Welcome to the new version of CaltechAUTHORS. Login is currently restricted to library staff. If you notice any issues, please email coda@library.caltech.edu
Published March 4, 2016 | Accepted Version
Report Open

Ignorance and Bias in Collective Decisions

Abstract

We study theoretically and experimentally a committee with common interests. Committee members do not know which of two alternatives is the best, but each member can acquire privately a costly signal before casting a vote under either majority or unanimity rule. In the experiment, as predicted by Bayesian equilibrium, voters are more likely to acquire information under majority rule, and attempt to counter the bias in favor of one alternative under unanimity rule. As opposed to Bayesian equilibrium predictions, however, many committee members vote when uninformed. Moreover, uninformed voting is strongly associated with a lower propensity to acquire information. We show that an equilibrium model of subjective prior beliefs can account for both these phenomena, and provides a good overall fit to the observed patterns of behavior both in terms of rational ignorance and biases.

Additional Information

March 17, 2014. The authors thank CONACYT (grant 61501) and Asociación Mexicana de Cultura for financial support. Palfrey is grateful for financial support from NSF (SES-0962802) and the Gordon and Betty Moore Foundation (GBMF#1158). The paper benefited from comments and suggestions from audiences at the 2012 Latin American Meetings of the Econometric Society at University of the Pacific, Lima, the University of Technology, Sydney, the Department of Industrial Engineering at University of Chile, George Mason University, and the 2014 Santa Barbara Conference on Experimental and Behavioral Economics. Kirill Pogorelskiy provided excellent research assistance.

Attached Files

Accepted Version - SSWP1386.pdf

Files

SSWP1386.pdf
Files (811.9 kB)
Name Size Download all
md5:5bacbc0a2166ce0460cfaab03c8bfc9e
811.9 kB Preview Download

Additional details

Created:
August 20, 2023
Modified:
January 13, 2024