We're doing quite a few things. One of the things we're doing is....
First of all, in reality, in the national household survey there will be very little non-response, because we have the responses from the 2011 census itself, so we have a certain amount of data for all of the households that are sampled for the national household survey. That, right there, is a major leg-up compared with what you might have in another non-response situation.
We're going to be focusing our efforts on follow-up across the country in trying to get a very uniform response rate. We're going to sub-sample non-respondents and focus on certain areas in order to get the highest possible response rates for those individuals. We're also going to be directing samples towards areas in which we have known low response rates from previous censuses.
Once we get the data to Statistics Canada, we'll be using all of the auxiliary information that we have from other sources, such as, for example, income tax records, to help us identify any potential problems and correct for them.
Without getting into all the technicalities of it, there is quite a large amount that we can do to mitigate against possible bias in the data. We cannot eliminate it completely, but we can certainly improve it.