Canadians must be able to count on Statistics Canada-Information allows us to make informed decisions

I have cut an pasted an article from the Montreal Gazette  below. It was written by Richard Shearmur.
Richard Shearmur is professor of Urban and Regional Economics and holds the Canada Chair in Spatial Statistics and Public Policy, INRS-UCS, at the Universite du Quebec. His opinion is very important so I am hoping to give it a wider audience of people who will appreciate his points and act if they concur. I consider the proposed changes odd in light of the fact that we live in a era when the entire process could easily be computerized to gain more and not less understanding.



The government's recent decision to do away with the census's mandatory long questionnaire might appear to many people as a minor technical matter. However, it is a major decision that will substantially reduce the validity of the information that we have about Canada, its citizens, and the way society is changing. This move, which will increase our ignorance about ourselves, will have long-term political consequences: As society becomes less informed, it will be easier for the government to manipulate it and to use its authority to circulate specious arguments and ideological positions.

In a fundamental sense, the move away from valid information about Canadian society is an assault on democracy, but one which is unlikely to lead to mass protest given its somewhat technical nature.

This decision to reduce Canadians' access to information appears to be part of a wider trend toward secrecy and obfuscation that characterizes this government, and is a decision of the same order (though in a different field) as that concerning the secrecy around information on Afghan detainees.

Without information, opposition and criticism, as well as well informed praise and accurate policy appraisal, are not possible. With this decision, Canada is moving away from having access to accurate and valid information about itself. In the long run, weakening the census will weaken Canada.

Why is it such a major decision? Two arguments can be put forward. The first concerns privacy: Despite the government's assertions to the contrary, the mandatory questionnaire is not an invasion of privacy, at least no more than that which is inflicted on Canadians each time they use a bank card or a cellphone. Of course, there is a big difference between a mandatory questionnaire, which people are obliged to respond to, and a voluntary questionnaire, but it has nothing to do with privacy. A mandatory questionnaire can be distributed to specific populations and sub-populations in such a way that statisticians are certain that the final census represents as accurately as possible current social and economic conditions. A mandatory questionnaire is distributed in such a way that it is representative of all Canadians.

The obligation to respond is necessary, not for Machiavellian or Orwellian purposes, but so that statisticians -and anybody else who cares to read their sampling framework -can be sure that the questionnaires are actually filled out, and that the final census information is accurate and representative.

Mandatory responses therefore have nothing to do with privacy invasion, and the current government should explain this rather than pandering to understandable, but ill-informed, fears.

Census data serve only, and exclusively, to obtain accurate overall pictures of Canadian society and of its social and regional sub-groups, and are never associated with the respondents themselves. The confidentiality of these data is taken extremely seriously, and the databases, as well as being kept under tight security by Statistics Canada, carry no names or identifying characteristics of the respondents. As a researcher, I have occasionally worked with micro data (the detailed records from the 20 per cent of Canadians who respond to the long questionnaire), but this can only happen in secure data laboratories, after a personal security check, and after signing a confidentiality agreement that opens me up to criminal prosecution if I divulge any of the information I acquire.

The probability of associating any information with a particular individual is slim to non-existent (only vague geographic information is given, and of course no names), and criminal prosecution will occur if ever, by happenstance, a researcher such as myself manages to do so and then divulges it.

In short, the government is playing a demagogic game, and would do better to inform citizens of the reasons why mandatory questionnaires are important, and of the strict security that surrounds these data. Frankly, I am far more worried about confidentiality every time I use a credit card or a cellphone, or even when I fill in my tax forms, than when I reply to census questions.

The second argument revolves around the value of census information. Why bother to collect census information at all? Well, in this information age, enlightened policy decisions can only be taken if the government and its advisers have a good idea of what is happening in Canada. This can be good news, such as employment growth in St. John's and Alberta, but also bad news, such as increasing income gaps between rich and poor, and maybe housing difficulties in certain neighbourhoods and for certain people (for instance new immigrants).

But how do we identify these trends, and how do we know if region A or social group B is doing well or not? We know this because of the census. And the census tells us this because Statistics Canada has, until now, made great efforts to ensure that all regions and subgroups are represented. This will no longer be the case because statisticians will no longer have any control over who responds and where they live.

In, say, 10 or 15 years it will be impossible to identify whether or not the data acquired from voluntary questionnaires is representative -for the 2011 census the 2006 census can still be used as a rough test, but as society evolves it will become more and more difficult to know exactly what is happening in Canada.

A final argument might be used against mandatory questionnaires, but it is also invalid: the argument would be that there are plenty of small surveys, run by Statistics Canada, that document employment and social trends. This is true, and these small surveys are valuable: but they are only valuable because once every five years we have a well-structured and complete census against which we can calibrate these smaller surveys. Without a good census, even the many small (and voluntary) surveys will become less valuable, and Canadians will know less and less about themselves.

So, although the census questionnaire, and whether it is mandatory or not, appears to be a narrow technical problem, it is in fact highly political. Canadians need to ask themselves: Why does this government not value accurate information about Canadian society and its regions? Is it because it wishes to protect people's privacy? Or is it because an ignorant and ill-informed society is easier to manipulate and govern?




By RICHARD SHEARMUR, Freelance July 4, 2010

© Copyright (c) The Montreal Gazette



Read more: http://www.montrealgazette.com/news/Canadians+must+able+count+Statistics+Canada/3234040/story.html#ixzz0sjXGq9lv