, 1968-81." Bulletin of Indonesian Economic Studies 22 (3): 40-69. .1988. "Indonesia's Slow Economic Growth, 1981-1986." Bulletin of Indone- sian Economic Studies 24 (1): 37-72. Sundrum, R. M., and Anne E. Booth. 1980. "Income Distribution in Indonesia: Trends and Determunants." In R. Garnaut and P. McCawley, eds., Indonesia: Dual- ism, Growth and Poverty. Canberra: Australian National University, Research School of Pacific Studles. 446 The Microfinance Revolution: Lessons from Indonesia Sutoro, Ann Dunham, and Roes Haryanto. 1990. KUPEDES Development Impact Survey. Jakarta: Bank Rakyat Indonesia. Suyono, Haryono. 1997. "Indonesian Family Empowerment to Alleviate Poverty." Indonesia Times, 6 September, p. 7. Tilly, Richard. 1994. "A Short History of the German Banking System." In Man- fred Pohl, ed., Handbook on the History of European Banks. Frankfurt: Main. Timberg, Thomas A. 2000. "What We Know About Small and Medium Enterprise (SME) Finance in Indonesia." International Monetary Fund,Jakarta. Tinuner, C. Peter. 1975. "The Political Economy of Rice in Asia: Indonesia." Food Research Institute Studies 14 (3): 197-231. .1987. The Corn Economy of Indonesia. Ithaca, N.Y.: Cornell University Press. U.S. Central Intelligence Agency. 1968. Indonesia-1965: The Coup That Backfired. Washington, D.C. Van Nell, Robert. 1984. The Emergence of the Modern Indonesian Elite. Dordrecht, Holland: Foris Publications. Vatikiotis, Michael. 1993. Indonesian Politics under Suharto. London Routledge. Versluysen, Eugene. 1999. Defying the Odds: Bankingfor the Poor. West Hartford, Conn.: Kumarian Press. Wardhana, Ali. 1994a. "Designing and Implementing Structural Adjustment Pro- grams." Mimstry of Finance,Jakarta. .1994b. "Financial Reform: Achievements, Problems, and Prospects." In Ross H. McLeod, ed., Indonesia Assessment 1994. Singapore and Canberra: Insti- tute of Southeast Asian Studies and Australian National University. 1997a. "Indonesia Today." Speech presented 20June,Jakarta. 1997b. "Sustaining Economic Growth in Indonesia."Address to the Con- ference on Sustaining Economic Growth in Indonesia, 17 December. .1998a. "Economic Reform in Indonesia: The Transition from Resource Dependence to International Competitiveness." In Henry S. Rowen, ed., Behind East Asian Growth: The Political and Social Foundations of Prosperity. London and NewYork: Routledge. .1998b. "Overcorming the Current Econornic Downturn."Address to the Capital Market Conference, 25 August, Jakarta. Warren, C. 1993. Adat and Dinas: Balinese Communities in the Indonesian State. Kuala Lumpur: Oxford University Press. White, Benjamin. 1976. "Population, Involution and Employment in Rural Java." Development and Change 7 (3): 267-90. .1979. "Pohtical Aspects of Poverty, Income Distribution and Their Mea- surement: Some Examples from Rural Java." Development and Change 10 (1): 91-114. . 1983. "The Agricultural Involution and Its Critics:TwentyYears After." Bul- letin of ConcernedAsian Scholars 5 (2): 18-31. White, Benjamin, and Makall. 1979. "Wage Labor and Wage Relations in Javanese Agriculture: Some Prelimnary Notes from the Agro-Economic Survey." Institute of Social Studies,The Hague, Netherlands. Wiradi, Gunawan. 1978. "Rural Development and Rural Institutions." Rural Dy- namics Series 6. Agro-Economic Survey, Bogor. Wirach, Gunawan, and Chris Manning. 1984. "Landownership, Tenancy and Sources of Household Income: Community Patterns from a Partial Recensus of EightVil- lages in Rural Java." Rural Dynamics Series 29. Agro-Econonuc Survey, Bogor. Wolff, HenryW 1893. People's Banks:A Record of Social and Economic Success. Lon- don: P S. King. Bibliography 447 World Bank. 1980. World Development Report 1980. NewYork: Oxford University Press. 1982. World Development Report 1982. NewYork: Oxford University Press. .1984a. Agriculture Credit: Sector Policy Paper. 2d ed. Washington, D.C. .1984b. Employment and Income Distribution in Indonesia. World Bank Coun- try Study.Washington, D.C. . 1987. Indonesia: Rural Credit Sector Review. Two vols. East Asia and Pacific Regional Office, Washington, D.C. - 1993. 7the EastAsian Miracle: Economic Growth and Public Policy. NewYork: Oxford University Press. .1996a. "A Worldwide Inventory of Microfinance Institutions."World Bank, Sustainable Banking with the Poor project, Washington, D.C. .1996b. "KUPEDES: Indonesia's Model Small Credit Program." OED Pre- cis 104. Operations Evaluation Department,Washington, D.C. .1996c. World Development Report 1996: From Plan to Market. NewYork: Oxford University Press. .1997a. "An Inventory of Microfinance Institutions in East Asia and the Pa- cific." Sustainable Banking with the Poor project,Washington, D.C. 1. 1997b. "An Inventory of Microfinance Institutions in East, Central, and South Africa." Sustainable Banking with the Poor project,Washington, D.C. 1997c. "An Inventory of Microfinance Institutions in Latin America and the Caribbean." Sustainable Banking with the Poor project,Washington, D.C. .1997d. "An Inventory of Microfinance Institutions in South Asia." Sustain- able Banking with the Poor project,Washington, D.C. .1997e. "An Inventory of Microfinance Institutions in West Central Africa." Sustainable Banking with the Poor project,Washington, D.C. 1997f. World Development Indicators 1997. CD-ROM.Washington, D.C. 1997g. World Development Report 1997: The State in a Changing World. New York: Oxford University Press. .1998a. East Asia: The Road to Recovery. Washington, D.C. 1998b. "Indonesia in Crisis: A Macroeconomic Update." Executive Sum- mary of the 1998 Country Economic Memorandum.Washington, D.C. - % 1 998c. Regional Databasefor East Asia. Statistical Information and Analysis System.Washington, D.C. 1998d. World Development Indicators 1998. CD-ROM.Washington, D.C. 1999. World Development Report 1998/99: Knowledgefor Development. New York: Oxford University Press. - .2000. World Development Report 1999/2000: Entering the 21' Century. New York: Oxford University Press. .2001. World Development Report 2000/2001:Attacking Poverty. NewYork: Oxford University Press. .2002. World Development Report 2002: Building Institutionsfor Markets. New York: Oxford University Press. World Bank News. 1996. "KUPEDES: A Success Story in Microlending."Washing- ton, D.C. Wright, Graharn A. N. 2000. Microfinance Systems: Designing Quality Financial Services for the Poor. London and NewYork: Zed Books; Dhaka: University Press. Yaron,Jacob. 1992a. Assessing Development Finance Institutions:A Public Interest Analy- sis. World Bank Discussion Paper 174.Washington, D.C. .1992b. Successful Rural Finance Institutions. World Bank Discussion Paper 150.Washington, D.C. Yaron,Jacob, and McDonald Benjamin Jr. 1997. "Developing rural Financial Mar- kets." Finance and Development 34 (4): 40-43. 448 The Mlcroflnance Revolution: Lessons from Indonesia Yaron,Jacob, McDonald Benjamin Jr., and Stephanie Charitonenko. 1998. "Pro- moting Efficient Rural Financial Intermedcation." The World Bank Research Ob- server 13 (2): 147-70. Yaron,Jacob, McDonald Benjamin Jr., and Gerda Piprek. 1997. Rural Finance: Issues, Design, and Best Practices. ESD Monograph 14. Washington, D.C.: World Bank, En- vironmentally and Socially Sustainable Development Studies. Yeats,Wilhiam Butler. 