So maybe the banks need to set a higher threshold on who they loan to. This doesn't say much about the effectiveness of microcredit for well-qualified applicants.
Excerpt: "The lender charged its normal rate: 200% APR. The remaining, just-below-the-normal-approval-bar applicants (the "control group") were rejected in line with the lender's normal credit policy. We then tracked both groups over the next six to 27 months, measuring their well-being based on a range of economic, social, health and mental health measures. Applicants who were randomly approved for a loan had higher incomes, less hunger, better credit scores and more positive outlooks than their control group counterparts -- even after paying the high interest rate. Though they had higher than normal default rates, the borderline loans were also profitable for the lender."