In the June 2009 issue of Enterprise Development and Microfinance, I summarized Freedom from Hunger’s experience with village banking offered by credit unions and rural banks. Here I excerpt the section titled “Poverty status of clients.”
For most of the past two decades, widely accepted poverty measurement methods that can be used by microfinance practitioners have not been available. Moreover, the institutions featured in this paper had little incentive to invest in poverty measurement because they were not targeting their services to people below a certain poverty line or criterion. Therefore, the poverty measurements reported here are made, in some cases indirectly, in the context of a broader impact research design. For example, Freedom from Hunger’s 1998 Ghana study rigorously compared participants with nonparticipants in communities randomly selected to receive the Credit with Education services and also with residents of matched communities randomly selected to serve as control for the participant communities. Before the introduction of the Credit with Education service by the Lower Pra Rural Bank, there were no important differences found in the general characteristics of the people who later became participants compared to those who became nonparticipants (by choice in the participant communities and by chance in the control communities), including no differences in indicators of poverty level such as household assets and consumption. The conclusion is that women residing in participant communities were equally likely to use the Credit with Education service regardless of their poverty status. In other words, the distribution of poverty status among the women served by Credit with Education is the same as that in the general population of eligible women in the participant communities. The village banking design did not discriminate against poorer women in this case.
What this conclusion does not yet address is whether the “poorer women” in these communities were in fact “very poor.” However, Freedom from Hunger defines “very poor” as being poor enough to be chronically food-insecure. The impact study did show that incidence of food insecurity was equally high in both the participant and control communities and was substantially reduced only among the participants (and their households) after at least one year of Credit with Education participation. The conclusion is that many “very poor” (food-insecure) people resided in these coastal Ghanaian communities, and they joined the Lower Pra Rural Bank’s village banks in the same proportion as the general local population. Moreover, the status of the very poor was substantially improved, for many to the point of leaving the category of “very poor.” The village banking design did not discriminate against even the very poor in this case.
Corroboration of the conclusions of the Ghana study is found in a study of poverty outreach in two Malian credit union federations (Nteziyaremye and MkNelly, 2001). Participatory Wealth Ranking was used to sort all households in nine communities into four relative wealth groupings that local informants defined as:
- Households that are food-secure
- Households vulnerable to food insecurity
- Households with periodic food insecurity
- Households that are chronically food-insecure
The wealth distribution of Credit with Education members mirrors the overall wealth distribution in the communities in general. Despite the program terms, a certain number of women from better-off households will join Credit with Education. And despite their extreme poverty, a surprising number of women from the poorest, most food-insecure households will also join.
Women from the poorest one-third in the community were interviewed—women who had never joined the program, women who were current members and women who were ex-members—in separate focus groups. The interviews revealed little to no evidence of the poorer women being systematically excluded either by better-off members or by program representatives. However, some poor women were self-excluding, choosing not to join out of fear for their already precarious economic situations.
Does Credit with Education serve poorer clients than the other products of local financial institutions?
The same study of the two Malian credit union federations also used a basic-needs methodology, which creates a poverty index based on the population’s own perceptions, as the interviewees themselves define what is a “basic need” (that which no household should have to live without). Credit with Education clients were in the relatively poorest client category for both credit union federations. The same result was found in an IFPRI study of the OTIV credit union federation in Madagascar using the CGAP Poverty Assessment Tool (Lapenu, Sharma and Ralison, 2000—I cannot find this research report on the web anymore). Credit with Education members appear to be the poorest group in the sample collected; “classical” members of OTIV were significantly better-off compared to both non-clients and “caisse feminine” (Credit with Education group) members.
Extending credit union services beyond the towns and large villages in which the branches are typically located seems to be more important for poverty outreach than loan terms (including joint-liability groupings) or even preferential lending to women. For example, in Mali, borrowers of credit union financial products designed for farmers were not significantly less poor than Credit with Education clients, though they were mostly men borrowing as individuals. The rural borrowers (farmers and Credit with Education clients), however, were significantly poorer than urban borrowers. The IFPRI study in Madagascar also found that the poorest households were much more likely to be rural than urban, and two-thirds of the Credit with Education clients were rural while the opposite was true for the “classical” members of OTIV credit unions. Products brought to the villages reached a relatively poorer clientele.
The value of Credit with Education for poverty outreach by local financial institutions may not be so much that it attracts poor people as that its group-based methodology makes it affordable for the institution to offer the service in rural areas where the poor are a higher proportion of the population.