I had the rather odd experience of having been obliged to fill out the Australian census while visiting that country last month. The census process there seemed to pass without controversy and most Australians merrily complied. It was quite a contrast to events here in Canada.
As many readers will know, in late June of last year, the federal government announced that it was eliminating the mandatory long-form census, sent out to one in five households, replacing it with a voluntary National Household Survey. That decision, which came as a total surprise to almost everyone, elicited months of heated criticism and debate.
In Australia, the census was conducted on August 9, and all respondents were instructed to fill it out based on their current situation on that exact evening. The cheerful exhortation, seen and heard in ads everywhere across the country for weeks prior to August 9, was to “Shed some light on census night.”
Unlike in Canada, where most households are required only to fill out a short, basic eight-question form – with two additional questions on language added this year – every household in Australia is required to complete a full census form of roughly 60 questions. (Unfortunately, I’m unable to find an online version of the Australian census form to link to, but this paper from the Australian Bureau of Statistics explains in detail the nature and content of the 2011 form.)
The questions asked in Australia are similar to those in the former long-form census in Canada and the new NHS that replaced it. And, as I mentioned, it is mandatory for all Australian households to complete the full survey. Additionally, I was surprised to discover that this requirement extends to anyone who spends census night in Australia, including non-citizens and visitors – hence my obligation to fill out the Australian census. I happily complied.
In Canada, the controversial changes to the census went into effect for this year’s census conducted in May. The changes were fiercely opposed by nearly all users of census data, including citizen’s groups, associations, municipal governments and academics, many of whom rely on census data for their research. Their main concern was that the voluntary nature of the new NHS will reduce the response rate and lead to self-selection bias – i.e., the likelihood that certain groups will be less inclined to fill out the survey – skewing the results and lessening their reliability.
(That prophecy, by the way, appears to be coming true. This blog post by the Canadian Association of Public Data Users notes that completion rates for the NHS for people who filled out the census online was about 60 percent and predicts that completion rates will be even less for those who filled out paper census forms. Groups “exhibiting or likely to exhibit non-response bias” included people with lower socio-economic status, immigrants and aboriginals as well males under 24.)
In a public statement on Canada’s census controversy, posted on July 13, 2010, then industry minister Tony Clement cited “complaints about the long-form census from citizens who felt it was an intrusion of their privacy” as the reason why the government did away with the mandatory long form. “The government does not think it is necessary for Canadians to provide Statistics Canada with the number of bedrooms in their home, or what time of the day they leave for work, or how long it takes them to get there,” he said.
Interestingly, the Australian census did ask the number of bedrooms in our dwelling. The question gave me pause, as I was on vacation there with my wife and children, and on August 9 we were travelling the Northern Territory in a camper van. We decided the correct answer was “one.”