Sure. The one reason you forecast is to get a lead time and to anticipate what you can do.
The secret there is to recognize that the very first thing is that the future is not forecastable. None of us know what will be the demand for anything 10 years or 20 years from now. But what you can do is lay out a set of assumptions and look at what we call alternative scenarios.
For example, we supported the construction industry for one of their advisory councils to HRSDC and generated four scenarios of how the construction industry might go ahead over the next decade. That was done about five years ago. That provided the backdrop for them as they began looking at what policies they might pursue to improve labour markets in those areas.
As they came up with an idea, let's say to expand the apprenticeship program, they asked how well that would work under all four scenarios. We asked them if there were some things they could do that are what we call robust, that work under almost anything you can imagine, or if there are some things very particular and very peculiar, and to understand that difference. They were asked to see whether, as time goes on, they could pick up which of these scenarios they thought was the one that's actually going to evolve.
It is, I think, possible to do a lot. It does require the statistical base to develop these models. Fortunately, we have that with Statistics Canada. Certainly, though, in this case, in this field, perhaps most complicated of all, it requires the interaction between employers, economists, government planners, and the education system, all of whom have their own ways of looking at things and their own ways of defining things. So it represents a real challenge for you.