Stylized Facts and Experimentation

In this comment, we clarify and extend Hirschman’s (2016) discussion of 'stylized facts.' Our focus is on the relationship between stylized facts and experimentation. Given the continued increase in experimentation across the social sciences, we think that it is important to consider the exact role that experiments play in the production and testing of stylized facts. We make three related contributions here. First, we describe how experiments can both provide new evidence to support existing stylized facts as well as produce new stylized facts. Second, we argue that the stylized facts produced through experimentation differ from other stylized facts. Third, we extend Hirschman's (2016) definition of “stylized facts” so that it distinguishes between those that describe correlation relationships and those that describe causal relationships.