1921. "The Second Coming." In Michael Robartes and The Dancer. Dundrum: Cuala Press. Zainu'Ddin, A. 1970. A Short History of Indonesia. NewYork: Praeger. Bibliography 449 Index Abri (Angkatan Bersenjata Republhk Indonesia, armed forces), government credit to agricultural households, 100-01, 40-43, 80n21, 80n31 See also Golkar; Soeharto 140n17 1965 coup attempt and, 29-30 green revolution and, 90-92 dwifungsi (dual function) doctrine of, 29, 40-41 varieties, 88-89 Golkar and, 43 AlgemeeneVolkscredietbank (AVB), 89, 99 human rights and, 30, 58-59, 61, 67-68, 74 BRI unit desas and, 179 independence struggle and, 27 Fruin and, 174 Megawati and, 72-73 merger with BRI, 175 separatist movements and, 61, 66-68, 80n33 Anderson, Benedict, 33 Soeharto and, 40-44 Arief, Kamardy, xlvi, 226, 258n3, 400 Wahid and, 61, 64 on BRI culture, 307-08, 311 Accounting and audit on changing the culture of the BRI unilt desa system, 229 1997 financial crisis and, 48, 51-52 Arnold, Wayne, 372 Bank Shunta Daya, 128 Asian Development Bank, 49, 71, 365-66,369,371-72 BKD, 408n8 Asia-Pacific Economuc Cooperation (APEC) forum, 41 BKK/CJ, 118 "AssessingVulnerability to Financial Crisis: Evidence from In- BPR, 112 donesia" (Kenward), excerpts from, 54-55 BRI urut desas, 111,218,236-37,244,299,301,312,315, Association for Social Advancement (ASA) (Bangladesh), xXii, 398-99 xxxvii corporate culture and, 380-81 Association of Indonesian National Private Banks comparison of Indonesian and international standards KUK funds and, 104 for, 377-78 Association of Indonesian Rural Banks IBRA, 379 KUK funds and, 104 state audit of Bank Indonesia's emergency hquidity loans, Association of Southeast Asian Nations (ASEAN), 41 70, 373-75 sustainable mucrofinance and, 354 Badan Koordinasi Keluarga Berencana Nasional. See BKKBN Aceh special region, 66, 74 Badan Kredit Desa. See BKD Afdeelingsbank (District Bank), 93-94 Badan Kredct Kecamatan of Central Java. See BKK/CJ Agriculture. See also BIMAS; Brown planthopper; Insecticide Badan Kredit Kecamatan of South Kalhmantan. See BKK/SK damage to rice production; KUPEDES, KUT; Unit desas of Badan Perencanaan Pembangunan Nasional. See BAPPENAS BRI Badan Usaha Milik Daerah (BUMD, Provincial Government- in four villages, 194-208 owned Enterprise), 118-19 Fruin's analysis, 89-90 451 Badan Usaha Mulk Negara (BUMN, State-owned Enterprise), BDB (Bank Dagang Bali), xxv, xxx, xl, xll, xliv, xlviIl, 85, 105 366-67, 407n4. See also Oka, I Gusti Made, Oka, Sri Adnyanu Baird, Mark, 69, 370, 378-79, 381, 408n7 balance sheet (1970-96),152 Baker, Susan L., 407n2, 407n5 balance sheet (1996-2000), 386 Bank Bah, 375, 379, 407n4 branches opened after PAKTO 88,101-02 Bank Central Asia, 377-78 customer relations, 155-56 Bank Dagang Bal. See BDB development and performance (1970-96), 150-55 Bank Dagang Negara, 150 development and performance (1996-2000), 384-88 Bank Danamon, 51 high-income savers, 163 Bank desas, 98-99,140n16 influence on BRI unit desas, 232, 235, 299 Bank Duta, 51 loan products, 156-57 Bank Ekspor-Impor Indonesia, 118-19,141n28 lotteries, 158 Bank Indonesia (central bank) lower-muddle- and middle-income savers, 162-63 BRI unit desas and, 180,187,220,232,234, 260nl9, 266 low-income savers, 161-62, 164nl0 financial reforms and, 37-38, 229 mobile savings teams, 158-59,164n7, 164n8, 187 IBRA and, 379 organization and management, 156-57 investigation of category A banks, 376 outstanding loans (1970-96), 151-53, 164n5 LDKP estimates by, 141n24 outstanding loans (1996-2000), 384-86 PAKJAN limits on subsidized credit, 38 overview, 144-46, 163nl role in rural credit, 98-103, 106-07, 119, 132-33, 135, philosophy, 147 137, 141n21, 141n24, 344, 347,399 profitablity during financial crisis, 114,384-88, 406,407 regulations, 276, 277,299, 301 profit and loss statement (1970-96), 151, 164n4 state audit of emergency liquidity loans, 373-75 profit and loss statement (1996-2000), 385 supervision, 380-81 public relations, 159-60 Banking Law (1992), 84 savings (1970-96), 151-53 BPR regulation under, 99, 102-04 savings (1996-2000), 385-87 Bank Mandirl, 377 savings customers, 160-61, 164n9 Bank Pasar Umum (BPU), 149-50,164n4 savings products, 157-58 Bank Pembangunan Daerah. See BPD scholarship program, 155 Bank Pembangunan Indonesia (BAPINDO), 51, 377 size distribution of loans and savings, 154 Bank Perkreditan Rakyat See BPR Benjamun, MacDonald,Jr., 361, 362n3 Bank Permagaan Umum, 149 Bergier,Jean-Franqois, 95 Bank Rakyat Indonesia. See BRI BIMAS (government-subsidized agricultural credit program), Bank Shinta Daya, 127-30,134,138 cv-}xxvi, 179-84, 210n2, 211nl6, 212n26 See also KUT Banks and banking. See also Banking Law; IBRA; KUPEDES, amount of outstanding loans (1970-84), 182 Loans; Rural financial institutions and programs; Savings; en- area covered (1970-84), 183 tries by bank name BRI unit desas and, 166,169, 220 BPR links, 112 creation, 178, 211nl3-15, 213n38 central bank supervision, 380 credit repayment, 170-71, 182, 184, 194-208 European background, 95-96 crop destruction by insecticides in BIMAS packet, 191-93, financial crisis and closures, 373, 377 212n27, 212-13n29 government recapitalization after financial crisis, 69, end of, 208-10 376-78,381-82, 407n5 examples of in four vlllages, 194-208,213n30,214n40-43 IBRA rehabilitation plan, 376-78 GERTAK loan collection program, 184,202 loans and savings in financial system (1995), 108-09 incompatible goals for rice production under, 188 moral hazard, 51 input packets, 188, 199, 206-07, 212n26 proliferation, 52, 84,107 Kredit Mini and Kredit Midi loans compared with, reforms, 37-38, 57, 375-78 185-86, 212n20-22 restructuring of corporate debt, 378-80 KIJPEDES compared with, 220 rural credit (1985), 100 loan rescheduling and forgiveness, corruption and, 190, BAPPENAS (Badan Perencanaan Pembangunan Nasional, Na- 194-208 tional Development Planning Agency), 94-95, 106 loans to rice fariners, 180-83,194-208 Barclays Bank, 378 number of outstanding loans (1970-84), 183 participation, 194-208 452 The Microfinance Revolution: Lessons from Indonesia performance, 167-68,187, 208-10 development, 109-10,141n25 poor rice farmers and, 189-90, 205-06 influence on KUPEDES, 232, 235 reasons for failure, 167-72, 187-210 profits and losses, 112-13, 349 Results inVillage C,WestJava, 198-200 public and private, 111-14 Results in Vllage G, East Java, 195-98 umts and clients compared with BRI units and clients, Results in Village P, South Sumatra, 202-06 109-10 Results in Vllage R, South Sulawesi, 200-02 various definitions, 94, 98-99 service delivery program, 179-80 Bresnan,John, 44, 46, 58-59, 78n3, 79nl7, 80n22, 80n36 Bina Swadaya (Self-Reliance Development Foundation), 106 BRI (Bank Rakyat Indonesia), xlvi-xlvii, 324, 362n1, 366. See BKD (Badan Kredit Desa,Village Credit Organization), XXxI, also Unit desas of BRI xli, 2, 12, 94, 99, 103-04, 114,115, 343-52, 363nl9, 364, Asset and Liability Coordmnating Committee, 298 366-67. See also BPR, BRI branch offices, 310-15,317-18,323-25,327-29,335-36, accounting, 408n8 376 Banking Law (1992) effects on, 344 branch staffing, 328, 362n13 deposits at BRI, 345-46 branch supervision of units, 314-15, 327-28, 362n6, financial crisis and, 382-84, 406-07 362n12 history, 343-44 Cooperative and Program Credit Division, 313-14 interest rates, 345-46 creation, 96-97 Ine of credit from BRI, 346 failure during crisis, 375-76 loan defaults and arrears, 345-46, 384 financial