Report No. 6917-IND Indonesia Rural Credit Sector Review (In Two Volumes) Volume 1: The Main Report Apnl 29, 1988 Asia Regional Office FOR OFFICIAL USE ONLY Document of the World Bank This report has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. CURRENCY EQUIVALENTS Before November 15, 1978 US$1.00 = Rp 415 Annual Averages 1979-85 1979 US$i.00 = Rp 623 1980 US$1.00 = Rp 627 1981 US$1.00 = Rp 632 1982 US$1.00 = Rp 661 1983 US$1.00 = Rp 909 /a 1984 US$1.00 = Rp 1,026 1985 US$1.00 = Rp 1,125 September 12, 1986 US$1.00 = Rp 1,644 /b FISCAL YEAR Government - April 1 to March 31 Bank Indonesia - April 1 to March 31 State Banks - January 1 to December 31 PRINCIPAL ABBREVIATIONS, ACRONYMS AND LOCAL TERMS Adat - customary ASKRINDO - PT Asuransi Kredit Indonesia (Credit Insurance Company of Indonesia) Bank Pasar - Market bank/petty trader bank BAPIND3 - Bank Pembangunan Indonesia (Development Bank of Indonesia) BBD - Bank Bumi Daya BDB - Bank Dagang Bali BDN - Bank Dagang Negara BEII - Bank Export Import Indonesia BI - Bank Indonesia BIMAS - Bimbingan Massal: "mass guidance" BKD - Badan Kredit Desa: village credit body BKK - Badan Kredit Kecamatan: subdistrict credit unit BKPD - Bank Karya Produksi Desa: village production bank BNI '46 - Bank Negara Indonesia 1946 BPD - Bank Pembangunan Daerah: regional development bank BPR - Bank Perkreditan Rakyat: secondary bank BRI - Bank Rakyat Indonesia /a On March 1, 1983, the Rupiah was devalued from US$1.00 = Rp 703 to US$1.00 = Rp 970. !b On September 12, 1986, the Rupiah was devalued from US$1.00 = Rp 1,125 to US$1.00 = Rp 1,644. FOR OFFCIAL USE, ONLY BUKOPTN - National Cooperative Bank BULOC - Board of Logistic Affairs DMBs - Deposit Money Banks IPTW - Insektif Pembayaran Tepat Waktu: incentive for prompt payment GOT - Governmt'it of Indonesia Kecamatan - Subdistrict KEPPMES - Presidential Decree KI - Kredit Investasi: investment credit up to Rp 70 million KIK - Kredit Investasi Kecil: small investment credit KK4K - Kredit Modal Kerja: working capital credit up to Rp 70 million KMKP - Kredit Modal Kerja Permanen: permanent working capital credit Kredit Mini - Small borrowers program Kredit Midi - Small credit program KUJD - Koperasi Unit Desa: village cooperative KUPEDES - Kredit Umum Pedesan: general village credit program KURK - Kredit UJrusahan Kecil: small credit program KUT - Kredit Usaha Tani: farmers credit i4KK - Lembaga Jaminan Kredit Koperasi LPD - Lumbung Perkreditan Desa: village storage and credit unit LWK - Lumbung Perkreditan Kecamatan:o subdistrict storage and credit unit LPN - Lumbung Pitih Negari: subdistrict and village credit body LbugDesa - Paddy banks NDLF - National Developmeni Loan Fund YES - Nucleus Estate and Smallholder PERUM PKK - Perum Pemgembangan Keungan Koperasi (State Corporation for Expanding Cooperatives Financing) successor to LJKK provides collateral for BRI lending to coope':atives PIR - Tree Crops Lending Program based on Nucleus Estates PMU - Project Management Unit PRPTE 4- Credit for improvement and rehiabilitation of tree crops with export potential PSN - Kredit Perkebunan Swasta Nasionall national credit for private estates PTP - Publicly owned estates PUSKUD - District head office for the subdistrict KUDs RCP - Rural Credit Project (IDA-827) "FI - Rural Financial Institutions SCDP - Smallholder Coconut Development Project SEDP - Small Enterprise Development Project Simpan Pinjam - Savings and loan association SIMPEDES - Simpanan Pedesan (Village Savings Program) SRDP - Smallholder Rubber Development Project SUSENAS - National household surveys TABANAS - Tabungan Nasional (Small Saving Program) TASKA - Tabungan Asuransi Berjangka (National Insurance Savings Scheme) Unit Desa - Village subbranch of BRI IThis document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authoriztion. - i - INDONESIA RURAL CREDIT SECTOR REVIEW Table of Contents Page No. D.s ................. ........ .i................................... * iv EXECUTIVE Si.............................................. I. OVERVIEW OF THE RURAL CREDIT SECTOR ............. ..... .......... 1 A. Recent Macroeconomic and Financial Developments0...0..0.. 1 B. GOI Credit Policy Framework.............................. 6 C. Organization of the Rural Financial System*.............. 9 Indonesia's Financial Sector........................ 9 Rural Financial System ............................. 12 Rural Financial Institutions ....................... 12 D. Trends and Patterns in Agricultural Credit............... 14 Availability of Rural/Agricultural Credit........... 14 Allocation of Agricultural Credit by Commodity...... 14 Regional Distribution of Total and Agricultural Credit................................. 16 Maturity Pattern of Agricultural Credit............. 18 Primary Agricultural Borrowers...................... 18 S. Sources of Funds for Agricultural/Rural Credit ........... 20 BI Funding of Rural/Agricultural Credit ............. 20 Rural Financial Savings............................. 20 Adequacy of Funds in Rural Credit Sector ............ 21 II. PERFORMANCE OF SELECTED CREDIT INSTITUTIONS.......... **..... 22 A. Effectiveness and Efficiency of Selected Credit Institut.................. ...... 0 ............. 00000000 22 Bank Rakyat Indonesia and Bank Bumi Daya 22 Scope of Operations. .............. .0.... .. 22 Organizational Structure.......................... 22 Loan Portfolio and Arrears.... 22 Sources of Funds........... . ............... oo.... 23 Efficiency measures......... e. ............ e...... 23 Selected Rural Financial Institutions*************** 24 Costs of Iiitermediation........................... 24 Interest Icc o m e 24 A r r e a rs.............. ................ ............. 24 Resource Mobilization............................. 25 Sustainability of Rural Financial Institutions.... 25 - ii - Page No. B. Factors Influencing the Performance of Rural Financial Institutionsu................................. 26 Availability of BI Liquidity Credits................ 26 Interest Rate Ceilings .......,.................... 26 Branching Regulations.......... .......e......... ..... 26 Economies of Scale and Homogeneity of Risks......... 27 Existence of Technical Assistance and Sponsoring Organizations.......................... 27 Assigned Role of the Rural Financial Sector......... 27 III. ASSESSMENT OF SELECTED CREDIT PROGRAMS.o..... ...........00.. 29 A. Main Features of Selected Credit Programs................ 29 Bank-assisted Tree Crop Credit Schemes.............. 29 KIK/KMKP . ........... 31 BKK Lending Programs....................................... 32 B. Measures of Program Success.............................. 32 Access to Credit. ....o.......................000.00 32 Loan Recovery Rates...... ..... .... ................... 33 Resource Base of Credit Programs.................... 34 Costs of Lending, Borrowing, Intermediation and Soubsidies ...o..o.oo..o...oo.ooo..o...oe.......oooo..o..oooo 35 C. Lessons Learned from Successful Credit Schemes........... 37 Future Strategy............................................. 38 IV. FINANCIAL RESOURCE MOBILIZATIONo...o.......................... 39 A. Assessment of Selected Instruments and Mechanisms for Mobilizing Savingso........o....o..........o........... 39 Savings Account........................................... 39 Lottery Schemes............................................ 40 Door-to-Door Savings and Its Cost Effectiveness..... 41 B. Factors Affecting Financial Savings Mobilization......... 42 Availability of Concessional Funds.................. 42 Choice of Savings Instruments....................... 43 Institutional Capacity for Mobilizing Savings....... 44 Regulatory Restraintso....... . . .o..... ..o......o ...o..o.oo 44 V. WOMEN IN THE RURAL FINANCIAL SECTOR.ooO.......... oooo.o....oo 45 A. Factors Affecting Women's Demand for Credito dito.oooo.... 45 B. Factors Affecting Supply of Credit to Rural Women en..o... 49 Transaction Costs.ostooooos.o..oooo..oooo....eoooooo 49 Targeted Len d i ng.... ..................0......00.0.0 49 Collateralo l atooeeroalo..oooo.ooo..o.o.o...oo......o 49 Cosigningoo.oo..o...oo oooooooo*oo* 50 - iii - Page No. One Loan Per Household............. .*00Oe*e*....... 50 Education ....... ........ ................................ .. ....... 51 C. Role of Women in Rural Savings........................... 52 VI. CONCLUSIONS AND RECOMMENDATIONS............................... 53 A* Main Conclusions...000.0.. ... .. ... ..... *....... .. .......... 53 B. Recommendations ..................................... ... 54 To Promote the Development of Financial Institutions.c ................................................ 54 To Improve Credit Delivery Mechanisms .............. 56 To Accelerate Savings Mobilization ..................9* 57 To Continue and Expand Women's Participation in Rural Finance ..................................5& To Strengthen Research Base......................... 59 World Bank Support to Rural Financial Reforms*...... 60 TEXT TABLES 1.1 Distribution of Agricultural Credit by Type of Bank ...... 11 1.2 Estate, Tree Crop and Food Crop Lending.................. 16 1.3 Regional Distribution of Total and Agricultural Lending.. 16 1.4 BRI Programs for Small Agricultural Borrowers............ 18 3.1 GOI Subsidies to Selected Special Credit Programs ........ 36 4.1 Resource Structure of Different Types of Banks........... 43 4.2 Growth of Demand, Time and Savings Deposits of Different Types of Banks ................o . .............. 43 Figures 1.1 Real Rates of In6erest of Deposits........................ 2 1.2 Nominal and Real Bank Deposits........................... 3 1.3 BI Liquidity and Direct Credits.......................... 5 1.4 Structure of Indonesian Financial Systemt................. 1.5 Outstanding Bank Credits by Sector ....................... 15 1.6 Regional Distribution of Total and Agricultural Credit... 17 1.7 Maturity Pattern of Agricultural Credit................... 19 - iv - Foreword (Objectives, Rationale and Scope of Study) i. This report analyzes the key issues affecting the rural financial system in Indonesia since the June 1983 financial reform and suggests policies and institutional arrangements for tapping savings in the rural sector and im- proving the availability and efficient use of credit for agricultural and re- lated activities. 'the study was undertaken in the light of Indonesia's grow- ing resource constraints and the consequent need to develop a viable financial system which could efficiently mobilize and allocate resources especially in the rural sector of Indonesia's economy. This sector accounts for more than three fourths of population and nearly 60% of employment. The efficient ope- ration of the rural credit and financial markets is a critical element in the provision of credit for the continued growth and efficient diversification of the agricu -e sector. ii. The report develops several recommendations from its analysis. The most urgent of these attempts to remedy the failure of the rural financial syster to charge borrowers enough for loans-both because interest rates are too low to cover lending costs and because loans are not effectively col- lected. Indonesia loses from this practice: every such investment will represent an added burden to the economy when returns do not cover costs. Nor could the report find a bias in favor of low income borrowers; these programs cannot be justified because they help the poor. Accordingly, the report recommends broadly; - immediately instituting sound collection procedures for outstanding and new loans; - raising interest rates on BI liquidity credits and onlending rates on agricultural and rural program loans by 2% per year until market rates are attained. The report makes numerous supporting recommendations that suggest more precisely how these changes might be made. It sets these out in the executive summary and discusses them at greater length in the body of the report. iii. The review focused on the following aspects of the rural financial market: performance of selected rural credit institutions, lessons from se- lected rural credit schemes, and factors affecting financial resource mobili- zation. A special topic investigated was the role of Indonesian women in ru- ral finance. Although the informal rural credit system is significant in size and coverage, this report was limited to the formal rural credit institutions - v - due to resource constraints.!1 A qualifying assumption of the study is the definition of rural credit and savings according to the urban/rural iocation of the institutional sources of credit as defined by the household budget sur- veys (SUSENAS). This is a necessary assumption since no precise data exist segregating the rural financial market from the overall economy. The study also estimated the share of agriculture in general, noncommodity-specific cre- dit schemes based on the purpose of the loan although the finai use of the loan cannot be ascertained due to the fungible nature of credit. Finally, the rural credit system is an integral part of the country's financial sector; thus, some findings and recommendations impact not only on rural finance, but also on the whole financial system. iv. The report is based on the findings of a mission that visited Indonesia in January 1987 and builds upon the results of studies undertaken by the Bk on the rural credit market in 1982 and the financial sector in l985.,' Additional information was obtained from the supervision of the Small Enterprise Development Project III, the appraisal of the Bank Rakyat Indonesia (BRI)/Small Enterprise Credit (KUPEDES) Project and the Project Completion Report on the Rural Credit Project in 1985. The outline of the report is as follows. The first section presents an overview of the rural credit sector against the ba^kground of recent macroeconomic and financial developments and focuses on the institutions of the rural financial system, trends and patterns in agricultural credit and financial resource mobilization, and government po- licy framework in the rural credit market. The second section assesses the effectivenest and efficiency of selected rural credit institutions in order to determine the factors that influence their success or failure. The third sec- tion analyzes the performance of selected special credit programs in order to determine the factors that affect performance and derive lessons from the suc- cessful programs. The fourth part focuses on cost-effective mechanisms for mobilizing rural savings and the factors that affect financial resource mobi- lization in rural areas. In the fifth chapter, the participation of women in rural credit and savings and the factors influencing their demand for and supply of credit are examined. The final section recommends measures to im- prove the effectiveness and efficiency of existing credit policy instruments and institutions, mobilize rural savings and improve the availability of credit. 1/ In addition, the Asian Development Bank is scheduled to complete, by December 1987, a regional study on the informal rural financial markets that will include Indonesia. 2/ The mission consisted of Ms. C. Tanchoco, Messrs. 0. 0. Nyanin (Bank), R. Hommes, M. Skully, D. Lucock and Ms. M. Berg (Consultants). It is deeply indebted to the cooperation and assistance of GOI and bank officials especially those from the Ministry of Finance, BAPPENAS, Bank Indonesia, Bank Rakyat Indonesia, Bank Dagang Bali, Regional Development Banks of Central and East Java and members of the Interagency Group on Rural Credit. - vi - EXECUTIVY SUMMARY I. Overview of the Rural Credit Sector Recent Macroeconomic and Financial Developments 1. The Indonesian banking system expanded rapidly in 1978-82 due to the Government's comfortable financial position, its policy of channeling surplus government funds through the banking sector and rapid economic growth (8% p.a. growth in CDP). However, as oil prices weakened, the Government of Indonesia (GOI) introduced a major reform of the financial sector in June 1983. The Monetary Board through Bank Indonesia (BI) deregulated the banking system, removed credit and interest rate ceilings and began to phase out BI liquidity credits, except for limited priority programs including certain agricultural and rural programs. Financial and tax reforms, fiscal austarity and a flexible exchange rate policy helped maintain macroeconomic stability. In 1986, overall GDP grew by 2.51, up from 1.11 in 1985 but lower than the 6.61 growth rate in 1984. The growth in agricultural output is estimated at 2.01 in 1986 declining from 3.21 in the previous year. However, agriculture remains the backbone of the Indonesian economy accounting for 241 of GDP, half of employment and over 601 of the value of non-oil exports. 2. The expansion of domestic credit, private sector credit, domestic liqaidity and money supply were all lower in 1986 than in the past year reflecting Government's aim of minimizing inflation. The growth of credit to state enterprises also slowed down due to budgetary stringency and tight credit policy, and net government deposits declined by indicating larger Central Government's recourse to banking credits. Nominal deposit and lending rates declined slightly while the rate of expansion of financial savings decelerated, pointing to the possible diminishing influence of the 1983 interest rate deregulation on stimulating financial savings. 3. In the face of a worsening external environment, the principal challenges for the Indonesian economy remain the adjustment to lower oil revenues and maintenance of financial stability in the short run, as well as reviving growth in the longer term. In agriculture, the major issues are how to maintain rice self-sufficiency and diversify agriculture on an efficient basis and how to sustain growth, employment, income and exports in the sec- tor. The rural financial sector has played an important role in agricultural development but its continued and expanded role would require major changes in policies and the design and management of credit programs to take account of the changing needs of the sector and external resource constraints. CO Credit Policy Framework 4. During the 1980s, OI has made significant progress in the area of financial and credit policy. The 1983 financial reforms which deregulated the banking system have led to a dramatic improvement in the level of financial savings. The rural credit market below the district level is market-oriented and subject to few interventions. A major change in 1984 was the shift in the - vii - rate of the Bank Rakyat Indonesia (BRI) village subbranches (unit desas) from being a conduit of subsidized and targeted credit to a financial intermediary mobilizing savings and providing general nonsubsidized rural credit. Despite these gains however, more than 75X of total rural and agricultural loans out- standing are provided under specific credit programs of the Government. These remain subject to GOI interventions such as directed credit to priority sec- tors, subsidized costs of funds through BI liquidity credits (refinance facil- ities), prescribed interest rates and liberal credit insurance. GOI also directly intervenes in the formal financial system through state ownership of most of the financial institutions in the country and regulation of the number of banks and bank branches in Indonesia. The main objective is to increase the supply of credit to priority sectors and "economically weak" groups on the assumption that credit is a precondition for investment and growth thereby accelerating development in the targeted sectors and improving income distribution. GOI has devoted a substantial amount of resources towards achieving this objective including agricultural credit subsidies to financial institutions estimated at Rp 334 billion in the period 1983-86. Although the objective of GOI interventions to promote growth, equity, and stability is laudable, satisfactory results are not guaranteed and the comparative effectiveness, efficiency and sustainability of these instruments need to be continually reviewed especially in the light of GOI's tight resource cons- troints. GOI would need to strengthen the research base for planning, deci- sion making and policy formulation in rural finance and conduct a regular assessment of the overall state of the rural credit sector and the performance of credit programs. Organization of the Financial System 5. At present, the Indonesian financial sector consists of BI, five state commercial banks, 69 private banks, 29 development banks, two savings banks, 11 foreign banks, 141 other financial institutions and over 17,000 rural financial institutions (RFIs), defined as credit organizations operating at or below the subdistrict (kecamatan) level. The RFIs can be classified into four groups: (a) 2,272 BRI unit desas; (b) secondary banks including 175 petty trader banks (Bank Pasars), 3,364 village banks (Badan Kredit Desas - BKDs), 217 village production banks (Bank Karya Produksi Desa - BKPDs), and 2,065 paddy banks (Lumbung Desas - LDs); (c) 479 government-owned pawnshops and 6,786 village cooperatives (KUDs); and (d) about 2,000 nonbank financial institutions such as the subdistrict credit units (Badan Kredit Kecamatans - BKKs) in Central Java, and the small credit program (Kredit Urusahan Rakyat Kecil-KURKs) in East Java. The state banks dominate the credit market and the capital market is in the early stages of development. 6. Indonesia has in place a large number and a great variety of BfIs at the kecamatan and village levels that are market-oriented and serving a ritatively large number of rural borrowers. Total assets and loans outstand- ing of these RFIs have steadily grown although their number have declined in recent years. The BRI unit desas and Bank Pasars are the largest lenders among RFIs while the most numerous village and paddy banks accounted for only 3.7Z of the total. The loan portfolio of RFIs is predominantly short-term and nonagricultural, comprised largely of trade and commerce credits. Only the BRI unit desas and Bank Pasars mobilize voluntary savings; the rest have -V - viii - compulsory savings schemes. User costs of credit are high, 20-2002 p.a., depending on the size of the loan, maturity, collateral requirement and type of institution. The large financial institutions serve the upper income segment of the credit market while the RFIs deal with the rural poor, the less profitable and more risky segment of the credit market. 7. In spite of the large number an4a complexity of financial institu- tions, the density of banks and nonbank officers is low compared with other developing countries. The large financial institutions except BRI have minimal presence or none at all below the district level. Private sector activity is also rare; the share of private banks in rural loans outstanding was only 1.3% in 1985. Of all the RFIs, only 175 Bank Pasars are privately owned, all the KUDs and 19 Bank Pasars are owned by cooperatives, and the rest are owned by the national, provincial or local governments. While the RFIs have grown at an impressive pace, they account only for 2.6% of total bank credits, 15% of rural credit, and 2% of total bank deposits. They deliver a limited range of finarncial services primarily of short-term, nonagricultural, and collateralized credit and are ill equipped, at their present stage of development, to handle term finance. Trends and Patterns in Agricultural Credit 8. Agricultural credit has increased in current and real terms and in relation to total credit and agricultural CDP. Its share of total credit is 8.4% compared with the 331 and 34Z shares of industry and trade respective- ly. The estimated 7.31 ratio of agricultural credit to agricultural GDP is also lower than the comparative ratios for industry at 64% and trade at 50%. This may reflect a number of factors including a high level of self-finance and credit from informal sources, lower returns and higher risks associated with agricultural investments compared with other sectors of the economy, in- adequate institutional outlets of formal credit, and interest rate restric- tions on agricultural credit. Agricultural lending has concentrated on the estat-A and tree crop subsector and North Sumatra, with a large area under tree crops, accounts for the highest share of agricultural loans among the dif- ferent provinces. Bank Bumi Daya (BBD) and Bank Rakyat Indonesia (BRI) have been the primary lenders for agriculture and rural areas while public estates and large borrowers have been the main beneficiaries. Of the total rural credit, less than 3% is granted on an unsecured basis which limits the access to credit of low income groups who do not possess land and property for collateral. An increasing proportion of agricultural credit is devoted to term credit (24% in 1980 to 531 in 1985) due to the expansion in estate and tree crop lending. The downward trend in short-term lending is primarily attributed to the decline of the Mass Guidance (BIMAS) program which provided short-term production credit to farmers for rice production in the 19709. Sources of Funds 9. The main sources of funds fot agriculture and rural credit are BI liquidity creit-3 and savings deposits. Despite the financial reforms of 1983, BI credits to agriculture have increased in nominal and real terms and relative to agricultural loans outstanding, indicating heavy reliance of the agricultural sector on BI funds. The amount of rural financial savings has - ix - doubled over the period 1983-85 but its share of total deposits has remained constant at 11X. The rural sector is still a relatively untapped reservoir of resources given the relatively high savings ratio of the Indonesian economy, the large number of people deriving income from the rural economy, and the large potential for transferring physical asset savings into financial forms. Adequacy of Funlds 10. If the volume of rural lending is compared with the total resources available from BI, rural savings and equity, the estimated surplus of funds for rural credit indicates a possible backflow of resources to the urban areas. The estimate implies that BI credits may have been used in place of internally generated funds (savings and equity) and rural credit may not be drastically reduced with the gradual withdrawal of BI/GOI support. Although effective credit demand in the rural areas may be weak and funds cannot be totally absorbed in the rural economy, it is also likely that funds are leaving the agricultural sectors for areas of higher returns, lower costs and less risks. At present, low and fixed lending rates on most agricultural and rural lending do not provide sufficient incentives for banks to lend to these sectors. II. Performance of Selected Rural Credit Institutions 11. Against the background of the credit policy and institutional framework and trends and patterns in rural/agricultural credit, the performance of key rural credit institutions (BRI, BBD, and selected RFIs) is analysed in order to identify the factors bearing upon their effectiveness and efficiency. Effectiveness and Efficiency of Selected Credit Institutions 12. BRI and BBD. As a mass supplier of retail credit and a collector of small savings, BRI is reaching a larger number of borrowers (about 2.4 million loan accounts at end 1985) than BBD whose loans numbered only 32,000 in 1985, of which 405 were to large public enterprises. BRI has extensive network of 326 branches and subbranches, 2,272 unit desas and 1,226 village posts throughout Indonesia, while BBD has only 128 branches (30 in Jakarta). Al- though BBD's mandate is to provide credit for estate agriculture and forestry, it could play a greater role in the rural financial market by expanding its network in the rural a-eas and lending to smallholders. 