Chapter 3: Arrangements for direct support

As noted in chapter 2, it is now generally accepted that microfinance institutions (MFIs) need to become self-sufficient if they are to expand their outreach to cover a significant proportion of poor people in Asia. Public resources, from governments and donor agencies, are strictly limited and MFIs cannot rely on ever increasing amounts of public resources to expand their outreach. Increasing self-sufficiency enables MFIs to reach more poor people without a commensurate increase in the amount of public resources devoted to microfinance. Indeed, where MFIs are operating at full financial self-sufficiency, direct support by governments and donor agencies is no longer necessary. Fully self-sufficient programs are able to meet all administrative costs, loan losses and financing costs from operating income, and do not need any grants or concessional loans from government or donor agencies. In the long term, the ideal situation would be for governments and donor agencies to withdraw all direct support from microfinance, and redirect such expenditures to other poverty reduction programs that are not able to operate on a commercial basis.

Nevertheless, such a situation is a long way off. In all countries in this study except for Pakistan, governments and/or donor agencies have spent large sums of money to support the development of microfinance. This support has been critical in enabling MFIs to become established and to increase their outreach. It is no accident that in Pakistan, where there has been little direct support for microfinance, the outreach of MFIs is far less than in any other country in the study. And even in Pakistan, those MFIs that do exist have relied heavily on the support of various donor agencies. The experience of the countries in the study suggests that active support from government and/or donor agencies, including direct financial support, is a necessary catalyst to the establishment of a viable microfinance sector.

This is largely because achieving self-sufficiency takes a considerable period of time. The Guiding Principles for Selecting and Supporting Intermediaries (Committee of Donor Agencies 1995) establish indicative ‘timetables’ for the period over which MFIs should be able to achieve self-sufficiency. Based on some assumptions about the rate of expansion of outreach, they suggest that successful MFIs should be able to achieve operational self-sufficiency within three to seven years, and financial self-sufficiency within five to ten years. To set up their programs in the first place, and for a number of years thereafter, they need to be subsidised. Initially, they will require external funding for both operating expenses and loanable funds. Once they reach operational self-sufficiency, they will be able to meet operating expenses from their income, but loanable funds will still need to be subsidised until they attain full financial self-sufficiency. It should also be noted that, in practice, most MFIs have not been able to attain self-sufficiency in accordance with the timetables proposed in the Guiding Principles.

To reach self-sufficiency, it is generally necessary for an MFI to realise significant economies of scale. A larger institution is generally better able than a small one to develop streamlined and cost-effective systems in areas such as administration and staff training, and to achieve high ratios of clients per employee. To achieve the scale necessary for self-sufficiency within a reasonable time frame, it is necessary to go through an expansion phase relatively early in the life of the institution. This requires a considerable input of resources for capacity building and institutional development. In addition to meeting ordinary operational expenses, governments and/or donor agencies will need to meet the costs of scaling up operations if large, viable MFIs are to emerge.

For these reasons, direct financial support from governments and donor agencies is critical to the development of a microfinance sector. Governments and donor agencies may also support microfinance in other ways, such as providing technical assistance, requiring banks and other regulated financial institutions to engage in microfinance, or through pilot programs designed to demonstrate the viability of microfinance and thereby encourage the private and non-government sectors to become involved in this sector.

Direct support for microfinance can also be channelled through different types of institutions. This chapter looks at the advantages and disadvantages of providing support through a number of different types of institutions, and how support through each may be improved. Section 3.2 discusses support for individual MFIs, while section 3.3 considers the tendency for support to be channelled increasingly through second tier microfinance institutions. Section 3.4 looks at support to microfinance through the banking system. Such support may be provided in different ways, such as requiring banks to allocate some of their portfolio to microfinance, channelling funds through the banking system to final borrowers, or supporting linkages between banks and MFIs. Section 3.5 looks at programs operated directly by government agencies, while section 3.6 summarises the key findings.

As noted above, MFIs in all countries have received considerable support from the government and/or donor agencies. In the past, most of this has been in the form of financial assistance from the government or donor agency direct to an individual MFI.

In Bangladesh, the country with by far the greatest outreach of microfinance, MFIs have received large amounts of support. The Grameen Bank itself has received considerable support from the government, Bangladesh Bank and donor agencies. Following the success of the initial experimental project by the University of Chittagong, the Grameen Bank project was established in 1979 with support from Bangladesh Bank and later from nationalised commercial banks. By end-1995, Grameen Bank had received donor grants amounting to $84 million for its revolving loan fund. It has also continued to receive support from the government and government agencies. For instance, Bangladesh Bank provided a line of credit of $145 million in 1993–94 at an interest rate of 6 per cent, although less than half of this was drawn down. Grameen Bank has also issued bonds guaranteed by the government, and the nationalised commercial banks purchased around $360 million of such bonds in early 1995 with the encouragement of the government.

Many other MFIs in Bangladesh have also relied on donor funds, especially during the establishment phase, with around 250 MFIs currently receiving foreign donations for both operational costs and loan capital. Large inflows of subsidised donor capital have been instrumental in the proliferation and expansion of MFIs. Nevertheless, since the late 1980s and early 1990s, donor funds have become harder to come by.

This important role for government and donor agencies is mirrored in most other countries. Few if any specialist MFIs with significant outreach to poor clients have been established without the support of such funding agencies. In Indonesia, where the autonomous role of non-governmental organisations (NGOs) is less firmly established, most MFIs have received assistance in the context of particular government-initiated programs rather than support for independent operations. In the Philippines, the major MFIs have received support from a large number of government programs which operate through the non-government sector. Moreover, numerous official and private donor agencies have seen microfinance as an effective mechanism for poverty reduction. In India, the myriad of MFIs and NGOs involved in promoting self-help groups (SHGs) rely heavily on funding from various donor agencies to meet their administrative expenses.

The situation is similar in Sri Lanka and Nepal, where MFIs have been supported by a wide range of external donors as well as some government programs. In Malaysia, the government has provided extensive support to the dominant MFI, Amanah Ikhtiar Malaysia (AIM), while the smaller institutions have relied on support mainly from domestic private sources. Even in Pakistan where very few MFIs exist, all significant initiatives have received support from official and/or private donor agencies from abroad.

This support for individual MFIs has been critical in enabling them to establish in the first place, and subsequently to expand their outreach and increase their self-sufficiency. Without this support, the microfinance sector would not have developed to anything like its current size. Indeed, it is unlikely that there would be a microfinance sector at all, in the sense of specialist MFIs providing financial services to the poor in an innovative and commercially viable manner.

Nevertheless, it is probably fair to say that the development of microfinance and the success of leading MFIs owes more to the vision of their founders and management, than to the particular criteria used by government and donor agencies for supporting MFIs. Support has been largely ad hoc in nature, and has generally not been explicitly designed to maximise the twin objectives of outreach and sustainability in the most cost-effective manner.

The Guiding Principles for Selecting and Supporting Intermediaries agreed by major donor agencies in October 1995 provide a number of suggestions as to how support by donors and government can best be directed to maximising outreach and sustainability. Most importantly, governments and donor agencies should only support institutions which demonstrate a capacity to reach poor clients on a sustainable basis. The Guiding Principles suggest that such MFIs should have three clusters of characteristics, either in their current operations or through credible plans underpinned by concrete measures.