deregulation (1983) and, 37 P4K compared with, 350-52 head office organization and management, 324,399-400 performance, 108,110-13, 345, 383 hustory (1895-1970), 172-75, 210n3, 211n4, 211n10, services, 344-45, 382-84 211nll, 388-402, 410-13 stablity during crisis, 402-05 interdivisional coordination, 355-58, 400 supervision by BRI, 99, 343-47, 382-84 losses compared with unit desa profits, 375-76 sustamability, 346-47 losses during financial crisis, 317-18 units and clients (1998), 110 microcredit evolution in (1970-96), 223 village ownership, 99, 110-11 Ministry of Finance as owner, 362n2 BKKBN (Badan Koordinasi Keluarga Berencana Nasional, Na- old and new cultures, 309-311 tional Family Planning Coordinating Board), 39, 80n29, organization and management,306-32,355-58,400-02 105,134. See also KUKESRA;TAKESRA P4K, 347-52 BKK/CJ (Badan Kredit Kecamatan of CentralJava, Subdistrict recapitalization and restructuring, 377, 390 Credit Organization), 113, 115-17 regional offices, 314-15, 327, 335, 337 distribution of profits, 141n30 savings and deposits, 390, 408n12 early loan programs, 117-18 subsidized and commercial mucrocredit programs, 222-24 influences on KUPEDES, 235 347-52, 358 lessons, 119-20 supervision of BKDs, 99, 343-47, 382-84 as provincial government enterprise, 118-19 Training Division, 339-42 BKK/SK (Badan Kredit Kecamatan of South Kalimantan, Sub- "Bank Rakyat Indonesia (BRI) (Case Study)" (Maurer), excerpts district Credit Organization) from, 301 development, 120-23, 141n31 Brown planthopper, 192-93,212n27. See also Insecticide dam- lessons, 123-25 age to rice production Boediono, 73, 77, 399 BUD (Business Unit Desa). See Umt desas of BRI Boeke,J. H, 173, 21 1n6, 227, 399 Budianto, Eko Santoso, 374 Bowen,John R., xlix, 212n30 Building a Modern Finanaal System. The Indonesian Experience (Cole BPD (Bank Pembangunan Daerah, Regional [also Provincial] and Slade), excerpts from, 31, 103 Development Bank), 107,117,119-22,124-26,133, 141n27, 150 Camdessus, Michel, 50 BPR (Bank Perkreditan Rakyat, People's Credit Bank), 84-85, Caragata,Warren, 69-70, 381 173. See also Bank Shinta Daya, BKK/CJ; BKK/SK Catholic Relief Services, 107 average loans and deposits (1997), 111 Center for Policy and Implementation Studles (CPIS), xlvii, 189. Banking Law (1992) and, 102-04 See also Development Program Implementation Studies; BRI unit desas compared with, 110-11 Ministry of Finance; Unit desas of BRI capital requirements after 1999 regulatory changes, 103 Rural Savings Group, 304n3 Index 453 Rural Savings Reports, 269, 304n2-3, 304n6-7, 304n13, CPIS. See Center for Policy and Implementation Studies 304nl7, 304n19 Credit. See BDB; BIMAS; BKD; BPD; BPR; BRI; KUPEDES; unit desa savings program, planning and research by BRI KUT; Interest rates; Loans; Rural financial institutions and and, 267-96, 304n4, 304n9 programs; Subsidczed credit programs; Unit desas of BRI CGAP. See Consultative Group to Assist the Poorest Credit unions, European, 172, 21 1n5 Chaves, Rodrigo, 132 Chinese, ethnic Danoesputro, Mariyanto, 102 conglomerates and, 45 Dapice, David 0., 259nlO economuc role, 31-32, 45-46 Darusman, Marzuki, 63, 64 financial crisis and, 48 de Jonge, B. C., 27 government credit programs of 1990s and, 102 Demographics, 25-27, 78n9, 78nll, 86-87 moneylending practices, 172 Demonstrasi Massal Swa Sembada Bahan Makanan (DEMAS, victims of attempted coup (1965-66), 30 Action Research Pilot Project for Rice Intensification), victims of riots (1998), 48 211n13 Cole, David C., xlix, 31-32, 46, 52, 55, 103, 407n2 Deposito Berjangka (time deposit instrument) Collateral at BDB, 152-54,158,160, 385-87 BDB loans, 157 at BPRs, 109, 114, 119, 141n26 KUPEDES loans, 243-44, 260-61n23 at BRI unit desas, 270-71, 276-77, 280, 284-85, 292, Computerization, unit desas and, 313 296-98, 302-03, 391-94 Conglomerates at district banks (1926), 97 corporate debt restructuring and, 378-82 Development Program Implementation Studies (DPIS), xlvll, corruption under Soeharto and, 45-46 189, 204, 212n29, 213n31 government credit programs of 1990s and, 102,105 field studies on BIMAS by, 194-210, 210nl, 211n17, IBRA and, 69-70, 373-78 213n37, 213n39, 214n40-43, 214n45-48, 215n49, technocrats and, 46 215n51-52, 215n54, 258n2 Conroy,John D., 132 Harvard Institute for International Development project, Constitution. See also Sukarno 210nl, 212n29, 229-32,263 of 1945, 28, 40 Rice Intensification: Report 2, 210nl, 229-30, 263 of 1949 and 1950, 28, 79nl3 Dewan Perwakilan Rakyat (DPR, Indonesian parliament), 41, Constitutional Assembly (Konstituante) (1955), 28 62-64 MPR and, 80n32 District banks, 97 proposal for independent commission on, 74 nonperforming loans during world depression, 140nl5 Consultative Group to Assist the Poorest (CGAP), 299-301 DPIS. See Development Program Implementation Studies Consumer price index, financial crisis and, 370-71,382,410-13 DPR. See Dewan Perwakilan Rakyat Cooperative banks, Germany, 96,140n13 Dual economy, theory of, 173, 211n6, 227 Corporate bankruptcies and debt restructuring Dutch colomal rule, 26-27 financial crisis and, 370, 372-75 BRI, district banks, and lumbung desas started by, 96-97 IBRA and, 376, 378-80 ethnic Chinese and, 31 JITF and, 378-80 lack of education for Indonesians under, 27 nonperforming loans of politically favored borrowers New Ethical Policy of Queen Wilhemrna, 172 and, 374 plantation agriculture and, 88 Corporate culture, lack of transparency and accountability in, rural financial institutions, 93-94 380-81 Dwifungsi (dual function) doctrine of military, 29, 40-41 Corruption BIMAS loan delivery and, 190,197, 316 EastAsian financial crisis (1997), 20, 47-48, 53, 81n40. See also conglomerates and, 45-46 Indonesian crisis courts and, 380 East Timor independence, 20, 76 IBRA and, 372, 379 Habible and, 58 in LKPs of Nusa Tenggara Barat, 127 massacres after vote for, 59 Megawati on, 74 Megawati acknowledgment, 74 nonperforming loans and, 373-75 political reforms leading to, 58 and subsidized loans, 316 unrest over, 67-68 under Soeharto, 44-46, 48-53 Wiranto and, 61 after Soeharto, 57, 63,75, 373-75, 378-81 454 The Microfinance Revolution: Lessons from Indonesia Economy. See also Exchange rates; Indonesian crisis; Inflation on collateral, 244 1966-96,34-40 Popular Credit System and, 174 1996-2000,47-58,73-74,81n43,365-73 Furman,Jason, 53, 81n41, 407n2 compared with other developing country economues, 39-40 GDP (gross domestic product), 21,34,48,57,68-69,78nl,370, economic incdcators, 35,410-13 404,407nl economuc reform program with IMF and other donors, Geertz, Clifford, 91, 139n5 49-50,57-58,68-71,365-73 General banks (bank umum), Banking Law (1992) and, 102 Elections Geographic diversity, 86,139nl Constitutional Assembly (1955), 28 Gerakan Serentak (GERTAK, government loan repayment controlled during Soeharto regime, 43 program), 184,202 general (1955), 28 German Agency for Technical Cooperation (Gesellschaft fur general (1999), 20-21, 58-59, 80n37 Technische Zusammenarbeit, GTZ), 106,132 Pancasila and, 33 Giro accounts pohtical pressures on lending policy and, 121 BDB, 152-54,387 of president by MPR, 43, 59-60,71, 79nl9, 366 BRI, 270,277,280,292,298 reformns, 58, 62 Golkar (golongan karya, functional groups), 41, 43-45, 58-60, of vice president by MPR, 60,72 62, 72, 80n34, 80n35, 102, 375. See also Abri; Soeharto Estate (plantation) agriculture, 88 Gonzalez-Vega, Claudio, 132,189 Ethnic conflicts, 30-32, 66-68 Gotong Royong (Cooperation) Cabinet, of Megawati, 73 Pancasila tolerance doctrine and, 34 Government mucrofinance initiatives, 102-07,225-35. See also Exchange rates BIMAS; BPR; Subsidized credit programs; Umt desas ofBRI financial crisis and, 51, 53,365,370,371,382,410-13 Bank Indonesia and, 106-07 mucrobanking clients and, 382, 404 for BPRs, 102-04 under New Order government, 36 BRI mucrocredit programs and, 222, 224 nonperformung corporate loans and, 373-75 corporate and wealthy incdviduals' levy for subsidized cred- Exports, 29,34-37, 80n24, 91 it programs, 105 External debt, financial crisis and, 3, 50-54, 365, 379 IDT, 105-06 KUK, 104 Fertilizer, BIMAS introduction of synthenc, 198, 200, 204, Government, structure of, 41, 59 205,212n37 Grameen Bank (Bangladesh), xxxv Fidler, Peter J., xli, I Green, Marshall, 79n17 Financial intermediation. See also BDB;BPR; Unit desas of BRI; Green revolution, 83,90-91,184,191-94. See also Agriculture; entries by bank name BIMAS; Rice production by rural financial institutions, 95, 137-38 debate over benefits, 90-93 as nucrobanking strength, 397, 407 in four villages, 194-208 Financial reforms, xxvi-xxvii, 37-38, 375-80 Guided Democracy (Sukarno), 28-29 after 1997 crisis, 20, 57-58,372, 375-80 Gus Dur. See Wahid PAKDES I, 37,80n25 Guterres, Eurico, 67 PAKJAN, 37-38, 80n27, 80n28 PAKMAR, 99 Haas,Wilhelm, 140n13 PAKTO 88,37-38,80n26,98-99,101-02,150 Habibie, B.J., 20, 50, 80n37 technocrats and, 35, 46 accountability speech to MPR rejected, 59 Financial sector, weaknesses in, 50-56, 365-66, 369-75 financial crisis and, 372 Fischer, Stanley, 71 as president (1998-99), 58-60,366 Fish production, 88-89,139n3 political reforms, 58-59 Foster Parents Plan International (FPPI), unat desa savings ac- Wahid and, 60 counts for, 290-91 Hamzah Haz, 72 Fox,James J., 87, 192, 212nl9, 212n30, 259nl0 Harvard Institute for International Development (HIID). See Freeport-McMoran Copper & Gold, 66 also Center for Policy and Implementation Studies; Devel- Friedman,Thomas L., 77 opment Program Implementation Studies; Ministry of Fi- Fruin, Thomas A., 89-90,174-77,211 n8, 21 1n9, 399 nance; Umt desas of BRI BRI unit desas and, 179 BRI unit desas' International Visitors Program and, Credit analysis by, 89-90,176-77 260nl7,362n1l Index 455 KUPEDES development and, 221, 263 economic reform program with IMF and other donors, Ministry of Finance and, 210nl, 212-13n29, 213n31, 49-50, 57-58,68-71,365-73 229-32, 258n2, 304n6 effects, 32, 48-50, 365-75 rice intensification report (1983), 229-30 foreign debt and, 69, 307, 321 Rural CreditWorking Group, 231-32, 260n22 nonperformmg corporate loans and, 373-75 Rural Savings Group, 304n3, 304nl9 poverty during, 93 Hatta, Mohanmuad, 27 roots, 20, 50, 55-56,370 Healy,Tim, 378 warnung signs, 52-55 "History Present Situation, and Problems of theVllUage Cred- Indonesian Democratic Party. See PDI it System (1897-1932)" (Fruin), excerpts from, 174-75, Indonesian Democratic Party-Struggle. See PDI-P 210n3, 211n9 Indonesian National Education Party, 27 Hook, Richard M., 260n17, 261n33, 305n23 Indonesian National Party. See PKI Hornick, Robert N., 380 Inflation, 57, 259nl 1, 365, 370, 404 Hulp-en Spaar Bank der Indlandsche Bestuurs Ambtenaren Infrastructure development, technocrats and, 35 (Support and Savings Bank for Domestic Civil Servants), 172 INMAS, cash program for farmers' nputs, 21 1 n 1 4 INPRES Desa Tertinggal program. See IDT IBRA (Indonesian Bank RestructurngAgency), 57,69-70,366, Insecticide damage to rice production, 191-93. See also Brown 374, 376-81 planthopper and Bank Bali, 374-75 inVillage C,WestJava, 199 bank rehabilitation plan, 376-78 in Vilage G, East Java, 197 governance structure, 69-70, 379 national pest management strategy for, 191-93 political difficulties, 372, 374-75, 378-81 relation to severe decrease in Indonesian rice gene pool, "Idea of Power injavanese Culture,The," (Anderson), excerpts 192-93, 212n29 from, 33 Interest rates IDT (INPRES Desa Tertinggal program, presidential instruc- Bank Shinta Daya loans, 129 tion for"left behund" vllages), 105-06, 136-37,212n31,273, Bank Shinta Daya savings products, 129 305nl4 BDB loans, 157, 384-85 Income Generating Program for Small Farmers and Fisherman. BDB savings products, 158, 327-28 See P4K BIMAS loans, 180 Impact studies, mucrofinance, xxxx BKD loans, 345-46, 383 Indonesia. See also Abri;Agriculture; Banks and banking; Cor- BKD savings products, 383 ruption; Dewan PerwakUlan Rakyat; Economy; Habibie; In- BKK/SK loans, 122 donesian crisis; Megawati Soekarnoputri; MPR; Soeharto, BRI unut desa KUPEDES loans, 7-8, 240-41, 259n12, Sukarno,Wahid 260nl9, 260n20-21, 312, 351, 360-61,394 compared with other emerging economies, 39-40, 80n30 BRI unit desa savings products, 272-73,275-79,298,393 environment, demographucs, early history, independence deregulation (1983), 7, 37 struggle of, 25-27, 78n9-11 IDT loans, 106 history, 18-22, 46-47, 78n3 KUKESRA loans, 105 lessons on commercial nmcrofinance from, 402-07 informal moneylenders'loans, 101, 140n19 rural development, 88-93 LKP loans, 126-27 rural finance, 93-139 Mitra Karya loans, 131 simultaneous implementation of opposite approaches, P4K loans, 348-49 233, 259nI3 PHBK loans, 133-35 Indonesian Communist Party. See PNI Internationale Nederlanden Bank (ING Barings), 156,164n6 Indonesian crisis, 47-56, 81n40, 369-73, 407n2. See also East International Fund forAgricultural Development (IFAD), 347 Asian financial crisis International Monetary Fund (IMF), xxix, 20, 22, 49-50, bank closures and, 373 365-66, 371-72 bank paralysis and, 375 International Rice Research Institute (IRRI), 184, 191-92 bank reforms and, 375-78 International Visitors Program of BRI unit desas. See Unit challenges from, 380-82 desas of BRI commercial mucrofinance institutions' stabilhty during, IPTW (prompt payment incentive) of KUPEDES, 242-43 12-14,85,114,139,364-65,367-68,382-407,410-13 Irian Jaya See Papua corporate debt restructuring and, 375, 376-80 Irish loan funds, 96, 140n12 456 The Micrefinance Revolution: Lessons from Indonesia Java, 23-25, 30, 33, 78n2, 86-92, 94, 97-98 financed by unit desa savings, 230-35,262-67,270,274, Jenkins, David, 79n15 279,293-96, 304, 410-13 JITF aakarta Initiative Task Force), 379-80 Fruin influence, 89-90 Johnston, Don E.,Jr., 260n17, 408n9 geographic coverage, 255-56 J. P Morgan, 378 household living standards and, 251-52 Judiciary, 65, 67-68, 71, 379-80, 383 implementation, 221-22, 237-46 Indonesia's enabling conditions, 219,221 Kalimantan, 48-49,67,120-25, 141n31-33 interest rates, 240-41, 259n12, 260n19, 260n20,260n21, Kartasasmita, Syafiuddin, 65-66, 71 312,351,360-61,394 Kenward, Lloyd R., 36, 53, 54-55, 407n2 KreditMidi and Kredit Mim influence, 184-86,232,235, Kenya Rural Enterprise Programme (K-REP), xlviii 237 Kern,James R., 305n19 lessons, 397-98, 402-03, 405 KIK (Kredit Investasi Kecdl), xxvi,101, 140n18 loan application process, 244-45 Kindleberger, Charles P., xxxLx loan product, 239, 260n17 KMKP (Kredct Modal Kerja Permenen), xxvi, 101, 140n18 loan purposes, 239-40, 260n18 Koentjaraningrat, R. M., 87 loan sizes, 240, 241 Koperasi Unit Desa See KUD loans, number and value (1984-96), 253 Kredit Midi, 184-86, 212n20-22 loans, number and value (1996-2000), 