13. Since the state banks are expected to perform a dual function of being a profit making institution and a development agency subject to Government initiative and directives, GOI has supported BRI and BBD in all aspects of their operation. On the resource side, the Government has provided the banks with equity, deposits, liquidity credit and guarantees for external borrowings at low costs. On the lending side, banks have had access to credit insurance to minimize their default risks and guaranteed sources of income to enhance their financial viability. However, the Government has utilized the state banks as channels of targeted and subsidized credit and has mandated interest rates on their program credits. Thus, the interest income of BRI and BBD was insufficient to cover their lending costs in 1985 and their profits were low and largely attributed to cross-subsidies from other areas of operations (foreign exchange operations, general lending, BULOG working capital credits, KUPEDES, etc.). Both banks are dependent on GOI and external funds, although BRI is intensifying its rural savings mobilization effort through the unit desas. With high arrears and low provisions for bad debts, BRI and BBD would suffer financial losses and possible decapitalization if realistic debt write-offs were undertaken. BRI and BBD will need to increase their interest income and reduce their arrears and operating costs to maintain their financial viability. 14. Selected RFIs. Of the RFIs reviewed, the BRI unit desas, Bank Pasars, BKKs and Bank Dagang Bali (BDB), a privately owned commercial bank in Bali have the best prospects for long-term growth and sustainable operation. They are well managed, efficient and financially viable institutions providing access to credit for a relatively large number of borrowers. BDB and Bank Pasars set a satisfactory standard for a well run provincial credit institu- tion. On the other hand, BKDs have high operating costs and arrears that are directly related to their small transactions, scale of operations, concen- trated risks, nature of clientele and inadequately trained staff. Their resource base is weak since they do not mobilize voluntary savings but depend on initial capital endowments and loans from BRI. The KUDs have similarly high level of rrears and operating costs due to weak management and lax credit collection; they are also almost entirely dependent on GOI funds. To develop into full banking organizations, the BKDs must expand their size of operations, diversify their markets, improve staff training, and mobilize voluntary savings. The KUDs will need to strengthen their management, inten- sify collection effort and improve their accounting and financial systems in order that they could play an active role in the rural financial market. Factors Influencing the Performance of Credit Institutions 15. Concessional funds from BI have reduced the incentives for financial institutions to establish an adequate and self-sustaining resource base (consisting mainly of deposits and equity) which is necessary for the long- term development and sustainability of a financial institution. Low, fixed and administered interest rates have endangered the financial viability of institutions involved in subsidized credit programs and discouraged the banks from penetrating the rural credit market, since their interest rates do not cover the high transaction costs and high risks of lending to agriculture and the rural areas. Additional factors that affect the success of RFIs are their economies of scale, degree of homogeneity of risks, a sponsoring organization that provides technical assistance for management, supervision, control and the initial thrust and commitment to success. A sponsoring organization is especially relevant to the secondary banks, since BRI, the supervisory body of BI, has no inherent interest in fostering the development of its natural competitors. OI policies and regulations that inhibit the development of the rural financial system are restrictions on bank branching and the number of banks. Such policies reduce competition and the access to efficient financial services for much of the rural population. Finally, the development of the rural financial system depends on the role assigned to it by GOI. Recognition of the development of the rural financial market as an objective in itself - xi - will spur efforts to develop the rural financial system and thus improve the yield and productivity of capital, increase savings and investment, and improve credit availability. III. Assessment of Selected Credit Programs 16. The bulk of institutional credit to the rural sector is provided under numerous credit programs. BRI alone, for example, has 51 special credit programs and keeps accounts for 304 loan categories with different interest rates, maturities, and eligibility requirements. Most of them are commodity- specific and do not take into account the multi-enterprise nature of the farm household. The plethora of credit schemes impedes the efficiency of financial institutions and increases the administrative costs of programs and the transaction costs of borrowers. They will need to be rationalized and consolidated, possibly under a single loan fund, to improve the credit delivery system and enhance operational efficiency. Credit should also be provided on a farm system basis rather than to specific commodities only. Although credit may still be required within a project framework to provide an integrated package of inputs to borrowers, the future credit delivery system should be geared towards market demand, provide credit on a farm system basis rather than to specific commodities only, and minimize the distortionary impact of subsidies on the rural credit sector. 17. Among the credit programs, the review focused on the following to determine the factors affecting their performance: (a) World Bank-assisted tree crop credit scheme; (b) Small Investment Credit (KIK) and Permanent Working Capital Credit Programs (KMKP); (c) General Village Credit Program (KUPEDES); and (d) Badan Kredit Kecamatan (BKK) lending program. These programs are not strictly comparable since the tree crop credit schemes and KIK are long-teo-m credit programs while KUPEDES, KMKP, and the BKK programs mainly provide short-term financing. However, they represent two distinct types of credit schemes-supply-led and subsidized credit in the case of the tree crop and KIK/KMKP programs, and market-oriented and non-subsidized credit for KUPEDES and BKKs--which could yield valuable lessons for future credit delivery mechanisms. Main Features 18. Initiated in 1977, the Bank-assisted tree crop credit schemes util- ize the credit mechanism to recover costs from smallholder loan repayments after the expenditures in establishing tree crops by the state owned estates (PTPs) and project management units (PMUs) are converted into loans to individual smallholders. The banks act either as channeling banks under the Nucleus Estate and Smallholder (NES) projects or as executing banks for the PMU projects. However, in both cases the borrowers are selected by the PTPs and PMUs and the loans are practically risk free investments to the banks as they are fully guaranteed by GOI. The credit terms and conditions differ among projects depending on the commodity and the repayment capability of the borrower. They range from 0-12X interest rate p.a. with varying levels of subsidy from GOI and 10-20 year -naturity with 3-8 years grace period. Of the total smallholder credit component amounting to US$1.0 billion, about Rp 173 billion of expenditures have been incurred, and only 0.1X have been - xii - converted into loans. A number of problems is encountered in loan conversion related to the procedural requirements for debt conversion, quality of tree crop stands, smallholder credit amounts and debt repayment capacity of smallholders. A fundamental weakness in the credit program for NES projects is the lack of incentives for the participating parties to convert the loan. A major issue is how to maintain the supply of term credit for Government predetermined, economically desirable activities (e.g. tree crops) that are not financially viable at market interest rates, outside of the subsidized asd liquidity credit-financed priority programs of Government. 19. KIK and KMKP were established in 1974 to provide finance at subsi- dized rates (12%) for indigenous, small-scale labor-intensive enterprises unable to raise their own funds. The funds for KIK/KMKP come from BI liquid- ity credits (55%), IBRD (25%) and participating banks (20%) and are insured with the Credit Insurance Company of Indonesia (ASKRINDO) for up to 75% of the loan value. While the growth of lending under KIK/KMKP was spectacular in the early years, the amount of loans has declined in recent years. Loans out- standing as of December 1985 stood at Rp 1,223 billion, about 34% of total rural credit. 20. KUPEDES, introduced in February 1984, is the lending arm of the BU unit desas. In contrast with other credit programs, it provides general rural credit at effective interest rates of 22-32% p.a. with maturities ranging from 3 months to 3 years. Loans are usually secured by land, fixed assets, or other forms of property and are funded from BI liquidity credits (53%) at 12% p.a. interest rate, savings and other deposits (28%) and a grant from Government (19%). The growth of the KUPEDES program has been remarkable; the number of loans increased by 55Z from 640,000 in 1984 to 992,000 in 1985, and the value of loans almost doubled from Rp 171 billion to Rp 339 billion. 21. The BKKs were started in 1975 by the Central Java Provincial Government and provide general credit to the rural population on a non- subsidized and non-collateralized basis, with maturity periods ranging from 22 days to six months. Currently, there are 497 BKKs covering all kecamatans in Central Java and 2,609 village posts, where the BKK employees (three young school leavers per BKK) lend and collect repayments, usually on market days. BKKs have increased their lending program remarkably from Rp 213 million in 1972 to Rp 12 billion in 1985, and since inception, have extended about 4.7 million loans %534,000 in 1985). Program Performance 22. The KUPEDES and BKK loan programs have performed well compared to the KIK/KMKP and Bank-assisted tree crop credit schemes. Given the small loon sizes of the BKK and KUPEDES loans, the absence of collateral requirement of BKKs, and accessible locations of the BKKs and BRI unit desas, they have provided wide access to credit for the rural population. In contrast, most KIK/KMKP lending have been in large loans to established entrepreneurs with proven experience. Small borrowers have effectively been screened out by banks that are unable to cover their lending costs by altering the adminis- tered price of credit. The loan repayment rates under the KUPEDES and BKK programs have been high. Although a survey on the impact of KIK/KMKP - Xiii - indicated that employment, rural investment and value added have increased, the low collection rates of these programs have undermined their viability and sustainability. The dependence of KIK/KMKP and tree crop credit schemes on low cost government funds has not provided any incentives for the banks to recover payments and maintain the financial viability of the programs. It also reinforces the perception of borrowers that loans are from the Government and may be considered grants. Although the lending rates of KUPEDES and BKK are high compared to the subsidized rates of KIK/IMKP and tree crop financing program, the borrowing cost of KUPEDES and BKK borrowers may be lower because of the accessibility of BKK and KUPEDES offices and lower transaction costs from simple loan documentation and shorter loan processing time. The inter- mediation margins for the tree crop credit schemes and KIK/KMKP have not been sufficient to cover the operational cocts and default risks of the banks and provide little incentives for banks to participate in these credit schemes. The cost of annual credit subsidies to the financial institutions are estimated at about Rp 80 billion p.a. for KIK/KMKP and Rp 42 billion p.a. for the tree crop projects and exclude subsidies related to loan defaults, direct interest rate subsidies to borrowers, and the share in the loan insurance premium paid by GOI. 23. The major lessons learned from the successful operations of the KUPEDES and BKK lending schemes are as follows: (a) a successful credit pro- gram can be designed to reach low income groups through market interest rates, absence of collateral, accessibility of bank offices, small loan sizes, simple lending procedures; (b) interest rates that cover the costs of lending to small borrowers have enabled the KUPEDES and BKK programs to provide wider access to credit for the rural population, maintain the financial viability of the programs, and ensure the continued interest and participation of banks in the programs; (c) availability of credit and lower transaction costs through simple administrative and lending procedures are likely to be more attractive features for an "economically disadvantaged" borrower than "below market" interest rates; (d) successful repayment is attributed to interest rates that adequately cover lending and collection costs, financial incentives to borrowers for prompt repayment, staff bonuses based on efficient and profit- able operations, character based and repeat lending, and full accountability of the bank for the lending risks; and (e) liberal credit insurance facilities encourage lax credit collection and lead to higher defaults; and (f) an adequate resource base of internally generated funds from earnings and savings deposits is essential to maintain incentives for sound loan management. IV. Financial Resource Mobilization 24. In view of the vital role that financial resource mobilization play in the development of viable and sustainable rural credit institutions and programs, the experience of selected savings instruments is analyzed to determine the factors that influence financial savings mobilization. 25. Availability and effectiveness of savings instruments. Rural savers have a limited choice of savings instruments in terms of liquidity, risk, maturity and other features. Savings accounts are the most common instru- ments, usually the Small Saving Program (TABANAS), the national savings - xiv - scheme, or more recently, the Village Savings Program (SIMPEDES), a BRI unit desa savings program. The current National Savings llrive Committee has concentrated on the promotion of TABANAS to the exclusion of other savings instruments. However, TABANAS deposits have shown little progress in recent years while SIMPEDES has grown at a spectacular pace. A combination of zero withdrawal restrictions, a consumer goods lottery plan, and relatively high interest rates has made it successful. 26. The commercial banking sector has not been particularly int.ovative in raising savings in either rural or urban areas. The concept of paying salaries directly into a savings account is quite n*'w and there are few examples of goal savings accounts, local savings plans, and door-to-door savings. While existing savings instruments have offered high positive real returns over the last few years, the net return to all savers may not have been positive due to high costs of rural dwellers outside of Java and Bali who have poor access to banks. Rural savers have also limited access to longer term instruments which entail a larger initial investments than most rural savers could afford. There is considerable scope for expanding the range of savings instruments for rural savers including door to door savings program, target savings plan, life insurance policies, bond issues etc. 27. Factors Affecting Financial Resource Mobilization. Perhaps, the most important disincentive to mobilizing rural financial savings is the access to cheap BI funds, because savings are more expensive and difficult to tap than BI liquidity credits. The greater the access of a financial institu- tion has been to BI credit the lower the amount of savings mobilized. Another constraint is the limited number of financial outlets at the village level and their narrow scope of operations which increase the transaction cost and lower the expected returns of the rural savers. The narrow range of financial instruments available in the market and their terms and conditions limit the degree of choice for potential investors and the pace at which physical asset savings are converted to financial forw'. Finally, restrictions on savings deposit rates and maturities tend to limit the freedom of banks to decide the terms and conditions attached to their savings instrument, and reduce the incentives for developing their own savings mobilization mechanisms. V. Women in the Rural Financial Sector 28. To assess the role of women in the rural credit market, the participation of women in rural ctedit and savings and the factors influencing their demand for and supply of credit are analyzed. The limited data available indicate that Indonesian women participate actively in the rural financial sector. Women are 23.4% of KIK/KMKP borrowers; 25% of KUPEDES borrowers (at BRI Unit Desas); and 29% of borrowers at one Bank Pasar. Women are a higher percentage of borrowers in non-bank financial institutions such as BKK (60%) and KURK (57X). They may be as many as 801 of borrowers from government pawnshops. Village-level studies show that women also borrow from informal sources, such as suppliers, traders, neighbors, money lenders, etc. and have organized traditional savings and loan associations, such as "arisans" and "simpan pinjams", in rural areas. - xv - Factors Affecting Women's Demand for Credit 29. Women's demand for credit is determined by their economic activities as well as their role in the household. The high proportion of women among the self-employed suggests that women's demand for credit will not be limited to consumption, but will include demand for working capital and investment credit for their businesses. Women's lower incomes probably translate into demand for smaller size loans, on average, than those demanded by men. Cultural factors and lack of information about lending may also serve to limit women's demand for credit. But given the 'evel of their economic activities and the multiple time constraints that women face because of their dual productive/reproductive responsibilities, it is the transaction co,ts associated with borrowing that are likely to be the most important factor limiting women's demand for formal sector credit. Based on women's high level of participation in financial institutions which maintain lending operations at the village level and keep loan application forms very simple, women's demand for credit is likely to be greater when transaction costs are low, even if interest rates are much higher than those charged by the commercial banks. Factors Affecting Supply of Credit to Rural Women 30. At the sites and institutions visited, there is no evidence that negative perceptions of women on the part of lenders limit the supply of credit to women, just because they are women. Other more important factors are those that affect small borrowers, regardless of their sex. Lenders are reluctant to lend to small borrowers including women, because of the high costs of making small loans. Government restrictions on lenders, such as interest rate controls, may lead them to ration credit away from small borrow- ers, and those borrowers they perceive to be more risky. Women's lower educa- tional level in rural Indonesia also puts them at a disadvantage in preparing loan applications, investment plans and financial statements that lenders require. If loans are targetted to specific economic activities or crops, if land is required as collateral, a husband or male relative is required as a cosigner for loans, complicated application forms are used, and/or only one loan is permitted per household, lenders are in fact channeling lending away from female borrowers. Therefore, while lenders interviewed in Indonesia do not appear to discriminate against women because oi their sex, some of their policies have the effect of limiting women's access to financial services. In cases where lenders had made a conscious effort to make credit available to smaller borrowers, by eliminating some of these structural barriers, such as in the pawnshops and BKK-type institutions, the participation of small borrow- ers and women is high. The Role of Women in Rural Savings 31. There is no data for income and expenditure in the household by sex but it can be assumed that women's incomes are lower than those of men. About 202 of rural households are headed by women, but a high proportion of these are in the lowest income groups. In addition, women who are heads of house- holds tend to be older and thus less important as savers. Thus, women would probably account for a smaller proportion of rural savings. Staff of banks visited indicated that women are more important as savers in small saver - xvi - accounts. Studies also show that Indonesian women help make decisions about the use of incomes and disposal of household assets. Therefore, their potential role in savings mobilization could be greater than their individual income levels would indicate. To the extent that women are holding assets in non-financial forms (such as jewelry, batik, and animals), expanding their access to financial institutions could promote the conversion of these assets into bank deposits. VI. Recommendations 32. Significant gains have been achieved by GOI in deregulating the banking system, increasing agricultural credit, improving rural savings deposits and establishing a network of RFIs that is market-oriented and subject to few restrictions. However, there is substantial scope for developing a viable and sustainable rural financial system which would mobilize savings and build up capital from efficient operations, and provide at reasonable user costs a broad range of financial services to the majority of the rural population. In order to enhance the development of the rural financial sector and thus improve the availability and efficient use of credit in agriculture and the rural economy, the following measures are recommended. 33. Gradually increase the interest rate on BI liquidity credits for agricultural and rural credit programs (say, 2% per year) to the marginal rate paid on savings deposits over a specific time period (e.g., three or four years), to encourage financial institutions to enter the rural credit market and mobilize savings. 34. Raise onlending rates on agricultural and rural program loans (e.g. 2% per year) to enhance the efficient allocation of resources, improve the viability of financial institutions and credit programs, encourage banks to lend to agricultural and rural enterprises, ant provide wider access to credit for low income groups. GOI should consider the introduction of variable rate lending with limits on the interest change per year especially for the provision of term credit to agricultural investments with long gestation period, e.g. tree crops. 35. Liberalize policies on bank branching and number of banks to promote competition among financial institutions, increase efficiency of the system, and achieve denser penetration of the rural financial market by financial institutions. Bank Bumi Daya should be encouraged to establish a subbranch network like BRI's and the use of subbranches, agency operations, village posts and mobile units should be actively promoted by MOP and BI. GOI should consider opening the banking industry to new firms which would infuse addi- tional capital into the sector and promote competition. Safeguards from fraud and mismanagement can be installed through fidelity and deposit insurance. 36. Strengthen the small RFIs by providing technical assistance, improv- ing supervision, allowing them to mobilize voluntary savings, and encouraging them to establish linkages with national level financial institutions through agency, subcontracting and/or co-financing arrangements. This would provide - xvii - capital and expertise to RFIs and expand and diversify their markets. Train- ing and incentive schemes will need to be developed to upgrade the staff bank- ing skills of the small RFIs. The financial operations of KUDs should be con- centrated on their savings and loan associations and their management, accounting and financial systems strengthened through technical assistance. BI should take over the direct supervision of the secondary banks or delegate the responsibility to another entity other than BRI who is the natural competitor of the secondary banks. 37. Improve design of future credit programs by: (a) adoption of market rates for the institution's cost of funds and onlending rates; (b) simplifying lending procedures; (c) establishment of lending criteria on the basis of character, savings and repayment record rather than on collateral; (d) full accountability of credit risks by lending institutions; (e) provision of financial incentives for borrowers for prompt loan repayments and imposition of sanctions on defaulting bor.owers; and (f) introduction of a staff bonus system based on efficient and profitable credit operations. Credit programs should also take into consideration the multi-enterprise nature of farm households. If there is a demonstrated need for project financing, e.g. for tree crops, the banks should perform a financial intermediary role, assume a major part of the credit risks, charge market and variable onlending rates, and borrow funds at market rates. Subsidies, if found necessary and desir- able, should minimize the distortionary impact on the financial system and be severely limited, transparent, temporary and sharply focused on target groups. If the amount of subsidies envisaged is large and highly distor- tionary, GOI should consider using an alternative means of extending subsidies outside of the credit system such as grants and cess taxes, or channeling and recovering funds through project implementing units. 38. To consolidate existing special credit programs, GOI should consider the establishment of a National Development Loan Fund (NDLF) which would be capitalized by the total outstanding liquidity credits of BI for program loans and any outstanding agricultural and rural credits directly extended by GOI. BI should manage and control NDLF but additional resources should come from the national budget to ensure transparency of credit assistance for GOI. NDLF can also be authorized to borrow from international organizations and issue bonds or other debt instruments to sustain its credit operations. The details involved in creating NDLF and phasing out low-cost BI liquidity credits need further study and preparation but the objectives would be to promote efficiency of financial institutions, accelerate financial resource mobiliza- tion, rationalize the structure of credit programs and provide flexibility to banks in designing their own credit schemes. Since BI liquidity credits are funded from increases in money supply and government deposits, the separation of NDLF from the BI and GOI accounts would isolate the growth of credit from the expansion of the monetary base and reduce the possible inflationary impact of credit expansion. 39. In the short run, increase the premium rates and decrease the coverage of loan insurance for BI/NDLF loans to account for the risks of default and improve the financial position of the credit insurance agency. However, the effectiveness and long-term viability of the credit insurance policy should be assessed and the justification for its existence, reviewed, - xviii - in the light of the observed adverse consequences of policy on loan collec- tions. Loan insurance coverage should not be provided for future credit pro- grams to encourage sound loan management and stringent loan recovery by banks. 