Indian community-based housing finance systems: potentials and pitfalls for urban development and housing improvement

ABSTRACT

In developing and emerging countries, housing finance is mainly mortgage finance in combination with Western unsecured credit, both of which do not serve the bottom of the income pyramid. Poorer sections of society depend on incremental finance for building their houses. A combination of different sources is usually used including individual servicing of housing finance and community-based housing finance. Community-based financing covers a relatively small section of the housing finance market. In the formal financial sector, such initiatives include mainly credit unions and cooperatives. In the informal sector, a much broader kaleidoscope of community-based financing schemes and links with banks exists. This paper provides an overview of three kinds of informal community-based housing finance associations: rotating savings and credit associations, savings associations and accumulating savings and credit associations. These examples can be seen as a source of good practices. However, schemes that are commodified can have a huge potential for the development of informal settlements, but can also become instruments of institutions, benefiting the state and the better off, who profit at the expense of those at the bottom of the income pyramid.

KEYWORDS:

• Housing finance
• community-based
• informal
• formal

1. Introduction

Despite the global spread of neoliberalism, access to institutional housing finance did not succeed in improving the habitat conditions of the poorer sections of society. Mortgage finance was to a large extent irrelevant for a majority of households, especially those at the bottom of the income pyramid (BOP). Moreover, low-income groups could not afford commercially-produced housing. In reaction, governments constructed a small number of complete housing units, but they had to shift to less expensive habitat solutions such as slum upgrading, sites and services programmes and core expandable unit projects. As a reaction, the World Bank promoted an enabling housing market approach in the early 1980s (Ferguson, Smets, & Mason, 2014; Smets, 1997).

The enabling housing market approach implies that governments should change their approach from housing construction to guiding the housing sector. Their role should include strengthening legal property rights, developing mortgage financing vehicles, rationalizing housing subsidies, providing infrastructure for residential land development, regulating land and housing development, organizing the building industry and creating a national institutional framework for managing the housing sector (Mayo & Angel, 1993). For housing finance, the enabling housing market approach suggests mortgage finance and subsidies for private developers. Along with this new approach, the World Bank simultaneously stopped funding serviced sites and slum upgrading projects and excluded them from the housing market (Ferguson et al., 2014).

The enabling housing market approach resulted in governments reducing their production of affordable housing, but some governments such as India’s continued providing affordable housing for the poor. A lack of interest in mass market housing among large housing corporations in the 1980s and 1990s led to many households having to depend on high cost supplies of housing and urban services. It is surprising that the enabling housing market approach, which came 25 years ahead of its time, focused almost exclusively on American and European models of housing production and finance, which did not match the needs of BOP populations in developing and emerging countries (Ferguson et al., 2014). These Western models were based on values that emphasize individualism and discourage cooperation. In them, collective interests are only possible when individuals aim at gaining profit for themselves. Neoliberalism has made this individualistic, selfish behaviour even more extreme (Gilbert, 2014; Sennett, 2013, p. 8). The adoption of such models implies that financial institutions prefer individual clients. In the Indian context, housing mortgage finance institutions include mainly banks and housing finance corporations.

In the past decades, financial institutions have had to deal with many financial crises around the globe. Caprio and Klingebile (in Lietaer, Ulanowicz, Goerner, & McLaren, 2010, p. 3) report that the World Bank has traced more than 96 banking crises and 176 monetary crises since the 1970s, which stem from US President Nixon’s floating exchange regime. According to Sassen (2014), such crises easily cross national borders due to globalization. For example, the 2008 financial crisis, which started in the United States, also had a detrimental impact on emerging cities in the Global South (cf. Ferguson et al., 2014).

Interventions can be put in place to cope with financial crises, but doing so does not renew financial institutions. In this respect, Lietaer, Ulanowicz, and Goerner (2009, p. 4) report system:

Whenever a bank that is too big to fail is in real trouble, the recipe has been the same since the 1930s: the taxpayers end up footing the bill to bail out the banks, so that they can start all over again.