389-91, 410-13 influences on KUPEDES, 232, 235, 237 loss ratios (1984-96), 253, 254-55, 395 Kredit Mini, 184-86, 212n20-22 loss ratios (1996-2000), 388-90, 410-13 influences on KUPEDES, 232,235, 237 maturities and repayment terms, 241-42 initial financing for KUPEDES from capital of, 232,234 outreach, 254-55, 388-96, 410-13 Kredit Umum Pedesaan See KUPEDES overview, 216-21 Kredit Usaha Tani. See KUT performance (1984-96), 252-58, 322, 360-61 KUD (Koperasi Unit Desa), 99, 215n54, 259nl3 performance (1996-2000), 388-91, 394-95, 410-13 Bank Indonesia and, 106,137 performance, reporting of, 332-37 BIMAS loans and, 208-9, 215n52-55 policy issues in creation, 231-35 BIMAS service delivery program and, 180-81 portfoho quality measures, 247, 261n24 Kuiper, Klaas, 179, 211n3, 211n7, 211nM5 principles, 235-37 KUK (Kredit Usaha Kecil, Small Business Credit Program), prompt payment incentive (IPTW), 242-43 80n28 repayment history, 254-55, 395, 397-98, 410-13 channeling arrangement with BDB, 157 Rural Credit Working Group and, 231-32, 260n22 commercial banks and, 104 staff training, 340-42 KUKESRA (Kredit Usaha untuk Kesejahteraan Rakyat, Cred- subsidized loans of BRI compared with, 220,222-24,350 it for a Prosperous Family), 105,134, 136-37 three crucial decisions m creation, 232-43 KUPEDES (Kredit Umum Pedesaan, BRI unit desas' Gener- urnt desa profitablity and (1984-96), 256-58, 258nl, al Credit Program),xxviu,xxx, 8-10, 15-16,259nl5,259nl6, 261n31-33, 410-13 260n17, 262-64. See also Unit desas of BRI umt desa profitability and (1996-2000),395-96,410-13 BIMAS compared with, 220, 222-24 unit desa reorganization and, 237-38, 309-22, 390 BKK/CJ influence, 118 unit desa transition from BIMAS to, 229-35, 259n7-10 borrower enterprise growth and, 248-52 in urban areas, 319-20 borrower selection procedures, 238-39 World Bank loans to fund, 259-60n16 borrower surveys, 211-12nl 8,246-48,249-51, 261n26, Kusumaadmadja, Sarwono, 75 261n27, 261n29 KUT (Kredit Usaha Tani), 106, 137, 233-34, 259n14 borrowers' views, 249-52 Kwik Klan Gle, 70, 378 BRI unit desa organization and, 310, 312-17, 320-21, 355-56 Ladang (swidden) agriculture, 88, 139n2, 203 collateral, 243-44, 260n23, 352 LDKP (Lembaga Dana dan Kredit Pedesaan, Rural Fund and as commercial rricrocredit program, 222,224 Credit Institution), 98-99,109-15, 141n24. See also BKK/CJ; creation, 22 1-24, 229-38, 234-35 BKK/SK; BPR; LKP delinquency measures and management, 245-47 performance, 110-11 expansion as priority, 401 Lehman Brothers, 378 Export-Import Bank ofJapan loan to fund, 260n16 Lev, Daniel S., 61-62 financial crisis and, 394-97 Lieberman, Ira, xXI-xxiii, I Index 457 Linguistic diversity, 86-87 Mimstry of Finance, xxx, 73-74, 77, 105, 118, 213n29 See also Literacy rate, 26, 78n9 Boediono, Prawiro;Wardhana LKP (Lumbung Kredit Pedesaan, Rural Credit Organization) Banking Law (1992) and, 102 of Nusa Tenggara Barat, 125-27 BRI and, 180, 210nl, 221,225,226,230-32, 234, 258n2, Loans See also BIMAS; BPD, BPR; Credit; Interest rates; KU- 259nlO, 268, 288, 308, 361, 362n2 PEDES; KUT; Nonperforinng loans; Rural financial insti- Commercial microfinance and, 94,102,230, 258n2, 288 tutions and programs; Subsidized credit programs; Umt desas Financial reforms, 35,37-38 of BRI, entries by bank name Mitra Karya Grameen replication project, East Java, 84-85, BDB, 151-57, 384-87 130-32 BIMAS, 180-86 Mobile savings teams, BDB, 157-58, 164n7, 164n8 BKDs, 343-47, 382-84 Monteverde, Richard T, 305n16 BPRs, 112, 114-30,407 Moral hazard, 51 by financial intermediaries compared with credit-focused MPR (Majehs Permusyawaratan Rakyat, People's Consultative organizations, 137-38, 407 Assembly) categories of nonperformmng, 407n3 structure and powers, 41, 59, 80n32 KUPEDES, 239-44, 262-56, 383-90 vice presidential election (2001) and, 72 Lotteries unanimous elections of Soeharto by, 43 for BDB savers, 158 Wahid election by, 20-21, 59-60 for BRI unit desa savers, 273, 275, 276, 299-300, 320 Wahid impeachment by, 63-64 LPD (Lembaga Perkreditan Desa,Village Financial Organiza- Mulyono, Sri Djojosupadmo, 24-25, 77 tion), Bah, 113 Mydans, Seth, 44,45. 61, 65, 67-68, 72,74,78 Lumbung desa (village granary/rice bank), 97-99, 173-74, 343, 363nl8. Nahdlatul Ulama, 59. See also Wahid Lumbung Kredit Pedesaan of Nusa Tenggara Barat. See LKP NASAKOM (Nasionalisme,Agama, Komunisme) coalition, 29 Nasution, Abdul Harls, 29 Majelis Permusyawaratan Rakyat. See MPR National Awakening Party. See PKB Majelhs Permusyawaratan Rakyat Sementara (MPRS), 30, 79n19 National Development Planrung Board. See BAPPENAS Markovitch, Patricia, 260nl7, 305n23 National Family Planmng Coordinating Board. See BKKBN Masters, Edward, 26-27 Neill,WilfredT, 88 Maurer, Klaus, 301, 376 Nestor, Camilla, 107 McBeth,John, 75 New Order government, 19-20, 34-47. See also Abri; Golkar; McCawley,Tom, 378 Soeharto McGuire, Paul B., 132 agricultural policies, 189-94 Mears, Leon, 212n19 BRI unit desas and, 166-68, 170-71, 175-80, 219-26, Media, reform of, 20, 61, 76 229-37,399-400,403-05 Megawati Soekarnoputri, 71-75, See also PDI; PDI-P; Sukarno conglomerates and, 45-46 assumption of presidency afterWahid impeachment, 22, crisis under, 30-32, 47-56 64,366 economy under, 34-40 economics and finance under, 73-74 financial reforms, 37-38 factors in presidency, 71-73 Golkar and, 41, 43-45, 102 general elections (1999) and, 58-59 rice intensification program under, 178-79 government priorities, 74-75 role of military, 41-43 Javanese concept of power and, 72 pohtucs, 43-46 poltical affiliation, 81n42 poverty reduction under, 38-40 presidency, 71-77 technocrats and, 34-37, 46, 80n22, 102 rift with Wahid, 62 Non-Aligned Movement (NAM), 41 Sukarno and, 58, 65-66 Nongovernmental organizations (NGOs) Tommy Soeharto arrest and, 65-66 PHBK borrower groups and, 132, 133, 137, 141n21 as vice president, 60 Nonperfornung loans. See also BRI; IBRA,JITF; Loans MicroBanking Bulletin, The (MBB), I conglomerates and, 36, 50-56 Mlitary See Abri debt restructuring, 375, 378-81, 407n6 extent of during crisis, 373-75 Nonrotating savings and credit associations (ASCAs), 94 458 The Microfinance Revolution: Lessons from Indonesia Oil development, 132-34 income during 1970s and 1980s, 35-36 lessons, 135-36 income from invested in rural development, xxvu, 39 Piprek, Gerda, 361, 362n3 share of exports, 36-37 PKB (Partai Kebangkitan Bangsa, National Awakenung Parry), Oka, I Gusti Made, xlviii, 144-45. See also BDB; Oka, Sri Ad- 60, 62-63 nyani PKI (Partal Kommunis Indonesta, Indonesian Communist Bank Pasar Umum and, 149-50 Party), 29, 30 customer relations philosophy, 155, 156 PNI (Partat Nationalls Indonesia, Indonesian National Party), early life, 147-48 27 nucrofinance focus, 388 Pos pelayanan desa (village service posts of BRI's unut desas), public relations, 159-60 317-18, 331 savings lotteries and, 158 Poverty Oka, Sri Adnyam, xlviii, 144-45,147. See also BDB; Oka, I Gusts BIMAS and, 189-90, 196, 198,199-200, 201-02, 206 Made in D.I.Yogyakarta, 128 Bank Pasar Umum and, 149-50 financial crisis and, 57, 93 customer relations philosophy, 156 government response to during financial crisis, 404 early life, 148 green revolution and, 90-91 as informal commercial moneylender, 148-49 in NusaTenggara Barat, 125, 141n34 Kartini award (1994), 155 reduction under New Order government, 38-39 rnucrofinance focus, 388 