40. Review and possibly restructure terms and conditions for existing tree crop credit schemes to provide financial incentives for banks to convert and collect the loans. 41. To accelerate savings mobilization, (a) continue policy for positive real interest rate; (b) remove interest rate controls on TABANAS, TASKA and government time deposits of 24 months, and restraints on the maturity of time deposits of secondary banks and approval of pension and super annuation funds; (c) expand the role and responsibilities of the National Savings Drive Committee to promote a wider choice of savings instruments, to assist banks and RFIs in developing savings campaign, and to improve the data base on rural financial savings; (d) increase the range of financial assets available in rural areas, particularly ones witb longer maturities such as government agency debt securities, life insurance policies, contractual goal savings plans (for retirement, home, education) and equities, local savings plans for local needs, bond issues, etc.; (e) encourage door-to-door financial services; (f) allow for better competition among savings instruments by giving banks greater freedom to set the terms and conditions of these accounts; and (g) introduce institutional incentives for savings mobilization through staff bonuses or prizes. 42. To continue and expand women's participation in rural finance, (a) strengthen and support financial institutions and lending programs that use alternatives to collateral through technical assistance, improved linkages with and supervision by larger financial institutions and encouraging nonbanks to mobilize voluntary savings; (b) restrict targeting credit programs or financial institutions by sex or client's specific activity; and (c) on loan procedures, standardize cosigning requirements for male and female borowers, relax regulation of one loan per family and streamline paperwork required of borrowers. 43. To strengthen the research base and assist policymakers on rural financial issues, (a) establish a unit within BI, MOF or NDLF with the responsibility and capability to do policy research on rural finance, to monitor and evaluate credit programs, and to provide technical assistance to small RFIs. This should be closely coordinated with the planned research component of the USAID-supported Financial Institutions Development Project; and (b) collect sex-disaggregated information on borrowers and savers and conduct research on priority issues surrounding the role of women in the rural financial sectors. World Bank Support to Rural Financial Reforms 44. The Bank could assist GOI in formulating a detailed action plan for carrying out the reforms outlined above. Based on an agreement for adoption and scheduled implementation of the action plan, the Bank could possibly provide BI/NDLF a line of credit to finance agricultural and rural enterprises and future Bank projects in different subsectors such as secondary food crops, - xix - livestock forestry, agroprocessing facilities, etc. The Bank could also assist in the institutional development of RFIs and the strengthening of the research base for policy formulation in rural finance through technical assistance programs. Through its support, the Bank would help mobilize savings resources, provide capital to rural and agricultural activities, and contribute to rural financial policy reform. I. OVERVIEW OF THE RURAL CREDIT SECTOR A. Recent Macroeconomic and Financial Developments 1.1 In spite of the external shocks to the Indonesian economy over the past few years, as declining prices of oil and primary export commodities seriously weakened the country's resource position, macroeconomic stability has been maintained. Financial, tax and trade policy reforms, fiscal auster- ity and a flexible exchange rate policy helped contain the external deficit to less than 3% of GNP in 1985 and inflation at about 4%. However, overall GDP decelerated considerably to an estimated 1.1% from the 6.6% growth rate in 1984 (Annex 1, Table 1). In 1986, GDP grew by an estimated 2.5% and agricultural output growth was 2.01, declining from the previous year's rate of 3.21. However, agriculture remains the backbone of the Indonesian economy accounting for 241 of GDP, half of employment and over 601 of the value of non-oil exports. Over the last 15 years, agricultural output has grown by about 41 p.a., above that achieved in most other East Asian countries and major oil-expurting countries. This impressive record is primarily due to the rapid expansion of rice production as well as other crops over the past decade. 1.2 Over the period 1978-82, the combination of Government's comfortable financial position due to high oil prices, its policy of channeling surplus government funds through the banking sector, and rapid economic growth (8X p.a. growth in GDP) provided the basis for a rapid expansion of the banking system. Total assets of the deposit money banks (DMBs)--the dominant finan- cial institutions in Indonesia--almost quadrupled and experienced a real growth rate of about 11 p.a. The Indonesian financial system was tightly controlled by the authorities. BI set deposit and lending rates of state banks and detailed credit ceilings for all banks, and guided the allocation of credit through the liquidity credit mechanism. However, as oil prices weakened in 1982/83 and the Government's financial position came under pressure, the Government of Indonesia (GOI) introduced a major reform of the financial sector in June 1983. The Monetary Board, through Bank Indonesia (BI), the nation's central bank, partially deregulated the banking system, removed credit and interest rate ceilings and began to phase out BI liquidity credits, except for limited priority programs including certain agricultural and rural programs. 1.3 The deregulation allowed government banks to set their own interest rates and resulted in a marked increase in the real interest rates available to savers. With the open nature of Indonesia's financial sector, interest rates are largely a function of overseas interest rates combined with the expectation of Indonesian savers of the rupiah devaluation. Real interest rates on normal savings deposits (up to Rp 1,000,000) 1- increased from 3.5% in 1983 to 10.71 in 1985 while time deposit rates climbed from 6.81 to 16.7% over the same period (Figure 1.1 and Annex 1, Table 2). Savers have responded 1/ Under the National Small Saving Program (Tabungan Nasional--TABANAS). Real Rates of Interest of Deposits i7- I | I| 14 12 4J~~ 1981 1982 19B3 IQa4 1185 ism YEAR 1X Savins Deposit + Timn DepositX 11~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~1 * S Nominal and Real Bank Deposits 14 r 13 00 X~~~~~~~~I 1 f84 IM 3 Z X * {13X 0.~~96 89S 9; 941 s 7~~~ ¢ >in - 4 - accordingly and between 1982 and 1985 deposit levels increased by 51% p.a. in nominal and 33% p.a. in real terms (Figure 1.2 and Annex 1, Table 3). Depo- sitors' preference for various types of deposits and maturity also changed over this period. In 1981 demand deposits were the largest single source of bank funding, but during 1983-85 savings and time deposits grew by an average of 581 p.a. and in 1985 amounted to 74% of total bank deposits. 1.4 In spite of the 1983 reforms which envisaged a reduction of BI li- quidity credits, total liquidity and direct credits from BI increased from Rp 6,515 billion in 1983 to Rp 8,635 billion in 1986, a 10 nominal growth rate or a 3% growth rate at 1983 prices (Figure 1.3 and Annex 1, Table 4). 8I liquidity credits comprised 4if of total private credit in 1983 and 44% in 1984. The share of priority I credits in total BI credits also increased from 31% in 1983 to 49% in 1986 due mainly to the expansion in credit to exports and tree crops. 1.5 In 1986, the monetary authorities faced a difficult challenge due to the uncertainty about the value of the Rupiah and financial pressures arising from the slowdown in market growth. Exchange rate speculation led to two epi- sodes of capital flight during February/March and December, but financial sta- bility was maintained due to BI's intervention. Adjusting for valuation changes resulting from the devaluation of Rupiah on September 12, 1986, the expansion of domestic credit (11%), private sector credit (14%), domestic liquidity (12%) and money supply (12%) were all lower than in 1985 reflecting Government's aim of avoiding an inflationary increase in liquidity (Annex 1, Table 5). Credit to state enterprises rose by only 3.2%, compared to a growth rate of 9.8% in 1985 due to budgetary stringency coupled with a tight credit policy. Net government deposits declined by 21.2%, compared to a growth rate of 7.1% in 1985 and 58.8% in 1984 indicating larger Central Government's recourse to banking system credits. Nominal deposit rates were slightly lower during 1986 than in the previous year due to the comfortable liquidity posi- tion of many commercial banks and the decline in international interest rate (Annex 1, Table 6). Nominal lending rates also declined slightly during the year, reflecting lower deposit rates, slowdown in credit demand and greater price competition among banks for quality customers. As a result of the sharp decline in real deposit rates, the rate of expansion of financial savings has slowed. Time, savings and foreign currency deposits rose by only 11.6% in 1986 compared to 39.5% in the previous year (Annex 1, Table 5). These trends would seem to indicate that after three years, the impact of interest rate deregulation on stimulating financial savings is weakening. 1.6 In the face of a worsening external environment, the principal chal- lenges to the Indonesian economy remain the adjustment of the economy to lower oil revenues and the maintenance of financial stability in the short run, as well as reviving growth in the longer term. In the agricultural sector, the 2/ Priority or program credits are largely subsidized credit for specific target groups and activities (except the General Village Credit Program, KUPEDES) while nonpriority credits are generally larger loans and to specific enterprises, mostly government owned. B I Liquidity arnd Direct C:redits 9~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -~~~~~~~~,~.- -v/- Prssaiority Nr . n-Pri*rity Direct C,di* '' -~~~~~~~~~~~~~~,~0 -, I- '~~~~~~~~~~~~~~~~~~ /, ~ ~ ~ ~ ~ L - 6 - major issues confronting GOI are how to maintain rice self-sufficiency and diversify agriculture on an efficient basis given the deterioration in the resource position of the Government, and how to sustain growth and employment, improve farm incomes, and expand exports in the sector. The rural financial sector has played an important role in agricultural and rural development, but its continued and expanded role would require major changes in policies and the design and management of credit programs to take account of the changing needs of the sector and external resource constraints. B. GOI Credit Policy Framework 1.7 During the 1980s, GOI has made significant progress in the area of financial and credit policy. The financial reforms in June 1983, which were intended to stimulate private savings, improve resource allocation and reduce the dependence of the banking system or BI liquidity credits, have had a dra- matic impact on the banking system. Nominal (and real) deposit rates paid by state banks have increased sharply and there has been a rapid increase in bank deposits of state banks in the post-June 1983 period. The rural credit market below the district level is largely unregulated. The role of the Bank Rakyat Indonesia (BRI) village subbranches (unit desas) has changed from being a channel of subsidized targeted credit to a financial intermediary mobilizing savings and providing general unsubsidized credit to the rural sector. How- ever, despite these gains, a significant number of agricultural and rural credit programs remain subject to GOI interventions including directed credit to priority sectors, subsidized cost of funds through BI liquidity credits, mandated interest rates and a liberal credit insurance policy. At present, BI has about 18 credit programs for investment and working capital loans to vari- ous sectors of the economy, 11 of which are for agricultural and rural activi- ties (Annex 1, Table 7). In addition, there are no less than 49 specific cre- dit schemes for crops, livestock or fisheries under foreign-assisted projects, supported mainly by the World Bank and the Asian Development Bank and 43 tree crop projects financed by GOI with credit components. BRI alone has 51 diffe- rent credit schemes and keeps the account for 304 loan categories with diffe- rent interest rate, maturity and eligibility requirements. 1.8 The present credit schemes have evolved from the Mass Guidance (BIMAS) credit program established in 1970 when short-term production credit was provided to farmers as an incentive to adopt the improved package of tech- nology and increase rice production. Initially limited to rice cultivation, the program was extended to secondary crops, including corn, groundnut, soy- bean, dryland rice, sorghum and green peas. Although the objective of rice self-sufficiency was achieved, studies reveal that this was due to the favo- able price incentives effected between 1976 and 1982 rather than to credit._/ Due to large defaults and declining number of eligible borrowers, BIMAS was later allowed to lapse. 3/ Rice Intensification, Development Program Implementation Studies Report No. 2 (December, 1983). - 7 - 1.9 The objectives of the present credit schemes include: (a) acceler- ating development in priority sectors (e.g. exports, tree crops); (b) provid- ing credit to economically weak groups (e.g. small- and medium-scale entrepre- neurs, cooperatives, farmer groups) who are normally screened out of commer- cial credit markets; (c) increAjing production and rural incomes; and (d) ge- nerating employment through assistance to small scale labor intensive enter- prises. The terms and conditions, borrower eligibility requirements, report- ing and monitoring conditions vary among different credit schemes. However, except for a few programs such as the General Village Credit Program (Kredit Umum Pedesan--KUPEDES), interest rates are invariably set by the Government at 12% p.a.; about 50-100% of the loans are funded by BI or GOI at 3% interest rate p.a.; and over 75% of the credit risks are insured by the Government (Annex 1, Table 7). 1.10 Low and fixed lending rates are intended to reduce the cost of cre- dit for disadvantaged borrowers (farmers, smallholders, small enterprises, etc.). It is assumed that these groups need below market interest rates to make their enterprise bankable and to compensate for the low levels of return found in vital areas such as agriculture. The 3% interest rate on BI and GOI credit to the financial institutions is designed to direct credit into prede- termined priority areas for which the banks would not provide credit under normal circumstances. In the same vein, the credit insurance policy attempts to induce the financial institutions to lend to specific high risk target groups and activities by insuring the banks from risks of loan defaults. The Credit Insurance Company of Indonesia (PT Asuransi Kredit Indonesia--ASKRINDO) was established by GOI in 1970 to provide loan insurance for specific credit schemes and the State Corporation for Expanding Cooperatives Financing (Perum Peagembangan Keungan Koperasi--PERUM PKK) was created to cover cooperative loans. About 75% to 95% of the credit risks involved in nonrepayment is co- vered by these credit insurance agencies. The insurance premium rates amount to 3-5% of the loan value shared equally by the participating bank and BI/GOI. The Government share in the premium payment represents an additional subsidy to the credit syste-m. 1.11 GOI also directly intervenes in the formal financial system through state ownership of most of the financial institutions in the country, and the overall regulation of the development of the financial sector. The large fi- nancial institutions and rural credit entities have been created and sponsored by the national, provincial and local governments in order to increase the supply of credit to the agriculture and rural sectors on the assumption that the supply of credit was a precondition for investment and growth. The deci- sion to intervene directly in the financial institution development is based on what was perceived as market failure, represented by the absence of private formal lending institutions in rural areas. Such absence was interpreted as an indication that rural finance was unprofitable due to the high level of po- verty of rural dwellers and to the immaturity of rural financial markets. Be- cause the initial costs of setting up rural credit institutions are too high for the private sector to undertake, the Government was willing to underwrite the initial subsidies and establish its own financial institutions. - 8 - 1.12 GOI policy also regulates the number of banks and bank branches in Indonesia. Because of previous bank failures and frauds, as well as the limi- tations of BI's supervisory capabilities, GOI's policy has been directed to- ward fewer and bigger banks that are easier to manage, regulate and control.4/ Thus, the banking sector has effectively been closed to new entrants, Private banks are required to undertake several mergers before being granted permis- sion to expand, and requirements for obtaining a foreign exchange license are very stringent. GOI policy is also more lenient towards state commercial banks, as they are governed by different rules from those of the private banks. Within the state banks, approval of a new branch presents little dif- ficulty, the Ministry of Finance (MOF) requiring only a favorable feasibility study of potential deposits, lending and profitability. For private national banks, however, the requirements are more demanding. In order to open a branch in Java or Bali, the bank must first open another branch outside these islands. Thus, private banks face limits to expanding their network and com- peting with state banks. 1.13 Although the objectives for GOI's interventions are laudable, suc- cessfus. results are not guaranteed. Government sponsored, subsidized credit programs ere usually plagued by arrears which threaten the sustainability of institutions and programs, and represent a drain on the budgetary resources of the Government. Interest rate subsidies may undermine the viability and sustainability of financial institutions, lead to misallocation of resources, and aggravate the inequitable distribution of income as less profitable enter- prises and bigger borrowers would tend to receive the cheap loans. Subsidized cost of funds reduces the incentives for savings mobilization and financial deepening in the rural areas and the benefits of the credit insurance policy may not outweigh its costs. The subsidies to agricultural borrowers implicit in below-market lending rates are estimated at Rp 372 billion over the period 1983-86 assuming a conservative market lending rate of .8X (Annex 1, Table 8) while GOI costs of subsidizing the funds of financial institutions amounted to Rp 334 billion in the same period, an average of Rp 83.5 billion annually (Annex 1l Table 9). Subsidies can play a positive role ia the development strategy of a country. However, to be effective and efficient they should be sharply focussed in order to reach the targeted beneficiaties; limited in du- ration and subject to review at regular intervals; and transparent so that the costs and benefits are clear both to the budget decision makers and to the recipients. 1.14 The most difficult obstacle to the success of a government financial institution lies in the inherent conflict of being both a financial institu- tion and a development agency subject to Government initiative and directives. With boards dominated by Government officials there may be a strong political 4/ This fear seems to be unfounded since the major part of the financial system is owned by government and the safety of depositors funds can be protected from mismanagement by fidelity and deposit insurance. For example, the recent failure of a Bank Pasar due to fraud and mismanage- ment indicates the need to strengthen prudential regulation rather than to restrict the number of Bank Pasars. - 9 - focus to programs with adverse consequences on managerial efficiency and fi- nancial prudence. Heavily dependent on cheap and accessible resources from Government and external donors, there is less pressure to be efficient, to mo- bilize savings or to enforce a more stringent loan recovery program. The managerial structure is sometimes fairly rigid, resembling that of a Ministry rather than of a bank, and staffing standards usually suffer from low salary scales and inflexible personnel policies; hence there is little incentive to contain administrative costs and generate financial innovations. The result- ing structure of the financial system may not be particularly healthy or dyna- mic. The dominance of state enterprises and a fixed number of private banks that are content with their protected market are less likely to produce compe- titive pressure to improve the efficiency and effectiveness of the financial system. 1.15 Ideally, the rural financial system should consist of a number and variety of viable and self-sustaining institutions located at accessible places, mobilizing savings and building up equity from efficient operations, and providing a broad range of financial services (loans of various sizes and maturities, guarantees, money transfers, etc.) to the majority of the rural population. User costs of financial services should be reasonable and at levels which cover the institution's costs of funds, risks and operations which change over time. Competition among institutions is necessary to gener- ate financial innovations, increase operational efficiency and improve access to credit of "economically weak" groups. While significant progress has been made in policy reform towards the achievement of this ideal system, there is still considerable room for improvement. The effectiveness, efficiency and sustainability of existing policy instruments and institutional structure need to be reviewed especially in the light of GOI's resource constraints. There is also a need to strengthen the research base for planning, decision making and policy formulation in rural finance. There has been limited research on the problems of the rural financial market and the work has mainly consisted of studies of limited value on measuring the impact of loans on borrowers. In view of the large amount of resources devoted to rural credit, GOI needs to conduct a regular assessment of the overall state of the rural/agricultural credit sector, a systematic review of credit programs and detailed studies on priority issues affecting the rural financial system. C. Organization of the Rural Financial System 1.16 Indonesia's Financial Sector. BI is at the apex of the finkancial system composed of 5 state commercial banks, 69 private banks (ten of which are licensed for foreign exchange transactions), 11 foreign banks, 29 develop- ment banks, 2 savings banks, 14 nonbank financial institutions, and 127 insur- ance and leasing companies (Figure 1.4). The capital market is still in the early stages of development. In addition, there are more than 17,000 rural financial institutions (RFIs) in Indonesia, defined as entities performing some kind of financial intermediation at or below the subdistrict (kecamatan) level. They can be classified into four groups: (a) 2,272 BRI unit desas su- pervised by BI; (b) secondary banks supervised by BRI on behalf of BI which include 175 petty trader banks (Bank Pasar), 3,364 village banks (Badan Kredit Desas--BKDs), 217 village production banks (Bank Karya Produksi Desa-BKPD), and 2,065 paddy banks (Lumbung Desas--LDs); (c) 479 pawnshops ana 6,786 vill- - 10 - Fi8ure 1. 4 W II t i,n;i - 11 - age cooperatives (KUDs) supervised by the Ministries of Finance and Coopera- tives, respectively; and (d) about 2,000 nonbank financial institutions, such as the Subdistrict Credit Units (Badan Kredit Kecamatans-BKKs) in Central Java and the Small Credit Program (Kredit Urusahan Rakyat Kecil-[ URKs) in East Java, supervised by the regional development banks (Bank Pembangunan Daerah-- BPDs). A brief description of the institutional components of Indonesia's fi- nancial sector as well as the summary features of a number of informal credit arrangements are found in Annex 6. Annex 1 Table 10 indicates the sources and uses of funds for the various categories of banks and financial institutions. 1.17 The five state banks 5/ dominate the credit scene, accounting for about 73% of the total assets of deposit money banks in December 1985, 74% of total loans outstanding, and 67% of total funds (Annex 1, Table 11). In spite of the diversity and number of RFIs, their loans outstanding amount to only Rp 660.5 billion at end June 1986, or 2.6% of total bank credits (Annex 1, Table 12). Most rural credit is disbursed by the commercial and development banks (85%) while the RFIs account for the remaining 15% (Annex 1, Table 13). The primary lenders for agriculture are BBD (46%), BRI (28%) and BEII (18%) mostly for estates and tree crops (Table 1.1). BRI, the main conduit for ru- ral credit, operates 297 branches, 2,272 unit desas and 1,226 village posts nationwide and has the most extensive banking network in rural Indonesia. Table 1.1: DISTRIBUTION OF AGRICULTURAL CREDIT BY TYPE OF BANK /a (December 1985) Rp X of (billion) total Subsector BRI 544 28 General agriculture and rural BBD 912 46 Mostly tree crops and estates BEII 353 18 Mostly tree crops BNI '46 45 2 General BDN 31 3 General Regional Development Banks (BPDs) 35 2 General Private banks 46 2 General Total 1,966/b 100 /a Annual Reports of individual banks and BI. 7b The total indicated here is more than Rp 1,656 billion shown by BI because of differences in the classification of agricultural credit between the banks and BI. Most of the difference is in lending for estate crops. 5/ The state connercial banks specialize according to specific areas of responsibility: (a) Bank Bumi Daya (BBD) estate agriculture and fores- try; (b) Bank Rakyat Indonesia (BRI) agriculture, fisheries, coopera- tives, rural development; (c) Bank Export Import of Indonesia (BEII)--exports and imports; (d) Bank Negara Indonesia (BNI) 1946--industry; and (e) Bank Dagang Negara (BDN)--trade. - 12 - 1.18 Rural Financial System. The rural financial system in Indonesia is characterized by the absence of commercial and development banks in the rural areas except for BRI which operates in the rural credit market because of its mandate. Even BBD which must disburse rural credit for estate and tree crops financing has a very low rural profile. Primarily an urban bank, BBD lends most of its resources to large, mostly public estates and a small amount to small, mostly urban borrowers. Foreign banks have branches only in Jakarta and even most regional development banks are in urban renters. Thus, in spite of the large number of financial institutions, the density of banking offices and nonbank financial institutions in Indonesia is low (Annex 1, Table 14). The number of people per bank office in Jakarta is 18,287 compared to the urban average of 9,826 in Thailand and 6,910 in India. Outside Java, the average is about 40,220 people per bank office compared to the rural average in Thailand of 20,278 and 23,449 in India. 