6 Small Loans: Shelter Microfinance – Financing Urban Shelter





Shelter has become a commodity for increasing numbers of low-income households, especially those living in urban areas of developing countries. Those who build incrementally (or progressively) are a very significant group in many countries in the South. However, loan finance for shelter-related investments in incremental dwellings made by low-income households whose income comes from the informal economy is rarely available through the formal commercial financial sector. In the vast majority of cases, these households are ineligible for commercial mortgage finance. Households seeking to invest in their shelter (land, infrastructure and housing) have been forced to use their own limited income, seek additional resources from family and friends, and borrow on informal credit markets or, in some cases, from groups such as credit unions. Sources of longer-term finance are extremely limited and interest rates may be high. Box 6.1 illustrates sources of finance used by low-income households in Hyderabad, India.

shelter microfinance and community finance mechanisms have grown considerably during recent decades

There have been several institutional efforts to assist these households in obtaining secure access to some kind of loan finance. In particular, shelter microfinance and community finance mechanisms have grown considerably during recent decades. Based predominantly in Asia and Latin America, there have been multiple explorations and innovations over the last 20 years.2 Initial activities were developed by non-governmental organizations (NGOs) working in housing and urban development, and by microfinance organizations interested in supporting housing investment. Agencies responsible for these activities now span the voluntary and public sectors. There are now a small number of larger programmes that involve multi-sectoral initiatives, with some also having a role for the private sector. As the effectiveness of small loans has become more evident, some innovative state programmes have sought to secure similar development benefits by replicating such programmes, albeit within different structures and systems. A small number of private-sector initiatives have been launched, generally building on microfinance approaches and seeking to expand into what is perceived to be a potentially profitable market. Urban dwellers in a wide range of different countries may now be offered such opportunities related to savings and loans for shelter investments.

What is characteristic of these initiatives is that they involve small-scale lending for shelter improvements. In many cases, they also encourage savings (although this may be constrained by the rules and regulations of the financial system). The growing interest in such programmes is reflected in the launch by the Cities Alliance of the Shelter Finance for the Poor Initiative in 2001, which focuses on emerging practices of providing housing finance to poor clients on commercially viable terms.

These initiatives are of particular significance in the urban sphere where land, housing, infrastructure and basic services are all marketed commodities. However, in some cases, lenders have extended into rural areas, notably in Bangladesh, where traditional materials are not sufficient for a secure dwelling. Community funds for utilities such as electricity and water management may also be associated with rural areas.

The programmes share a common perspective in that they work with the realities of urban development in the South (and, in some cases, in transitional economies), rather than the Northern model of urban development in which a house is constructed and then sold (often through a financing package) to a family or individual. Their underlying model of housing investment is one of incremental shelter development. Housing is secured over time as improvements are made when funding is available. The dwelling is gradually consolidated, made more secure and services and infrastructure are obtained. As is the case with microenterprise development, there is not a big market for such lending in the North.3 This is partly due to much higher levels of affordability, but also arises because building regulations prevent the extensive use of incremental shelter strategies. In the North, conventional modern housing is complete in one single stage even if later investments expand, renovate and/or modify the dwelling.

Low- and many middle-income urban households in the South use incremental strategies. Underlying incremental housing development is the issue of affordability as already noted in Chapter 4, many struggle to afford mortgage finance and lack the capital to purchase a completed house outright. In 1991, one study in nine Asian countries concluded that between 40 and 95 per cent of all households had no possibility of living in a dwelling produced by the formal sector.4 Estimates of cement producers conclude that 70 per cent of housing investment in Mexico is occurring incrementally.5 In Tanzania, it is estimated that 98 per cent of the housing stock in urban areas is constructed on an incremental basis.6 This is unchanged from the figures quoted for 1978.7 In the Philippines, a similar estimate is that 93 per cent of owner-occupied houses have been built through an incremental building process.8 Such figures emphasize the significance of incremental development.

This chapter considers a distinct approach to delivering small loans to low-income households (almost universally in the South): microfinance. Another significant approach – community funds – is discussed in Chapter 7.9 Microfinance loans are almost universally to individuals, generally to those with some security of tenure, for investment in housing (construction, improvement and extension). Chapter 7 examines the community fund approach in which lending is typically to communities for land purchase, infrastructure and service investment, and (in some cases) for housing construction. In both cases (although most likely in the first), shelter lending may be accompanied by opportunities to borrow for micro-enterprise development. Prior to considering these lending strategies, the strategy of incremental development of shelter and the strategies used by the poor to secure finance are introduced.

Incremental development

For individuals or households with limited incomes, the only possibility of homeownership (even in an illegal settlement) is through shelter investment made in several stages. Land purchase, service installation and upgrading, as well as housing construction, consolidation and expansion, are all made at separate times. For higher income households, the land purchase may be first, with subsequent investments made as incomes increase and assets accumulate over a period of years. In the lower income families, the first investments may be in shelter on a piece of land with uncertain security. Subsequent investments are made as security increases. Infrastructure may be installed (perhaps with state assistance). A shack may be transformed into a more robust dwelling, with rooms being added, and flooring and roofing improved with the use of permanent materials.

Such incremental shelters, often initially built of temporary materials, frequently require repairs because of damage – for example, from natural forces. In Hyderabad, about one quarter of a sample of 224 households had recently repaired their house.10 No less than 64 per cent had repaired the roof (essential against monsoon rains). Box 6.15 on the Grameen Bank's loan package highlights the high cost of repeat repairs for houses built of traditional materials with loan finance, scarce funds can be allocated more effectively.

Box 6.1 describes the strategies that are used in Hyderabad, India. The lack of contact with formal financial institutions that is illustrated within this example is evident in many parts of the world.

Despite its significance, incremental development may be discouraged by more formal housing finance agencies. The Kenyan Banking and Building Societies Act explicitly forbids financial institutions from lending for plots of land with no or partially constructed housing on it.11 Households allocated land by the state Self-help Housing Agency in Botswana were expected to replace traditional building materials within two years – a very short period of time for those with low incomes to accumulate sufficient funds.12 One study of housing strategies for the poor in Zimbabwe also highlights the resistance of some politicians and residents to incremental housing.13 The lending conditions of the Housing Finance Company of Uganda require land title, together with a number of further conditions: the development must be located in an urban area, have full services, be constructed of permanent materials and have local authority approval for construction.14

In general, this resistance to incremental housing by formal finance companies is because of the risks associated with the building processes (particularly potential illegality) and because of uncertainty about house value and, hence, problems of mortgage valuation. However, one general policy concern about incremental strategies is that investment is wasteful because small improvements are made that might have to be repeated when a further extension is added.15 However, the financial implications are also clear. Low-income households cannot afford to pay the high interest charges on a complete loan, but are more likely to be able to cover the relatively small interest charges from repeat borrowings of much smaller amounts of finance.

Low- and many middle-income urban households in the South use incremental strategies

Access to financial services

What research and practice during the early 1990s emphasized was that the quality of self-help investment could be enhanced by financial institutions that enabled the accumulation of savings and/or offered small loans. However, little finance is available for the poor in the South. Several examples from different countries all point to the high dependency of the poor upon non-mortgage sources of housing finance.

In India, according to the National Statistical Survey's (NSS's) 44th round survey, more than 80 per cent of housing finance comes from private savings, sale of assets and non-formal sources of credit.16 In a number of households studied in Hyderabad, 45 per cent of those living in 13 low-income settlements were in debt for housing (but less than 2 per cent borrowed from formal financial institutions).17 This is higher than that reported in low-income settlements in Amritsar, where it was estimated that 10 per cent of the credit taken out by low-income households was for housing.18 A further example comes from South Africa, a country widely noted for having an extended financial sector. As noted in the discussion of mortgage finance, within one group of low to lower-middle earners in South Africa, only 38 per cent had applied for finance, with 13 per cent being successful.19 For those unable to secure mortgages, in the non-mortgage housing finance sector 89 per cent of loans by value are personal loans secured by ceding a pension and payroll deduction (only available to formally employed persons). The remaining 11 per cent of housing loans by value are unsecured personal loans. Approximately 60 per cent of South African households fall into income and employment categories that would make them potentially eligible for only this kind of loan under current South African conditions.20 And in a further African example, the overwhelming source of housing finance in Tanzania during the 1980s were people's own savings, and this was true for the formal and informal sectors.21

Box 6.1 Housing finance for low-income households in Hyderabad, India

What are the strategies of low-income households for obtaining housing finance? This question was studied in detail by Peer Smets in Hyderabad, India. This Indian city was chosen because it is in a state that had moved away from a managed economy towards liberalization. The information was collected between 1993 and 1996 in 13 low-income neighbourhoods, each with between 76 and 530 households. There were no housing schemes by external agencies in the chosen settlements. The focus was on low-income groups, notably the economically weaker sections (which at that time had a monthly income of below 1250 rupees) and the lowincome groups (with a monthly income of between 1250 and 2650 rupees).

In 2002, the city population was about 5 million, with considerable numbers of the poor squatting or living in illegal subdivisions. Despite the interest during this period by the national government in exploring the role of innovative savings and lending instruments, both with regard to housing and, more generally, the housing finance systems, practices and outcomes in Hyderabad could only be understood with reference to local land and housing markets. Important factors included how competing elite interest in land development have been reconciled, the presence of ongoing ethnic tensions between Muslims and Hindus, local political interests, and the changes that resulted from Hyderabad being made the capital of the newly formed state of Andhra Pradesh in 1956. By the 1970s, the land market in Hyderabad was uncertain, with an inadequate registration process and many disputes. In 1970, 60 per cent of residents lived in rental accommodation, and by 1981 this percentage had fallen to 55 per cent. In 1981 the population living in illegal settlements was 19.6 per cent but ten years later it had increased to 29 per cent.

Considering the urban poor in the 13 study settlements, 53 per cent of the households are above the poverty line and 47 per cent below. Some 38 per cent are tenants and 62 per cent are homeowners with no significant differences in income. The physical quality of the shelter can be divided into kaccha (traditional materials, corrugated iron, cloth and wood) semikaccha (partly or completely constructed with concrete or cement-plastered walls, with asbestos or iron sheets or tiles on the roof) and pucca (concrete or brick masonry with a concrete roof). Over 80 per cent of tenants and homeowners are living in semikaccha houses, with only 4 per cent of the sample living in pucca houses.

A sample of 242 households has been surveyed in greater detail. These are either homeowners or tenants who have bought elsewhere in the city. Sixty-five per cent of those with land in the low-income settlements (illegal land) have bought their rights. About half made a single payment for the land and the other half paid in instalments. In terms of construction materials, just under 50 per cent have made some use of second-hand materials. Fortyfive per cent are in debt because of an investment in housing, of which the majority live close to the poverty line. The biggest single source of funding for the first step of incremental building is savings and the second most significant source is friends/relatives/neighbourhoods. This is closely followed by the third source of finance: chit funds. Chit funds involve a given number of participants who each commit to paying an equal monthly amount. There are a number of different systems for selecting the order according to which members receive the funds. For second investments in incremental housing, savings and money lenders/pawnbrokers are the most important, and for further steps, savings is the source of finance in 75 per cent of cases. Considered across all financing stages, employers are a fourth source of funds. Finance or credit co-operatives and banks are used very rarely, if at all.

Source: Smets, 2002.

The importance given to savings is repeated elsewhere. A study of 198 households in low-income settlements in Pereira, Colombia, found that savings was the most common method of financing land purchase and construction, with only 10 per cent of households using loan finance.22 In Botswana, savings were once more found to be a critical source of funding for housing investment, with few other alternatives being used.23

This information points fairly clearly to a lack of housing finance. Would more finance increase investment in low-income areas and assist in a more speedy development of incremental housing? The evidence is somewhat mixed. In one area in Colombia, about one third of these households could secure public-sector loans as they had plots in a sites-and-services project.24 Despite this possibility, these groups did not have a higher incidence of borrowing, and even within these projects those who secured loans did not appear to be faster at consolidating their housing than other households. Another scheme in Mauritania provides land security to the urban poor, together with further assistance for development.25 The assistance programme includes housing finance, technical assistance for enterprise development and literacy and skills-enhancement classes for residents. The housing finance package is divided into three components: room, latrine and perimeter wall. Participants make a deposit of 25 per cent of the cost, the municipality gives a subsidy of 25 per cent and the remainder is repaid over two years at 0 per cent interest. Due to poor uptake, subsidies were increased for the second phase (which started in 2002). During the first 18 months, the programme was successful in increasing housing development. Those who did not obtain shelter finance but who secured land from the programme invested an average of US$178 in their housing, while those who secured the loan and subsidy invested an average of US$349. Clearly, affordability remains an important issue. At the same time, the speed with which some microfinance initiatives for shelter have grown suggests that, in at least some households, there is a considerable demand for loan capital.

What is also notable is that limited access to mortgage finance means that there are few alternatives to incremental development for households wishing to secure housing. The problem is exemplified by the low-income settlements in South Africa where, even if someone wants to sell a house in a low-income settlement (even with legal tenure), it is difficult to secure mortgage finance. There are no financial products in South Africa appropriate for those who wish to purchase housing that has been developed incrementally from a sites-and-services programme, or that has been recently constructed and financed under the capital subsidy programme or formal houses built for Africans between 1948 and 1960.26 Current lenders to low-income households offer secured and unsecured microloans and pension-backed loans of between 5000 and 15,000 rand (US$775–$2325). This is not enough to purchase existing houses in any of the housing sub-markets mentioned above: the mean selling price for houses in each is between 13,000 and 52,000 rand (US$2000–$8000). The specific problems resulting in mortgage refusal include lack of adequate land title, insufficient income by the purchaser, lack of formal employment and red-lining (the refusal to issue mortgages in specific areas) due to generalized problems of foreclosure.