rural, 92,140nlO public relations, 159-60 Power, Javanese concept of. See also Anderson Operation Work for Prosperity (Operasi Karya Makmur, OKM), as an independent entity, 23 204-05 Soeharto's presidency and, 33, 41-45 Otero, Maria, xxxv PPP (Partal Persatuan Pembangunan, United Development "Overcoming the Current Economic Downturn" (Wardhana), Party), 43 excerpts from, 51-52 Prawiro, Radius, XIvi, 78n PRODEM (Fundaci6n para la Promocidn y Desarrollo de la P4K (Pembinaan Peningkatan Pendapatan Petani-Nelayan Microempresa) (Bolvia), 1 Keci], Income Generating Program for Small Farmers and Program to Link Banks and Self-Help Groups. See PHBK Fishermen), 12,115, 347-48, 363n20 Provincial Development Bank See BPD BKD compared with, 350-52 Provisional Manualfor the Credit Business of thie General Poptilar competition with PHBK, 134 Banke (Fruin), excerpts from, 89,174, 210n3, 211n8 loan repayment, 349 excerpts on agricultural loans from, 176-77 participation, 349 Proyek Bimbingan Massal (early version of BIMAS), 21 ln13 program costs, 349-50 PT Freeport Indonesia, 66 Pancasila as the five guiding principles of the Indonesian govern- Rabobank (the Netherlands), 96, 140nI4 ment, 23 Radelet, Steven, 35 Soeharto's presidency and, 33-34 Raiffeisen, Friedrich W, 172 Sukarno's declaration of, 28 savings and loan model, 96, 97, 140n13 Papua (Irian Jaya), 66, 74 Rais,Amien, 65,74 Parhusip, Uben, 128, 129,131,137-38 Ramayana, 24-25,45 Patten, Richard H., 259n10, 362n5 Ravicz, R. Marisol, 120 PDI (Partai Demokrasi Indonesia, Indonesian Democratic on BKK/SK, 120-23, 141n33 Party), 43, 81n42 on interest rates for BKD loans, 346 PDI-P (Partal Demokrasi Indonesia-Perjuangan, Indonesian on interest rates for P4K loans, 349 Democratic Party-Struggle), 58, 81n42 on LKP performance, 125-26 Pekarangan (mixed garden) agriculture, 88 on NGO-supervised credit groups, 133 People's Consultative Assembly. See MPR on P4K, 349-50 People's Credit Bank. See BPR on PHBIK, 132,134 PHBK (Program Hubungan Bank dan Kelompok Swadaya Regional Development Bank See BPD Masyarakat, Program to Link Banks and Self-Help Groups), Rhyne, Elisabeth, xxxv 122, 141n21 Riantiarno, Nano, 47 at Bank Shinta Daya, 128 Index 459 Rice Intensffication: Development Program Implementation Studies Re- adrrumnistrative costs of at BRI unit desas, 299, 305n24 port 2 (Harvard Institute for International Development), Bank Shinta Daya, 128-30 210nl,229-30,263 BDB (1970-96),151-53 Rice production. See also BIMAS; Brown planthopper; Insec- BDB (1996-2000),385-87 ticide danage to rice production; KUT BDB customers, 160-61,164n9 area covered under BIMAS loans for, 1970-84, 183 BDB products for, 157-58 cultivation, 88-93 BDB mobile savings teams, 158-59, 164n7, 164n8, 187 in four vlllages, 194-208,212n30 BRI unit desas, 186-87,262-305,388-94,398,410-13 green revolution and, 83,90-91 commercial microfinance institutions during financial insecticide damage, 191-93,197, 199 crisis, 405-7 intensification program, 178-79 compulsory, 117-18,122,126,131,133-34,137,141n29 national commutment to, 210 LKPs, 126 national pest management policy for, 193-94 lotteries and, 158,299-300 Robinson, Marguerite, 164n9,212n25,259n8,259nlO,305nl9, market penetration of by BRI unmt desas, 264, 293-94, 362n1,362n5,363n23 305n22 Rotating savings and credit associations (ROSCAs), 94,101,265, overview, 262-65 268-69 products, demand for, 268-95, 304n9 Rural Credit Working Group, 231-32, 260n22 propensity of households for, 212n23 Rural development, 39,82-83,86,88-89. See also Agriculture; reasons for providing services at BRI umt desas, 265-67 BIMAS, Economy; KUPEDES; Rural financial institutions performance by instrument at BRI umt desas, 302-04, and programs; Savings 390-94 during 1997 financial crisis, 93 planmng for by BRI unit desas, 269-70 green revolution and, 83,90-93, 140n9 principles of at BRI umt desas, 266 Rural financial institutions and programs, 83-85. See also Banks at public and private BPRs, 112 and banking; BIMAS; BKD; BPR; KUPEDES; LDKP; Sav- systernc approach to mobilization of at BRI unit desas, ings; Subsidized credit programs; Umt desas of BRI 294-95 1980s, 99-102 transfer price and, at BRI unit desas, 300-301, 305n25 1990s, 105-36 Umt Achievement Competitnon at BRI unit desas and, Bank Shinta Daya, 127-30 295-96 Banking Law (1992) and, 102-04 Savings casebook for BRI unit desas, 305n19 BKD, 94,115,343-52,366-67,382-84,406-07 household savings mobilization, 282-87, 305nl9 BKK/CJ, 94,113,115,117-20,124,138,141n30,235 industry linkage cases, 282-84 BKK/SK, 120-25,138,141n31 listening to savers' needs, 284 early development, 93-99 locality-specific cases, 290-91 financial intermediation, 137-38,407 nationwide cases, 287-90 government rnucrofinance initiatives in 1990s, 102-7, organizations' and institutions' savings mobilization, 141n22 287-91,305n19 LKP, 125-27 remittances, 284-85 loans (1985),100-02 from rural land sales, 285-87, 305n20 loans and savings (1995), 108-09, 141n23 from schools, 288-89 Mitra Karya Grameen replication project, 130-32,138 Sawah (irrigated riceland) agriculture, 88, 90, 212n33. See also PHBK, 132-36,138 Agriculture; BIMAS; Rice production perceptons of bureaucrats of potential for, 226-28,258n4, Schnmt, Leo Th., 172,175,210-1 1 n3, 21 lnlO, 362n8 259n5,259n6 Schultze-Delitzsch cooperative bank model, 96 perceptions of policymakers of potential for, 225-26 Schwarz,Adam, 31,45,79nl2,79nl4,79nl7-18,79-80n20-21, perceptions of villagers of potential for, 228-29 80n36 proliferaton, 108 on Golkar 43 public and private, 95, 112-14,138 onJavanese concept of power, 33 self-help groups and, 128-36 on massacres after coup attempt, 79n18 on Pancasila, 28 Savings,304nl,408nl1. See also Bank Shinta Daya;BDB,BKD; on Soeharto, 42-43 BPR; Deposito Berjangka; Savings casebook for BRI unit on Sukarno's Guided Democracy, 28-29 desas; SIMASKOT; SIMPEDES;TABANAS Selbel, Hans Dieter, 128,129,131, 137-38, 376,391 Self-help groups, 128-36 460 The Microfinance Revolution: Lessons from Indonesia Separatist movements in provinces, 61, 66-68, 80n33. See also MPR presidential elections, 43, 79n19 East Timor independence mysterious kllings and, 80n21 SIMASKOT (Simpanan Kota, urban savings in BRI unit desas), Pancasila and, 33-34 280, 293, 296-98 305n26, 319 poverty reduction under, 38, 57 adaptation of SIMPEDES for, 362n8 presidential instructions, 105-06, 179, 210 during financial crisis, 392-94 resignation, 49-50, 365-66 interest rates, 298 rural development under, 92-93, 178-79, 193-94 lotteries, 299-300 Wahid's attempts to bring to trial, 63-65. performance (1984-96), 302-03,334-36 Soeharto family, 19-21, 50, 69-70, 372-73, 374, 379, 381 performance (1996-2000), 392-94, 410-13 Tommy, 65-66 SIMPEDES (Simpanan Pedesaan, rural savings in BRI unit Soeharto supporters desas) See also Savings; Savings casebook for BRI unut desas; and IBRA, 372, 374,377, 381 SIMASKOT; Unit desas of BRI role after Soeharto resignation, 57-58, 60, 64-72, 76, account sizes, 274-75, 302, 305nl6, 335-36 381 BKK/CJ's TAMADES compared with, 118 Soekarnoputri, Megawati. See Megawati Soekarnoputri economic ministers' approval, 279-80 Sombart,Werner, 173 estimating demand for, 268-69,271-72, 304n9, 305nlO Steinwand, Dirk, 94,140nll, 211 n3, 407 expansion to all unit desas, 280-82, 291-96 on banking system after 1999 regulatory changes, 103 