I think my colleague Mr. Beaud is more qualified than me to answer this question in detail. However, I can tell you that I personally do not know a single statistician, not a single person working seriously in statistics who will consider data collected on a voluntary basis as being essentially representative of a population. By doing so, we inevitably select people who are more docile, or compliant. So we get a slightly or a very distorted picture—we cannot be sure of how distorted it is—that describes the compliant sub-population within the general population. Just because some Canadians are resistant, lazy or careless and will not respond to the census, it does not mean that we have to refrain from taking the necessary steps to gather information to better understand and govern the general population.
It is a fundamental dilemma. If we do not get good representation in the census, a ripple effect will affect many other surveys, including the Labour Force Survey, which is conducted every month and is weighted based on census data. We will lose all those benefits. It is unfortunate that things are unfolding this way. The procedure has been split into two. In fact, my understanding is that there are now two methods of data collection: one of them is the official census, which barely includes some ten key issues, or eight or nine, and the other is the household survey, which essentially uses the 2006 census formula. All that is now supposed to be done voluntarily. We could make an argument that, by producing the statistics based on the results from those two methods, we would get an effectiveness index, which would help us determine whether the sample is representative, but I do not believe in that. I do not think it would be possible. It is very unlikely.