Randomizing the "Last Mile": A methodological note on using a voucher-based approach to assess the impact of infrastructure projects

This methodological note discusses the potential and limits of using voucher-based experiments to randomly evaluate the micro-level impact of infrastructures on households' well-being. We argue that such methods are policy relevant, statistically robust, and ethically correct. A number of conditions regarding the vouchers' design and level, as well as allocation methods and household sampling, must be taken into account, however. We illustrate the discussion with an ongoing voucher-based impact evaluation of a rural electrification program in Ethiopia.