Noisytalk.com: Broadcasting Opinions in a Noisy Environment

Stanford GSB Working Paper

43 Pages Posted: 8 Jan 2001  

Anat R. Admati

Stanford Graduate School of Business

Paul C. Pfleiderer

Stanford Graduate School of Business

Date Written: December 2000

Abstract

This paper is motivated by the vast quantity of public broadcasting that occurs on the internet, e.g., customer reviews of products on various sites. We analyze a model of information transmission in which a sender attempts to communicate information to a population of receivers using a finite number of messages. Noise in this process arises from the possibility that the sender is uninformed. Receivers are rational in interpreting the message they observe; the sender is either rational or overconfident concerning the informativeness of his signal. We show that if the sender is rational, there is an equilibrium outcome that is informationally efficient, but this equilibrium may be neither unique nor symmetric. This is true even when the underlying distributions are symmetric. If the sender is overconfident, he tends to exaggerate, and equilibrium involves extreme messages being sent with a relatively high likelihood. Overconfidence generally reduces the number of equilibria, and this might lead to an improvement in the informational efficiency of the equilibrium. We analyze the model with costless as well as with costly broadcasting, and show that with costly broadcasting equilibrium may involve a subset of the messages that are not used in equilibrium. Our results are generally consistent with the casual empirical observations such as that the customer ratings on the internet have a highly asymmetric distribution, with a preponderance of extreme messages, and with some available messages essentially unused.

Keywords: Asymmetric Information, Internet, Efficiency of Communication, Overconfidence

JEL Classification: D00, D80

Suggested Citation

Admati, Anat R. and Pfleiderer, Paul C., Noisytalk.com: Broadcasting Opinions in a Noisy Environment (December 2000). Stanford GSB Working Paper. Available at SSRN: https://ssrn.com/abstract=255318 or http://dx.doi.org/10.2139/ssrn.255318

Anat R. Admati

Stanford Graduate School of Business ( email )

655 Knight Way
Stanford, CA 94305-5015
United States
650-723-4987 (Phone)
650-725-6152 (Fax)

Paul C. Pfleiderer (Contact Author)

Stanford Graduate School of Business ( email )

655 Knight Way
Stanford, CA 94305-5015
United States
650-723-4495 (Phone)
650-725-7979 (Fax)

Paper statistics

Downloads
272
Rank
88,718
Abstract Views
1,978