Emphasizing improvements in the operating efficiency of financial institutions in crisis has led to the formation of similar kinds of institutions within the financial landscape, but ones with a more efficient approach. However, financial institutions have tended to continue their business in more or less similar ways as before a crisis, which implies that they are just waiting for the next crisis. A more fundamental change is required (e.g. Lietaer et al., 2009).

Apart from the impact of financial crises on the banking institutions and their methods of operation, there is another problem with the large size of housing loans. Remarkably, incremental housing and financing was widespread among those who lacked access to mortgage models. Moreover, the largest commercial banks discovered this market and began funding ‘unsecured’ home loans through different means such as ‘building materials retailers/hardware stores, microfinance institutions, credit cards, unsecured personal loans and more rarely micromortgages for account holders’ (Ferguson et al., 2014).

Still the majority of housing in developing and emerging countries has been built and financed incrementally, a housing construction process that has been described by Turner (1976). Such housing is built step by step as sufficient financing is available and a household is willing to invest in their house. Moreover, the financing must fit the livelihood strategies of the poor. This suggests that the poor could borrow with a maximum three-year term but that a shorter term would be preferable (Smets, Bredenoord, & van Lindert, 2014, p. 11). In practice, incremental housing finance is often a combination of different sources. In this respect, a distinction can be made between individual and community-based housing finance. Individual housing finance comes from friends, neighbours, relatives, moneylenders, pawnbrokers, employers and colleagues. Apart from these sources of housing finance, people can join hands to collect savings and offer credit facilities. Community-based finance can be used for many purposes of which housing is often one of the options. Housing purposes include funding individual house construction, housing maintenance, buying land and obtaining communal infrastructure such as a road, drainage, water distribution and connection, and community empowerment (Ferguson & Smets, 2010; Smets, 2004).

Although there are community programmes, such as credit unions, financial cooperatives (Ganapati, 2014), community-led development of affordable housing (Chitekwe-Bitti, Patel, & Mitlin, 2014), community contracting in Indonesia (Steinberg, 2014), community networking of savings groups in Thailand (Yap & de Wandeler, 2010), and the Community Mortgage Programme in the Philippines (Llanto, 2007), community-based initiatives are less widespread than the regular housing finance sector. More ideas can be generated from the informal sector, and these can expand ideas about good practices concerning community-based housing finance. To explore opportunities for a community-based approach to initiatives, the informal financial sector provides an interesting pool of good practices.

This paper explores the question: What inspirations for community-based housing finance initiatives can be gained from practices in the informal financial sector? To answer this question, I look at chit funds, a community-based housing-finance system in the informal financial sector in India. Chit funds can be in the form of rotating savings and credit associations, savings associations and accumulating savings and credit associations. After gaining insight into the different organizational forms, I will reflect on the operation of chit funds in relation to finance for housing and urban development.

2. Community-based finance: chit funds introduced¹

Community-based finance offers the possibility of mobilizing grassroots-level financing that can be used for urban developments such as improvements in land, housing and infrastructure for services. Community-based finance can have a complementary role to private and public sector organizations that face budget constraints or are not eager to invest in informal settlements (Shand & Colenbrander, 2018). Before looking at this complementary role, different types of community-based finance systems – called chit funds in India – will be explored below.

In a chit fund, people pool money in order to deposit their savings in a relatively safe place and gain access to a lump sum of credit or savings. The way finances are allocated to participants can take place in a democratic or authoritarian way. In a democratic chit fund, participants allocate the fund or parts of it, while in an authoritarian chit fund, the organizer makes the decisions. In other words, in a democratic chit fund, the collective is responsible for coping with default risk. However, in an authoritarian chit fund, this is the responsibility of the organizer(s) (Smets, 1992). Apart from different allocation mechanisms, chit funds can also have different organizational set-ups, which will be discussed below.

The origin of chit funds can be found in non-monetary mutual aid arrangements such as joint labour or production among peasants and urban craftsmen. Furthermore, financial self-help organizations have evolved from the need for financial assistance. Collective funds have been created for certain social events such as weddings and funerals. In other cases, such funds have been created for times of insecurity and emergency or for improving the living conditions of a certain group of members, for example, by laying a water pipe. An individual in need of cash may start an association by asking friends and neighbours to join and lend him or her money. It is repaid by inviting the lenders to gatherings, for which the borrower pays for the food and drinks (Schrader, 1991, p. 2). These financial self-help organizations provide security or insurance, but they also have an economic and socializing function (Bouman, 1994, p. 375).

The next three sections provide an overview of the three main organizational types of community-based finance, including options for housing finance: the rotating savings and credit association, the savings association, and the accumulating savings and credit association.

2.1. Rotating savings and credit associations

A rotating savings and credit association (ROSCA) consists of a group of participants who make regular contributions to a fund that is given, in whole or in part, to each member in turn. The first collector receives an interest free loan from all other participants, which can be used for housing, but also for other purposes. The last in line saves money while extending credit to fellow members. The other members alternate between net debtor and net creditor positions. They save until they receive the fund and start paying back. In short, a ROSCA is an intermediary between savers and borrowers (Bouman, 1978, pp. 36–37). The simplest form of a ROSCA operation is one with twelve participants, each contributing Rs. 100 monthly. Each participant in their turn will receive Rs. 1200 at the monthly meetings. Who will receive the pot each month is chosen by, for example, drawing lots. After twelve months, everyone will have received Rs. 1200 and the cycle is closed. Participants may decide to start a new cycle, possibly with new members and regulations. In this example, it is assumed that all members will pay their contributions to the kitty, even after they have received the fund. Furthermore, no deductions are made to compensate the organizer. More elaborate ROSCAs may include bookkeeping, issuing of receipts, written rules and regulations and contracts. Besides drawing lots, the rotation order may be decided by auction, seniority, negotiation, consensus, bribery or the decision of the organizer (see e.g. Ardener, 1964; Bouman, 1978, p. 38; Nayar, 1986; Smets, 1992, pp. 13–23). The lottery and auction systems will be discussed more extensively below.

In a lottery ROSCA, the rotation order is decided by drawing lots at every meeting or drawing the complete sequence at the beginning of the cycle. Suppose all twelve participants each pay Rs. 100 into the common pot, which is distributed to each member in turn. Each member receives Rs. 1200, but at different times (see Table 1).

Table 1. Financial transactions in a lottery ROSCA.

Months	1	2	3	4	5	6	7	8	9	10	11	12	Total
Contribution	100	100	100	100	100	100	100	100	100	100	100	100	1200
Funds received	1200	1200	1200	1200	1200	1200	1200	1200	1200	1200	1200	1200	–

Source: own research.

Although the organizer in this example does not receive payment for organizing and managing the ROSCA, he or she may receive a fixed share of the fund each month or the first draw during the ROSCA cycle. This is because, if there are unsolved default cases, the organizer has to make up the deficit. Since the organizer has to bear the losses, maintaining a certain reserve amount is necessary, which may legitimize the organizer receiving the first draw (cf. Schrader, 1991, p. 13).

The auction system sometimes used for determining rotation order is the Dutch auction system. In this system, participants can compete for obtaining the fund. The person who offers the highest discount on the kitty obtains the fund and the remaining amount, called the auction amount, is commonly distributed equally among all participants. At the beginning of the cycle more people are competing for the fund, which increases the auction amount and at the end of the cycle the auction amount is zero because competition lacks. This is the simplest form, but more complex models are possible (Radhakrishnan et al., 1975, pp. 6–7). Table 2 shows an overview of the financial transactions in an auction system and a cash flow analysis for the first and last recipients. It should be noted that the auction amounts listed equal the interest recipients have to pay.

Table 2. Financial transactions in an auction ROSCA.

Months	1	2	3	4	5	6	7	8	9	10	11	12	Total
Contributions	100	100	100	100	100	100	100	100	100	100	100	100	1200
Collected in fund	1200	1200	1200	1200	1200	1200	1200	1200	1200	1200	1200	1200	–
Auction amount	600	600	500	400	400	300	400	300	200	200	100	0	–
Fund re­ceived by winner	600	600	700	800	800	900	800	900	1000	1000	1100	1200	–
Dividend received by each person	50	50	42	33	33	25	33	25	17	17	8	0	333

Source: own research.

