REBA case study. Rural Micro Finance, Zimbabwe. About REBA Case Study Briefs. Overview NUMBER 20 NOVEMBER 2007

Transcription

1 REBA case study NUMBER 20 NOVEMBER 2007 About REBA Case Study Briefs What is RHVP? The Regional Hunger and Vulnerability Programme (RHVP) supports improvements in policy and programme approaches to hunger and vulnerability in southern Africa with particular emphasis on the role of social protection. The Regional Evidence Building Agenda (REBA) Evidence-building, together with capacity-building and policy change, is one of RHVP s three interlinked activities. The Regional Evidence Building Agenda (REBA) is a cohesive framework that has guided the Programme s cross-country evidence-related activities between April 2006 and September The REBA consists of individual case studies of 20 ongoing social transfer programmes together with a further 18 thematic studies of various cross-cutting design and implementation issues. The studies were carried out by a network of national consultants in the six southern African countries covered by RHVP (Lesotho, Malawi, Mozambique, Swaziland, Zambia and Zimbabwe). This network was supported and guided by a core team of international mentors which included Stephen Devereux (IDS Sussex), Frank Ellis (ODG East Anglia) and Lionel Cliffe (University of Leeds) and was coordinated and managed by Philip White (ODG). REBA Aims The REBA aims to support RHVP s efforts to promote improved policy and programme approaches to social transfers as a means of addressing hunger and vulnerability. REBA findings are feeding into a range of policy, advocacy and research outputs and processes, including policy briefs, best practice guidelines, national and regional learning events for policy-makers, practitioners and civil society, a film series and research publications. In addition, by working through a network of national consultants, the REBA aims to increase national capacity to carry out analytical research on hunger and vulnerability within the six focus countries. REBA Case Study Briefs This series of briefs summarises the main findings from the 20 individual case studies undertaken under the REBA. Based on these findings and those of the accompanying thematic studies, a series of consolidated thematic briefs are currently in preparation. The full reports of each case study will be made available in early Rural Micro Finance, Zimbabwe Overview It is arguable whether microfinance can be regarded as social protection. For one thing, the use of external funds for microfinance is typically restricted to start-up activities, with self-financing being the long run intention. For another, savings and loans are to do with financial markets and small business development, and are therefore aimed at stimulating economic opportunities rather than securing minimum acceptable levels of consumption through social transfers. Of course these are matters of interpretation and degree, but some contemporary uses of the term social protection seem to expand its remit to cover almost all development policy, and this is not necessarily helpful for advancing the case for improved social protection coverage for the vulnerable in southern African countries in the future. Nevertheless, this review of a rural microfinance scheme in Zimbabwe is included as the final case study in this social transfer series. Readers can make up their own minds whether microfinance constitutes a social transfer. The scheme under consideration is called the Kupfuma Ishungu Rural Micro-Finance Programme (KI- RMFP). The goal of this scheme is stated as the sustainable protection and promotion of the livelihood security of rural economically active households. The scheme is managed by the international NGO CARE through local partners. It began in 1998, with funding from Austrian aid, and has expanded substantially since then, with funding in different phases also from the DFID Protracted Relief Programme (PRP), and the Swedish International Development Agency (SIDA). KI-RMFP currently works in 20 districts across 5 of Zimbabwe s provinces. At the most recent count in October 2006, its total outreach comprised 7,600 savings and loans groups with 72,000 members, of which 87 per cent were women. Cumulative savings across groups in mid-2005 amounted to Z$4.2 billion, and cumulative total small loans made across all groups had reached over 1.1 million at that time. Difficulties for participants in the scheme are created by the accelerating hyperinflation of recent years, due to the speed at which the real purchasing power of money savings declines. This means that cash savings must be turned into tangible assets almost instantly in order to minimise erosion in their value.

