In last few years, the randomized program evaluation has become the gold basis for evaluating the development programs and also the bread and butter of many development economists. The evaluations sometimes uncover the valuable new information, but are contentious, and can also be prohibitively expensive to implement, for small NGO’s. Appreciating how the microfinance affects clients is not straightforward for the reason that there are several possible explanations for why, say, a borrower is doing well as compared to her non-borrowing peers. The credit may be helping–or conceivably the borrower was already comparatively prosperous and would have fared better even without the loan. These new papers clarify cause and effect by performing controlled experiments, in which a few parameters are arbitrarily varied and the effects measured. The idea of ‘experiments’ sounds good because it gives scientific legitimacy to what are in actual fact social phenomena marked by complex correlations, multiple variables, and extraneous and often invisible factors. Let us look at these reports to see just how ‘controlled’ they are.