during financial crisis, 393-94 on BKDs, 343-44 first-stage pilot project, 270-76, 305n13 on BPR profitablity, 112-13 interest rates, 272-79, 297-98, 335 on BPR self-sufficiency, 114 lotteries, 299-300 on BPRs and financial crisis, 115,367 performance, 302-03, 334-36, 393-94 on creation of BRI, 173 second-stage pilot project, 276-79, 305n15, 305nI7, on European mucrofinance, 95-96 305nl8 on comparison of financial intermediaries and credit- unhmnited transactions, 266, 268-70, 274-76, 292 focused organizations, 137-38, 349 Sims, Calvin, 63 on inconsistency of BPR statistics, 109, 141n26 Sjahrir, Soetan, 27 on mandatory savings as share of total savings at BKK/CJ, Slade, Betty F, 31-32, 46, 52, 55, 103, 407n2 141n29 Small Business Credit Program. See KUK on performance of public LDKPs, 113 Small Investment Loan Program. See KIK on PHBK and earlier Indonesian mistakes, 134-35 Small Permanent Workmg Capital Loan Program. See KMKP on public and private BPRs, 112-13 Snodgrass, Donald R., xlvii, 21 Onl, 212n25, 212n30, 259nlO Stern,Joseph J., l, 56, 80n23 Soeharto, 3, 19-20, 32, 76, 80n35, 80n38. See also Abri; Con- Stiglitz,Joseph E., 53, 81n41, 407n2 glomerates; Corruption; Economy; Indonesian crisis; Ja- Subsidized credit programs, 84-85, 101, 106-07, 222-23, 354, vanese concept of power; New Order government; Soeharto 358, 360. See also BIMAS; Government mucrofinance ini- famrly; Soeharto supporters tiatives; INPRES Desa Tertinggal; KUJD; KUKESRA; KUT; 1965 attempted coup and, 29-30, 79nl7 PHBK; Mitra Karya Abri and, 41-43 BIMAS program at BRI unut desas, 169-215 BIMAS and, 178-79 commercial credit compared with at BRI, 220,222-24 conglomerates and, 45-47, 57, 381 history, 94-95 corruption and, 19-29, 45-47, 69, 373 INPRES Desa Tertinggal, 105-06, 136-37 crisis and, 47-56, 370 KIK, xxvi, 101, 140n18 on divine guidance, 79-80n20 KMKP, xxvi, 101, 140n18 early life and military background, 42 KUDs and, 106,137,208-09 economy under, 34-37, 40, 44, 76, 81n4l KUKESRA, 105, 134,136-37 education under, 39 KUT, 106,137,233-34 endorsement of BRI unit desa system, 224,279 levy on state-owned enterprises and wealthy individuals family planning under, 39 for, 105 financial reforms under, 37-38,102-04 PHBK, 132-36 Golkar and, 43-45 Subsidy dependence index (SDI). See also Ravicz;Yaron governance, 41-47, 77 BKK/SK, 123 Javanese concept of power and, 33, 41 BKK/SK and LPK compared, 121 Megawati and, 58 BRI unit desas, 360-61, 363n24 Index 461 measurement, 363n24 Tegalan (dryland) agriculture, 88, 198 P4K, 350 Thapa, Ganesh B., 132 PHBK, 134,135 Tllly, Richard, 96 Sudarsono,Juwono, 43,67-68 Timberg, Thomas A., 113-14 Sugianto, xlvli, 226, 311, 326, 362nl, 362n4, 363n23 Time deposits. See Deposito Berjangka pohcy concerns, 359 Trnner, C. Peter, 259nlO Suharto, Pandu, 97 Sukabumi District,West Java, 305n13 Umt Achievement Competition in BRI unut desas, 295-96,332, early SIMPEDES success in, 273-75 338-39 estimating demand for savings, 271-72,304n9,305n10 Unit desas of BRI, xx, xxiii, xxv-xxxi, xl-xlu, xlvil,84,94,367. follow-up survey, 275-76 See also Arief, BIMAS; BRI; Center for Policy and Imple- SIMPEDES pilot, 271-76 mentation Studies; Deposito Berjangka; Development Pro- Sukarno, 19, 24, 46, 76. See also Megawati Soekarnoputri gram Implementation Studies; Harvard Institute for 1965 attempted coup and, 29-30, 79n17, 70nl8 International Development; Kredit Midi; Kredit Mini, KU- constitutions and, 28, 79n13 PEDES; SIMASKOT; SIMPEDES; Sugianto;TABANAS; economy under, 29, 79nl 6, 178 Wardhana Guided Democracy, 28,29 BIMAS compared with, 220,222-24 Javanese concept of power and, 33 BPRs compared with, 109-10 as leader of independence movement, 27 BRI corporate structure and, 315, 331-32, 352-62, organization of PNI and, 27 399-400 Pancasila proclamation, 23 Business Umt Desa (BUD) as orgamzational unit in BRI, presidency, 28-30, 46 314,326-26,322-32,390 Suksesi (Rlantiarno), 47, 80n39 challenges, 398-402 Sumitro Djojohadikoesoemo, 21 lnl employment by organizational level (1996), 325 Sundanese, 78n2 Deposito Berjangka (time deposits), 270-71, 276-77, Survei Sosial Ekonomu Nasional (SUSENAS, National Socio- 280,284-85,292,296-98,302-03 Economic Survey), 271, 304n8 field studies about, 208-10,210nl,211-12nl8,213n30, Suryatmaja, Putu Indra, xlviii 213n39, 214n42, 214n47, 229-32, 238, 242, 244, Suryohadiproyo, Sayidiman, 43 246-52,254,258n2,259n5,259n7-8,260nl7,260n22, Sutarto, Endriartono, 63 261n26, 261n28-29, 267-70, 274-82, 292-95, Swidden agriculture. See Ladang 304n4-6,304n17,304nl9 Swift,Jonathan, 96 financial crisis and, 12-14, 85, 114, 364-65, 388-96, 408n9,408nl2-13 TABANAS (Tabungan Nasional, national savings product), 18. financial self-sustainability incdcators, 360,410-13 See also Unit desas of BRI InternationalVisitors Program, 248,260nl7,362nl1 availability in banks, 297-98 Kredit Midi and Kredit MNm, 184-86 in BRI umt desa pilot project for savmgs, 270-71,273-78, KUPEDES credit program, 216-61 280 KUPEDES during financial crisis, 394-95, 397-98, interest rates at BRI unit desas, 232, 234, 273, 280, 410-13 297-98,305nl2 KUPEDES geographic distribution, 255-56 limutations on withdrawals, 229, 268, 292 KUPEDES loan loss ratios, 253-55, 388, 410-13 lotteries at BRI unit desas, 300 KUPEDES loan number, 253,380,388-89,410-13 performance at BRI umt desas, 302-3, 334-36, 392, KUPEDES loan value, 253, 388-89, 391, 410-13 410-13 KUPEDES principles, 235-37 savers' use at BRI unit desas, 286,289, 296 KUPEDES prompt payment incentive (IPTW), 242-43, SIMPEDES compared with, 292 363n21 withdrawal restrictions lifted at BRI umt desas, 296 history, 95-98, 172-87 Tabungan Nasional. See TABANAS location, 316-19 TAKESRA (Tabungan Kesejahteraan Rakyat, Savings for a management, 190-91,221-22,226-28,237-38,316-18, Prosperous Famuly), 105 323-24, 258n3, 306-08, 310-16, 322-32, 352-59, TAMADES (Tabungan Masyarakat Pedesaan, Rural People's Sav- 390,397,399-402 ings at BKK/CJ), 118 marketing and public relations, 273-79, 287, 299-300, Tandjung,Akbar, 72 319-21 Technocrats, 34-37, 46, 80n22, 102 microsavings evolution in, 264 462 The Microfinance Revolution: Lessons from Indonesia mission, 15-16, 235-37, 265-67, 396-402 Village P, South Sumatra, 194-96, 202-08, 212n32, 214n43, iustakes as lessons, 316-22 214n46,214n47,214n48,215n49,215n50,215n51 old and new cultures, 309-11,316 Village R, South Sulawesi, 194-96,200-02,205-08, 214n41, organization, 237-38, 307-16, 322-32, 390 214n42,214n43 overview, 5-14 Volksbank (People's Bank), 93-94,173 performance, 167-68,187,252-58,322,360-61,388-96, Volkscredietwezen (Popular Credit System), 89, 173-75 410-13 policy issues related to, xxv-xxxi, 102-09,208-10,219, Wahid,Abdurrahman, 20-22,60-71 224-26,229-35,265-67,359-62 censure and impeachment, 22, 63-64 pos pelayanan desa (village service post), 317-18, 331 corporate debt restructuting under, 378-80 principles of savings mobilization, 266 DPR and, 62-64 profitability, 114,212n24,217,220,256-57,334-37,360, early reforms under, 61-62 369,388,395-98,408nl3,406-07,410-13 economy under, 68-71,369-72 returns on assets, 256, 261n32, 388, 408nlO, 410-13 efforts to bring Soeharto to trial, 64-66 rice intensification program and, 178-80, 210 financial crisis and, 369-72 SIMASKOT (urban savings), 280,293,296-303,334-36, IBRA under, 69-71, 376-80 392-94,410-13 MPR presidential election