Beyond The Commons

 Beyond The Commons

Aaron Wherry covers all the goings-on in and around Parliament Hill. Follow Aaron on Twitter: @aaronwherry

Categories: Beyond The Commons, Capital Read

The collection of numbers by the numbers

by Aaron Wherry on Wednesday, July 21, 2010 1:03pm - 0 Comments

According to Mr. Clement’s office, the 2006 census cost $567 million, while the 2011 census is expected to cost $660 million.

The 2001 census was sent to 11.8 million households—roughly 9.4 million receiving the short form, 2.4 million receiving the long form—and resulted in 52 cases of non-compliance being referred to the Department of Justice. According to a National Post report in 2006, “44 were resolved before trial, three pleaded guilty and four were found guilty, while one was given an absolute discharge.” A Statistics Canada official at the time told the Post that fines typically ranged between $50 and $100.

The 2006 census was sent to 13.6 million households—roughly 10.9 million receiving the short form, 2.7 receiving the long form—and resulted in 64 cases of non-compliance. A July 2008 report from Canadian Press cited a Statistics Canada official as saying most of those people had, by then, complied. A Kingston man, who objected to the use of software purchased from arms manufacturer Lockheed Martin to process the census, was found guilty and fined $300. The case of a Saskatchewan woman, who also boycotted the census because of Lockheed Martin’s involvement, is still before the court.

Tags: , , ,
Bookmark and Share
  • MostlyCivil

    Maybe it's just me…but do at least 10 of those mentioned above have twitter accounts?

  • http://intensedebate.com/people/tobyornotoby tobyornotoby

    I somehow doubt the people who were protesting the selection of a foreign arms dealer as the data base provider are twitter friends of Tony Clement.
    .

  • Lord Kitchener's Own

    It does go to show that the Tories aren't totally partisan. In addition to catering to right-wingers who think the government is out to get them, removing the mandatory long form of the Census also caters to left-wingers with an extreme paranoia regarding the U.S. military-industrial complex.

    The Tories are clearly willing to cater to ANYONE's irrational fears.

  • http://intensedebate.com/people/Jenn_ Jenn_

    Thanks for that, Aaron. What we needed to know.

  • Mike T.

    The preferred unit for the measuring government spending is a "gun registry", which is equal to two million Canadian dollars. For example, "the 2011 census is expected to cost [330 Gun Registries]." Or, "The costs of the initial outlay for the 65 fighter jets will be 4500 gun registries"

    "

    • Olaf

      "the 2011 census is expected to cost [330 Gun Registries]."

      Sorry, what's the dollar figure you're using for one gun registry unit?

      • http://intensedebate.com/people/Crit_Reasoning Crit_Reasoning

        Judging by Mike T.'s math, he seems to think that one "gun registry unit" is a paltry 2 million dollars.

        • Olaf

          I came to the same conclusion. Either that, or he's labouring under the impression that the census will cost $660 billion dollars, which I admit would appear quite excessive

          • http://intensedebate.com/people/tobyornotoby tobyornotoby

            If you want to use the total cost of establishing the gun registry, would you not also have to compare that with establishing Stats Canada from scratch in 1996 dollars? Otherwise it would seem fair to compare the operating costs.

        • Lord Kitchener's Own

          Is that perhaps the cost to keep it running once it existed?

  • Dee

    Re: 'According to a National Post report in 2006, “44 were resolved before trial, three pleaded guilty and four were found guilty, while one was given an absolute discharge.” A Statistics Canada official at the time told the Post that fines typically ranged between $50 and $100.'

    So, Tony Clement and Maxime Bernier's "large" numbers of complaints turn out to be miniscule. For the 2001 census a total amount of fines around a few hundred dollars are handed out. Meanwhile, for the upcoming census Clement proposes spending tens of millions of dollars more to obtain a deeply flawed, inaccurate census to allay these "large" numbers of complaints. Good to see the Conservative government is so careful with taxpayer's money, eh?

  • wilson

    Q: Since then, has that belief been challenged? Has Statistics Canada given you any reason to believe they are not happy or not satisfied with this?

    A: I think it's clear there are some bureaucrats in StatsCan who are not happy with this. It's obviously clear because they are speaking to media. That's my evidence: they're speaking to media. They haven't spoken to me but that's not necessarily where they would go to speak.

    Q: Is Statscan an independent agency? I am unclear on that.

    A: It operates pursuant to legislation and it does report to a minister who is responsible and accountable to the public.

    Q: So it's not independent like [Auditor-General] Sheila Fraser?

    A: No. No.

    Q: So it's not arm's length

    A: No.
    http://www.theglobeandmail.com/news/politics/otta…

    ''Statscan cancels census town hall; agency chief ‘reflecting on his future’ ''
    Statscan cancels census town hall; agency chief ‘reflecting on his future’

  • Charles H.

    Thank you, Aaron. This makes a comment from another thread a little closer to what I was curious about.

    Some quick back of the envelope calculations:

    The population of Canada as reported by the 2001 Census was 30,007,094.
    The number of dwellings reported by the same Census was 12,548,588.
    This is an average of 2.39 people per dwelling.