1.19 Private sector activity in rural finance is also rare, and usually limited to areas with commercial-agricultural activity. Despite the rapid growth and profitability of the private banks, they operate almost exclusively in urban markets. At end 1985, 99% of their loan portfolio was in short-term loans to predominantly large manufacturing, trade and service sector borrow- ers. In the same year, the share of private national banks and foreign banks in rural loans outstanding was only 1.3% (Annex 1, Table 13). Of all the RFIs, only 175 Bank Pasars are privately owned; the rest are owned by the National Government (BRI unit desas and pawnshops) provincial governments (BKKs, KURKs, etc.) and local governments (BKDs, BKPDs, LDs). All the KUDs and 19 Bank Pasars are owned by cooperatives. 1.20 Rural 'inancial Institutions. Indonesia has in place an impressive array of RFIs that are market-oriented, subject to few credit controls and interventions and serving a relatively large number of small borrowers (about 19 million loan approvals in 1985) (Annex 1, Table 15). The BRI unit desa system accounted for 46% of the total loans outstanding of the RFIs in June 1986, while Bank Pasars, the second largest lenders, accounted for 34%, about Rp 210 billion (Annex 1, Table 16). About 6.5% (Rp 40 billion) was lent to agriculture and rural borrowers by the pawnshops, 5.0% (Rp 31 billion) by BKKs, KURKs, etc.; and 4.5% (Rp 28 billion) by the KUDs. The most numerous village and paddy banks accounted for only 3.7% of the total. 1.21 The number of RFIs has not expanded significantly in the last six years and even declined in total number in 1986 due to the decrease in the number of unit desas and secondary banks (Annex 1, Table 17). However, the reduced number of unit desas reflect BRI's efforts at consolidating the unit desa financial position by phasing out the unprofitable units. In the case of the secondary banks (consisting of Bank Pasars, village and paddy banks), their total assets have increased in current and constant prices and in relation to total assets of financial institutions despite the decline in numbers (Annex 1, Table 18). Their loans outstanding have also grown from Rp 78 billion in 1983 to Rp 214.3 billion in 1985 (Annex 1, Table 13). The performance of the unit desas has been equally strong in recent years as loans outstanding climbed from Rp 89 billion in 1984 to Rp 285 billion in June 1986. The pawnshops had a less impressive growth; although total assets have - 13 - increased in absolute terms, their share of the total assets of all financial institutions declined from 0.22% in 1981 to 0.16% in 1985 (Annex 1, Table 18). The BKKs have expanded at a dramatic pace. The amount of their equity has more than quadrupled in the last five years from Rp 1.3 billion in 1981 to Rp 6.3 billion in 1986 (Annex 1, Table 19). 1.22 The BRI unit desa system and the BKKS, KURKs, etc. have multiple branches while the Bank Pasars, pawnshops, village and paddy banks, and KUDs have single unit outlets. Only the BRI unit desas and Bank Pasars have volun- tary savings programs. The village and paddy banks, BKK-type institutions, and KUDs have compulsory savings features in their lending programs while the pawnshops do not accept any deposits at all. The range of financial services available at or below the kecamatan level is thus limited compared to the full financial services offered by large financial institutions at the district, provincial and national level. 1.23 The loan portfolio of RFIs is predominantly short-term, nonagricul- tural and made up of trade and commerce credits (Annex 1, Table 15). Only the BRI unit desas offer investment loans of 3-year maturities and these comprise only 7% of their loan portfolio. The hierarchy of RFIs serve different parts of the credit market. The upper income segment is served by the BRI unit desas and Bank Pasars which require land, fixed assets and inventory as colla- teral and lend an average amount of Rp 460,000-Rp 735,000 per borrower. The BKK-type institutions and village and paddy banks serve the lower income groups of the rural population since they require no collateral for an average loan amount of Rp 30,000- Rp 57,000 per borrower and they are located at the village level nearest to the rural borrower. Women are over half of the bor- rowers of the Central Java BKKs and the KURKs of East Java. The government pawnshops, although located at the Uistrict and subdistrict levels, probably reach the largest number of urban and rural poor (in FY85/86 they processed 16.5 million loans) and women with loan amounts ranging from Rp 2,500 to Rp 300,000 against the security of household goods or valuables. The KUDs provide mostly subsidized credit to their members financed from GOI funds. The user costs of credit of smaller borrowers are very high (up to 194% p.a. from the village banks) compared to the rates charged by BRI unit desas (22-32% p.a.) and the Bank Pasars and pawnshops (42-60% p.a.) (Annex 1, Table 15). 1.24 Above the district level, the large financial institutions serve the upper-income segment of the total credit market lending to small-, medium- and large-scale enterprises as well as public and foreign enterprises. The branches of state banks (cabangs) dispense short-, medium- and long-term credit at interest rates of 12-33% p.a. fully secured by land or fixed assets. The cabangs are the main conduits of GOI sponsored programs but they also have a general ("umum") credit portfolio at market terms and conditions funded from the banks' own funds and deposits. Thus, the large private or public finan- cial institutions are serving the upper strata of the credit market, leaving the lower-income rural lenders and savers to small financial institutions. The BRI with its unit desas is the only large banking institution reaching rural areas but even this system focuses on the upper income groups of the rural population, because its credit program requires land as collateral. The numerous village banks, the BKK-type institutions, and the cooperatives are - 14 - left to deal with the rural poor, the less profitable and more risky segment of the credit market. D. Trends and Patterns in Agricultural Credit 1.25 Availability of Rural/Agricultural Credit. The amount of rural credit - in Indonesia is estimated at Rp 4,046.3 billion in June 1986, grow- ing at 8% p.a. in real terms from Rp 2,672.0 billion in 1983 (Annex 1, Table 13). Agricultural credit outstanding increased from Rp 526 billion in 1980 to Rp 2,071 billion in September 1986, representing a 27% nominal growth rate p.a. and a 15% growth rate p.a. at constant 1983 prices (Figure 1.5 and Annex 1, Table 20). Over the same period, there was major growth in credit to trade and services, 16.5% and 30.3% real growth p.a., respectively, while credit to industry grew at 10.8%. The share of agricultural credit to total credit rose from 6.7% in 1980 to 8.4% in 1986 but its share remains far below the 33% and 34% shares of the industry and trade sectors, respectively. The ratio of agricultural credit to agricultural GDP also climbed over the period 1980-86 from 4.6% to 7.3% but is again lower than the comparative ratios for industry at 64% and trade at 50% (Annex 1, Table 21). This low gearing ratio probably reflects a high level of self-finance and credit from informal sources and indicates a large potential for expanding institutional credit. 1.26 Allocation of Agricultural Credit by Commodity. Although agricul- tural lending could not be disaggregated by commedity, the evidence indicates a growing concentration of agricultural credit on the estate and tree crop subsector. For example, Bank Bumi Daya's estate agricultural creuit alone increased from Rp 208 billion in 1981 to Rp 912 billion in 1985, representing an increase in its share of total agricultural credit from 26% to 55% (Table 1.2). This figure understates the amount of credit going to tree crops; there are additional loans from BRI, BEII and the smallholder tree qyop loans financed by external donors and GOI which are still to be converted.' Tree crops liquidity credits from BI have also increased from Rp 134 billion in 1983 to Rp 513 billion in March 1986, accounting for an increasing propor- tion (19% to 43%) of total agricultural liquidity credits. The share of the food crop subsector in total agricultural credit could also not be segregated; the only identifiable BI liquidity credit channeled to food crops are to BIMAS Program which provided short-term credit for rice production and to a new credit program for paddy field establishment. These are estimated to have been Rp 161 billion in 1983 and Rp 53 billion in 1986. Since there have been no other major credit initiatives related to food crops, it is reasonable to assume that the share of food crops of total agricultural liquidity credits, and in aggregate agricultural lending, has declined steadily from 23% in 1983 to only 4% in 1986. 6/ This is estimated as the sum of the agricultural loans outstanding by commercial and development banks, the rural nonagricultural estate credit of BBD and rural loans outstanding of the RFIs. 7/ This is estimated at Rp 990 billion in 1985 (BRI-Rp 221 billion; BEII- Rp 353 billion; tree crop loans financed by external donors and COI- Rp 416 billion). OUTSTANDING BANK CREDITS BY SECTOR 4- 4 *. OTHERS AGRICULllJR 1 /~~~~~~~~~~~~1 - 16 - Table 1.2: ESTATE, TREE CROP AND FOOD CROP LENDING (Rp billion current prices) December March 1981 1982 1983 1984 1985 1986 A. Total agricultural credit 813 1,025 1,226 1,318 1,656 - B. Bank Bumi Daya (BBD) estate agricultural credit 208 487 516 662 912 - B/A (X) 26 48 42 50 55 - C. Total agricultural liquidity credit - - 710 901 995 1,191 D. Tree crops liquidity credit - - 134 225 324 513 D/C (X) - - 19 25 32 43 E. Food crop liquidity credit /a - - 161 159 131 53 E/C (%) - - 23 18 13 4 /a BI liquidity credit to BIMAS and paddy formation. 1.27 Regional Distribution of Total and Agricultural Credit. As of end- March 1986 the distribution of total loans outstanding for all banks shows that 50.3% of total loans outstanding was extended in Jakarta (Figure 1.6 and Table 1.3). Java and Bali, with 63% of Indonesia's population, account for about 76% of total banking credits. Table 1.3: REGIONAL DISTRIBUTION OF TOTAL AND AGRICULTURAL LENDING X Agricultural X Total lending Distribution lending Distribution (Rp billion) (Rp billion) Java 15,988.1 75.6 841.0 46.1 Jakarta 10,644.4 50.3 182.7 10.0 Rest of Java 5,343.7 25.3 658.3 36.1 Bali 180.0 0.9 16.0 0.9 Sumatra 2,403.5 11.4 753.2 41.3 Kalimantan 1,047.6 5.0 98.9 5.4 Sulawesi 572.9 2.6 38.9 2.1 Others 944.2 4.5 77.2 4.2 Total 21,136.3 100.0 1,825.2 100.0 REGIONAL DISTRIUUtON Total Credit LAfricultural Credit ¢mQtm~~~~~~~~~~~~~~~~~~~~~~** -~v a .-, a 14-__ Z.'JrnQtr~~~~~~~~~~~~~~~~~~ ii 4*~~~~~~~~~~~~~ ~~~eh 4 - 18 - 1.28 North Sumatra, with a large ari.a under tree crops, accounts for the highest share (36.9%) of agricultural lending among the different provinces (Annex 1, Table 22). East Java follows with 17.82, then Jakarta with 10X. All Java plus Bali accounted for 47Z of total agricultural lending followed by Sumatra (412) and Kalimantan (5.42) (Table 1.3). For all of Indonesia, the average agricultural loan outstanding per capita was Rp 11,052, ranging between provinces from Rp 164 in East Timor to Rp 71,160 in North Sumatra (Annex 1, Table 21). 1.29 Maturity Pattern of Agricultural Credit. Investment credit as a percentage of total agricultural loans outstanding has risen from 24% in 1980 to 53% in 1985, while the share of short-term credit has declined accordingly (Figure 1.7). The increase in investment credit may be attributed to the expansion in estate and tree crop lending, while the downward trend in short- term lending is primarily due to the demise of the BIMAS credit program. 1.30 Primary Agricultural Borrowers. Data on the allocation of agricul- tural credit by income group is not available but the evidence on agricultural borrowers suggest that estates and large borrowers are the main beneficiaries of agricultural credit. About 55X of total agricultural credit is extended to estates from BBD (Table 1.2). BEIt which accounts for 18% of total agricul- tural loans also lends mostly to large borrowers (Table 1.1). Even in the case of BRI which caters to smallholders, the number of borrowers in targeted programs has declined by 2.8 million over the period 1983-85 largely because of the decrease in the number of borrowers under the BIMAS program (Table 1.4). Table 1.4: BRI PROGRAMS FOR SMALL AGRICULTURAL BORROWERS (Rp billion) 1983 1985 Change in Loans No. of Loans No. of Loans No. of out- borrowers out- borrowers out- borrowers standing ('000) standing ('000) standing ('000) BIMAS 143 3,929 15 327 -128 -3,602 Kredit Mini/Midi and KUPEDES /a 25 198 229 1,034 204 836 KHKP 48 80 35 50 -13 -30 KIK/RCP/ IDA 480,400 106 93 107 79 1 -14 Total 322 4,300 386 1,490 64 -2,810 /a Figures in 1983 refer to Kredit Mini and Midi loans of the unit desas and to KUPEDES for 1985. Maturity Pattern of Agricultural Credit 40~~~~~~~~r4 "7 1/ /~192 9 Ia 18 tfflShj7rt-tormn YEA tn'vestnnt -~~~~~~~~/1 / /~~~~~~~~~~~~~~~~~~~9 o - /~~~~~~~~~~~~ I- /~~~~~~~~~~~~~~~~~~~~ // //~~~~~~~~~~~~ 4~~~ ~~/ - "' ~~~~~~~" - 20 - As of 1985 most loans outstanding and borrowers of BIMAS were in arrears and no program has replaced BIMAS to provide seasonal financing for food crop production. The volume of medium-term lending and number of customers for the Permanent Working Capital Credit Program (Kredit Modal Kerja Permanen--KMKP) have also decreased by 13% and 30% respectively. The number of borrowers under the Small Investment Credit Program (Kredit Investasi 0 ecil--KIK) and past Bank-assisted credit projects (RCP, IDA 480, IDA 400) 8'has similarly declined. The only growth area in credit for small borrowers is lending under KUPEDES. Of the total rural credit extended by various financial institu- tions, less than 3% is granted on an unsecured basis by the BKK type institu- tions, village and paddy banks and KUDs, which limits the access to credit of low income groups why do not possess land and property for collateral (Annex 1, Table 13).!' E. Sources of Funds for Agricultural/Rural Credit 1.31 BI Funding of Rural/Agricultural Credit. Agricultural and rural credit are financed primarily through BI liquidity credits and savings depos- its. The outstanding amount of BI liquidity credits for rural credit programs are estimated at Rp 1,909 billion in March 1986, a 31% increase in real terms from the 1983 level of Rp 1,456 billion (Annex 1, Table 23). On the other hand, BI liquidity credits to agriculture have grown from Rp 710 billion in 1983 to Rp 1,191 billion in 1986, an 11% real growth rate p.a. (Annex 1, Table 24). Although BI is funding a declining proportion (43% to 35%) of total outstanding credits for all banks, BI liquidity credits account for a high percentage (62-65%) of total agricultural lending, indicating heavy reliance of the agricultural sector on BI funds. 1.32 Rural Financial Savings. While financial savings cannot be accu- rately classified as "rural" or "urban", available evidence suggests that only a very small proportion of financial savings can be considered rural in ori- gin. For example, about 72% of total time deposits are located in Jakarta. Based on the deposits of the financja' institutions operating in rural Indonesia, rural financial savings - rose from about Rp 1,025 billion in 1983 to Rp 2,210 billion in 1985 (Annex 1, Table 25) due mainly to the increase in real deposit rates. BRI and the unit desas account for about 88% of total rural deposits and the group of secondary banks, KUDs, BKKs, KURKs, and saving and loan associations for the remaining 12%. Despite the doubling of rural deposits over a three year period, the share of total deposits has remained constant at 11%. If the deposits mobilized by the RFIs are used as an indicator of the rural savings mobilization performance, the amount, estimated at about Rp 338 billion in 1985, is only 2% of total bank 8/ Rural Credit Project (RCP-IDA 827-IND), Smallholder Private Estate Tea Project (IDA 400-IND) and Fisheries Credit (IDA 480-IND). 9/ In the 1982 Rural Credit Study, it was estimated that about 6.9 million households were landless. 10/ Rural financial savings is estimated as the sum of the bank deposits mobilized by BRI and all the RFIs. - 21 - deposits. The rural sector is still a relatively untapped reservoir of resources given the relatively high savings ratio of the Indonesian economy (0.25 in 1985), and the large number of the people deriving their livelihood from the rural economy. Although the ratio of quasi-money (M2-Ml) to GNP doubled from 0.7 in 1980 to 0.14 in 1985, this is still low compared with Malaysia at 0.48 and Thailand at 0.50 indicating a large potential for finan- cialization of savings (Annex 1, Table 26). Also, BRI marketing surveys suggest that hoarding in kind, gold and currency is still significant in rural areas and there is therefore scope for attracting physical asset savings into financial forms. 1.33 Adequacy of Funds in the Rural Credit Sector. If the volume of rural lending is coTY7red with the total resources available from BI, rural savings and equity,- the estimated surplus of funds for rural credit (Rp 495 billion in 1985) indicates a possible backflow of resources into urban areas (Annex 1, Table 27). A part of either deposits or liquidity credits are being channeled to uses other than rural credit. An implication of the resource surplus is that BI liquidity credits may have substituted partly for internally generated funds (savings and equity) and rural credit may not be drastically reduced with the gradual withdrawal of BI/GOI support. Although effective credit demand in the rural areas may be weak and funds cannot be totally absorbed in the rural economy, it is also likely that funds are leaving the agricultural and rural sectors for areas of higher returns, lower costs and less risks. At present, low and fixed lending rates in most agriculture and rural credit programs, although compensated by lower cost of funds, do not provide sufficient incentives for the financial institutions to lend to agriculture and small rural enterprises. Given Indonesia's open capital account and fungibility of financial resources, the banks are able to invest their funds abroad at rates equivalent to the nominal rate in foreign money markets (6-8.4%) plus an expected rate of depreciation of the Rupiah (e.g. 10%). Thus, in June 1986, the national foreign exchange banks which include the five state commercial banks and 10 private commercial banks with foreign exchange license had a net foreign exchange position of Rp 2,778 billion which was roughly 44% of their borrowings from BI (Annex 1, Table 28). The returns to foreign fund placements are attractive if sourced from BI liquidity credits at 3%. However, the foreign asset holdings of banks would likely be lower if cheap BI liquidity credits were not available since it would not be profitable to invest abroad funds mobilized from savings and time deposits at 12-18%. 1.34 Against the background of these trends and patterns in rural/agri- cultural credit and financial resource mobilization, the next chapter will examine the performance of key institutions providing credit to the rural sector. ll/ Estimated from total equity of RFIs. - 22 - II. PERFORMANCE OF SELECTED RURAL CREDIT INSTITUTIONS 2.1 This analysis of selected institutions covers BRI and BBD which are the primary rural lenders in Indonesia and the RFIs which have the greatest presence in the rural areas and aims to identi'y the factors bearing upon their effectiveness and efficiency. A brief description of these institutions as well as the summary features of a number of informal credit arrangements is found in Annex 6. A. Effectiveness and Efficiency of Selected Credit Institutions Bank Rakyat Indonesia and Bank Bumi Daya 2.2 Scope of Operations. At end-1985, BRI and BBD held assets of Rp 5,652 billion and Rp 5,802 billion, respectively (Annex 2, Table 1). However, BRI is a much larger organization with 31,330 employees (including 13,000 in the unit desa system), compared with 7,634 at BBD. BRI has a net- work of 326 branches and subbranches, while BBD has only 128 (30 in Jakarta). Furthermore, BRI operates 2,272 unit desas throughout Indonesia and has, in addition been given the responsibility by GOI and BI, to supervise the the secondary banks and the KUDs. 2.3 As a mass supplier of retail credit (about 2.4 million loan accounts at end 1985) and a collector of small savings, BRI is reaching a much larger number of borrowers than BBD, whose loans numbered only 32,000 in 1985, of which 405 were to large public enterprises. Only 2.5% of the value of BBD's loans outstanding went to small, and mostly urban, borrowers (59X of the total number of loan accounts), and about 35% of the value of loans outstanding was for very large credits (2.5% of the total accounts). About 24% of BBD's loan portfolio is for agricultural lending and most of it (90%) goes to large estates (Annex 2, Table 2). The ratio of the number of loan accounts per employee for BRI is 77 while BBD's is 4.2. Although BBD's mandate is to provide credit for estate agriculture and forestry, it could play a greater role in the rural financial market by expanding its network in the rural areas, lending to smallholders and mobilizing rural savings. 2.4 Organizational Structure. Both banks are organized in the tradi- tional British/Dutch pattern of a managing or executive directorate respon- sible for specific areas of administration to a board of supervisors. The organization of divisions is along functional areas: auditing and control, credit, international transactions, accounting, banking services, and adminis- tration. Reflecting their emphasis on lending rather than resource mobiliza- tion and their ready access to concessionary BI/GOI funds, both banks lack sections responsible for marketing bank services, particularly for the design, promotion and marketing of savings instruments. Collections and the adminis- tration of arrears is a function of the credit divisions, but since most loans are to public enterprises and/or guaranteed fully by GOI, credit collection policies and procedures have not received priority attention. 2.5 Loan Portfolio and Arrears. At end-1985, the proportion of loans outstanding to total assets of BRI was 74.3% and 67.8% for BBD, declining from 82% and 70% respectively in 1982 (Annex 2, Table 3). BRI's leverage (total - 23 - assets/net worth) of 48 times is very high and more than double that of BBD's. Provisions for bad debts at BRI are 3.5% and arrears represent 10.8% of loans outstanding. At BBD, the provisions are 5.0% and arrears are 19.3% of the loans outstanding. In both cases, the provisions are low which may be due to the fact that most BRI Loans are insured by ASKRINDO and PERUM PKK while many of the borrowers in arrears at BBD are large public enterprises fully suppor- ted by the GOI. As mentioned before, about 90% of BBD's agricultural credit goes to public sector estates that have arrears representing 25% of their loans outstanding. 2.6 Sources of Funds. BRI is more dependent on government funds than BBD. While BRI held credit from GOI and El amounting to 54% of its assets, BBD's borrowings were 35.5% (Annex 2, Table 3). BRI's deposits in 1985 were only 34% of assets while BBD's were 42% although this has increased from 29% (BRI) and 38% (BBD) in 1982. The capitalization of both banks is low repre- senting only about 2.1% of total assets for BRI and 4.6% for BBD (Annex 2, Table 1). The share of capital to total assets has actually declined from 5.7% for BBD and 2.4% for BRI in 1982. 2.7 Efficiency Measures. There are few meaningful measures of the efficiency of the state commercial banks who are expected to perform a dual function of being development agencies and profit making institutions. GOI has supported BRI and BBD in all aspects of their operations. On the resource side, the Government has provided the banks with equity, deposits, liquidity credit and guarantees for external borrowings at low costs. On the lending side, credit insurance has been provided to minimize the risks of financing target groups. The banks have access to guaranteed sources of income which enhances their financial viability. BRI's profit, for example, is partly attributed to the financing of the Board of Logistic Affairs (BULOG) working capital credits which comprise 38% of its loan portfolio and yield a fixed and guaranteed interest income for BRI. 2.8 The ratio of interest income to average loans outstanding of BRI (13.0%) and BBD (10.1%) is low reflecting their arrears problem and low interest rate ceilings on a major part of their loan portfolio. The share of interest costs to average loans outstanding are 8.4% (BBD) and 7.3% (BRI) in 1985 rising from 7.6% (BBD) and 3.4% (BRI) in 1982 as a result of the increase in the share of deposits in their total assets (Annex 2, Table 3). Personnel expenses of BRI as a percentage of average loans outstanding (3.4%) are more than twice those of BBD (1.3%) in 1985, although this ratio has declined from 4.0% (BRI) and 1.6% (BBD) in December 1982. Othet administrative costs of BRI (2.9% of average loans outstanding) were also higher than those of BBD (1.8%) reflecting the higher costs of reaching a larger number of smaller borrowers. In spite of the higher operational costs of BRI, its gross profit (0.8% of average assets) is higher than that of BBD (0.4% of average assets) because of a larger share of lower cost funds from the government. The difference in financial performance of the two institutions is shown in the net profit after taxes (0.43% of average assets for BRI and 0.26% for BBD), which is very low in both cases. - 24 - Selected Rural Financial Institutions 2.9 The RFIs that have been reviewed include the following: (a) BRI unit desa system; (b) secondary banks, such as Bank Pasars and village banks (BKDs); (c) pawnshops; (d) nonbank rural financial institutions, such as the BKKs, LPN, and KURKs; and (e) village cooperatives (KUDs). In addition Bank Dagang Bali (BDB) has been examined as an example of a privately owned commercial bank. 2.10 Costs of Intermediation. Operating cost as a proportion of loans outstanding has been used as an indicator of the efficiency of intermediation. In most of the smaller institutions these costs are very high. The lowest cost of intermediation is that of the Bank Pasars which operate near city mar- kets, or at the district and subdistrict level. They are probably che most efficient secondary banks, and may be used as a standard for comparison. The operating cost of the Bank Pasar as a percentage of outstanding loans is about 3-3.6% per year (Annex 2, Table 5), which compares well with the cost of intermediation of Bank Bumi Daya at 3.1% in 1985. In comparison to this stan- dard, other rural credit institutions have a very high cost of intermediation, with the exception of BDB (4.8% of outstanding loans). Their costs of opera- tion range from 15% of loans outstanding for KURKs, to 23% for 11 model KUDs operating with Dutch technical assistance, and 34% for the government-owned pawnshops that require costly storage facilities. The Central Java BKKs have a 10% ratio which is a realistic standard for small scale operations serving a high cost and risky clientele. The BKDs which lack the technical assistance of BKK, but operate in similar conditions of risk and high unit cost of lending, have operating costs that are 14% of loans outstanding. BRI unit desas, at an early development stage with excess capacity, hase operating costs of about 16.2% of loans outstanding, but are expected to lower costs to about 10%, the level of BKKs. 2.11 Interest Income. Interest income as a percentage of loans outstand- ing varies widely for the sample of institutions analyzed. It ranges from 21% for the KUDS and 23% for the Bank Pasar, to 42% for the pawnshops. The inter- est earnings of BRI unit desa were 28% of loans outstanding in 1985. The BKK- type institutions have interest income in the neighborhood of 26% of loans outstanding per year, while the BKDs show higher interest income (34%). These accrued levels of interest are much lower than the stated effective rates of interest that most institutions claim to charge. The difference between the claimed rate and the accounting interest income may be related to the high arrears of most of these institutions. BDB, with low arrears, shows interest income of 33.5% of loans outstanding which is near the levels of its effective interest rates. 