Complexity Theory and Organizations

Fundamentals

The complexity theory in management is largely based on the Kauffman NK model. Stuart Kauffman, a biologist, designed the model to study how interactions between genes affect the fitness of a species. Within the context of organizations, the model has been used to explain how the interactions among decisions within and across organizations affect organizational performance. The effect of the interactions on organizational performance has been studied within the context of various organizational structures, incentive systems, learning processes, technological regimes, industry characteristics, and environmental dynamics.

The key construct of the model is the notion of interdependence. Interdependence between two decisions exists when one decision influences not only its own performance contribution but also the performance contribution of another decision. The overall organizational performance is assumed to be a function of the performance contributions of all decisions that the organization makes. The organizational performance is conceptualized either as the organizational adaptation or as the ability of the organization to solve a given problem. Superior performing organizations are those that achieve a better fit with the external environment or discover a better solution to a problem. The organizational performance is emergent in the sense that it cannot be deduced from the analysis of each organizational unit in isolation but rather depends on the interactions within the system as a whole. The model is most relevant, and its predictions are most likely to hold, in contexts where the outcomes are driven by the interactions among the decisions as opposed to being dominated by individual decisions.

The model assumes that decision makers within organizational units have bounded rationality. Bounded rationality implies that the decision makers are unable to select the best possible set of decisions but must proceed through an iterative search. An iterative search consists of trial-and-error steps. After each step, the decision makers change a limited number of decisions and observe whether the changes lead to an increase in performance. Typically, only performance-enhancing choices are retained.

The key relationship predicted by the model is that, due to bounded rationality, an increasing density of interdependencies complicates the search of the decision makers. With few interdependencies, changes in a small number of decisions have a small impact on the overall organizational performance. When the interdependencies are dense, however, even changing a small number of decisions can have a dramatic effect (positive or negative) on the overall organizational performance as the focal decision may affect the performance of many other decisions. Limited in their ability to consider a wide range of decisions, the boundedly rational decision makers tend to settle on less than optimal outcomes when facing interdependent choices. A higher density of interdependencies, thus, potentially leads to lower organizational performance. The research has examined a variety of factors that interact with this relationship and could potentially allow the organization to achieve a higher organizational performance. These factors include organizational centralization versus decentralization, differences in the cognitive mechanisms and imitative abilities, and technological modularity and environmental turbulence. The NK model has been recently extended to study the effect of interdependencies across organizational units and entire firms.

...