In the work that we've done, it's really similar to the work that's done across the country by researchers. The data you get in the census is often quite different from the data you get in a voluntary survey. This is why people are talking about non-response bias. For example, if we look at age, younger people tend to be less likely to participate in voluntary surveys. Older people have more time, as you say, and they get to them.
If we look at the allophone community, we get lower participation in voluntary surveys by that community, which is one of the communities we are worried about missing when we miss the mandatory long-form census. Their participation is actually reversed. It's the younger allophones who are more likely to participate in these surveys than the older allophones.
The issue is that you don't know where that non-response bias is. It's there. You can't measure it and we know it varies across the sectors. By making it mandatory, you are encouraging everyone to participate, so that when you do the voluntary surveys, you then have a base upon which to understand how to adjust and weigh back to the known reality of the country.