CHAPTER 4

Getting Started with Saving for Change

With the project in Zimbabwe in tatters, my colleagues and I began scouting for other places where introducing Saving for Change would make sense. Several countries in West Africa met our criteria: those with a large portion of the population living in rural areas, high rates of rural poverty, and a tradition of using groups to manage their finances. I reached out to the Oxfam America West Africa regional office in Dakar, Senegal, to develop a savings group project there. Mamadou Biteye, the senior program officer, agreed to take the project on. He later told me it was against his better judgment, but the savings concept was so counterintuitive that it intrigued him. “I was not sure how [saving] could have such a positive impact on people,” he stated dryly. “I have always thought that poor people probably need more money.”¹ I flew out to meet him in Senegal.

We were offered funding from the Norway-based Strømme Foundation to launch Oxfam’s savings group program in Mali.² Mali certainly met our criteria. At that time, in 2005, 70 percent of the population lived in rural areas and 80 percent of the labor force earned their living through agriculture and fishing.³ We later found that 82 percent of the population where we would eventually work lived below the international poverty line of $1.25 per day.⁴ Malaria was endemic. The population had tripled over the last fifty years, forcing farmers to cultivate the same land year after year, reducing crop yields.⁵ An increasingly unstable climate further threatened food security—some villages experienced droughts one year and floods the next.

The Cultural Context

From the start, we made the strategic decision to focus on women. At that time, the UN Human Development Index ranked Mali fifth from last of all countries in its Gender Inequality Index.⁶ More recent data indicates that in Mali, the majority of people marry in their teens and 46 percent of girls have given birth by age eighteen.⁷ The average woman gives birth to six children over her lifespan, and one in twenty-two women dies in childbirth.⁸ Less than one-third of women can read and write (less than that in rural areas), compared with half of Malian men.⁹ Women have little access to political power: in 2013, women made up just 10 percent of Mali’s parliament.¹⁰

At the household level, social and economic norms divide family members’ roles by gender.¹¹ Men farm the bulk of the land, although many women are allocated plots on which to grow their own crops. Depending on ethnicity and location, men may also earn income through animal husbandry, tailoring, butchery, fishing, or laboring for others. Increasingly, younger men may migrate outside their village for work, returning during the planting and harvesting season when their labor is required. In rural Mali, it is common for families to live in extended households headed by a patriarch with several of his younger brothers or sons, perhaps each with multiple wives and their children. Though it is a cultural expectation that no man will marry an additional wife unless he can provide equally for all his wives and children, in reality it is common for different wives and their children to experience very different economic and social circumstances. Older women may find they hold power over younger wives; conversely, they may be neglected if their husbands turn more of their attention (and income) toward the new brides. In female-headed households, women must take on greater responsibilities and may face greater hardship as they navigate single parenthood amid cultural norms that often bar women from property ownership and equal inheritance and limit the earning opportunities available to them.

As a rule, men spread at least some of their harvest across the household as a whole, providing staple foods such as millet, corn, or rice, which make up the bulk of the typical Malian diet. Women are tasked with managing their households. They cook, pound millet, weed their husbands’ fields, carry water from an often distant well or stream, do the laundry, shop in the market, care for infants, and educate young children. To finance their share of the household expenses, many women harvest wood for charcoal and shea nuts to refine into shea butter (both laborious processes), cultivate produce in a garden or fields, or sell produce or prepared food in the local markets. In general, women provide the “sauce” ingredients that flavor the staple grain—growing or buying tomatoes, onions, okra, and perhaps fish or a little meat. Women generally also take care of their children’s clothing, school supplies, healthcare, and other everyday expenses, often paying costs themselves. Women might take turns preparing meals for their extended family or for their husband, if he has more than one wife.

Many women manage their household finances through tontines, where they can save for larger purchases. They may also keep money at home or, if they are a bit more prosperous, own a few chickens or goats to sell if need be. When these reserves run out, some take loans at high rates from moneylenders, while others report begging or going hungry.¹² Regardless, most rural Malian women, as well as men, lack access to formal financial services. Delivery costs are too high and credit needs are too small to make serving these Malians profitable or sustainable.¹³ Mamadou and I felt this was the group with which we wanted to work: financially experienced, independent, and savvy women who faced great challenges and who could probably benefit most from better tools to manage their finances.

