From The StarPhoenix, Tuesday October 11, 2011

An ill-advised ministerial decision that put political pandering ahead of sound scientific methodology likely will continue to hamper Canada's ability to make wise public policy decisions and conduct good social research for at least a decade or more.

A study commissioned by Statistics Canada says there's not enough time to embark on a new type of census before the 2016 census or even the 2021 tally. Consultant Don Royce concluded the only viable course remains to juggle the components of the current census, the methodology of which has come under attack by statisticians and academics after a political decision to make completion of the long-form census non-compulsory.

While Statistics Canada reported in May that there were no major issues with its 2011 census that saw 98.1 per cent of respondents complete and return the short form, the agency hasn't yet ascertained the quality of the information it has received from the longer voluntary National Household Survey, which had a response rate of 69.3 per cent.

The agency's former chief statistician, Munir Sheikh, quit after then industry minister Tony Clement wrongly claimed the agency had backed the government's decision on moving to a voluntary long form. Sheikh noted bluntly that such an approach will provide inadequate data or biased results, particularly on such groups as aboriginal people and recent immigrants who might be less inclined to fill out the forms if doing so is voluntary.

And Mr. Royce notes in his report whether elements of the current survey can be used in 2016 depends upon the quality of the results Statistics Canada obtained from this year's survey. What will prove interesting is what options the agency will have for the 2016 or even 2021 censuses should the fears of social scientists and statisticians prove to be well-founded.

In the wake of the government's abrupt change to the census, among the options for revamping the process was a suggestion to create a national repository of citizen data, as Scandinavian countries have done by using central population registries.

However, as Mr. Royce notes, it would take decades for Canada to create such a registry that's complete with data on everything from dwellings to employment to education and income.

And if, as Conservative politicians such as Mr. Clement disingenuously suggested in making the change, many Canadians are opposed to sharing personal information ranging from the number of rooms in their homes to their annual income, imagine the reaction to having such data gleaned from sources such as income tax files and then stored in a centralized location. The cost alone would make one's head swim, never mind the security and privacy implications.

Even to adopt the process used by countries such as the U.S. and France, which conduct short general surveys every decade, supplemented by ongoing surveys of a percentage of the population on more detailed questions, couldn't be implemented before 2021, Mr. Royce suggests. Again, this would take at least 10 years to implement, be more expensive than the model that was scrapped, and requires mandatory compliance, which the government abhors.

So, a government that claims to be cost conscious is spending $30 million more now to gather census data that shortchange federal and provincial agencies on information needed for decisions on everything from pension programs to labour shortages to helping aboriginal people and immigrants to adapt and thrive in Canada.

Call it the cost of political expediency.

The editorials that appear in this space represent the opinion of The StarPhoenix. They are unsigned because they do not necessarily represent the personal views of the writers. The positions taken in the editorials are arrived at through discussion among the members of the newspaper's editorial board, which operates independently from the news departments of the paper.