We have to ask the client how much she or her household holds in total savings in whatever variety of mechanisms she and other household members use. Accurate answers are quite hard to get. This post looks at the answers some researchers have gotten.
Sebstad and Cohen (2000) wrote that “findings from studies on the impact of microfinance programs on savings are surprisingly limited.” Still, they cite a study of BRAC in Bangladesh that found “the savings by members, including mandatory member savings, were twice as high as savings for a control group.” Another study showed that savings by CVECA members in Mali complement rather than substitute for informal savings mechanisms.
Sebstad and Cohen also cite some interesting uses of microcredit to build households’ informal savings. In the Philippines, CARD loans “indirectly helped clients join ROSCAs, because their membership in CARD Bank improved their credit rating with other ROSCA group members.” In Uganda: “Many clients, especially poorer clients with fewer resources of household income, put aside part of their loans in the form of cash savings.” Such “precautionary” savings to safeguard against risk (including risk of non-repayment of the loans) might be expected to increase as poorer people use microfinance services but then decline as these same people gain access to other, more direct risk- management tools, such as emergency loans and insurance.
Both in Ghana and Bolivia there were significant differences in the percentage of participants with their own cash savings between years versus controls and nonparticipants. There was also a significant difference in log value of amount saved between years for participants versus nonparticipants and controls. The amount that participants had in savings also varied, again indicating the range in economic success among those borrowers living in the same communities and participating in the same program. For example, in Ghana, participant savings varied from $0 to $118. In Bolivia, participant savings varied from $0 to $132. And in Mali, the impact survey found evidence of positive impact on whether women reported having personal savings.
The respondents in Freedom from Hunger’s “impact stories” (Jarrell et al., 2011, p. 24) revealed “that participation in Credit with Education [credit-led group microfinance] or Saving for Change [savings-led group microfinance] can make the difference between … knowing that some unexpected event in the future will endanger the financial position of the family even further and knowing that there are sufficient savings put away to protect the family from that financial crisis.” The great majority of respondents in Bolivia, Burkina Faso and Benin described their economic benefits partially in terms of adding to or contributing to savings, which perhaps indicates that microfinance participation added to savings rather than simply substituted for other, less formal saving mechanisms.
Focusing on the impacts of savings groups sponsored by INGOs like CARE, Catholic Relief Services, Pact, Oxfam America and Freedom from Hunger, Megan Gash (in chapter five of Savings Groups at the Frontier) summarized the results of several impact studies:
Since all savings programs require members to save, the accumulation of savings is considered a process outcome of the program, although when members recoup these sums varies. Some groups distribute only the earned interest and put the bulk of their money back into the next cycle to build more loan capital, while others distribute the entire pot. Smaller sums received at share-out tend to go toward consumption-smoothing, and bigger sums are often used for large purchases or investments, such as furniture or significant home repairs.
It seems that savings groups (and similar semi-formal, group savings mechanisms) build savings, almost by definition, but the lump sums accumulated seem to be transformed quickly into consumption or physical assets. That means there is no long-term build-up and preservation of a stash of cash readily available to meet unexpected needs or to build toward purchases that are very large or toward meeting needs in the distant future. On the other hand, the savings group is a form of social capital that can persist for years, dependably building for each member a modest stash of cash every year (or whatever the term to payout, or cycle, may be). And some groups choose to keep rolling over their savings to accumulate over several cycles.
But do the savings in the savings groups add to or substitute for savings that would have been kept elsewhere? Given the research to date, Megan asked but could not answer:
Is the amount still relatively small compared to their overall household assets? Do members save in other ways during the savings group’s cycle, and if so, what percentage of their total savings is invested in the groups?
Kathleen Odell (2010) nicely summarized the first randomized controlled trial (RCT) of the impact of microfinance on savings, by Dupas and Robinson (2009) . They opened savings accounts in a village bank in rural western Kenya for a randomly selected sample of poor daily-income earners: women market vendors and men who operated bicycle taxis. Despite the accounts’ effectively negative interest rate (no interest paid and substantial withdrawal fees), 89% of those 122 men and women offered an account actually opened one. Fifty-five percent of the 122 actively used their accounts (59% of the women, 51% of the men). These savings accounts did not crowd out other forms of saving, such as purchasing animals and participation in ROSCAs. Therefore, “total savings appear to have increased significantly thanks to the treatment” (see p. 13 of the 2012 version of the paper). Dupas and Robinson concluded that savings constraints were real, and these constraints were relieved by providing opportunities to save with a village bank (a form of microfinance provider). If so, access to microfinance encouraged and allowed these rural Kenyans to save more than they would have saved without the access to microfinance.
Only one RCT in one country (Kenya) with one type of microfinance client (women market vendors) under artificial conditions (the researchers paid the account opening fee for randomly-selected self-employed people who chose to accept the offer of a savings account), but it does give us a credible anchor in reality for less rigorously sourced evidence that microfinance access can increase total savings.
The next post takes a look at careful RCT research delving into the psychology of saving and pointing to effective design of savings products.
Comment from Laura Fleischer Proano, Director, Savings Group Methodololgies at Freedom from Hunger (I added this comment here, because the comment feature of The Evidence Project blog site is under reconstruction to deal with literally thousands of spam comments! Fortunately, I was moderating comments and blocked them. For the time being, if you have a real comment, please send it by email to firstname.lastname@example.org). Here is Laura’s comment written Nov. 29:
Great questions, Chris. The RCTs conducted by IPA on the Savings Group programs of Freedom from Hunger/Oxfam in Mali and CARE in East Africa (results forthcoming in early 2013) will surely help to shed light on your question about savings in Savings Groups adding to or substituting for savings that would have been kept elsewhere.
Financial diaries and qualitative research conducted with Savings Group members can also help us to understand some of the nuances that the RCTs may not reveal. For example, in addition to knowing if the savings amounts increased or if savings shifted from one place to another, it will be important to understand how members use savings services, which services meet which needs and the overall impact of the different types of savings services. As a Savings Group practitioner, some of my questions include: What needs do Savings Groups meet? Do Savings Groups appropriately meet these needs? What is the impact of meeting these needs? How does this compare to other savings options available? What other needs are unmet? How can those needs be better met to achieve our goal of poverty alleviation and food security?