program in Adaptive Economic Dynamics
The course consists of 3 hours
of formal lectures per day for two weeks, supplemented by tutorials and
laboratory instructions in the use of SWARM and Mathematica
for the computable modelling of economic processes, agents and institutions.
This is the first of a series of intensive courses to be offered by the Computable and Experimental Economics Laboratory (CEEL).
Course in Computable Economics
The main theoretical applications will be to finitely repeated games of a special class, rationality, learning, production, decentralization and dynamics. The novice will be introduced to the art of introducing computability assumptions and the science of devising recursion theoretic proofs. Where possible comparisons and contrasts will be made with the mathematical formalisms and techniques of standard economic theory. Thus, for example, instead of emphasising equilibrium existence proofs using non-constructive methods, in computable economics an attempt is made to highlight the nature of the processes that may lead to equilibrium configurations.
The formal introduction of computability constraints makes it almost natural to replace optimisation paradigms by decision problems and the concomitant technicalities: computational, diophantine and algorithmic complexities of algorithms. It will be shown how these aspects enables the computable economist to formalize, in a perfectly sensible, non-suboptimal sense, notions such as bounded rationality and satisficing.
The students will also be introduced to the proof techniques, methods and concepts of recursion theory: diagonalization, reduction to diophantine decision problems, the incompressibility method, the use of Rice's theorem, the Halting Problem, computation universality, Church's Thesis, the Busy-Beaver problem, recursive enumerability vs. recursiveness, undecidability and uncomputability. It is hoped that students will, at the end of the course, have considerable facility in manipulating such techniques, methods and concepts to construct their own models of particular economic problems for theoretical and empirical investigations. Thesis topics with such aims in mind will be suggested for the participants and their adoption will be encouraged by a promise to monitor subsequent progress by competent supervisors.
(*An outline of the nature and scope of the field can be gleaned from Computable Economics by Velupillai (Oxford University Press, Oxford, UK; January 2000) ).