What I'm doing

Well, the blog is called Social Performance and Salteñas so I guess, the appropriate thing to do would be to follow up the expository post with one that explains what social performance is and what I’m doing here (I’ll let you guess what the third post will be about).

Without going into too much detail, FONCRESOL is a development-oriented microfinance institution (MFI). This means that, unlike profit-oriented institutions which see microfinance simply as way of tapping into previously underserved markets, FONCRESOL provides credit with the primary aim of developing local sustainable economies. This is not to say that FONCRESOL is a charity. Making a profit is still a goal, but profitability is viewed more as a way to ensure the sustainability of the enterprise than as a goal in and of itself. This is what some people call a double bottom line: FONCRESOL seeks social returns as well as financial ones.

This double bottom line, however, makes it very tricky to gauge the success of the institution. While measuring financial returns is well understood (easy even), it is notoriously difficult to measure an institution’s social return.

This is where I come in.

Measuring social performance has been a hot topic in microfinance circles over the last 15 years. Previously, it was common to assume that if customers returned for subsequent loans, that this was enough of an indicator that the loan programs were working and that clients were happy. But with the proliferation of microfinance institutions around the globe, and the growing sophistication of donors and social investors, this method of gauging social impact ceased to cut the mustard. People wanted to really know if these programs were having an impact or not, and maybe even how much.

There are a million different approaches that MFI’s have taken to evaluate how good they are at encouraging development, some internally focused (evaluating mission statements and procedures), some externally focused (long client surveys about how use their loans).

Here in La Paz, I am working on two things: measuring the poverty levels of clients and getting feedback from clients about what’s working for them and what’s not. To accomplish the first goal, I am helping FONCRESOL to implement a tool called the Progress out of Poverty Index. This tool was developed by the Grameen Bank and I think it’s really interesting and useful.

Basically, the tool involves giving clients a poverty score. This is done using a short survey which is tailored specifically for different countries. The questions, such as “What material is your floor made of?” or “What type of fuel do you use for cooking?” are multiple choice and responses are easily verifiable by staff (which ensures the integrity of the data). Points are assigned for each answer and the tallied score can be translated into a likelihood that the client lives under a certain poverty-level (for all my nerdy econ friends, the scores are calibrated to various poverty lines using national household survey data and a logit regression). When the tool is administered over a population, the percentage of that population living under a certain poverty-level can be estimated. Cool, huh?

While this information is pretty basic, it’s also pretty useful. The goal is to integrate the tool as part of the loan application process so that data is gathered for every loan. With this information one can observe a couple important things:

1) How poor is the average client? How poor is the average new client? This is important information to make sure the services are reaching the intended population. Lending to (relatively) rich people doesn’t do much good if your goal is to reach the poor.

2) Does the client’s poverty level change after each loan cycle? This is where one might be able to crudely measure social impact. By tracking changes in the poverty level of clients over time, this is one way to observe impact. (I’m going to pre-empt my nerdy friends by admitting there is an identification problem here. Clearly, without a measurement of the counterfactual, causality cannot be determined. However, I’d argue that this information is still valuable to management, and given the small cost of implementation, is a great first step).

The other thing I’m working on is developing a program of client satisfaction focus groups. In addition to finding out what’s working and what’s not (from the perspective of the clients), this tool will also increase the amount of client input when it comes to program design and management decisions (which I believe is valuable in and of itself). I don’t really have much more to write about that (it doesn’t exist yet), but this post is already really long and conducting these focus groups in the different regions that FONCRESOL serves (five in total) should make for some good blog material in the future.

So that’s what I’m doing.