Over the last 15 years agent-based models have also come to develop into an alternative or complementary approach to the modelling of complex systems in the sphere of economics. However, there is so far – except in the case of more specialised applications in the sectors transport and energy - little experience in applying this type of model to economic policy advice.

The event emphasises two central objectives. They consist of contributed papers aiming at answering the questions "What can be achieved by agent-based models in the field of economic policy advice?" and "What changes are necessary in both the methods and the communication of results to develop this model into a standard tool for applied economic research?". In addition to the issue of applying the models there is the methodical key question of what methods may be introduced to validate the models.

The idea of mapping whole economies using agent-based models has only taken on a concrete shape over the last years, when the required computing capacities were available. Two such approaches will be jointly presented and discussed in the Workshop for the first time. Both the modelling itself and the application to economic policy advice in turn will be up for discussion.

The question of how robust the achieved results are, i.e. what influence the choice of the parameter has, arises for both applications. Typically, agent-based models show a high number of parameters, which generally have a non-linear influence on the results. This gives the aspect of validating the models more importance, particularly regarding their application in policy counselling. Several approaches to this issue will be discussed in the course of the Workshop


24.03.2009 - 25.03.2009


ZEW, L 7,1 D-68161 Mannheim


Senior Researcher