of, 59,80n37 SIMPEDES (rural savings), 118,270-303,393-94,410-13 presidency (1999-2001), 60, 71,366 staff incentives, 295-96,315-16, 332,337-39 separatist movements and, 66-68 staff traimng, 190, 237, 280-91, 294-95, 307, 315-16, Wallace, Alfred Russel, 86 339-42,363nl5,363nl6 Wardhana,Alh, xxv-xxxi, xlvi, 22,35-37, 53, 80,81,349,407. strengths, 396-98 See also Ministry of Finance subsidy dependence index, 360-61, 363n24 on 1997 financial crisis, xxvii-xxix, 22, 50-52,78 TABANAS, 229,232,234,268,270-71,273-78,280,292, on BRI unit desas, xxv-xxxi, 231-35 297-98,300,334-36,392,410-13 on financial reforms, xxvi-xxvii transfer price, 270,300-301,312,348,397 on KIK and KMKP, xxvi, 140n18 transition from subsidized lending to commercial finan- KUPEDES and, 234 cial intermediation, 229-35, 259n7, 259n8, 259n9, technocrats and, 35 265-67,307-22,328-31 Warwick, Donald P, 210nl unit development reports, 332-37 Wayang shadow puppet theater, 23-25,32-33,77,78n4,78n5, United Nations Development Programme (UNDP), 347 78n6,78,n7,78n8 U.S.Agency for International Development (USAID), xlvm, 118, "What a Mess!" (Friedman), excerpts from, 77 316,362nll Widjojo Nitisastro, 35 U.S. Central Intelhgence Agency, 29,79nl7 Wiranto, General, 61 U.S International Executive Corps Service, 150 Wiriamaadya, Raden, 172-73 Usman, Marzuki, 1, 259nlO Witoelar,Witmar, 59,77 Wolff, Henry W, 96 VanWesterrode, DeWolff, 172-73,363nl8 World Bank, xxiii, xxix, xlviii, xlix, 1, 26,30-32,36-37,39-40, Van Zorge,James, 75 49,53,69,71,78,80,81,86,107,108, 212n23, 259-60nl6, Varley, R. C. G., 1, 305nl5 365,371-72,407n2 Vietnam Bank for the Poor, xlix Vlllage C,WestJava, 194-96,198-200,205-09,212n38,212n39 Yacob, Sofyan, 65 Vlllage G, EastJava, 194-98,212n34,212n35,212n36,214n40, Yaron,Jacob, 257, 259n 15, 361, 362n3 214n43 Yusuf, Glenn S., 374 Index 463 Order Your Copy of The Microfinance Revolution Use this convenient order form to reserve your copy of The Microfinance Revolution. Customers outside the U.S.: Contact your local distributor for information on prices in local currency and payment terms, induding acceptable credit cards. Title Stock # Price Oty. Total US$ The Microfinance Revolution: Sustainable Finance for the Poor D14524 $35 Volume I Publication Date June 2001 (ISBN 0-8213-4524-9) The Microfinance Revolution: Lessons from Indonesia D14953 $35 Volume 2. Publication Date: June 2002 (ISBN 0-8213-4953-8) The Microfinance Revolution: The Emerging Industry D14954 $35 Volume 3. Publication Date: June 2003. (ISBN 0-8213-4954-6) PLEASENOTE:Prcesarinuss.Specialdscount ply forfcustomers in developincounriesPleasevist Subtotal our website at www worldbank.org/publicat,ons and click on 71o.w t.o *.ider fpr ad itional infoimation Ol Shipping and avilable dis ut n Worl Ban dis rutors-lj Shippin g and Idandlg charges are uS$8.00 oer d Handling * For air=mail delivety outside the United 11States, charges are -S$7.00 per order p U 6.00 per ite Total US$ O YES, please send my copy of The Microfinance Revolution PAYMENT METHOD (Please Print) Orders from individuals must be accompanied by payment or credit card information Credit cards accepted only for orders addressed to the Name World Bank. Please do not send cash. Title __ Check no. in the amount of $ is enclosed. When ordering Organization directly from the World Bank, make check payable in U.S. funds drawn on Address a U.S bank to: The World Bank. Please send your check with your order E Charge $ . . to my: city E2 Amencan Express E Mastercard [1 Visa State __ Card no Zip/Postal Code Expiration date / Country Name Phone as it appears on the card Signature Fax required for all credit catd charges Institutional customers in the U.S. only: E-mail ___ Bill me. Please include purchase order Mail order to World Bank Publications PO. Box 960, Herndon, VA 20172-0960, USA, or Fax to 703-661-1501 Order by phone: 703-661-1580 or 800-645-7247. Order online: www. worldbank.org/publications Ouestions E-mail us at books@worldbank.org DMCR World Bank Publications 1 14~J Visit our website at www. worldbank. org/publications Jeannlette Marie r~,rnith _ 1 rne 7 i'~~~~~~~~~~~~,N I . . C3-301 WASHINGTON DC i ¢ (m icrollFE>jU,;lX .inance "A ."'FZ4geSu' rayf: -. | ~~ Around the world, a revolution is The Emerging Industry (much of which E"ziF *;;;r 2 qE occurring in finance for low-income was written with Peter J. Fidler), explores ''~'~~"" - ~ ~ people. The rnicrofinance revolution refers commnercial microfinance around the to the delivery of financial services to the world and offers a preview of it in 2025. i<\, ,-J-,5J.~~ J~-r-,7 p j- i. I-Jv ZIJ economicaUy active poor on a large scale This second volume focuses on TC-r1- 5 *-,- dithrough competing, financially self- Indonesia, the first country to develop E-S~ fT4 )T,,,, r f'irm7 TPT-jjc rsufficient institutons. In a few countries large-scale commercial microfinance. this has already happened; in others it is Chapter 8, an introduction to Indonesia __ t:under way.The emerging rnicrofinance through mid-2001, and chapter 9, a ;\ JJvS-J- Ri8w,t ,i,,f,-,v5J] industry has profound implications for review and critique of the country's rural '; ~ . -r . social and economnic development. For the develbpment and rural financial f;;iZ~ B~Th.ic. < ,-;v,,Jo ,- first time in history, capital is well on its institutions, provide background for fiX~21 P?* cwn~~ ____ way to being democratlzed, understanding microfinance in Indonesia >v597wFr57 T-jw ff7)1j2i',. j-rj~ r~ The Microfinance Revolution, in three today. Bank Dagang Bali, the oldest volumes, is aimed at a diverse licensed general bank specializing in ''''7ire J. f/r-rv- .} -., ,* ,J':, ;K.:r;- -; -, 1 l- readership-economists, bankers, commercial microfinance currently policymakers, donors, and social operating in a developing country, is 7t -' t (VI *LuUVTjIFA= scientists; microfinance practitioners and discussed in chapter 10. R rY f . )4J specialists in local finance and rural and The development and growth of '+~ '~ 9X/grm$7 r urban development; and members of the BRI's microbanking (unit desa) system general public interested in development. from 1970-96 are considered in chapters >A <5iuHX 1i,tr The first volume, Sustainable Financefor 11-15. Chapter 11 examines the system as rMpg-<-i7efs7rC cm fsC-t the Poor, focuses on the shift from it developed mto a massive, failed >W m, ^-ff r J>1,s r ~ government- and donor-subsidized subsidized rural credit program (197083). credit systems to self-sufficient Chapters 12-14 explore the unit desas' '2J:-Af js v- r o...-. ! -n ' m icrofinance institutions providing transformation into a nationwide -I .ZzIU-J7v . IF(t AS jr>irr voluntary savings and credit services.This commercial nucrobanking system 34-,(>; t;i6-l.llJ;; i1_t, Sri; microfinapce-its st4biliry, outreach, and analyzes the extraordinary stability of the _. profitability even during the recent unt desas through 2001-with some 20 C 7 severe crisis-and focuses on Bank million clients and more than 15 years of ifrk)7J.m tA i7tfM F rrn> faniIt Rakyat Indonesia's (BRI's) unit desa profitability-aand other commercial system, the world's largest sustainable microfinance organizations during @,!? c*rW-,, f>¢rc>;t microbanking system.The third volume, Indonesia's'recent crisis, P'in A: C fL/ FtrnI IfgT Yb.r -,/;J? T,rw JW7TI Jt>f0 m jLjru9/ l l11111