First, they should be strong institutions. They should have institutional cultures and structures that can support sustained service delivery to a significant and growing number of low-income clients. This includes sound governance, freedom from political interference, and strong business plans for expansion and sustainability. They should also have accurate management information systems that are actively used to make decisions, systems that manage small transactions efficiently, and meaningful, transparent financial reporting that conforms to international standards.

The second cluster relates to quality of services and outreach. MFIs seeking support should focus on the poor and have lending terms and conditions that meet the needs of their clients, such as quick, simple and convenient access to small loans. Where possible they should offer savings services. They should also demonstrate significant progress in expanding outreach.

Third, institutions should have sound financial performance. They should charge interest rates that are sufficient to cover the full costs of lending after a reasonable start-up period. They should maintain low arrears. They should have clear plans for steadily reducing dependence on subsidies and for attaining operational and financial self-sufficiency within a reasonable time period. And they should have plans for building a solid and growing funding base through mobilisation of commercial funds from depositors and/or the financial system.

The Guiding Principles also consider the types of support that donor agencies should provide. They argue that grants should only be given to cover operating losses during a clear, time-limited, start-up or expansion phase. On the other hand, support for institutional development is appropriate at all stages of an institution’s life, but should become more selective and specialised as institutions evolve. Support through soft loans for on-lending is appropriate for MFIs that meet performance standards. However, it is also important that MFIs seek loan capital from commercial sources.

The Guiding Principles also note that grants for equity can be of strategic importance in enabling MFIs to build a capital base. A sound equity position is important to generate investment income, build the loan portfolio, and leverage funds from commercial banks and capital markets. As such, they can be a catalyst and complement to domestic mobilisation of funds. Grants for equity can also help MFIs seeking to become formal financial institutions to meet minimum capital requirements.

In brief, if governments and donor agencies wish to ensure that their support for microfinance has maximum impact, they need to look very carefully at the institutions they support and ensure that such institutions meet a number of key criteria. While the major donor agencies have endorsed these principles, it is fair to say that they are not always followed in practice. It would be appropriate for all governments and donor agencies to ensure that their support for MFIs is consistent with these principles.

Most countries included in the study have some form of ‘second tier’ microfinance institution which channels funds from the government and/or donor agencies to individual MFIs. The largest and most successful second tier institution is the Palli Karma Sahayak Foundation (PKSF) in Bangladesh. PKSF was established as a private non-profit organisation in 1990, and has received a total of $17.9 million from the government of Bangladesh. It is governed by a general body of fifteen people, the majority of whom are from outside government. By December 1996 PKSF had disbursed loans of $31.6 million to 135 MFIs with a demonstrated capacity or potential capacity in microfinance, for on-lending to final borrowers. The repayment rate by MFIs on loans from PKSF was above 98 per cent. A total of 546,000 final borrowers had obtained loans through these facilities.

By all accounts, PKSF has operated very successfully. A recent World Bank review (1996) found that its operating policies and procedures were quite satisfactory, with sound and conservative credit approval policies, intensive credit supervision, and comprehensive external audit. The World Bank poverty alleviation microfinance project provides a line of credit of $109 million to PKSF for on-lending to MFIs, and a further $6 million for technical assistance for capacity and institution building for MFIs, PKSF and the Bangladesh Bank. It should be noted that PKSF only lends to MFIs for on-lending to final borrowers. It does not provide funds, whether in the form of grants or soft loans, for scaling-up operations, institutional development or increasing equity. It does, however, provide training and advisory services, and conducts research.

In India, there are no fewer than three second tier microfinance institutions which channel funds to specialist MFIs at concessional interest rates. The largest of these is Rashtriya Mahila Kosh (RMK), established by the government in 1993 with a corpus fund of $8.6 million. By March 1997 it had disbursed around $5.7 million to NGOs for on-lending. The Small Industries Development Bank of India (SIDBI) has also operated a microcredit scheme since 1994, and by March 1997 had provided loans of $2.8 million to 79 NGOs. Finally, the National Bank for Agriculture and Rural Development (NABARD) provides a small amount of revolving fund assistance to NGOs for on-lending to self-help groups (SHGs).

A major weakness in the programs of RMK and SIDBI is that their interest rate policies make it very difficult if not impossible for their partner MFIs to be self-sufficient. The interest rate charged by RMK to its partners is 8 per cent, and they are required to lend to SHGs at 12 per cent, a margin of only 4 percentage points. While MFIs can obtain some interest rate relief from RMK, this only enables them to increase the margin to 6.5 percentage points. SIDBI lends to MFIs at 9 per cent, and the MFIs are required to on-lend to SHGs at a rate not higher than 15 per cent, a mark-up of 6 percentage points. These margins are not enough to enable MFIs to cover their costs, and MFIs participating in the programs are only able to operate through grants from donor agencies and other sources, which meet a large proportion of their costs. This greatly inhibits their ability to expand, and reduces the scope for new MFIs to be established. On the other hand, one very positive feature of all three Indian programs is the availability of capacity building funds to enable MFIs to strengthen their capabilities and scale up their operations.

There is also a second tier microfinance institution in Sri Lanka. The National Development Trust Fund (NDTF), originally known as the Janasaviya Trust Fund (JTF), was established in 1991. It is governed by a 20-member board of trustees, selected by the government and comprising around half public sector and half private sector representatives. Its main function is to manage a credit fund for disbursing credit to partner organisations, but it also manages a human resources development fund for providing skills training to beneficiaries, a rural works fund, and a nutrition fund. A credit fund of $35 million was established with support from the World Bank, the government of the Federal Republic of Germany and the government of Sri Lanka. By March 1997, it had provided funds totalling $21.5 million to 157 partner organisations, including commercial banks, regional rural development banks, thrift and credit cooperative societies and NGOs. NDTF also provides assistance to partner organisations for institutional development.

NDTF has achieved a solid repayment rate of 90 to 95 per cent on its loans to partner organisations. However, it has experienced a number of difficulties, to the extent that the mid-term project review by the World Bank in May 1995 classified the project as a ‘problem’ project. Difficulties included inadequate processes for selecting partner organisations, inadequate monitoring and supervision of partner organisations, and concerns about the targeting of the program. These failures appear to reflect a number of factors. Some commentators stated that NDTF was under pressure to disburse funds too quickly, and that there were not enough sound MFIs available to use as partner organisations. It would also appear that NDTF did not have sufficient staff with the technical skills to implement the program properly. Finally, some commentators pointed to politicisation of the organisation. More recently, some measures have been implemented to try to overcome these problems, and there appears to have been a marked improvement in its performance.

In the Philippines, the People’s Credit and Finance Corporation (PCFC) was established in 1995 as a government finance company for lending to the poor. It is incorporated under the Corporations Act and regulated by the Securities and Exchange Commission. PCFC defines MFIs broadly, lending to NGOs, rural banks, cooperatives and other intermediaries as ‘conduits’ for on-lending to the poor. It aims to ensure that such intermediaries are replicable, self-sustaining and operationally viable. The National Credit Council envisages that the corporation should gradually replace many of the other lending programs operated by line agencies of government.

