We have some models in the department that help us understand the dynamic of these decisions. These models are very limited in their capacity to estimate changes in the behaviour of people in face of such changes.
The other aspect to keep in mind is that we're looking at changes 10 years from now, in a context where there's tremendous or deep change in the way people contribute to the labour market. When you look at the current trends, you look at people who are changing their pattern of work. It's very difficult for us to pin down the impact this change will have on the work efforts of older Canadians 10 years down the road. It's something we'd like to do. We have models that can help us, but the modelling is partial because it doesn't take into account all the realities. It is not dynamic; it is static in time, so we have to use historical data to help us understand what the dynamic could be. It's very difficult for us.