Evidence that Microfinance Empowers Women – Schuler and Hashemi

In my first post on empowerment (# 55), I listed the six domains developed by Sid Schuler and Syed Hashemi to describe and measure women’s empowerment. Here I present results from their 1994 report of the effects of Grameen Bank and BRAC membership on an empowerment score (based on these domains) and on use of modern contraceptive methods in rural Bangladesh.

In Due Diligence (p. 203), David Roodman briefly reviewed a different report by Schuler, Hashemi and Riley (1996) based on the same field work and using similar non-experimental quantitative analysis. He was skeptical of the value of the study design to determine cause and effect, and I agree with his point. The results below demonstrate association or correlation between microfinance access, women’s empowerment and contraceptive use, but they do not establish causation. Still, I single out this study for presentation, because it is carefully done and describes a highly plausible causal pathway. The different effects of Grameen vs. BRAC membership are particularly interesting.

Schuler and Hashemi (S&H) did their study and analysis in two parts. First, they compared randomly selected villages served by Grameen Bank with nearby randomly selected villages without the Grameen Bank or BRAC. They found that:

    • Grameen Bank membership was highly correlated with both the woman’s empowerment score and the woman’s financial contribution to the family. S&H proposed that the positive effect of membership on women’s level of empowerment was “attributable, in part, to its effectiveness in strengthening women’s economic roles through credit, and in other ways as well.”
    • Grameen Bank membership was highly correlated with both the woman’s empowerment score and the woman’s use of modern contraceptives. S&H proposed that “women who are empowered and women who contribute substantially to their family’s support are more likely than others to use contraceptives.” They further propose that Grameen membership’s effect “on levels of contraceptive use in program villages can be attributed to its effects on the women’s levels of empowerment.”

In other words, the evidence presented by S&H is consistent with the hypothesis that Grameen membership leads to credit-induced increase in ability to contribute financially to the family, which leads to increased empowerment, which leads to increased use of contraceptives. However, S&H’s reference to “other ways as well” admits to the possibility that the causal pathway is not so linear and that Grameen Bank membership, independent of the credit effect on financial contribution to the family, may increase empowerment as defined by the score they applied. Likewise, membership and empowerment may both contribute directly to increased contraceptive use.

The second part of their analysis compared the effects of Grameen Bank and BRAC membership on empowerment and contraceptive use.

    • Grameen Bank membership, BRAC membership and being a nonmember living in a Grameen Bank village (Grameen and BRAC were present in different villages, with little if any overlap) were all highly correlated with the women’s empowerment score. However, the score was substantially lower for nonmembers in Grameen Bank villages (did not look at nonmembers in BRAC villages, because BRAC tended to include everyone in the village while Grameen did not).
    • Grameen Bank membership is highly correlated with contraceptive use (59 percent usage rate compared to 43 percent in control communities), but BRAC membership was not.
    • Over an 18-month period, the rate of contraceptive usage increased much faster among Grameen Bank members than among BRAC members and residents of control communities.
    • Controlling for respondents’ use of contraception before they joined the program, Grameen Bank membership still has an effect significantly greater than BRAC membership.
    • Remarkably, nonmembers in Grameen Bank villages also had significantly higher contraceptive use than the control communities but not nearly as much as the Grameen Bank members. Even when controlling for level of empowerment, the effect of being a nonmember in a Grameen Bank village was significant; this suggests that the mechanism by which Grameen Bank influences contraceptive use of nonmembers does not directly involve empowerment.

It is frustrating that the study design cannot be conclusive in teasing apart the effect of participation in a group from the effect of credit for microenterprise, separately or together, on women’s empowerment and on contraceptive use, separately or together. The causal pathways are there but the picture is too out of focus for us to see what leads to what.

Nonetheless, I find this study compelling for three reasons:

First, the measure of empowerment is multidimensional; a single score embraces five of the six domains S&H had identified. This obscures the relative importance of different domains, but it is more likely to register a change in empowerment than more narrowly defined measures. I suspect that several negative attempts to find correlation between microfinance and women’s empowerment are due to the narrow definition of empowerment adopted by the researchers. Moreover, it seems clear that participation in a group has an empowering effect beyond any direct economic impact due to credit.

Second, the study shows that microfinance participation (whether it is the group membership or the credit or both) seems to empower women in some way that leads to tangible, convincing evidence – contraceptive use – that this microfinance-stimulated empowerment can lead to meaningful, presumably positive, change in women’s lives.

Third, the contrast in results between Grameen Bank and BRAC members shows that program design matters, but in surprising ways. Both programs actively promoted empowerment through staff interactions with members; the difference was that BRAC provided fairly traditional training in women’s rights, health and literacy while Grameen Bank members were focused on enterprise credit but were also required at each weekly meeting to march, salute and chant from memory the “Sixteen Decisions,” including a simple exhortation to keep their families small.

In both programs, women’s identification with their group gave them a socially legitimate reason to meet and interact in the public sphere, which served to increase their exposure to new ideas, their self-confidence and their mobility; hence both programs seemed to empower their members. S&H concluded that the more regimented and ritualistic approach of Grameen (compared to BRAC) developed a more intense identification and bond that made it easier for women to resist the restrictions of traditional family life and adopt new family planning norms. Moreover, Grameen Bank sent its members a very clear, simple message about family planning (which dovetailed with widespread messaging from family-planning field workers from other organizations, from radio and from neighbors) and reinforced this message at every weekly meeting.

From this study emerges a picture of several interacting elements: women from a very restrictive culture join groups organized by and regularly interacting with field officers from a microfinance provider that offers credit and information that can be used to increase each woman’s financial contribution to her household and to become aware of new options, which together can lead a woman to increased self-confidence, autonomy and connections outside her household that can be called empowerment. Altogether, these elements allow and even encourage the women to dare to adopt the use of contraceptives and other bold innovations.

I don’t know of any field experimental study that incontrovertibly demonstrates the validity of this picture or the pathways of cause and effect, but S&H come intriguingly close with their careful regression analyses. I am thinking that we need to look particularly closely at the roles of the groups and the field officers to find further confirming evidence.

Onward to the next post – on the value of groups.