By December 1996 PCFC had lent a total of $10.8 million to 76 conduits, benefiting some 32,000 final borrowers. PCFC is also the executing agency for a $26.3 million program funded by a loan from the Asian Development Bank (ADB) and the International Fund for Agricultural Development (IFAD). These funds are to be provided exclusively for PCFC to support MFIs replicating the Grameen Bank approach, even though there is a wide variety of models being used in the Philippines. In mid-1997, PCFC was funding 58 conduit institutions from its other reserves, of which only 17 were Grameen replicators.

PCFC also loans funds to conduits for capacity building. These are soft loans and although they are limited to 20 per cent of total outstandings, they are a source of concern given the need for PCFC to show a profit in the lead-up to privatisation (the ADB is requiring PCFC to complete a privatisation plan by mid-1998, with a view to completing the process by end-1999). After privatisation, the government may need to accept greater responsibility for capacity-building requirements of MFIs.

In Thailand there are two institutions, the Government Savings Bank (GSB) and the Urban Community Development Office (UCDO), which lend to cooperatives and community organisations involved in microfinance. Unlike most of the NGOs and partner organisations which borrow from second tier institutions in other countries, the cooperatives and community organisations which borrow from GSB and UCDO are owned by their members and tend to be small. Most of them are not registered and operate on an informal basis. By February 1997 GSB had provided loans to some 1,360 organisations with a total membership of 188,000. UCDO had approved loans of $4.9 million to 103 organisations. As well as providing loans, both GSB and UCDO provide the organisations with technical support. The government is currently planning to establish the Community Organisation Development Institute (CODI) which will take over the overall policy responsibility for the two programs. But it is expected that, at least initially, GSB and UCDO will continue to administer the schemes.

Nepal also has a second tier institution for channelling funds to NGOs engaged in microfinance, but it is much smaller than comparable institutions in other countries. The government established the Rural Self-Reliance Fund (RSRF) in 1991 with a one-off contribution of $400,000. By January 1997, RSRF had disbursed around $300,000 to some 54 NGOs. Eligible NGOs are those that are registered with the government and permitted by the central bank to engage in limited banking.

There is currently no second tier microfinance institution in Pakistan, but the government is proposing to establish the Poverty Alleviation Fund, with support from the World Bank, to perform this role. In Malaysia, the Credit Guarantee Corporation (CGC) has performed the role of a second tier microfinance institution to a minor extent as an adjunct to its primary responsibilities. The small number of MFIs and more limited demand for microfinance in that country reduce the need for a specialist second tier institution.

Interestingly, there is no institution in Indonesia that performs the role of a second tier microfinance institution, despite the prominence of microfinance. This reflects greater emphasis in that country on a financial systems development approach to microfinance, with microfinance flowing as a result of the development of effective formal financial institutions. NGOs play a relatively minor role in the Indonesian situation, although government-supported community groups are important players.

Where they operate well, second tier microfinance institutions are a very effective means for governments and donor agencies to channel support to MFIs. Such institutions have a number of advantages over the traditional approach of providing support to individual MFIs on a largely ad hoc basis. As noted above, if governments and donor agencies are to maximise the effectiveness of their support for microfinance, they need to look very carefully at the MFIs they support, and ensure that such institutions meet a number of key criteria. This requires a detailed appraisal of the institution, and regular monitoring against the various criteria. Clearly, it is much more efficient for one institution to conduct this analysis than for different government and donor agencies to do their own appraisals and evaluations. Moreover, if all the support for a particular MFI can be channelled through one agency, it is much easier to ensure that it is complementary and directed to the same objectives. This can avoid the situation where different government and donor agencies have different and possibly inconsistent objectives, as is often the case at present.

An effective second tier institution can also ensure that all MFIs operate on a level playing field, with support available to all MFIs meeting predetermined standards on the same terms and conditions. Moreover, second tier institutions can ensure that the MFIs that borrow from them face common performance and reporting standards, rather than different standards imposed by different donor agencies, or no standards at all as is often the case at present. By setting appropriate standards, second tier institutions can contribute significantly to a more efficient and professional microfinance sector. These issues are discussed in more detail in chapter 4.

Second tier institutions also offer advantages for MFIs. They are a one-stop funding source, and can avoid the delays that often arise when relying on other funding sources. They can also facilitate training to MFIs, and enable exchange of best practice between them.

For all of these reasons, it would be desirable for governments and donor agencies to channel their support for microfinance, as far as possible, through efficient and well-managed second tier microfinance institutions. In countries with more than one second tier institution, it is important to ensure that the different institutions adopt consistent standards and policies.

At the same time, there is evidence both from Asia and from other regions that second tier institutions do not always operate on a sound basis. For instance, Fruman and Goldberg (1997) comment that second tier institutions only work well when there are high standards, clear objectives and managerial authority. Where these conditions are not met, they are fraught with administrative, political and technical difficulties. They also point out that second tier institutions are most effective when there is already a group of strong retail MFIs, such as in Bangladesh. In countries that do not have a well-developed microfinance sector, it may be very difficult to establish an effective second tier institution.

If second tier institutions are to play an effective role, it is clear that there are a number of critical issues that need to be addressed. First and most important, it is important that second tier institutions establish and enforce appropriate performance and reporting standards for the MFIs that they fund. In some countries second tier institutions have not been sufficiently rigorous in establishing and enforcing such standards. Insufficient attention to the performance of partner MFIs represents a missed opportunity to improve the outreach and sustainability of the microfinance sector. In the extreme it can also reduce the ability of the partner organisations to repay their loans and undermine the viability of the second tier institution itself. However, it should be noted that even the institutions discussed above which are not performing as well have maintained repayment rates of at least 90 per cent on loans to their partner organisations, while the best institutions have maintained repayment rates of close to 100 per cent.

The standards to be enforced should be based primarily on outreach and financial criteria. Second tier institutions should not restrict support to any particular model or models of microfinance, as this may inhibit the development of new approaches to microfinance capable of satisfying appropriate financial norms. There is scope for a diversity of models and institutional forms, including financial intermediaries, social intermediaries and mixed financial/social intermediaries.

Pressure to expand

Second, pressure to expand too rapidly may lead to problems. While in Bangladesh there is a large number of well-managed MFIs able to meet fairly rigorous performance and reporting standards, this is not the case in most other countries. Pressure to meet unrealistic disbursement targets may cause institutions to lower their standards and lend to some ineffective MFIs, as in the case of NDTF in Sri Lanka. In most countries, it will take time for large numbers of MFIs to reach the standards required by a prudent second tier institution. This is particularly relevant in countries such as Pakistan, where there are currently very few MFIs operating on a sound basis.

Restrictive interest rate policies

Third, second tier institutions should avoid restrictive interest rate policies. In India, the major second tier institutions impose ceilings on the interest rates that their partner MFIs can charge borrowers, at levels which make it difficult if not impossible for the MFIs to achieve self-sufficiency. This means that they have to rely on donor grants to fund their operations, and severely restricts their ability to expand their outreach. In most cases, there would not appear to be any need for second tier institutions to impose interest rate ceilings on their partners. If they do impose ceilings, they should ensure that the margins permitted are sufficient to enable MFIs to meet all of the costs of their operations.

Support for capacity building

Fourth, some second tier institutions, most notably PKSF in Bangladesh, only lend to MFIs for on-lending to final borrowers. PKSF does not provide funds, whether in the form of grants or soft loans, for scaling-up operations, institutional development or increasing equity. In many countries, funds for these activities are at least as important in enabling MFIs to increase their outreach and sustainability as funds for on-lending. It would therefore be appropriate for second tier institutions to fund scaling-up, institutional development and equity as well as lending. Appropriate criteria would need to be established for accessing support for each of these activities.

