I'd like to make the first attempt at answering that question. And in doing so, I'd like to refer the committee to a very helpful article or paper that was recently published in a prominent journal called Canadian Public Policy. It's volume 36, number 32010, I believe, and it was published in September of this year. The article is entitled “The Importance of the Long Form Census to Canada”written by David A. Green and Kevin Milligan, from the department of economics at the University of British Columbia in Vancouver. To quote a short excerpt from that, it says that
All voluntary Statistics Canada surveys come with a set of weights of this type that researchers need to use to obtain accurate statistics. But constructing those weights requires having a “true” population benchmark, and the census is that benchmark. Thus, without the census, both the stratification and weighting stages of all other surveys would be affected. For the LFS this would mean inferior statistics on unemployment and employment. Beyond the set of surveys collected by Statistics Canada, privately collected (e.g., by polling firms) surveys
--the members of our association--
must also be compared to some standard to ensure they are providing unbiased statistics. Comparing them to some other voluntary survey (such as the NHS) which has its own, unknown, response biases is obviously of limited usefulness. Thus, to ensure the quality of these surveys, the mandatory census short and long forms are important.