Suchergebnisse

1. Non-equilibrium dynamics in the evolution of industries -- 2. Models of industry dynamics -- 3. A dynamic model of Schumpeterian competition -- 4. Growing an industry in silico -- 5. Shakeouts : limited foresight, technological shocks, and transient industry dynamics -- 6. Industry dynamics in the steady state : between-industry variations -- 7. Firm dynamics in the steady state : within-industry variations -- 8. Cyclical industrial dynamics with fluctuating demand -- 9. Conclusion.

The dynamic choice between individual and social learning is explored for a population of autonomous agents whose objective is to find solutions to a stream of related problems. The probability that an agent is in the individual learning mode, as opposed to the social learning mode, evolves over time through reinforcement learning. Furthermore, the communication network of an agent is also endogenous. Our main finding is that when agents are sufficiently effective at social learning, structure emerges in the form of specialization. Some agents focus on coming up with new ideas while the remainder of the population focuses on imitating worthwhile ideas.

To explore the efficacy of a corporate leniency program, a Markov process is constructed which models the stochastic formation and demise of cartels. Cartels are born when given the opportunity and market conditions are right, while cartels die because of internal collapse or they are caught and convicted by the antitrust authority. The likelihood that a cartel, once identified, is convicted depends inversely on the caseload of the antitrust authority due to an implicit resource constraint. The authority also chooses an enforcement policy in terms of the fraction of non-leniency cases that it prosecutes. Using numerical analysis, the impact of a leniency program on the steady-state cartel rate is investigated. Holding the enforcement policy of the antitrust authority fixed, a leniency program lowers the frequency of cartels. However, the additional caseload provided by the leniency program induces the antitrust authority to prosecute a smaller fraction of cartel cases identified outside of the program. Because of this less aggressive enforcement policy, it is possible that the cartel rate is higher when there is a leniency program.

An organization is a collection of agents that interact and produce some form of output. Formal organizations - such as corporations and governments - are typically constructed for an explicit purpose though this purpose needn't be shared by all organizational members. An entrepreneur who creates a firm may do so in order to generate personal wealth but the worker she hires may have very different goals. As opposed to more amorphous collections of agents such as friendship networks and societies at large, organizations have a formal structure to them (though informal structures typically emerge as well) with the prototypical example being a corporation's organizational chart. This structure serves to define lines of communication and the distribution of decision-making. Organizations are also distinguished by their well-defined boundaries as reflected in a clear delineation as to who is and who is not a member. This boundary serves to make organizations a natural unit of selection; for example, corporations are formed and liquidated though they can also morph into something different through activities like mergers.

Scientific progress is driven by innovation—which serves to produce a diversity of ideas—and imitation through a social network—which serves to diffuse these ideas. In this paper, we develop an agent-based computational model of this process, in which the agents in the population are heterogeneous in their abilities to innovate and imitate. The model incorporates three primary forces: the discovery of new ideas, the observation and adoption of these ideas, and the endogenous development of networks. The objective is to explore the evolving architecture of problem-solving networks and the critical roles that different agents play in the process. A central finding is that the emergent network takes a chain structure with innovators (those most skilled at generating new ideas) being the main source of ideas and those most skilled at imitating acting as connectors between the innovators and the masses. The impact of agent heterogeneity and environmental volatility on the network architecture is also characterized.