Nonmarket Interactions

Abstract

A large body of recent research argues that social, or non-market, interactions can explain a wide range of puzzling phenomena from fashion cycles to stock market crashes. This paper attempts to connect the range of these papers with a general model and a broad empirical overview. We establish conditions for existence and uniqueness of equilibria in social interactions models. The existence of multiple equilibria requires sufficient non-linearity in social interactions and only moderate heterogeneity across agents--strategic complementarities are neither necessary nor sufficient for multiple equilibria. We establish conditions for the existence of a social multiplier, which is the ratio of the aggregate outcome-input relationship to the individual outcome-input relationship. Models with multiple equilibria are empirically indistinguishable from models with significant social multipliers. Finally, we show the formal relationship between three known methods of empirically estimating social interactions, and suggests the plusses and minuses of these three approaches.