Changing market structures under changing resource spaces: an agent-based computational approach

PhD ceremony: C.E. García Díaz, 16.15 uur, Academiegebouw, Broerstraat 5, Groningen

Thesis: Changing market structures under changing resource spaces: an agent-based computational approach

Promotor: prof. A. van Witteloostuijn

Faculty: Economics and Business

 

García Diáz’s thesis blends ecological processes and industrial organization assumptions to model specific aspects of industry evolution. It focuses specifically on the role that both consumer heterogeneity and its spatial distribution play in shaping market structures. He uses a computational (agent-based) modelling approach to study evolutionary industry dynamics. This work develops a framework within which ecological processes take place among firms that reveal profit-seeking behaviour. The computational models consider both exogenously and endogenously defined consumer distributions.

The results not only show that the replication of ecological market-level regularities from individual-level behaviour and interaction is possible. They also reflect that the ecological theory of market processes would be complemented if additional aspects were taken into account. Findings reveal both demographic effects (size and firm type) on hazard rates, as well as ecological implications of different consumer distribution shapes. García Díaz shows how specific firm type founding processes and different consumer mobility mechanisms influence the so-called market-partitioning process. Consumer heterogeneity is also explored by studying the effects of changing market dimensionality of the product space. Results indicate that large firms take advantage of product differentiation and price discrimination as dimensionality increases. These advantages are offset by the increasing costs of niche coverage (scope disadvantages) and by market crowding, making the small firms better performers. The cross-fertilization of organizational ecology and industrial organization not only represents a better-integrated framework to study industry evolution, but also opens the door to new developments in computational social science.