2 Organisation The activities of CARE Zimbabwe are organised into four main programmes, one of which is the Small Economic Activity Development (SEAD). The KI-RMFP is a rural savings and loan project that comes under SEAD. Since its inception in 1998, the project has evolved several different models of implementation. One is direct implementation down to community level by CARE itself, with extensive hands-on involvement by CARE s own field staff. A second is through local partner organisations, and one of the principal partners has been the Tongogara AIDS Support Organization Zimbabwe (TASOZ). The third, more recent approach, is through the creation of Community Based Trainers to handle most implementation issues at community level, and this is thought to have good prospects for ensuring the sustainability of the scheme beyond the timescale during which CARE or a partnership organisation would normally be involved. KI-RMFP uses a widely tested approach to creating selfsustaining rural financial services in low income countries, called the internal savings and lending model (sometimes abbreviated to ISAL). This model was originally developed by CARE in Niger in the 1990s, and is based on the traditional rotating savings and credit associations (ROSCAs) that are found in many Sub-Saharan African countries and elsewhere. The approach creates village level savings and loan associations or groups. The process of association creation starts with familiarising the community with the basic concepts using local stakeholders such as traditional leaders and communitybased organisations. Once this has taken place, interested individuals are trained in clusters of 25 people over five continuous half-day sessions. Key topics covered in training include individual self-screening, groups and leadership, constitution, group fund development, savings and lending record keeping, and selection of income generating activities. After training, individuals form voluntary self-selected associations, and members then set their own goals, internal by-laws, level and regularity of saving, and loan amounts and procedures. Members also determine a rate of interest payable on loans, that helps to ensure that their group fund grows. After an agreed period of saving and lending the group fund is wound up and shared between members, after which a new cycle starts (or the association may decide to disband). CARE or one of its partner organisations monitors the progress of the associations in any particular locality for 16 months, through a programme of routine visits and meetings. Vulnerability The Zimbabwe economy has been in deepening crisis for five or more years. The outward manifestations of this crisis are hyperinflation (estimated to have reached 6,600 per cent in mid-2007) and a rapidly depreciating external value of the domestic currency (see footnote 1 above). Hyperinflation affects most severely those whose income and assets do not adjust upwards in value as fast as the rate of inflation, and the poorest and most vulnerable members of society are always the hardest hit. Almost all recent macroeconomic data about Zimbabwe has to be treated with a great deal of caution, since accurate measurement is practically impossible when events are moving so fast. According to official UN data, Zimbabwe s GDP fell by more that half, from 7.4 to 3.4 billion US$ between 2000 and 2005; while per capita GNI declined by 25 per cent from US$460 to US$350. The discrepancy between these two rates of decline is explained by remittance income from outside the country underpinning the income levels of a great number of Zimbabweans. It is estimated that out of a total population of 13 million in 2005, roughly 3 million Zimbabweans are living outside their country. Rapid economic decline increases the proportion of the population in poverty, as well as their vulnerability to food shortages. Additional factors are rainfall failures, causing production failures and food security stress in rural areas; and AIDS causing loss of able-bodied labour, chronic illness, medical and funeral costs. Zimbabwe currently has an estimated overall HIV prevalence of 20.1 per cent in adults aged years. This has apparently come down from higher previous levels, although quality of reported data may be an issue. At any rate some 1.7 million Zimbabweans are thought to be living with HIV (2005 estimates) and there are an estimated 1.1 million AIDS orphans (UNAIDS, 2006) In the choices it has made concerning where and with whom to work, KI-RMFP has set out to address these various vulnerabilities. Places with high climate variability, high HIV and AIDS prevalence, growing numbers of AIDS widows and orphans, and difficulties of access to farm inputs have been prioritised by the scheme. As 2 REBA CASE STUDY BRIEF NUMBER 20 NOVEMBER 2007

