The Globe’s Experiment in Census-Taking

The Globe and Mail has allowed their poll about Census “intrusiveness” to stand for the third day.

By 10 a.m. on Sunday over 32,000 people had registered their vote with the newspaper. Undoubtedly this is some kind of record for the daily Globe poll, which got 222 votes for the previous one.

The question is highly leading – “Do you think the long-form census questionnaire is an intrusion on the privacy of Canadians?”

This is hardly the stuff of scientific research.

It does not tell us much about what Canadians “really” think. The question is not substantially different from the typical the tax cut polling question: “Would you like a tax cut?” gets a “yes” response from most people. Such one-off polling questions are void of context or implication on other parts of our lives.

Similarly, asked if I mind the intrusive nature of XXX on my privacy, the question is likely to garner a hearty “yes”. … unless I know there is a trade-off, and both costs and benefits to the consequences of my opinion being acted upon.

Limitations aside, the Globe’s decision to keep this thing going, turning it into a broader sounding of public opinion on the limited terms of this highly emotive question, still yields some fascinating results.

The swing reported earlier has now settled into a 50-50 proposition. That rough measure has held since yesterday morning, at least when I’ve periodically checked it, with slight variations that have tended to be in favour of the government’s view, that the Census long-form questionnaire intrudes on people’s privacy.

The rate at which more votes have been added has, reasonably, slowed. It’s the weekend all across this suddenly tropical land of ours, and summertime leisures are more likely taken by water or air-conditioned malls for most Canadians.

Perhaps they’ll let it run for a few more days, and we can see how uptake changes when people are back at their desks.

Hopefully the Globe will produce a graph of how opinions changed over the course of the past few days, and how take-up has varied. It would be interesting to see the results of this social experiment over time.

The original swing factor may be mirroring the way people initially got the word out to friends who agree with one position or another, urging them to sign up. In this case, poll results would be partly a reflection of which “side” was a better organizer. That might have initially produced an east-west political flavour, as mentioned in the comments to yesterday’s blog.

But while political preferences sometimes have regional skews that might effect the results of a 24 hour poll, given how long the thing has been running, the numbers now speak to different patterns of human nature.

From here on in, if the poll continues to stand for part of the work-week, some of the increased number of votes will be random, some because of active organizing.

No matter how big the numbers are by the end of this process, precisely because it is a voluntary survey and not an explicitly randomized sample, what is being measured is the opinion of those who volunteered to share their view – people who, for whatever reason, have become engaged with the story.

If they had run this polling question a few weeks ago, when the Conservatives announced their decision, the pro-government-decision faction would have probably been much higher.

Since then, we have had nothing short of a public education campaign as a result of massive media attention.

But, alas, there is as much mis-education going on as education, and from our own government to boot.

And it is likely that the Conservatives will crow that, whatever the results of the Globe poll, this is proof that a remarkably high proportion of Canadians agree the government should not insist that people respond to government questionnaires, through Census and, by extension, anything else.

Of course the results from this poll is nothing but an interesting social experiment that underscores exactly what the experts have been saying about the huge new National Household Survey – voluntary processes deliver, effectively, junk, and this one comes with a price tag of something upwards of $35 million.

Perhaps the Globe could use this moment to advance public understanding about the differences between polls, randomized surveys and census. It would be interesting to hear the science and views of experts from all three disciplines of inquiry into the lives and minds of Canadians.

Hopefully the Globe will write something about this exercise in on-line newspaper polling, acknowledging the highly political nature of the question they posed.

Presumably it was designed to divide the public into clear camps. It certainly is no assessment of Canadians’ understanding of this hotly contested issue. But it has told a remarkable story.


  • I wouldn’t be too worried over the government trying to claim that. They know as well as anyone this is an unscientific poll, and everyone knows that online polls can be “freeped” very easily to try and make it look like their POV is the “common” viewpoint out there.. which is what I don’t doubt is what happened when Conservative supporters saw the early poll going 70-30 against them.

    Lets see what a couple of “scientific” polls say (and preferably the telephone polls, not the online ones) before we get too worried.

