Persuading an audience: Testing information design in the laboratory

Governments, central banks, and private organizations frequently face the challenge of convincing their audience to take a specific action. One key choice is whether to send a public message that can coordinate the audience's actions or to rely instead on private messages that may differ across audience members and thereby miscoordinate actions. This paper uses a laboratory experiment to test whether public or private messages are more persuasive and how this depends on the audience's strategic environment. In the experiment, public signals are most persuasive. The results match the theoretical prediction that public persuasion works best when the receivers' strategic environment features strategic complements. However, contrary to theory, public signals are equally persuasive as private ones under strategic substitutes. Senders respond to this pattern by engaging more frequently in public communication, especially when the receivers' environment features strategic complements.