Universität Bielefeld

Institut für

Computational Economics

Our main interests lie in the following subfields of Computational Economics: CE4

Agent-based Modeling

Agent-based models of economic systems have in recent years attracted increasing attention as a tool to rigorously analyze properties of economic dynamics emerging from (local) interaction of bounded rational individuals. The approach allows to capture dynamics on different time scales without relying on assumptions of instantaneous market clearing, rational expectations or representative agents. Bielefeld University has for years been on the fore-front of the development of agent-based models with a clear conceptual basis and empirical grounding. Using the Eurace@Unibi model a range of economic policy issues have been examined including the effects the spatial distribution of human capital policies, implications of different degrees of spatial labor market frictions on convergence between regions and the effectiveness of different types of cohesion policy measures employed by the EU.

Numerical Methods in Dynamic Games

Many relevant issues in Industrial Economics (as well as in many other areas of economics) are characterized by a combination of strategic and inter-temporal effects that decision makers have to take into account. Differential Game Models are a suitable tool for gaining insights into such problems, where typically Markov-Perfect Equilibria are the most convincing solution concept to be applied to such models. Apart from differential games with rather special structure the characterization properties und induced dynamics of such equilibria by analytical means is however hardly possible. Hence, numerical approaches, like collocation or homotopy methods, are needed. However, a straight forward application of such methods to games with large state spaces or non-standard structures is typically not straight forward and there exists little general guidance which type of methods is most suitable for which type of games. In the framework of the DFG-funded project 'Dynamic Oligopolistic Competition between Innovating Firms' numerical methods for the analysis of a class of oligopoly games with evolving market structure, which from a technical perspective are multi-mode differential games, are developed.