W. Brian Arthur explains agent-based modeling.
What if legislators could foresee how people might find loopholes in the bills they’re working on–and head it off? That was one of the themes in a fascinating presentation made by the Santa Fe Institute’s W. Brian Arthur at IBM’s Smarter Health Through Modeling and Simulation conference last week in San Jose, Calif.
Arthur, an external professor at the institute, which focuses on the study of complex systems, said that by using agent-based modeling, researchers could have been able to identify some of the unintended consequences of the partial repeal in 1999 of the Glass-Steagall Act, which had prohibited bank holding companies, whose accounts are insured by the federal government, from owning securities businesses. Some economists have blamed the repeal of the act for worsening the effects of the global financial crisis.
The techniques could have headed off some of the problems that arose with Massachusetts’ health care reforms, as well. One of the problems with the the Massachusetts system, he said, is that too many people signed up for health insurance only when they anticipated they would need it in the not-too-distant future, so they took benefits from the system but didn’t contribute their fair share to the insurance pool. Arthur said agent-based models could have spotted this loophole ahead of time and allowed legislators or the insurance industry to put in place rules that could close it.
In this video clip, Arthur explains the magic of agent-based modeling.