DEPARTMENT OF ECONOMICS Uppsala University
D-level thesis Author: Erika Calles Supervisor: Javad Amid Spring 2005
Microfinance according to SafeSave - a better way to target the poorest?
A Minor Field Study from Bangladesh
Abstract
Poor people often lack collateral, which is one of the reasons that they have no access to formal financial institutions. Microfinance institutions (MFIs) provide financial services to poor people. Traditional MFIs have received some criticism, for instance that they do not target the poorest of the poor. This paper, with a field study from Dhaka, takes a closer look at SafeSave, a new MFI working in a quite different way than the traditional MFIs in
Bangladesh. The conclusion of this paper is that SafeSave’s more flexible services are able to reach the poor better than the services of traditional MFIs, but might not be the best solution seen from a long-term development perspective.
Keywords: microfinance, group-lending, savings, Bangladesh.
Acknowledgements
I would like to take this opportunity to acknowledge all those people who helped me with this paper.
First, I would like to thank the Department of Economics at Uppsala University and the Swedish International Development Cooperation Agency for giving me the financial resources necessary to make this study possible.
Many people in Dhaka provided great help for my research. I would especially like to thank my field supervisor Nipun Sangma for sharing his time and knowledge, and explaining about microfinance and the life in Bangladesh in general. I am greatly indebted to him. Also, many thanks to Ripon Islam for his invaluable help during the days in the field. I would also like to thank all the collectors and clients, no one mentioned and no one forgotten, that I interrupted in their daily work with sometimes probably very strange questions.
In Dhaka I would also like to thank all the people at Viator, staff as well as guests, for
information, discussions and help in so many different fields. Special thanks of course to Siv- Helene and Elin, Monica, Gunilla and Becky for making the time in Bangladesh a good memory.
In Sweden I would especially like to thank Karin Edmark at the Department of Economics at Uppsala University for all the encouragement and support both before and during my time in Bangladesh. Also, I would like to thank Javad Amid for help in finishing the paper.
Uppsala 2005-05-22
Erika Calles
TABLE OF CONTENTS
1. INTRODUCTION... 5
2. BACKGROUND: FORMAL FINANCIAL MARKETS AND MFIS ... 6
2.1LACK OF ACCESS TO FORMAL FINANCIAL MARKETS... 8
2.2MICROFINANCE AND ITS MECHANISMS... 10
2.2.1 Group-lending ... 12
2.2.2 Dynamic incentives... 14
2.2.3 Regular repayment schemes ... 15
2.2.4 Collateral substitutes... 15
3. MICROFINANCE IN BANGLADESH ... 16
3.1PROBLEMS WITH AND CRITICISM OF TRADITIONAL MFIS... 17
4. A FIELD STUDY OF SAFESAVE... 20
4.1CLIENTS AND ORGANISATIONAL STRUCTURE... 23
4.2SAVINGS... 25
4.3LOANS... 25
4.4THE PERFORMANCE OF SAFESAVE... 27
5. COMPARISON OF SAFESAVE AND TRADITIONAL MFIS... 28
6. CONCLUDING REMARKS... 32
APPENDIX. SAFESAVE PRODUCTS ... 37
1. Introduction
Poor people need and, contrary to a common impression, use a variety of financial services including savings deposits and loans. However, poor people often do not have access to formal financial institutions. One of the reasons is their lack of collateral, which means that getting access to for instance loans is often impossible. Also good facilities for savings are often also missing. The alternative for borrowing is for instance private money lenders, who may charge very high interest rates. This makes it difficult for poor people to access funds for starting small income generating activities. Microfinance institutions (MFIs) give small loans to very poor people, often for self-employment projects. This helps them generate income that allows them to care for themselves and their families and climb over the poverty line.
According to the 1997 Microcredit Summit Declaration, the availability of financial services to poor people will empower them to end their poverty. It is also stated that microfinance will contribute to virtuous circles for women, helping them not only to get economic
empowerment, but also in a wider sense social and political empowerment. All these questions are crucial to the wider development issue.
MFIs, like the well-known Grameen Bank, have received global recognition for their group lending model and their loan repayment rates, which are almost always above 95%. The programs have also proven to be able to reach poor individuals, particularly women, who have been difficult to reach through alternative approaches (Morduch, 1999b).
During the 1970s and much of the 1980s MFIs were characterised by an integrated package of credit and training. Since the 1980s the field of MFIs has grown substantially. Today a large number of institutions exists that focus on providing financial services only. Many MFIs have moved a long way from the Grameen Bank model, which pioneered the field of microfinance with its very standardized loans to rural women, towards more commercial institutions with a wider range of financial services beside credit. The traditional Grameen Bank model has received a lot of attention due to its high repayment rates. It is also claimed that it has helped many people to improve their situation and contributed to empower women in social,
political, and economical ways. Today most MFIs in Bangladesh work according to the traditional Grameen Bank group-loan model (Matin et al, 2000).
However, many people, especially the very poor, still lack access to basic financial services.
The challenge of providing financial services to the poorest still remains. SafeSave is a fairly new organisation working in the slums of Dhaka, with methods that are quite different from those of traditional MFIs in Bangladesh, for instance the organisation does not lend to groups.
SafeSave claims that they are better than the traditional organisations in reaching the poorest.
It is thus interesting to find out how they work compared to the traditional organisations in Bangladesh. The focus of this study is therefore: are the methods used by SafeSave more efficient in reaching the poorest? Would their methods also work in other settings than the present? And what about the gender perspective?
This paper will be based on earlier research in the field, as well as a minor field study conducted in Dhaka, Bangladesh during February and March 2005. The aim with the field study was to get a first-hand knowledge about the methods and means of SafeSave in order to be able to analyse and compare it to traditional MFIs.
Following this brief introduction, Chapter 2 is a background chapter that will analyze the demand of the poor for financial services. It includes a discussion on the lack of a formal financial market for the poor and where microfinance can step in. A discussion on different mechanisms used by traditional MFIs in lending to poor households is also included. Chapter 3 discusses the microfinance situation in Bangladesh, and problems and limitations with the traditional models and mechanisms. It also gives an introduction to and method for the field study. Chapter 4 contains the field study of SafeSave’s financial services. Chapter 5 analyzes the findings of the field study and compares SafeSave and traditional MFIs. Finally, Chapter 6 concludes the paper and tries to put the results in a broader perspective.