Participation in a ROSCA is not always based on expected profit. The social ambience of a ROSCA can be a decisive factor for some participants. Others may want to establish multifunctional relations with other participants or organizers to secure future help, financial and otherwise. Other factors may also be decisive for participation, such as limited or no access to other financial institutions, but also sociocultural, religious, and gender factors.

In general, sociocultural activities can work as a kind of glue to keep people together in a ROSCA. That glue is instrumental for the ROSCA’s operation. ROSCAs have a flexible nature and can adapt themselves to a changing environment. They often have informal, individually-based procedures and contain effective mechanisms that regulate membership eligibility, credit rating and repayment (Bouman, 1978, pp. 37–39; Nayar, 1986, p. 54). However, proper operation of a ROSCA requires mutual trust among its participants. Participants who have not yet received the fund have to trust that those who have already received the fund will keep paying their contributions until the end of the cycle. In addition to a more or less shared set of norms and values among members of a specific ROSCA, its cycle should not be too long so that observability and social control can be ensured. The most popular length of a cycle is about one to two years, which depends on context and location (Smets, 1992). In a ROSCA, people defaulting and people harmed by others’ defaulting know each other personally, which turns such incidents into personal affairs and possibly into big quarrels (Van der Harst, 1974, p. 69). Social control seems to be a formidable barrier against fraud and defaulting. This is easier in a small ROSCA because the small size ensures members’ knowledge of each other’s character (Bouman, 1978, pp. 37–38).

In general, the compulsory contribution to a ROSCA enables people’s access to a lump sum which can be used for improvement of their housing. Furthermore, membership in a ROSCA may be considered as a guarantee for (housing) loans from, for example, relatives or friends (see Smets, 2000; Van der Harst, 1974, p. 73).

2.2. Savings associations

In 1960, 227 members of the seminomadic Mandula tribe settled in Hyderabad. Nine years later they started a chit fund called ‘Sangam Chittie’. Each member was asked to contribute Rs. 0.5 a month to repay a loan of Rs. 5,000, which had been taken out for the construction of a community hall. After three years, the loan was repaid, but the payments continued. In 1972, the monthly contribution was raised to Rs. 1. This money was pooled in a common fund for community purposes such as the construction of a temple. By 1986, the common fund had grown enough that it could offer Rs. 25,000 for the temple’s construction. For additional financing, Rs. 25 was collected from each house (based on Smets, 2006a, p. 157).

As described above, people can join together to save for a common purpose such as the construction of a community hall and temple. In a basic savings association, or SAVA, a term introduced by Smets (1996), people pay regular contributions of a fixed or variable size to a common fund, from which no credit will be provided. The fund accumulates for a predetermined period, after which all participants receive their deposits back (Schrieder & Cuevas, 1992, p. 46). However, there are also savings associations whose cycles will not be terminated. Here, the pooled savings accumulate in a fund that will be used for emergencies, improving public facilities, or another defined purpose. For the duration of the cycle, the fund is kept at the treasurer’s home or in a savings account at a bank (Schrieder & Cuevas, 1992, p. 46).

Joining a SAVA can be made attractive to potential participants by introducing a prize that one or more people can win. The prize could be a specific sum of money or a good such as a house, scooter, or refrigerator. This lottery system differs from the type of lottery in which there are winners and people who lose their entire contribution. In a SAVA, all participants win a prize, but their input differs. This can be illustrated by the following example of a SAVA in which participants can win an apartment among other prizes. However, one should realize that the target group of this SAVA does not include the poor.

The 650 participants in the lottery SAVA are supposed to deposit Rs. 500 a month to a common fund. The SAVA’s cycle is twenty months. Every month a lucky draw takes place. Only the people who have made their monthly payment before the 15th of the month are allowed to participate in the lottery. The winner gets a prize and does not need to continue payments. In addition to the monthly prize, ceiling fans, scooters, mopeds, TV sets and refrigerators can be won. In the 19th month, the only prize is a one-bedroom flat (Author’s research).

