Editorial: Loss of long-form census hurts our ability to know

By Andrew Bates - Editor-in-Chief

7/15/10

Opinions are just gossip when they aren't backed up with facts.

In the most recent example of governing-by-announcement in Canada, Tony Clement has announced that in 2011, the mandatory long-form census will be scrapped in favour of a compulsory eight-question questionnaire. Instead, Statistics Canada will collect a voluntary National Household Survey that asks the same questions as the old document.

Basic survey theory states that your data is shaped by the people who answer it, and it is necessary to attempt to control that skew as much as possible. The sample data collected from the new survey will be slanted in favour of people who are more likely to want to respond to surveys. Up until now, that bias has absent, as collection was mandatory.

Much of the uproar on this issue has focused on the ideology of the Conservative government, claiming that gutting the country's premier public policy tool allows them to rule on coffee-shop wisdom and that they want to keep the people in a state of ignorance. I won't tread those partisan waters, because I think the question of why it's done is less important than why it should stay.

What we're talking about is knowledge. Governments attempt, generally, to gather information and make a decision based on it. Sociologists attempt, generally, to use detailed statistics to make conclusions on how society works. We as newspapers attempt, generally, to gather knowledge and present issues to people so they can be better informed and form an opinion.

But if hard facts aren't available, we have to rely on what people say. The Office of the Privacy Commissioner of Canada says they received just two complaints about privacy in the census in the last four years. Statistics Canada say they received no feedback about privacy concerns after the 2006 edition.

Tony Clement says that he has recieved "dozens and dozens" of letters supporting the decision, and that people haven't been complaining through those channels, but they have been complaining to MPs. How can that be refuted? There's no official statistics on MP complaints. There's no way to accurately measure the amount of people Clement talked to who don't like the census, pushing him to make a decision which could affect statistics history.

The only way for us to know why or what Clement is up to is to guess. But that's not really reliable, is it? Denouncing that claim as bogus is as fact-based as making a major national decision because you heard some people might be unhappy. It's gossip all around. That's what will happen more and more, if there is no mandatory census, as our pool of knowledge will shrink and shrink. We need to know. We need to. That's why we're here; to inform you on what matters.

Governments that eliminate reliable information impede our ability to gather knowledge and yours, as well as public servants, sociologists, health care professionals, and anyone else who needs to know the hard facts before making their decision. The mandatory long-form census needs to be restored, because you can neither run a country nor inform it on gossip.

Tags: editorial, Andrew Bates, long-form census