Pure Agent-based Computational Economics of Time, Knowledge and Procedural Rationality with an Application to Environmental Economics
The mainstream of economic research follows the paradigm of neoclassical economics with its strong emphasis on equilibrium. Since this approach depends upon analytical solutions (existence-proof), it cannot represent time as a (historical) process. Unfortunately, it is thus limited to substantive rationality (maximisation principle) and does not allow for an explicit knowledge representation. The computer simulation technique of agent-based modelling allows for these properties and embraces complexity. By the name of pure agent-based computational economics (pACE), such an approach to economics with a strong emphasis of time, knowledge and procedural rationality is described. The potential of pACE is highlighted with a first modelling exercise: the bakery model. The analysis shows that the neoclassical competitive equilibrium can be an emergent result of pACE, but as a special case only. Furthermore, in an extended version of the same model, the classical environmental economics’ question of “prices vs quantities” (taxes vs. quotas) is analysed. The duality known from comparative static analysis does not hold when it comes to the analysis of the adjustment process, and in the “prices” setting it becomes important who bears the tax. The adjustment process under an effluent charge differs from that under an equivalent sales tax, and both differ again from that under a quotas regime. As a side-effect of the research conducted, some new methods for the analysis of ACE models are introduced.
Nutzung und Vervielfältigung:
Alle Rechte vorbehalten