Hey there, time traveller!
This article was published 4/2/2015 (780 days ago), so information in it may no longer be current.
OTTAWA — The Liberal MP pushing parliament to bring back the long-form census says he’d be content if the federal government would agree to expand the short-form census slightly.
"If the government wanted to take a lot of the wind out of my sails they would add six or eight more questions to the short-form mandatory census," said Ted Hsu, MP for Kingston and the Islands.
Hsu’s private member’s bill to reinstate the long-form census comes up for a vote in the House of Commons today. Although he has support from groups such as the Canadian Chambers of Commerce and the Canadian Medical Association, the bill is largely expected to be defeated because the government has indicated it does not back it.
However, Hsu said if the government added just a few extra questions to the handful on the short form, it would mean that survey could be used by anyone else producing a survey to provide a baseline of data. That would enable responses to be weighted to account for sample bias when an insufficient cross-section of people respond to a survey.
"You could construct a census with slightly more mandatory questions and use them make that sample adjustment," said Hsu.
That, he said, would solve most of the problems people are having with the voluntary National Household Survey (NHS).
Prime Minister Stephen Harper cancelled the mandatory long-form census in 2010, saying it was too intrusive to demand Canadians answer such specific questions about their lives. Instead, the government introduced the NHS, a voluntary questionnaire about things such as income, ethnicity and labour. They kept the short-form census, which is still mandatory for all Canadians, with eight questions including name, address, birthdate, marital status and languages spoken.
The NHS is similar to the long-form census, but it was voluntary rather than mandatory and the response rate was uneven among certain groups or geographic locations. It means some segments of the population are over-represented, others are under-represented, and the data is not of the same quality. Statistics Canada warns the NHS should not really be compared to data from previous mandatory censuses.
As well, more than 1,100 communities, including one-third of those in Manitoba, have no data available because not enough people in those locations answered the survey to make the data useful.
Across the country, economists and market research experts, municipal leaders and health-care authorities have expressed concern about the difficulty using the NHS to make decisions because it is less reliable, and most say the census now can’t be used to adjust for sample bias on other surveys.
The Manitoba Centre for Health Policy last year announced it wasn’t going to use the NHS at all, and would instead rely on the 2006 long-form census.
"We can’t have confidence that the NHS data are reliable," said Randall Fransoo, the associate director of research for the centre. "It’s a fact that the people who respond to the survey are systematically different from people who don’t."
Fransoo said whether expanding the short-form census would fix the problem depends on what questions are asked and how many are added.
Hsu said as a scientist and now a politician, he understands the importance of having good quality data.
"Decisions should always be made with evidence," said Hsu. "This is an example of where we have reduced the amount of good data available to governments and businesses.
Last November, Manitoba Conservative MP James Bezan, the parliamentary secretary for defence, said the government could not support the bill, saying the government is committed to "balancing the need to collect reliable statistical data while protecting the privacy of Canadians and reducing costs for our taxpayers."