This is the situation in which low-income householders find themselves. Unable to afford fully developed houses with established legal title, they develop housing incrementally. In the absence of commercial or state finance for complete houses, they invest when and as they are able. To fill this gap, a number of different initiatives have developed. This chapter looks particularly at the provision of small loans for shelter to individual households provided primarily, but not exclusively, by the group of microfinance agencies that emerged during the 1980s and 1990s to supply enterprise finance. It concentrates on the financial sector that works primarily with individuals, while Chapter 7 considers the separate but related tradition of community finance. This chapter focuses largely on microfinance institutions because that is where considerable innovation has taken place, and where there are indicative signs that further value might be added. However, the discussion also considers other small-scale lending through civil society groups, such as credit unions, as well as commercial loans from small commercial lending firms and building material suppliers.27 Chapter 7 then turns to community funds, a further financial strategy developed within socially orientated development agencies working particularly on urban poverty and/or addressing housing need.

Table 6.1 highlights the differences between these two strategies and those of mortgage and microfinance for enterprise development. In essence, community funds seek to address the needs of poorer groups and, thus, use collective loans both to build the capacity of the poor to act together and because the priorities of secure land tenure and infrastructure cannot be afforded individually. Shelter microfinance responds to the needs of the poor with reasonably secure tenure to upgrade their dwellings using strategies that have developed for lending to small and medium-sized enterprises.

Table 6.1

Lending strategies for housing development

Mortgage finance Microenterprise finance Shelter microfinance Community funds
Objective Provide long-term housing finance Provide investment finance for enterprise development and enable income growth Provide housing improvement and improve well-being Enable the poor to secure shelter assets, particularly land and infrastructure
Borrowers Upper- and middle-income households Micro- and small entrepreneurs Those with land who need to improve the dwelling Those without secure tenure, basic services and adequate housing
Use of loan funds Acquisition of property Development of business Housing improvement Land, infrastructure and occasionally housing improvement
Role of savings Deposit required; savings process not important May be required Savings may be required; deposit may be required Savings generally essential; deposit may be required
Additional support Irrelevant Generally not Possible Nearly always considered necessary because of complexities of land development
Attitude to the very poor Avoid Generally avoid; some specialist programmes Depends upon orientation; but requirement for land likely to exclude the poorest Generally seeks to help the very poor if they are residentially stable
Purpose of the collective (community organization) None May be used as guarantor May be used as guarantor; sometimes additional community support is a part of the process Lending is collective and the role of the group is seen as essential to address the exclusion of the poor
Amount Generally over
Generally under
Generally between
Generally under US$1000
Interest rate Inflation plus a margin
of 8–15%
Inflation plus a margin of
Inflation plus a margin to cover costs of 10–20% Inflation plus administration
Term 15–30 years Less than 1 year 1–8 years 3–20 years (generally shorter)
Collateral Mortgage Personal guarantees,
goods, co-signers
Personal guarantees,
goods, co-signers, mortgage
Can be title deeds but emphasis placed on collective loan management
Financial sustainability Generally considered essential, but may be state subsidies Desired – support for product development Desired – support for product development; occasionally integrated with subsidies for land development Seek state support to offer subsidies for land development and services in order to include lower income families
Linking role None To other financial institutions To other financial institutions; may involve the municipality in slum upgrading programme To state and municipality
Source: adapted from ACHR, 2002, p6, and Ferguson, 2004b, p5.

Microfinance: what is it?

As emphasized above, many in the South develop housing incrementally and the need for small loans is considerable. An estimated 70 per cent of housing investment in developing countries occurs through such progressive building.28 Microfinance for shelter addresses a gap that larger-scale mortgage lenders are unwilling to provide for and, arguably, for which they lack the skills and capacities.

Table 6.2

Consultative Group to Assist the Poor (CGAP): typical terms and conditions for shelter microfinance

Size Varies, but generally two to four times larger than average working capital loans
Term Usually 2 to 24 months for home improvements and two to five years for land purchase or construction
Interest Same as standard working capital loans or slightly lower
Delivery method Almost always provided to individuals rather than to groups
Collateral Mostly unsecured; co-signers often used; real guarantees may be used; formal ownership of dwelling or land may be required; savings sometimes used as guarantee (may be compulsory)
Target clientele Low-income salaried workers; microentrepreneurs primarily in urban areas; poor people
Other services Sometimes accompanied by land acquisition, land registration and construction (including self-help building techniques)
Source: CGAP, 2004.

The main issues discussed in this section are the growth of shelter microfinance, including the sources of funding, the terms and conditions of lending and the challenges that this sector faces.

Microfinance for shelter offers small loans suitable for significant housing improvements. Terms and conditions are summarized in Table 6.2. Loan sizes are between US$1000 and US$5000, although they may be smaller in some countries where construction costs are lower and/or building standards do not prevent low-cost housing options. Loan terms are generally between one and eight years, although in most cases they are at the shorter end of this range. Hence, although these loans are often given by existing microfinance lenders and are seen as falling within this category of financial services, they are often considerably larger than enterprise loans (especially those taken by new borrowers when entering this market).29

Shelter microfinance responds to the needs of the poor with reasonably secure tenure to upgrade their dwellings

Security conditions vary considerably depending upon local circumstances. In some cases, they are similar to those required for enterprise development (that is, group guarantees and co-signers). In other cases, they involve holding the para-legal documents to the property and other non-mortgage collateral. Some shelter microfinance lenders follow a process similar to that of a conventional mortgage for larger loans. The Consultative Group to Assist the Poor (CGAP) is a consortium of 28 public and private development agencies seeking to expand access to financial services (microfinance) for the poor in the South. In 2004, they recognized the significance of shelter microfinance with a briefing for members.

Microfinance for shelter addresses a gap that larger-scale mortgage lenders are unwilling to provide for

Loans are generally taken to build additional rooms (often turning space constructed using wood and traditional materials into concrete built structures), improve roofs and floors, and add kitchens and toilets. Investing in improved facilities is very popular and the Self-employed Women's Association (SEWA) in India estimates that ‘almost 35 per cent of housing loans from SEWA Bank are utilized for installing infrastructure, such as a private water connection or toilet’.30 The emphasis is very much on improvements for homeowners. The terms and conditions of microfinance lending in the context of incremental development favour those who already have some degree of tenure security and housing structure. For this reason, these loans are often referred to as housing loans, or housing microfinance. In some cases, they are also for land – for example, the Grameen Bank will lend for land purchase if the borrower does not have legal tenure. However, lending for land purchase is much less likely because of the high costs and other problems with individualized solutions to tenure and infrastructure needs, and because some degree of land security may be a prerequisite for such a loan.

There is a vibrant rental market in many low-income settlements in most Southern cities. Such rental activities are, in general, informal in addition to the fact that the income is not taxed or declared, the rental agreements are managed outside of the formal legal system. Tenants may be particularly vulnerable and may face difficult terms and conditions, with few alternative affordable options. They generally enjoy restricted access to services.31 In some cases, microfinance loans are used by the landlords to construct additional rooms for rent. However, there is not much information about such purposes and it does not appear to be happening at scale. In one housing loan scheme in low-income settlements in Mauritania, two-thirds of households used the home for some kind of enterprise activity, including renting space. The percentage renting space to others among the group who took housing loans is twice the percentage of those who did not take loans – but it still remains low at 6 per cent of households.32 In a few cases, small loan programmes have been orientated towards the landlord sector to improve the living conditions of tenants. However, there are relatively few intentional initiatives of this kind. One difficult issue is that, although the project may be intended to improve living conditions for tenants, in practice, the improvements may be associated with rent increases and the displacement of one (poorer) group of tenants by another (higher income group). Box 6.2 describes a small revolving fund in Kitale, Kenya, which offers loans for improved sanitation to plot owners, many of whom are also renting rooms. The objective is to improve environmental health, although the risk of potential rent increases and the displacement of tenants is recognized.


The growth of microfinance agencies since their inception during the 1980s has been considerable and there are now many such organizations. To exemplify the situation in one country, in India the number of such grassroots-level organizations engaged in mobilizing savings and providing microloan services to the poor is estimated to be in the range of 400 to 500 organizations.33 However, some 60 million families in India (approximately 36 per cent of the country's population) are in need of financial services, while the cumulative outreach by microfinance agencies is no more than 1.5 million households (2.5 per cent).34

The developments in shelter microfinance follow the development of a growing microfinance sector. During the 1980s, several agencies demonstrated success in offering small loans for enterprise development. The underlying and emerging argument was that small entrepreneurs were constrained by a lack of credit. The availability of credit, it was argued, would enable businesses to expand and development opportunities to emerge from within the small-scale, invariably informal, business sector.

Early and continuing evaluations demonstrated that, whatever the loans were taken for, a proportion as large as 25 per cent could be diverted for shelter investments. For example, Centro de Fermento a Iniciativas Economicas (FIE), a Bolivian microfinance agency, estimates that 20 per cent of its enterprise loans are allocated to housing investments.35 An assessment made by the Association for Social Advancement (ASA) in Bangladesh suggests that 15 per cent of borrowers for income generation use these loans for improving housing.36 Findings such as these have encouraged the exploration of microfinance lending specifically for shelter. Box 6.3 describes how one microfinance agency sought to develop a specific product to address the housing needs of borrowers.

Although much emphasis of the early microfinance lending was on enterprise development, shelter has been a possible reason for lending since the mid 1980s, much the same time as enterprise lending was expanding. For example, the Grameen Bank started lending for housing in 198437 while in 1985, for example, the US Agency for International Development (USAID) offered finance for the Co-operative Housing Foundation to implement credit programmes in a number of Central American countries.38 The foundation developed its work with local organizations such as CACIEL (a credit union in Honduras) to expand shelter lending. Between 1985 and 1990, US$11 million was invested in activities with 28 organizations to offer 4653 home improvement loans and a further 2828 construction loans.39 Within the same programme, experimentation also occurred (on a much smaller scale) with community loans for infrastructure improvements (such as water systems).

There are a considerable number of NGOs that have been working with housing issues, generally for lower income groups, and that have been drawn into loan financing in order to scale up their activities and/or to provide assistance to residents who have been successful in acquiring land. In such cases, NGO loan programmes are part of a more substantive programme to improve housing conditions that may involve the provision of technical assistance community development training grants for improving infrastructure and services building materials production and support in negotiations with local authorities.40 Most of these initiatives emerged from Southern NGOs seeking to address the needs of the poor more effectively. Agencies working with housing, urban poverty and urban development issues were aware that self-builders faced major problems in securing the finance they needed for incremental development, and the NGO professionals were also aware of the long-term cost of short-term temporary improvements.

Shelter NGOs looked to the examples of microfinance agencies seeking to bring financial markets to those who traditionally had been excluded from opportunities for savings and credit. Others concentrated on the individualized lending systems of enterprise microfinance, but orientated the loans towards housing improvements. One example of this heritage is Proa in Bolivia, an NGO that started work in 1988 with a concentration on urban development and which evolved a programme of housing loans (see Box 6.4).41

Box 6.2 Sanitation revolving fund in Kitale, Kenya

The sanitation revolving fund has been initiated by the Intermediate Technology Development Group in two settlements in Kitale (Tuwani and Shimo la Tewa). The first phase has included 23 loans, all to plot owners, some of whom rent rooms within their plots. Many plot owners wished to take loans and the successful applicants were selected on the basis of their willingness to accept the loan in the form of materials, as well as according to their capacity to contribute towards the cost. The loans are to be repaid over two to three years. The amounts loaned are between 27,000 and 60,000 Kenyan shillings, and the interest rate charged is 12 per cent (if the repayment period is two to three years), or 11 per cent for a one-year repayment. A one-month grace period on repayments is offered. To assist in securing repayments, an affidavit has to be signed by each recipient. A further incentive for repayment may be that people have bigger dreams (better housing) and seek further opportunities to borrow. A remaining question is whether they see the additional facilities as an opportunity to raise rents.

The Catholic Diocese of Kitale has agreed to manage the sanitation revolving fund on behalf of Intermediate Technology Development Group. The diocese already has some expertise in microfinance. A board of trustees oversees the loans and includes three members from the diocese, along with community members.

Source: L. Stevens, pers comm, 2004.

Even within the housing NGO sector, there are two distinct groups of such NGOs working in housing finance in Mexico.42 The first group is professional urban development NGOs who have primarily been drawn into finance programmes in order to influence state policies and the demands of low-income communities. Such programmes are illustrated in Box 6.6.43 The second group are humanitarian agencies who have worked to improve housing conditions in low-income areas. Recognizing that families are able and willing to invest in their own dwellings, they have directly developed small loan programmes at scale. Their work is illustrated in Box 6.5 (it is estimated that households below five minimum salaries cannot afford a fully completed dwelling paid for with unsubsidized mortgage finance).44

Box 6.3 Launching a housing microfinance product: Mibanco, Peru

With 70,000 active borrowers, Mibanco in Peru is one of the largest microfinance institutions (MFIs) in Latin America. The organization started as a non-governmental organization (NGO), but became a commercial bank in 1998. The conversion into a deposit-taking institution gave Mibanco the funding necessary to expand from microenterprise lending into other areas. During mid 2000, Mibanco added a housing product, Micasa, in the form of a loan for improvement, expansion, subdivision, or rebuilding or replacement of existing housing.