It should be noted that while PCFC in the Philippines makes soft loans to MFIs to finance capacity building, the need to become more profitable in preparation for privatisation imposes a strain on its ability to undertake capacity building. Where a second tier institution is privatised, government should accept some responsibility to finance capacity building of MFIs.

Importance of independence

Fifth, it is critical that second tier institutions operate as independent entities and are not politicised. Some commentators suggested that the performance of NDTF in Sri Lanka had been affected by politicisation, and this also appears to have been a factor affecting institutions in some other countries. By contrast, one of the strengths of PKSF in Bangladesh is that it appears to have been able to operate independently with little or no political interference. The extent to which institutions are able to operate independently partly reflects the overall political culture of the country, and in some cases it may be extremely difficult to avoid political interference completely. Nevertheless, possible techniques for minimising politicisation include spelling out the objectives of the institution as simply and clearly as possible, making the board of directors directly responsible for achieving these objectives, limiting the number of public sector representatives on the board, appointing board members for fixed terms, and enabling various private and non-government bodies to appoint representatives to the board directly.

3.4 Support through the banking system

In her study of the role of commercial banks in microfinance, Goodwin-Groen (forthcoming) notes that banks can engage in microfinance in a wide variety of ways. She presents a taxonomy of twelve different ways, nine of which involve a government mandate or government support. While this section considers government support through the banking system in a much more simplified framework, it is important to recognise that there is considerable overlap and blurring between the categories discussed below.

First, this section looks at directed credit schemes where banks lend to fulfil a specific government mandate. While such schemes are considered together in this section, they may take a variety of different forms. In some cases banks may be required to lend at subsidised rates, while in other cases they may lend at commercial rates. Moreover, they may lend directly to individual borrowers, or they may make use of non-governmental organisations (NGOs) or self-help groups (SHGs) as financial or social intermediaries.

Second, it considers government programs where subsidised government money is channelled through the banking system. Again, banks can lend to individual borrowers under such programs, or they can involve NGOs and SHGs in the program. Finally, it goes on to look at government programs to encourage linkages between banks and specialised MFIs. While such programs may be promoted by government, their objective is to facilitate such linkages on a purely commercial basis, without involvement from government.

Seven of the nine countries included in the study impose directed credit requirements, whereby banks are required to lend a certain proportion of their loan portfolio to particular sectors. Only Bangladesh and Sri Lanka have abolished all such sectoral lending requirements.

In Thailand and the Philippines, the requirements relate to broad sectoral allocations only. Commercial banks in Thailand are required to lend 20 per cent of their total deposits from the previous year to the rural sector. In the Philippines, banking institutions are required to allocate at least 25 per cent of loanable funds to certain specified sectors, including agriculture. In these countries there is no requirement on banks to lend to the poor. Similarly in Malaysia, Bank Negara has three lending guidelines requiring banks to lend to priority sectors of the economy, but none of these directly targets poor borrowers.

In Indonesia since 1990 the government has instructed commercial banks to allocate 20 per cent of their advances to ‘small’ business. Small-scale business credits (KUK) consist of investment and working capital loans extended to small-scale entrepreneurs, with a maximum overall credit ceiling of Rp350 million ($150,000). While microcredit lending is included in this category, the relatively high ceiling on assets below which the entrepreneur is classified as ‘small’ and the generous ceiling on individual loans make it clear that this government requirement is concerned only incidentally with stimulating support for microfinance and has very little direct relevance for poverty alleviation. Its primary purpose is to satisfy pressures from indigenous Indonesian entrepreneurs who feel disadvantaged relative to ethnic Chinese business interests which are over-represented in medium- to large-scale enterprise. It is relevant to our concerns to note that commercial banks which have difficulty in reaching the 20 per cent target are entitled to count for that purpose credit extended to the small rural banks (BPRs) which deal primarily with small business and microenterprise. These banks are discussed in chapter 5.

The remaining countries make more attempt to target poor borrowers through directed credit. India has perhaps the most elaborate requirements. All commercial banks and regional rural banks are required to lend 40 per cent of net bank credit to priority sectors. In broad terms, at least 18 per cent of lending must be to agriculture. A further 10 per cent must be to weaker sections, such as small and marginal farmers, rural artisans and agricultural labourers. The remaining 12 per cent is unmarked, and can be for either of the above categories or for small-scale industry. Of the lending to small-scale industry, 40 per cent must be for tiny industry. Most lending to priority sectors is eligible for refinancing from the National Bank for Agriculture and Rural Development (NABARD). The repayment rate on priority sector lending is generally low, in the range of 60 to 65 per cent.

In Pakistan, the State Bank administers a scheme requiring some banks to lend a certain proportion of their loans to agriculture, and within that to small farmers. The scheme applies to the three longstanding state-owned commercial banks, the two privatised commercial banks, the Agricultural Development Bank and the cooperative banks. There is a complex procedure for determining the allocations for individual banks, but in the case of commercial banks it is understood that the requirements mean that around 2.5 per cent of the value of loans outstanding must be allocated to small farmers. However, it is unlikely that many poor farmers benefit from this allocation. It is understood that around 70 per cent of farmers fall within the definition of small farmers and, even so, some large farmers have managed to find ways of obtaining loans. Repayment rates have increased in recent years, but are still only in the 70 to 75 per cent range.

In Nepal, commercial banks are required to set aside 12 per cent of their loan portfolio for priority sector lending under the Intensive Banking Program (IBP). Within this, 3 per cent of their loan portfolio must be devoted to the ‘hard core poor’. However, there is no definition for the ‘hard core poor’, and loans up to Rs15,000 ($270) to individual borrowers are regarded as meeting this requirement. The recovery rate under the IBP was estimated at 46 per cent in 1995–96, reflecting weak supervision, untrained bank personnel and lack of incentives. There are also suggestions that most of the resources have gone to the non-poor.

Other things being equal, directed credit schemes which target the poor cause banks to lend more to poor borrowers than they otherwise would. Nevertheless, such schemes tend to be an inefficient means of reaching the poor. Many of the considerations discussed below relating to the implementation of government microfinance programs through commercial banks also apply here, with most participating banks not having the commitment to make the far-reaching changes that are necessary to operate successful microfinance programs. As such, their priority sector lending to the poor tends to be poorly targeted and to suffer from low repayment rates. While such schemes do not necessarily impose a cost to the budget (although where refinancing is involved they may), they reduce the efficiency of the financial system, and therefore impose resource costs on the economy as a whole. Again, experience with such schemes suggests that there are much more efficient ways of encouraging microfinance than through directed credit requirements, and that a more effective way for governments to encourage regulated banks to become involved in microfinance is to ensure an appropriate regulatory and prudential framework.

In some countries, the government operates specific microfinance programs in which funds are channelled through the banks. By far the largest of these is the Integrated Rural Development Programme (IRDP) in India. This program involves the provision of credit to the poor by banks, with a cash subsidy provided to borrowers by the government. Beneficiaries are selected by local village assemblies on the basis of a survey by local officials of households below the poverty line, and are then screened by the banks. It is designed to assist the poorest of the poor, with reservations for scheduled castes and tribes (50 per cent), women (40 per cent) and the physically handicapped (3 per cent).

