Small is smart

Written by Maitreesh Ghatak | Updated: Aug 25 2009, 03:19am hrs
Microfinance is big these days. It may be called micro but there is nothing small-time about it. It serves more than 100 million poor people (mostly women) worldwide. About $25 billion (roughly the size of the US current account trade deficit) is lent out worldwide through various microfinance programmes (MFPs). While its true origins are lost in the mists of time, Muhammad Yunus of Bangladesh is viewed as the leader of the movement for singlehandedly creating the most famous and successful microfinance institution (MFI) of the modern era, the Grameen Bank of Bangladesh. In 2006, Yunus and the Grameen Bank won the Nobel Peace Prize for reducing world poverty.

Why is microfinance needed Why cannot standard finance (i.e. banks) do the job Because of transactions costs (screening and monitoring borrowers, ensuring repayment) credit markets are imperfect. After all, credit is an odd transaction: one party gets money and the other party simply gets a promise (to repay).

A standard solution is to use collateral, which makes the promise to repay credible. But there are several problems. First, a large fraction of the population in developing countries is poor and does not own any assets. This means they are shut off from the credit market, so they are unable to improve their economic condition and accumulate assets: a classic poverty trap. Second, even if someone owns assets he or she does not necessarily have a formal title to it. Third, even if someone has an asset and a formal title to it that can in principle be used as collateral, foreclosing on collateral is costly because of inefficient judicial systems.

Thus, there is an obvious case not to rely on the operation of market forces. However, the evidence on government lending programmes targeted towards the poor is not very encouraging. Low repayment rates, mistargeting of loans, corruption and leakage are common. This is not surprising. The reasons why markets do not work well apply to government organisations too, and there are the usual problems of bureaucracy and political interference. In this gloomy scenario, the Grameens record is quite striking. Only 5% of their borrowers were outside the target group, and even according to conservative estimates, its repayment rate is 92%.

Starting in the mid-seventies, the Grameen now lends to about 8 million people, most of whom are rural, landless women, and operates in 84,000 villages covering most of the country. It gives small loans for self-employment projects (poultry, weaving, grocery, tea shops). No collateral is charged and interest rates are comparable to government programmes.

The key innovative feature of Grameen is that borrowers are asked to organise themselves into self-selected groups of 5 people from the same village. Earlier Grameen used explicit group or joint liability: loans were given for individual project, but the group was jointly liable for each others loans. Now it has shifted away from explicit joint liability. However, in their own words, Repayment responsibility rests solely on the individual borrower, while the group and the centre oversee that everyone behaves in a responsible way, and none gets into repayment problems. This is consistent with implicit joint liability. If any member of a group defaults, other members fear they might become ineligible for credit in the future even if the lending contract does not specify this punishment. One form in which this can happen is if the branch of the microfinance organisation itself chooses to fold its operations when faced with delinquency.

Some economists see the group aspect of microfinance as a key to its success. Members of a community know more about one another than an outside institution such as a bank.

While a bank cannot apply financial or non-financial sanctions against poor people who default on a loan, their neighbours may be able to impose powerful non-financial (e.g. social) sanctions at low cost.

Therefore, an institution that gives poor people the proper incentives to use information on their neighbours and to apply non-financial sanctions to delinquent borrowers can outperform a conventional bank. These programmes might be inducing group members to select their peers carefully, monitor each other, and put pressure on delinquent group members. Others have argued that even without joint liability, the threat of losing access to future loans is the mechanism that explains the success of MFPs. The problem with this argument is that conventional banks can also use these dynamic incentives, and so the question is why arent they doing it Also, whether or not explicit joint liability is used, almost all MFPs are group-based and so its reasonable to infer that groups must be performing some role.

Microfinance represents a big watershed in how economists think about anti-poverty programmes. It is decentralised, largely independent of the government, and the underlying philosophy is to view the poor as potential entrepreneurs as opposed to passive recipients of government subsidies. While all this sounds good in theory, the key question is, has microfinance been successful in doing what it aims to do, namely, relaxing credit constraints, and improving the standard of living of its clients

Some studies have compared the performance of MFPs with conventional lending programmes and found that the latter in general have better outcomes. However, this runs into the standard problems of selection bias and endogenous placement that beset empirical evaluation of any programme or policy. In particular, borrowers who are more dynamic and reliable may be more likely to join a MFP, and as a result, the impact of such a programme on them is not representative of what its impact would be on an average member of the target group. Similarly, if MFPs choose to operate in poorer and backward villages (i.e. where the need is the greatest) then again, the outcomes in these areas will underestimate its potential impact in an average village.

A number of recent studies avoid these problems. They follow the method of randomised control trials that are increasingly being used in economics. When feasible, they have the great virtue of being both simple and powerful. They follow experimental trials in medicine: you select two groups that are similar and then randomly select one to receive the treatment (a drug, or a policy) and then compare the outcome of this group (treatment group) with the other group (control group). If the difference is statistically significant, then that is attributed to the treatment.

For example, MIT researchers Abhijit Banerjee et al* have studied the impact of access to microfinance on the creation and the profitability of small business as well as various measures of standard of living. Working with Spandana, a MFI, they randomly selected 52 of 104 slums of Hyderabad for opening one of their branches, while the other 52, who were eligible and similar to the former group, were not selected. About a year and a half later they conducted a detailed household survey of an average of 65 households in each of the slums, with a total of 6850 households.

In this study, the control and treatment slums look pretty similar before the programme was carried out in terms of population, average debt outstanding, businesses per capita, per capita expenditure, and literacy (Table 1). What about the effect of the programme The study finds that the treatment groups were significantly more likely to start a new business, and existing business owners experienced a significant increase in profits, but employment did not increase significantly (Table 2). However, other than expenditure on durables, including durables used in business, the programme didnt increase per capital expenditure significantly (Table 3). The likely explanation for this is that for the effects of increased investment to show up in increased consumption, a longer time horizon is required. Also, the study shows that households change the pattern of consumption and cut down on non-essential consumption to augment loan money and finance their new business ventureanother reason why total consumption doesnt increase immediately. The study also finds that there was no significant impact of the programme on education and health. These effects are unlikely to kick in immediately.

The evidence suggests that microfinance is no magic bullet to solve the problem of poverty. But perhaps the fault lies not with microfinance (or, for that matter, any anti-poverty programme) but our impulse to look for magic bullets that will solve the problem of poverty overnight. After all, microfinance is supposed to relax credit constraints faced by the poor who are shut off from formal credit markets and allow them to create and expand businesses through investment. The evidence suggests that it is successful in doing that. That sounds like a promising start to me.

*Abhijit Banerjee, Esther Duflo, Rachel Glennerster, and Cynthia Kinnan (2009): The Miracle of Microfinance Evidence from a Randomised Evaluation (Available at

The author is professor of economics, London School of Economics.