InterJournal Complex Systems, 794
Status: Accepted
Manuscript Number: [794]
Submission Date: 2004
A Dynamic Model of Inter-Firm Competition
Author(s): Duncan A. Robertson

Subject(s): CX.4



Complexity and agent-based models have recently been introduced to strategy research, and management science more generally. However, these models and frameworks have not realized the promise that they initially commanded. This is partially due to the fact that there are no simple and accessible models in management science that incorporate the benefits of considering inter-firm competition as complex systems. One of the most promising models to be introduced into management science is the NK model, pioneered by Kauffman (in a biological setting), and transferred to organization science by Levinthal. The NK model produces a complex ‘fitness landscape’ which firms can explore and thereby attempt to maximize their fitness. Whilst this model offers an enticing introduction to complex models, there may be problems when considering the application of the model to management science. Several comments and criticisms are offered in this paper, including the fact that the NK model offers a static representation of a firm’s decision set, and therefore is of little use in turbulent or high-velocity environments. The concept of landscape models is extended in the model presented in this paper. However, instead of basing the model on the NK framework, a competitive landscape model is produced from the interaction of firm and customer agents co-existing in a product space. (This is extended to more general firm – ‘value-generating agent’ interactions.) From this simple model, a complex landscape is generated. This has the advantage over the NK model in that it can be represented as an actual landscape rather than relying on a schematic representation. The model produced by this interaction is dynamic, in that the movement of firms or customers produces a change in the landscape. Furthermore, the landscape in which a firm operates is different for each firm. This leads to true inter-firm heterogeneity, rather than the homogeneous firms intra-group as in the strategic groups literature or that of economic models more generally. Easily representable 'Profit Lanscapes' are generated from the model, rather than the schematic representations as in the NK model. As firms and value-generating agents move in this space, the landscape deforms and produces a dynamic, changing landscape. Simple strategies such as ‘hill climbing’ algorithms as suggested by agent-based modellers (such as Rivkin / Rivkin and Siggelkow) may not be appropriate under this model, whereas they may be appropriate when exploring a static NK landscape. Entry decisions, the effect of movement of competitors, and optimal strategies for such landscapes are considered. Extensions to the model, for example customers with different value are suggested as extensions to the model. Agent-based modelling can be used to explore the landscape introduced in this model, and is introduced as a further extension to this research. Overall, the model introduced in this paper provides a dynamic model of inter-firm competition, extends the extant models, and provides a basis for further research for revitalizing the positioning literature.

Retrieve Manuscript
Submit referee report/comment

Public Comments: