New macroeconomic modeling approaches: Hierarchical dynamics and mean-field approximation

Abstract

This paper proposes hierarchical structures and mean-field approximation schemes as two potentially useful ways of modeling collective behavior in macroeconomics. Economics systems with hierarchically organized state spaces or agents can exhibit the sluggish response patterns typical of economic time series with unit or near-unit roots. We describe how such hierarchies may arise in several ways, and how dynamics can exhibit behavior with multiple time scales. Then, we describe the use of the mean-field approximation method to model a system composed of many (heterogeneous) agents subject to nonlocalized externalities, and give two brief examples.