The Feasibility Study

Our first step was to make sure the needs we had identified resonated with the women we wanted to reach. Mamadou and I traveled to Mali to conduct a feasibility study in mid-2004. Before launching our project, we needed to understand how our idea, crafted from studying similar projects elsewhere, would be adapted to fit the Malian context. We needed to understand the interests and priorities of the women with whom we would be working. How did they manage their money? Did they have access to sufficient credit? What type of income-generating activities were they engaged in? After learning all we could before we got there, Mamadou and I arrived in Bamako, Mali’s capital. Before we traveled to Mali, Mamadou Biteye, along with Edouine François and Boubacar Diallo from Freedom from Hunger, had traveled to Niger to visit CARE’s impressive savings group program, Mata Masu Dubara, which translated from Hausa means “Women on the Move.” This provided Mamadou, Edouine, and Boubacar an intensive introduction to how these groups work.

We set out with local Strømme staff to a nearby village with a permanent market. With its proximity to Bamako and its central square overflowing with market women displaying goods piled on brightly colored cloths spread out on the ground, this village was a hub of economic activity that we wouldn’t find everywhere. Outside Bamako and a few other hubs, the majority of Mali’s cities are barely more than large towns. Most of the country is made up of small villages, often clustered within walking distance of a larger village that holds a regular market, as many sellers transport their goods by donkey cart, by bicycle, or on foot, balanced on a woman’s head.

The Strømme staff, working through village leaders, had organized a group of about fifty women in a circle of chairs and benches set up under a frayed canvas roof. The staff introduced us, and after formal greetings and small talk, and with the help of a translator, Mamadou and I took it from there.

“Are any of you part of a tontine?” I asked the group. More than half of the women raised their hands.

“How many tontines are represented here?” I asked. There were five. We asked the members of each tontine to sit together and then asked them to explain how each group was organized.

We were surprised by how diverse they were. In two of the tontines, the women simply made weekly payments to a group fund, eventually redistributing the amount they had saved back to each member. According to Stuart Rutherford, a scholar on financial models of the poor, individual savers use groups such as this to “create some gentle peer pressure to help ensure the savings get made.”¹⁴

Two other tontines operated like the ROSCAs, which I expected. Each member contributed a fixed amount every meeting, with a different member receiving the entire sum collected at that meeting until all had received the payout once. ROSCAs replace the risk of storing cash over time (a potential pitfall for the first type of tontine) with the risk that if you are last in line those ahead of you may lose the incentive to continue contributing to later payouts. Rutherford notes that saving over time, receiving rotating payouts from a group, and taking out a loan all serve the same function: the user is able to control when she has a large lump sum of cash on hand, created out of small, regular payments.¹⁵

The fifth tontine used the cash it collected to buy large cans of oil and other commodities in bulk that it could resell at a profit. This buying and selling club functioned like the other groups in that it helped its members commit to saving regular, small amounts of money that might otherwise have been spent on small purchases—an extra half-kilo of rice here or a cup of tea there—until the money accumulated into a usefully large sum. Members could make a tidy profit, though they could also lose their savings in a bad investment.

We were impressed by the variety of financial instruments used by Malian women. The three types of tontines were fine-tuned to fit different people’s needs and comfort with risk. Portfolios of the Poor, a landmark book exploring the strategies poor people employ to survive on tiny sums of money, makes the point that the less money you have to work with, and the less reliable or steady your income, the less secure your ability to store cash, then the more precisely you must manage every aspect of your finances: “If you are poor, managing your money well is absolutely central to your life.”¹⁶ Income must be carefully divided among everyday needs such as food, healthcare, and education as well as investments in business and agriculture to ensure future income. All this while somehow preserving quick access to the larger sums of cash (usually under fifty dollars) necessary for coping with emergencies—emergencies that are, of course, compounded by poverty: a preventable illness, for example, can become an untreatable catastrophe if you cannot afford preventive care.¹⁷ To make our program worthwhile, we would need to understand and build on the careful, complex financial strategies already in use.