2. Background: formal financial markets and MFIs
This chapter comprises two sections. The first section discusses the demand of poor people for financial services and reviews the inability of formal financial markets to meet poor people’s demand. The second section defines microfinance and discusses what microfinance can do to solve the previously discussed problems. It also reviews different mechanisms used by traditional microfinance organisations.
An assumption in the promotion of microfinance is that the poor, just as everyone else, have a demand for credit and that they have a need for savings facilities.
Poor people sometimes need to spend amounts, that are from their horizon large, but that are too small to be of interest for conventional banks and other financial organisations.
Rutherford (1999) mentions three primary categories within which poor people need access to amounts of money that are larger than what they normally have at their disposition. These are life-cycle events, emergency needs, and investment opportunities. Among life-cycle needs can be mentioned weddings, child birth, education and old age as well as the desire to leave something to heirs. Then there are the recurrent festivals. In each case the poor need to be able to get their hands on sums of money which are much larger than the amounts of cash which are normally found in the household. Emergencies that create a sudden need for a relatively large sum of money can be personal, like sickness or injuries, or impersonal, like war, floods and fires. Also, poor people, like all people, have a wish to invest in costly items that make life more comfortable, like better roofing or a fan. There may be opportunities to invest in a business, or to buy land or other productive assets. (Rutherford, 1999) The same categories are valid also for people with larger assets. The difference is that poor people have greater difficulties in handling such problems. If you have assets you may have reserves to use on such occasions, but poor people do not usually have any reserves. It is difficult to have large amounts of money available to handle costly events, when you need all your assets to take care for the daily survival.
All over the world, people save in a variety of forms. It could be at home, in rotating savings and credit associations (ROSCAs), and in other savings and loan associations. Some people pay informal collectors to hold their savings safely. The savings are not always in cash, they can also be for instance animals, crops, or jewellery. (Robinson, 2001:224)
The vulnerability of the poor is another important factor when it comes to have access to financial services. Structural factors, such as seasonality, inflation, or the weather are things that can make a person vulnerable. Operating an enterprise can also increase peoples’
vulnerability. With access to financial services households can transform small sums of savings into usefully large sums, and therefore can go a long way in decreasing vulnerability.
To reduce their vulnerability, people can use a variety of preventive strategies. These strategies include diversifying income sources, building up physical, financial, human and
social assets, and focusing on good money management. All these strategies rely on access to financial services (Consultative Group to Assist the Poor, 2000).
2.1 Lack of access to formal financial markets
Financial services are rarely accessible for the poor through the formal financial sector. Credit is widely available from informal moneylenders but typically at a very high cost to borrowers, especially poor borrowers1. Informal savings clubs does also exist.
The lack of access to formal financial markets for the poor is due to a number of factors. The most important are small sums of credit required by the poor, location, risk-avert lenders, lack of collateral, and that benefits to society are not equal to benefits to the lender.
The behaviour of a lending institution is affected by the risk. It is difficult for lending
institutions to determine the default risk of borrowers (the screening problem). It is also costly to make sure the borrower repays (the monitoring problem), as well as it is costly to compel repayment (the enforcement problem).
An MFI must charge interest rates high enough to cover the cost of its loans to maintain and increase its services over time. The costs the MFI must cover are the cost of the money it lends, the cost of loan defaults and transaction costs. MFIs must charge rates that are higher then in formal banks to be able to keep their services available for risky borrowers. These rates are, however, far below those of informal money-lenders. (Consultative Group to Assist the Poor, 2003b)
The cost of lending is high when it comes to poor people. This is one of the reasons why formal financial institutions choose not to involve in microfinance. Poor people often live in rural areas. The customers are often geographically dispersed. Most banks are located in cities or large villages. Branch offices are not located in remote rural areas because such locations would be expensive for the bank. This results in a lot of travelling for the rural poor, just to deposit a relatively small sum of money. Obviously, this is not practical and in many cases not possible either. For example, rural women must generally care for their family’s basic
1 According to Morduch (1998), the interest from money lenders is often as high as 10% per month.
needs, so there is often no time to spare to travel several hours to visit a bank. They may not even be able to leave home due to cultural rules. (Dankelman & Davidson, 1988)
The cost of collecting information about the clients is also high. Usually poor people lack collateral. A number of studies (e.g. Robinson, 2001 and Hulme & Mosley, 1996) points out that poor people are usually unable to inform financial markets about their creditworthiness when they cannot present any collateral. If banks do not know enough about their customers they can not trust the customers to pay back. Very few people are known to a bank, which is an important reason why banks do not want poor people without collateral as customers.
Poor, rural borrowers are often considered risky. Their living conditions are marked by insecurity and risk. The default risk is high. Banks are unwilling to lend money because of failure to repay seems to be high. Lending to groups instead of individuals solves this issue by distributing the responsibility for the loan, which from a bank’s point of view proves a kind of insurance cost-free.
Lending institutions may also fail to lend to socially beneficial projects when they are not profitable for the lenders themselves (Hulme & Mosley, 1996). This could result either from a difference between prevailing prices and free market prices or from the inability of lenders to consider the social benefits of loans.
There are above all two different informational problems in lending too poor people. Adverse selection arises when the borrowers know more than the lenders. Markets may not function properly when the loan providers can not set fair premiums based on accurate measures of expected loss. A loan provider would have to constantly monitor each borrower to get the premiums right (Nicholson, 2002). When using high interest rates, risky borrowers will be attracted, as only risky projects will have returns high enough to cover the interest costs.
Moral hazard occurs when the loan contract affects the actions of each borrower. Individuals may undertake a variety of actions to influence that a risky event will occur. If a person is fully insured against losses she will have a reduced incentive to undertake costly precautions and may therefore increase the likelihood of a loss occurring. (Nicholson, 2002) A high interest rate will raise the lenders return on successful loans, but will also give each borrower an incentive to use riskier techniques. When a borrower does not bear the downside risk of
her actions, she will undertake riskier projects, making it less likely that the loan will be repaid (Perkins et al, 2001:509). The borrower may undertake riskier activities than they normally do, in order to earn enough money to pay the instalments.