The type of SAVA described above is also known as a share game, prize chit or lottery SAVA. Since the number of participants is much larger than the number of draws, there is no relation between the large number of participants and the lifespan of the association. As soon as people receive their prize, they drop out of the scheme (see also Schrader, 1991, p. 80).

The organizer in a lottery SAVA controls a large sum of money, which accumulates throughout the cycle. At the end of the cycle, the organizer has to pay out to the non-winners. Providing a certain percentage of interest or a present can increase the amount of savings. Saving is compulsory for participants in a lottery SAVA until they have won a prize. Thus, defaulters are not a problem, because players who drop out do not affect other participants. As noted in the example, an incentive for members to pay their monthly savings in time is exclusion from the prize draw for late payments. Depending on the organizational set-up of a lottery SAVA and its operation, the organizer can make a profit by putting the total contributions minus prize expenditures into a savings account or by using them as working capital (Schrader, 1991; Smets, 1992).

Shopkeepers or land developers can also use these schemes as marketing mechanisms. In this way, shopkeepers can increase their sales and get access to relatively cheap working capital (see Smets, 1992, p. 19). Land developers also make use of these schemes to sell plots and use the lottery mechanism to attract clients, as the following example illustrates:

In a suburb of the twin city of Hyderabad-Secunderabad, a scheme has been organized to sell serviced plots of 250 square yards, which are only allocated after the required sum has been completely paid. A plot costs Rs. 22,500–25,500, which can be paid in a lump sum or in instalments. An additional membership fee of Rs. 200 is required. For a period of fifty months, 2,000 applicants have been grouped together in a scheme with prizes worth Rs. 5,000,000. There are monthly lucky draws for a free plot, a colour television and a washing machine. Every sixth month the prizes consist of a motorcycle, a refrigerator, a video recorder, and ten prizes of Rs. 1,000. And once a year, a car, a gold ornament worth Rs. 50,000, a video recorder and twenty consolation prizes of Rs. 1,000 are awarded. There is also a mega draw for a car worth Rs. 350,000, gold ornaments (Rs. 100,000, a four-door refrigerator and thirty prizes of Rs. 1,000 each (Author’s research).

The incentive of possibly winning a prize and still obtaining the desired item can encourage the participation of rational actors enormously. Moreover, such schemes could also be used for selling building materials, which would help the poor.

2.3. Accumulating savings and credit associations

From noon onwards onwards, chit fund organizers sit at four places outdoors to collect money from their clients. Meanwhile, four 20-member groups come together in a small room to repay their credits to the treasurer. In addition, loans are provided from the common fund, which holds about Rs. 20,000: three loans of Rs. 300 and one of Rs. 500 are available. The participants of each group have to bid for these loans, with a minimum bid of Rs. 15 for a Rs. 300 loan, and Rs. 25 for a Rs. 500 loan. The highest bidder gets the loan and has to pay the bid as interest. The next month, new loans can be provided after the outstanding credit and its interest has been repaid (Author’s research).

This chit fund with its four groups has its roots in 1976, when twenty people joined and came to the house of a slum leader to pay their monthly contribution of Rs. 300 each. This made a fund of Rs. 6,000 from which loans could be provided to the participants. When sufficient money was available in the fund, the members decided to stop paying contributions. Later on, new members were accepted one by one, but each had to be guaranteed by one of the members.

The chit fund in this example is an accumulating savings and credit association (ASCA). Rutherford (2000) came up with this name. The ASCA resembles credit cooperatives or credit unions. In an ASCA, just as in a SAVA, the pooled savings are not directly redistributed, but they accumulate in a fund for a specific time. The difference is that the fund created in an ASCA can be used partly or completely to provide loans either to members only or less commonly to non-members. Part of these savings may be kept in a fund for emergency cases or used for community development or joint investment in e.g. upgrading a settlement (Bouman, 1994, p. 376). In contrast to a SAVA, loans can be provided from the common fund.