“What’s missing?” I asked the group. The response was immediate. They wanted a way to access flexible amounts of money immediately without putting their long-term savings in danger. With credit, they would have the money they needed right away but pay interest. With savings, they would pay with patience. Saving for Change, we posited, would give members access to the upfront cash they wanted in the form of short-term loans while they built assets through convenient, weekly savings. Putting aside a bit of money every week is less of a burden if members know they can take out loans when they need money. When the fund is divided at the end of the cycle, each woman would receive all that she had saved plus her share of all the interest collected. Loan interest payments would become interest returned on savings down the road. Data later proved that a 40 percent rate of return on savings was not uncommon.¹⁸

We explained the Saving for Change method, drawing diagrams on the ground with sticks and piling pebbles to illustrate savings and loan payments. “You would keep your own records,” I remember Mamadou concluding. “You would elect your own group officers. You decide on your own rules. All the records and decisions are left with the group. You decide how much you will save each week, how much interest you will charge, and what you will do if someone misses a payment. We train you on the basic structure and offer support to make sure your groups run smoothly. What do you think?”

“When can we start?” several people responded. We had to explain that this was just a feasibility study to gauge interest and to see whether the idea would work. I explained that measuring local interest and getting feedback for a project is a crucial first step in its design. If the program did indeed launch, we would send trainers to work with them a few months later. Mamadou told them that what we had learned from talking to them that day was that many women in rural Mali already have the skills to organize groups and manage their own finances. We would add to what they already knew.

As Mamadou and I traveled back to Bamako, we talked about how member ownership would be key to success. I remembered the mantra, “Dependency is not empowering.” How far could we go to avoid dependency and increase ownership? “Tontines have spread from village to village, from market to market, long before roads, mass communication, and NGOs showed up,” I said. “How can our program spread like that?” We needed to provide groups with the tools and training that would enable them to manage themselves. Our staff would need to be teachers and grassroots organizers, not financial managers. We would need an exit strategy so that groups would have no alternative except to function on their own: the trainers could work in a region for a few years organizing independent groups and then move on to train groups in new regions.

Looking back, I am proud to say that this “in their own hands” strategy worked. After the first year of training, most groups were visited only once every quarter, some even less frequently. Six years later, we could report that 95 percent of the groups trained in the first two years were still functioning, while closer to 98 percent of the groups formed later continued to save and lend.¹⁹

Selecting Partners

Oxfam’s policy is to build local capacity by working through local NGOs rather than by implementing a project with its own staff. We decided that we would work with Malian NGOs that had already been working in their villages for some years, as they would be more likely to continue working in those villages even after Strømme, Freedom from Hunger, and Oxfam America stopped providing support. Mamadou and I prioritized local NGOs that delivered services well and at scale—for example, those related to literacy, health, education, agriculture, human rights, and finance. Saving for Change would capitalize on those organizations’ connections within their communities and their knowledge of culture, traditions, and language.

After considerable searching, we found two NGOs with substantial operations in villages where we wanted to work. Le Tonus was based outside Kati, just north of Bamako, and CAEB had its central office in Bamako.²⁰ Both embraced the idea of our project as an interesting additional approach that would complement their ongoing microfinance work. Little did they know that in just a few months, the number of members in their savings groups would dwarf that of their struggling MFIs.

Participatory Design and Training

Meanwhile, the project’s design was underway across the Atlantic at Oxfam America’s headquarters in Boston, Massachusetts, and in Davis, California, where Freedom from Hunger is based. Freedom from Hunger stands out among international NGOs for its expertise in designing engaging, empowering adult education programs that it pairs with other development initiatives, such as microfinance or health. I reached out to Freedom from Hunger knowing that its expertise in project design, implementation planning, and evaluation and its deep knowledge of the Mali context would prove invaluable, as the organization already had a strong presence in the country. Happily, they were eager to work with us, and we became three partners. Strømme provided contacts and funding for the local NGOs’ implementation, while Oxfam would oversee the execution and management of the initiative. Kathleen Stack from Freedom from Hunger, along with Vinod Parmeshwar, my deputy director at Oxfam’s Community Finance Department, took on the nuts and bolts of curriculum design.²¹ Vinod had considerable experience developing curricula from his work at CRS in India and had worked with Freedom from Hunger there.