In many developing countries, banks tend not to mobilise small sums of savings for poor people either. Reasons for this could be that banks assume a lack of demand for this as well as they think it would be unprofitable. (Robinson, 2001:229)
Poor peoples’ lack of savings and difficult access to capital makes it difficult to become self- employed and to undertake productive employment-generating activities. Informal money- lenders often exist and play an important role. However, they usually charge very high interest rates that are too high for poor people. Informal groups, such as savings and credit associations, can meet occasional needs for the household; however, they can not be regarded as a reliable source of finance for income-generating activities. Microcredit programs have been developed to reach the poor at an affordable cost and can thus help poor people for instance to become self-employed. (Khandker, 1998:2)
2.2 Microfinance and its mechanisms
Microfinance refers to the provision of small scale financial services, primarily credit and savings to poor and disadvantaged people. (Robinson, 2001:9) Generally, the provision of credit services has been more thoroughly promoted than the provision of savings facilities to the poor. Thus the emphasis on microfinance has been the support of credit. Microcredit has been specifically defined in the Microcredit Summit Declaration (1997) as programs which give loans to very poor people for self-employment projects, which will generate income, allowing them to care for themselves and their families. Savings services allow people to store excess liquidity for future use and to get returns on their investments. On the whole, microfinancial services intend to help low-income people to reduce risk, improve
management, raise productivity, obtain higher returns on investments, increase their incomes, and improve the quality of their lives and those of their families. (Robinson, 2001:9)
The typical clients of any microfinance institution are poor people who do not have access to formal financial institutions. Microfinance clients are typically self-employed micro
entrepreneurs. In rural areas, they are usually small-scale farmers and others engaged in small
income-generating activities such as food processing and petty trade. In urban areas, microfinance activities are more diverse and include for instance shopkeepers, service providers, artisans, and street vendors (Citigroup Microfinance Forum, 2004).
‘The poor’ is a much larger group of people than for instance small-scale farmers. Often forgotten in the discussion are women and children. (Yunus, 1999:70) Women in developing countries usually belong to the poorest people in society. For different reasons they seldom have any assets of their own. For example, in Bangladesh, purdah2 keeps women inside their houses and their social value is regarded as low. Traditionally women have been the main clients in MFIs.
Participation in microfinance programs provide women with access to knowledge and information that help them interact with the outside world and permits the building and strengthening of social networks. By increasing women’s economic contribution to the household, participation in a microfinance program can improve their sense of self-esteem and control over assets (Consultative Group to Assist the Poor, 2000). Loans given to women create better effects than those given to men. It has been shown that when women get access to financial services their nutrition status improves, child mortality declines, the sanitation facilities improve and the use of contraceptives increases more than when men receives the loan (Yunus, 1999:24). These are all important steps in poverty reduction.
If microcredit is an appropriate tool for reducing poverty or not depends on local
circumstances. Poverty is often the result of low economic growth, rapid population growth and very unequal distribution of resources. Unemployment and the low productivity of the poor are among the determinants of poverty. One way of reducing poverty could be by creating jobs, another to invest in human and physical capital to increase workers’
productivity. Khandker (1998:1) states that in many countries, among others in Bangladesh, poverty is caused by a lack of both physical and human capital and thus, the best way to reduce poverty is to deal with both problems; increasing productivity by creating employment and developing human capital. This is where microfinance can play an important role.
2 Literally a ‘curtain’ or ‘veil’. Refers to a range of practices in response to the Koranic injunction to guard women’s modesty and purity. (Yunus, 1999:93)
Microfinance institutions employ a diversity of approaches in lending to low-income households. The method that has been given the most attention and is the most important is group-lending. Group-lending is one way for a formal lending institution to reduce the costs of screening, monitoring and enforcement. There are also a few other measures often used in combination with group-lending. The following sections present some of the different approaches used by traditional microfinance organisations. These methods are also used by the dominating MFIs in Bangladesh, and these are the methods that will later be compared with the methods of SafeSave.
2.2.1 Group-lending
One of the reasons poor individuals do not receive credit is that lenders have few means to monitor the use of borrowed amounts to ensure repayment. Joint liability is the central feature in group-lending. This allows lenders to provide credit to people without physical collateral.
Lenders need some kind of collateral to enforce repayment. With joint liability clients use each other as guarantors of loans; the screening, monitoring and enforcement of the loans are passed to the peers. Each member of the group is liable for the loans of any other member in the group. People lend on “social” collateral instead of economic assets. Furthermore, loan delivery costs fall dramatically due to the simultaneous delivery of multiple loans and reduced time needed for credit approval. (Karlan, 2001) Among the most important mechanisms behind group-lending are the informal advantage mechanism, the peer pressure mechanism and the informal insurance mechanism.
MFIs, as other creditors, face the problem of lending with imperfect knowledge about the creditworthiness of their clients. The clients themselves are the best source of information regarding the creditworthiness of their group members and can solve problems of asymmetric information between the MFI and the borrowers. Members of the same community generally have good knowledge about who is an economically reliable person and who is not. This is the informal advantage mechanism and it works if clients are allowed to select their group themselves. (Morduch, 1999b and Ledgerwood, 2001:70) The reason for a borrowing group to form itself rather than being formed by the MFI is that the group solidarity will be stronger in this way. (Yunus, 1999:105). A client tends to select group members that he/she thinks would be able to manage their credits in a good way. Ghatak (1999) demonstrates that this sorting process can contribute to lower interest rates and to raise social welfare. Group-
lending reduces certain institutional costs; it is cheaper to give a loan to a group than to each individual in the group. MFIs can save money by using a hierarchical structure (Ledgerwood, 2001:70). Most of the time loan officers from the MFIs do not deal with individual group members, but with a group or village leader.
By screening and monitoring costs to the group, a MFI can reach a large number of clients even when the asymmetric information through the self-selection of groups exists.
(Ledgerwood, 2001:70)
Once group members have been selected another mechanism must exist to ascertain that individual borrowers repay their loans. (Ledgerwood, 2001:70) The self-selection process is not enough to create high repayments. Although a commercial bank can attempt to monitor business and life outcomes for individuals, it is both difficult and costly to do so. Group- lending mechanisms provide incentives to the borrowers to monitor each other and this is where the peer pressure mechanism steps in. (Karlan, 2001) Pressure from other group members can act as a repayment incentive, since members do not want to let their fellow group members down or suffer social exclusion from the group as a result of failure to pay.
(Ledgerwood, 2001:70). Through groups, members also gain a sense of collective protection and empowerment. (Nguyen Tien Hung, 2004) Even if this mechanism can be effective in stopping entire groups from failing, it does not solve the problems of moral hazard or free- rider incentives from the individual borrower that arises from joint liability. (Stiglitz, 1990).
Moral hazard, that people are likely to engage in riskier ventures than they would if the money belonged to them, is often apparent in lending programs. There is also the incentive to free-ride on other group members since the borrower has no physical collateral which the lender can take to enforce repayment. A group lending program will not work properly if borrowers behaved like this. Group members have an incentive to monitor their peers, when their behaviour affects the entire group.