After 12 months of operation, Micasa had 3000 clients, with portfolio at risk greater than 30 days of 0.6 per cent and a return on loan portfolio of 7 to 9 per cent. Loan size ranged from US$250–$4000, and averaged US$916. Interest rates were 50 to 70 per cent per annum. These rates are less than those Mibanco charges on microenterprise loans. Loan periods were as much as up to 36 months but most households preferred loans of 6 to 12 months, and the average loan period was 11 months. Mibanco uses its analysis of repayment potential and household assets to guarantee most loans. Mortgage liens are sometimes taken, but only on larger loans (those above US$4000) if the client already has clear legal title. In total, mortgage liens secure only 7 per cent of Mibanco's home loans. The housing loan product has strong profitability and demand, and Mibanco expects such loans to represent half of its portfolio within three years.

Source: Ferguson, 2003.

Box 6.4 Proa: learning how to offer shelter loans

When the non-governmental organization (NGO) Proa in Bolivia moved into housing finance in 1991, its original strategy was small loans for home improvements using solidarity groups to guarantee repayments. For this and subsequent strategies, it secured funding from Mutual La Paz, a mutual savings association. This first strategy failed and Proa was forced to cover some of the losses with Mutual La Paz however, there was an enthusiasm to carry on.

The second strategy was to use some form of individual guarantee using landownership. The costs (and time) of registering a mortgage with the Office of Property Rights were considered too high but even without this measure some claim over the property could be secured. In addition, procedures to follow up repayments were strengthened. This system worked relatively well and a refined, but broadly similar, strategy was introduced in 1993.

Most loans are for housing improvements including access to water and sanitation services. However, some are for the regularization of properties and new construction by small contractors. One measure of success has been that foreclosure and late payment rates are now below those for Mutual La Paz's overall mortgage lending and are low for Bolivia. Repayments that are more than 90 days late account for 1.09 per cent of the portfolio compared to 4.1 per cent for Mutual La Paz's middle- to higher income mortgage lending.

Source: Ferguson, 1999, p193.

However, it is also important to recognize that not all housing and urban development NGOs have chosen to develop financial services for housing. In Mumbai, for example, at the end of the 1990s there were 18 NGOs addressing issues broadly related to housing and community finance, but only four specifically providing housing loans.45 Urban development NGOs in Mexico have tended to develop housing finance initiatives as exemplar projects, not necessarily intending to take them to scale, but seeking to use the experiences to influence the ways in which housing policy is being developed.46 This strategy extends well beyond Mexico, and many other NGOs who are advocating for more successful housing policies and strategies have introduced demonstration projects to show the effectiveness of small loan provision.47 Another significant and influential group are the community organizations and their representatives, and NGO-initiated programmes have sought to switch community demands away from clientalist favours and towards effective development interventions that can go to scale. The Step-by-Step programme in Ecuador and Peru is a recent example of such a programme (see Box 6.6).

Box 6.5 Humanitarian housing interventions in Mexico

One of the most significant of the humanitarian housing agencies operating in Mexico is Habitat para la Humanidad México A.C., the Mexican branch of Habitat for Humanity. This agency began operating in Mexico in 1987 and currently works in the federal district and 13 other states, with 20 active affiliate groups. The organization provides credit to previously formed mutual aid groups of selected families, who supply the labour for their own and other group members’ house construction on their own land. Until now, it has financed 14,388 houses in 600 communities in both rural and urban areas.

Another non-governmental organization (NGO) which provides home improvements finance for workers in the bonded industries (maquiladores) in Ciudad Juárez, on the Mexico–US border, is Fundación Habitat y Vivienda A.C. (FUNHAVI). This was set up in 1996 as a branch of Co-operative Housing Foundation International with the help of a Ford Foundation donation as seed capital. It is also sponsored by the Inter American Foundation, from which it received two loans for a total of US$500,000 last year and donations from local businesses. Its target population comprises homeowners earning between two and eight times the minimum wage, although the average income of beneficiaries was four minimum wages in 2001; 38 per cent were women. The same source quotes that loans ranged from US$500 to $2500 (average loans of US$1623), interest rates were 2.5 to 3 per cent a month, and a 2 per cent commission is charged by the organisation, as well as a US$20 mandatory technical assistance fee. The technical assistance provider or ‘architect’ decides what sort of loan is needed. Loan terms vary from 6 to 36 months, with repayments being paid monthly at the local supermarket chain, with which FUNHAVI has a special arrangement. Another special arrangement with construction material distributors enables FUNHAVI to purchase them at wholesale prices, although recipients of the loans have to buy them from FUNHAVI at retail prices this covers 11 per cent of FUNHAVI's running costs.

Source: Connolly, 2004b.

As illustrated by the Step-by-Step programme, in the case of some NGO programmes the desire to influence policy is combined with a wish to respond to the needs of those seeking to improve their housing and to improve access to loan finance. Despite widespread discussions about the value of microfinance, need remains acute, and in many cases there are few providers. In Ecuador there is a subsidy programme but as illustrated in Box 6.6, the requirement for a savings contribution means that it cannot be accessed by many low-income households.

A further illustration of the continuing responsiveness of the NGO sector is given by the launch of South Africa's Kuyasa Fund (see Box 6.12). In South Africa, there is considerable state subsidy for housing provision and a commercial banking sector that has been under significant pressure to expand lending to the poor. Hence, the context appears to be one in which there are opportunities for low-income communities to secure both housing improvements and financial services. Despite this apparently favourable context, there is a further need for small housing loans. A Cape Town-based NGO, the Development Action Group, launched the Kuyasa Fund in 2001 after beginning trial housing loans in 1999. It did so because the communities with whom it was working needed finance to upgrade their dwellings, and there were few alternative accessible and affordable sources of finance.

Links to formal financial agencies

During the 1980s, some programmes had the explicit intention of preparing their clients for entry into formal housing finance either in the short or the longer term.48 There was an underlying expectation that the poor could borrow from the formal financial systems once appropriate modifications had been identified and implemented. For example, in the case of the Central American programmes supported by the Swedish International Development Agency (SIDA), links have been sought. However, in practice, it has proved difficult to convince such formal financial institutions that they should participate in direct lending this is due, in part, to the small loan size and associated high administration costs.49 Generally, this expectation has changed and there is now greater recognition that it might be preferable to build significant institutions that specialize in small loans. Such institutions might link to the more formal commercial financial institutions to secure capital but the formal financial institutions would not be expected to interact directly with the poor.

Box 6.6 Step-by-Step programme in Peru and Ecuador

The Step-by-Step programme is located in Peru and Ecuador and is being implemented by Centro de Investigaciones (CIUDAD) and Centro de Escudios y Promoción del Desarrollo (DESCO), with contributions from the European Union (EU), Centre for Ecology and Development (KATE), Instituto de Estudios Políticos para América Latina y Africa (IEPALA), Alternativas Sostenibles de Desarrollo, España (ASDE) and La Asociación para la Cooperación con el Sur/Las Segovias (ACSUD-Las Segovias). Activities include the establishment of a revolving loan fund with the related construction of safe and affordable housing through incremental development and the promotion of savings among participants. The revolving loan fund for shelter production seeks to establish a credit system that is adaptable to, and appropriate for, the needs of self-building families. Technical assistance in the building of affordable, healthy and safe houses is also being provided. Training programmes for the people involved in the construction of low-income housing are offered, together with the dissemination of good practices on progressive shelter financing schemes through a training and dissemination centre. In addition to the direct benefits, a further intention is to improve housing policies for low-income families through the targeted dissemination of the programme's achievements.

The total budget for the programme is 1.8 million Euros. In Peru, the project is taking place in Villa el Salvador (part of metropolitan Lima), with its 1 million residents. In Ecuador, the project is located in Quito, Riobamba, Alausí and Cotacachi. In total, 0.54 million Euros are allocated to the revolving fund and it is anticipated that just over 2000 loans will be provided.

In Ecuador there is a national housing subsidy system that offers families a grant of US$1800. However, families have to be able to save 10 per cent of the value of the house to qualify, and experience suggests that it is difficult to save the required amount. Generally, they are not eligible for loans and they do not trust formal credit institutions. As a result, one use of the revolving fund is for the down payment to secure the subsidy. Additional uses are part payment for new houses in existing housing programmes, housing improvement, down payments or deposits for commercial loans and providing community facilities.

Local financial strategies involving the fund can be exemplified through the women's association Luchando por la Vida. The 36 families have an average income of US$185 per month, with 94 per cent of households falling below the official poverty line (US$360 per month). The housing programme involves the construction of 6 buildings and 36 apartments (of 60 square metres at a unit cost of US$5100). The total costs are paid thus: 33 per cent by the government housing subsidy and 67 per cent by the families (using a combination of savings, commercial loan and a Step-by-Step deposit to access the housing subsidy).

From 2001 to 2004, Step-by-Step in Ecuador has granted more than 930 loans and 550 families have secured new houses of good quality. In addition, 62 per cent of the users of the loans are women, and 72 per cent of families who have secured houses through the programme have incomes below the poverty line. Step-by-Step's loans (for a total amount of US$750,000) have already mobilized more than US$2.5 million from government subsidy (25 per cent) and private bank loans (75 per cent).

Source: M. Vasconez, pers comm, 2004.

There remains the tradition of guarantee funds, although their use is somewhat limited to a few specific examples, and scaling up such examples into regular practice appears difficult.50 A number of NGOs have specifically sought to use guarantee fund strategies to release financial capital from the formal (mainly commercial) financial sector. Such guarantee systems have a dual rationale. On the one hand, they are intended to build links between the formal and informal financial systems, encouraging further lending (with no or lower guarantee ratios) once a positive experience has occurred. Second, they are a way of leveraging finance if the guarantee is accepted to be less than 100 per cent. Examples of guarantee funds include the Latin American and Asian Low-income Housing Service (SELAVIP), the Society for the Promotion of Area Resource Centres (SPARC), Homeless International and a number of other Southern NGOs. Such guarantees can be illustrated by SPARC (an Indian NGO), the state-financed Housing and Urban Development Corporation (HUDCO) and a housing co-operative in Dharavi (a large low-income area in Mumbai).51 In this case, the funding was only released after SPARC guaranteed the repayments (with the financial support of a Northern NGO) and 25 per cent was withheld from the first loan instalment as a contribution to interest payments. More recently, SPARC has had more successful experiences based on the increasing interest of the private sector to find a way of investing in the development of lowincome urban areas.52 One experienced commentator concludes that despite difficulties around the release of additional finance for the local development activities of NGOs and community organizations ‘it is too early to give up on banks yet’.53

In this context, the more recent interest by commercial financial agencies, such as the Colombian Banco Davivienda, in developing a small loan facility is notable.54 However, what is not yet clear is the extent to which the state will have to support such initiatives. What is evident from the following discussion is that the commercial banking sector in some countries is seeking more involvement in what is considered to be a potentially profitable sector.

Microfinance agencies appear to be diversifying rapidly into housing micro-credit in at least some regions

The microfinance institution (MFI)
experience: enterprise to housing loans

In addition to NGO initiatives, there has been considerable interest in housing lending shown by the microfinance sector. It is difficult to assess the significance of the growth of microfinance agencies into small shelter lending, but it appears to be significant. For example, three significant microfinance agencies were profiled in a 1996 study when, at that time, none of them were working in housing.55 Four years later, two of the three were working in this area. Microfinance agencies appear to be diversifying rapidly into housing microcredit in at least some regions.

Box 6.7 Assessing the demand for housing finance

Demand for housing and small-scale lending for housing investment is likely to increase, as is illustrated by the following examples:

  • •     Peru: 82 per cent of the 8 million people living in greater Lima are classified as poor. At least half of poor households and 60 per cent of the poorest households express a strong desire to expand or improve their home within the next 12 months. Only 10 to 15 per cent are borrowing from formal or informal sources. The potential market in metropolitan Lima for housing finance loans is estimated at 610,000 home improvements annually.
  • •     Indonesia: during 2000, the country's urban population of 85 million already represented 40 per cent of the total. By 2010 it will represent 50 per cent, with 120 million people. Annual projections for housing needs for the next ten years are approximately 735,000 new units and an additional 420,000 in need of improvement. An estimated 70 to 80 per cent of all housing in Indonesia is constructed informally and incrementally, with minimal access to formal financial markets.
  • •     Morocco: two surveys found that 88 per cent of households have or are planning a productive activity in the home, and more than 83 per cent of households are willing to take a loan to finance home improvement. Ninety-two per cent of urban and 94 per cent of rural households constructed their own homes without access to formal finance.
  • •     Mexico: one market study of microfinance in three Mexican cities (Tijuana, Matamoros and Juarez) bordering the US found that 14 per cent of all households both qualified for and wanted housing microcredit at terms of 35 per cent amortized over three years. The effective demand for shelter microfinance (US$122 million) amounted to five times that for microenterprise loans (US$20 million) in these cities.

Source: Malhotra, 2003, pp218–219.

A further potential role for shelter microfinance is within more comprehensive slum upgrading programmes

Over the last three years, most leading microfinance agencies in Latin America and the Hispanic Caribbean have established a housing product. Cases in point include Banco Sol in Bolivia, Banco Solidario in Ecuador, Mibanco in Peru, Banco Ademi in the Dominican Republic, Calpia in Honduras, and Genesis Empresarial in Guatemala.56

One study funded by the International Finance Corporation (IFC) identifies 141 institutions providing shelter-finance loan products to the poor.57 Another, focusing on Latin America, identifies 57 microfinance agencies as offering housing loans, just under 30 per cent of the total number of such institutions.58 Of these agencies, about 18 per cent of their total loan portfolio is related to housing loans, amounting to about US$160 million.59 Among the 27 financial institutions in the Accion Network, seven have housing portfolios totalling almost 10,000 active clients and US$20 million in outstanding balances.

