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 (II)

by Aaron Wherry on Monday, July 26, 2010 9:02am - 0 Comments

The Industry Minister has claimed that 160,000 Canadians “refused” to complete the long-form census in 2006. Statistics Canada now says 275,000 households failed to return a short-form census and 160,000 failed to return the long form, but only that census officials were unable to “make contact” with the majority of those households.

If a breakdown of the reasons for not responding exists, I’m not aware of it. (Requests for information in this regard have so far failed to produce a response.) It was reported in 2008 that 35,000 Aboriginals had refused to participate in the process. And there was something of a campaign to boycott that year’s count on account of Lockheed Martin’s involvement.

As noted previously, roughly 10.9 million households in 2006 received the short form, while 2.7 million received the long form. The response rates then were 97.5% for the short form and 94.1% for the long form.

Tags: aboriginals, census, Lockheed Martin, long form, Statistics Canada, Tony Clement

More by Aaron Wherry

http://intensedebate.com/people/SeanStok Sean

Aaron, a small quibble: I don't think your response rate numbers can be exactly extrapolated as you've done above. From the CP story: "About 435,000 households were missed in the last census, the agency says. Of those, some 160,000 were supposed to fill out the long form."

I think 435,000 households did not return the short form, not 275,000 as you write. It sounds like those failing to return the long form also didn't do the short form.

Which suggests, to me, that the friggin' long form was not necessarily the issue, as some citizens would refuse to participate in *any* version of a census – from brief to long.

But someone please set me straight if I'm short on coffee and comprehension!

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

I don't think we get both, it's one or the other. So of the 435,000 non-respondents, 275K were asked to complete short forms and 160 K were given long forms.

http://intensedebate.com/people/SeanStok Sean

Ah – thank you for setting me straight!

practical mom

I thought all housholds received the short form. This is from the StatsCan website under "2006 Census: Taking a Census of Population."

Data collection
This stage of the census process ensured that each of the 13.5 million dwellings in Canada received a census questionnaire. The census enumerated the entire population of Canada, which consists of Canadian citizens (by birth and by naturalization), landed immigrants, and non-permanent residents together with family members living with them. Non-permanent residents are persons living in Canada who have a Work or Study Permit, or who are claiming refugee status, and family members living with them.

lgarvin

The short form census is included in the long form. The first 8 questions of the long form are the short form.
practical mom

Thanks for that.
Marion

The short form questions are included in the long form. I got the long form last time (my first census), and I only got one form. I didn't even realize other people were getting something different.

This is where I think the government's idea of a voluntary survey will become a big mess. If it's voluntary, they will actually have to send the short form to ALL households, and then send the second, voluntary form to one third of households again.

Can you imagine the confusion?
Anon 001

Yes, I can. I actually thought there were two forms, one short, one long.
practical mom

The plan seems to be that the voluntary survey would follow the short form census a few months later. People may move in that time so this plan just gets more confusing if the short form questions will be on the long form and the forms are sent out at two different times.

CAPS

In 2006 I received both long and short forms.

Anon 001

As noted previously, roughly 10.9 million households in 2006 received the short form, while 2.7 million received the long form. The response rates then were 97.5% for the short form and 94.1% for the long form.

Right, which would make it a response rate of 95.6% for the short form, instead of 97.5%. So the response rates are still very good, much better than if it were voluntary.

Loraine Lamontagne

La Presse has a very interesting bit of information this morning.
http://www.cyberpresse.ca/actualites/quebec-canad…
http://intensedebate.com/people/SeanStok Sean

The more I think about it, we can't take the 160,000 long form refusals as a strict measure of unwillingness to fill out a long form.

Presumably, some proportion of those households would have not returned a short form either – perhaps the vast majority.

How do we know the long form was the issue for those households, rather than any form of census participation?

http://intensedebate.com/people/LynnTO LynnTO

If StatsCan was unable to make contact with them, we don't. It's speculation. They could be in the process of moving. They could be on vacation. They could be blind. They could also have a fundamental distrust of the Census. All of the above.

Emily

Yes, moving about or new immigrants…possibly illegal ones….or some of the lost aimless people we have wandering the streets.

