To experiment or not to experiment in tax policy?

Abstract

Despite the increasing use of experiments in policy-making the suitability of experimenting in the public sector context is still under debate. In this article, we focus on experimenting in the field of tax policy and ask: what are the promises and pitfalls of using experimental approaches in tax policy? While the existing discussions on tax policy experimentation focus on randomized controlled trials from a legal perspective, we adopt a broader view and provide a more comprehensive discussion by synthesizing insights from the fields of political science, public policy, public administration, and governance. Our analysis encompasses randomized controlled trials, non-randomized policy pilots and design experiments. We summarize the existing knowledge on using experimental approaches in policy-making and discuss the implications of the knowledge for experimenting in tax policy. We seek to offer a more holistic and critical take on whether we should promote the use of experimental approaches in this domain.