2.12 Arrears. The lowest observed level of arrears, as a percentage of loans outstanding, was 0.4% for BDB. Their lending practices are based on first-hand knowledge of prospective borrowers, their assessment of the borrow- ers' ability to pay, and collateral. Other institutions mostly rely on village authorities for the assessment of credit risks although some approved all credit applications. The BRI unit desa system which has 4.3% of its loans outstanding in arrears requires that the prospective borrower supply collateral in addition to a character reference from local authorities. The - 25 - arrears of the pawnshops are 0X according to official records but staff of a shop visited claimed that 20Z of the merchandise is not reclaimed and has to be sold at auction. The Bank Pasars that operate in the city or farm markets, and give commercial loans to traders, have moderate arrear levels of 5-7Z of loans outstanding. Noncollateral systems, such as the BKDs and BKK-type oper- ations, have arrears ranging from 16-20% of outstanding loans. RUDs, which are subject to local pressures and less able to collect, have arrears of 37% of loans outstAnding. 2.13 Resource Mobilization. RFIs have not been generally effective in savings mobilization and depend on endowments and funds from BRI or from the regional development banks. The exceptions are BDB with a 9.6 ratio of sav- ings to net worth, and the Bank Pasars (7-9 ratio). The BRI unit desa system is increasing its saving effort through the Village Savings Program (Simpanan Pedesan-SIMPEDES) mechanism (para. 4.4) and can be expected to increase its savings to net worth ratio, which is moderately low at the present time. The BKK-type institutions and BKDs rely on compulsory savings schemes and have not tapped the voluntary resource mobilization potential indicated by the relative success of the SIMPEDES program, and by the performance of BDB and the Bank Pasars. 2.14 Sustainability of Rural Financial Institutions. The BRI unit desas, BKKs, Bank Pasars and BDB seem to be well managed, financially viable institutions capable of long-term growth and sustainable operations. BDB and the Bank Pasar set a satisfactory standard for a well run provincial credit institution which should be expected to yield a 20-25% rate of return on net worth per year. Other RFIs need to reduce their arrears and operating costs and increase their size and interest income in order to survive indepen- dently. For example, the BKDs have high operating costs that are directly related to the small size of their transactions and the scale of their operations. At the same time, their costs are increased by a high level of arrears. In some cases, this leads to negative operating margins which could prevent the growth or development of these institutions. They can survive from income derived from the investment of their capital endowments, or by ignoring the cost of bad debts, and allowing arrears to go on indefinitely without writing off the unrecoverable debt. However, this situation is not conducive to efficiency or the future development of these institutions. To play an active role in the financial markets, they must expand their scale of operations, diversify risks, mobilize voluntary savings, and reduce their operating costs. 2.15 The KUDs are multi-purpose organizations that have taken on more financial responsibilities than they can handle. Their credit programs are mostly financed by BRI loans which in turn are funded from GOI sources (e.g. BI credit for 100% of food procurement loans to KUDs) and insured by PKK for about 90% of their value. Their savings and loan program constitutes less than 10% of their financial resources and does not show a significant saving mobilization effort. Most of the credit programs carried out by the KUDs are conceived at the national government level and imposed on the KUDs. Thus, arrears are high: 22% of BRI loans to KUDs and 30% of their small borrowers programs (Kredit Candak Kulak-KCK). Of the total number of KUDs, only 922 (14%) are classified under Class "A' which are well run, self-sufficient and - 26 - financially sound. Neither a bank nor a non-bank financial institution, the KUDs have not been adequately prepared for performing a financial intermedia- tion role and would not be able to sustain their credit operations without continuous GOI support. B. Factors Influencing the Performance of Credit Institutions 2.16 Availability of BI Liquidity Credits. BI has played a central role in the provision of credit to the rural sector in Indonesia. Instead of being a "lender of last resort" as in other countries, BI is a development bank that contributes significantly to the credit market, and directs the allocation of credit in Indonesia through its own credit decisions. However, the provision of BI resources to the state commercial banks at low interest rates has reduced the incentives for financial institutions to mobilize savings deposits and expand their capital resources through retained earnings or profitable operations. With cheap and accessible resources from Government and external donors, the banks have been unable to build an adequate and self-sustaining resource base (consisting mainly of deposits and equity) which is necessary for the long-term development and sustainability of a financial institution. A self-reliant rural financial system capable of mobilizing domestic resour- ces, operating efficiently and building up internal capital is essential especially in the light of tight budgetary constraints of GOI and the worsen- ing external environment. 2.17 Administered Interest Rates. An important factor that impacts directly on the performance of credit institutions is the interest rate ceiling imposed on the program credits channeled through these institutions. Although compensated by lower cost of funds, the administered rates do not adequately cover the risks and administrative costs of delivering the credit. Thus, the financial standing of BRI and BBD have been endangered since these program credits form a major part of their loan portfolio. In 1985 their ratio of interest income to average loans outstanding (13.0X for BRI and 10.1 for BBD in 1985) failed to cover interest, personnel and other operational expenses (13.6% of average loans outstanding for BRI and 11.5X for BBD). If realistic write-offs and provisions for bad debts were undertaken, the negative operating margins of BRI and BBD would increase even further. Their profits in 1985 (0.43Z of net wcrth for BRI and 0.26Z for BBD) are largely due to cross-subsidies from other areas of operations (foreign exchange operations, general lending, BULOG working capital credits, KUPEDES, etc.). Besides their negative impact on the viability of financial insti- tutions, "below market" interest rates have also discouraged the banks from penetrating the rural credit market and contributed to the poor representation of banks in the rural areas. 2.18 Branching Regulations. The commercial banks are presently con- strained in expanding their branch network. To expand in Java or Bali, a private bank must first open another branch outside these islands. The priv- ate banks are reluctant to expand into the Other Islands as the level of economic activity may not warrant a full branch. Since a full branch may be uneconomic and GOI regulations which do not distinguish between full branches, subbranches, mobile units or private agency are stringent, there are no incentives for private banks to branch out and penetrate the rural areas. The - 27 - state banks are not similarly hampered as a favorable feasibility study of potential deposits, lending and profitability is the only requirement of MOF. However, with the ready access of the state banks to cheap GOI resour- ces, and lucrative channels of lending (large scale enterprises, foreign credit market, etc.) in the urban areas, they have little interest in expand- ing into unfamiliar and riskier areas of agriculture and rural lending. 2.19 Economies of Scale and Homogeneity of Risks. Some RFIs are too small and their risks too concentrated to be able to offset the high trans- actions costs and lending risks they must bear. One BKD office visited by the mission had only eight active borrowers who were paying effective interest rates of 194% p.a. to sustain the operations of the institution and pay the salaries of the part-time BKD staff. Many of the RFIs operate exclusively in one village, or lend only to the rural poor. Thus, their loan portfolio is skewed to one type of risk: the risk of a small farmer or trader whose income comes from a high risk activity. Moreover, if the borrowers are farmers from one area, or if they trade farmer's products, the same adverse conditions may affect them simultaneously. A village bank with high transaction costs, prevented from branching out and diversifying their risks, or from compen- sating the high costs of lending with more profitable activities, may be a practical impossibility in the absence of concessional support by a higher entity. Isolated village banks, or even a system of BKKs operating exclu- sively in homogenous credit markets, would need access to the different levels of the credit market and to a greater variety of borrowers, in order to balance the risk of the rural poor with other lower-cost and lower-risk lending, or investing. 2.20 Existence of Technical Assistance and Sponsoring Organizations. Based on the experience of the unit desas and the BKKs, their successful per- formance seems to be heavily influenced by an entrepreneur or sponsoring orga- nization who provides management, supervision, control and the initial thrust and commitment to success. This sponsor in the case of BKK operations is the provincial Government through the BPD and BRI. for the unit desas. The suc- cess of the units depends on the strength of the sponsoring organization; its ability to offer lower-level customers the opportunity to progress from small loans to medium- and even large-scale business; and the quality and intensity of technical assistance available for the units, such as the USAID assistance to BKKs and BRI for the unit desas. For the village banks and KUDs, there is no sponsoring organization and little technical assistance except for two limited Dutch government-sponsored programs of 12 KUDs under the National Cooperative Bank (BUKOPIN). A sponsoring organization is especially relevant for the secondary banks, since BRI, the supervisory body appointed by BI, has no inherent interest in fostering the development of its natural competitors. 2.21 Assigned Role of the Rural Financial Sector. The development of the rural financial system also depends on the objectives assigned to it by GOI. Currently, the rural credit sector is considered only an instrument for attaining higher ends, not a desirable development objective in itself. Its part in the services sector contributing to GNP and employment not unlike the agriculture, industry or trade sectors has not yet been recognized. The development of the financial system is perceived widely to mean the growth of profits for the institutions, not a broader set of development results which - 28 - would include the improvement in the yield and productivity of capital, the increase in the overall level of savings and investment, and broader access to credit. The recognition of the appropriate role of credit and the financial sector in national development is necessary to any effort to develop the RFIs and to strengthen their ability to serve the financial needs of the agricul- tural and rural sectors. - 29 - III. ASSESSMENT OF SELECTED CREDIT PROGRAMS 3.1 More than 75% of total rural and agricultural loans outstanding is provided under numerous credit programs funded by BI and GOI. Most of them are commodity specific and does not take into account the multi-enterprise nature of the farm household. The plethora of credit schemes impedes the efficiency of financial institutions and increases the administrative costs of the programs. In addition, they increase the transactions costs of borrowers as the farmer has to take out three separate loans if he wants to finance his rice, soybean, and livestock enterprises. These multiple credit schemes need to be consolidated, possibly under a single loan fund, in order to rationalize their terms and conditions and improve the operational efficiency of financial institutions. Although credit may still be required within a project framework to provide an integrated package of inputs to borrowers, the future credit delivery system should be geared towards market demand, provide credit on a farm system basis rather than to specific commodities only, and minimize the distortionary impact of subsidies on the rural credit sector. 3.2 This chapter analyses the selected credit programs to determine the factors affecting their success or failure and to derive lessons from success- ful programs. The programs examined are as follows: (a) World Bank-assisted tree crop credit schemes; (b) Small Investment Credit/Permanent Working Capital Credit Programs (KIK/KMKP); (c) General Village Credit Program (RUPEDES); and (d) the lending programs of the Subdistrict Credit Units (BKKs) in Central Java. Annex 7 contains a detailed description of these programs as background to this discussion. A. Main Features of Selected Credit Programs 3.3 Bank-Assisted Tree Crop Credit Schemes. GOI with the assistance of external donors has been engaged in a large investment and production program in the tree crop sector, through both publicly owned estates (PTPs) and the development of the smallholder sector. PTPs or project management units (PMUs) have been used to develop tree crop stands (rubber, oil palm, coconut, etc.) for smallholders, landless and low income farmers in project areas, and transmigrants. The credit system is primarily a cost recovery mechanism from the loan repayments of beneficiaries after the expenditures by PTPs and PMUs are converted into loans to individual smallholders. The banks, mainly BRI, act as a channeling bank under the Nucleus Estate and Smallholder (NES) projects or as an executing bank for the PMU projects, such as Smallholder Rubber Development Projects I and II (SRDP I and II), and Smallholder Coconut Development Project (SCDP). The two credit arrangements imply different functions for the banks. Whereas in executing loans, banks act as a financial intermediary; in channeling loans, they act as an agency and collect a fee for handling the accounts. However, practice, the distinction becomes blurred since in both cases the borrowers are selected by the PTPs and PMUs and the loans are risk free investments to the banks as they are fully guaranteed by GOI. The credit terms and conditions differ among projects based on the commodity financed and the estimated repayment capacity of borrowers. They range from 0-12% interest rate per year with varying levels of subsidy from Government and 10-20 year maturity with 3-8 years grace period. Although - 30 - aware of the potentially distortionary effects of credit subsidies, GOI has considered the development of tree crops for export expansion and improvement of smallholder incomes as an overriding objective. 3.4 Since 1978, World Bank loans to 11 tree crop projects amounted to US$828 million, about 73% of the total external assistance for tree crops (Annex 3, Table 1). Total project costs are estimated at about US$1.6 billion and the smallholder credit component at about US$1.0 billion. Of the total smallholder credit component, about Rp 173 billion of expenditures have been incurred and only about 0.1% or 611 loans worth Rp 2.5 billion under the NES scheme and 94 loans worth Rp 52 million under SCDP have been converted. A number of problems is encountered in loan conversion related to the procedural requirements for debt conversion, quality of the tree crop stands, smallholder credit amounts, and debt repayment capacity of the smallholders. A fundamental weakness in the credit program for NES projects is the lack of incentives for the participating parties to convert the loan. Smallholders will not enter into debt unless obliged and since their savings deposits ea!n 15% a year, they have no incentive to pay off their loans at 10.5-12% p.a. PTPs have nothing to gain from the process, because their costs have been covered by GOI and the conversion process brings out inadequacies and high cost of their operations. BRI does not have any funds at risk since it is merely a channeling bank and is therefore indifferent to the outcome. GOI is the only interested party as it has to recover about Rp 416 billion of government funds which have been released so far. 3.5 The issues affecting tree crop credit relate not only to loan conversion problems but also to financing Government predetermined, economically desirable activities that are not financially viable at market interest rates. The bank,, are reluctant to provide term credit for these activities from their own resources given the shortening maturities of their savings deposits, the higher risks and costs of term lending, and fixed interest rates which provide inadequate protection from interest rate fluctuations over time. In addition, interest rates on investment credit have tended to be lower than on short-term loans reducing further the incentives to provide term finance. The Government's response to this problem has been to provide low cost funds for institutions to lend to priority sectors and to lower the onlending rates for borrowers to make their enterprise bankable. Thus, term credit for tree crop projects has been primarily provided through state banks under BI and GOI targeted and subsidized credit programs. How- ever, this strategy has had distortionary effects on the financial system as subsidized cost of funds reduces incentives for deposit mobilization, public estates and large borrowers become the main beneficiaries, and the financial position of banks are undermined. The main issue is therefore how to maintain the supply of term credit outside of subsidized liquidity credit-financed priority programs in line with investment demand in the future. To induce banks to increase term lending, long-term lending rates may need to rise and vary over the loan maturity period to reflect the banks' costs -nd risks of lending. Sources of long-term funds e.g. pension and superannuation funds, bonds, life insurance policies, etc. will have to be tapped to improve the deposit mix of banks and lengthen the maturities of available financial resources. - 31 - 3.6 KIK/KMKP. The Bank has also been deeply involved with KIK and KMKP through three Small Enterprise Development Projects (SEDP). KIK/KMKP was set up in 1974 to provide term finance for indigenous, small-scale, labor- intensive enterprises unable to raise their own funds. It is implemented through the state commercial banks, Development Bank of Indonesia (BAPINDO), and selected BPDs and private commercial banks. The normal terms are 8 years for KIK loans and 5 years for KMKP loans at 12% p.a. with grace periods of 4 years (KIK) and 1 year (KMKP), a loan ceiling of Rp 15 million, and collateral requirement of about 150% of the loan amount. The Government pays to the handling banks a 1.5% p.a. direct subsidy on KIK/KMKP loans under SEDP III. The funds for KIK/KMKP programs come from BI liquidity credits (55%), World Bank (25%) and participating banks (20%). Through ASKRINDO, GOI insures 75% of total credit risks of the banks. The Government also bears the interest rate and foreign exchange risks on World Bank funds made available to handling banks. The growth of lending under KIK/KMKP was spectacular in the early years, but the amount of loans approved has declined by 12%-35% in the years 1982-84 (Annex 3 Table 2). Loans outstanding as of December 1985 stood at Rp 338 billion for KIK and Rp 885 billion for KMKP. 3.7 KUPEDES. KUPEDES was introduced in February 1984 to replace the faltering small borrower (Kredit Mini) and small credit (Kredit Midi) programs of the BRI unit desas. These programs were designed to provide term finance to small borrowers who could not qualify for KIK/KMKP. In contrast with other special credit programs, KUPEDES is a general, noncommodity-specific and market-oriented, rural credit program, charging effective interest rates of 22-32% p.a. with maturities ranging from 3 months to 3 years. The loan limits per borrower are Rp 25,000 minimum and Rp 2 million maximum with collateral requirement of at loast 100% of the loan amount in the form of any property. The growth of the KUPEDES program has been remarkable; the number of loans increased 55% from 640,000 in 1984 to about 992,000 in 1985, and the value of loans almost doubled from Rp 171 billion to Rp 339 billion (Annex 3, Table 3). As of October 1986, total cumulative loans amounted to Rp 901 billion aud loans outstanding were about Rp 313 billion. 3.8 As of June 30, 1986, working capital loans accounted for 93% of loans outstanding. By maturity, 12 month working capital loans accounted for 57% of the total number of loans and 45% of the value. The purpose listel4for 69% of the loans outstanding was trading, while 27% were for agriculture.-/ For the same period, 75% of the Loans outstanding were in Java, where 64% of the BRI unit desas are located.- Approximately 25% of all borrowers and 75% of cosigners are women. The sources of funds for KUPEDES loans come from BI liquidity credits (53%), savings and other deposits (28%) and a grant of Rp 66.7 billion (19%). 1/ Since credit is fungible, the listed purpose for a loan is mainly to determine ability to pay. 2/ BRI has Unit Desas in 14 of the 15 provinces in which it has branches. - 32 - 3.9 BKK Lending Programs. The BKK ir.stitutions were initiated in 1972 by the Central Java Provincial Government to provide capital to the rural poor to help them raise their incomes and standard of living, to protect them from usurious moneylenders, to educate them on the benefits of financial savings, and to create employment. Currently, there are 497 BKKs covering all kecamatans and 2,609 village posts, where the BKK employees (three young school leavers per BKK) lend and collect loan repayments, usually on market days. BKKs have six different types of loans based on the maturity period, from 22 days to six months, with nominal interest rates of 2Z-4.8X per month on the original loan amounts (Annex 3, Table 4). With a commitment fee of 1%, the effective monthly interest rates are estimated at 2.2%-10.8%. The loans contain a forced savings feature (6.5% to 20% depending upon the type of the loan) but no collateral is required. Loans under Rp 25,000 can be approved by the BKK staff and over this amount, by the kecamatan head. BKK loans out- standing have grown at a tremendous pace from Rp 213 million in 1972 to about Rp 12 billion in 1985 (Annex 3, Table 5). B. Measures of Program Success 3.10 The objectives of the selected credit schemes encompass a broad spectrum of aspirations including: (a) creation of employment opportunities; (b) expansion of foreign exchange earnings from tree crops; (c) raising the income and welfare of smallholders and the poor; (d) geographic dispersion of productive investment and development of entrepreneurial opportunities and technical skills; (e) provision of markets for local small-scale enterprise products and raw materials; (f) protection from usurious practices; and (g) credit education. KUPEDES has the objective of developing the unit desas into viable entities capable of providing efficient financial services (savings instruments and loans) in the rural areas. qowever, the direct causal rela- tionship between credit and these development objectives is not obvious because of problems such as fungibility, attribution, and the interdependence of farm and household decision-making. Given the fungibility in cash flow management within farm households (i.e., the ability to shift money from one purpose to another), it is difficult to identify the effects of the loans on the farm versus the household. Moreover, attributing the observed outcome to credit does not take into account other factorL such as technology, prices, infrastructure, and extension. Available field surveys and studies on the impact of credit also has theoretical shortcomings in the methods employed to measure the outcome of credit. Thus the most meaningful performance indica- tors of special credit programs are the access of borrowers to the program, loan repayment rate, resource base, and efficiency or cost of intermedia- tion. 3.11 Access to Credit. Since their inception, the cumulative number of loans/participants under the selected credit schemes have totalled about 9.1 million (Annex 3, Table 6). At end June 1986, about 105,000 (53%) of the total planned number of participating smallholders have been recruited under the Bank-assisted tree crop projects and 805 (0.7%) of the recruited small- holders have converted loans. The participants consist mainly of settlers, transmigrants and smallholders but an indeterminate percentage of ineligible participants such as government civil servants, ex-military personnel, etc., (40% in NES I Tebenan site) is also included. For the KIK/KMKP programs, the - 33 - number of borrowers has declined from a peak of 312,000 in 1981 to 96,000 in 1985 (Annex 3, Table 2). The bulk of KIK/KMKP lending is in large loans (Rp 5.2-7.7 million per borrower) to existing entrepreneurs with proven expe- rience and maximum net worth of Rp 40 million. As a result of the low and administered lending rates, small borrowers have been screened out by banks participating in KIK/KMKP who are ug)ble to cover their lending costs by altering the fixed price of credit." KIK/KMKP loans can only be disbursed by full bank branches generally located in larger towns at the district level which means that they have not reached large portions of the rural areas. KUPEDES is reaching the highest number of rural borrowers with 992,000 loans in 1985 and 2,272 unit desas spread throughout Indonesia at the subdistrict level. Given an average loan size of Rp 460,000 and its collateral require- ment, KUPEDES tends to serve the upper income segment of the rural population who has land and property. The rural poor, particularly women, (at least of Central Java) are being served by the BKKs given the small loan per borrower (Rp 57,000), the absence of a collateral requirement, and the location of the BKK outlets at the village level. In 1985, the BKKs extended 534,000 loans and since 1972 to June 1986, about 4.7 million loans. Based on the extensive outreach of the KUPEDES and BKK lending programs and the rapid growth of the amount of unsubsidized loans under these programs, the availability of credit seems to have been more important to the "economically weak" entrepreneurs and farmers than below-market interest rates. 