The curriculum design team reviewed a pile of manuals, including those from Freedom from Hunger’s Credit with Education program, Pact’s WEP, and India’s self-help groups, and the savings group manuals developed by the now-CEO of VSL Associates, Hugh Allen, for CARE’s widespread VSLA program, which had by then become the standard savings group model. While drawing heavily from the existing howto guides, we decided to try something different. “We wanted to develop our own in-house capacity to do it ourselves,” explained Vinod. With our own manual, we could create “a self-reinforcing learning loop,” he said, “where we were constantly updating our manuals based on what we learned from the field, so they never got stagnant.”²² For Vinod, the process was as important as the result, a principle that would become core to every aspect of Saving for Change.

marc bavois, an expert on training savings groups, worked at Freedom from Hunger when Saving for Change was just starting. In explaining this active approach to curriculum design, he said that the ultimate goal was for each group “to have a genuine, participatory discussion to come to the group’s decisions,” such as the savings rate or loan conditions.²³ Organizing a group that becomes “a unique institution that is making its own decisions” requires members to engage in “genuine reflection” about their own rules, instead of just copying the interest rates or late fees that work for another group.²⁴ Development scholar and senior lecturer in the Department of Social and Policy Sciences at the University of Bath, Dr. Sarah White, explains that participation in design profoundly changes people’s experience of a program. Participation “is both a means to empowerment and an end in itself [that] transforms people’s reality and their sense of it.”²⁵ When everyone understands how the group works and why, its mechanics become transparent and leaders become accountable to members. It is much harder for someone with wealth, education, or social status to co-opt the group’s resources for his or her own, what we call “elite capture.”²⁶

To put the plan into action, the Saving for Change team organized a three-week workshop to train the first set of animators—the staff of our local NGO partners who would organize groups. We needed to ensure that the animators understood the participatory engagement at the heart of Saving for Change—training the animators had to follow that same practical, “learning conversation” style so that the animators too felt ownership over the core principles of the training process. The animators also functioned as a critical bridge between international funders (like us) and project managers and the poor, rural women who would eventually call the shots.²⁷Animators would need to have the cultural expertise to navigate the complex internal power dynamics of their villages and groups, ensuring that everyone could understand and engage on equal terms.

As the training progressed, the trainees were sent out to put into practice the lessons they had learned the day before as they formed groups in nearby villages. This practice helped develop each animator’s skills while their feedback informed the program design. “It is a principle of adult learning to be able to immediately apply the learning,” explained Vinod. “So as we trained the animators, we were gathering input and making real-time adjustments.”²⁸

After three weeks of training in January 2005, the animators were ready to start work on their own. When I returned to evaluate the program that August, I was enormously relieved to find there were already 216 groups with five thousand members in place in just seven short months. We were ecstatic.

Introducing Saving for Change and Oral Record Keeping

Not that the work was easy. Early on, many animators found that villagers were skeptical of their efforts to organize groups. “Sometimes I faced opposition,” Fatoumata Traoré, one of the first animators, reflected. “Sometimes people do not believe the information of the development program—they do not believe you are who you say you are.”²⁹ Villagers had learned from prior experience to be wary of anyone seeking to manage their finances.

Kanimba Samake, an early member and a Saving for Change volunteer responsible for training new groups in her own village, later told us about another concern. “One of the doubts I heard from other village women,” she said, “was that they were worried [that] if they would get the loan, and if they had trouble paying the loan back, that they’d be put in jail. … To convince others, I told them that Saving for Change is different from other types of [microfinance], this is our own program.”³⁰ With Saving for Change, members manage their own money, they never turn it over to the animators or NGO, and the groups themselves assess whether a member is qualified for a loan and what fines to impose on late payments or default.

Member ownership required transparency. We soon found that the written record-keeping systems we had come up with were out of place in an environment in which few women could read. Written records were not transparent to the illiterate majority. Written records would concentrate power in the hands of one or two members who could read or would force groups to depend on a literate volunteer, like a member’s husband. Sooner or later that person would dominate the discussions and decisions of the group.³¹ I had seen how written records often slow everything down. At the low levels of literacy and numeracy found in rural Mali, ledger entries would generate a lot of waiting time and frustration in meetings while a few people struggled over computations and other members waited for results that they only partially understood.

