Every day in Dhaka, Bangladesh, bankers with SafeSave visit their small-scale savers. They can make deposits, get a small loan or draw on their savings there and then without paperwork.

As a result, they do save. SafeSave's 5 228 active clients had an average of just over US$21 on deposit at the end of 2001. A conventional bank would not want sums that small. Meanwhile, the bank had outstanding loans of nearly US$175 000 in amounts as small as US$17. The value of loans written off last year was 1.55 percent or less of the total. Increasingly, this integration of loans with savings is seen as a blueprint for lending to the poor.

"If they can save, people who are just above the breadline may be able to build a small cushion against hard times," says FAO economist Sumiter Singh Broca. "This would make them more likely to invest in seeds of a new crop variety, or better tools or bicycles, because they would no longer have to fear becoming destitute if those investments should fail. Meanwhile, their deposits provide a line of credit for other people like them. So even if they don't take a loan, they are more likely to invest their existing funds -- because a safety net is available."

SafeSave, the better-known Grameen Bank (also in Bangladesh), and numerous other formal and informal savings and loan organizations have put this into practice to great effect. But they also address other key problems. For example, it is uneconomic to incur US$20 in administrative costs to make a loan of just $17. But these costs, meant to protect the lender from default, may not be necessary. Community-based lenders know who is creditworthy and who isn't -- so they don't need to demand collateral, avoiding a lot of paperwork.

Another way to avoid asking for collateral is group lending. If everyone loses when one defaults, the group will deal with the defaulter. For example, agricultural decollectivization in Kyrgyzstan after independence in 1991 left farmers with as little as one hectare each and no assets to use as collateral for investment. The European Commission and other donors funded a group-lending scheme to address this problem. In one instance when a farmer defaulted, the others had him put in jail. Group lending can be a good way for farmers to invest in marketing facilities -- or new technology (see box, "Farmers' finance for better bananas").

Lenders can eliminate even more paperwork if they don't have to ask what the loan is for. This is another area in which old assumptions about credit are now being challenged.

New thinking

"The way the development community looks at microcredit is changing," says Mr Broca. "In the past they assumed that loans should only be for productive investment, not consumption. But enforcing that policy has made micro-lending uneconomic."

It is pointless anyway, he says. First, research suggests loans for consumption are just as likely to be repaid. Second, loans for consumption are also an investment. To someone in a rich country, the words 'consumption' and 'consumer goods' indicate luxuries such as television sets. But in a developing country, they could simply mean purchasing food. And this is an investment in itself. As the Nobel Prize-winning economist Robert Fogel has pointed out, hungry people cannot work their way out of poverty. In England and France around 1790, he suggests, the bottom 20 percent of the population was effectively excluded from the labour force because they were too hungry to work. He thinks better nutrition accounted for about half of Britain's economic growth between 1790 and 1980. Says Mr Broca, "A loan for consumption may be an investment in the body -- perhaps not so different from buying a machine or a bicycle."

So, first, savings must be integrated with loans to create a single financial system that helps the poor. Second, controls on the use of small-scale loans should be eliminated. And third, lending bodies must be community-based so that they don't waste money on paperwork. That is what organizations like SafeSave have done.

Nothing new...

But are they reinventing the wheel?

Informal credit facilities have always existed. According to a 1998 report by the International Food Policy Research Institute, Rural finance and poverty alleviation, about 72 percent of the poor in Nepal engage in financial transactions. Borrowing during bad times or the planting season, and saving or repaying loans in good times or after the harvest, are integral to poor people's lives, said the report.

Credit sources include friends and family, village moneylenders and suppliers of seeds or fertilizer who make loans 'in kind' in return for part of the harvest. The informal sector also includes rotating savings and credit associations (ROSCAs). Each member contributes to a monthly lottery and the winner invests. Immigrant communities in the United States use ROSCAs to finance capital assets such as grocery stores and taxis. Generally based in a neighbourhood or ethnic group, ROSCAs are self-policing.

So are formal structures like the Grameen Bank or donor-financed schemes necessary?

Since local, informal arrangements don't have much capital to lend, they can't always finance costly investments such as irrigation or purchase of land and must usually confine their lending to a few borrowers in the local area. This increases the risk of most of their clients defaulting at the same time because a disaster like a flood or drought has hit that area. To allow for this risk, lenders have to charge high rates on loans, including those for necessities such as seeds, holding borrowers in a poverty trap. Also, in agricultural communities a disaster like a flood or drought usually hits lenders as well as borrowers, making credit scarce just when it is needed most.

The poor will always need informal credit arrangements. But they are insufficient on their own -- the poor need integrated savings and loans systems that can help them both escape immediate hunger and finance their futures. Moreover, institutions such as the Grameen Bank sometimes offer training or basic literacy programmes that can make the loans more useful to the borrower.

But such institutions have often succeeded only because they rejected old assumptions about lending money.

"Think of substitutes for collateral," says Mr Broca. "They can be found. And help people to save. Then they'll have the confidence to borrow. And, perhaps most important, don't be afraid to lend for consumption. Microcredit should be seen as both a safety net and an investment in economic growth -- because they reinforce each other."