Could this be a solution just for developing countries? And what are the behavioral consequences of the increased control over expenditures given by pay-as-you-go electricity when credit constraints are not binding? An interesting working paper to find out