In Nepal, Marcia Odell’s WEP program handled illiteracy by building reading and writing education into a fundamental part of the project. With a higher level of literacy, written records could work. CARE’s VSLA model was increasingly settling on passbook-style stamp cards to keep track of member savings amounts and loan payments, with considerable success.³² In the regions where we worked, only half of the men could read and write, and only 28 percent of women were literate.³³ We took the bold step of scrapping paper records altogether and developed instead an oral record-keeping system based on the members’ existing capacity to recall a limited number of transactions.

Vinod took the lead designing the system. “We created a series of steps that we then changed as it was implemented, and over time it solidified,” Vinod continued. “The key was that everything had to be announced loud and clear and it needed a regimented process.”³⁴ Each person would hold up her savings contribution for all to see and then put it in the cashbox. Eventually, the women started to sit in the same spot each week, leading to the buddy system—each member just needed to remember her own information and that of her neighbor, both to verify her memory and to fill in if someone was absent. Vinod boiled down the record-keeping system to five essential pieces of information: did they make the savings payments; did they owe any fines; and, if they had a loan, how much was it for, when would it be repaid, and what was this month’s payment? “Five points to remember for yourself and for your neighbor.”

Vinod explained that “this ensured that, if the group leader left, the group wouldn’t be broken apart, because the information was extremely decentralized. It also embedded participatory group management in a way that didn’t create a new elite. This is something I saw in India—that the group leaders often got extra training, leading to the creation of a new elite within the group—and we didn’t want that to happen. Our groups are more resilient because they’re not dependent on a leader.”³⁵

Different groups improvised methods to make oral record keeping work better. Animator Lamine Coulibaly explained that he saw groups keep pebbles in their cashbox to count the number of meetings that had taken place and to record the number of interest payments an individual had made on a loan.³⁶ A group I once visited showed me how they used shea nuts and sticks to represent loans of different sizes that had been given out, allowing the treasurer to easily keep track of both the amount of cash on hand and the portion of the group’s funds that was out on loan. Animators systematically spread these new ideas to other groups so that all could benefit from the inventions and innovations.

In keeping their own records, group members gained new skills and independence. Fatoumata described what a woman had told her:

Before Saving for Change, women in our village were not able to calculate and make differences, but today women are able to count money from one to one million CFA. Before Saving for Change, men didn’t believe women could go or meet with microfinance [loan officers] alone because men were afraid women wouldn’t understand the rules, or someone would steal from women. Now men are more confident that women can go to meetings, go meet an MFI to get a loan; men do not feel obliged to accompany them. Before they never let women go alone.³⁷

As time went on, Vinod and I continued to build the framework necessary to shift the locus of power in Saving for Change away from Oxfam America’s Boston headquarters and to the village groups. We set up a technical unit of our best animators and NGO supervisors, led by Paul Ahouissoussi as regional program coordinator from the Oxfam West Africa regional office in Dakar. Fatoumata Traoré, the most skilled animator from CAEB, and field coordinator Soumaila Sogoba of Le Tonus had been working directly with groups for months. Their charisma and track records made them obvious picks for the technical unit. They fully understood the method, were passionately committed to it, and had credibility and respect among the other animators. Hiring from within created an internal career ladder, allowing those who excelled room to move up. We chose the next best animator at Le Tonus to replace Soumaila as supervisor there. Although Vinod and I continued to give advice and overall direction, our role moved increasingly to implementing Saving for Change in other countries. Vinod said, “I call this the backward dance.… How do you push the responsibility more and more to the people and back out of the system?”³⁸

Voluntary Replication

The technical unit and partner NGOs engaged in their own backward dance, gradually expanding the role of the animator from “trainer of groups” to “trainer of trainers.” In our vision, volunteer replicating agents would take on the bulk of the work of organizing new groups. At first, we imagined a spontaneous process of inspired group members bringing the savings group idea to their families and friends, or neighbors who witnessed a group’s success approaching existing groups and copying their methods, as I had seen in Nepal and India. Reports of unprompted, unlinked new groups were gratifying and exciting. I anticipated that this process of spontaneous replication would cause Saving for Change to spread virally across Mali, without need for our support.

