The insight of pay-as-you-go schemes as a response to variable incomes is a great one.  Its applicability to things like mobile phone minutes is obvious, and I’m very curious as to how it could be applied to financial services for the poor.  The big stumbling blocks appear to be the questions of credibility and enforceability.  Or, rather, the fact that enforceability tends to engender credibility – banks believe their clients when they say they’ll pay, as they can easily check up on their payments and enforce them if they fall behind.  A pure pay-as-you-go plan in microcredit – say, a three-month loan that can be repaid in any amount at any time before the three-month period is up – would be unenforceable over the period of payment, as the bank couldn’t demand payment at any given time before the deadline.

Actually, that’s a separate (and interesting) question – how different types of enforcement may affect the repayment of loans.  Perhaps you could look at peer pressure in lending groups, vs. occasional visits from a loan officer, vs. frequent visits from a loan officer, vs. financial incentives for timely payment, or something like that.  That might shed some more light on designing systems of incentives for ultimate on-time payment of a pay-as-you-go loan.