New type of modelling to guide policy development
13 Jul 2011
After the failure of traditional economic models to predict the GFC, many organisations, including the US government, have turned to a new type of modelling as a way to understand where things went wrong and guide new policy.
Agent-Based modelling is an emerging computational method increasingly employed by the world's top scientists which focuses on the interaction of complex systems adopted within many sciences including business and social sciences.
Simulation programs such as SIMlife, The SIMS, SIMPark and Second Life are all based on ABM. The same methods can be used to create models of the real world and offer ways to better understand the economic environment and inform policy and management.
The growing interest in ABM can be seen in the award of the Nobel Prizes in Economics to Elinor Ostrom (2009), Tjomas Schelling (2005) and Vernon Smith (2002); all for work in complex systems.
Examples of how this modelling is being used can be found in research by leading biologists who are using ABM to investigate resource scarcity and climate change and leading economists who are investigating why poor decisions were made in the lead up to the GFC.
In recognition of this emerging research tool, The University of Sydney Business School this month hosted the first intensive three week course on Agent-Based modelling outside the US.
The simulation course will be led by Agent-Based modelling expert Professor David Earnest from Old Dominion University in Virginia, and is based on a similar course he offers there. The course combines lectures with hands on experience in lab sessions building familiar with theoretical and methodological foundations of ABM. Participants will be able to implement their own simulations in the programming language NetLogo to support their research.
Course convenor, Professor Ian Wilkinson from the Marketing Discipline at the University of Sydney Business School says: "We are in the midst of a sea change in research methodology that complements existing methods - it does not replace them - but we are also facing major resistance from entrenched interests and mainstream academics."
"Agent-Based Modelling role in future research is exciting and the course we are presenting is a terrific way to introduce Australian academics to this innovative new way of modelling."
An example of Agent-Based Modelling
David Earnest is using network data on interbank transactions in the US and Europe. The study uses an ABM system to simulate negotiations between the US and Germany modelling the negotiations as a "Battle of the sexes" game of mixed cooperative and competition incentives. The model illustrates how a small number of centrally located banks in the German network could shape the aggregation of social preferences to reflect their preferred outcomes. The study contributes to our understanding of international regimes by illustrating how market-driven networks among firms create relationships. This suggests that the methods of social network analysis can help explain the origins if state preferences and the prospect of cooperation among states.
Professor David C Earnest is an Associate Professor of political science and international studies at Old Dominion University Norfolk, Virginia USA. His substantive research focuses on the political incorporation of migrants in democratic societies.