There is no specific interest rate set under the program, but as with all small loans by commercial banks, there is a ceiling of 12 per cent. The maximum loan size under the IRDP is Rs25,000 ($690), with the average loan size around Rs15,000 ($420). Lending by banks under the IRDP is eligible for refinancing from NABARD. As at August 1996, over 49 million loans had been advanced under the program since it began in 1978.

Numerous evaluation studies have shown that the IRDP has not met its objectives. For instance, the Interim Report of the Expert Committee on the IRDP, constituted by the Reserve Bank of India (1995), found the following shortcomings:

(1) There was inappropriate identification of borrowers, with the subsidies often being captured by better-off households.

(2) The provision of the subsidy at the beginning of the loan had caused leakages and malpractices, such as encouraging beneficiaries to dispose of the assets purchased with the loan and simply pocket the subsidy.

(3) Repayment rates were very low, with the cumulative repayment rate having declined to 30.9 per cent by end-June 1993.

Recently, some changes have been made to the operation of the scheme, in response to this report. It remains to be seen to what extent these changes will lead to an improvement in the performance of the program.

Indonesia has a number of programs in which credit is channelled through the banking system, with varying degrees of effectiveness and varying degrees of relevance to poverty alleviation. Possibly the most successful, the P4K project (income-generating project for marginal farmers and landless) is conducted by the Ministry of Agriculture and has a mass character. It has employed Bank Rakyat Indonesia’s field staff, community facilitators and ‘small farmer groups’ to develop a savings and credit scheme which the International Fund for Agricultural Development (IFAD) is said to regard as the best it has supported in Asia. Bank Rakyat Indonesia covers its costs, and final borrowers (who number some 480,000) pay market-related interest rates.

The central bank, Bank Indonesia, has a targeted microcredit project which is based on a poverty line, with one-third of borrowers ‘poor’ and the remainder ‘nearly poor’. Financial development at the grassroots level is at least as important an objective as poverty alleviation for this project. The rural banks (BPRs), other small financial institutions and NGOs are involved in the disbursement process and interest rates are set to cover costs. Bank Indonesia also channels ‘liquidity credits’ to cooperatives, with funds from commercial banks, in the ratio of 3 to 1. The central bank credits are subsidised and provide generous margins to support the cooperatives, with interest rates to final borrowers (few of whom are below the poverty line) below market rates. Another Bank Indonesia initiative is the PHBK (program linking banks with self-help groups), discussed below in the section on linkages between banks and specialist MFIs.

A much smaller scheme operates in Malaysia and Sri Lanka. In Malaysia, commercial banks serve as a conduit for a number of loan schemes operated by CGC. The main scheme is the loan fund for hawkers and petty traders. The scheme is funded through a one-off soft loan from the government to CGC of RM120 million ($48 million). This money is on-lent to commercial banks and finance companies free of interest, and banks lend a maximum of RM10,000 ($4,000) to final borrowers at an interest rate of 4 per cent. While the scheme has achieved high repayment rates, it was reported that it does not generally reach poor borrowers.

In Sri Lanka, the government has recently introduced the Samurdhi Development Credit Scheme, targeted at welfare beneficiaries. The scheme is implemented by the Ministry of Youth Affairs, Sports and Rural Development in collaboration with two state commercial banks. The government has provided Rs500 million ($9.1 million) to the two participating banks for 1997. Borrowers are selected by village task forces, and receive loans of between Rs2,500 ($45) and Rs10,000 ($180) at an interest rate of 10 per cent. Because of the newness of the scheme, no data on its performance are available at this stage.

Thailand

In Thailand, one of the largest microfinance programs is the credit program for rural development, administered by the Government Savings Bank (GSB) and referred to in section 3.3. The program provides loans to savings and credit organisations in rural areas at an interest rate of 7 per cent. To date 1,360 organisations with a total membership of 188,000 have obtained loans, and repayment rates are reported to be high. The program appears to have been more successful than most other microfinance programs channelled through the banking system. This is very likely because GSB operates through linkages with savings and credit organisations, and does not lend to final borrowers in its own right.

Nepal

Nepal also has programs channelled through the banking system. The Production Credit for Rural Women (PCRW) scheme was introduced with assistance from IFAD, and channels funds through three public sector banks direct to self-help groups. It has been reasonably successful with a cumulative repayment rate of around 82 per cent, but delivery costs are quite high. The Micro Credit Project for Women (MCPW) is supported by the ADB. It is channelled through two state commercial banks, and involves NGOs as financial intermediaries. Repayment is reported to be close to 100 per cent, very likely reflecting the use of NGOs.

Pakistan

In Pakistan the First Women Bank, a state commercial bank, administers a small fund of Rs48 million ($1.4 million) provided by the Department of Women’s Development. From this fund, the bank provides small loans at a mark-up of 12 per cent to poor women. Loans are generally provided to individual borrowers, although in some cases borrowers organise themselves into groups and provide group guarantees. However, 22 NGOs are associated with the program as social intermediaries, in identifying and training borrowers. The bank has also lent to five NGOs as financial intermediaries for on-lending to final borrowers. To date, the program has provided loans to around 5,040 women and achieved a cumulative repayment rate of around 97 per cent. This successful performance may reflect the fact that First Women Bank is a relatively small bank with a particular focus on women, and the use of NGOs as financial and social intermediaries.

Conclusions

Based on experience in Asia and elsewhere, it is fair to say that the implementation of government microfinance programs through commercial banks has generally not been successful, at least where commercial banks have lent directly to final borrowers without using NGOs as financial or social intermediaries. While it is clear that banks can establish successful microfinance programs, it is equally clear that this takes a great deal of effort and commitment. In the past at least, most banks have not had the commitment to make the thoroughgoing changes in organisational design and financial technologies that are necessary to operate successful microfinance programs.

Moreover, participation in government microfinance programs is unlikely to induce them to make these changes. Such schemes are generally implemented through very large, often state-owned commercial banks, and account for only a small proportion of their total loans outstanding. They are therefore seen as marginal to their overall operations. Often the loanable funds come from a specific government allocation or are guaranteed by the government, reducing the incentive for intensive follow-up to ensure high repayment rates. There is little incentive for the banks to ensure that such programs reach genuinely poor borrowers. And interest rates are generally kept too low for commercial banks to operate such schemes on a sustainable basis, even if they had the commitment to do so. Hence, it is not surprising that such schemes tend to be poorly targeted and to suffer from low repayment rates.

Such schemes may also ‘crowd out’ other microfinance programs, not only by absorbing scarce public funds, but also by making it more difficult for other programs to charge market interest rates and by eroding repayment discipline. Where public funds are available for microfinance, it would appear much more appropriate to channel them through second tier microfinance institutions and specialist MFIs, rather than through the banking system. There is also no reason why small banks with a commitment to microfinance and which meet the required performance and reporting standards cannot borrow from second tier institutions. In the Philippines a number of rural banks have borrowed from PCFC, and in Sri Lanka NDTF has lent to a number of regional rural development banks.