3.12 Loan Recovery Rates. The loan recovery performance of credit pro- grams is critical in maintaining the resource base of these prorrams. If initial funds are not recovered and recycled into the next lerb s phase, resources are steadily eroded endangering the sustainability of the credit program. The repayment performance of KIK/KMKP has been poor compared to the fairly high levels of loan recovery in the KUPEDES and BKK programs. The tree crop smallholder loans which have been converted recently are under grace periods and thus the recovery performance for these programs cannot be assessed. KIK arrears amount to Rp 99 billion in January 1985, about 30% of total loans outstanding and more than 90% of these arrears are one year old and over (Annex 3, Table 7). Data on KKKP loan arrears are limited and incon- sistent but BRI, the largest lender in the program, showed an average arrears rate of 30% for KMKP as of end August 1986, affecting about 60% of the KMKP loan accounts (Annex 3, Table 8). 3.13 In SEDP supervision reports, KIK/KMKP loan arrears have been attributed to low lending rates, liberal credit insurance policy, and borrowers' perception of loan funds as government grants. The 12% interest rate charged on KIK/KMKP loans which is below the deposit rate of 15-18% discourages repayments by borrowers in favor of alternative uses and provides little incentive to the handling banks to undertake detailed appraisal and supervision of these projects. Moreover, generous ASKRINDO coverage provided to the handling banks causes the banks to accept borrowers without adequate 3/ This conforms to the "iron law of interest rate restrictions" which predicts that loan portfolio concentration increases under interest rate restrictions. - 34 - review of their credit history or proposed projects and to rely on ASKRINDO's guarantee rather than on their own collection efforts. The premiums charged by ASKRINDO are very low compared to the guarantee coverage provided and do not allow ASKRINDO to be commercially viable. Total claims have increased over the years (Annex 3, Table 9) and have led to financial losses totaling Rp 169 billion in the period 1983-85. Given the mounting arrears despite the demonstratC4 positive impact of KIK/KMKP lending on sales/production of borrowers,_ willful default could be a major cause for arrears stemming from the perception that KIK/KMKP loans are grants from the Government. 3.14 In contrast, the KUPEDES program arrears ratio (total amount overdue as a percentage of total loans outstanding) stood at about 4.3% as of October 1986 (Annex 3, Table 10). The long-term bad debt ratio (total overdue pay- ments to total payments due) and short-term bad debt ratio (payments missed in a month to payments due in that month) were also small at 2.2% and 2.9%, respectively, although they have risen from 1.0% and 1.4% in December 1984. The successful repayment performance of KUPEDES is attributed to market interest rates, an Incentive for Prompt Payment ("Insektif Pembayaran Tepat Waktu"--IPTW) of 0.5% per month collected with the monthly payment and returned if installments are in full and on time, financial incentives to the unit desa staff for profitable unit desa operations, and the absence of risk sharing arrangement with GOI and loan insurance cover. Prompt attention to the emerging arrears problem is also important. With a review showing that over 50% of all arrears are concentrated in 22 branches, special BRI teams have been formed from both head office and regional staff to investigate the reasons for the arrears and to recommend measures to correct the problem. 3.15 The loan recovery rates of the BKK lending programs have also been satisfactory taking into account its poor clientele and absence of collateral. Although the arrears ratio (overdue payments to total loans outstanding) was about 16% at end 1985, both the ratios of the total amount of overdue payments to the total payments due and the overdue payments in a year to the repayments due in that year have declined from 12% in 1972 to about 2Z in 1985 (Annex 3, Table 11). The arrears stem from the early years of the program and consist primarily of overdue repayments in the BKK Khusus program under which the BKKs acted as collection agents for the special technical service agencies of the Provincial Government. The factors that seem to account for the successful recovery of BKK loans are interest rates that adequately cover costs, character-based lending and repeater loans, staff bonuses based on profit sharing, and the 100% accountability of the financial institution of the lend- ing risks. The BKK system does not have a write-off policy yet, but the mana- gers estimate that 40% of the overdue repayments are unrecoverable and are taking steps to write off these bad debts. 3.16 Resource Base of Credit Programs. An adequate resource base of internally generated funds from earnings and loan repayments and savings mobi- lized in the area determines the life of a program. External sources such as 1/ An evaluation study on KIK/KMKP indicated increases of sales and/or production for borrowers. - 35 - the Government or foreign donors may prolong the life of a credit program by pumping more resources into it, but are subject to resource constraints and alternative rates of return in their investment possibility frontier. An indicator of the adequacy and stability of the resource base of a program is the percentage share of equity and savings deposits to total resources. Among the programs reviewed, the BKK programs are the most self-reliant with 78% of funds generated from retained earnings (60X), savings (141) and equity (4X) (Annex 3, Table 12). KUPEDES follows with 47% from equity (19%) and savings (28X) but a turnabout is expected if the SIMPEDES savings program of the unit desas continues to be successful. It should also be noted that both the BKK and the KUPEDES program borrow funds from external sources at rates close to market levels (12-15X). The KIK/KMKP programs show a heavy dependence on external sources (80%), about 55% from BI and 25% from the World Bank. At the other end of the spectrum, the handling banks under the Bank-assisted tree crop projects have a 0-7% stake in the credit program and depend almost entirely on GOI funds. The resource pattern of KIK/KMKP and the tree crop credit schemes provides less incentives for the banks to recover payments and maintain the financial viability of the programs. Also, it reinforces the perception of borrowers that loans are from the government and may be consid- ered grants. 3.17 Costs of Lending, Borrowing, Intermediation and Subsidies. The effective lendi ng rates of the selected special credit schemes range from the mandated and highly subsidized rates of the Bank-assisted tree crop credit schemes of about 3.0-7.7% to 129.6% p.a. under the "daily payment" loan scheme of the BKKs (Annex 3, Table 13). The high lending rates of BKKs reflect the costs of small loans for the rural poor without collateral and with a high potential for default. However, the actual interest income earned by BKKs on their loan portfolio (26% p.a.) is way below the nominal and effective lending rates and is caused by the accumulated arrears of the program (Annex 3, Table 13). On the borrower side, the low subsidized lending rates do not necessarily translate into low borrowing costs because of transaction costs which include transport, opportunity and other miscellaneous costs. The non- interest costs may be highest for a KIK/KMRP borrower who lives in a village because the loans require considerable documentation and are dispensed at the branch level located in the district. These costs are low for the unit desas which are accessible in the kecamatan level and even lower for the BKK posts located at the village level. Furthermore, the lending procedures and approval process of the KUPEDES and BKK programs which stress the prompt and speedy delivery of credit reduce the transaction costs for the borrowers. The average time lag between loan application and disbursement in KUPEDES is about two weeks for first-time borrowers and about two days for repeat loans. The BKK process takes an average of two days and requires a simple one-page application with the signature of village head as character reference. 3.18 The intermediation margin (the difference between the effective lending rates and cost of funds) in the case of the tree crops and KIK/KKKP programs is fixed because of mandated rates on the cost of funds and the lend- ing rates. The rates for tree crops (-1.8% for SRDP Il, 1.4% for SCDP and 5.1% for SRDP I) are not sufficient to cover the operational costs and default risks of the banks and provide no incentives for the banks to participate in the credit schemes. Likewise, the 6.3% margin for KIK/KMKP handling banks - 36 - does not cover the administrative expenses and collection costs of the program resulting in unprofitable operations and lack of interest of the banks. The intermediation margin of the BKKs is 13.8%, which covers its 10.4% ratio of operational expenses to loans outstanding leaving a small percentage for bad debts and profits. For KUPEDES investment loans, the margin (14.2% p.a.) is lower than that of working capital loans (20% p.a.) deupite its cheaper and subsidized cost of funds. A KUPEDES investment loan is thus a losing proposi- tion since it does not cover the 16.2% ratio of operational expenses to loans outstanding. This may explain the preponderance of working capital loans (93%) in the KUPEDES loan portfolio. The operational costs of the BKK and KUPEDES programs may reflect the high cost of small-scale, high risk lending operations, but there is scope for reducing intermediation costs through financial innovations such as group lending, joint liability groups, diversi- fication of risks and expansion of scale of operations. 3.19 Compared to the saving deposit rates of 12-15%, the cost of funds under the tree crop and KIK/KMKP programs of 0-8.3% imply a net real subsidy. The implicit subsidies involved in the credit programs under review are sum- marized in Table 3.1. Additional subsidies from bad debts that need to be written off, financial losses of the credit insurance agency and the share in the loan insurance premium paid by GOI have not been estimated. The KIK/KMKP subsidies have accrued mainly to established entrepreneurs and large borrowers possessing land and property for collateral which have not improved income distribution. Table 3.1: GOI SUBSIDIES TO SELECTED SPECIAL CREDIT PROGRAMS (Rp billion per year) Subsidy to cost Subsidy to Direct interest of funds /a borrowers /b subsidy Bank-assisted tree crop credit schemes 42/c 190/d 17/e KIK/KMKP 80/f 73/g 18/h KUPEDES investment loans 1.5/h /a Assuming savings deposit rate of 12%. 7T Assuming market lending rate of 18%. 7E Estimated by multiplying the planned amount of smallholder credit by the difference between the savings deposit rate and the effective cost of funds. /d Estimated by multiplying the planned amount of smallholder credit by the difference between the market lending rate and the effective lending rate. /e Direct interest subsidy to borrowers: SRDP I and II = 14.5Z Phase I, 3.5% Phases II and III; SCDP = 3.85% Phase I, 7.35% Phase II, 1.35% Phase III. /f Estimated by multiplying the outstanding amount of BI liquidity credits by the difference between the savings deposit rate and the effective cost of funds. /g Estimated by multiplying KIK/KMKP loans outstanding by the difference between the market lending rate and the effective lending rate. /h Direct interest subsidy to banks of 1.5% p.a. of the KIK/KMKP loans outstanding. -37- C. Lessons Learned from Successful Credit Schemes 3.20 The major lessons learned from the success of the KUPEDES and BKK lending programs are as follows: (a) Successful credit programs can be designed to reach the lower income segment of the rural population through the absence of collateral requirement, small loan sizes, market interest rates, and low transaction costs through simple lending procedures and accessibility of bank offices. (b) Interest rates that cover the costs of lending to small borrowers have enabled the KUPEDES and BKK programs to provide wide access to credit for the rural population while the low, fixed and adminis- tered rates of the KIK/KMKP credit programs have favored large and wealthier clients over smaller borrowers with higher risk and search costs that cannot be adequately covered by the mandated lending rates. Market lending rates have also provided adequate financial margins for the banks to maintain the viability of the credit programs and ensure their continued interest and participation in the programs. (c) The availability of credit and lower transaction costs through simple administrative and lending procedures and less reliance on collateral are likely to be more attractive features for the "economically disadvantaged" borrowers, including women, than "below market" interest rates. (d) The successful repayment performance of the KUPEDES and BKK programs is attributed to interest rates that adequately cover lending and collection costs, financial incentives to borrowers for prompt pay- ment, staff bonuses based on efficient and profitable operations, lending on the basis of character and repayment, and full accounta- bility of the financial institution for the lending risks. (e) Liberal credit guarantee facilities, as provided by ASKRINDO for KIK/KMKP encourage lax credit collection, lead to higher defaults, and endanger the long-term viability of the program, the lending institution, and the credit insurance agency. (f) An adequate resource base of internally generated funds from earn- ings and savings deposits is essential to the sustainability of a credit program. A financial institution should have a high percentage of its own funds in the program as incentive for sound loan managment, a stringent loan recovery program, and continued financial viability of the program. 3.21 Future Strategy. Supply-leading and commodity specific credit programs have provided the bulk of agricultural credit in the past. At the early stages of credit development, these were justified by the failures and imperfections of the credit market and the objective of Government to accelerate development of priority sectors and groups through subsidized financing. In the future, there will probably be a continuing need for - 38 - directed credit within a project framework in order to provide an integrated package of inputs to borrowers. They would finance activities, e.g., tree crops, which are considered economically viable but are financially nonbankable given the market conditions for the supply and demand for funds. However, their design, terms and conditions, and subsidy structure (if needed) should be improved to minimize the distortionary impact on the rural financial system. Along with targeted credit schemes, an alternative approach should be developed similar to the KUPEDES program which should be demand-oriented and nonsubsidized enhancing the development and viability of rural financial institutions and providing credit on an integrated farm system basis rather to specific commodities only. At a later stage of financial and economic development, these two tracks (directed and general credit) might converge with financial institutions taking the lead in cultivating the rural credit market and clientele and developing credit and savings programs consistent with their cost structures and suitable to local needs and conditions. - 39 - IV. FINANCIAL RESOURCE MOBILIZATION 4.1 As shown in the previous chapters, the efficient and effective mobilization of financial resources is vital to the development of viable and sustainable rural credit institutions and programs. Although overall bank deposits have grown dramatically, past efforts of rural savings mobilization have been limited both in the scale of operations and the range of mechanisms utilized. This chapter examines the experience of selected savings instru- ments and the factors that influence financial savings mobilization. A. Assessment of Selected Instruments and Mechanisms for Mobilizing Savings 4.2 Savings Account. The full range of financial assets commonly available to Indonesian investors is shown in Annex 4, Table 1 and their specific terms and conditions discussed in Annex 8. Saviygs accounts are the most common and generally take the form of TABANAS/TASKA,_ the national savings and insurance scheme, or SIMPEDES, a recent BRI unit desa savings program. All deposits in state banks, BRI unit desas, the Postal Service, or TABANAS/TASKA accounts are guaranteed by BI and the GOI, and the interest on most savings and time deposits is exempt from tax. The current National Savings Drive Committee has concentrated on the promotion of TABANAS to the exclusion of other saviLgs instruments. 4.3 In contrast to the general growth in time and other savings cate- gories, the two national savings schemes, TABANAS and TASKA, sponsored by BI and offered by state banks and some private banks, have shown very little progress (Annex 4, Table 2). The TASKA deposits have hardly moved over the last few years and are of little importance in overall savings. TABANAS has only maintained, not expanded, its overall share of deposits, because banks have been unable to differentiate their TABANAS accounts from their competi- tors' and because some institutions, most notably BRI, have created their own proprietary savings schemes. Many banks also view TABANAS as a public ser- vice, in that the cost of handling these accounts is high compared to the level of deposits (an average balance of Rp 67,973) and the interest rate paid (15X on balances up to Rp 1 million and then 12X). Banks prefer to compete with time deposits which raise larger amounts of funds for longer time periods and at lower administrative cost. TABANAS is in effect a form of restricted demand deposit. 4.4 A significant change in the savings deposit mix is the exceptional growth of BRI SIMPEDES scheme (Annex 4, Table 2). The SIMPEDES scheme is offered only through the BRI unit desa system and hence primarily in rural areas. While accounting for only 0.07Z of total deposits, it went national in June 1986. It has since grown exceedingly fast and by December 1986 amounted to Rp 82.41 billion (see Annex 8 for more detail). A combination of no with- 1/ TABANAS (Tabungan Nasional) is a Small Saving Program, TASKA (Tabungan Isuransi Berjangka) is National Insurance Savings Scheme. - 40 - drawal restrictions, a consumer goods lottery plan, and relatively high interest rates has made it particularly popular with rural savers. This exceptional growth is expected to continue for the immediate future through improved marketing in rural areas. 4.5 Probably because of the relative ease of funding through BI liquid- ity credits, the commercial banking sector has not been particularly innova- tive in raising savings in either rural or urban areas. The concept of paying salaries directly into a savings account is just being introduced through TABANAS in parts of Sumatra, and the only examples of goal or target savings accounts and local saviaLgs plans are through the efforts of Bank Dagang Bali (BDB). Except in some less formal institutions, such as self help groups and informal credit unions, savings is seldom linked to any subsequent loan applications. Within the banking system, only the National Savings Bank and BDB have a down payment savings account that uses the potential of loan finance as an additional incentive for savers. This is surprising since the borrowers' ability to repay, as proven through a good saving record, could play an important role in personal credit evaluation and, as a condition for lending, help reduce arrears. While some institutions (BKKs and KURKs) do have compulsory savings, most of these are a compensating balance funded out of the loan itself, rather than from the borrower's regular savings efforts. 4.6 In general, while the existing savings instruments have offered high positive real returns over the last few years, the net return to all savers may not have been positive due to high costs of rural dwellers outside of Java and Bali who have poor access to banks. Another problem with these instru- ments is their short-term maturity. Those with slightly longer terms, such as time deposits, Certificates of Deposits (CDs) or securities, entail a larger initial investment than most rural savers would be able to afford. The potential for life insurance policies and long-term bond issues has also not been explored. Thus, for both medium- and longer-term investment, there is still a bias in favor of physical assets. Also the exposure to risk is limited, since most financial instruments offered by the formal sector are relatively risk free and provide fixed rate returns. The lack of risk/rate of return alternatives means that small investors wishing a slightly riskier investment are forced into physical assets or lottery style gambles. 4.7 Lotterv Schemes. Lotteries have been important in encouraging savings accounts with Indonesian banks. A lottery feature was added to the TABANAS/TASKA scheme in l197. When BRI developed its SIMPEDES scheme, market researc% argued strongly for a lottery component as part of the savers' total return.- The market surveys also found a demand for prizes in the form of tangible goods, such as motorcycles rather than for money, and for the ability to increase their chances of winning through higher savings balances. The SIMPEDES scheme was designed specifically to meet these demands. 2/ This was an interesting finding since a rational investor should prefer an additional return in the form of higher interest and the option to purchase lottery tickets with the additional funds. - 41 - 4.8 While TABANAS emphasizes prize winning in its promoti.onal efforts, the chances of winning a major prize are quite small (one in 11 million) and the amount allocated for prizes each year (Rp 1,274 million in 1986) is a small percentage (0.12%) of total deposits (Rp 1,045,300 million). In con- trast, the SIMPEDES lottery provides prizes worth about 1.25% of the average funds on deposit. Besides the greater prize money, the tangible prizes from SIMPEDES are preferred over the straight cash payments from TABANAS. 4.9 The cost of operating the lottery also differs considerably between TABANAS and SIMPEDES. TABANAS involves relatively little administrative cost. The numbers are drawn and the results are published in the press and posted in offices of the handling banks. The handling bank claims the amount from BI and pays it into the customers' account. There is no other contact with the customer. SIMPEDES is much more complex. Each unit desa issues one coupon per month per account. The coupon contains as many lottery numbers as there are multiples of Rp 5,000 in the account's minimum balance. BRI estimates this process costs 0.12% p.a. of the total SIMPEDES deposits. 4.10 Given that TABANAS pays 15% p.a. on balances of up to Rp 1,000,000 and SIMPEDES pays a maximum of only 12% p.a., most smaller savers should prefer TABANAS. SIMPEDES' additional 1.25% in prizes hardly offsets its lower interest rate. Moreover, SIMPEDES pays only 9% p.a. on SIMPEDES balances under Rp 200,000 and no interest on balances under Rp 25,000. The majority (55%) of SIMPEDES savers have balances earning only 9% p.a. with the average account balance being Rp.196,696. Of the remaining accounts, 26% have bal- ances under Rp 25,001 and earn no interest, and 19% are over Rp 200,000. 4.11 Both TABANAS ard SIMPEDES are offered at each unit desa, but in November 1986 SIMPEDES deposits exceeded TABANAS deposits. Thus, rural depos- itors are either poorly informed about the returns on the two accounts or are willing to accept SIMPEDES' lower interest in return for the possibility of winning tangible prizes. While Indonesia has national sports lotteries, in rural areas participation may be more difficult, or culturally there may be negative associations with gambling ir sports lotteries. Also, physical goods seem to have a greater appeal for rural savers than monetary prizes as reflec- ted in their preference for physical rather than financial assets. Another reason savers may prefer SIMPEDES accounts is that they perceive the maximum two withdrawals per month restriction under TABANAS as a potential problem although only one in four SIMPEDES accounts has a withdrawal each month. Finally, BRI unit desa staff may actively discourage TABANAS deposits in favor of SIMPEDES accounts. 4.12 The BKKs will begin to accept voluntary deposits later in 1987 and will test which feature draws savings, whether interest rates, lotteries or withdrawal restrictions. Like SIMPEDES, BKK accounts will have no restric- tions on withdrawals. However, instead of a lottery, they will pay a higher interest rate, possibly as much as 15%. 4.13 Door-to-Door Savings and its Cost Effectiveness. To date, rela- tively few commercial banks use mobile banking units and door-to-door banking. Most commercial banks prefer to wait for customers rather than seek out sav- ings. The secondary banks, particularly the Bank Pasars, are much more active - 42 - in encouraging deposits. For example, in one district in Bali, some villages are served at least weekly by nine different financial institutions including the BRI unit desa, two private national banks and six secondary banks. 4.14 Among these institutions, BDB provides the most extensive mobile bank service with 213 of its 426 staff using 28 mini-vans and 42 motorcycles to visit customers in both urban and rural areas. In some cases, the staff conduct business in a set place within the village (a savings post), but more often they visit their clients' homes or businesses. In urban areas bank staff walk to nearby clients, mainly shops and other businesses, on a daily basis to collect deposits. They also help clients with loan applications, accept loan repayments and allow withdrawals. As an added incentive, mobile unit employees receive a commission of 0.15% on ordinary savings collected, 0.25% on time deposits or new loans, and Rp 100 for each new account opened. 