People usually have 7 jobs in a lifetime these days, or more. They don't stay in the same place anymore.
http://intensedebate.com/people/Stewart_Smith Stewart_Smith

Yet again (still) Tony is all wet.

The 160,000 number is remarkable small. It is quite a bit smaller than the number of those households that would have had a recent death or catastrophic illness. It is quite a bit smaller than the number of those households that would have recently moved or be in the process of moving. It is less than the number of Canadians who take extended holidays out of country each year. It is less than half the number of Canadians who went to work this morning with mismatched socks. (I made the last stat up… why should Tony have all the fun)

BCer in Mtl

I think Tony's dried off by now after his weekend heroics

http://www.calgarygrit.ca CalgaryGrit

If the short form response rate was 97.5% and the long form rate was 94.1%, we can safely assume that about 3.4% of people objected to do the long form BECAUSE it was the long form (i.e. they would have filled out the short form).

So we're looking at about 90,000 actual refusals. Some of those may be because people were lazy and some may be because they actually objected to the form.

http://intensedebate.com/people/Stewart_Smith Stewart_Smith

To quibble, I would say 3.4% did not fill it out because it was the long form.

I am well known at work to immediately do paperwork that will take 5 minutes or so just to get it out of the way. For things that take much longer than that, I usually set them aside until I fine the time. Truth is, I very seldom find the time. So you need to add incompetence as a reason for not filling out the longer form.
Anon 001

Still a very small refusal rate i.e. there was no issue until Harper decided to invent one.

Sigh

As anecdotal evidence of another reason why some forms are not returned:
I was living in residence at a large university during one census. The census forms were put in the mailboxes for all the units in the residence. However, as it was summer, most of the units were unoccupied. To the best of my knowledge, there was no follow-up on the forms that were not returned. If this scenario were repeated at universities across the country, that alone would account for a fair number of unreturned forms.

Joffré

That seems really weird. The census instructs respondents to count students who live away for university but come back home for any part of the year as living with their parents, so handing out forms at university dorms seems counterproductive.

Sigh

It does seem confused. It may have been because it was a graduate student residence, where students tend to be older and not all of them live with their parents. Some students lived in the residence year-round, while some returned to their home cities during the summer.

CAPS

" it was a graduate student residence, where students tend to be older and not all of them live with their parents."

But eventuall they will be moving back into their parents' basement … ;-)

Sigh

Ah, the circle of life.

Gayle

I believe that all residences have to receive a form, When there is no response, someone from the census follows up and if it is discovered no one lives there the form is marked accordingly.

Sigh

As I said, the evidence is anecdotal, but I do know that in this particular case, there were individuals who received the form but did not send it in. There was no follow-up in those cases.

Dave

"University summer", maybe. The census is usually done in May, except in some northern communities.

Sigh

The spring semester usually runs from January to April.

Sigh

I should have been more specific, but I simply recalled it as being in the "summer semester", or "university summer".

jarrid

It's Monday morning and the left/lib partisans are still in a rage about the government abandoning a policy of forcing people to answer census questions under threat of jail

Get a life folks. It's summertime and summers are short in this part of the world.

http://intensedebate.com/people/SeanStok Sean

You guys really miscalculated this one, eh? I, for one, am not a lefty nor a lib partisan (Unless a desire to see a government free of duplicity, dishonour, and ignorance makes me one of those).

I see even Mr. Flanagan could not abide this idiocy. Do you write him off as a left/lib partisan too?

jarrid

Sean, you're a reasonable person, I won't say otherwise, but you are left/lib, so why deny the obvious?
From a purely pragmatic political angle, this is a proverbial tempest in a teapot, whatever one's take on it.

What I do find interesting however, is that any step by the Conservatives to disturb a liberal or leftist legislative or policy status quo is met with howls of outrage from the left/lib media. I good example is Paul Wells. Instead of debating the pros and cons of such a change, all we hear is outrage. Again your average Canadian, as I stated below, does not share the left/libbers undying loyalty to left/lib policy choices and legislation. That's why, at the end of the day, the Scott Tribes of the world will be in blue funk because the latest left/lib media-manufactured outrage won't move the poll numbers.

Sean

There's an awfully broad spectrum of perceptions and positions possible outside the narrow Harper-neocon approach to things. Not all of them are automatically left or classically liberal.

