Logarithmic Preferences, Myopic Decisions, and Incomplete Information

Abstract

This paper examines a dynamic production economy with incomplete information and shows that the set of myopic preferences, those that induce myopic decisions, depends on the representation of the information flow. For example, logarithmic preferences are nonmyopic when some of the economic state variables are unobservable. The analysis offers a broader definition of myopic behavior, termed “generalized myopia,” which is independent of the representation of the information flow. Allowing for any smooth concave utility function, logarithmic preferences endogenously emerge as necessary for generalized myopia in incomplete information economies; and when combined with restrictions on the information structure, they become sufficient.