Even if the informal advantage mechanism and the peer pressure mechanism exist it is unavoidable that failures to repay will sometimes occur. This can be solved through the informal insurance mechanism. The poor are very vulnerable and can experience unavoidable external impacts which will affect their ability to repay, i.e. droughts, fires or accidents. In
cases like this, the other group members cover the repayment, giving the client time to recover. He/she then repays to the group.
To summarize; group-lending is used as a tool in microfinance because it solves the problems with adverse selection and moral hazard, reduces administration costs for the organisation and by its use of peer pressure also reduces monitoring costs.
2.2.2 Dynamic incentives
Dynamic incentives are often used in combination with group-lending to secure high repayment rates in MFIs. Many programs start by lending small amounts and then increase loan size upon satisfactory repayment. This repeated pattern– and the threat to cut off any future loans if not repaying– can be used to overcome information problems and improve efficiency. (Morduch, 1999b)
Incentives are further increased if the borrower knows that he or she can receive a stream of increasingly larger loans, often called progressive lending. As in the group-lending case, keeping interest rates low are important, since the advantage of microfinance programs lies in their offering services at rates that are more attractive than those of the competitors’, e.g., money lenders (Morduch, 1999b). Another advantage of progressive lending is the ability to test borrowers, by giving them small loans as a start. This makes lenders develop relationships with their borrowers and forecast what their reactions will be before expanding the loans.
Morduch (1999b) claims, however, that dynamic incentives will work better in areas where the mobility is relatively low. Mobility is often lower in rural areas where people more often remain in the same place compared to people in cities. In urban areas it may not be easy to catch defaulters, who simply move to another part of town and start with a clean record at a different bank or program.
To focus on women, as most traditional microfinance organisations do, have some obvious advantages. MFIs will judge the lower mobility of women as an advantage since this will make them less likely to take the money and never pay back. Also, since women have fewer alternatives from whom to borrow dynamic incentives are something positive. Ironically, the
success of many programs with the focus on women may come from the lack of economic assets of women. (Morduch, 1999b)
2.2.3 Regular repayment schemes
One unusual feature of most microfinance credit contracts is that repayments must start nearly immediately after disbursement. In a traditional loan contract, the borrower receives the loan, invests it, and then repays the money in full with interest at the end of the term. But at
Grameen-style banks, terms for a year-long loan are likely to be determined by adding up the principal and interest due in total, dividing the sum by 50, and starting weekly collections a couple of weeks after the disbursement. Some microfinance programs tend to be a little more flexible than Grameen Bank, but even they do not drift far from the idea of collecting regular repayments in small amounts. (Morduch, 1999b)
Initially, the reason for regular repayment schemes was that the management of Grameen Bank feared that the borrower would not want to part with a large sum of repayment at once, and therefore they found weekly collections better. This increases the likelihood of
repayment. (Yunus, 1999:113)
The advantages of regular repayment schemes are several, according to Murdoch (1999b).
First, the regular repayments screen out undisciplined, and thus unwanted, borrowers. Second, because the repayment process begins before investments show results, weekly repayments force the household to have an additional income source on which to rely. That is, with weekly repayments, the bank is lending partly against the household’s steady, diversified income stream and partly against the risky project. This provides advantages for the bank and for diversified households. This means, however, that there are many threats to microfinance institutions in areas focused on highly seasonal activities, like agriculture, which is a common feature of some of the poorest regions of the world.
2.2.4 Collateral substitutes
Few programs require collateral but many of them have substitutes instead. Forming groups is one kind of collateral, when people in that setting are jointly responsible for repayment.
Programs working according to the traditional Grameen Bank model also require for instance that borrowers contribute to an “emergency fund”. The borrowers pay 0.5 percent of every
unit borrowed, beyond a given scale, to this fund (Morduch, 1999b). The fund gives insurance in the cases of for instance death and disability in amount proportional to the length of
membership. An extra 5 percent of the loan is also taken out as a “group tax” that goes into the account. Up to 50 % of the emergency fund can be used by the members of the group if all members agree.
3. Microfinance in Bangladesh
A large number of countries have created microfinance programs with the clear goal of reducing poverty. This is supposed to be done by providing small amounts of credit to the poor, in order to generate self-employment in income-earning activities. Bangladesh is a leader among these countries. Using a group-based approach to lending, the country’s small scale microcredit programs have proven more effective than formal financial institutions.
(Khandker, 1998:3) Bangladesh is one of the poorest countries of the world, which makes microfinance a very important topic. About 50% of the total population in Bangladesh still lives below the poverty line- as measured by income, consumption, and ability to meet basic human needs. (World Bank, 2003b)
Both private and public institutions are involved in microfinancial activities in Bangladesh.
(Hazan Abed, 2000). Around twelve hundred MFIs are currently operating in the country, although the industry is heavily concentrated in a handful of large organisations, namely Grameen Bank, BRAC3, ASA4, and Proshika (Rahman, 2000). These four organisations cover around eleven and a half million people, or 90 % of all clients. (World Bank, 2003a)
Microfinance products tend to be uniform across large geographical areas in Bangladesh.
Most MFIs offer some variety of the product pioneered by Grameen Bank– a loan with a term of about a year, repaid in frequent (usually weekly) instalments, given in a group context, ostensibly for micro-enterprise use, and with a compulsory saving element5. (Matin et al, 2000) The original Grameen Bank focused on the poorest women and used travelling loan
3 Bangladesh Rural Advancement Committee (NGO)
4 Association for Social Advancement (NGO)
5 What in this paper is meant by the ”Grameen model” is the workings of Grameen Bank 1976- approx. 2000.
The Grameen model has recently undergone changes in order to better suit the needs of the clients, different niche markets and life-cycle events. This “Grameen II model” will not be evaluated at all in this paper.
officers who organised rural women into five-person groups already knowing each other but from different households. (Yunus, 1999:14) Eight such groups, i.e. 40 women, met each week in their own village to form a ‘centre’ and loan officers visited them weekly to conduct outdoor banking.