An apparent exception to this analysis is in Indonesia, where some microfinance programs channelled through banks, in particular Bank Rakyat Indonesia and the BPRs, appear to have been successful. This is because these banks have a genuine commitment to outreach, and the success of their efforts in this regard brings them into the market for microfinance. Their appropriate organisation, products and procedures contribute to their success at the bottom end of the market. Moreover, their role as a conduit for various government programs is only one aspect (and in many cases a relatively minor one) of their overall involvement in the microfinance sector. Hence, in Indonesia, Bank Rakyat Indonesia and the network of BPRs may be suitable vehicles for the implementation of microfinance programs.

In general, however, the most effective way for governments to encourage regulated banks to become involved in microfinance is to ensure an appropriate regulatory and prudential framework. With an appropriate framework, those banks which wish to become involved in microfinance will be able to do so on a commercial basis, while those that do not will be able to concentrate on other activities. These issues are discussed in more detail in chapter 5.

One important way in which commercial banks can become involved in microfinance is through linkages with NGOs and SHGs. This approach combines the strength of commercial banks in financial management with the ability of NGOs to reach the poor. In many cases, community-based NGOs may have an advantage over commercial banks in reaching the poor, reflecting factors such as proximity, trust, commitment, flexibility and responsiveness. Linkages between banks and NGOs and SHGs may take a variety of forms. For instance, NGOs may act as financial intermediaries, with banks extending loans to NGOs and the NGOs on-lending to final borrowers. Alternatively, NGOs may act as social intermediaries, selecting and motivating groups of borrowers on behalf of banks, but with the loan contract being between the bank and the final borrower. Commercial banks may also lend to institutions involved in microfinance that are not NGOs, such as second tier microfinance institutions, rural banks, cooperatives and others.

The Foundation for Development Cooperation (1992 and 1995) has demonstrated that such linkages have considerable potential to provide financial services to the poor on a sustainable basis. Nevertheless, it must be said that progress in establishing such linkages has been slow, and there are relatively few examples of linkages except where they have been promoted by specific government programs as discussed below. In Bangladesh and Nepal, some state commercial banks have established linkages with MFIs on a commercial basis and outside the framework of specific government programs, but to date no private commercial banks have done so in those countries. In Sri Lanka a private commercial bank, Hatton National Bank, has established a very successful microfinance program which includes using NGOs as social intermediaries (this program is discussed in detail in Gallardo, Randhawa and Sacay, 1997). Such examples, however, are still comparatively rare.

In most countries included in the study there are no specific regulatory impediments to linkages. In some countries, the use of NGOs as social intermediaries may be constrained by interest rate ceilings which prevent banks from lending on a commercial basis to microentrepreneurs. These issues are discussed in chapter 5. Some other general regulations affecting banks or NGOs may also reduce the scope for linkages. Nevertheless, the major obstacles to linkages relate to matters other than the regulatory environment.

The only exception to this is Pakistan, where banking regulations make it extremely difficult for commercial banks to establish linkages. First, banks are only permitted to make unsecured loans up to Rs100,000 ($2,850). This rule greatly restricts the scope for banks to lend to NGOs, village organisations or SHGs for on-lending to poor borrowers, as such organisations generally have little to offer by way of security. Second, other than individuals, banks are only allowed to lend to structured bodies which meet certain conditions. For instance, structured bodies are required to maintain audited accounts, and are subject to regulations concerning debt-equity ratios and other matters. This makes it very difficult for banks to lend to relatively informal group structures such as village organisations and SHGs. The Aga Khan Rural Support Program is currently holding discussions with the State Bank to amend these rules or to exempt loans to NGOs and village organisations from these requirements. Nevertheless, as they stand, these restrictions are clearly a major barrier to banks establishing linkages with NGOs and SHGs. In other countries there are no major policy barriers to linkages. However, other than in India and Indonesia, there is little positive support for linkages from policymakers.

In India NABARD, which was established by the Reserve Bank in 1982 as the apex body for rural credit, has an active program for linking banks with SHGs, and has issued detailed guidelines to banks about how to facilitate such linkages. The main features of the program are as follows:

(1) The banks extend loans to SHGs either directly or through NGOs. In around 94 per cent of cases, the SHGs have been initiated and promoted by NGOs. In around two-thirds of these cases the NGOs have acted as social intermediaries, and the loan contract has been between the bank and SHG. This is NABARD’s preferred model, and is the one adopted by major NGOs such as the Mysore Resettlement and Development Agency (MYRADA). In around one-third of cases, the NGOs have acted as financial intermediaries, and have borrowed from banks and on-lent to the SHGs. NABARD, however, has sought to discourage this model.

(2) The banks lend at an interest rate of 10.5 per cent if lending through an NGO, or 12 per cent if lending directly to an SHG. Where the bank lends to an NGO, the NGO on-lends to SHGs at 12 per cent. The banks receive refinancing from NABARD at 6.5 per cent.

(3) Loans from banks to NGOs and SHGs are generally for three years, but loans from SHGs to members are flexible, with terms determined by the members. SHGs normally charge interest rates of between 24 and 36 per cent.

As of March 1997, a total of 8,546 SHGs, with a total membership of more than 150,000, had been linked to banks. Banks had lent Rs111 million ($3.1 million) to these groups, and had received refinancing from NABARD of Rs103 million ($2.9 million). Twenty-six (26) state commercial banks, two private commercial banks, 67 regional rural banks, seven cooperative banks and 166 NGOs had been involved in the program. As well as providing refinancing, NABARD has also supported linkages between banks and SHGs in a variety of other ways. For instance, it has funded exposure and training to commercial bank staff to enable them to understand the program and the advantages that it offers to banks. Training is normally provided by NGOs and training institutes. By September 1996, around 2,200 bank officers had received training and exposure.

While the NABARD program is a positive initiative, it should be noted that it is not sustainable in its current form. In particular, the mark-ups do not provide anywhere near a sufficient margin for NGOs to recover the costs of forming and motivating SHGs. Where NGOs act as social intermediaries, the mark-up between bank and SHG is only 5.5 percentage points. In practice, banks retain this margin and do not reimburse the NGOs for their costs, which in any case are likely to be above 5.5 percentage points. Where NGOs act as financial intermediaries, the mark-up is only 1.5 percentage points. Hence, NGOs are only able to participate in the program if they can obtain donor funds to subsidise these activities. This is likely to mitigate against expansion of the program.

In Indonesia, the PHBK (project linking banks with SHGs) has been implemented with German technical assistance. It originally concentrated on nurturing linkages between branches of various state commercial banks (although private commercial banks are now involved to some degree) with NGOs and state-sponsored agencies which are active in organising the poor into SHGs for savings and credit management. The growing involvement of rural banks in the PHBK scheme has permitted an acceleration in recent times in the numbers of clients and volume of credit disbursed via linkages, at a rate of 10 per cent per month. Some 400 banks, 183 ‘self-help promoting institutions’ (including NGOs) and 6,800 groups were involved in September 1996. Interest rates are market based and transaction costs are covered by interest margins. The project is a painstaking attempt to build a sustainable system for credit and savings mobilisation, with limited outreach as yet.

The freedom of banks, NGOs and SHGs to set interest rates at sustainable levels in Indonesia’s deregulated financial environment, and the absence of central bank refinancing, are other factors which distinguish the Indonesian situation from that in India. The preferred model of linkage is one in which the role of the NGO or other intermediating agency is social; financial transactions occur directly between the bank branch and the SHG. Some rural banks are moving directly to organise their own groups, cutting out the role of the ‘social intermediary’ in some cases.