4.15 BDB estimates its cost of mobile services at Rp 2 for every Rp 1,0l0 raised. The total cost of this service is difficult to quantify in percentage terms as the amount collected in each village varies considerably. One rural savings route serviced daily by a two-man motorcycle team, consisted of 350 customers and had an average savings balance of Rp 6 million while another had 1,000 customers with some Rp 300 million in deposits. On average, BDB's vil- lage routes had outstanding savings of slightly less than Rp 50 million. With a monthly cost of Rp 300,000 per team and one team servicing an average of five areas, the cost of mobile savings is estimated at 0.12% p.m. or 1.4% p.a., which makes it a profitable activity for financial institutions to undertake. B. Factors Affecting Financial Savings Mobilization 4.16 Availability of Concessional Funds. Perhaps the most important disincentive to mobilizing savings in the rural sector is the availability of low cost credit resources from BI, GOI and other external sources. The bulk of BI liquidity credits in the last four years has gone to state commercial banks (75-80X) or to publicly owned banks (95X) (Annex 4, Table 3). A comparison of the resource structure of different types of banks in Table 4.1 reveals an inverse relationship between the extent of borrowings from BI and the percentage of resources mobilized from deposits. The greater the access of a financial institution has been to BI credit, the lower the amount of savings mobilized. Deposits of private commercial banks that are not dependent on BI liquidity credits have also grown much faster in the last three years than those of the national foreign exchange banks consisting primarily of the state commercial banks and the development banks (Table 4.2). The latter rely on BI liquidity credits at 3X and captive public sector deposits, rather than compete actively for ordinary savings and time deposits at 12-18X. - 43 - Table 4.1t RESOURCE STRUCTURE OF DIFFERENT TYPES OF BANKS (Z Distribution) National foreign exchange banks National (5 state banks and Development Foreign private 10 private banks) banks banks banks Deposits 51.7 34.0 84.C 73.8 Borrowings from BI 22.0 25.0 0.8 7.5 Government deposits 6.4 3.2 0.4 - Others 19.9 37.8 14.8 18.7 Total 100.0 100.0 100.0 100.0 Table 4.2: GROWTH OF DEMAND, TIME AND SAVINGS DEPOSITS OF DIFFERENT TYPES OF BANKS (X) 1983/84 1984/85 1985/86 National foreign exchange banks 22 37 4 Development banks 33 24 -3 Other commercial banks 43 61 14 4.17 Choice of Savings Instruments. An important factor affecting savings mobilization is the range of financial instruments available in the market and their terms and conditions. Rural savers have a limited choice of three types of bank accounts: demand deposits, savings deposits and time deposits. At the village level, the options are more limited to compulsory savings accounts with the BKKs, BKDs and KUDs and SIMPEDES at a few BRI unit desa savings posts. Most of these compulsory savings take the form of a com- pensating balance funded out of the loan itself rather than from the voluntary savings efforts of the borrower. Longer-term assets such as time deposits, CDs or securities, life insurance policies, contractual goal savings plans, etc. are not available or entail a large initial investment which is beyond the reach of most small savers. With respect to equity instruments, GOI has chosen not to issue its own government securities for sale in the domestic market. This policy has not assisted the development of an active capital market. Pensions and superannuation funds, potentially the largest source of long-term funding in the financial sector, have not grown rapidly either due to difficulties in obtaining MOF approval. The small number of pension folnds also contains the number of bond issues which can be absorbed in the market - 44 - and has ruled out the bond markets' expansion. While these instruments may have minimal direct effect on savers, the lack of growth in these areas limits the degree of choice for potential investors and the pace at which savings are transferred from physical to financial assets. 4.18 Institutional Capacity for Mobilizing Savings. While the rural sector accounts for nearly 94% of the offices of Indonesia's financial institutions, they hold only 11% of total deposits (Annex 1, Table 17). This is due in part to differences in income levels between urban and rural dwellers and also to the quality and range of financial services offered by a full commercial bank branch in urban areas as compared to a village bank. There is also a marked difference in the availability of bank offices/units between regions and this is particularly true for the numerous secondary banks which are almost all located in Java and Bali. 4.19 The rural savings market is dominated by BRI through its branch structure and unit desa operations. Neither of these is located at the vil- lage level, but rather at the district and subdistrict, respectively. Some BRI unit desas serve larger villages through village savings posts (a form of part-time bank agency), but at present most savings posts were created by closing those unit desas that were not operating economically on a full time basis, and hence are still at the subdistrict level. However, expansion of saving posts into other villages is planned. The Postal Service also accepts savings as an agency for the National Savings Bank, but post offices are norm- ally at the district, and seldom at the subdistrict level. Of the numerous secondary banks, only the Bank Pasars take voluntary savings deposits; the rest (BKDs, BKPDs, LDs) have compulsory savings schemes. Many KUDs accept deposits, subject to certain restrictions, while the BKK-type institutions are considering the expansion of their current compulsory savings to a voluntary savings program. The pawnshops which are the most active RFI in the lending market do not mobilize savings. Finally, a range of informal institutions (rotating credit societies, savings and loan associations, and self-help groups) conduct some savings and lending operations but they comprise a minor portion of total rural savings. The limited number of financial outlets at the village level and their narrow scope of operations constrain the rural savings mobilization effort by increasing the transaction cost and lowering the expected returns of the rural savers. Whereas savers have low transaction costs in some areas (in Bali, banks literally come to their door), in other regions (such as parts of Sumatra) their travel time and costs to reach a financial institution are considerable. 4.20 Regulatory Restraints. As of January 1987, the remaining interest rate controls imposed on commercial, savings and development banks include the 12% and 15% ceilings on TABANAS savings accounts, the 9% ceiling on TASKA insurance/savings deposits and the minimum 12% rate on government bank time deposits of 24 months. At present secondary banks are not restricted on their deposit rates, but are limited to a maximum maturity of 3 months on their time deposits although in practice some have circumvented this restriction through informal rollover agreements with their customers, thus providing in effect 12 month time deposits. These restrictions limit the freedom of these insti- tutions in deciding the terms and conditions of their own savings instru- ments. Coupled with the fact that only TABANAS accounts and state Bank deposits are guaranteed by GOI, the private banks have less incentives in developing their own savings mobilization mechanisms. - 45 - V. WOMEN IN THE RURAL FINANCIAL SECTOR 5.1 This chapter addresses the issue of the role played by women in the rural credit market. There are a number of difficulties in assessing the degree and type of participation by women in terms of assessing both demand for, and availability of, credit. On the demand side, one difficulty is the interrelationship of the numerous production and consumption decisions within the household, which makes it hard to disentangle the uses of borrowed funds and to identify the end-recipients or beneficiaries of a loan. The formal loan applicant or applicants are not always the actual users of the loan, and funds will not always be used for the purpose stated in the application. The second difficulty arises from the fact that most lending institutions and programs do not disaggregate data on their borrowers by sex of borrower. As a result, much of the information presented here is based on special studies and the impressions of the managers and staff of banks which were visited during field trips to rural areas. 5.2 The limited data indicate that Indonesian women participate actively in the rural financial sector (Annex 5, Table 1). Women are 23.4% of KIK/KMKP borrowers; 25% of KUPEDES borrowers (at BRI Unit Desas); and 29% of borrowers at one Bank Pasar. Women are a higher percentage of borrowers in nonbank financial institutions such as BKK (60%) and KURK (57%). As many as 80% of the borrowers from government pawn shops may be women. Village-level studies show that women borrow extensively from informal sources, such as suppliers, traders, neighbors, friends, and money lenders, for both consumption and business purposes. Women also have organized informal savings and loan asso- ciations, such as "arisans" (rotating savings and credit associations) and "simpan pinjams" (savings and loan associations). 5.3 In the first section the factors that affect women's demand for credit are discussed. This is followed by an assessment of factors affecting supply of credit to women and concludes with a look at the role played by women in savings mobilization. A. Factors Affecting Women's Demand for Credit 5.4 Women in Indonesia constitute one-third of the rural labor force, and some studies indicate that their role in the management of household re- sourcelsmay be even greater than their participation in market activities sug- gests.- Yet it is difficult, and perhaps misleading, to attempt to quantify their demand for credit. First, as mentioned above, it is hard to separate demand for credit on the part of women from the demand of rural households. Second, in so far as demand for credit in the rural areas (particularly that of women) is satisfied by informal sources, there is no adequate source of da- ta to estimate current and future demand, even with interest rates and econo- 1/ Hanna Papanek and Laurel Schwede, "Men Can't Handle Money But Women Can Be Trusted: Earning and Spending in an Indonesian City", 1984, mimeo; Benjamin White and Eudany Lestari Hastuti, "Different and Unequal: Male and Female Influence on Household Affairs," Agro-Economic Survey, Working Paper No. 6 (Bogor: Bogor Agricultural University, 1980). - 46 - mic conditions given. Despite these cautions, however, it is possible to iso- late certain factors that will affect both the magnitude and the direction of women's demand for credit. In addition, the recent figures on women's parti- ci.pation in different credit programs and institutions, summarized above, pro- vide an indication of women's current demand for credit. 5.5 Since the role of Indonesian women in household consumption deci- sions is widely acknowledged, vomen's demand for credit is expected to be largely for consumWion purposes. A study of rural credit in three West Javanese villages - supports this view to a certain extent; a larger propor- tion of loans to women (61.2%) are used for consumption, than men's loans (on average, 41.5% of loans to men are used for consumption) (Annex 5, Table 2). Women in the villages studied also borrowed in kind more frequently than men. 5.6 But women's participation airong rural borrowers extends beyond their role within the household. As income-earners, women also represent an impor- tant market for lenders, both for consumption and business purposes. While agriculture is still the major source of employment for rural women, over the past 15 to 20 years, changes in the structure of agricultural production have led to a relative decline in opportunities for women in agriculture. But there has been an increase in women's participation in trade and, to a lesser extent, services*.3 A number of village-level studies have highlighted the high prou?rtion (over 50%) of women traders at the village level, specifically on Java.- 5.7 Another factor in determining women's demand for credit (particular- ly commercial credit) is women'a ownership and operation of businesses. About 28.3% of all the self-employed and employers in rural Indonesia are women (Annex 5, Table 3). While women are less than one-fourth of self-employed/ employers in rural agriculture, they ari 42% in the tertiary sector (again, trade and services) of rural Indonesia.-/ These figures suggest, first, that women's demand for credit will not be limited to consumption, but will include 2/ Karl f. Jensen, "Rural Credit in Banten, West Java: A Snapshot," Report to USAID/Indonesia (Jakarta, December, 1986), mimeo. 3/ Mayling Oey, "Changing Work Patterns of Women in Indonesia During the 1970s: Causes and Consequences," PRISMA, No. 37 (Sept. 1985), p. 18-41. 4/ Nancy Peluso, "Occupational Mobility and the Economic Role of Women," Rei3rt. No. 39, Population Studies Ctr., Gadjah Madah University, Yogyakarta, 1982; Jennifer Alexander and Paul Alexander, "Finance and Credit in a Rural Javanese Market: An Anthropological Perspective", (August 1986). 5/ There is a great deal of variation among different sources of data. For example, the national labor force survey of 1977 shows only 15.2Z women among the self-employed/employers in agriculture (vs. 22.5% for the 1980 census). More women in agriculture were classified as unpaid family workers by the labor force survey. - 47 - demand for working capital and investment credit for their businesses. Second, one would expect women entrepreneurs in the trade and service sectors to be a significant share of commercial borrowers in those sectors, although given their lower levels of income and returns, the amounts borrowed will probably be smaller than those of male borrowers. 5.8 Given their level of economic activity, women's demand for credit is likely to be highly sensitive to transaction costs as well as the interest rates associated with borrowing. The negative effect of transaction costs of borrowing on women's demand for credit is largely related to their status as small borrowers and the location of their businesses at the village level. Observers of the rural financial sector in Indonesia have argued that the demand for credit by small borrowers is relatively interest-inelastic; trans- action costs often represent a greater proportion of borrowing costs on small loans.- The major elements of transaction costs for the borrower are trans- portation, opportunity costs of time required to complete the transaction, and special fees. Most of these factors will affect the borrower's demand for credit in a similar way, regardless of sex. However, opportunity costs of foregone labor have different implications for women, who on average work a longer day for a lower wage. While the opportunity costs are perhaps lower than those of men, there are indirect costs in terms of household tasks foregone, and substitution of child labor in the home, that play a role in determining demand. 5.9 The highest transaction costs for rural women are associated with loans from commercial banks located at the provincial capital and district (kabupaten) levels because women must travel greater distances to reach them. Women's econoric activities tend to be concentrated at the village level, particularly in agriculture, handicrafts and trading. Commercial banks also tend to have more time-consuming paperwork requirements. For special government crqit svcemes at state-owned banks, these requirements may be even more onerous.- 5.10 The lowest transaction costs are found with the money lenders and other informal intermediaries who go to the borrower. The convenience of borrowing from these sources is cited repeatedly by borrowers as their reason for dealing with money lenders and the like, despite the high interest rates they charge. After the informal sources, the most convenient lenders appear to be financial institutions that are easily accessible, such as the BKKs. Women in particular appear willing to pay the higher rates these institutions 6/ Claudio Gonzalez-Vega; World Bank Rural Credit Study, 1983. 7/ At one BRI branch visited, for example, one loan application file contained two inches of documents, including business licenses, insurance certificates, tax certificates, a mortgage on collateral posted, notarized loan agreements, notarized authority to sell collateral, financial statements, quarterly reports on the business (balance sheets, income statements, etc.), and a number of signed documents written in Dutch. - 48 - charge. The BRI Unit Desas have incorporated some features of the BKK, but their locations are further removed from the village leptl, and their clien- tele is economically better off than the BKK borrowers.- 5.11 Social and cultural factors may also influence women's demand for credit in Indonesia. On the negative side, cultural norms may limit women's ability to undertake credit transactions with men. One source in Central Java claimed that women were actually a higher proportion of borrowers than the figures for the program with which he was associated suggested. He argued that cultural restrictions prompted women to seek credit through their hus- bands. In a recent anthropological study of market women in Central Java, some informants argued they 37eferred to deal with female informal lenders, rather than male bank staff..- Of course, these restrictions will vary widely by cultural group and class. Women who are already economically active, and deal with men outside their family in business relationships will probably be less constrained by such restrictions. Cultural factors also influence the structure of information channels. Women may not know about credit programs or how to apply for loans if information is spread through male networks such as cooperatives or farmer's associations, or if marketing is directed to men. On the other hand, there may be cases where cultural influences increase women's role in credit transactions. If household expenditures are in the hands of women, for example, their demand for credit may be increased, partic- ularly for consumption. 5.12 To summarize, women's demand for credit is determined by their eco- nomic activities as well as their role in the household. The high proportion of women among the self-employed suggests that women's demand for credit will not be limited to consumption, but will include demand for working capital and investment credit for their businesses. Women's lower incomes probably trans- late into demand for smaller size loans, on average, than those demanded by men. Cultural factors and lack of information about lending may also serve to limit women's demand for credit. But given the level of their economic activ- ities and the multiple time constraints that women face because of their dual productive/reproductive responsibilities, it is the transaction costs associa- ted with borrowing that are likely to be the most important factor limiting women's demand for formal sector credit. The evidence provided by women's high level of participation in financial institutions which maintain lending operations at the village level, making it possible to obtain credit without numerous trips to the subdistrict (kecamatan) or district (kabupaten), and keep loan application forms very simple, supports the conclusion that women's demand for credit is greater when transaction costs are low, even if interest rates are much higher than those charged by the commercial banks. 8/ US Agency for International Development, Financial Institutions Development Project: Project Authorization Amendment (Jakarta: US AID/Indonesia, 1986). 9/ Alexander and Alexander, 1986. - 49 - B. Factors Affecting Supply of Credit to Rural Women 5.13 One reason often cited as an explanation for women's limited parti- cipation in formal credit programs is sex discrimination by lenders. In fact, the problem is more complex. At the sites and institutions visited in rural Indonesia, there was no evidence that negative perceptions on the part of lenders limit the supply of credit to women, just because they are women. Other factors appeared to be far more important in determining the supply of credit to women. A number of these are factors that affect all small borrow- ers, regardless of their sex. Since women are concentrated among this group, a greater proportion of them would be affected by these constraints. 5.14 Transaction Costs. Many lenders are reluctant to lend to small bor- rowers, including women, because of the high unit costs of making small loans. If similar procedures are used for all borrowers, there will be little difference in loan processing time for small and large borrowers. This means that, proportionately, costs will be higher for small loans. Data on the operating costs of programs that lend very small amounts--BKKs and pawn shops, for example--support this conclusion even though these institutions have taken compensating steps to streamline their procedures and tailor them to small borrowers. To the extent that interest rates are regulated, banks cannot compensate for higher costs and/or risks of making small loans by increasing their lending rates. Therefore, the removal of interest rate ceilings should have the effect of making it easier for small borrowers to obtain loans. 5.15 Targeted Lending. If lending is targeted to specific economic activities, this may serve to reduce the supply of credit to women. As men- tioned earlier, women are a smaller proportion of self-employed business oper- ators in agriculture, but they are over 40X of the self-employed in trade and services. Therefore, special credit programs which target agriculture only, or particular crops, would not include as many women borrowers. Unfor- tunately, there is no sex disaggregated data available on agricultural lending to test this hypothesis. However, it can be supported to an extent by data on lending under programs not targeted by activity (such as KUPEDES and BKK), where women's participation is high. 5.16 Collateral. Most formal lenders will not make business loans with- out collateral. This is a problem especially for small borrowers, but even relatively high-income rural households in Indonesia may not have,glear title to land or other property that lenders will accept as collateral.- Under BRI's KUPEDES program for example, bank regulations permit a variety of assets to be used as collateral, including land, buildings, vehicles, and even furni- ture. In practice, however, unit desas generally require land as collateral. More specifically, land ownership must be proven by possession of the haq milik, a clear, government-registered title that is very costly and time- consuming to obtain. 10/ The 1983 Rural Credit Study (p. 6) pointed out that 37X of households in the top quintile by household receipts were landless. - 50 - 5.17 The ability for women to own land is perhaps less restricted in Indonesia than in many other developing countries, because officially, property acquired prior to marriage continues to be individually owned and any property acquired during the marriage is jointly owned. However, secondary sources indicate that it is more common for land titles to be registered in a man's name, even with property which is jointly owned.- In addition, while the civil law on inheritance provides for equal shares of property to go to surviving children, inheritance is also influenced by adat customary law and Islamic law which mandate a larger share of inheritance for male heirs. The Minangkabau of Western Sumatra provide a notable exception to these customs. In that matrilineal society land holdings are customarily passed on through the women in the family and registration of land in the woman's name is more common. 5.18 Cosigning. BI claims that there is no formal, legal requirement that spouses cosign for business or personal loans. Local custom and/or the practices of specific banks may, however, dictate this practice in a number of areas or institutions. According to its staff, the BRI, the largest rural lending institution in Indonesia, has a policy of requiring a spouse to cosign an application for a loan by the other spouse. 1ata on KUPEDES lending indi- cate it is not being applied across the board.- If it is the case that female borrowers are required to have their spouses or other male relatives cosign for their loans more often than men, this discriminatory practice might discourage borrowing by women because of the added transaction costs involved in getting a spouse to agree tc cosign, and transportation and time costs for the cosigner on top of those already associated with borrowing. 5.19 One Loan Per Household. According to BRI staff interviewed in Jakarta, BRI has a policy of granting only one KUPEDES loan per household, even if household members have separate business activities. It is not known whether other state banks have a similar policy. The rationale behind this policy is the assumption that businesses are interconnected within the house- hold, and even when a loan is used for a specific activity, additional income generated from that business may be invested in another business or used for consumption by other household members. This assumption may hold true for 11/ Pauline Milone, A Preliminary Study in Three Countries: Indonesia Report, Report to USAID, Office of Women in Development, Washington, D.C., 1978. Personal interview with Nani Yamen, Director, the Institute for Consultation and Legal Aid for Women, January 19, 1987. 12/ A study by the Center for Policy and Implementation Studies (CPIS), based on data for all KUPEDES borrowers at an unspecified number of unit desas revealed that only 57% of the loans had consigners. Since the number of male borrowers is greater than the total number of cosigners, we can conclude that some men received loans without cosigners. But although the ratio of male cosigners to female borrowers is higher that the reverse case, one cannot tell from the data provided by CPIS whether men's loans were made to unmarried individuals, whether some women also qualified for loans without cosigners, and so on. - 51 - many small business loan applicants in Indonesia, but studies in other devel- oping countries have shown that there may be substantial indeyg?dence between men's and women's income-earning activities in the household.-'/ In households where individual economic activities are relatively autonomous, the one-loan-per-household rule may limit access to loans for creditworthy borrowers. In the case of KUPEDES and other BRI lending, lack of collateral is likely to be the binding constraint to women's access to credit. However, in the case where each spouse owns property in his or her own name, KUPEDES would still not allow for lending to the businesses of both under existing rules. 5.20 Edueation. Women's lack of education in rural Indonesia puts them at a disadvantage in preparing complex loan applications, investment plans and financial statements that lenders require. Over the past 20 years educational attainments have been rising for both males and females in Indonesia, but edu- cational attainment levels remain very low for older women in rural areas. Only 15.2% of women aged 30 to 49, have completed primary school (as compared to 35.8% of men in that age group). Among the rural self employed/employers who operate their own businesses, 16.7% of women and 38.4% of men have comple- ted primary school and 48.9% of rural women, compared to 27% of rural men are illiterate. However, the proportion of illiterates is lower for both women and men in the tertiary sector. 5.21 In some programs, such as KUPEDES, loan officers provide assistance to borrowers in completing the application requirements, thus making the pro- gram more accessible to less educated women. Others, such as BKK, use a very simple application form that is filled out by the loan officer and signed by the borrowers. For KIK/KMKP, on the other hand, the application procedures are relatively complicated, and over half of the borrowers had a high school education and above (although only 1.4% of the rural self-employed had such a level of education). 