The speed with which housing loans have been integrated within such agencies appears to have been facilitated by the similarity of lending practice. For example, in the case of one Peruvian agency, Mibanco, adding a home improvement loan product was easier than originally anticipated.60 Traditional microfinance agencies treat housing loans broadly as they treat microenterprise lending, with small repeat loans that are not (in many cases) conditional upon collateral.

Reasons for expansion of MFIs into housing

One reason for the diversification of microfinance agencies into housing is commercial advantage. Such diversification may increase the financial stability of their loan portfolio, enable them to take advantage of opportunities for growth, and avoid losing clients to other microfinance agencies that provide housing loans.61 A further notable advantage is that the longer repayment period associated with housing loans helps to draw the borrowers into a longer-term relationship with the lending agency and increases the likelihood that further loans will be taken (for example, for enterprise development). Thus, lending for land and housing has commercial benefits for a microfinance industry seeking to extend its niche and strengthen performance. The need for diversification may be particularly important in countries such as El Salvador and Bolivia, in which microfinance agencies are facing considerable competition for clients.62

It appears likely that there is significant scope for expansion, at least in most of Latin America and Asia. Given the scale of housing need, microfinance for shelter remains significantly underdeveloped in many countries in which market conditions appear favourable, such as Mexico and Brazil.63 In Central America, the SIDA has been financing a number of market assessment studies to identify what people want, both in terms of demand for housing loans and other financial services beyond credit.64 Generally, demand has been diverse and has included infrastructure loans, as well as demand for microinsurance and housing. The market may also be significant in Africa but it is likely that the income group will be different. In Africa, where many of the middle class may not be able to access formal loans due to land title problems, microfinance may not reach down the income groups so far and scale may be smaller but still valuable. Box 6.7 summarizes a recent analysis of the potential for growth in a number of countries. These assessments are only indicative of potential. However, they illustrate some of the reasoning that lies behind new initiatives in shelter microfinance.

Neighbourhood improvement
(slum upgrading)

A further potential role for shelter microfinance is within more comprehensive slum upgrading programmes. There appears to be a growing interest in using microfinance agencies to provide specialist financial services within more comprehensive neighbourhood improvement and poverty reduction programmes. Within this strategy, the development agency, central government and/or municipality finances a process to upgrade the low-income area with components to regularize tenure and provide and/or upgrade infrastructure and services. The upgrading programme then contracts with an organization to offer small-scale housing loans for those who wish to upgrade their homes. At the broadest level, such programmes are similar to best practice elsewhere, involving local government and public–private partnerships to address housing and community development activities.65

Box 6.8 The local development programme (PRODEL) in Nicaragua

In order to address the need to improve the physical environment and the socio-economic conditions of the poor in Nicaragua, Programa de Desarollo Local (PRODEL), a local development programme, established the following kinds of support:

  • •     infrastructure and community works, including the introduction, expansion, repair and improvement of infrastructure and services through small-scale projects costing up to US$50,000
  • •     housing improvement through small loans (of between US$200–$1400) targeted at low-income families who can afford to enlarge and improve houses and to repay their loans
  • •     financial assistance to microenterprises with small short-term loans (of between US$300–$1500) for fixed and working capital these loans are directed, in particular, at microenterprises owned and operated by women and
  • •     technical assistance and institutional development to strengthen the capacities of local governments and encourage institutionalized financial entities to become involved in non-conventional lending programmes for housing improvements and microenterprise loans.

Between April 1994 and December 1998, 260 infrastructure and community projects were carried out in 155 different neighbourhoods, benefiting more than 38,000 families. Total investment has been US$4.4 million (an average of US$16,972 per project). Contributions from municipal governments and the beneficiary communities (in kind, cash, materials, tools, labour, administration and supervision) totalled 43.1 per cent, with the remaining 56.9 per cent coming from the programme. Thirty-five per cent of the projects were for improving roads, gutters and sidewalks 10 per cent for improving and expanding potable water and sewage systems 14 per cent for rainwater and storm water drainage 18 per cent for electrification (public lighting and/or household connections) and 23 per cent addressed community infrastructure (including construction, improvement, expansion and repair of primary schools, daycare centres, health centres, parks and playgrounds). The communities contributed approximately 132,000 days of work to these 260 projects, both volunteer and paid, using their own resources.

In five years, more than 4168 loans were given for housing improvements (total disbursed funds reached US$2.7 million). By 2003, the total had grown to over 11,000 loans and annual disbursements during this year exceeded US$2.5 million. These benefited approximately the same number of families. Families contributed their own resources, construction materials, labour, transportation and project administration to an amount equivalent to at least 15 per cent of the total value of the labour, transport and building materials. Seventy per cent of the families have monthly incomes of US$200 or less, including many with monthly incomes below US$100.

More than 12,451 loans to microentrepreneurs were allocated to communities in which PRODEL is active, with almost US$5.5 million being disbursed, benefiting approximately 2400 existing families. Seventy new microenterprises have been created, giving jobs to some 210 people.

Source: Stein, 2004, pp117–118 PRODEL, 2004.

A number of different variants of this model have developed. Box 6.8 describes the Local Development Programme (Programa de Desarollo Local, or PRODEL) in Nicaragua that was set up to enhance development in smaller towns and cities with a number of components, including infrastructure improvements, housing loans and loans for microenterprises. The activities received the support of SIDA, who signed an agreement with the Nicaraguan government in 1993 for the implementation of a programme to address basic needs and support development in a number of urban centres. In this case, the programme worked with Banco Crédito Popular, a state commercial bank, and selected two existing NGOs, Asociacíon de Consultores para el Desarrollo de la Pequeña, Mediana y Microempresa (ACODEP) and Nilapán-FDL, both of whom were already active in lending for microenterprises and who wished to expand their activities.66 Although the physical areas for the different components of the programme do not necessarily overlap exactly, the cumulative effects are illustrated by the change in the number of those receiving housing loans who have land titles. In 1994, only 15 per cent of those receiving housing loans had title deeds in 2002, the figure had increased to 73 per cent as the titling programme expanded.67 Although communities do not pay directly for the improvements in basic services and infrastructure, they contribute self-help estimated at 13 per cent of the costs.

A more focused (and smaller-scale approach) is illustrated in Ahmedabad, India, where the Slum Networking Project (undertaken within the municipality) wished to include a credit component to help households afford to contribute to infrastructure improvements. In establishing this programme, they drew upon the local expertise of SEWA, a local agency lending to the poor. More recently, the Parivartan Programme has been established to upgrade slums in and around Ahmedabad through the joint participation of government entities, NGOs, the private sector and low-income residents themselves. The programme was initiated by the Slum Networking Cell within the Ahmedabad city government. Parivartan means ‘transformation’ in Gujarati and Hindi. The programme seeks to offer improved infrastructure and better communication between the local residents and the authorities. It provides a water supply to every house, an underground sewerage connection, toilets in the home and an efficient storm water drainage system. Further benefits are street lighting, paved roads and pathways and basic landscaping, together with solid waste management. Costs are divided between the residents (2000 rupees, or US$42) and the municipality (8000 rupees, or US$170). SEWA helps the lower income residents with loans.68

Box 6.9 Community-led Infrastructure Financing Facility (CLIFF)
and bottom-up neighbourhood development

The Community-led Infrastructure Financing Facility (CLIFF) is an urban poor fund capitalized by donors that has been designed to act as a catalyst in slum upgrading through providing strategic support for community-initiated housing and infrastructure projects that have the potential for scaling up. The overall goal is to reduce urban poverty by increasing the access of poor urban communities to commercial and public-sector finance for medium- to large-scale infrastructure and housing initiatives. The first initiative is in India with the Society for the Promotion of Area Resource Centres (SPARC), the National Slum Dwellers Federation (NSDF) and Mahila Milan.

Scaling up citywide requires an engagement with the formal development process and the establishment of working relationships with formal-sector institutions. This is usually problematic, largely because public-sector financing is severely constrained and has a proven record of being reluctant to lend to the poor. A further problem is that the formal sector has continued to be unable to adapt their systems to accommodate non-formal investment processes. In December 2002, Cities Alliance approved a proposal to establish CLIFF with a seed capital of US$10 million from the UK Department for International Development (DFID) and an additional US$2 million from the Swedish government. Homeless International (a UK NGO) is the implementing agent and works with Samudaya Nirman Sahayak. The main function of CLIFF is to:

  • •     provide bridging loans, guarantees and technical assistance
  • •     initiate medium-scale urban rehabilitation in cities in the South
  • •     work in partnership with community-based organizations (CBOs)/and non-governmental organizations (NGOs) who have or can be assisted to develop a track record in delivering urban rehabilitation
  • •     seek to attract commercial, local and public-sector finance for further schemes, thus accelerating or scaling up the response to the challenge of urban renewal and
  • •     establish local CLIFF agencies that can operate as lasting local institutions.

The example of SEWA gives some indication of the potential for housing finance agencies to work in alliance with groups seeking sources of funds and organizational potential for upgrading. One further programme is the Comprehensive Kampung Improvement Programme (KIP) introduced in Surabaya, Indonesia, during the late 1990s and following on from earlier improvement strategies for these low-income areas. In these earlier strategies, the experience was that housing investment took place as the local environment was upgraded. In the Comprehensive KIP, revolving funds within communities have been capitalized to provide a source of finance for income generation and housing investment.69 In Comprehensive KIP 2003, some 30 per cent of the initial investment revolving fund (US$33,000) per area was used to capitalize a revolving fund specifically allocated to housing. Between 2001 and 2003, an estimated 860 households had borrowed for housing improvements.70 The delivery of housing loans was integrated with the provision of enterprise lending, as well as physical improvements to the area. Similar strategies have been used in a number of other programmes, including the Programme for Integrated Urban Renewal in El Salvador to assist in the rehabilitation of mesones in San Salvador after the earthquake. These are old houses now subdivided with tenants in each room. The programme provided for the improvement of infrastructure (with substantial finance) and then offered loans for housing improvement and microenterprise development.71

Although most slum upgrading initiatives have been led by the state, an alternative approach is that developed from an Indian alliance of SPARC (an NGO), the National Slum Dwellers Federation (NSDF) and Mahila Milan (a network of women's collectives). Their strategy is to develop the capacity of local communities to manage a comprehensive upgrading and redevelopment process that is financed primarily by the state (through subsidies), with additional monies through loans taken by communities and repaid by individual members. Through a not-for-profit company, Samudhaya Nirman Sahayak, communities draw down the funds they need to pre-finance land, infrastructure and housing development. The scale of activities has resulted in additional donor finance being drawn into the process through the Community-led Infrastructure Financing Facility (CLIFF), which is described in Box 6.9.

Land development

A further model offering a more comprehensive development strategy than shelter microfinance is the strategy of combining small loans for housing improvement with land development.72 One illustration is the case of El Salvador where low-cost subdivision regulations established during the early 1990s have helped to stimulate a low-income land development industry of 200 firms.73 After developing the area and selling the household a serviced plot, many of these developers offer a small loan (often around US$1,000) to build an initial core unit. It appears that this strategy has resulted in affordable secure tenure over the last decade, and – with greater supply – has lowered real estate prices in real terms. However, there are concerns about housing quality, and households who face difficulties at the end of the period may fail to secure a legal title.74 A similar system is used by the Salvadoran Integral Assistance Foundation (FUSAI), a microfinance agency that has also started to be operational in land development in order to address the needs of clients. FUSAI acquires the title to the land and undertakes the cost of infrastructure development. Once serviced, the land is subdivided and allocated to families who have been accepted by FUSAI according to income and capacity to pay criteria. Families receive the land title once they pay back the loan. The amount to be financed by the loan equals the price of the house, including road and infrastructure development, minus the subsidy received from the state, minus the value of the self-help contribution by the family.75 Similar initiatives are ongoing in Bolivia where Banco Sol has an agreement with a major developer and construction company in Santa Cruz.76

Figure 6.1

The Community-led Infrastructure Financing Facility(CLIFF) process

Note: DFID = Department for International Development (UK)
SPARC = Society for the Promotion of Area Resource Centres
HI = Homeless International SSNS = Samudaya Nirman Sahayak

Source: D'Cruz, 2004b.

The discussion here highlights the growing diversity of approaches that are grouped together within shelter microfinance. This final discussion on neighbourhood development (slum upgrading), together with the servicing of greenfield sites, has suggested a number of distinct neighbourhood and housing strategies that include a role for small-scale housing loans:

  •      Improvements of existing housing units: this is the dominant approach at present within shelter microfinance. Small-scale loans are provided to households with reasonably secure land tenure to enable the extension and/or improvement of accommodation.
  •      Linked land purchase and housing loan developments: private development companies prepare serviced land (and, perhaps, basic housing units) for sale with additional loans for housing development.
  •      Linked land development and/or upgrading paid for with a capital subsidy and housing loan developments: this has been discussed in Chapter 5 in the context of complete housing (paid for by saving, subsidy and loan) but the subsidy funds might be used to prepare a serviced plot with additional loans being taken as the household can afford to improve the dwelling.
  •      Linked settlement upgrading and housing loan: a further option may be for the government (either development agency and/or municipality) to upgrade the area, with households then taking additional loans to improve the dwelling.

