Experimental and non-experimental methods to study government performance: contributions and limits
In recent years, field experiments using randomized trials have gained increasing popularity in the field of development economics. In particular, scholars have argued strongly for their use as the best means of identifying ‘what works’ in foreign aid. Field experiments have also become increasingly popular in political science, particularly in the study of democracy and elections. The Symposium brings together leading economists and political scientists to discuss the use of field experiments in the study of governance, and their limitations with regards to ‘what works’, ‘what is scalable’, and ‘what is transferable’ in development policy. Project contributors employ new and existing data to analyse precisely (1) what is gained and lost through the use of randomized controlled trials versus other approaches (e.g., quasi-experimental designs, survey techniques) and (2) the possibilities of employing new techniques in the design and analysis of randomized experimental trials to improve the external validity, i.e. their ability to speak beyond the narrow context of the experiment.