3 already indicated earlier, however, project design has in the past assumed that these vulnerabilities were embedded in relatively stable contextual economic conditions, which is preferable for the orderly working of saving and loan transactions. The advent of hyperinflation places the design in unstable circumstances that it may not be sufficiently robust to adapt to, placing scheme success under considerable stress. Targeting At beneficiary level the KI-RMFP is designed fundamentally as a self-targeting programme. It therefore avoids the manifold problems of beneficiary selection, and the associated community tension, that tend to accompany targeting using proxy indicators of poverty and vulnerability. Nevertheless at different levels, KI-RMFP uses criteria that influence where and by whom savings and loan associations eventually form in rural communities. A broad concept of geographical targeting allied to spatial manifestations of vulnerability in Zimbabwe has informed the provinces and districts in which the scheme has been rolled out. In rural communities, particular attention has been paid to women, widows and families containing orphans and vulnerable children (OVC), and special subprojects have focused on savings and lending for children (for example, to enable them to buy school uniforms). Overall, about 85 per cent of participants in associations are women. This has had the positive effect of empowering women in many communities, but the lack of male involvement in the scheme has been of some concern to CARE, which has tried to encourage the formation of mixed associations. These, though, throw up particular social interactions of their own that are not always conducive to stability. For example, women quickly lose interest when men seek to dominate discussions and decision making, and it is often found that men prioritise the purchase of different assets from women, for example, livestock in preference to cooking utensils. The encouragement of children s associations has generally been rather successful, despite the potential for intra-household conflict between children and adults regarding the priorities of such groups. Children s associations evidently require the active participation of parents or guardians in their functioning, and the latter are normally expected to provide the initial seed money to enable such associations to get going. Table 1 summarises the total number of households participating in children s savings and lending associations in the seven districts where these have been encouraged for the year Table 1: Households Participating in Children s Saving and Loan Groups District No. of No. of % Female Groups Households Participation Zaka Bikita Gutu Gweru Masvingo Mberengwa Zvishavane Total Source: CARE Zimbabwe Coverage The KI-RMFP started in 1998 in the Mberengwa, Shurugwi and Zvishavane districts of Midlands Province in Zimbabwe. Due to the widely perceived success of the pilot, the project was replicated in Masvingo Province from May 2000 to August With further sponsorship from DFID, the project outreach was extended to further parts of Masvingo and Midlands Provinces in the period September 2001 to March More recently, under SIDA funding, the project has extended to other provinces and to urban areas. The project is currently operating in 20 selected districts in five provinces in Zimbabwe, as shown in Table 2. Table 2: Geographical Coverage of KI-RMFP at the end of 2006 Province Midlands Masvingo Mashonaland West Manicaland Harare Source: CARE Zimbabwe Districts Mberengwa, Zvishavane, Shurugwi, Chirumhanzu, Kwekwe, Gweru Zaka, Chivi, Mwenezi, Gutu, Bikita Kadoma, Makonde, Zvimba, Hurungwe Nyanga (East and West) Chitungwiza, Mabvuku, Epworth REBA CASE STUDY BRIEF NUMBER 20 NOVEMBER

4 Additional data on the coverage of the project between 2003 and 2005 is provided in Table 3 below. This shows a steep rise in number of groups (from 2,694 to 7,595) and in total group members (from 17,636 to 47,644) between December 2003 and March This rate of expansion has continued since 2005 as indicated by the participant figures given in the Overview above. Over the past two years, the spiralling rate of inflation has made it very difficult for groups to keep savings in cash, and they have moved towards converting cash into collectively prioritised tangible assets on a routine basis. This possibly compromises the rural financial market rationale of the scheme, a point taken up further below. Coordination The broad picture of the coordination of KI-RMFP has already been described under Organisation above. There are two aspects that can be further delineated. First, KI-RMFP has been supported by a succession of different donors, and donor priorities as well as the situation on the ground have determined some of its switches in scheme emphasis over the years. Thus under Austrian government funding, it was concerned with providing options to reduce rural vulnerability, primarily by expanding the range of different livelihood activities available to poor rural families. Under DFID funding after 2001, the emphasis switched to AIDS widows and orphans, and this is the main period when coverage of women and households containing orphaned children was greatly expanded. Under more recent SIDA funding, another switch in emphasis has occurred to take in poor urban areas in addition to its previously exclusive rural remit. The other area of coordination worth saying a little more about is Community Based Trainers (CBTs). One of the problems any scheme like this confronts is the cost and duration of commitment at field level of salaried field officers, whether of the principal NGO or of its local partners. Given that CARE considers that associations need monitoring and nurturing for up to 16 months after start-up, this is potentially quite an onerous burden. The idea of CBTs arose from suggestions made within participating communities themselves. It is that community members volunteer to train others to start their own savings and credit groups, using the knowledge and materials that they have obtained in their own training. In this way the scheme can potentially replicate itself with only minimum oversight by the original catalysts for group formation. By December 2005, 403 CBTs were in place across project communities, and 655 new groups have been trained by CBTs (CARE Zimbabwe, 2006). Cost Effectiveness Quite a lot is known about the budgets and costs and outcomes of KI-RMFP in the period up to March 2005, when an evaluation of the project was carried out (see Table 3). The data in Table 3 suggest that about US$1.6 million had been spent on the project by March 2005, working out at US$33 per graduated participant in the scheme i.e. excluding participants who dropped out along the way. The overall savings level across the 7,595 groups that existed in March 2005 was roughly Z$4.9 billion (US$808,000). The cumulative loans achieved by then were Z$12.7 billion (US$2.1 million). Loans are a multiple of savings, due to the rotating character of an ISAL scheme. This means that the scheme cost about US$0.75 in seed money per US$1.00 in loans achieved. This ratio would tend to improve in efficiency over time, since rotating savings and loans can continue indefinitely once a successful association is up and running, while the set-up cost remains the same. The table also contains data indicating the rising efficiency of field staff involved in the setting up and monitoring of saving and lending associations. This rise is partly artificial in that if business expands and more associations are created long after the field staff have moved on, then this will automatically appear as if rising productivity is occurring. More relevant would be comparisons of new groups created per field officer on an annual basis. In addition, to the extent that there is a rise in efficiency this may be to do with the creation of the CBTs, which will have reduced the burden of downstream field visits to project locations after the initial start-up phase in each place. Market Effects It is widely agreed that increased savings and loan activity generally implies increasing cash-based activity in a rural area overall, thus helping to expand and consolidate the working of rural markets. In the case of KI-RMFP this is all the more so, since certainly earlier on in the project 75 per cent of loans made by associations were for trading activities (see Table 4). These data apply to 2002, and could misrepresent loan patterns in later years. 4 REBA CASE STUDY BRIEF NUMBER 20 NOVEMBER 2007