In this particular case, I would heartily endorse a reasoned and intelligent examination of the census. One should never waste respondents' time and efforts with needless questions, and it's fair that every question on the census be included for a genuine need. And it's always fair to ask if any government policy or program satisfies fundamental criteria of necessity – the census should be no different.

But Clement/Harper didn't approach it that way, did they? All of sudden, this became the exemplar of state intrusion in individual privacy (even though the census isn't even mentioned in the 2008 Conservative policy declaration – odd, no?). Clement then proceeded to demonstrate an appalling ignorance of statistical science (more surveys cannot overcome proper representation, for example). Outright lies have been uttered regarding the long form content. An honourable minister is reduced to citing the half dozen messages of support on Twitter, where even the identiy of this handful of Canadians cannot be verified. Irrelevant cases like Denmark have been cited, even though they maintain ongoing databases of individual information that almost no Canadian would tolerate (thus making the census relatively redundant). All in all, a perfectly valid policy discussion became a train wreck. I lived through this approach with the Harris neocons in Ontario, and it's bad governance.

And I don't think I'm alone here. Nothing is sacred, and I sure don't argue for the existence of government programs because they're already there. But I do demand some reasoned and transparent process to accompany the jettison of existing programs that were created, and continue to be used, for particular reasons.
Gayle

"Instead of debating the pros and cons of such a change, all we hear is outrage."

I disagree. What you hear is that the experts say this is a real bad idea. If you really want to have a debate, you have to be prepared to hear from the opposing side without trying to de-legitimize their point of view by characterizing it as "outrage".

"Again your average Canadian, as I stated below, does not share the left/libbers undying loyalty to left/lib policy choices and legislation. That's why, at the end of the day, the Scott Tribes of the world will be in blue funk because the latest left/lib media-manufactured outrage won't move the poll numbers."

Again, an adult discussion means you discuss the policy change on its merits. It is obvious you are far more concerned about how this issue affects polling numbers than you are about how it affects our country.
http://scottdiatribe.canflag.com/ Scott_Tribe

I'm honoured to be used as an example of the left/liberal constituency here in Canada, as I'm sure Paul is to be an example of "media-manufactured outrage"

Jan

Badges of honour are in the mail.

http://scottdiatribe.canflag.com/ Scott_Tribe

Yet you feel compelled to comment on it, Jarrid, despite it being such a boring topic?

If conservatives are out there trying to claim no one cares, I find it because they're very worried that people do care, hence their comments trying to claim otherwise.

jarrid

Left/libbers think it's just peachy to force people to answer general census questions under threat of jail, small c conservatives and libertarians do not. People who are up in arms about this are government bureaucrats, the left/lib media and Liberal partisan bloggers such as yourself Scott.

Surely we can disagree on this little ittly bitty issue without running around like chickens with our heads cut off?

The left is in a tizzy, not your average Canadian. The average Canadian getting a long-form census is as pleased at receiving such a form as a person accidentally stepping in doggy-do.

http://intensedebate.com/people/SeanStok Sean

Do you know something about Tom Flanagan that I don't?

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

Well, of course he does. He also knows what every 'average' Canadian is thinking on a daily basis. No wonder he's not worried about the census!

practical mom

And experiences about the same minor consequences as stepping in doggy-doo.

jarrid

You don't go to jail for stepping in dog sh*t.
http://intensedebate.com/people/TJCook TJCook

And apparently not for spewing it, eh jarrid?
http://intensedebate.com/people/SeanStok Sean

So change the penalty, and leave the census as is?
jarrid

Taking away the threat of jail would be an improvement over the status quo, although many would disagree.
http://intensedebate.com/people/Jenn_ Jenn_

Uh, you can't change the penalty to no-one goes to jail anymore.

Because no-one has ever gone to jail!

Anon 001

You're inserting the same comment everywhere. If anybody needs to get a life, seems to me it should be you.
http://intensedebate.com/people/Thwim Thwim

It's Monday morning and the right/con partisans are still in a rage about the public not abandoning a story of government forcing Statistics Canada to spend more money for less valid results.