    2.4 million households (approx) received the long form in 2001.
    Assuming the above average holds, this means that approx. 5,700,000 people were surveyed by the long form.
    As certain MPs and commenters like to repeat: 21 thousand (again, approx.) of these peoples gave "Jedi" as their religion.

    That's 0.37% of the total number of people surveyed or, assuming a 95% completion rate, 0.39% of those who responded.

    Now, there are two major assumptions in the above calculations:

    The first is that the compliance rate in 2001 was the same as 2006. That's unlikely. As wilson likes to point out, the non-respondent rate has been increasing — though whether the quoted increase is in absolute numbers or percentage of forms sent out isn't something I've seen mentioned. In either case, the percentage of "Jedis" is actually smaller than those numbers arrived at above.

    The second is that the average number of people per dwelling holds true for those calculations. If it doesn't, then the number may be off in either direction.

    Though it does illustrate why we keep hearing about 21,000 Jedi rather than a percentage — 21,000 sounds so much more impressive than "less than 0.4%".

    (Curiosity now strikes however: if I were to pay for the detailed religion breakdown, rather than the free "250,000 or higher" data on the Stats Can website, would I see Jedi as a separate category, or would it be lumped into another category — likely "no religion"?)

    • Dave

      "As certain MPs and commenters like to repeat: 21 thousand (again, approx.) of these peoples gave "Jedi" as their religion."

      I just checked the StatsCan summary table for religion in the 2001 census.

      That table has a lower cutoff of 20,000. Serbian Orthodox, at 20,520, is the smallest religious affiliation to have made the cut. Next-higher are Brethren in Christ, Pagan (which includes Wicca), Methodist, Hutterite, and Aboriginal Spirituality, all less than 30,000.

      Where are the 21,000 (approx.) "Jedi"?

      • Charles H.

        Hence my last question. From what you've just said it would sound like the data was either thrown out as obviously false or it was in fact rolled into another category.

        (I had missed the more detailed table when I posted that, having only spotted the one that cuts off at 250,000 members. Thanks for pointing it out that it's online. Link for the curious.)

        • Dave

          Strong the Force with you is.

  • bergkamp

    "…. and resulted in 52 cases of non-compliance being referred to the Department of Justice."

    I don't know where you got your stats from, Wherry, but does it say how many people declined to comply and weren't referred to DoJ like both Kinsella and Levant claim to have happened to them. I would like to see stats on who gets the exemptions and how do they factor that in to keep the data reliable.

    If making census mandatory is to keep stats sound, wtf do exemptions do to the numbers. Seems like well connected lawyers get a break and I wonder who else does as well.

    "Ultimately, he said, the census takers simply gave up trying to make him complete it. "When I got the long form census many years ago, I thought it was very intrusive," he said.

    He said he challenged census staff to punish him. "I said 'I am not going to change my mind. Put me in jail. I dare you'," Mr. Kinsella explains. "So they gave me an exemption." Globe/Mail, July 16, 2010

    • wilson

      Clement was quoted , 168,000 Canadians refused to fill in the form.
      But no details on if that was before or after the census police were done with them.

      • http://intensedebate.com/people/tedbetts tedbetts

        Making it 0.06% of the forms were not returned.

        • wilson

          Which does not reflect all those law abiding citizens that fill it out, and did not like it.
          Polled twice, half of Canadians support the govt in making the census voluntary.

          • http://intensedebate.com/people/tedbetts tedbetts

            I've only seen one poll that was an online poll (which also showed that the more you understood how statistics worked your disapproval response went way up).

        • Crit_Reasoning

          That would actually make it 6%, not 0.06%.

          • http://intensedebate.com/people/tedbetts tedbetts

            DOH!

            That's why I went into history and law instead accounting!!

    • Standing By

      Sorry, but Kinsella makes up sh&t all the time.

  • tedbetts

    One of the odd claims by defenders of Harper on this idiocy claim the long form is already effectively voluntary and certainly inaccurate now since 160,000 long forms were not returned.

    So the long form went to 2.7 million Canadians. That would make the response rate equal to 99.94%.

    What is it with the far right and numbers? They can't do balanced budgets, they can't count supporters, they can't calculate response rates.

    UPDATE: Sorry, just reread. 2.7 million Canadian households received the long form which means about 5.7 million people received it. But then 160,000 non-respondents probably also refers to households so the response ratio either goes up even further or remains the same very high response rate.

    What are the chances the response rate on the voluntary forms is going to be anywhere near there?

    • wilson

      Didn't recent polls state a 70%+ would still comply….not sure, but it was close to that.

      The US experiment with voluntary census concluded that an increase in voluntary sampling would be as reliable as the mandatory census.

      • http://intensedebate.com/people/tedbetts tedbetts

        Um, in a word, no.

        "While Canada debates the merits of a compulsory long-form census, its neighbours to the south have already decided a national voluntary survey doesn't work.