In the case of the final three options, there are two distinct strategies that are considered in this chapter and Chapter 7. Shelter microfinance considers those strategies that are based on individual lending to the household by the microfinance agency. Without community capacity (and in the absence of state upgrading programmes), it is not possible for shelter microfinance to do more than loans for housing improvement.77 Chapter 7 looks at an alternative approach, community funds, which places more emphasis on collective capacity and which lends to groups of low-income households within a defined area and/or group. In some cases, the approaches within community funds have led into much larger-scale upgrading or land development strategies with the involvement of a much greater number of agencies. Notable examples are the Baan Mankong programme in Thailand (which has emerged from the work of the Community Organization Development Institute, or CODI) and the upgrading of 100 settlements in Phnom Penh, Cambodia, which was catalysed by lending from the Urban Poor Development Fund.


The preceding discussion concentrated on the growth in provision of small-scale loans for housing by NGOs and microfinance agencies. However, there are numerous sources of finance for small loans, although there are few large programmes that offer opportunities to finance incremental housing development at scale. One reason for the lack of scale is a lack of capital. The discussion of the development of this sector has concentrated on microfinance agencies and NGOs, both of whom receive external development assistance. Many of the providers considered below have had no external source of finance. The lack of loan capital is relevant to all providers and is discussed further in Chapter 8 as one of the challenges facing the sector.

There are numerous sources of finance for small loans, although there are few large programmes that offer opportunities to finance incremental housing development at scale

Small loans tend to be offered by less formal financial markets and they may have a number of characteristics that differ from formal financial markets.78 Access to finance may depend upon social networks based on religion or ethnicity. In some cases, households secure finance from neither formal nor informal financial markets, but borrow or otherwise obtain from friends and family. In this case, there may be further obligations in addition to repayments, with the loan being simply one component within a dense set of reciprocal exchanges. The following discussion focuses primarily on small loans being offered by institutions and organizations, rather than those being offered through entirely personal networks. Table 6.3 draws on one recent analysis that identified several key types of providers.79

The potential significance of commercial micro-lenders can be illustrated in the case of South Africa. Like most countries in the South, South Africa has always had informal money lenders who ignore official interest rate restrictions. During the mid 1990s, however, revisions to the Usury Act created the possibility of formal commercial micro-lending at unregulated interest rates. These commercial micro-lenders, shown in Table 6.4, now comprise 64 per cent of registered institutions with the Microfinance Regulatory Council. Table 6.4 also illustrates the importance of larger banks. These commercial micro-lenders serve a market that is predominantly formally employed, with access to a bank account. The council estimates that about 11 per cent of disbursements from such institutions are used for housing.

Table 6.3

Sources of small housing loans

Institutional type Area of focus
Microfinance agencies (large) Those already working with more than 100,000 clients. Housing loan programmes may have emerged from disasters and may be a ‘reward’ for successful enterprise lending.
Microfinance agencies (medium) Medium scale with 10,000 to 100,000 clients. May use similar principles for housing as for enterprise lending. May be short of suitable longer-term capital for such lending.
Northern NGOs Some lend directly and some provide wholesale funds. They may provide limited technical assistance.
mutuals and municipals
Locally owned and locally started housing programmes. May be part of networks for the sharing of experiences.
State housing programmes May have limited capacity to offer small loans. Major source of second-tier funding, but limited outreach.
Commercial agencies Some downscaling to housing generally faster than microcredit. Security and collateral is a major issue. Could mobilize large amounts of capital.
Local NGOs Mainly involved in housing from a community perspective. Most are small with less than 1000 clients.
Source: adapted from Escobar and Merrill, 2004, pp38–39.

In South Africa and other countries, there are also examples of commercial banks seeking to reach the lower-income market with smaller loans for housing. One example is the Banco de Desarrollo in Chile, which has a small lending programme for housing with 15,000 current loans and an average size of US$1200 per loan.80 As noted earlier, Banco Davivienda in Colombia is now considering developing a small loan facility for housing.81 However, such initiatives appear to be limited since many banks do not see it as profitable to develop lending into the small-loan housing finance sector.

In addition to the commercial microlender industry, there are some alternative forms of housing finance that have emerged, including lines of credit from building materials suppliers and hire purchase of individual items such as sanitary ware. In some cases, there are longstanding practices within these or associated industries (such as furniture). In Chile, companies such as Easy, Homecenter and Home Depot provide people with building materials and have credit systems to which it is very easy to have access, providing that proof of income can be offered. Home building materials supply chains (such as Elektra in Mexico) may also enter this business on a more significant scale. Elektra (a large electrical appliance chain) has now formed a bank that provides credit for building material packages suitable for starter homes. A further Mexican programme, Patrimonio Hoy, is run by Cemex and encourages women to save together for the purchase of building materials. At the end of five weeks, the programme will advance raw materials worth ten weeks of savings. After three years of operation, Patrimonio Hoy had 36,000 customers and over US$10 million in extended credit the customer base is reported to be growing at the rate of 1500 to 1600 individuals per month.82

Table 6.4

Registered micro-lenders,2003

Type Branches Percentage of total
Banks 2387 32.7
Publicly listed, non-bank 15 0.2
Private commercial micro-lenders 4687 64.2
Trusts 116 1.6
Section 21 (non-profit) companies 54 0.7
Co-operatives 42 0.6
Total 7301 100.0
Source: www.mfrc.co.za/detail.php?s=95.

Although remittances are not a provider of small-scale investments in housing finance, they are emerging as a significant source of finance for housing investment. Their current scale is estimated to be US$200 billion a year, placing remittances as the second largest inflow to the South after foreign direct investment.83 The largest receivers of remittance income are India, Mexico, the Philippines, Morocco and Egypt. Their growing scale has resulted in a number of institutional innovations to capture these financial flows and to more efficiently enable housing investment.

For example, Mexico's remittance income equalled 1.5 times the tourist income in 2002.84 Although precise data is hard to come by, it appears that a significant proportion of such remittances are invested in housing.85 An indication of the scale of such funds is the interest shown by financial agencies and building material companies in facilitating such investment. Box 6.10 describes the commercial systems that have been established to assist in housing investment in Mexico for workers based in the US.

The other set of institutions that may be concerned to provide small-scale loans are traditional home lenders. Traditional home lenders face substantial barriers in engaging in microfinance due to the relatively high costs associated with lending, which have been noted earlier.86 Such institutions may require a mortgage lien as security for their home loans, while most shelter microfinance agencies work with other sorts of collateral. The culture and underwriting standards of traditional home-lending institutions suit lending to the middle class and those with higher incomes, while these institutions have difficulty with the practices required for lending to low-income households, such as reconstructing informal income and securing alternative forms of collateral.

State programmes offering small loans are potentially important, although they have not featured much in the development of the sector. In general, there has not been large-scale state finance for small-scale lending to support incremental housing development, although there are some exceptions to this situation, including the programmes discussed in Chapter 7 in which small loans are offered through collective mechanisms. Further exceptions are where small housing improvement loans have been associated with larger-scale upgrading (slum improvement programmes). In other cases, governments have sought to provide capital for NGOs interested in providing small loans for housing development. In India, the government has sought to provide capital through HUDCO from the early 1990s. A number of NGOs have taken up these funds, while some have struggled to manage the restrictions within the programme.87 The Colombian government has recently taken a loan from the Inter-American Development Bank (IADB) that includes financing for 10,000 microloans for housing improvement. There may be a significant number of other programmes. Households buying serviced land from the city of Windhoek in Namibia can ask to repay over eight years at an interest rate of 15 per cent.88

There are signs that there is a growing interest in financing these approaches and more groups interested in participating in activities. In Peru, the state housing authority is channelling housing funds to microfinance agencies, municipal savings and loan co-operatives, as well as some microfinance banks, in an effort to provide appropriate finance.89 There also appears to be increasing interest at the municipal level in Latin America. The municipal funds in Peru offer little direct lending for housing, although the scale of their activities suggests that they have a major impact upon the financial choices of many of the residents. There are now 14 such funds throughout the country, with total deposits of US$200 million and an annual growth in deposits of US$40 million.90 Belem in Brazil provides a further example of the potential role of the municipality. Collaboration between the municipality, civil society, the Banco de Povo and the community itself has resulted in a flexible loan programme offering loans of up to US$500 for a variety of activities, and housing loans have also been made available through a new programme.91 One quarter of borrowers have improved their sanitation provision, reflecting urgent and pressing needs in the low-income neighbourhoods.

There have been some deliberate attempts to draw formal financial institutions closer to the microfinance sector. The discussion of social housing in Chapter 5 highlights the programme Tu Casa in Mexico, and there is a very similar component within an IADB loan to Colombia.92 In both cases, small home improvement grants are a minor part of loan and subsidy programmes that are primarily concerned with funding complete houses.

Cooperatives and other voluntary sector

There is a range of voluntary sector agencies, such as cooperatives, and credit unions, that seek to extend credit to their membership and that may offer small loans for housing. These may also include less formal rotating savings and credit associations (ROSCAs). In general, the loans offered by such providers are not intended for housing improvements but in some cases they are used for this purpose. A significant problem for such small-scale lenders is that the size of the loans is generally not sufficient for housing improvements. The issues are illustrated by an analysis of the Women Credit Union in Sri Lanka.93 The housing needs of the members led to external finance being raised to enable the union to offer housing loans. However, such credit was limited and, thus, few loans could be allocated. The Kenya Union of Savings and Credit Co-operatives established a housing fund in 1998 through an agreement with the National Co-operative Housing Union (NACHU). However, funds also appear limited, and by 2003 the fund had extended 33 loans valued at 40 million Kenyan shillings.94

Although informal financial mechanisms are used for incremental improvements in Hyderabad, such finance often cannot be accessed by the poorest.95 Many of the ROSCAs require regular payments that are difficult for the poor to meet. The more flexible systems that do not require monthly payments have higher participation from the poor.96

Box 6.10 Mexico and remittances

Cemex is a Mexican company and the world's third largest cement producer. Since 2002, Mexican residents in the US can buy cement and other building materials directly in eight Cemex branches in the US (a subsidiary called Construmex) and have the materials delivered directly to a chosen address in Mexico. Since it began this service (early 2002 to October 2004), US$3 million have been taken in construction sales. The company estimates that the building materials needed for a 100 square metre completed two-bedroom house cost US$6700.

The new mortgage banks, the Sociedad Finaciera de Objeto Limitado (SOFOLES), have also sought to capitalize on similar funds and two have opened branches in the US, with a third operating in the US via an intermediary. However, the sales of mortgages have been slow, in part, because Mexican migrants come from communities with self-build traditions. Presidents Bush and Fox launched a bilateral organization, Partners in Prosperity, in 2001. In November 2004, the programme stated:

BANSEFI [Banco del Ahorro Nacional y Servicios Financieros], together with housing finance public institutions INFONAVIT [Instituto del Fondo Nacional de la Vivienda para los Trabajadores] and FONHAPO [Fondo Nacional de Habitaciones], have continued working under the Programa de Recepción de Aportaciones de Mexicanos en el Extranjero to allow Mexican migrants living in the US to transfer money to their families in order to obtain housing benefits and pay mortgage debts. Sociedad Hipotecaria Federal (SHF) is funding the Raíces programme where mortgage intermediaries (SOFOLES) grant loans to Mexicans living in the US to acquire a house in Mexico.

Source: Connolly, 2004a, b.

Although remittances are not a provider of small-scale investments in housing finance, they are emerging as a significant source of finance for housing investment

Housing and/or savings and loan co-operatives and mutuals are a further source of loans in Latin America.97 Also notable are the housing and mutual aid co-operatives of Chile (Federacíon Unificadora de Cooperativas de Vivienda por Ayuda Mutua, FUCVAM), which provide loans and assist with construction. Although it might be anticipated that housing co-operatives would provide appropriate sources of finance, in practice many seem to concentrate on the provision of complete houses. This might be explained by their need to build ‘officially’ and conform to building regulations and/or by their own need for collateral. Box 6.11 discusses a scheme in Kenya to provide both housing and income support to low-income groups in Nakuru and highlights some typical problems of affordability that have been experienced elsewhere. Housing People in Zimbabwe faced very similar difficulties and found that many of those turning to housing co-operatives had higher incomes. Although such organizations often make considerable efforts to reach down to low-income groups – for example, Housing People helped one group of domestic workers – this tends to be exceptional. NACHU in Kenya has made some efforts to offer loans for land purchase and (household-level) infrastructure development to its member co-operatives. However, it is hard to assess the scale and affordability of this programme, and other loans are orientated towards those who have landownership.98 The Nala Makazi Housing Co-op in Dodoma, Tanzania, has also managed to raise capital for housing construction and is currently developing housing for those living in informal settlements.99 However, the scale is again very small. Similar problems appear to be prevalent in Latin America where credit unions will extend loans for housing improvement and purchase to lower income households but they require that households have savings deposits equal to about 25 per cent of the loan, which the poorer households are unlikely to find affordable.100 One exception to lending for incremental housing is the cooperative Jesus Nazareno in Bolivia, which provides small loans with a solidarity group guarantee for the purchase of land.101 However, titles are held by the co-operative until repayment is completed, and this suggests that the loans are to those able to afford secure tenure and title.

Box 6.11 From self-help group to housing co-operative in Kenya

Between January and February 2003, members of the Nakuru Housing and Environment Cooperative (NAHECO) in Kenya have accessed seven housing loans from the National Housing Co-operative Union (NACHU) amounting to 360,000 Kenyan shillings and microcredit loans amounting to 140,500 Kenyan shillings. The membership of the co-operative has increased from the initial number of 15 groups from the three low-income settlements to 30 groups drawn from seven low-income settlements. The increased membership to NAHECO has resulted in increased savings. NAHECO has taken up a role of coordinating local self-help activities and people show confidence and trust in the operations of the group.

