Ideas to Action:

Independent research for global prosperity


David Roodman's Microfinance Open Book Blog


I first connected with Hugh Allen when he commented on my post about the WSJ "bubble" story. Visiting the site of his company, VSL Associates, I realized I needed to talk to him in order to learn more about the Village Savings and Loan Association (VSLA) model of microfinance. My first exposure to these ideas came last November, when Jeffrey Ashe of Oxfam America visited me to tell me about Oxfam's community finance program. VSLA seems to be about helping people start the kinds of informal financial groups they tend to do anyway all over the world. The potential to form savings groups is ambient. Trainers from the outside bring a particular, honed model and serve as the nucleus around which a crystal grows.

Hugh and I spoke Wednesday, he I think from Germany. Since I am blogging in order to share the learning experience of writing my book, I am posting the notes from the conversation, which Hugh has edited. This is a first for me, so let me know what you think of this experiment.


In 1991, CARE funded a small project in Niger to help women doing handicrafts for income. A Norwegian woman named Moira Eknes ran the project. She fairly quickly realized that the women knew more about handicrafts than she did. What they lacked was capital. So Eknes learned about the various indigenous, informal forms of community finance in the region, such as tontines/ROSCAs and arrived at the idea of a group that would accumulate savings in a box with three locks (presumably three different people held the keys, to prevent theft) and then lend out money to the same members. In westerners' technical terms this made it a kind of ASCA.

One huge difference from conventional microfinance is that the supporting organization has no permanent presence. Training and guidance is provided for perhaps a year to get the group going. Then the trainers leave. It is essentially informal finance. He agrees with my characterization of VSL as a subtle elaboration on extant ASCA models. [In fact English friendly societies, which go back to the 1600s, sometimes used lock boxes.]

Slowly the model spread. He was at CARE then. He was as skeptical as everyone else at the organization. You put money in a box, lend it back out to the same people, and charge interest??

But it became clear it was working. 20,000 people had joined these village savings and loan (VSL) or "community-managed microfinance" groups by 1995.

He took quite an interest in it around that time. Over several years, in Zimbabwe, Mali, and Zanzibar, CARE deployed the model. It worked everywhere. He quit CARE in 1998 because mainstream microfinance, as distinct from VSLAs, was leading with the wrong product (credit) and not reaching the truly poor. He focused exclusively on spreading the VSL model. Microcredit is based on several assumptions, all highly questionable: debt=investment=profit=net extra income. By about 2000, CARE and other groups had pushed VSL out to about 15 countries (now 24). PACT took it to Nepal around 2000.

VSLAs have spread to most major international development organizations. Plan International is doing it—the only major microfinance form Plan does in Africa. CARE, CRS, World Vision. The model has taken on a life of its own. If you include what’s been done in India it's reaching ~2.6 million people. However, he prefers not to count the 1 million in India because they are essentially in self-help groups, which are distinctively India form, different in spirit. They open formal bank accounts and need ongoing technical assistance to maintain complex written records, usually in order to comply with the bank's procedures to access group loans.

VSLAs are running in 28 countries in Africa, five in Asia (Afghanistan, Cambodia, India, Indonesia, Nepal), two in Latin America (Ecuador, El Salvador). And his data are for CRS, CARE, Oxfam, Plan, and World Vision only. Lots of small organizations are doing it too, and are not counted. CRS reports 600,000 members, CARE 1.6 million, Oxfam 260,000, Plan 150,000. World Vision 20,000.

CARE gives credit to Eknes, but the model has evolved significantly since the early days and all of the organisations cited have developed brand identities, which they protect and promulgate. Yet the differences between them are more those of style than substance. By 2005 the practitioner organisations had nailed down the principles and methods, which vary across organizations approximately as follows: Oxfam in Africa does it with nonliterate people. Transactions, debts, and credits are all memorized: nothing is written down. At the other end of the spectrum, CRS's groups write all transactions in ledgers, which in Allen's view is too complex for most of the poor. In the middle, CARE and Plan use passbooks, which he favors: they blend some formal record-keeping with a process of witness familiar to oral cultures and reduce training and supervision time.

Oxfam works for viral replication: members of a group are trained from the start to promote the model informally. But Plan and CARE promote the user of Village Agents, who are formally trained to create new groups, receiving fees for this service from the groups themselves. In one of the best documented examples, CARE worked for 3 years in Zanzibar, creating 43 groups. They left behind a group of community-based Village Agents who created new groups and, as of December 2008, there were reported to be 250 VSLAs.

The groups do not work much above 30 people. Beyond that, the meetings become too long and the transactions become too numerous to remember, so they become opaque. With opacity comes the opportunity for theft, distrust, or error.

If VSLAs are not so different from other forms of informal finance, what's so special about the effort to promote them?

1. Spread the idea of a model that works. After someone invented the wheel, that idea had to be spread too.

2. They have developed particular, highly structured approaches to training and supervision for the first year. The essential details are ones of process. Meetings run certain ways in order to ingrain certain habits and routines, and eventually reach autopilot. The groups are built on written/memorized constitutions and run by elected officials. They develop policies for managing the savings fund, granting credit, governance, etc., which are all part of the constitution. Follow-up studies show that more than 94% of these groups survive more than 5 years on their own and there is little tendency to elite capture or inequitable distribution of benefits.

During periods in which the demand for individual credit is low, the groups often make opportunistic short-term investments, such as storing grain for a couple of months, till a time when it is more needed.

The VSL industry is splitting into several camps. He says: let’s just spread this simple model as far as we can. Others want to improve, elaborate on, the model, e.g., to link to banks to get additional loan capital as in India, to federate VSLAs into larger organizations that can support more sophisticated financial services. He opposes. Historically, refining this reliable ASCA model was a feat in itself. Isn't that achievement enough?

Credit unions are much larger---say, 250 people. For that size, they have to be formal. Have to have literate, numerate membership. Have to build in financial controls. Need physical premises. But this dramatically raises the costs and complexity of providing services, which reduces the number of basic "entry-level" VSLAs that can be created. To some extent, VSLAs go back to the 19th century roots of credit unions.

NYT had a photo of a lock box belonging to a savings group made up of German immingrants in a bar around 100 years ago.

Contrast the simple technology of the 3-lock box with cell phones. People are mesmerized by the latter. And this can lead to safer and more flexible ways of making deposits and loans. But cell phones or no, 60--70% of people will not be served by formal institutions until they are better off.

Most VSLAs are rural, but they have also worked extremely well in urban areas, such as in the Kibera slum outside Nairobi, as well as in Accra and Luanda.

He agreed with my interpretation that when the value of people’s time is low, it makes more sense to do it yourself, through informal finance, as opposed to paying someone else to manage your money, as in formal microfinance. [I remembered Jeffrey Ashe describing microcredit to me from the point of view of poor people in rural Niger: "that is for rich people." The poorest cannot afford to pay, through interest, for the wages, computers, and buildings of microcreditors.] If people are poor enough, they don’t mind walking a mile and sitting under a tree for an hour to save $1.

Also, the social experience of microcredit and VSL groups is important. He's skeptical of many claims for microcredit, but he's been impressed over the years with the empowerment that comes out belonging to these financial groups.

Average return on people's investments: 40%/year. Average family of VSL participant has generated slightly less than one new income-generating activity, eats better and more regularly; sends more children to school, has improved access to medical services and improves its social status. Setting up a VSL costs $50/head of donor funds at most. Compare that to cost to set up a sustainable African MFI, at $200--350 per participant.


CGD blog posts reflect the views of the authors, drawing on prior research and experience in their areas of expertise. CGD is a nonpartisan, independent organization and does not take institutional positions.