Get a life, jarrid. It's summertime and summers are short in this part of the world.
CAPS

Do you include Tasha Kherridan amongst your let/lib partisans who think this is a really, really dumb idea?

http://intensedebate.com/people/Scott_Tribe Scott_Tribe

..and the National Post editorial board, not to mention Kelly Mcparland of the same paper.. as well as the afoermentioned Tom FLanagan.

How about the Canadian Chamber of Commerce or the Canadian Jewish Congress, or the Christian Fellowship of Evangelicals? Those orgs are RAGING left/liberal orgs, right?

wilson

This may help jarrid.
At the 11th hour, before Minister Clement goes infront of committee,
'the national body that advises Statistics Canada and the chief statistician' has spoken….

His recommends : (my read, but I am sure many here will have a diff take)
-StatsCan must balance privacy/intrusion concerns with reliable data
(restating exactly what Clement asked for, as a must)

-Take jail time out, except for those that collect data and break privacy laws.

-Some of the questions are indeed intrusive and totally unneccesary and should not be in the census, take them out
http://www.nationalpost.com/news/Statistics+counc…

http://intensedebate.com/people/SeanStok Sean

Also suggests leaving the 2011 census alone, no?

wilson

Yes, because there is no time to make any changes.
But there still is time.

http://intensedebate.com/people/LynnTO LynnTO

They also recommend going ahead with a mandatory longform Census, with the removal of time use questions, because (they claim) the time-use questions raised the greatest number of privacy concerns.

I'm not sure the time-use questions are a privacy issue so much as they are a "let's be reasonable, here, how much can you really expect me to remember?" issue. And this notwithstanding, the PComm's audit didn't really take issue with the longform Census.

I would argue time-use questions are helpful for establishing tracking data on gender roles and that generational crunch that's getting so much sociological airtime (read: boomers and late-boomers caring for children as well as aging parents); such numbers contribute to the rationales behind homecare, daycare, and equity programs.

Nonetheless, if the data quality is affected by the ability to recall accurately the information required to respond, then it may be better to scrap those particular questions anyhow. I'd have to take a better look at the data to assess…

wilson

By removing the jail time, and leaving the fine as the consequence, may be an acceptable compromise.
Kinda along the lines of a late fee for not filing WCB forms on time.

But questions that do not relate 'directly' to government decisions and where answers are available elsewhere, should be OUT.

http://intensedebate.com/people/LynnTO LynnTO

I suspect if the answers were available elsewhere, to the same degree of confidence that the Census offers through its breadth and methodology, we wouldn't be having this discussion.

Though I too wonder why people are threatened with jail if they don't fill out a form, particularly when the threat is all bark and no bite.

Anon 001

And now for what the Council actually said:

“We believe that the changes will harm the integrity and quality of the Canadian statistical system,” Ian McKinnon, chair of the council, said of the Conservative government’s plans. “At the same time, the council recognizes that concerns about intrusiveness and confidentiality should be addressed.”

They are giving Harper a way out. Naturally, he'll refuse to take it. But will Clement dive into this cesspool and save his boss?

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

I predict that Clement will begin to wade in, decide it's too dangerous and scamper back to shore.
wilson

Giving Harper a way out?
No, the Council is conceeding that Harper is right, and they need to offer up an acceptable compromise or they are stuck with the Government of Canada's decision.

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

"Harper is right"

Wilson, I don't think I've seen anyone on any of these threads say the long-form census was perfect, and so too was the way it was mandated. By which I mean I think everyone was happy to take a look and see what could be improved. This would have come out, I feel sure, if any consultations were conducted whatsoever. That, and the digging in of the heels, slagging 'elitists' and 'special interest groups' misleading the public on advice given, etc. etc., is why this is bad governance.

You have a funny idea of ":right ", although, he sure ain't left. :)

http://intensedebate.com/people/charlesh Charles H.

So, what they recommend are the same solutions that everyone else has been suggesting as potential answers to the government's complaints, rather than scrapping the mandatory long form? Quelle surprise!

wilson

Sheikh quit, said it can't be done!
His advisors say a compromise can be done.

It's just a matter of the Government of Canada accepting a fine still be attached to the census.

Charles H.

Sigh. Wilson, wilson, wilson.

1. The National Statistics Council is not proposing a voluntary census, which means that they aren't agreeing with the government.