However, the poorest within the area are unable to benefit from this programme. The participatory needs assessment results showed that 92.6 per cent of people living in the three project settlements were tenants living in dilapidated housing 70.9 per cent of them were very poor, with no land on which to construct own housing. One of the criteria for accessing credit for housing through NAHECO is possession of land or the ability to save enough to buy some. This is a major weakness in identifying the target group and formulating the guidelines for accessing credit through NAHECO. This implies that the poorest of the people in the target area may be excluded from benefiting from the project.

Source: Ng'ayu, 2003.

Despite such difficulties, the significance of many small providers is emphasized by a recent assessment of the microfinance sector in Peru.102 Major institutions offering small loans include self-managed communal funds and cooperatives, NGO programmes, local microfinance agencies, municipal funds, rural funds and the protection funds of some workers’ unions. An estimated US$25 million may be loaned each year to housing, of which about 67 per cent comes from the savings and loan co-operatives.

Box 6.12 Acquiring loan finance in South Africa

The Kuyasa Fund is a non-profit microfinance institution based in Cape Town, South Africa. Since 2001, it has reached more than 2643 clients with US$1.8 million of housing loans. Portfolio at risk is 15 per cent and write-offs are 5 per cent of cumulative disbursements. Women constitute the vast majority of Kuyasa borrowers at 72 per cent, and account for 70 per cent of the value of loans taken.

The Kuyasa Fund has been unable to obtain any loan equity locally, and the wholesale equity and start-up grants have all come from offshore donor sources. Although the parastatal National Urban Reconstruction and Housing Agency (NURCHA) has assisted with loan guarantees, none of South Africa's housing-related parastatals have been willing to lend or grant Kuyasa any funds on the grounds of ‘high risk’. Kuyasa, however, has already demonstrated conclusively that its lending performance is better than that of mortgage lenders operating in the same market. Recently, the National Housing Finance Corporation has courted Kuyasa management with the offer of a loan, but on terms that made it unviable for Kuyasa. Once again, the parastatal cited risk as its main concern. Kuyasa's difficulty in attracting local equity, even in the face of solid performance, reflects the continuing dominance of the mortgage mindset and risk aversion in South Africa's parastatal housing finance sector.

Source: Baumann, 2004 Van Rooyan, 2004.

Sources of capital finance

How do microfinance agencies secure capital for their lending? Some providers draw on their own capital, notably the private sector and, for the most part, the small-scale voluntary organizations, such as credit unions. However, most agencies who wish to expand their lending have to find significant sources of capital.

Although consumers in South Africa have been successful in accessing and using small loans and targeted savings for incremental housing improvement, the policy and regulatory environment has not been developed with this approach in mind, and there is no source of wholesale finance or technical support for such institutions.103 The NGOs who have developed this model cannot drive the development of a pro-poor housing finance sector alone. Groups such as the Kuyasa Fund now face a major constraint in the lack of capital to expand lending (see Box 6.12).

Such NGOs and other microfinance agencies have four sources of funds: deposits, development assistance, governments and the private sector. The problem remains even in countries with a well-developed microfinance sector, such as Bangladesh. Despite the creation of an apex financing institution, the Palli Karma-Sahayak Foundation (PKSF), agencies such as the Grameen Bank remain short of capital to finance microloans for housing.104

Although many agencies encourage deposits and, as noted in Box 6.14, in SEWA's case these savings provide 80 per cent of capital, availability of medium-term capital is recognized to be a constraint. This is a problem even in the context of the Central American agencies funded by SIDA that generally receive medium- to long-term support (an average of nine years per programme).105

Microfinance organizations, for the most part, seek to be viable commercial enterprises. The small number of agencies studied by Cities Alliance are broadly successful in this aspiration.106 Micasa (the housing programme of Mibanco in Bolivia) broke even on a cash-flow basis, including the initial investment in adjusting the management information system, within nine months if performance continues at current levels, it is expected to generate a return on loan portfolio of between 7 and 9 per cent, compared with its overall return on loan portfolio of 3.4 per cent. FUNHAVI, the Mexican agency, was operationally self-sufficient after six years of business and moving towards full financial sustainability.

However, both these agencies appear to have had sufficient capital to expand their activities to a profitable level. Proa (also in Bolivia) has a model that would work without a subsidy only if volumes increased.107 The programme has money from a mutual savings association at 9 to 10 per cent and on-lends at 13.5 to 15 per cent. Given current volumes, a higher fee (margin) is required but this is not allowed by the mutual association providing the funds. The expansion of the programme from US$175,000 to US$500,000 of new loans per month would allow costs to be covered. The success of this strategy is critically dependent upon securing adequate capital to expand lending.

This aspiration to be financially viable without access to financial support has a number of implications for the nature and development of microfinance. Many microfinance organizations face a balance between reaching down to the poorer households with smaller loans and minimizing administration and management costs by offering larger loans. In general, the emphasis has been greater on cost-effective lending. There is a widespread belief (supported by many experiences) that access to credit is rather more important than the price of credit and, hence, that microenterprise lending can charge interest rates that are relatively high in comparison to the formal financial markets (although low compared to informal money lenders). However, housing loans are often considerably larger and therefore the interest rate charges are more significant. In some cases, lenders have developed specific housing products with lower interest rates these are generally commercially viable even if they are not fully market based.108

Some bilateral donors have funded shelter microfinance activities for a considerable period (almost 20 years) including Swedish Assistance (Box 6.13) and USAID. However, the multilateral donors – such as the IADB and the World Bank – have only begun to learn about and develop programmes in this area over the last few years. In their absence, Northern NGOs have played a very significant role in supporting such initiatives. These NGOs have included Misereor (Germany) and CordAid (the Netherlands), as well as specialist housing and urban development groups such as SELAVIP (Belgium) and Homeless International (the UK).

Shelter microfinance and subsidies

There is a difference of opinion between microfinance agencies about the need for housing subsidies. On the one hand, there is a belief that subsidies are necessary because of the traditional association between subsidies and lowincome housing and because of the larger size of housing loans.109 On the other hand, it is widely accepted that microfinance needs to perform without subsidy finance in order to be able to expand as market conditions permit. Sector commentators suggest that subsidies should not be offered through interest rates or permitted defaults, and that subsidies, if offered, should be managed separately outside of the loan operation.110 For example, subsidies might be provided through capital grants for housing investment or through the provision of water and sanitation services. Chapter 5 discussed the use of small loans to top up housing subsidy finance.

Despite such recommendations, this is not necessarily common practice. In situations in which there is no state support, there appears to be an effective cross-subsidy from enterprise to shelter lending, as the interest rates are lower in the latter. In some countries, particularly in Asia, subsidies are available through reduced interest rates and microfinance agencies have become a conduit to deliver state support to the poor. In some cases, the subsidy is provided in the form of an interest rate reduction. Grameen Bank and SEWA have both accessed low-interest sources of funds and pass on this subsidy.

Box 6.13 Swedish International Development Agency (SIDA)
assistance to low-income housing in Central America

Since 1988, the Swedish International Development Agency (SIDA) has financed housing and local development programmes in Central America with total resources of US$50 million. By the end of 2003, the programmes had helped approximately 80,000 low-income families, or about 400,000 people, in the main urban areas of the region to improve their habitat conditions. The resources from SIDA have been channelled through different institutions and programmes – namely, the Foundation for Housing Promotion (FUPROVI) in Costa Rica, the Local Development Programme (PRODEL) in Nicaragua, the Salvadoran Integral Assistance Foundation (FUSAI) in El Salvador, the Urban and Rural Social Housing Development Foundation (FUNDEVI) in Honduras and the Local Development Trust Fund (FDLG) in Guatemala.

SIDA's policy throughout the region has been that housing subsidies are primarily the responsibility of national governments, who act as counterparts to the international agency. That is why most of the funds allocated by SIDA have been channelled to finance three main components of these programmes: loans (including microloans for housing improvements and new housing), technical assistance (both to executing agencies and the target population) and institutional development, especially of those institutions that manage the Swedish funds.

Source: Stein with Castillo, 2005.


There is a considerable diversity in the nature of shelter microfinance as provided by the many different organizations who are active in this sector. One commentator illustrates such differences thus:

In Mexico, CHF International and FUNHAVI [Fundación Habitat y Vivienda A.C.] have developed a home improvement loan that features an average loan amount of US$1800, a repayment period of 18 months for first-time borrowers and a 54 per cent effective annual interest rate. The Grameen Bank's housing loans typically are repaid over ten years. They are offered at an interest rate that is 10 per cent below rates assessed for microenterprise loans, and first-time clients are not eligible for such loans.111

The average size of the Grameen Bank housing loan is 13,386 Bangladesh taka (US$224).112 This contrast demonstrates the significance of local context in developing appropriate housing finance solutions. The difference between these approaches reflects the type of housing solution that is acceptable and affordable to the borrowers, and the solution that is likely to be approved by the authorities if they have a significant presence. The contrast also reflects the target group for lending activities and the operating constraints and choices of the agency. For example, as is sometimes the case for microenterprise lending, some microfinance agencies prefer to give fewer larger loans, thereby reducing their administration costs and increasing their financial returns for a given amount of loan capital.


The link between housing investment and savings extends well beyond the microfinance sector. In the North, traditionally families have saved for several years simply to access conventional mortgage finance. Savings is a particularly significant component of the contract-savings schemes in Western Europe, notably the German Bausparkasen and UK building societies.113 Similarly, many microfinance programmes for housing, particularly in Asia and Africa, have savings requirements.

Savings has a place in microfinance for many reasons. Savings is a strategy to assist with repayments in which borrowers have to demonstrate a capacity to make regular payments and accumulate sufficient funds for the required down payment or deposit. Microfinance agencies may try to get would-be borrowers to save at a rate equal to loan repayments, in part to reduce the risk to lender and borrower. The required savings period typically lasts between 6 to 12 months before a loan is granted.114 One notable example is Bank Rayat Indonesia (BRI), which has mobilized more than US$2.7 billion in voluntary savings through 16.1 million savings accounts however, saving on this scale is very unusual. A further reason to encourage savings is to assist the agencies themselves in acquiring funds.115 In SEWA's case, the bulk of the bank's loan portfolio arises from client deposits, although additional finance for housing loans is provided by the government though HUDCO.116 The importance of saving can be illustrated for the case of SEWA:

In order to be eligible for any loan from SEWA Bank, for example, the would-be borrower must have a regular savings record at SEWA Bank for at least one year. What is important for SEWA is that the savings history is stable and consistent. SEWA Bank's experience is that facilitation of a strong savings habit correlates significantly with high loan repayment rates – hence, a client's savings record serves as the main form of collateral for loans.117

The significance of savings to the clients of microfinance agencies has long been recognized. The experience of the Kuyasa Fund in South Africa is that clients use their savings to augment the subsidy that they receive from the state. A very notable estimated 65 per cent of Kuyasa clients only save and do not take loans.118

Collateral and security

Collateral is an asset pledged to a lender until the borrower pays back the debt. Its major role is in reducing lender risk and it is widely recognized that a key challenge for shelter microfinance is that of loan security.119 Many microfinance agencies seek to minimize the need for collateral by using existing client history (enterprise lending). A further strategy used for lending for income generation is small repeat loans as a way of building up repayment skills and capacities and providing an incentive for repayment. However, the larger size of shelter microfinance makes this strategy more difficult to follow.

Another strategy used by microenterprise lenders is that of group guarantees. However, this strategy has been found to be problematic for housing loans, again because of the bigger loans and longer loan period.120 This may explain the problems faced by the Group Credit Company in South Africa (which tried and failed to replicate Grameen Bank strategies in offering small loans). Difficulties are related to the longer period of the loans and, hence, the lack of need for the group unless repeat income-generation lending is also taking place. The use of group guarantees should not be confused with group loans, which include a collective responsibility to manage and repay the loan (see Chapter 7).

In the absence of such strategies, a wide range of collaterals are used, including mortgages, personal guarantees, group guarantees, fixed assets and/or pension/provident fund guarantees.121 Pension fund collateral is used particularly in South Africa and Bangladesh, and more recently in Namibia, but is not significant elsewhere. In a recent study of microfinance agencies’ practices, the following are identified as collateral:122

  •    land title and buildings
  •    chattel mortgage/lien on assets
  •    obligatory savings
  •    assignment of future income (wages)
  •    personal guarantees (co-signers)
  •    joint liability and group guarantees (character-based lending) and
  •    other financial assets (for example, life insurance policies and pension funds).

One difficult area is the extent to which legal title is a requirement of lending. One commentator argues that ‘Client ownership of the home or land is preferred: it is against the policy of some lenders to provide credit for housing on squatted land.’123 Moreover, in some countries such as the Dominican Republic, lenders may not be legally allowed to extend housing loans without a formal property title.124 However, despite an emphasis on land ownership, the use of title deeds as collateral for microfinance loans is limited, and one study of 80 such organizations found that only one quarter use it.125 For example, the experience of Mibanco in Peru is also to avoid the use of land titles. The agency relies on the same informal collateral of household assets and co-signers used for microenterprise loans (despite the mass land-titling programme that has taken place in Peru, discussed in Chapter 4). Mibanco found that land titles are expensive to use as guarantees, and that poor clients do not want to use title as collateral for a loan of less than US$1000.126 Banco Sol uses such collateral but considers that there are major risks because of the poor standard of deeds and title documentation.127

Alternative strategies are varied. In some cases, such as the Grameen Bank, home loans are only given to those who have experience in enterprise lending and a good repayment record. Alternatively, social collateral such as guarantees from other residents involved in the programme may be used. A further option is holding the para-legal documents to the property, or other non-mortgage collateral such as jewellery. Some lenders take a mortgage lien when the costs and legal structure permit for larger loans, such as for the construction of a core unit. In the case of PRODEL in Nicaragua, experience suggests that for loans under US$700, there are other types of collateral as effective as a mortgage.128 PRODEL gives loans to families who do not have full land ownership, but that are able to demonstrate security of tenure – for example, co-signers who could put up their properties for mortgage, valuable objects and municipal certificates that show security of tenure, although not necessarily land title. Only half of the more than 5000 loans provided up to the year 2000 were mortgaged, and delinquency rates were still very low.