Grameen Bank has tried to maximize operational simplicity. Their way of working is quite simple. (Yunus, 1999:14) Each member contributed with a small weekly saving which could not be withdrawn until membership ended. Within weeks of forming a centre, members began to obtain small loans (typically worth USD 50), with disbursement staggered and continued as long as those already holding a loan were repaying regularly. (Rutherford, 2004) Repayment starts one week after the loan, and amounts of 2 % per week for 50 weeks. (Yunus, 1999:14) The interest rate was 20 percent and it was divided into equal instalments and paid along with principal repayment each week. As soon as they had paid off a loan, members usually took a fresh one, approximately USD 20 larger than the previous one. (Rutherford, 2004)
According to Zaman (2004), the literature agrees that access to these micro loans has
significantly reduced the vulnerability of poor households in Bangladesh. Poor households are able to smooth their consumption more reliably, thereby limiting the hardships arising from seasonal shortfalls of income. Unexpected shocks such as natural disasters can be better absorbed by building up assets. Moreover, female borrowers became less vulnerable and more empowered within their households and the larger community. The availability of microcredit has indirectly affected social conditions, for instance children of borrowers are more likely to attend school, have better sanitation facilities, and better nutrition. These impacts are because of the “increased income effect” of microcredit as well as the “social mobilization effect” of borrower group meetings. (Zaman, 2004)
3.1 Problems with and criticism of traditional MFIs
Grameen Bank and BRAC are the largest MFIs in Bangladesh. Both have received a lot of attention. They have also received a lot of criticism for their neo-liberal developmentalism, their social control, the disciplinary imperatives, and the subsidised system of lending applied.
(Develtere & Huybrechts, 2002)
All researchers do not agree that micro-credit helps in fighting poverty, but most of them are convinced that the vulnerability of the poor is reduced. Caution is needed when referring to the poor as a homogenous group. There are a lot of different kinds and degrees of poverty. A criticism often heard is that MFIs do not reach poor at the bottom of the society. (e.g.
Develtere & Huybrechts, 2002 and MacIsaac, 1997) Both Grameen Bank and BRAC have defined their target group as rural people who have less than an acre of land. However, authors like Zaman (1999) have shown that this criterion is often neglected. 15 to 30 % of the members in BRAC do not belong to the target group. Grameen Bank shows similar figures.
(Morduch, 1998) Some of Bangladesh’s lead MFIs, among them BRAC and ASA, have begun to admit their failure to reach the extremely poor people. (Hickson, 2002)
The membership requirements are sometimes seen as a considerable obstacle in reaching the bottom poor (Evans et al, 1999). To pay registration fees, savings, attending group meetings and participating in different training sessions are often impossible and too time-consuming for the poor people. These people often need all their time to work to survive the day. It is also stated that the group-mechanism does not make things easier for such households. For instance Hulme and Mosley (1996:130) maintain that MFIs often do not want to enrol the bottom poor in their organisations since they are regarded as too risky. MFIs tend to prefer less poor people as clients, since this makes it easier to form a sustainable organisation.
Other obstacles are self-selection or self-exclusion. Evans et al (1999), Morduch (1999a) and Hulme and Mosley (1996) all agree that the poorest of the poor, who do not have a lot of assets, often consider taking a loan too risky. They fear to end up in debt and, thus, do not want to take part in microfinance. Even when they do take a loan, they can end up in
difficulties when the returns from the investment fall short of the cost of borrowing. This gap can be filled by one or more of the following options; borrowing from money-lenders, using savings or selling assets, reducing consumption levels, negotiating to re-schedule loans, or simply defaulting. This means that the poorest often fall through the net. (Develtere &
Huybrechts, 2002)
Some criticism has also focused on the system of group-lending. All mechanisms described in section 2.3.1 may break down. First, if a client group is formed by an outside agent, like a bank, the group will consist of people who are not acquainted. Then they do not have adequate knowledge about each other’s businesses and the informal advantage mechanism
will break down. (Zeller, 1998) Under such circumstances, a borrower may end up in a group that will engage in riskier activities than he or she desires.
Second, the peer pressure mechanism can break down. For a rational actor, it can sometimes be worth the opportunity cost of loosing access to services or to receive a negative reputation to misuse the credits. Dynamic loans can be effective to encourage repayment, but when a client has decided that he or she does not want another loan, this incentive disappears. When other MFIs are available, they may provide more attractive loans. (Morduch, 1999b).
Third, the informal insurance mechanism may also break down. The bank exerts pressure on the group to repay the loans. Clients who want to get a larger loan have an interest in the entire group repaying. But they are rational actors and do not want to cover the costs of their failing group members, since this may limit their own benefits. They can either cover the costs of the failing member or take that individual’s personal assets to cover the cost of repayment. In this case the failing borrower is worse off than before the loan. (Grootaert et al, 2003 and Montgomery, 1996)
Clients with a high mobility, primarily those in urban areas, have a diminished opportunity cost for misusing credits when it is easy for them to leave their residence. (Morduch, 1999b) Moreover, external shocks may cause the borrower to stop the repayments in order to meet immediate needs. (Morduch, 1999a)
Another criticism of the traditional MFIs in Bangladesh is that they are not exactly what the clients demand. Hickson (2002) made semi-structured interviews with 41 slum dwellers in Dhaka and demonstrated that both poorer and richer households were expressing strong dislikes for collective loan responsibility and compulsory meetings. They preferred loans and savings that were flexible with respect to deposits, withdrawals and repayments of loans as well as maturity and frequency of opportunities to make transactions. The majority of the people disliked the standard loan product offered by traditional MFIs (ibid).
On the whole, it seems like a large majority of the bottom poor have not been reached yet by the two dominating MFIs in Bangladesh. (Develtere & Huybrechts, 2002). Hashemi et al (1996) claim this does not mean that Grameen Bank and BRAC are unable to work with all
poor people, but that microfinance is not something for everyone. Poor people who do not have the needed assets, social relations or self-confidence have to be targeted in other ways.
4. A field study of SafeSave
The criticism of traditional MFIs in Bangladesh led to the establishment of new MFIs such as SafeSave in Bangladesh. SafeSave works in a clearly different manner from the traditional MFIs. Their activities are in many aspects more flexible than traditional models for saving.
Also, people can not borrow money until they have a certain amount of money in their savings account. Contrary to many other MFIs SafeSave does not work with groups either.
In order to find out how SafeSave works, I chose Dhaka for a field study during February and March 2005. The questions to be answered were: How do SafeSave handle problems that other MFIs solve by group lending? Do the methods applied by SafeSave work? Do they reach and involve the people not targeted by other organisations? How do their services perform? Can their methods be applied successfully in other places than the slums of Dhaka?
In addition the gender perspective is interesting since SafeSave does not exclusively focus on women.