5 Table 3: Financial and Operating Data, KI-RMFP, PARTICIPATION Reporting Period To: 31 Dec Dec Mar 05 Total Members 17,636 37,433 47,644 Total Men 2,249 4,079 6,517 Total Women 15,387 33,354 41,127 Total Groups 2,694 6,013 7,595 Mean Group Size SAVINGS Savings per Group Z$ 65, , ,065 Savings per Member Z$ 10,017 50, ,150 Total Savings Z$ bn* LOANS Total Loans Z$ bn (cumulative) * Loans per Member Z$ 12, , ,978 No. of Loans per Member Average Loan Size Z$ 12,484 94,020 11,102 Total Loans To Date 53, ,895 1,141,486 Interest on Savings % 36% 95% 82% EFFICIENCY Ratio Field Staff to Total Staff 75% 83% 75% Groups per Field Officer Individuals per Field Officer Cost per Graduated Participant ($US) US$18 US$39 US$33 Source: Chigara (2005:, p.11) * calculated from other figures in the Table Note: Financial data in this table is in old Z$. The exchange rate in March 2005 was old Z$6,082 = US$1.00. Table 4: KI-RMFP Loans by Sector of Investment, 2002 Sector No. % Manufacturing Trade 12, Service Agriculture School Fees Food 1, Health Bus Fares Maize Milling Funeral Costs Source: Allen et al., 2002 Asset Building Typically, asset building under a microfinance project might be expected to occur due to the use made of loans to buy tangible capital (e.g. livestock, bicycle, sewing machine). This has certainly occurred in KI- RMFP, but what is perhaps unique to the scheme in recent years has been the substitution of cash for assets in the savings portfolio of associations, in order to overcome the continuously eroding value of cash under hyperinflation. This is called the asset-based saving innovation. Group members alternate in buying assets for one another instead of paying cash into a savings account, and then these assets are shared out at the end of the pre-arranged cycle. In communities directly supported by CARE, asset-based saving had reached REBA CASE STUDY BRIEF NUMBER 20 NOVEMBER