Loans are made available at weekly, monthly or other time intervals. The provision of credit can be in a predetermined order, on a first-come-first-served basis, or decided by consensus, negotiation, seniority, bribery or the organizer. Loans may even be granted to a member who offers the highest interest payment (Mehta & Mehta, 1991, p. 1110). Normally, the loan amount is such that it corresponds with the repaying capacity of the recipient. To ensure repayment of the credit, a guarantor is often needed. During a cycle, some members may take out more than one loan, while others do not need to borrow from the fund and use the ASCA more as a savings club, with the option of making profits or saving for future consumption or security. Loans can be repaid in a lump sum or in instalments within a prescribed period. If the loan cannot be repaid in time, the savings are taken from the borrower’s savings account and that of the guarantor(s) if necessary. At the end of the cycle, the accumulated fund is partly or completely redistributed among its members. The most common cycle is about one year, during which participants may use the fund for a specific purpose, for example, paying taxes and school fees or for a religious ceremony such as Diwali, Eid-Ul-Fitr or Christmas (Bouman, 1994, p. 376; Smets, 1992, p. 67).

To increase the common fund, entry fees, shares, (high) interest rates on loans, and penalties on delayed payments of contributions and loan instalments may be charged. Interest rates may be up to 13% per month, which makes the fund accumulate rapidly (Smets, 1992, pp. 29–32). However, a share in the profits will often compensate for the high rates. ASCA participants may decide to distribute only part of the fund at the end of the cycle and start a new cycle with the remainder. Consequently, the accumulation process extends over a longer period and the financial mediation becomes more permanent in nature, which creates opportunities for long-term credit. However, risks of safekee­ping, embezzlement and loan defaults, as well as administration and transaction costs, increase (Bouman, 1994, p. 376; Smets, 1992, p. 67). If many members pay high contributions, large amounts of money may circulate in an ASCA. To manage this properly, strong discipline on the part of the organizer(s) and members is necessary (Smets, 1992, p. 72). Compared with ROSCAs, ASCAs are less socially and more financially oriented. The value orientations of an ASCA’s participants do not need to be homogeneous to ensure its smooth functioning. The social control mechanism is partly replaced by a system of rules and regulations and control by an organizer or board (Smets, 1992, pp. 66–68).

So far, different types of financial self-help organizations have been described. The next section will focus on what insights can be derived from the operation of chit funds for establishing community-based housing finance.

3. Community-based housing finance: potentials and dangers

Chit funds can be organized on the basis of geographic location, age, gender, community, ethnic background or place of work (see also Mehta & Mehta, 1991, p. 1110). People join in a chit fund to meet common needs such as housing and facilities and having a safe place to deposit savings and access lump sums. Such savings help the poor cope with irregular incomes and pay for needs such as health care, education, emergencies, and income-generating activities. Generally, working together in a chit fund is instrumental for meeting individual needs unless resources are being pooled for a common purpose, such as housing improvements or settlement upgrading. Gaining access to a chit fund usually requires that potential members belong to a network because participation is dependent on introduction by a known person or on being known as trustworthy by the organizing body. In other words, access to a chit fund is restricted to those who are considered trustworthy or are being guaranteed by those who are known as being trustworthy. In addition, pooling collective savings can also create horizontal relations. In larger chit funds, face-to-face relations decline. Consequently, formal control mechanisms are increasingly needed for the survival of larger groups (cf, Hechter, 1988, p. 105). Controlling access to a chit fund is essential because whether the chit fund’s collectively preferred goals, such as settlement improvement, can be met depends on participants’ behaviour. For an overview of the main characteristics of community-based housing finance systems see Table 3.

Table 3. Overview of the main characteristics community-based housing-finance systems.

	ROSCA	SAVA	ASCA
Allocation mechanism of finance	Rotating	Accumulating savings	Accumulating
Number of participants	Small	Small – large	Small – large
Cycle	Short	Short – long-term	Short
Orientation	Financially and social oriented	Financially, to lesser extent socially	Financially, to lesser extent socially
Control mechanism	Mainly face-to-face in combination with rules and regulations	Mainly rules and regulations	Face-to-face in combination with rules and regulations
Management system	Manly democratic	Mainly authoritarian	Mainly authoritarian

The question is, What could community-based finance systems imply for housing and urban development? In addition to the insights provided above, below are some lessons learned for setting up community-based (housing) finance systems (see also Lehmann & Smets, 2014).

