In order to understand what I'm doing, this summer, let me explain my adventures, a year ago. I had a wonderful opportunity to spend last June and July in Kyrgyzstan and Azerbaijan, respectively, working for the Foundation for International Community Asistance (FINCA). This NGO provides micro-loans to people in poorer countries. To see FINCA's website, go to www.villagebaking.org. Anyway, my job was to interview a representative sample of FINCA clients to gather demographics, business, and family data. The idea is to get a snapshot of the clientele in a country, in order to make some generalized statements, and even predictions as to the effectiveness of FINCA's program.
Upon arrival in Kyrgyzstan, my two partners and I realized that following FINCA's established methodology of conducting individual oral interviews with each client, we would never gather a statistically significant sample size of a country's clientele. We realized that with 99% literacy rates in former Soviet states, we could devise a written questionnaire, and administer the tool en masse. Our results turned out to be very successful, and we were able to gather data for approximately 450 clients in Kyrgyzstan, and 350 in Azerbaijan.
When I arrived home, I began investigating the datasets with assistance of local BYU professors and my FINCA partner, Kris Johnson. We discovered that we could put together moderately robust econometric models to predict family business income as a function of demographics data and the number of loans that a client has received. This is a potentially very interesting finding, because it suggests some quantifiable results of real positive impact of micro-loans on family income. The generally acepted real impact of the micro-finance industry is somewhat ambiguous, at this point, due to a paucity of rigorous quantitative studies. In spite of these encouraging results, it became apparent that our data was incomplete. It represented a snapshot in time, which means that we can only compare clients to each other. This is a problem because clients must be measured using a standard of their own progress, not someone else's. For example, a more mature client might make less money than a less mature client. This does not necessarily suggest that the each progressive loan has caused the mature client's income to decrease, but could rather suggest that the clients each started from different initial income levels, and that both have increased over time. Kris and I realized that in order to develop a truly robust model, we would have to follow these clients over time, and develop a panel-data econometric model (sorry for the jargon...).
We pitched a proposal to the founder of FINCA, John Hatch, when he was visiting Provo, back in March. We suggested that we could return to one or two FINCA offices in the former Soviet Union (hereafter referred to as the NIS, or Newly Independent States), and train country staffs to administer the questionnaire over time. John Hatch enthusiastically endorsed the proposal, but decided to make the project much more ambitious than we initially intended. He suggested that Kris and I travel to ALL of FINCA's offices in the NIS to conduct our tool. Since that time, Kris has left the project to pursue a career in the corporate world, and I'm now on my own. I've subsequently streamlined the project to include only 5 NIS countries. Next week, I will leave on a trip that will include two-week stints in Tajikistan, Armenia, Azerbaijan, Uzbekistan, and Kyrgyzstan. To get a cursory look at each of these countries, go to the following links:
Tajikistan, Armenia, Azerbaijan, Uzbekistan, Kyrgyzstan
I don't have a final departure date, because I'm having some visa troubles. As soon as I have some final information, I'll be sure to let you know. This has gotten a little long, so I'll continue my ramblings, later.