As computers have become more powerful and less expensive, the ability to simulate reality is changing the way we engage in science and technology. Aviation engineers, for instance, would not think of testing a new airplane wing in a wind tunnel until they developed a numerical simulation of how it should work.
Michael Wolfson’s career has been defined by the role of simulations. The Canada Research Chair in Population Health Modelling/Populomics and professor in the University of Ottawa’s Faculty of Medicine first encountered their vital function as an undergraduate physics student, when he compiled a model for tracking the motions of subatomic particles through the magnets of a cyclotron.
After obtaining a PhD in economics at the University of Cambridge, Wolfson joined the federal government as a policy analyst. In the late 1970s, he was employed by the federal Task Force on Retirement Income Policy before being seconded to the Privy Council Office to work on the government’s pension policy Green Paper. Then, as today, there was a vigorous political debate about whether the federal government could deliver adequate retirement incomes.
“Pension policy has multiple objectives,” explains Wolfson. “There’s an anti-poverty objective that has been a major success story, so that Canada has one of the lowest elderly poverty rates among countries belonging to the Organisation for Economic Co-operation and Development. The other main objective has been to maintain continuity of consumption, which is a middle-class issue.”
As a firm believer in the importance of evidence-based policy, he began developing a life course simulation model for the PCO to detail the major economic and personal decisions people make throughout their lives, though the essential longitudinal data were almost completely lacking. “With this rather simple model, we could foresee the challenges facing Canada’s retirement income system,” Wolfson observes. “Now, with a far more sophisticated model, our projections show that about half of middle-income earners over 40 today will see a significant decline in their standard of living post-retirement.”
Wolfson’s career took a sharp turn when he was recruited to Statistics Canada in 1984 and given a new research division by the chief statistician. He championed the development of microsimulation models for applied policy analysis, including a key one called LifePaths, which was designed for pension and related analyses. In his current role as Canada Research Chair, he has re-entered this major public policy debate.
“It’s long past time to expand public pensions because the private sector has proven unable to deliver the goods for even a large minority of workers,” argues Wolfson. “The opposing rhetoric makes it seem that the government is some evil ‘other’ forcing Canadians to do something they don’t want. But the government is us. Expanding public pensions is no more than an efficient and simple way of forcing ourselves to save for something that is in our own interest.”
The LifePaths computer simulation model at the heart of this analysis is complex and, like other sophisticated dynamic microsimulation models, requires a multi-skilled team for its development, maintenance and ongoing use. But Wolfson points out that the current government in Ottawa is well-known for disregarding evidence when making policy. He worries about the steady erosion of analytical skills within federal government departments as people with expertise in this field have left without being replaced.
Still, Wolfson remains passionate about the future development and potential of simulation modelling. “The most powerful new insights are coming at the interface between traditional subject matter areas,” he says.
In fact, the multidisciplinary methods that make it possible to predict the performance of government pensions or an airplane wing can also help us determine the fate of a distant star, anticipate the spread of disease in crowded cities or chart the course of plant species in a rapidly changing landscape. As researchers continue to explore the frontiers of simulation modelling, they are just beginning to learn how powerful information can be.
by Tim Lougheed