How do groups speak and how are they understood?

Abstract

We experimentally study an environment where a group of senders communicates with a receiver by disclosing or not disclosing a realized outcome. Group members have distinct pref[1]erences over disclosure/non-disclosure, and aggregate their interests into a collective disclosure decision via a given deliberation procedure. In line with theoretical results, our experimental evidence establishes a relationship between the procedure used by the group and the receiver’s interpretation of the group’s “no disclosure messages:” group members who have more power over the group’s disclosure decision are regarded with more skepticism when the group fails to disclose. We further document that in a group disclosure setting, the observer is typically not as skeptical about group members’ values upon seeing no disclosure, relative to theoretical predictions; and that the interpretation of communication from a group differs from that of in[1]dividual communication, even when the two are theoretically equivalent. We argue that these observations are consistent with group members having social preferences; and contrast them with previous literature on the “romance of leadership.”