The method of investigation used has been observation during participation. Traditional methods for field studies are either quantitative or qualitative in their design. Quantitative methods focus on information expressed in numbers whereas qualitative methods focus on non-quantifiable qualities. The method used for this study is more qualitative than
quantitative. The reason for choosing participatory observation as a research method instead of a more quantitative method was that the primary interest of the study was the methods used by the organisation. This was best studied following the organisation’s employees in their work and learning from them for a better understanding.
When a study based on participatory observation is performed it is important to keep in mind that we only get access to behaviour and information that are voluntarily presented in the situations where we are present. It means that we mainly see such behaviour and information that they want us to see (Eneroth, 1984). It was decided that this was only a minor problem
since the study focused on methods. All parties contacted during the study were informed that I was a student and of the reason I was present.
I chose to combine the field study with data and information already available from the organisation. This information was regarded as a reliable source since it is presented in the organisation’s external papers. The available time for the field study made it impossible to personally cover all aspects of the activities of SafeSave, but resulted in an overview of the organisation’s activities. All the information in Chapter 4 origins from SafeSave.
SafeSave was initiated as a research experiment in 1997 by Stuart Rutherford and Rabeya Islam. The former is a microfinance researcher and the latter was a Dhaka housewife with years of experience in running successful savings clubs for poor people. Together they started an ultra-flexible money-management service for poor and very poor slum dwellers in Dhaka.
The purpose of the experiment was to explore sustainable methods for providing individual financial services to slum dwellers in Dhaka6. The experiment was registered as a cooperative under the Cooperative Societies Act of 1984 as cooperative number 199: Shahaj Shanchoy Prokolpa Samity Ltd, or SafeSave Cooperative Ltd. The organisation currently offers four different products in Dhaka (P2, P3, P4, and P6) and another one (P5) in their rural
experiment7. In 2002, the objective of SafeSave was changed from a research experiment to a permanent, self-sustaining MFI. SafeSave reached operational sustainability at an effective loan yield of 37% in 2004. A summary of economic activities from the start in 1997 to the end of 2004 is shown in Table 1. It is evident that all factors are increasing. It is also evident that the amount of deposits as well as withdrawals and loan volumes has increased since the last year as well as in relation to the mean of the five preceding years. SafeSave’s clients were depositing more than one million taka (USD 16,000) per month in the early 20058.
6 In Dhaka city, 25% of the population is slum dwellers. (Rahman, 2000)
7 For a description of the different products, see Appendix.
8 Converted: 1 taka equals 0.016 USD or 0.115 SEK (2005-05-24)
Table 1. Cumulative transactions in millions taka.
Cumulative 1997-2002 2003 2004
Deposits 53.0 30.8 10.1 12.1
Withdrawals 41.3 24.0 7.8 9.5
Loans 108.7 60.3 18.1 30.3
Loan
repayments 89.0 49.6 14.9 24.5
Write-offs 2.1 0.8 0.3 1.0
Source: SafeSave, 2005
SafeSave is a very small MFI and practices outdoor banking in the slums of Dhaka, with an unconventional product that is more flexible than the dominating product in Bangladesh.
SafeSave works exclusively in low-income residential and market areas, and attracts the poor and very poor in those areas by offering products that are designed specifically for them.
According to the management of SafeSave it is this flexible product design and not the targeting of certain individuals which allows the organisation to work with the poor and very poor. Table 2 is included to give an overview of the size and variety of the organisation’s services during one month (December 2004). Table 2 reveals that savings deposits are the dominating service in SafeSave even though loans also exist.
Table 2. Transaction volume (Taka), one month (Dec. 2004), all branches.
Volume % of total Per fieldworker Per client
Deposits 56,891 56.4% 1,073 6.1
Loan repayments 25,693 25.5% 485 2.8
Loan fee payments 16,684 16.5% 315 1.8
Withdrawals 1,239 1.2% 23 0.13
Loan disbursement 345 0.3% 6.5 0.04
All transaction types 100,852 100% 1,903 10.8
Source: SafeSave, 2005
Through the end of 2004, SafeSave had issued 29,860 loans. 88% of these loans are for 6,000 taka or less (the equivalent of USD 100 at the end of 2004). Table 3 is included to give an overview of the averages of different transaction types in SafeSave in December 2004. The average loan amount was for instance 4,089 taka, an amount that may seem small but can make a big difference for people.
Table 3. Transaction averages by type (Taka), December 2004
Average loan amount 4,089
Average withdrawal amount 557 Average loan repayment amount 64 Average service fee payment
amount 38
Average deposit amount 18
Source: SafeSave, 2005
The figures presented in Tables 1 – 3 are intended to underline the fact that SafeSave is indeed involved in micro-finance. The following sections present more closely the structure, method, services and performance of SafeSave.
4.1 Clients and organisational structure
Clients of SafeSave are recruited irrespective of gender, health, occupation, ethnicity, or religion. They are recruited as individuals; there is no group-formation and no cross-
guaranteeing of loans. There are no weekly meetings. The organisation allows both men and women, as well as children, to become members. The age of the client does not matter, as long as he/she is over 15 years old. If the client is under 15 years old, he/she can open savings-only accounts. There is no mass-promotion of SafeSave’s services; all clients are recruited by a collector and enrolled by the branch manager in the branch office. Clients may close their accounts any time they wish, provided that any loans or fees due have been paid.
SafeSave works in seven different slum areas, “branches”, in Dhaka. The branches differ quite a lot from each other. They are everything from permanent housing to illegal occupation of land, to camps for internal refugees, with the common denominator that the people are very poor. The organisation’s head office is situated in Mohammadpur in Western Dhaka, and there is one sub office for each branch. The director and the general manager work at headquarters, and the rest of the staff works in the branch offices in the field. The branch offices will never be contracted in locations where land ownership is unclear, because of the slum clearance risk, even though the areas where clients live often are areas like that. The branch offices are small and very simple, and there are strict regulations on what furniture etc is allowed. The offices do not even have a land line telephone, but will communicate with the headquarters through publicly available telephones. Monthly expenses for rent and utensils for branch offices must not exceed a total of 6,500 taka without permission from the chairman. The organisation does not own any vehicles either. All this is done to minimize costs.
At every branch office a manager, an assistant manager and an office helper works. Each branch has 8-10 field workers, collectors, depending on the size of the area. All collectors are female and slum dwellers themselves. Because Bangladesh is a muslim country, it would not be possible to use male collectors in the organisation, since they would not be allowed to visit all female clients. Before collectors can start their job they receive one week of education.