Other countries have not adopted comparable programs. In Bangladesh, the central bank conducted a seminar on linkages in 1994, and it has issued a number of circulars to commercial banks encouraging them to establish linkages with NGOs for lending to the poor. Nevertheless, the commercial banks commented that it is necessary for the central bank to take a more proactive role. In Nepal, the proposed Financial Intermediation Act will be an important step in establishing the financial intermediary role of NGOs and MFIs. In these and in other countries, it would be appropriate for the central bank to take a more active role in encouraging linkages by documenting and publicising successful examples, issuing clear guidelines that the banks could follow, providing training to banks, encouraging ratings systems for MFIs, and more generally promoting change in the culture of banking. This would be conducive to an environment in which such linkages would occur on a commercial basis, outside the framework of government programs. The guidelines issued by NABARD in India may be a useful starting point, although they would need to be adapted to ensure that programs are sustainable and to suit the institutional framework in other countries.

The other main category of microfinance programs is those operated directly by government agencies. This is a diverse category, with few commonalities among schemes. While most countries have such programs, most are small and not well documented. Nevertheless, the information that is available suggests that most microfinance programs operated by government agencies suffer from high costs and low repayment rates, and in many cases do not reach poor borrowers.

Bangladesh

There have been a number of microfinance programs administered by government agencies in Bangladesh. One of the largest and most successful of these is the RD-12 program of the Bangladesh Rural Development Board (BRDB), with some 328,000 borrowers as at December 1995. BRDB is a semi-autonomous government agency with its own board of directors. In addition, the RD-12 project has a central coordinating committee, including representatives from the Canadian International Development Agency (CIDA), which meets once every six months to decide on overall policy matters relating to the project. The project borrows heavily from the experience of NGOs involved in microfinance, including the Grameen Bank and Bangladesh Rural Advancement Committee (BRAC), with the lending methodology based on the Grameen Bank model.

The project has performed very well compared to most other government microfinance programs both in Bangladesh and in other countries, with a recovery rate of 91 per cent in 1993. Its success demonstrates that by using best practice techniques and ensuring relatively independent management free from political interference, it is possible for government agencies to establish effective microfinance programs. Nevertheless, the repayment rates achieved by RD-12 are still below those achieved by leading MFIs in Bangladesh. RD-12 has also not been under as much pressure as the NGO programs to minimise costs and become sustainable. Moreover, it is understood that most other government programs have not been nearly as successful as RD-12.

Indonesia

In Indonesia, there are two major new microfinance programs of a mass character operated by government agencies. These represent something of a reversal in the trend towards sustainability in government programs in that country. The IDT (Inpres Desa Tertinggal or Presidential Instruction relating to backward villages) is a presidential initiative programmed in Indonesia’s sixth Five Year Plan (1993–1998). Prompted by perceptions of growing relative poverty and regional income inequality and coordinated by the National Development Planning Board, it calls on interdepartmental resources to channel funds to more than 20,000 ‘backward’ villages, primarily through grants to provide seed capital for the economic activities of the poor who are organised in self-help groups. The groups themselves decide the terms on which funds will be on-lent to members. Other resources are provided to facilitate the preparation of these groups, and to improve infrastructure in the chosen villages. It is hoped that after some period of funding, the groups will ‘graduate’ to more commercial sources of funding (such as the PHBK program of Bank Indonesia). It is not clear that sustainability is a primary goal, in that groups hold what is essentially a revolving fund and substantial leakage may be occurring. Coverage of the program was some 107,000 groups and 2.9 million people at end-1996. By mid-1997 some 3 million households were said to have received seed capital averaging $81 per household.

The Prosperous Family program was initiated in March 1996 and over a twelve-month period channelled funds to the mothers of some 9.8 million families. This was possible because it employed the infrastructure of Indonesia’s national family planning agency, which is far-reaching and effective, and the existing structure of groups organised for family planning and maternal and child health. The program is another Presidential initiative, and a direct reaction to political concerns about inequality. Women beneficiaries are stratified by economic status and perceived ‘readiness’ to handle credit effectively. The poorest start with savings, and initial small grants are designed to provide an incentive for individual saving. After some savings have been accumulated, women progress to receiving very small loans, at a highly concessional interest rate of 6 per cent. Much criticism is directed at the small size of these loans and at the capacities of family planning and health field staff to generate economic activity among the poor. The scheme is still too new, and the numbers too large, for any careful evaluation. It remains to be seen (for both IDT and Prosperous Family) whether concerns for institutional development and sustainability will prevail over the attraction of short-term political gains from widespread disbursement of cash.

Philippines

In the Philippines, a recent audit of microfinance programs by the National Credit Council (NCC) uncovered at least 111 government credit programs. Many of these involve government agencies lending directly to final borrowers. These government credit programs have been criticised for being inefficient, highly politicised, uncoordinated and unsustainable. For instance, Llanto, Garcia and Callanta (1996, p.14) comment that:

. . . Philippine experience has shown the huge inefficiency and high costs of using government non-financial institutions to implement credit programs. Recent research has shown the unsustainability of government supply-led credit programs, the great capacity for leakage of the benefits of government credit programs to the non-poor, the duplication and overlapping of a number of credit programs leading to gross inefficiencies, the distortion of the financial market and weakening of private sector incentive to innovate.

The Department of Finance and the NCC are currently working to rationalise these programs. The NCC guidelines provide that government line agencies should focus on technical assistance and capacity building, and should channel all their lending activities through banks and MFIs. Political and bureaucratic resistance leaves the ultimate outcome in doubt.

Another large government microfinance program is the poverty alleviation project operated by the Community Development Department in Thailand. The program operates only in villages that are identified as poor, and individual borrowers are also means tested. A revolving fund is established in each village, and borrowers are identified by local government authorities using basic minimum needs data collected by the village. Loans to individual borrowers are interest free, with administrative costs met through the budget. By 1995, around B2.8 billion ($110 million) had been advanced to village welfare committees in around 10,000 villages, and nearly B1 billion ($40 million) had been on-lent to final borrowers. The program is clearly not sustainable, and analysis by the World Bank suggests that repayment rates are very low.

Comment

Based on these experiences, it appears that government microfinance programs that lend directly to final borrowers are rarely cost-effective. Moreover, they may ‘crowd out’ more effective microfinance programs. In general, it would be appropriate to abolish such programs and channel the funds through well-managed MFIs instead. Direct government programs should only be retained where it can be shown through a rigorous evaluation that, taking all direct and indirect subsidies into account, they are as cost-effective as leading MFIs.

In a number of countries, governments have forgiven certain categories of small loans under their own programs or by state commercial banks. Two particular examples are the 1991 decision by the government of Bangladesh to forgive all agricultural loans up to Tk5,000 ($120), and the periodic debt waivers by successive governments in Sri Lanka, most recently in 1995. However, debt forgiveness is not confined to these two countries. These episodes have created major problems for MFIs. Clearly, failure to ensure high repayment rates in other parts of the financial system reduces overall borrower discipline and makes it much harder for MFIs to maintain high repayment rates. Governments should avoid debt forgiveness for small loans in the future.

3.6 Summary and recommendations

Active support from government and/or donor agencies, including direct financial support, is a necessary catalyst to the establishment of a viable microfinance sector. While microfinance institutions (MFIs) should strive for self-sufficiency, attaining it takes a considerable period of time. To set up their programs and for a number of years thereafter, MFIs need to be subsidised. Expanding to achieve economies of scale also requires a considerable input of resources for capacity building and institutional development.

