The need for impact evaluation in electricity access research. (February 2020)

• Record Type: Journal Article
• Authors: Bayer, Patrick; Kennedy, Ryan; Yang, Joonseok; Urpelainen, Johannes
• Is Part Of: Energy policy. Volume 137(2020)
• Publication Date: 2020-02
• Subjects: Impact evaluation -- Electricity access -- Observational and experimental methods -- Causal inference -- Sustainable development; Energy policy -- Periodicals; Politique énergétique -- Périodiques; Electronic journals; 333.79
• DOI: 10.1016/j.enpol.2019.111099
• Languages: English
• ISSNs: 0301-4215

Abstract

Abstract: Universal household electrification is a key component of the United Nations Sustainable Development Goals, but the evidence base for social and economic impacts of electricity access remains unclear. Here we report results from a systematic review of impact evaluations of household electrification based on five key outcome measures. We only find 31 studies that conduct statistical hypothesis tests to assess impacts. Among these, seven draw on a randomized experiment designed for causal inference. The randomized experimental studies generate fewer positive results than observational or quasi-experimental studies, such as correlational, instrumental variable, and difference-in-differences designs. These results call for a reassessment of what we know about the impacts of household electrification. They also call for major investment in impact evaluation of electricity access using randomized controlled trials, with a particular focus on when and how energy access interventions can furnish large benefits to their intended beneficiaries. Large-scale impact evaluations using experimental methods will require close collaboration between policymakers and researchers. Highlights: We use a systematic review to identify rigorous impact evaluations of electricity access. We find only 31 studies that conduct statistical hypothesis tests to assess impacts. Experimental studies produce fewer positive impacts than observational studies. This result holds for different technologies, different outcomes, and inside/outside Africa. Large-scale impact evaluations require collaboration between policymakers and researchers.