Firstly, peer pressure in a community encourages participants to pay on time. In cities, it is easier to escape peer pressure. The chance of escaping peer pressure is lower when potential defaulters are tied to their place of living – once they reside in their own house.

Secondly, participation in a chit fund makes financial behaviour – saving regularly and paying on time – visible to other participants. This behaviour could work as unconventional collateral that helps obtaining money from other sources. In other cases, guarantees of trustworthiness can be requested to ensure the repayment of credit.

Thirdly, participants in a community-based housing finance scheme should consider the scheme as ‘ours’ and not ‘theirs’ (e.g. the government’s, bank’s, NGO’s or community leaders’). Once participants see a scheme as ‘ours’, their involvement and contribution to the success of the scheme may be enhanced. However, if participants become used to welfare schemes in which aid is provided free of cost, they may expect that credit will be remitted in the future. In this respect, Wright (2000) uses the concepts of hot and cold money. Hot money is seen as ‘ours’ and is often characterized by a face-to-face approach. Here personal interaction is practiced, and money is controlled by those who own the financial means. Hot money is used in many ways in finance involving organized and non-organized citizens partaking in financial arrangements. Cold money comes from outside agencies with whom participants do not have mutual responsible relationships. Here, the feeling of responsibility to repay a loan diminishes once that money is not considered ‘ours’, but ‘theirs’. The norms and values of the community can keep out unscrupulous people who go only for personal benefit even if it negatively affects the community.

Fourthly, a lottery SAVA offers possibilities for buying plots, building materials and facilities. Smets (1998) developed a toilet scheme, which unfortunately could not be implemented by UNICEF. However, the design, including a financial overview (see Table 4), can still be used as a source of inspiration:

The lottery SAVA scheme deals with low-cost sanitation units of the two-pit, pour-flush, waterseal latrine type costing Rs. 1,200 each, which would be provided on a matching grant basis. In the proposed scheme, 200 low-income households participate and pay Rs. 50 a month. After twelve months each participant has deposited Rs. 600, which is half the unit costs. To encourage timely payments, a lottery will be held monthly. Only the persons who have paid their contribution before a specific deadline are allowed to participate in the lottery. (…) Every month two winners get the latrine installed and can stop paying their contributions. In the final month there are also two prize winners, each receiving back their entire contribution paid during the cycle (Rs. 600). This last option is the best that can happen to a participant (Smets, 1998, p. 63).

Fifthly, chit funds can mobilize small amounts among participants, but these amounts are often too small for investing in substantial improvements in land, housing, and service infrastructure in informal settlements. Organizations such as Slum Dweller Federations, Shack/Slum Dwellers International, and Asian Coalition for Housing Rights use collective savings to mobilize additional finance from public and private sector organizations. Bridging formal and informal systems enables the formation of larger funds as well as co-production and the use of cross subsidies, whereby the better off contribute towards housing the poor. When informal and formal financial systems are combined, such co-production helps participants in community-based finance systems learn negotiation skills. In other words, the community-based housing finance collectives strengthen horizontal relationships and learn to cope with vertical relationships that enable a shift towards governance, making urban transformations possible. In addition, the linkage between informal and formal also leads to changing perceptions of legitimacy (Schermbrucker, Patel, & Keijzer, 2016; Shand & Colenbrander, 2018).

Table 4. Financial overview of a low-cost sanitation scheme.

Month	Number of participants	Beneficiary contribution	Total contributions	Cost latrines	Collections minus costs**
1	200	50	10,000	2,400	7,600
2	198	50	9,900	2,400	7,500
3	196	50	9,800	2,400	7,400
4	194	50	9,700	2,400	7,300
5	192	50	9,600	2,400	7,200
6	190	50	9,500	2,400	7,100
7	188	50	9,400	2,400	7,000
8	186	50	9,300	2,400	6,900
9	184	50	9,200	2,400	6,800
10	182	50	9,100	2,400	6,700
11	180	50	9,000	2,400	6,600
12	178	50	7,700*	213,600	−205,900
	Total		112,200	240,000	−127,800

*This amount is the collected contributions minus the payment of Rs. 600 to the two winners each in the last month of the cycle.

