Formation, Size, and Stability of Coalitions within Organizations

Abstract

This paper investigates the link between coalitions, i.e. informal organizational structures, and individual learning in organizations. Building on the canonical Garbage Can Model (Cohen et al., 1972), we develop an agent-based model that allows us to analyze how organizational members’ learning about others’ preferences enables the emergence of coalitions. Altering the concentration (extremeness) of individuals’ preferences, we find differences in the precision of learning. We observe how large and stable coalitions get as a result of this learning process and how the presence of this informal structure can benefit the overall organization. Subsequently, we find that the organizational benefit of coalitions is greatest for low concentration setups. In exploiting this interplay, this paper also takes a first stab at integrating two hitherto disjointed lines of thinking on learning and bargaining.