Unfortunately, spontaneous replication, though important, proved both insufficient and inefficient. Groups would often form, but they would be poorly organized, missing key elements of the structure that had been so carefully designed to ensure transparency, participation, and ownership by all members. We felt too that the rate of replication was low compared with what we had hoped for and what we would need to reach our desired scale. Animators, we learned, sometimes stopped volunteers from forming new groups because they were not sure whether they should be encouraging replication. The animators who did encourage volunteers to train new groups did so on an ad hoc basis, but their roles were confused—who did what? Should volunteers just teach from memory based on how their own groups work, could they perhaps shadow an animator while she trains and copy her process, or were they supposed to receive specific guidance?

We were encountering the tension between maintaining a high-quality program with clear structures in place to ensure effectiveness, accountability, and transparency and our desire to leave it all in the hands of the group members. The key to resolving this tension came in the form of an interactive training process that empowered participants to feel ownership over their knowledge. A later study would prove that formal training for replicating agents resulted in greater participation in Saving for Change, and better socioeconomic outcomes such as increased livestock ownership and food security.³⁹

By closely observing the group replication activity already underway, we designed a system to move from laissez-faire, spontaneous replication to the “replicating agent model,” discovering that we could vastly increase the number of groups trained in the process. Working closely with Freedom from Hunger, we simplified the Saving for Change manual even further, cutting text until it was almost entirely a set of simple pictures that illustrated each step in training a group. Our intent was that someone with low or no literacy skills, once trained on this manual, would have the tools to facilitate the participatory, discussion-based training sessions we felt were essential to creating member-managed savings groups.

Next, we adjusted the animator’s role in introducing Saving for Change to a village. Animators would train the first group and then move on after identifying one or two smart, energetic women who wanted to take on the replicating agent (RA) role in their village. By working at the same time in twenty or more villages in an area, animators could train many RAs at once. The adventurous women willing to join a village’s very first group would provide a good source of ambitious local volunteers who would then train more groups in their own villages. When the animator eventually pulled out to form groups in a new region, the RAs would continue the work independently. The endgame was geographical saturation—every village in a geographic area would have a Saving for Change group.⁴⁰ This would help ensure that even the poorest or most marginalized households in a village would be able to participate in Saving for Change.

Animator Basenji Coulibaly explained how he chose RAs from the women with whom he worked. “My criteria are that when I go to a village, I start working with a group. I try to notice and identify women who are open-minded and dedicated to the work, and whom the other women listen to. When I notice someone like that, I go to her and propose to her to become an RA. If she agrees, we go to her husband to seek his agreement before moving forward. After that, when everyone agrees (the lady and her husband), then we go ahead and I explain and teach how to create a group. ‘What are the rules? What are the steps?’”⁴¹

Another animator, Lamine Coulibaly, reflected on the differences between paid animators and voluntary RAs. “Maybe the RA is more advantageous because she is in the village. The fact that she’s always in the village [means] she knows the realities in the village. … She is living within the community so she knows the problems, speaks the same language, and speaks about the same things, so communication is better between them.”⁴²

A volunteer replicating agent, Fatoumata Bagayoga, explained the value of her work:

I told them about the roles of officers and that they have to put these [positions] in place in order to become operational. I explained how they save, how they give loans, how fines and repayment are done. I talked to them about all these rules so they could understand. … Before people understood how it worked, it was challenging. I do this training even if I’m busy with my own household work. I still do the trainings. It is important to me because of how people will consider me in the village—other people will see it is me who is doing this important role in my village.…

There were some women in the village who didn’t know anything about selling things or petty trading. But when Saving for Change came, and some women started small businesses, these women learned from them. The health situation of the people has also improved because it made money available to buy healthcare. This was not possible in the past because you could not get money from anybody to pay for healthcare. Another thing is if you have a child who is in school, and you have problems regarding school fees, you can borrow some money from the group and solve this problem. …

I take loans from my group to buy rice, cook it, and sell it. Also I use part of the money to buy and sell shoes. I was doing this before Saving for Change. Before Saving for Change, the quantity of rice was 3 kg, but now I can buy 10–15 kg to cook and sell. Before Saving for Change, I sold 19,000 CFA worth of shoes, but now I do 50,000 CFA. Before Saving for Change, I managed with the little amount of money I could get from agriculture. It was not enough money, I was really barely managing. I bought some sheep with my share-out—five sheep for myself.⁴³