They must be able to write and count a little, but many of them are almost illiterate. A collector has to make a thorough report to her branch office every day, and the general manager evaluates the collectors each month to see how the work is going, how many clients they have and so on.
Each collector is responsible for 150-200 clients. The clients are visited at least once a day, six days a week, in their home or workplace. Because the areas are densely populated, visiting so many clients in one day is possible. Savings deposits, loan repayments, interest payments, and small withdrawals will be handled during the field visits. Loan disbursement and large withdrawals will be handled in the branch office. All financial transactions between SafeSave and the clients take place during daily visits and clients do not have to visit a branch office, nor attend meetings. At each daily visit clients may save as much as they like, or nothing at all, or withdraw as much as they like, up to the limit of their savings balance. It is very important for the collector to always ensure that clients fully understand the product rules, to deliver the services according to the product rules in a professional and courteous way and that they encourage clients to save, repay, or pay fees.
New branch locations are proposed by the director for approval by the executive committee.
The general manager is responsible for establishing each collector’s working area within the radius of 0.5 kilometres or 5 minutes walking distance of the branch office. Care will be taken to ensure that each collector has an opportunity to develop a portfolio of 200 active clients, without pressure to open multiple accounts in the same household. The size of the working areas will vary, as density is much higher than in established areas. In better off areas, the working area will need to be much larger, as the density will be lower and fewer households will be interested in the services of SafeSave. Collectors’ working areas are arranged as efficiently as possible, and the collectors should not overlap into each other’s area. A client can become a member of SafeSave if he or she has been living in the slum area for one year.
There are two reasons for this restriction. It is more probable that a client will stay in the area and he or she understands how the area works after one year there.
4.2 Savings
In all of SafeSave’s products, there is a simple savings account in which clients may save as much as they like, when they like, at daily visits. If they do not hold a loan, they may
withdraw savings on demand. SafeSave’s products, described in the appendix, differ from each other in some aspects. In the most basic product, P2, interest of savings is earned at the rate of 6% per year, but only in months when the balance does not fall below 1,000 taka.
Clients choose when and how much to deposit. Requests for withdrawals of up to 500 taka will be handled by the collector when she visits. This is simply done by signing the collectors’
collection sheet with name/signature or thumbprint. Collectors in some branches use hand held computers, which is a way to increase the internal control in the organisation9. Larger withdrawals, more than 500 taka, are handled in the branch office within ten minutes. The collector must then accompany the client to the branch office where the manager or assistant manager fulfils the requests. The client must sign a withdrawal slip with his or her name or thumbprint and the collector must sign the manager’s cash disbursement voucher. There are no charges for withdrawals of savings, however rates on savings may favour clients who withdraw infrequently or save up large amounts.
If product rules have been met, requests for up to 5,000 taka made during branch hours must be fulfilled within ten minutes. Amounts larger than 5,000 taka will be paid immediately if money is available, but must in all cases be fulfilled within one working day of the request.
Withdrawals should always be processed as quickly as possible, and clients requesting withdrawals have first priority and clients applying for loans have to wait.
4.3 Loans
Even though savings are the main focus of SafeSave, clients may also take a loan if they want to. The loans are provided for any purpose. Loan purpose is not considered as part of the approval process.
9 The PalmPilot project is used as an experiment in some of the branches. When used right it is a way to improve the safety in the organisation and it is also simpler. The collector has a personal code and every client has their own code. The collector enters the codes in the PalmPilot and then she enters the saldo according to the client’s passbook. On her Palm Pilot the collector can see the symbols for savings, withdrawals and so on. She presses the right symbol and enters the sum. When the Palm Pilot beeps three times the client knows that the transaction is correct and registered. Even if the collector has only basic knowledge in reading and mathematics it is easy for them to use the technique. It is also very difficult for them to be able to/tempted to embezzle money.
Clients choose when and how much to borrow10. They are guaranteed loans when they meet product requirements, staff are not authorized to make subjective loan decisions. Loans are guaranteed within two working days of a request. There are no requirements for physical collateral or third party guarantors. Instead the savings the client already has works as a kind of collateral for the loans. The maximum permitted value of the loan is determined by the balance of the savings account and the clients’ previous history with SafeSave. Clients may borrow an amount that is equal to their savings balance plus a sum that increases with each loan. In SafeSave’s basic product, P2, the first loan may be for a maximum of two times the savings balance, up to a fixed maximum. Subsequent loans may increase in value by a set figure.
All loans are issued personally by the branch manager, and must be issued in the branch office to the account holder, with the collector present. Loans may only be issued directly to the account holder. If the account holder intends to lend the money to another person, that is his/her choice, but SafeSave considers the account holder to be responsible for repayment.
Loans do not have to be repaid according to a fixed schedule and there is no fixed maturity date. In all SafeSave’s products (except rural P5), clients may pay back loans with whatever amount they wish and when they like, during the daily visit by the collector. If clients choose to repay in a lump sum, they may do so by cutting the sum from their savings balance.
The clients do not have to repay according to a fixed schedule. They must, however, pay loan interest (“fee”), each month. The fee is calculated on the declining month-end balance, at 3%
per month, and is due within a specific period (3 months). In P6 however, interest is paid quarterly (and in P5 all interest is paid at loan disbursement for a fixed term). Other aspects of loan vary with the product. As long as loan interest is paid, repayment of loan principal is up to the client; there are no set schedules or terms for SafeSave loans. However, most get paid within 8-9 months. All loans are forgiven on the verified death of the client.
Clients have the right to take loans if they meet the requirements in the product rules. The collector and branch manager do not have the right to deny a loan on the basis of any other criterion than the published requirements. Valid reasons for denying a loan are, for instance,
10 Often the loans are very small. 90% of loans in SafeSave are for less than USD 100 and some are as small as USD 10; of 28 000 loans issued so far, the biggest ever was for USD 600.
that someone other than the client is asking to take the loan on account, or savings
requirements are not met or principal or fees are due from previous loans. A rumour that the client will be leaving the area, the client is a ‘difficult client’ and that the client seems risky because they are too poor cannot be used to deny loans.
4.4 The performance of SafeSave
2004 was SafeSave’s first profitable year. Operational sustainability11 was 102% and returns on average equity was 1,5%. The organisation is then not dependent on external donors to exist. The numbers of members is increasing and the number of total clients was 9,737 in April 2005. Most of the clients are women.