        At the behest of Congress, The U.S. Census Bureau conducted an experiment in 2003 using its compulsory American Community Survey to see what would happen if some people were given the choice of filling it out.

        The result? The data was degraded so much, fixing it would be too expensive. The idea was quickly abandoned."

        Response rates dropped by a third.

        Which is really partially besides the point: the reliability is based upon the randomness of the sampling more than the response rate. A voluntary census is not reliable because it is not random.

        • wilson

          Yes Ted, and the US Census Bureau concluded:

          'If the American Community Survey were made voluntary,
          the Census Bureau concluded that it would have to mail survey forms to a larger number of households
          and conduct a larger number of follow-up phone or personal interviews
          **** in order to assure the same quality of results that it promises from the mandatory survey.'

          (Which is what the Govt of Canada is doing)

          http://pewresearch.org/pubs/1655/should-census-su…

          • Charles H.

            There are two clauses in the quoted conclusion. "[M]ail survey forms to a larger number of households" is indeed what the government has suggested doing, but it is merely the first half.

            "[C]onduct a larger number of follow-up phone or personal interviews" is the second clause. There have been no reports that the government has suggested that StatsCan do this. Given that this would likely be disliked to a greater degree and seen as even more intrusive than the current mandatory long-form, I find it unlikely that they will suggest it. ("Why yes, we did make changes to require StatsCan to either phone you or visit you at home. After all, it's just like telemarketers and door-to-door salesmen, and everyone likes those, right?")

            They are joined by an "and", stating that in order to maintain the same quality of data both actions are required. Merely doing one of the two (larger number of forms but no follow-up interviews, or follow-up interviews but the original number of forms) is not enough to maintain the same quality of data.

            So no, the US Census Bureau's conclusions are not what the government is doing.

      • http://intensedebate.com/people/Thwim Thwim

        The hell are you getting that from? The only articles I can see are that the data degraded so much they decided they couldn't use a voluntary census.

        • wilson

          You must be reading regurgitated Canadian media,
          the US experiment changed the census to monthly, and came to the same conclusion that StatsCan came to,
          larger sample would equal mandatroy result.

          • http://intensedebate.com/people/tedbetts tedbetts

            Wilson: that is simply false. No statistician believes that. You can't make up for randomness by simply increasing the number of respondents. You just magnify the skewering that results.

            If a certain segment of the population is less likely to respond to a voluntary census, then how on earth can this be magically "fixed" just by increasing the number of forms issued? Makes zero sense. Better to defend the decision on the "coersion" grounds than to make stuff up.

          • Thwim

            Link it, please. I've tried to google, but all I find is the same article about how their experiment showed that a voluntary census isn't effective.

          • http://intensedebate.com/people/Halo_Override Halo_Override

            http://www.voicesinhead.net doesn't show up. I'm afraid you're on your own, unless wilson comes through.

            So you're on your own.

  • http://intensedebate.com/people/amherstvw amherstvw

    What's an election cost $300M? Maybe we should still be waiting out PM Harper's 4 year fixed date to save money?

    • http://intensedebate.com/people/amherstvw amherstvw

      Judging from recent results, any census would have more value than what elections have recently returned … and returned.

    • http://intensedebate.com/people/ChrisInKW ChrisInKW

      Fat chance. Parliament's likely to get 'dysfunctional' again before the time is up.

      • http://intensedebate.com/people/amherstvw amherstvw

        I was thinking of the time before – to measure the 4 year term ( … or was it the time before the time before?)
        And are they still "Canada's New Government" ? Modern politics in Canada is so confusing !
        I have learned that "extra" elections are "unnecessary, wasteful" … unless they are called ad hoc by PM Harper.

  • http://intensedebate.com/profiles/dougrogers dougrogers

    Freeing up the courts to get tough on crime.

  • Sigh

    It's not clear from the way it is worded above, but were the 52 cases non-compliance in filling out the long form only, or non-compliance in filling out the census forms in general?

  • http://intensedebate.com/people/tobyornotoby tobyornotoby

    We don't need to have everyone fill out the census, we just need one completely average person to fill it out. Diana-Moon Glampers where are you?

    • Sigh

      Which one?

      "…a sixty-eight-year-old virgin who, by almost anybody's standards, was too dumb to live".
      – God Bless You, Mr. Rosewater

      • http://intensedebate.com/people/tobyornotoby tobyornotoby

        Yeah, that one.

  • http://www.countmeout.ca D. Rogers

    I don't often agree with the Tories, but they have got it right on this one!

    The census was supposed to be a simple head count. But it has grown over time (along with its bureaucracy) into an invasive array of personally-intrusive questions.

    The idea of throwing someone in jail for three months and/or a $500 fine is abhorrent. One might expect to find such oppressive coercion in some third world countries, but not in Canada.

    Other polls, such as Ipsos-Reid produce statistically-accpetable results on a voluntary basis. So too can the Census.

  • http://intensedebate.com/people/TJCook TJCook

    Those stats are amazing!

    I keep trying to figure out why Harper would attack the census. Maybe he just can't stand to have such a stellar example of good government in action.

Previous

Next

From Macleans