The last few posts make a compelling but inconclusive case that group-based microfinance fosters women’s empowerment. But there is a contrary view. In Due Diligence (chapter seven, especially pp. 202–220), David Roodman examined the evidence that group microfinance increases the individual member’s “agency” or “freedom” to act in her own best interest (a major dimension of empowerment). Because of standardized loan terms and peer pressure to repay, he concluded tentatively that group-based borrowing is inherently less empowering than individual borrowing.

Helen Todd’s excellent ethnographic field work with Grameen Bank members in Bangladesh, reported in her 1996 book Women at the Center, seems to reinforce Roodman’s conclusion. She found that solidarity groups do not foster solidarity (mutual aid in hard times and standing together to confront those in authority).

Here is what seems to be the logic of those skeptical about the value of groups. If there is no solidarity value of group membership to offset the cost to individual agency, then (all else being equal) net empowerment should be negative. Net positive empowerment from group borrowing is presumed to be generated by the economic success that comes from borrowing money, not by being a member of the group. If both group and individual borrowers receive equal value in their loans, then credit should be equally empowering for group and individual borrowers. The group borrowers, however, shoulder the additional, dis-empowering cost of group membership. Therefore, net empowerment from individual borrowing should be greater than from group membership.

Are groups indeed just a burden on the members that is imposed solely for the risk- management benefit to the lender? Certainly this has become a popular narrative in microfinance circles in recent years. The natural response of believers is to do everything possible to free the borrowers from the burden of having to come together in groups. It makes so much sense to denizens of “modern” individualistic cultures that this narrative is hardly questioned.

But question it we must, because there is so much evidence to the contrary. To start, Todd’s solidarity is but one dimension of empowerment. See my post # 55 on the many dimensions of empowerment and posts # 56 and # 57 for evidence that group microfinance does empower women in other dimensions. The evidence I presented is inconclusive but no more so than Todd’s evidence. Moreover, in South Asia, social capital beyond the kin group is low, and the culture is notoriously restricting for women. In that setting, the solidarity dimension (covertly or overtly political in nature) is the dimension of women’s empowerment least likely to be affected by microfinance. But even regarding this solidarity dimension, Roodman observed that the Self-Help Groups (SHGs) of India seem to be “designed and implemented much more deliberately to help women perceive and attack the injustices done to them” (p. 214).

So, there are groups and there are groups! Roodman correctly agrees with the conclusion of Linda Mayoux “that microfinance does not automatically empower, but it can if designed to do so” (p. 218). Unfortunately, Roodman goes on to contend that the empowerment in SHGs is made possible by their lax financial discipline. This is likely to be a spurious correlation, not a causal connection—groups can be both empowering and self-disciplined. Again, it is a matter of design and implementation, as Roodman also concludes: “Fortunately, the short-term trade-off between the lender’s bottom line and the client’s freedom is not an iron law. In fact, MFIs have found many ways to dodge the trade-off—for example, by making loan repayment terms more flexible and taking savings” (p. 219).

Let’s now look at the randomized trial in Mongolia (XacBank) that compared lending to groups of women vs. lending to individual women—the first rigorous comparison of the two types of lending. This study found that access to group loans led to more enterprise formation, better enterprise success, and more and healthier food consumption, but access to individual loans did not. In both group and individual loan treatments, the less educated (presumably poorer) clients experienced these benefits more than better-educated clients—benefits were even greater when they were members of groups. Repayment rates, however, were the same for both treatments.

Clearly, something good, from the client viewpoint, is happening in these microfinance groups. Since there are groups, and then there are groups, it is important to inquire into the details a bit. These Mongolian group members were bound by joint liability for the loans to fellow members and responsible for vetting loan proposals as the first screen (in the second screen, the lender staff reviewed and approved each member’s loan proposal that had been pre-approved by the group). Regular repayment meetings were not required by the lender (the group leader collected payments from members and handed them over to the lender staff, so it seems the lender staff did not meet regularly with the group). Nonetheless, the groups organized their own meetings, roughly once a month.

The researchers speculate that the group imposed a greater discipline on members to invest in the enterprises proposed in the loan application. This view was reinforced by the finding that group members received and gave less money to friends and relatives than those with individual loans—less leakage of the loan capital to non-enterprise uses and fewer informal financial transactions with non-members of the group. What the researchers could not tell us is what kind of interaction, and possibly mutual assistance and encouragement, might have occurred among the group members.

What are we to make of these results in light of the presumption that group microfinance primarily benefits the lender and dis-empowers the client compared to individual microfinance? The effect of credit on the Mongolian clients should be the same (loan sizes and terms were roughly the same). The lender’s costs and revenues were effectively the same for both types of lending, so there was no apparent advantage to the lender one way or the other. The only apparent difference was in the internal dynamics of the group, including the joint liability. The researchers did not investigate any dimension of social capital or empowerment as either an input to or an outcome from the groups, so we are left to speculate about the internal dynamics.

Fortunately, some other studies by Dean Karlan and his associates look directly at these internal dynamics and offer insights into what may be happening inside the groups to help members do better. I will review these studies and their implications in the next post.