The branches of SafeSave generally reach profitability within 18 months. SafeSave charges 36% effective interest for its loans. While that is competitive on a global level, most MFIs in Bangladesh charge 27-32% effective rate. According to SafeSave, to reduce the loan interest rate would require a restructuring of their operations; increase the average loan size per client, average loan portfolio per collector, and average loan portfolio per branch. Demand for SafeSave’s loans is strong, even though the interest rate is 36%. The management maintains that changes required to increase lending volume to reduce the interest rate might increase losses, diminish customer service, and lead SafeSave away from its focus on the very poor when they would not be able to work in the same cost minimizing way.
Cost-benefit analysis appears favourable, and SafeSave’s management has begun to pay for the expansion of the system out of its own resources, instead as with donor funds as for the first two branches. Table 4 is a performance summary of SafeSave, showing a steady increase in number of clients as well as the number of clients who are also borrowers. The percentage of the clients who are borrowers is almost constant. As the organisation matures the number of branches show a slow but steady increase and the operational sustainability increases in a similar manner. Other indicators in Table 4 show the same pattern.
11 (Income/Expenses) before adjustments for inflation and subsidies
Table 4. Performance summary.
2003 2004 2005 (forecasted)
Clients 7,326 9,443 11,500
Borrowers 4,670 5,950 7,200
Branches 6 7 8
Total staff 70 76 98
Clients per total staff 105 124 117
Loan portfolio 13.9m 19.7m 23.0m
Savings portfolio 9.1m 11.7m 15.0m
Average deposit balance 1,242 1,242 1,304
Average loan balance 2,980 3,319 3,194
Average loan at
disbursement 3,404 3,641 3,605
Calendar year 2003 Calendar year 2004
Cal. year 2005 (forecast)
Operational sustainabilty 97% 102% 108%
Return on average equity 2.4% 1.5% 6%
Growth in total clients 11% 29% 22%
Growth in deposit portfolio 36% 30% 35%
Growth in total assets 14% 27% 24%
Growth of loan portfolio 30% 42% 17%
Revenue growth 15% 33% 34%
m= million
Source: SafeSave, 2005
5. Comparison of SafeSave and traditional MFIs
Table 5 summarizes the main differences I found between the traditional MFIs in Bangladesh and SafeSave.
Table 5. Main differences between traditional MFIs in Bangladesh and SafeSave
Organization Traditional MFI SafeSave
Size Big Small
Clients Women All
Setting Rural area Urban area
Main service Loans Savings
Product type Standardized Flexible
Loan requirements Use the loan to start No specific purpose
or expand small-
scale
businesses
Main mechanism Groups Individuals
Services Weekly Daily
Repayment policies Regular No set schedules
One of the major differences between SafeSave and traditional MFIs is that SafeSave does not use group loans. Traditional MFIs use the method of joint liability through lending to groups mainly because it minimises adverse selection, it minimises monitoring costs and
administration costs. The problems with adverse selection and moral hazard are also very important factors.
The management of SafeSave have chosen not to work with group-lending, since they believe that groups are of limited benefit and that group meetings are too time consuming for the clients. According to the philosophy behind SafeSave, groups often act as a barrier of entry of poor households. Either the poor are not courageous enough to participate in a group lending scheme because they are afraid to fail, or people who are already members of a group may deny access to those who are poorest and are believe to be too risky. There is thus no
requirement for group meetings in SafeSave. Borrowers I discussed with in Dhaka, found the individual services one of the best things with SafeSave. There are many other MFIs working in the areas where SafeSave works, that use the group mechanism, but this does not appeal to the clients of SafeSave. The main explanation the people I spoke to gave, was that they were afraid that they would have to cover the costs of failing group members. Another reason was that they did not want other people to know what they were using the money borrowed for, and involve too much in their personal decisions.
SafeSave finds working with individuals a more suitable way to provide financial services for poor people. The entire organisation is based on savings. SafeSave limits their risks in two main ways. First, a client must have a good history with the organisation before being able to get a loan, second the loan is linked to the savings. The client must build up a savings account before he/she is eligible for taking out a loan. In SafeSave clients use their own savings as kind of collateral for the loan. This implies that the organisation has some knowledge about the client before giving the loan, knows that he/she is trustworthy. When using this method it is thus not necessary with a group to cross-guarantee the loan. The asymmetrical information is minimised because the organisation already knows that the borrower is trustworthy, when he/she already has a well-working savings account. Instead of the mechanism where a client is chosen as trustworthy by his peers, the client shows he is trustworthy by being able to save.
The client’s value the services they have access to from SafeSave so highly that they do not want to loose them access to them.
Another reason for traditional MFIs to use groups is because it is difficult and costly to monitor a large number of individuals. SafeSave has solved this problem by collectors, recruited among the slums dwellers. These women have very few years of schooling, but they value their job highly and they have good knowledge about the people and the areas they live in. The travel cost of the collectors is zero and their salaries are very cheap for the
organisation. However, the high responsibility of the collector can be a problem. It is for instance very important how they explain product rules and regulations to the clients. The result depends very much on the capabilities of individual collectors. The collectors are, however, after what clients told me, highly trustworthy women, regarded as friends to the families that they work with.
Traditional MFIs can also use groups to lower administration costs. Giving a loan to a group is cheaper. SafeSave acts strictly cost minimizing for the entire organisation. With strict cost control and appropriate product pricing it has quite low administration costs. The interest is a bit higher than other organisations, but it is accepted by the clients. The organisation is entirely sustainable from year 2004, that is, the organisation’s money comes from the clients and they do not need any external money.
Another way in which SafeSave differs from traditional MFIs is in handling repayments. All traditional MFIs use regular repayments and the client starts to repay almost immediately.
This is used to screen out undisciplined borrowers and it is also a way to make them have an additional income source to be able to pay back. In contrast, the only demand SafeSave has when it comes to repayment of loans is that interest, 3% on declining month end balance, must be paid monthly. This increases the incentive to repay quickly. Even though clients do not have to pay back their loan in a certain time, most people in SafeSave repay their loans in 8-9 months. It seems that the fear of loosing access to the financial services will be enough to ensure repayment, even without deadlines. A client with an outstanding loan in SafeSave gets reminded of it every day when the collector visits him/her. All households know that they are economically vulnerable and have incentives to build up savings or/and have access to credit, which they have when they repay their loan. The loan must be fully repaid before the client is allowed to take another loan. This was not regarded as a problem for those borrowers I had the opportunity to talk to in Dhaka.
