How System Dynamics Can Help Governments Make Better Decisions
System dynamics was invented a half-century ago to use technology to analyze complex industrial processes. It's got a lot of potential for the public sector.
"Genius," someone once said, "has been defined as the ability to come to correct conclusions on insufficient information; I do not believe that the complex problems of American society can be solved without automation of information collection, storage and retrieval. There simply are not enough geniuses to go around."
That someone was me, testifying before a congressional committee almost exactly 40 years ago. What sparked this trip down memory lane was a recent conference I attended on "system dynamics" (SD), a technology-driven method for understanding the behavior of complex systems that was developed in the 1950s for managers of industrial processes.
SD increasingly is being used in the public sector for policy analysis and design, and the more it catches on in government, the better. America's public-sector problems are increasingly complex, critical and urgent, and SD can help us find the best ways forward.
The goal of SD is to help us make better decisions -- certainly better than those resulting from flipping a coin, guessing or playing "what-if" games. This, of course, is not a new problem. The ancient Greeks started us thinking about thinking. Aristotle identified two types of reasoning: "inductive" and "deductive." Aristotle certainly was a genius, but neither form of reasoning will help us figure out, for instance, what the effects of the 1,200-page Affordable Care Act will be in, say, 2023.
For that we need a very detailed model of the American economy, our health-care system and our changing demographics. We also need a model that incorporates the many feedback loops involved in the complex environment of health care: The federal law requires most people to participate, mandates coverage for pre-existing conditions and limits administrative costs. And all the while the population will be aging.
Sound like a job for system dynamics? A model of the role of biofuels in America presented at the conference, which was held in Cambridge, Mass., and attended by some 500 people from 50 countries who use SD computer simulations in their work, included 3,300 variables as well as 59 units of measurement.
Jay Forrester, now a retired MIT computer engineer and systems scientist, invented SD. He wrote "Industrial Dynamics" based on his early work, followed by "Urban Dynamics" after working with former Boston Mayor John Collins. The best known SD model is documented in "The Limits to Growth," published in 1972.
Many others continue to build on Forrester's work. System dynamics has its own membership society. At the conference, several programs for developing SD models were demonstrated, with names like Forio, iThink, Powersim, STELLA, Sysdea, Vensim and ViaSim. There's competition, and that's good.
I was fascinated by ReThink Health's simulation of an "Anytown" complete with data on population, health providers and facilities, along with a variety of options for saving and spending additional resources. What, you might ask, are the long-term savings if the community invests upfront in behavioral changes among the young and disadvantaged? If you cap fee-for-service, will many physicians simply will move away? ReThink Health can be calibrated for any geographic area and its health system. It has been used in Albany, N.Y., by local elected officials concerned with just these issues.
America and the world need to find the best possible ways to learn about and solve the many societal and planetary challenges we face. Until something better comes along, and with geniuses remaining in short supply, System Dynamics should and will be a big help. I think it's the best we've got.