Skip to main content Skip to local navigation

Professor Valerie Preston says making the long-form census voluntary could hamper research on Canada's vulnerable

A growing chorus of Toronto voices, including the director of York’s Centre of Excellence for Research on Immigration & Settlement (CERIS), is opposing Ottawa’s plans to change the national census, which gathers in-depth information from Canadians to form public policy, wrote InsideToronto.com and The Beach-Riverdale Mirror July 13:

Beginning with the 2011 census, held every five years, Industry Minister Tony Clement is doing away with the mandatory long form, which had been sent to one in five Canadians. The much shorter survey that goes to all Canadians will remain compulsory.

Instead of forcing 20 per cent of the population to fill out the long form under threat of jail time and fines, Clement said a third of Canadians will be sent the long form, which they can voluntarily complete.

But the change has prompted an outcry across the country from people who believe making the long form voluntary will result in less accurate information.

York University Professor Valerie Preston, director of CERIS, told Toronto Community News she is upset with the changes. “We have a 97 per cent compliance response to the (mandatory) long form and so it gives us a very complete picture of Canadians. A voluntary survey will not give us anything like that compliance rate. It won’t even approach it,” she said.

Communities such as recent immigrants and lower-income Canadians who move often will be less likely to fill out a voluntary survey, she added. And less reliable census information could mean the needs of the most vulnerable, such as newcomers to Canada, the elderly, the poor and single parents, will be ignored, she said.

“I’m very concerned,” she said. “Without that information, how are you going to transfer (government) funds where they are most needed?”

Preston argued reputable public polling firms, which rely on voluntary responses, can start out with a sample of 17,000 possible respondents but only end up getting answers from 1,000 people.

Republished courtesy of