I have already answered this question several times. The size of the sample—whether it is 10%, 20% or 30%—is not a problem. We spoke about this earlier. It is the strategy used to collect the information that is the issue.
All the scientific literature on this subject indicates that voluntary samples are more problematic. Given that, in practice, we have not used a voluntary long form, what we are saying is based on our knowledge of what took place previously. Will it be 70%, 75%, or 80%? We do not know. The problem is that the damage will have been done. There will be a break in the data series used to carry out analyses and enabling governments to make good decisions.
I do not believe that many statisticians or people interested in these matters would defend this. The quality will diminish. To what extent? It is difficult to say because we have not yet gone through the exercise.