6 48 per cent of all saving in 2005, while in TASOZsupported communities this proportion had reached 75 per cent. The most common assets acquired (see Figure 1 below) are household utensils such as plates and pots (42 per cent), and agricultural inputs and livestock e.g. chickens, rabbits, goats, cattle (31 per cent). A wide variety of other assets is in a mixed category (23 per cent); while property corresponded to 6 per cent and bedding to 2 per cent. Strengths KI-RMFP is widely regarded in Zimbabwe as a highly successful microfinance initiative. An evaluation carried out in 2005 for SIDA noted a number of positive impacts and outcomes of the project. Some strengths of the programme are summarised here as follows: (i) it is a scheme based on previous successful experience, and it has adapted itself to rapidly changing circumstances in Zimbabwe as time has gone on; (ii) since self-selection into savings and loan groups is a fundamental philosophy of the scheme, its continuous and at times rapid expansion over the past 9 years reveals that its microfinance model seems to work for a great number of poor and vulnerable people; (iii) it has adapted its basic model for specialist groups such as AIDS widows and families caring for orphans, and has successfully launched savings and loans for children in vulnerable families; (iv) it has innovated as it has gone along, in particular by creating community facilitators to support replication of savings and loans association at community level with minimal additional external assistance, and by encouraging associations to move from cash to asset-based savings as the rate of inflation picked up; Figure 1: Importance of Different Categories of Asset-Based Savings, Weaknesses Some weaknesses of the scheme were noted in the review of its functioning undertaken as background to this case study (Gandure, 2007) (i) Utensils Livestock Mixed Property Bedding Source: Chigara (2005) there was a sense that monitoring of outcomes at field level had tailed off somewhat in recent years, especially for associations created in earlier phases of the project: this risked making it difficult in the future to assess the functioning of associations in the period since hyperinflation really took off in 2006 and 2007; (ii) the shift from cash saving to asset-based saving that was beginning to occur already in 2005 is all very well as an innovation to overcome hyperinflation, but it does mean that the loan part of the model is substantially weakened: in effect, people save to pool their purchasing power over assets like kitchen utensils, not necessarily to invest in productive activity. (v) the 2005 evaluation found that many household level indicators of welfare and well-being had improved for individuals and families that had joined savings and loan groups, including overall asset levels, housing quality, access to health and education, and food security. 6 REBA CASE STUDY BRIEF NUMBER 20 NOVEMBER 2007

7 Policy Lessons As discussed at the start of this case study, KI-RMFP is not really a social transfer project and should not strictly be regarded as a form of social protection. However, it certainly shares a great deal in common with many social transfer schemes, being oriented towards the rural poor and vulnerable, those adversely affected by HIV/AIDS, and orphans and vulnerable children (OVC). It seems to be a successful scheme, despite the extremely difficult economic circumstances within which it has had to operate. It has expanded more or less continuously since inception in 1998, and appears to have found a means of replication that is low cost in terms of the external resources required. A new phase of the project has been underway since January This is a five-year initiative funded by SIDA and the private mining conglomerate Anglo- American, that promises a continuation of the previous focus on the economically active poor and vulnerable, particularly women headed households, HIV/AIDS infected and affected, and attention to be paid to special groups like the disabled. The project will continue to operate in four provinces and intends to reach an additional 72,000 households through to 2011, in other words to double in size during this phase. Not quite enough is known about the sustained outcomes of the project in terms of participants long run food security and livelihood status. The project is able to continue to demonstrate that there is a high demand for the model that it promotes, and that associations and their members continue growing across all districts covered. However, outcome data on real changes in living standards and well-being is fragmentary and indicators for this are not well developed. Sources Allen, H., Koegler, P., and Rushawa, J, 2002: End of Term Evaluation of Kupfuma Ishungu Rural Microfinance Programme (RMFP), Zimbabwe, Report No. EZA /00 CARE Zimbabwe, 2006, Kupfuma Ishungu Rural MicroFinance Project: End of Project Narrative Report, January to December , submitted to SIDA Zimbabwe Chigara, P., 2005, KI-RMFP Mid-Term Evaluation Report, May Gandure, S., 2007, Internal Savings and Lending Model: Lessons from CARE Zimbabwe s Rural Micro Finance Programme, Report prepared for the Regional Evidence Building Agenda (REBA) of the Regional Hunger and Vulnerability Programme (RHVP), June UNAIDS, 2006, 2006 Report on the Global Aids Epidemic: Annex 2 HIV and AIDS Estimates and Data, 2003 and 2005, Geneva: UNAIDS More Information REBA material, including these briefs and fuller case study reports, as well as information regarding the REBA process can be viewed and downloaded from: building_evidence REBA CASE STUDY BRIEF NUMBER 20 NOVEMBER

8 20 Girton Road, North Park, 3rd Floor, Parktown, 2193, Johannesburg, South Africa Postnet 307, Private Bag X30500 Houghton, Johannesburg, 2041, South Africa Tel: Fax: Further information is available from or by contacting RHVP 2007 Funded by DFID Department for International Development Implemented by ODG The Overseas Development Group DEV The School of Development Studies