Neighborhood-Based Information Costs

Working Paper: NBER ID: w26743

Authors: Benjamin M. Hbert; Michael Woodford

Abstract: We derive a new cost of information in rational inattention problems, the neighborhood-based cost functions, starting from the observation that many settings involve exogenous states with a topological structure. These cost functions are uniformly posterior-separable and capture notions of perceptual distance. This second property ensures that neighborhood-based costs, unlike mutual information, make accurate predictions about behavior in perceptual experiments. We compare the implications of our neighborhood-based cost functions with those of the mutual information in a series of applications: perceptual judgments; the general environment of binary choice; regime-change games; and linear-quadratic-Gaussian settings.

Keywords: Information Costs; Rational Inattention; Perceptual Judgments; Neighborhood-Based Cost Functions

JEL Codes: D83; G41


Causal Claims Network Graph

Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.


Causal Claims

CauseEffect
neighborhood-based costs (R20)behavior in perceptual experiments (C99)
mutual information (D83)behavior in perceptual experiments (C99)
neighborhood-based costs (R20)continuous response frequencies in decision-making (D91)
mutual information (D83)discrete jumps in decision-making (D91)
neighborhood-based costs (R20)unique equilibrium in regime-change games (C73)
nature of information costs (D83)decision-making process in regime-change games (D70)
neighborhood-based costs (R20)efficiency in games with rationally inattentive agents (C73)
neighborhood-based costs (R20)volatility in games with rationally inattentive agents (C73)

Back to index