Although not collateral, a further common requirement is to specify the maximum percentage of income that can be used for housing loan repayment. A maximum percentage of 25 to 30 per cent of income in housing repayment is widely used by agencies. However, the effectiveness of this constraint can be questioned as precise incomes are not that easy to establish. Lenders may have different conditions for salaried workers and entrepreneurs.129

Interest rates

In many cases, interest rates for shelter loans are lower than those for enterprise development, even when offered by the same agency.130 In most cases, the rates are fixed as the loans are for relatively short periods and it is very difficult for low-income households to cope with the uncertainty of variable rates. In a study of four Bangladeshi microfinance agencies offering loans for housing, the interest rate was lower in every case.131 Although the Grameen Bank's explanation rests on the social significance of housing, it is also notable that higher interest rates would be unaffordable for the target group, given loan size and repayment periods.132

Setting the level of interest rates is clearly a difficult issue. Interest rates must be acceptable to borrowers and one report on SIDA's experience suggests that interest rates cannot diverge greatly from (even if they are not identical to) mortgage rates.133 Most agencies seek to at least cover the cost of inflation and administration, with an allowance for defaults and bad loans. Box 6.14 summarizes SEWA's experience in setting interest rates for housing loans. An alternative approach used by Habitat for Humanity in Africa and the Middle East is to use a variable inflation index on the loan, which is pegged to the price of a bag of cement.134 This allows repayments to maintain their real value.

Loan periods

There is a very significant difference in the loan periods of different shelter microfinance programmes. One recent survey of 15 agencies offering small loans for shelter finds that the loan periods differ by between 20 months and 15 years.135 It might be anticipated that longer loan periods would be used for larger loans, potentially secured on the property, as the incentive for small repeat lending could not be used. In practice, this does not appear to be the case and some of the small loans have long repayment periods, with some larger loans featuring shorter repayment periods. However, the longer loan periods may be misleading. For example, one case is People's Dialogue in South Africa, where in most cases the loans are bridge financing for the state housing subsidy and are paid off rapidly once the subsidy entitlement has been accepted and finance released.

Box 6.14 Self-employed Women's Association (SEWA)
interest rates for housing, India

When the Self-employed Women's Association (SEWA) first started lending for housing in India, it did not differentiate between housing and enterprise loans (in practice, the housing loans were bigger and were often the third or fourth loan that was taken). However, due to the size of housing loans (and the fact that they did not necessarily generate an instant higher income flow), they have been differentiated as a separate loan product since 1999, since which time they attract a lower interest rate of 14.5 per cent. Income generation loans – which typically account for 50 per cent of SEWA Bank's total loan portfolio and are usually of a lower loan amount and generate faster returns, charge interest at 17 per cent, thus partially cross-subsidizing the housing loan portfolio. SEWA's average cost of capital is 8 per cent and this primarily reflects the interest that it pays on members’ savings. To secure housing loans, clients must have a regular savings record of at least one year. SEWA's experience is that a strong savings record correlates to good repayments and the regularity of payments is more important than the amount.

Source: Biswas, 2003.

Technical assistance

A further area related to the provision of subsidies is that of technical assistance. Many of those lending for shelter microfinance seek to provide assistance in construction activities. For example:

  •      FUSAI is an NGO in El Salvador that is working in housing-related activities. In 2002, it decided to separate its housing financing activities from construction support in order to maximize the efficiency of both operations.136
  •     Proa, a Bolivian NGO lending for housing
    improvements, has technical staff who prepare plans and budgets. They receive a commission on each loan (US$40) and secure additional payments from the households if required.137
  •     SEWA found that its members were increasingly asking for other services related to housing (in addition to loans). The Gujarat Mahila Housing SEWA Trust (MHT) was established to provide SEWA members with technical services related to housing, including advice on improving and extending existing houses, building new houses and infrastructural services. The MHT plays a key role as an intermediary with government departments in accessing schemes, including those related to infrastructural facilities and environmental improvement.
  •     FUNHAVI (Mexico) goes one step further and requires borrowers to buy construction materials from it (as well as providing technical advice). However, this is also a financial measure, as it buys at wholesale and sells at retail prices.138
Box 6.15 Grameen Bank loan package, Bangladesh

In Bangladesh, the families who are members of the Grameen Bank typically live in small shelters of jute stick, straw, grass thatch, bamboo and dried mud. Each year a family has to spend about US$30 to repair the house after the monsoons. For an equal amount of money, a family can repay a housing loan for a strong, well-constructed house with a floor area of 20 square metres. The bank views housing loans as investment rather than consumption since a secure and well-constructed house aids the health and well-being of the family and helps them to break the vicious circle of poverty. The house can be used for storage for their small businesses, and time and money are saved in not having to continually repair the jute-stick shelters.

The Grameen Bank has developed two standard house designs. The smaller one costs US$300 and a larger version costs US$625. In many cases, the family adds their own savings to the loan and spends up to US$800–$1000 on their home and its furnishings. The houses vary in appearance throughout the country, but have the same basic structural components. There are four reinforced concrete pillars on brick foundations at the corners of the house and six intermediary bamboo or concrete posts, with bamboo tie beams, wooden rafters and purlins supporting corrugated-iron roofing sheets. This provides stability in the flood and strong monsoon wind and protection from the heavy rain during the monsoon season. In cases of severe flooding, the house can be dismantled and the components stored and reassembled later. A sanitary latrine must be provided with each house. Families can build the houses themselves with the help of friends and neighbours. Local skilled carpenters carry out the roof construction.

Source:www.bshf.org and www.grameen-info.org.

Opinions differ about the viability of such services for microfinance agencies. One argument is that the more developed microfinance agencies do not offer such services.139 A related view is that ‘Construction assistances in the context of housing microfinance does not appear to be a predictor of financial performance.’140 Some, such as Associación para el Desarollo de Microempresas (ADEMI) (Dominican Republic), argue that it is up to clients to manage their own affairs. Groups such as the Co-operative Housing Foundation argue that it is a necessary service and the content helps to reduce default rates. Another position is that of the Kuyasa Fund in South Africa, which does not want to provide these services itself, but recognizes the need to work alongside those who can provide technical assistance around construction issues.

In some cases, such as the Grameen Bank in Bangladesh, the loan is for a defined package of building materials, which minimizes the need for technical assistance (or greatly eases its provision) (see Box 6.15).

Orientation towards women

There is an emerging preference to lend to women in many of these institutions, based on the reliability of repayment.141 Women borrowers are ‘current good practice’ and there is a particularly strong predisposition towards lending to women in Asia.142 The Grameen Bank, for example, argues that the title to the house constructed with loan finance is vested with the borrower, and in 95 per cent of cases this is the woman. By having title to the house, the woman obtains financial security and an improved status within the family and society. In the case of FUNHAVI in Mexico, 38 per cent of the clients are women.143 According to the Kuyasa Fund, South Africa, women are 72 per cent of the borrowers.144 In the case of PRODEL in Nicaragua, more than 60 per cent of the housing improvement loan recipients and 70 per cent of the microentrepreneurs are women.145 Such figures are indicative of the more general position: women are often predominant among borrowers, but few funds exclusively serve women.

In the case of shelter, the role of home carer is often defined by gender and given to women. Hence, women may have a greater interest in investing in housing even if they are less likely to be the formal ‘owner’ of the dwelling.

Income generation

Although the primary focus of the initiatives discussed above is on savings and lending for shelter improvement, some of these programmes recognize the evident links between shelter and livelihoods. Some agencies, such as SEWA, have long recognized the close connection between home-based enterprise lending and housing improvement loans.146 Improving the infrastructure in the areas in which SEWA is working resulted in an average 35 per cent increase in small enterprise earnings.147 Through experiences such as these, there is a growing awareness of the links between enterprise and shelter investment.

There are three notable ways in which these programmes are linked to enterprise lending. The first is through lending for income generation. In many cases, shelter microfinance is offered along with income-generation loans. In some cases, it is a condition of the lending organization that income-generation loans are taken first, in other cases, one or other might be taken. The justification for the first strategy is that successful income generation is needed to be able to afford housing investment and related loan repayments. The argument in favour of the second strategy is that many ‘enterprise’ loans are diverted to housing investments and repayments proceed successfully.

Second, housing investments are more directly linked to income generation in a number of ways. Housing construction activities may be to improve a business or production area, such as a small shop or a workroom. In some cases, they may not even be related to a productive or vending enterprise directly, but may be providing a room to rent. Finally, the more ambitious schemes have explored the possibility of creating commercial centres to improve local livelihoods and to strengthen the local economy. Generally, these strategies belong to initiatives with more ambitious development objectives (see Chapter 7).


Although shelter microfinance might not be effective in every context, there is now widespread experience and understanding of the process and considerable appreciation of the approach in many countries. There are two notable challenges facing the shelter microfinance sector. The first is the nature of the beneficiary group and the difficulties faced by very poor households due to problems of affordability and lack of secure tenure. The second is sources of funding. Although other issues may be of specific concern to particular programmes, these two subjects are those that appear to be the most significant.


Microfinance for shelter may contribute to a more holistic approach to development than that generally associated with microfinance. In so doing, it may be addressing some of the concerns raised about its ability to assist some of the poorer families.148 By reducing expenditure on basic needs (such as rent, repairs to housing and water costs), lending for land, infrastructure and housing may increase remaining income and reduce vulnerability. As demonstrated in the case of the Grameen Bank (see Box 6.15), housing investment reduced repair costs and essential expenditures.

These programmes appear, in general, to reach the income groups served by microfinance agencies lending for enterprise development and families with similar incomes in the formal sector. The bias of microfinance agencies towards the somewhat higher income groups has been recognized for some time. This bias reflects the need of the agencies to secure high levels of repayments and give out larger loans (with the administration costs therefore being a smaller proportion of the loan). It also reflects the self-selection of their clients, with the more vulnerable avoiding the problems of debt, or beginning and dropping out of the programmes. Many shelter microfinance programmes appear to be targeted at the higher income urban poor, sometimes those with formal employment (at least one member of the family) and often those with diversified household livelihood strategies. As is the case with SEWA, successful income-generation borrowing may be required prior to housing loan applications. In many cases, land tenure is required.

The target group of those agencies reviewed by Cities Alliance is profiled thus:

…these financial institutions describe their clients as the economically active poor in the informal sector. They are largely serving their existing poor clients with this new loan product, and most provide housing loans as a reward for good past performance on microenterprise loans.149

In the cases of the agencies considered, Mibanco's clients have an income that is around or below the poverty line for Peru (where 50 per cent of the population have incomes below the poverty line). FUNHAVI in Mexico serves clients who earn between two and eight times the local monthly minimal wage of US$125. SEWA Bank's clients are all poor self-employed women – predominately street vendors, labourers or home-based workers. In 1998, an estimated 76 per cent of SEWA borrowers had annual household incomes below US$415 and half of these had annual incomes below US$276. Clearly, the group that is being reached is poor and in need of housing investment. However, these are large income categories and they may say little about how far below the poverty line such programmes are able to extend.

In some cases, shelter microfinance is linked to state subsidy programmes (notably in Latin America), and this may extend their reach downwards towards lower income groups. The Step-by-Step programme in Ecuador, for example, helps households to raise the deposit they need in order to secure the direct demand subsidy and therefore afford improved housing. However, as noted in Chapter 5, such programmes may include further loans and, hence, the poorest may not be able to afford the costs of inclusion.

The use of other mechanisms and, notably, the requirement for secure tenure, may further define the client group as being the poor, but not so poor. The greatest difficulty faced by the poor is that, in general, these programmes offer small loans for housing improvement and therefore cannot address the large numbers who do not have tenure security (if not a full title). A further illustration of such restrictions is given for one housing loan programme in India in which only those households who were occupying the house on an ownership basis were selected and tenants on rent were excluded this was based on the consideration that such households would not be in a position to join the shelter upgrading programme.150

It might be argued that any household able to afford a loan is not going to be the very poorest therefore, the shelter microfinance programmes will inherently struggle to reach down to those with lower incomes. The group that is being reached by these programmes is clearly benefiting from the assistance. Moreover, without access to loans, housing investment is very inefficient. For those who do secure loans, the benefits can be considerable. In addition to the income benefits discussed above, Box 6.16 describes some of the health consequences. Shelter microfinance appears to be effective in improving the housing conditions of a group eager to invest in its own dwellings. It has a significant role in a system of housing finance, while, at the same time, there is a need to be realistic about the limitations of the strategy in reaching the poorest.

Securing capital

As noted above, securing sufficient loan capital is difficult. Lack of capital emerges as being a very significant constraint on expansion. Banco ADEMI (in the Dominican Republic) cited lack of capital as the principal challenge that the organization faces in providing housing credit, for which there has been substantial demand.151 These difficulties reflect a general constraint on the microfinance sector and usually do not appear to be specifically related to housing lending however, as illustrated in the example of Bangladesh, there may be even more limited sources in the case of housing. As noted earlier, in the case of some agencies, viability is related to the scale of activities, and capital for expansion will result in profitable lending and potentially an easing of capital constraints.

