Stochastic frontiers, productivity effects and development projects

A common objective of many development projects is to promote output growth as well as better management in order to improve incomes and reduce poverty. Technically, the purpose is to induce upwards shifts in the production frontier (i.e., technological change) while also promoting better management (i.e., narrowing the gap from the frontier). Given the link between managerial performance and technical efficiency, stochastic production frontiers are well suited for the task from a methodological point of view. Despite this suitability, work linking stochastic frontiers with impact evaluation methods has just begun and a major hurdle is resolving biases that might arise from selection on observables and unobservables. This article provides an overview of how impact evaluation and stochastic frontiers, two well- established areas in applied econometrics, are being brought together to shed light on the productivity effects of agricultural development interventions.