5.22 In summary, Government restrictions on lenders, such as interest rate controls, may lead them to ration cred:_ away from small borrowers, and those borrowers they perceive to be more risky. Women may find themselves disproportionately represented among these groups. Therefore, government policies which allow financial institutions to attract more deposits with higher interest rates and compensate for higher lending costs (to smaller borrowers) by charging appropriate interest rates, should expand the supply of credit to somen, and to other small borrowers. However, certain requirements of lending institutions serve to exclude or limit the participation of women in their lending operations. If loans are targeted to specific economic activities or crops, if land is required as collateral, a husband or male relative is required as a cosigner for loans, complicated application forms 13/ Nancy Folbre, "Hearts and Spades: Paradigms of Household Economics," World Development 14(2), 1986, p. 245-255; Christine Jones, "Gender, Technology and Development: The Household Production Approach,," Harvard Institute for International Development, 1987, mimeo. Amartya Sen, "Women, Technology and Sexual Divisions," paper prepared for UN IWSTRAW. (Oxford: Oxford University, All Souls College, 1S84), mimeo. - 52 - are used, and/or only one loan is permitted per household, lenders are in fact channeling lending away from female borrowers. In the absence of incentives for rationing credit, these measures may still be taken in the face of economic uncertainty or because they are standard banking practices. Therefore, while lenders interviewed in Indonesia do not appear to discrimi- nate against women because of their sex, some of their policies have the effect of limiting women's access to financial services. In cases where lenders had made a conscious effort to make credit available to smaller borrowers, by eliminating some of these structural barriers, such as in the pawnshops and BKK-type institutions, the participation of small borrowers and women is high. C. The Role of Women in Rural Savings 5.23 There is no data for income and expenditure in the household by sex but given the difference in wages for male and female employees, it can be assumed that women's incomes are likely to be lower than those of men. Thus, women would probably account for a smaller proportion of rural savings. As the 1983 Rural Credit Study showed, the average propensity to save of lower income households is less than the 20% estimated for rural Indonesia. Some 20% of rural households are headed by women, but a high proportion of these are in the lowest income groups. In addition, women who are heads of house- hold tend to be older and thus less important as savers. 5.24 Staff of banks visited indicated that women are more important as savers in small saver accounts. At Bank Dagang Bali, for example, it was estimated that women are at least half of all savers (but do not necessarily hold half of the value of savings deposits) in the passbook-type accounts. Bank Dagang Bali is unique among the banks visited in the extent of its savings mobilization efforts and the use of female collectors in those efforts, two factors which might make women's participation higher there. At a BRI unit desa in Bali, bank staff reported that women are a higher propor- tion of savers in TABANAS accounts (which are restricted to 2 withdrawals per month) than in SIMPEDES (an account that functions more like a demand deposit). The former also appeal to smaller savers because they pay a 15Z rate of interest on the minimum monthly balance, whereas SIMPEDES pays no interest on small balances. The staff speculated that women might have 20% of the SIMPEDES accounts. Women are a majority of savers among clients of BKK- type institutions since these programs involve a compulsory saving component. However, since it is difficult to withdraw savings from these institutions, voluntary savings is at a minimum. Studies also show that Indonesian women help make decisions about the use of income and disposal of household assets. Therefore, their potential role in savings mobilization could be greater than their individual income levels would indicate. To the extent that women are holding assets in nonfinancial forms (such as jewelry, batik, and animals), expanding their access to financial institutions could promote the conversion of these assets into bank deposits. - 53 - VI. CONCLUSIONS AND RECOMMENDATIONS A. Main Conclusions 6.1 Significant steps have been taken by GOI toward building a viable rural financial system. The financial reforms of 1983 particularly the interest rate deregulation have led to a dramatic improvement in the levels of total and rural savings deposits in the country, although recent evidence indicates a weakening of these effects after the initial surge in savings. Agricultural credit has increased over the last five years in current and real terms and in relation to total credit for all sectors and to agricultural GDP. Indonesia has established a large number and a great variety of RFIs operating at the kecamatan and village levels that are market-oriented and subject to few restrictions and interventions. The recent shift of the BRI unit desas from channeling subsidized and targeted credit to the provision of general nonsubsidized rural credit has had encouraging results. The RFIs have played an important role in providing broad access to credit for a relatively large number of small borrowers in the rural areas. 6.2 Despite these achievements, there is substantial scope for improving the rural financial system to efficiently service the credit requirements of rural and agricultural development. The bulk of rural and agricultural loans outstanding is provided under specific credit programs that remain subject to GOI interventions such as directed credit to priority sectors, subsidized cost of funds through BI liquidity credits, prescribed interest rates, and liberal credit insurance. Although agricultural credit has increased, the major part has gone to estates and large and wealthy borrowers for tree crop production and processing. Under KIK/KMKP, one of the largest credit programs of GOI, small borrowers have effectively been screened out by banks that are unable to cover their lending costs by altering the administered interest rate thus adversely affecting income distribution. Less than 3% of total rural credit outstanding is granted on an unsecured basis, which has limited the access to credit of low-income groups and women who do not possess land and property for collateral. As shown by the KIK/KMKP experience, the performance of subsi- dized credit programs has not been satisfactory. High default rates and dependence on GOI and external funds have impaired the viability and long-term sustainability of this credit program. On the other hand, there are some successful programs which have achieved relatively high repayments and a degree of financial integrity such as KUPEDES and the program of the BKKs. The lesson seems to be that programs with simple administrative arrangements, market interest rates, demand orientation and low transaction costs are more efficient credit delivery mechanisms. 6.3 In spite of the large number and complexity of financial institu- tions in Indonesia, the density of bank and nonbank offices remains low. The large financial institutions have minimal presence or none at all below the district level and lend only 16X of their loan portfolio to the rural sector. The scant penetration of the rural areas by national level financial institutions may be partly attributed to fixed, administered and low lending rates which reduce the incentives for banks to tend to agriculture and rural areas, and has resulted in less banking facilities and services available to - 54 - the rural populace. While the RFIs have grown at an impressive pace, they account only for 2.6% of total bank credits, 15% of rural credits and 2% of total bank deposits. They deliver a limited range of financial services consisting primarily of short-term, nonagricultural, and collateralized credit and are unable, at their present stage of development, to provide term finance. User costs of credit are high (up to 194% p.a.), reflecting the high operating costs of RFIs related to the small size of their transactions and high default risks of their clientele. The large number of village and paddy banks (BKDs, BKPDs, LDs) and KUDs need ca?ital, expertise, and economies of scale to develop into viable financial intermediaries. In addition, the financial position and resource base of BBD and BRI, the primary lenders for agriculture, are fragile and untenable without continuous GOI support as interest income reflecting "below market" interest rate ceilings has been insufficient to cover lending costs and realistic bad debt write-offs. 6.4 Although saving deposits have expanded remarkably, rural deposits still constitute a small percentage of total deposits. The availability of cheap liquidity credits from BI has reduced the incentives of financial insti- tutions to seek out savings resources from the rural areas and inhibited efforts to build a self-sustaining resource base. Rural savers have a limited choice of savings instruments and the institutional capacity for mobilizing savings, especially at the grassroots level, needs to be strengthened. 6.5 Despite the laudable objectives of increasing the supply of credit to priority groups and sectors, the credit insurance policy has in fact led to lax credit collections and consequently higher defaults. The claims against ASKRINDO have mounted over the years and have led to financial losses of the credit insurance agency over the last three years. The continued dominance of state financial institutions and the protected market of a fixed number of private banks have produced a noncompetitive environment which is not condu- cive to the growth and development of the financial sector. Finally, the costs of credit subsidies to the financial institutions, which represent about 19% of the 1986 agricultural development budget (excluding fertilizer subsidy), and accrue mainly to public enterprises and large borrowers, have to be weighed against actual benefits, especially within the context of GOI's scarce budgetary resources. 6.6 In the next section, suggestions are made to improve the rural financial system and thus enhance the availability and efficient use of credit in agriculture and the rural economy. B. Recommendations To Promote the Development of Financial Institutions 6.7 Gradually increase interest rate on BI liquidity credits for agricultural and rural credit programs. To encourage financial institutions to enter the rural credit market and mobilize savings, it is necessary to gradually limit their access to cheap liquidity credits by gradually increasing the BI rate on liquidity credits (say, 2% next year) to the marginal rate paid on savings deposits over a specific time period (e.g., three or four years). Although lending rates would increase to reflect the - 55 - increase in the banks' cost of funds, the gradual increase would allow the economy to adjust to higher lending rates and minimize any negative impact on credit demand and investment. As the supply of credit resources increases with the development of rural financial services and opportunities for greater financial savings, average lending rates should decline. The competition in the credit and savings markets would also promote improved efficiency among institutions and lower costs of intermediation. 6.8 Gradually raise onlending rates on agricultural and rural program loans (e.g. 2% per year). The interest rate increase would enhance the efficient allocation of resources and improve the financial viability and suste.nability of agricultural and rural credit programs and financial institutions. It would encourage banks to lend to agricultural and rural enterprises and small borrowers thereby increasing the supply of credit to these sectors and broadening the access to credit of low income groups. GOI should consider the introduction of variable rate lending especially for the provision of term credit to agricultural investments with long gestation period, such as tree crop projects. By setting limits on the interest rate change per year And/or allowing flexible maturity periods, the borrower can be protected from excessive interest rate fluctuations. 6.9 Liberalize bank branching policy and remove the barriers to entry of private firms into the financial sector. To promote competition among financial institutions, increase efficiency of the system and achieve denser penetration of the rural credit market by financial institutions, the branch- ing requirements for banks should be liberalized and the policy for the estab- lishment of new banks should be reconsidered. Based on a simple feasibility study of potential deposits, lending and profitability, the banks should be allowed to branch out to any area which they consider is financially viable. Bank Bumi Daya should be encouraged to expand its operation and establish a village network like BRI's aimed at better utilization of the existing manpower and financial resources. The use of sub-branches, agency operations, village posts and mobile units for smaller towns and villages should be encouraged by not imposing any requirements on their establishment and actively promoting their use by MOF and DI. This would expand the outreach of banks to the rural areas, strengthen their institutional capacity for mobilizing savings and provide a greater range of financial services to the rural populace at lower costs. GOI should consider opening the banking industry to new firms which would infuse additional capital into the sector, promote a competitive environment and enhance the efficiency and effectiveness of the financial sector. Safeguards from fraud and mismanagement can be installed by strengthening prudential regulation, requiring fidelity insurance as licensing requirement for all public dleposit taking institutions, and introducing deposit insurance. 6.10 Strengthen the small RFIs by providing technical assistance from GOI, improving supervision, and allowing them to mobilize voluntary savings. Training and incentive schemes will need to be developed to upgrade the staff banking skills of the small RFIs. The RPIs should be encouraged to establish linkages with the financial institutions at the national level through agency, subcontracting and/or co-financing arrangements which would expand the network and outreach of large banks and provide capital, expertise, and - 56 - expanded and diversified markets to RFIs. This will improve the access of small borrowers and women to financial services. The financial operations of KUDs should concentrate on their savings and loan associations and their management, accounting and financial systems strengthened through technical assistance. With the inherent conflict of interest between BRI's own unit desa operations and the secondary banks, BI should take over the direct supervision of the secondary banks or delegate the respornsibility to another financial entity. The improvement of BI's supervisory capacity needs priority attention. To Improve Credit Delivery Mechanisms 6.11 Improve design of future credit programs by: (a) adoption of market rates for the institution's cost of funds and onlending rates; (b) simplifying lending procedures; (c) establishing lending criteria on the basis of character, savings and repayment record rather than on collateral; (d) full accountability of credit risks by lending institutions, (e) providing finan- cial incentives for borrowers for prompt loan repayment and imposing sanctions on defaulting borrowers; and (f) introducing a staff bonus system based on efficient and profitable credit operations. Credit programs should as far as possible provide general nontargeted rural credit and take into account the multi-enterprise nature of the farm household. If there is a demonstrated need for project financing, e.g., for tree crops, the banks should perform a financial intermediary role and assume a major part of the credit risks. In addition, the costs of funds for the banks and onlending rates should be variable and at market rates. Subsidies, if found necessary and desirable, should be severely limited, transparent, temporary in duration, and sharply focused on target groups. If the amount of subsidies envisaged is large and potentially highly distortionary, GOI should consider using an alternative means of extending subsidies outside of the credit system such as grant and cess tax system or channeling and recovering funds through project implementing units. 6.12 To rationalize existing special credit programs, GOI should consider the establishment of a National Development Loan Fund (NDLF) which can consolidate all agricultural and rural credit programs under its umbrella and be capitalized by the outstanding liquidity credits of BI for agricultural and rural program loans and any outstanding agricultural and rural credits directly extended by GOI. Since BI liquidity credits are financed from increases in the money supply and government deposits, separating NDLF accounts from the BI balance sheet would isolate the growth of credit from the expansion of the monetary base and reduce the possible inflationary impact of credit expansion. BI should manage and control NDLF but additional resources for NDLF should come from national budget allocation to ensure the transpa- rency of credit assistance from GOI. The amount and source of financing credit subsidies need to be transparent so that policymakers, recipients and financiers can assess the benefits and costs of the programs. NDLF should be a financially viable and sustainable institution capable of covering its operating costs and earning positive financial margins on the resources it mobilizes. It could be authorized to borrow from international organizations and issue bonds or other debt instruments to sustain its credit operations. In structuring the accounts of dDLF, the outstanding amount of past special - 57 - credit programs and BI liquidity credit in arrears should be clearly identified and placed in a separate account. A debt write off policy for the unrecoverable debts should be formulated and carried out. The terms and conditions of the ongoing program credits of BI would be unchanged upon transfer to NDLF but the interest rate on NDLF loans should gradually increase over time to the marginal rate paid on savings deposits. Future credit programs should be along the lines outlined in previous paragraph. 6.13 The details involved in establishing NDLF including organizational and staff arrangement, transferring BI liquidity loans to the fund, and consolidating the present special credit programs need further study and preparation. The implementation would involve gradual phasing out of special credit schemes, mandated interest rates and credit subsidies over a planned time period. The objectives would be to wean the financial institutions from BI credit support except for emergency purposes, to price NDLF funds so that savings deposits would be a competitive source of funds, and to provide flexibility to banks so they can design their own credit programs to suit their needs as well as those of the borrowers. 6.14 In the short run, increase the premium rates and decrease the -overage of loan insurance for BI/NDLF loans to account for the high risks of default and improve the financial position of the credit insurance agency. However, the effectiveness and sustainability of the credit insurance policy should be assessed and the justification for its existence, reviewed, in the light of observed adverse consequences of the policy on loan collections. Loan insurance coverage should not be provided for future credit programs to encourage sound loan management and stringent loan recovery by banks. 6.15 Review and possibly restructure the terms and conditions for the existing tree crop credit schemes to provide financial incentives for banks to convert and collect the loans. The lesident Staff in Indonesia (RSI) is currently intensifying its supervision efforts to accelerate loan conversion by assisting GOI in the preparation of debt amortization schedules, and an action program for upgrading substandard plantings. GOI needs to reexamine the financial spreads provided to banks which are, at present, inadequate to cover their lending and collection costs. To Accelerate Savings Mobilization 6.16 Continue Policy for Positive Real Interest Rate. As shown earlier, a combination of low inflation and high positive interest rates over the last few years have helped increase financial savings in Indonesia and should con- tinue to ensure adequate savings mobilization in the future. 6.1i Remove Regulatory Restraihts. The remaining interest rate controls on TABANAS, TASKA and government time deposits of 24 months and the three month maximum maturity on time deposits of secondary banks should be lifted to allow the development of longer term savings instruments. The approval pro- cess for the establishment of pensions and superannuation funds should be facilitated to encourage their expansion. - 58 - 6.18 Promote Rural Savings and Financial Services Through the National Savings Drive Committee. The emphasis of the National Savings Drive Committee has been almost exclusively on TABANAS/TASKA accounts. The responsibilities of the Committee should be expanded to promote a wider choice of savings instruments, to assi4t banks and RFIs in developing savings campaigns, and to improve the data base and statistical reporting on rural financial savings. 6.19 Allow for Better Competition Among Savings Instruments. The present structure of the TABANAS/TASKA saving scheme allows the handling banks to compete for deposits only on the basis of customer service. While they are free to introduce their own proprietary savings schemes, such private schemes would not have the unlimited Bi guarantee provided on TABANAS/TASKA accounts. While in the long run TABANAS/TASKA might be replaced by other market related risk free assets (such as government securities), better financial services would develop if banks were allowed to develop their own savings schemes and to decide their own conditions for withdrawal and interest rates. Under such circumstances, BI might charge for its guarantee of these deposits. 6.20 Increase the Range of Financial Instruments Available in Rural Areas, Particularly Ones with Longer Maturities. At present there are a wide range of financial assets issued by formal and informal financial institu- tions, but in practice these instruments are similar and short-term. Longer- term assets such as government agency debt securities, life insurance policies, contractual goal savings plans, and perhaps equities should be available to provide rural savers/investors with more choice and thus increase the financial assets component of their savings. Local savings plans could be devised to suit specific local needs and urban savings could be tapped for rural development through long term bond issues. 6.21 Encourage Door-to-Door Financial Services. Door-to-door financial services such as deposit taking and loan collection are offered by some finan- cial institutions even at the village level, but this is the exception rather than the rule. While there are advantages to a simple payment and receiving operation, the cost benefit ratio might be improved by offering other finan- cial products through the same network which could also justify the expansion of these programs into unserviced areas. 6.22 Introduce Institutional Incentives for Savings Mobilization. The present banking system provides few direct incentives for either the institu- tions or their staff to expand their savings mobilization efforts. The BRI unit desa gives incentives to the staff for successful deposit raising and lending activities. Other institutions should consider introducing similar reward measures. To Continue and Expand Women's Participation in the Rural Financial Sector 6.23 Strengthen and support financial institutions and lending programs that use alternatives to collateral such as community/social pressure through group mechanisms or personal references, savings records, repayment records for small loans, nontraditional collateral--household items, jewelry, etc. This would be undertaken through technical assistance, improved linkages with - 59 - and supervision by larger financial institutions and encouraging nonbanks (e.g., BKKs, KURKs, etc.) to mobilize voluntary savings. 6.24 Restrict targeting credit programs or financial institutions by sex or by client's economic activity. Specifically, programs that single out agriculture or particular crops for lending are likely to have limited participetion of women as borrowers. There are other steps which could be taken in specific projects to improve womeni's access to savings and credit, such as the use of female staff in local savings collection efforts (as is done by the Bank Dagang Bali), marketing or channeling information on banking services directly to women, training existing extension staff and loan officers to assist women clients. 6.25 On loan procedures, standardize cosigning requirements within insti- tutions such that requirements are applied uniformly to male and female bor- rowers; relax the regulation allowing only one loan per family for cases where separate economic activities are carried on by different individuals in the family and where those individuals meet creditworthiness criteria; and sim- plify and streamline the paperwork required of borrowers so that women, who tend to have lower educational levels, will not be discouraged from applying for loans. To Strengthen Research Base 6.26 In order to assist the policymakers on rural financial sector issues, establish a unit within BI, NDLF, or MOF with the responsibility and capability to do policy research on rural finance issues, to monitor and evaluate credit programs and to provide technical assistance especially to small RFIs on the design and implementation of financiral innovations (such as, integrated rural financing schemes, group lending, etc.), credit programs and marketing schemes for loan and savings instruments. The proposed unit should be closely coordinated with the planned research component of the USAID- supported Financial Institutions Development (FID) project that is spearheaded by the Directorate of Financial Institutions of MOF. Under this project, a preliminary research agenda was outlined in September 1986 which could be a starting point. In addition, studies on the issues surrounding the participa- tion of women in the rural financial market (para. 6.28) and the effectiveness and efficiency of credit insurance policy involving ASKRINDO and PERUM PKK need priority attention. 6.27 Collect sex-disaggregated information on borrowers and savers on a regular or sample survey basis; number of borrowers, and amount of loans disbursed and outstanding and arrears by sex of borrowers; number of savers and amount of savings by sex of saver, in order to establish a data base for evaluating the participation of women in rural finance. 6.28 Conduct research on the folLowing questions: (i) fungibility of credit, household allocation of resources, and implications of intrahousehold allocation of credit for women; (ii) women's participation in the informal rural financial sector; (iii) borrower's transactions costs for men and women and their impact on the demand for credit by female and male borrowers; etc. - 60 - World Bank Support to Rural Financial Reforms 6.29 The Bank could assist GOI in formulating a detailed action plan for carrying out the reforms outlined above. Based on an agreement for adoption and scheduled implementation of the action plan, the Bank could possibly provide BI/VDLF a line of credit to finance agricultural and rural enterprises and future Bank projects in different subsectors (e.g., secondary food crops, livestock, forestry, agroprocessing facilities, etc). The Bank's contribution would increase the weighted cost of NDLF funds and help make savings more competitive. Through its support, the Bank would assist in mobilizing finan- cial resources and providing capital to rural and agricultural development. The Bank could also assist in the institutional development of RFIs and the strengthening of the research base for policymaking in rural finance through technical assistance programs. In contributing to rural financial policy reform, the Bank would help GOI strengthen the rural financial system and improve the availability and efficient use of credit in agriculture and rural economy.