Box 6.16 Improving shelter, improving health

The few impact evaluations conducted of shelter finance point to positive results for the poor. An evaluation of Plan International's Credit for Habitat programmes in Bolivia and Guatemala showed that clients invested their US$200–$800 loans in roofing, walls, floors, tiling, water, sewage and electrical connections, as well as additional rooms. Seventy-eight per cent of clients said that home improvements improved family health. Clients with Grameen-financed homes in Bangladesh – equipped with Grameen's construction standards of cement pillars and sanitary latrines – had 50 per cent fewer incidences of illnesses than those without Grameen houses. Their houses suffered far less structural damage during the devastating floods of 1987 and since, compared with non-Grameen homes. An impact assessment of the Self-employed Women's Association (SEWA) Bank's slum upgrading programme in India, which included progressive housing loans, reported increases in literacy (school children enrolment), productivity (increase in number of working hours), income and health (lower incidences of illness and, thus, lower health expenditures), and increased marriage opportunities, higher status and respect in the community for women borrowers. In sum, housing finance loans serve poor households and help them to improve their livelihoods.

Source: Malhotra, 2003.

It is not so evident that sheltermicrofinance can succeed in ensuring a growing andsecure capital base

Very little is known about the aggregate balance of sources of funding for shelter microfinance. A recent study of the total capital of the larger microfinance agencies in Bangladesh highlights some interesting trends.152 It is notable that finance from the commercial banking sector increased from 3 to 11 per cent of total capital between 1996 and 2002. Donor finance has dropped fairly dramatically through a similar period (from 58 to 17 per cent), although this partly reflects the growing significance of the Palli Karma Sahayak Foundation, a public–private apex body that channels funds to microfinance agencies, which has increased its significance by providing 12 per cent of capital in 1996 and 24 per cent of capital in 2002. However, the analysis suggests that the strategy used by these agencies may not be easy to replicate in other countries and it is not so evident that shelter microfinance can succeed in ensuring a growing and secure capital base. A further specific suggestion is that Palli Karma Sahayak Foundation should extend its activities and provide finance for housing.153

Shelter microfinance agencies may face a particularly difficult balance in setting interest rates that weigh borrowers’ demands against their own financial needs. Interest rates must be acceptable to borrowers.154 In some countries, subsidized interest rates for mortgage loans may increase pressure for reduced interest rates. Such factors, as well as longer-loan terms and required concessions for affordability, may explain the use of favourable interest rates in the case of small loans for housing.

As is evident from this discussion, microfinance agencies face an issue of scale. To be profitable, they have to increase the quantity of lending. There is evidence that this is driving their expansion into shelter microfinance but for the smaller agencies, lack of capital to expand operations appears to be a significant constraint. One view is that the shorter lending terms of shelter microfinance may better fit the short-term funding sources (with the bulk of financial liabilities often one year or less) available to financial institutions in the South therefore, more conventional housing lenders should be active in this area. Such a match of demand and supply may help to account for the strong interest being shown in this area. The greater interest demonstrated by the private sector may assist in reducing the capital constraint however, it is equally evident that this is unlikely to happen in all countries. The Banco Davivienda in Colombia is currently working with the government to examine the possibility of offering loans of less than US$2800 to be repaid in up to five years for homes valued at less than US$15,000.155

Nevertheless, it is equally apparent that longer-term loan repayment periods are also common in shelter microfinance agencies, despite the small size of the loans. Raising funds for shelter microfinance may be more complicated than for enterprise lending because of these longer loan periods. In the case of microenterprise lending, donor support has placed emphasis on building the institutional capacity of lending agencies and assisting in the accumulation of their capital base. There has been a resistance to providing concessional funds for on-lending. Despite this, it has been argued that one problem is that such agencies have had access to funds at a modestly concessional rate, which have been built into the cost basis of their operation. As a result:

… one recent ambitious effort to raise funding for major MFIs [microfinance institutions] on international capital markets ran squarely into this problem – lack of demand for the funds. Very few MFIs wanted funds on the resulting market terms.156

Shelter microfinance products continue to be developed, and there are reasons to believe that more agencies are entering this area and that those that are here already are expanding their activities. Can shelter microfinance continue to scale up? Lack of financial capital does appear to be a significant constraint. However, there are more agencies interested in this area in some countries, notably the private sector, municipal government and central government. In some cases, they are working with existing microfinance agencies in others, they are developing their own products. In part, the growth of shelter microfinance has been driven by the commercial interests of existing microfinance agencies and the need to consolidate and extend their own market base. In the Latin American context, this has happened in a number of countries in which direct-demand subsidies already exist or are being introduced. Microfinance can help to secure subsidies and add value to the construction process. In other cases, microfinance agencies have responded to their own analysis of need and have been able to secure funds from the state to extend their services. As a result, shelter microfinance as a sector is witnessing the expansion of existing agencies, new NGO and microfinance agency initiatives and new interest from groups that were not previously involved in offering small loans.


   1    This chapter is based on a draft prepared by Diana Mitlin, University of Manchester, UK, with assistance from a number of urban researchers listed in the Acknowledgements.

   2    See, for example, those
discussed in ESCAP, 1991 Environment and Urbanization, 1993 Arrossi et al, 1994 Mitlin, 1997 Jones and Datta, 1999 Center for Urban Development Studies, 2000 Environment and Urbanization, 2001 ACHR, 2002 Daphnis and Ferguson, 2004 Malhotra, 2003.

   3    There has been someexperimentation and the case for housing microfinance is explored in Daphnis and Ferguson, 2004.

   4    ESCAP, 1991.

   5    Ferguson, 2004b, p4.

   6    Mutagwaba, cited in Government of Tanzania and UN-Habitat, 2003, p31.

   7    See Okpala, 1994, p1572.

   8    Ballesteros, 2002, p3.

   9    The Center for Urban Development Studies (2000) divides the groups into two based on the nature and purpose of the originating agency (microfinance or shelter advocacy). The alternative division used here, which divides by the terms and conditions of the loan and, in particular, by whether loans are collective or individual, has the advantage of distinguishing those agencies able to lend for land and infrastructure (which must necessarily be collective) and those for whom shelter finance is primarily restricted to housing improvement.

  10   Smets, 2002, p77.

  11   Malhotra, 2003, p225.

  12   Datta, 1999, p204.

  13   Kamete, 2000, p254.

  14   Okwir, 2002, p95.

  15   Okonkwo, 2002, p97.

  16   Biswas, 2003.

  17   Smets, 2002, pp65, 87.

  18   Smets, 2002, p65.

  19   Moss, 2001, p33–4.

  20   Baumann, 2003, p88.

  21   Government of Tanzania and UN-Habitat, 2003, pp53–54.

  22   Gough, 1999.

  23   Datta, 1999, p203.

  24   Gough, 1999.

  25   Wahba, 2001.

  26   Nell et al, cited in Baumann, 2004.

  27   Civil society is a term used to refer to the not-for-profit and voluntary sector. It is also referred to as the third sector in some texts. The distinctive features of such agencies are that they are not part of the state, nor are they commercial companies. Although not all such organizations areregistered (for example, as charities and/or voluntary associations), many of those offering loans and/or savings facilities are likely to have some kind of registration. The exception is the large number of less formal savings groups including those that fall under the title of ROSCAs (rotating savings and credit associations). They are not discussed here as few offer sums substantial enough for shelter investment.

  28   Ferguson, 2003.

  29   CGAP, 2004.

  30   Biswas, 2003, p51.

  31   Smets, 2002, p75 Mohamed, 1997.

  32   Wahba, 2001, Appendix 1.

  33   SANMFI, cited in Biswas, 2003.

  34   Biswas, 2003.

  35   Ferguson, 1999, p191.

  36   Hoek-Smit, 1998, p41.

  37   Arrossi et al, 1994, p58.

  38   Co-operative Housing
Foundation, 1993, p38.

  39   Co-operative Housing
Foundation, 1993, p41.

  40   Jones and Mitlin, 1999, p27.

  41   Ferguson, 1999, p93.

  42   Connolly, 2004b, p9.

  43   A further Mexican example is FOSOVI, an NGO active in Mexico City. Between 1998 (when it was set up) and 2000, FOSOVI granted 1266 housing loans to families (SELAVIP, 2003, p43). The maximum loan per family is US$1000 and the loan period is 36 months. Capital has been secured through a partnership with local government, local savings and loan societies, NGOs and popular movements (residents’ associations). About one quarter has been for the construction of a core house and half for the extension of an existing house theremaining loans were for the improvement of facilities such as toilets or the improvement of housing quality. Families organize themselves into groups to receive the loans and this group helps in the material purchasing and construction process.

  44   World Bank, 2004a, p3.

  45   Smets, 2002, p181.

  46   Connolly, 2004b.

  47   Center for Urban
Development Studies, 2000.

  48   Arrossi et al, 1994, p54.

  49   Stein with Castillo, 2005.

  50   McLeod and Mitlin, 1993.

  51   Smets, 2002, pp195–197.

  52   Sheela Patel (director of SPARC), pers comm, 2004.

  53   McLeod, 2002, p204.

  54   Forero, 2004, p41.

  55   Mutua et al, 1996.

  56   Ferguson, 2003, pp26–27.

  57   Malhotra, 2004, p222.

  58   Escobar, undated, p21.

  59   Escobar, undated, p24.

  60   Malhotra, 2003, p222.

  61   Mutua et al, 1996, p182 Escobar and Merrill, 2004, p36 Ferguson, 2004a, pp24–25.

  62   Ferguson, 2003.

  63   Ferguson, 2003.

  64   Vance, pers comm, 2004.

  65   See Renaud, 1999, p760, for a discussion of such processes in the US.

  66   Stein, 2004, p116.

  67   PRODEL, 2002.

  68   Cities Alliance, 2002.

  69   Silas, 2004.

  70   Septanti, 2004, p8.

  71   Murcia de López and Castillo, 1997, p173.

  72   Ferguson, 2003.

  73   Ferguson and Haider, cited in Ferguson, 2003.

  74   Escobar, undated, p36.

  75   Stein with Castillo, 2005.

  76   Escobar, undated.

  77   Stein with Castillo, 2005.

  78   Smets, 2002, p9.

  79   Escobar and Merrill, 2004, pp37–38.

  80   Escobar and Merrill, 2004, p41.

  81   Forero, 2004, p41.

  82   Prahalad, cited in Connolly, 2004b.

  83   Ratha, 2003.

  84   Ratha, 2003.

  85   Connolly, 2004b, p3.

  86   Ferguson, 2003.

  87   Cities Alliance, 2002, p3 Smets, 2002.

  88   Gold et al, 2002.

  89   Escobar and Merrill, 2004, p40.

  90   Cabannes, 2002.

  91   Cabannes, 2004.

  92   IADB, 2003.

  93   Albee and Gamage, 1996.

  94   Gitau, 2004.

  95   Smets, 2002, p129.

  96   Smets, 2002, p130.

  97   Escobar and Merrill, 2004, p39.

  98   Gitau, 2004.

  99   Gitau, 2004.

100   IADB, 2003, p4.

101   Vance, 2004, p142.

102   Escalante, 2004, p59.

103   Baumann, 2004.

104   Hoek-Smit, 1998, p39.

105   Stein with Castillo, 2005.

106   Malhotra, 2003.

107   Ferguson, 1999, p196.

108   Escobar and Merrill, 2004, p58.

109   See, for example, Christen, 2004, pxiii.

110   CGAP, 2004.

111   Daphnis, 2004a, p3.

112   Grameen Bank, 2004.

113   Ferguson, 2004b, p4.

114   Vance, 2004, p141.

115   Escobar and Merrill, 2004, pp45, 48.

116   Biswas, 2003.

117   Biswas, 2003.

118   Van Rooyan, 2004.

119   Christen, 2004, pxiii.

120   Jones and Datta, 1999, p21 Escobar and Merrill, 2004, p58.

121   Escobar and Merrill, 2004, p61.

122   Vance, 2004, p139.

123   Escobar and Merrill, 2004, p62.

124   Daphnis, 2004b, pp108, 110.

125   Vance, 2004, p135

126   Malhotra, 2003.

127   Escobar, undated, p26.

128   Stein with Castillo, 2005.

129   Escobar, undated.

130   Escobar and Merrill (2004, p57) note that six of the eight agencies they profile offer lower interest rates, and one of the other two agencies may also offer lower rates.

131   Hoek-Smit, 1998.

132   Hoek-Smit, 1998, p38.

133   Stein with Castillo, 2005.

134   Escobar, undated, p19.

135   Escobar and Merrill, 2004, p56.

136   CGAP, 2004.

137   Ferguson, 1999, p196.

138   Malhotra, 2003, p221.

139   Escobar and Merrill, 2004, p62.

140   Daphnis, 2004b, p105.

141   Arrossi et al, 1994, p63.

142   Escobar and Merrill, 2004, p51. This emphasis on women is also true for microenterprise lending, although in some cases it has resulted in the woman in a family taking the loan but not being in control of theinvestment.

143   Malhotra, 2003.

144   Van Rooyan, 2004.

145   Stein, 2004, pp117–118.

146   Biswas, 2003.

147   Center for Urban
Development Studies, 2000.

148   See concerns expressed by Hulme, 2001.

149   Malhotra, 2003, pp219–220.

150   Lall and Lall, 2003, p15.

151   Davies and Mahony, 2001, p15 Pedro Jímenez, executive vicepresident, Banco ADEMI, Housing MicrofinanceQuestionnaire, 7 March 2001.

152   Zaman, 2004, pp9–12.

153   Hoek-Smit, 1998.

154   Stein with Castillo, 2005.

155   Forero, 2004, p41.

156   Ferguson, 2003, p27.