Miskatonic University Press

Census data matters


The Canadian census has been in the news a lot recently because Industry Canada Minister Tony Clement decided that for the 2011 census, instead of one in five households receiving a mandatory long-form survey (that goes into great detail, asking a lot of questions) now one in three will receive an optional long-form survey. He's doing this because he claims lots of Canadians find the form intrusive.

The decision caused an uproar, with good reason: it makes the data statistically useless. That sampling technique makes any analysis completely unreliable. The people at Statistics Canada (who, on another point, are scrupulous about privacy) know this, of course, but the government is ignoring them.

Today Munir Sheikh, Chief Statistician of Canada, resigned from Statistics Canada:

I want to take this opportunity to comment on a technical statistical issue which has become the subject of media discussion. This relates to the question of whether a voluntary survey can become a substitute for a mandatory census.

It can not.

Under the circumstances, I have tendered my resignation to the Prime Minister.

I'm an academic librarian, and every week I see that census data is crucial to work in universities, from student assignments to faculty research to country-wide planning about the education system. It's used in the public sector and the private sector in countless ways. It helps us understand the past and the present, and prepare for the future. Census data is at the core of all government planning. Should be at the core. Used to be at the core. Now, perhaps so that there is no evidence of the effects of their policies, the Conservative government is going to junk where that data comes from.

Throwing away the detailed 2011 census data is sheer stupidity. This decision is bullshit.