Chris Dunford: Freedom From Hunger has had quite a bit of experience with poverty assessments of its credit with education clients. And these are not our clients; these are the clients of organizations that are our partners who have learned to do credit with education with our help. And, just for some examples here—Madagascar and Bolivia and Mali—we have seen the application of the CGAP poverty assessment tool. We've measured poverty against the national poverty line, and we've used participatory wealth ranking. So, we've become familiar with the variety of methods. You'll notice the poverty assessment tool from CGAP gives you some sense of whether your clients are in the bottom third of the local area, as far as poverty is concerned. But, you can see that for Bolivia you get a very different result if you compare it to the Bolivian national poverty line. This national poverty line is based on an index of basic needs which is very different than the kind of poverty line you would get in looking at surveys of income and expenditure. So, there are a lot of problems with the comparability of the data. I should say that, for managers, a relative measure of poverty really is quite adequate. Let's say it's more the donors who need something that's going to be relevant to an international standard. And so if you're applying a measure like the poverty assessment tool of CGAP or the participatory wealth ranking which is considered to measure relative poverty relative to the local area rather than to some international standard or even a national standard, that's really probably fine for the MFI manager and his/her staff. The problem becomes when you try to compare the results from different MFIs, which is more of a donor problem or an academic's problem. When we did the participatory wealth ranking, we noticed, as John de Wit did in his presentation, that one of the characteristics that local people will use to classify their fellow citizens is their food security. And, we found this also in Mali. And, we found that some 70-75% of the households that were being served by credit education were households that suffered some degree of food insecurity. And, what occurred to us there, and I think this is true also of the CASHPOR House Index (CHI), is that you can, perhaps with some small alterations, take a measure like CHI and even PWR and relate it, calibrate it to some kind of international standard ... not the dollar a day, in this case, but to the fact that the very poor (and, if you're concerned about the very poor rather than trying to look at various gradations of poverty—the not-so-poor and the vulnerable non-poor) ... If you're just concerned with the percentage of the very poor in your clientele, then you can make a fairly quick distinction between those that are suffering from some kind of food insecurity and those that are not. Food insecurity [is] a very basic characteristic of absolute poverty.

That intrigued us and led us to kind of a radical proposition. Perhaps to construct a poverty assessment measure, it could be very simple and it was based on food security scales. Specifically looking at—and you might find this a little humorous but it actually does appear to work rather well, as I'll explain in a moment—actually asking people about the anxiety that they have about food being sufficient; the experience of running out of food; substituting fewer or cheaper foods for the ones that they would prefer to eat and having to actually reduce food intake, skipping meals and having smaller meal sizes. This kind of food security scale could be constructed like this. Where you have, in this case, four categories (this reflects the categories that local people use in their participatory wealth ranking in Mali). You have households that are food secure having access to food at all times; those that have significant anxiety about having enough adequate food and have limited variety or quality of foods; those that are experiencing reduced food intake by adults and having consequences like the actual physical sensation of hunger and, those that have more diverse and frequent instances of hunger for adults and also hunger for their children. Another way of saying it is that the first category group has adequate food, as far as they're concerned, all the time; those in the second category have some years when it's tough; and the third category, every year in the season right before the harvest of the next crop, they are hungry as defined here; and, those in the fourth group have a problem every year all the time. And, so you can pretty well say that the group four is the absolute poor, and groups two through four are subject to some kind of food insecurity.

…We wanted to go beyond this a little bit and we took a look at the USDA. The US Department of Agriculture is the unit of the US government that looks at food insecurity and poverty in the United States. They've developed food security scales to go along with studies of consumption and expenditure on foods. They have an 18-item scale, a 6-item scale and a 1-item scale. And, we were intrigued by the idea of a 1-item scale. But…they've demonstrated the validity of these scales—even the 1-item scale—in comparison to caloric and nutrient intakes, food inventories, economic measures, (such as household food expenditures and weekly household income). So, it seems to be a very powerful tool as simple as it may be. Maybe we've found something that's on the far side of complexity, simplicity on the far side of complexity, something that's been really checked out and it really works quite well. Perhaps, it could be one of the items on the menu. Researchers at the University of California at Davis have been refining this USDA set of food scales for use in Mexico and other developing countries. [Freedom From Hunger, also based in Davis] has really engaged in interesting conversations with them.

What's fundamental in coming up with a measure of poverty is to actually ask people questions or observe what the characteristics of those people's lives are. So, what you're measuring are the characteristics of their lives, the quality of their lives, not some tangentially related aspect like loan-size.

Syed Hashemi: [In summary] I would just like to make a couple of comments. First of all, every MFI does not have to reach the poorest. It depends on what your vision is. And, of course, all sections of the poor need financial services. Even a lot of the people above the poverty line are not served by formal financial institutions; they too require financial services. It depends again on your vision. For many of us, of course, that is the vision, [to reach the very poor]. And, for a lot of donors, there are mandates that tax-payer money be used to achieve the Millennium Development Goals, that tax-payer money be used for the poorest. So, there is a bottom- line imperative for many of us. The second point I want to make is on accuracy of measurement. You always hear this: "How accurate are the measurements?" Now, even the million dollar household expenditure surveys that are done or the million dollar World Bank LSMS surveys that are done are not accepted by equal numbers of specialized econometricians. And, you have esoteric econometric debates on how good poverty lines are, on how good dollar-a-day is as a measurement. So, at that level of accuracy, the jury is always out. What we are looking for, really, is something that is reliable, something that's meaningful, something that is low cost, something that makes sense in the majority of the cases.

Return to Workshop Transcript