**Note that the organizer obtains relatively cheap working capital in an incremental way.

Source: Smets (1998, p. 64).

Sixthly, community-based housing finance systems could also be promoted outside India. One way of spreading alternative schemes is by publishing good practices on websites, but a more powerful option is exchanging the experiences of groups who participate in such schemes. Organizations such as Slum Dweller Federations, Shack/Slum Dwellers International and Asian Coalition for Housing Rights exchange their experiences with collective saving, but they could also look into and exchange experiences with different community-based finance systems.

This paper has focused on housing finance from the perspective of community-based organizations. Today, some elements from community-based finance systems in the informal sector have been incorporated into banking schemes and insurance mutuals. For example, in Bangladesh, powerful NGOs used shame and honour to pressure women into participating in savings and credit groups, which resulted in increased indebtedness and problems with repayment. Moneylenders provided high-interest loans so the women could repay other credit and conflicts within households and communities increased. Moreover, multinationals sold consumer products through the financial groups. Thus, moneylenders and multinationals gained enormously at the expense of the BOP (Karim, 2011). Bähre (2014) reports that insurance mutuals in South Africa benefited the better off and the state. Instead of trickle-down mechanisms that favour the BOP, trickle-up economics shows that mutuals channel cash to the state and companies. Both cases show that money has been extracted from the poor to benefit the better off.

4. Conclusion

Neoliberalism goes hand in hand with the development of institutions that serve individuals and generally see community approaches as complicated and undesirable. As (housing) finance institutions follow suit, they suffer from their efficiency-driven approaches and forget to reflect on the effectivity and sustainability of their operating methods. Against this background, community-based housing finance tends to be taken less seriously. Housing finance tends to be trapped in a one-sided way of looking at customers. Though this article promotes community-based housing finance, such financing is not the solution for improving housing conditions and urban development. However, it is a necessary if insufficient factor. For a sustainable landscape for housing finance, a kaleidoscope of different kinds of arrangements and institutions is required. Also needed is access to elements of the housing process such as building materials, a plot to build on and construction skills, and access to other financial sources such as savings. Chit funds offer the possibility of obtaining housing financing. People join chit funds to meet their financial but also sociocultural needs. It is likely that participants’ sociocultural and psychological needs can be met more effectively in smaller chit funds, which are often of a ROSCA nature. When a chit fund is larger, financial and economic functions tend to dominate. These differences also have repercussions for the operation of particular groups. In smaller groups, face-to-face relations often enforce social control, while in larger organizations, these mechanisms are (partly) replaced by the introduction of guarantors and rules and regulations. In chit funds where face-to-face relations dominate, the mechanism of social pressure to exercise self-constraint dominates, but in leader-oriented or authoritarian chit funds, where participants do not know each other well, a mixture of external constraints and social pressure to exercise self-constraint is in operation. Moreover, if savings and credit are not linked, the potential for defaulting and free ridership increases. Not all types of financial self-help organizations closely fit the livelihood strategies of the poor.

The operation of chit funds has shown that there is a huge potential for financial self-help among the BOP, but though the mobilized financing can enable incremental building, it is insufficient for settlement upgrading and urban development. Therefore cooperation with public and private sector organizations could also play a role in pooling financial resources. One should be careful in co-creation and linkages with larger organizations. Especially when commercial organizations and others such as exploitative moneylenders use these groups, the poor do not benefit. In these cases, there is a trickle-up mechanism by which means and sources are extracted from the BOP to benefit the better off. Community-based housing finance is an alternative when the motto is ‘small is beautiful’ (Smets, 2006b), but if small units become encapsulated by commercial firms and unscrupulous entrepreneurs, exploitative activities tend to develop. This is a problem that should be taken seriously. A limitation of this study is that the legal setting for different types of community-based housing finance within specific countries has not been investigated. Future studies should focus on that aspect.

Acknowledgment

The author would like to thank Paul van Lindert and anonymous referees for their helpful comments during the preparation of this paper.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1 This section is partly based on Smets (2004).