By Beth Rhyne, “Microfinance Miracle or Myth” session moderator
A full summary on the “Microfinance Miracle or Myth” session can be viewed here.
After the workshop on impact, I received the following email from Dean Karlan, one of the researcher who conducts randomized control trials as an impact measurement tool. He writes,
“I ran into someone who attended your talk at the recent conference in San Fran [sic]. She told me that you stated several times that RCT’s cost $1.5 million in average, compared to the “free” market test. I thought you should know that you have that fact wrong. In fact, RCT’s are cheaper than non-experimental quantitative assessments that include non-borrowers (statistical power is improved, as you don’t need to model selection). But anyhow, $1.5mm is wrong. Our Philippine study, e.g., was probably about $100k or $200k. So if you could please correct that, that would be great. As Patrick Moynihan said, we can all have our own set of opinions, but not our own set of facts.”
I would like to note that there was a basis for the $1.5 million figure I cited: it was the price quote I received regarding the RCT taking place at Compartamos. But as with interest rates, maybe Mexico is just a high-cost country, and RCTs in other countries or with less complex designs cost less.



