The Use of Identity Primes to Explain Behavioral Differences between Groups : A Methodological Note

Economists are increasingly using primes that make group identity salient to overcome the inferential limitations of behavioral quasi-experiments involving pre-existing groups (e.g., males vs. females). However, while priming group identity provides powerful means for identifying a causal effect of group membership on individuals' preferences, so far, there has been little methodological reflection on the use of identity primes to identify the causes of group differences in preferences. This note's main contribution is to offer a framework for thinking systematically about the treatment effects of priming individuals' group identity and the identification of specific group traits explaining differences in preferences or behavior between pre-existing groups. The framework sets a high bar for studying the causes of group differences in preferences using identity primes but we clarify its usefulness using concrete examples