Construct Validity and the Experimental Microfinance Literature
There is an ongoing debate on the impact of microfinance. While the first randomized controlled trials find either underwhelming or weakly positive evidence, the more recent experimental work on microfinance shows that tweaks of the standard model may lead to different conclusions.
As with many interventions, the design space of a microfinance program is highly dimensional. Not only the different components of the intervention, but also who implemented it, may matter for the observed impact. In this study, we will map the design space of a microfinance intervention and look into how robust the results are across variations of the design space.
The aim of our study is twofold. First, we will provide a comprehensive overview of the microfinance literature by conducting a deep literature review and an expert survey with the authors of these studies. Second, using the microfinance literature as an example, we will demonstrate the importance of construct validity, and how variations of the design space may influence the results obtained. This study fits within the broader debate about how we use evidence generated by RCTs to draw conclusions about what works in a real-world setting.
Currently there are no publications available for this project
Prof. Dr. Nathan Fiala, Prof. Dr. Jörg Peters, Lise Masselus
University of Connecticut