  • The way the question is worded skews the result, typical Globe and mail, they probably thought about the question for less than 60 seconds and posted it. Hence the reason why there is such a thing as questionnaire designers. I refuse to believe that this is dumb ass poll is a measure of what we are debating here. We are not debating whether it is an invasion of privacy we asre debating, there is not doubt it is a privacy issue, I am surprised more people did not respond yes, it is a question about democracy. Essentially, does a citizen feel it is too much of the state to ask personal questions to allow the collection of reliable information so that many of the decisions that are made within a democracy can attempt to be effectively and efficiently addressed.

    The Globe poll, automatically misleads people to think this is an issue of privacy and invasion of privacy.

    If it were worded to include a short bit on the benefits, I would bet the Globe and Mail a large box of timbits, that the response would be at least 80% feel the mandatory census long form is acceptable.

    By the way has anybody mentioned to the Tories, that not a single person has ever went to jail for not answering the census, and never would. I have never heard of such idiocy in my life as that which is coming from this census debate and the ultra right reformers standing behind Harper. These guys are actually getting pretty scary in their attempt to destroy the country. I just do not understand how it is that they come to the conclusions that they do.

    This whole fiacso is going to do some real large damage to Statscan over the next coupel of years. this whole notion of privacy and jailing people does nothing but feed the fears of confidenctiality and identity theft and a whole lot of other issues that Statcan is fighting to keep its repsonse rates up on many others surveys.

    It is quite sad the damage Harper is causing Statcan and the country because this whole misinformed debate and we will see it down the road in terms of response rates to major surveys, and that needlessly cuts deep into the ability to quantify the nation on many levels and many surveys.

    At what cost is Harper going to drive the country into the ground, all because of his opinion of what privacy and democracy is defined as. You wanna talk about democracy and invasion, well lets start with a man that hides in a prorogued parliament with his tail between his legs every time he phracks up something.

    The functioning of a Democracy does actually demand participation once in awhile, ahhh, should I say, answer a few questions once every 25 years, is that really such a personal cost?

    I did like the Fraser institute suggesting that data users have been getting off free for years and that government should step aside to allow private companies to come in and provide the required data. (it was an article in the globe) Yeah, just like our credit information, available to anybody with a dollar, so confidentiality would never be assured.

    Also let’s assume the private sector dopes come in to fill the data gaps and somehow collects this information. People will still have to be asked these questions by somebody, and this information will be sold back to policy makers and other data users at a profit, and these policy makers will then have to incur these costs, which in the end make it back to the tax payer. So I am at a loss of how, somehow letting the private sector fill such data gaps will be lower cost to tax payers. Does anybody at the Fraser institute actually think about policy responses before writing them up. Makes you wonder what kind of thugs they have working over there.


  • Armine Yalnizyan

    Scott, I would agree with you if this government was acting rationally. They are acting politically.

    Dean Del Mastro’s comments on CBC’s The Current and Tony Clement’s tweets with Stephen Gordon show, if they do know about the science – and that seems unlikely – they don’t care.

    They are strictly speaking to their base, and hoping to whip up a sense of “yeah, that’s right, the government doesn’t need to know what’s going on in my bedroom” among those who aren’t firm supporters of any party.

    As of 7:20 this morning the count on the Globe poll was 38,862, and 52% of voters agreed with the government’s line.

    I seriously doubt the Conservatives would have any compunction about using the Globe numbers to prove their case.

    They haven’t been shy to mischaracterize and misrepresent information before.

    The potential for an accelerated campaign of miseducation of the public is real, which is why I wrote this piece and hope the Globe and the Star and the Postmedia (old CanWest) chain and Canadian Press take on the facts.

    Perhaps the Conservatives want to leave the Age of Reason behind, but most humans don’t.

  • unwarranted and blatant reputation damage to Statcan is the result of this entire fiasco.

    social and political trust is a very difficult and expensive cultural dynamic, and Mr. Harper along with his political minions have just cost Canadians millions of dollars. Lost response rates, damage to the reputation, mistrust and others are now stuck with statcan. And this is without even considering the loss of the census long form.

    just the bickering in the media, and this whole not in my bedroom, attitude has cost Statcan a lot of money in an increased cost to collect data from a hardened public.

    Who knows how far this will get pushed but the blood is gushing from what was once the top statistical agency in the world, over a hundred years in the works being single handed maligned by Harper’s quest to dumb done the nation.

    wow, just amazing leadership.

Leave a Reply

Your email address will not be published. Required fields are marked *