MFIs in all countries have received considerable support from the government and/or donor agencies. In the past, most of this has been in the form of financial assistance direct to an individual MFI. This support has been critical in enabling them to establish, and to increase their outreach and self-sufficiency. Without this support, it is unlikely that there would be a microfinance sector at all, in the sense of specialist MFIs providing financial services to the poor in an innovative and commercially viable manner.

Nevertheless, the development of microfinance and the success of leading MFIs owes more to the vision of their founders and management than to the criteria used by government and donor agencies for supporting them. Support has been largely ad hoc in nature, and has generally not been explicitly designed to maximise the twin objectives of outreach and sustainability in the most cost-effective manner.

The Guiding Principles for Selecting and Supporting Intermediaries, agreed by major donor agencies in October 1995, provide a number of suggestions as to how support by donors and government can best be directed to maximising outreach and sustainability. Most importantly, governments and donor agencies should only support institutions which demonstrate a capacity to reach poor clients on a sustainable basis. Support should be directed to institutions with demonstrated performance or credible plans in terms of institutional strengths, quality of services and outreach, and financial performance. The Guiding Principles also consider the types of support that donor agencies should provide.

While the major donor agencies have endorsed these principles, they are not always followed in practice. All governments and donor agencies should ensure that their support for MFIs is consistent with the Guiding Principles.

Most countries included in the study have some form of ‘second tier’ microfinance institution which channels funds from the government and/or donor agencies to individual MFIs. The largest and most successful second tier institution is the Palli Karma Sahayak Foundation (PKSF) in Bangladesh. Second tier institutions are also an important part of the microfinance sector in India, Sri Lanka, the Philippines and Thailand.

Where they operate well, second tier microfinance institutions are a very effective means for supporting MFIs. It is much more efficient for one institution to analyse and monitor the performance of individual MFIs than for different government and donor agencies to do their own appraisals and evaluations. Moreover, if all the support for a particular MFI can be channelled through one agency, it is much easier to ensure that it is complementary and directed to the same objectives.

For these reasons, governments and donor agencies as much as possible should channel their support for microfinance through efficient and well-managed second tier microfinance institutions. In countries with more than one second tier institution, governments should ensure that the different institutions adopt consistent standards and policies.

At the same time, there is evidence both from Asia and from other regions that second tier institutions do not always operate on a sound basis, and may be subject to significant administrative, political and technical difficulties. If second tier institutions are to play an effective role, it is clear that there are a number of critical issues that need to be addressed.

First and most important, second tier institutions should establish and enforce appropriate performance and reporting standards for the MFIs that they fund. They should support MFIs based on the standards they achieve, rather than the model they follow. Second, pressure to meet unrealistic disbursement targets may cause institutions to lower their standards and be counter-productive.

Third, second tier institutions should avoid restrictive interest rate policies. In India, for instance, the major second tier institutions impose ceilings on the interest rates that their partner MFIs can charge borrowers, at levels which make it difficult if not impossible for the MFIs to achieve self-sufficiency.

Fourth, some second tier institutions only lend to MFIs for on-lending to final borrowers. They should develop appropriate criteria for funding scaling-up, institutional development and equity, as well as providing loanable funds. Where privatisation is envisaged, other arrangements for such support must be assured.

Fifth, it is critical that second tier institutions operate as independent entities and are not politicised. Some commentators suggested that the performance of the National Development Trust Fund (NDTF) in Sri Lanka had been affected by politicisation, and this also appears to have been a factor affecting institutions in some other countries.

Seven of the nine countries included in the study impose directed credit requirements, whereby banks are required to lend a certain proportion of their loan portfolio to particular sectors. Other things being equal, directed credit schemes which target the poor will cause banks to lend more to poor borrowers than would otherwise be the case. However, they are an inefficient means of reaching the poor, and impose economic costs.

Channel public funds through second tier institutions and specialist MFIs rather than banks

In some countries, the government operates specific microfinance programs in which funds are channelled through the banks. The largest of these programs is IRDP in India. Such programs have generally not been successful, and may ‘crowd out’ more effective programs. Where public funds are available to support microfinance, it is more appropriate to channel them through second tier microfinance institutions and specialist MFIs, rather than through the banking system. The most effective way for governments to encourage regulated banks to become involved in microfinance is to ensure an appropriate regulatory and prudential framework, rather than through direct financial support. On the other hand there is also no reason why small banks with a commitment to microfinance should not borrow from second tier institutions.

An apparent exception to this analysis is in Indonesia, where some microfinance programs channelled through banks, in particular, Bank Rakyat Indonesia and the rural banks (BPRs), appear to have been successful. This is because these banks have a genuine commitment to outreach, as well as appropriate organisation, products and procedures. Hence, in Indonesia, banks already involved in microfinance may be suitable vehicles for the implementation of government microfinance programs.

Encourage commercial bank involvement via linkages

One important way in which commercial banks can become involved in microfinance is through linkages with non-governmental organisations (NGOs) and self-help groups (SHGs). The only country included in the study with specific regulatory impediments to such linkages is Pakistan. However, other than in India and Indonesia, there is little positive support for linkages from policymakers, and even in India the program is not sustainable because the mark-ups that banks and NGOs can charge do not provide anywhere near a sufficient margin for NGOs to recover the costs of forming and motivating SHGs. Central banks should take a more active role in encouraging linkages by documenting and publicising successful examples, issuing clear guidelines that the banks could follow, and other measures. This would be conducive to an environment in which such linkages would occur on a commercial basis, outside the framework of government programs.

The other main category of microfinance programs is those operated directly by government agencies. The largest and possibly the most successful such program is the RD-12 program of the Bangladesh Rural Development Board (BRDB). Its success demonstrates that by using best practice techniques and ensuring relatively independent management free from political interference, it is possible for government agencies to establish effective microfinance programs. Nevertheless, its performance is still not as good as that of leading MFIs in Bangladesh.

In the Philippines, a recent audit of microfinance programs by the National Credit Council (NCC) uncovered at least 111 government credit programs, many involving government agencies lending directly to final borrowers. These government credit programs have been criticised for being inefficient, highly politicised, uncoordinated and unsustainable. Fragmentary evidence suggests that these criticisms could also be levelled at most direct government microfinance programs in other countries.

In Indonesia, while there is not the proliferation of line agency programs observed in the Philippines, there are several such programs. Two have achieved mass outreach in a short period, with the Inpres Desa Tertinggal (IDT) reaching more than 3 million households in less than four years, and the Prosperous Family program reaching almost 10 million households in a single year. Considerations of institutional development and sustainability may have been casualties in the face of the government’s need to be seen to be responding to popular concern about regional and interpersonal income inequalities.

Abolish direct government microfinance programs

In general, governments should abolish direct microfinance programs and channel the funds through well-managed second tier microfinance institutions or MFIs instead. Direct government programs should only be retained where it can be shown through a rigorous evaluation that they are as cost-effective as leading MFIs, taking all direct and indirect subsidies into account.

Avoid debt forgiveness for small loans

In a number of countries, governments have forgiven certain categories of small loans under their own programs or by state commercial banks. These episodes have created major problems for MFIs. Governments should avoid debt forgiveness for small loans in the future.