In: Economics
No this is not a causal result as the experiment design has certain flaws. Although the randomization done between the person who gets the loan and who does not is desirable, the experiment does not control for any other factor that might lead to a difference in earnings. There has to be a formal estimation of the regression equation controlling for the education, gender, past employment, assets ownership, parental background etc. that might have themselves lead to a difference in total earnings. All these factors will on average cancel out of the sample size was large (if say 1000 people applied for the loan and 500 were given loans randomly. But wit only two people, the other factors will be different and they need to be controlled before making any causal statements about the impact of micro loan on earnings.