On the Efficiency of Social Learning
41 Pages Posted: 8 Dec 2017 Last revised: 12 Jan 2018
Date Written: November 9, 2017
Abstract
We revisit well-known models of learning in which a sequence of agents make a binary decision on the basis of a private signal and additional information. We introduce efficiency measures, aimed at capturing the speed of learning in such contexts. Whatever the distribution of private signals, we show that the learning efficiency is the same, whether each agent observes the entire sequence of earlier decisions, or only the previous decision. We provide a simple necessary and sufficient condition on the signal distributions under which learning is efficient. This condition fails to hold in many prominent cases of interest. Extensions are discussed.
JEL Classification: D83, D83
Suggested Citation: Suggested Citation