The interview with Dean Karlan is very good. The research paper on impact measurements can be found here. Over at the BBC (via Innovations for Poverty Action):
Academics have been trying to work out from the evidence whether microcredit does actually raise people's incomes.
But it's been hard to do a proper scientific survey, since you need to compare those who do get a loan with a control group of similar people who don't.
Dean Karlan, professor of economics at Yale University, has managed to do it – with a control group – in the Philippines. His results raise some serious questions about the effectiveness of microcredit in reducing poverty.