Two crucial problems when research agencies or donors need to assess empirically the microfinance/children education nexus on already operating organizations are lack of availability of panel data and selection bias. We propose an original approach which tackles these problems by combining...