2. We don't know precisely why Sheikh quit (since, due to confidentiality rules, he can't say what advice he gave), but it's assumed it was over the demand that he implement a voluntary survey and the government's characterization of his advice regarding such.

3. Given that methods already exist for adding or removing questions to the long form, it's probably safe to assume that Sheikh wouldn't have characterized that as not possible had he been given the option.

4. Furthermore, given the length of Sheikh's tenure in the public service, it's also probably safe to assume that he wouldn't have characterized changing the Statistics Act as impossible either. However, doing so wasn't within his remit in the first place, as it's the job of Parliament to make changes to the law.

Anon 001

The Council also says:

"Since the timeframe before the 2011 census is tight — the agency begins preparing for the next census as soon as the previous one is completed — the council recommends next year’s census go ahead as initially planned, with the mandatory long form sent to 20% of households. "

wilson

Due to timeframe.
This is the Harper government, do you think that they can't get this thing settled and into action in a couple of weeks?
The census forms have not gone to print,
there is lots of time.

Gayle

Heh. So the new spin/talking point is that this was what Harper was after all along.

My goodness – rewrite history much?

wascally wabbit

Too many people are jumping on all of this – claiming privacy concerns etc. without either understanding the facts or refusing to accept them.
These include not only our Prime minister and minister of industry – but stalwarts of the political talking head class such as Warren Kinsella – who is going through a mode of refusing to post comments which do not conform to his current spin!
Census forms ARE sent to every household.
I’ve been attempting to find the specific official reference on the Stats Canada site to support my understanding – but the short form package is sent to every household – and – I may have to correct this later – simply stuffed into mailboxes by Canada Post – addressed to current occupier (simply because around 10% of us change addresses each year and the confusion introduced by explicitly addressing a named occupier could result in a lower return rate than the ones above).
On both the short form – and the long form (the 2011 version has not been published yet but the first 8 questions or so are likely to be the same) – the “head of household” is asked to write in the name of all occupants on the Census date BUT THOSE NAMES ARE ORDINARY DATA FIELDS – not key / index / identifiable fields – which will be removed later in the working data bases in all but the archived images of the census forms – which are NEVER accessible to researchers or other government departments.
Contrast that with your personal health information – currently kept on paper chart files by your family physician – and indexed by your name and / or OHIP / QHIP / other province Health Card or your personal credit information – held by private sector companies and updated willy nilly by every commercial enterprise with whom you do business.

http://intensedebate.com/people/lgarvin lgarvin

I've been attempting to find the specific official reference on the Stats Canada site to support my understanding – but the short form package is sent to every household – and – I may have to correct this later – simply stuffed into mailboxes by Canada Post – addressed to current occupier.

Back in the day, when I worked the 91 census, it did work differently. We delivered the census forms by hand – but not addressed – with the long form going to every 5th house in strict sequence. This was to ensure that forms were delivered at random. It's possible that the delivery has subsequently been subbed out to others but I'm doubtful because it was our job to follow up on the forms that were not returned. We tracked the outcome of every form we delivered and were responsible for it's completion or resolution. Keep in mind, this was close to 20 years ago and I've done a lot of things between then and now. Still, I expect this is – more or less – how it still works.
wilson

Here's a fact for yah,
it's none of the governments business how many hours of housework I do,
and none of yours either.

Just because citizens are herded like sheep into the corral to be counted, and can't be identified individually,
doesn't mean they have the right to ask any questions they want.

http://intensedebate.com/people/SeanStok Sean

You may or may not be right for that particular question. But why aren't you advocating for the scrapping the census, and eliminating the keeping of all national statistics? I'm even surprised you keep mentioning the potential use of other databases. Never mind that it's logistically impossible, statistically unworkable – and generally in violation of existing terms – I should think you'd howl if one suggested sharing your health files with the feds, for example.

So, you apparently agree that some form of national statistics is required to guide the formation of policy, intelligent debate and discourse, and the appropriate allocation of resources and tax revenues. Where do you draw the line, and why? And why would using existing databases be less intrusive? (Point of fact, I can explain to you why it would be more intrusive!).

For all the talk of rights, citizenship also entails certain duties and responsibilities.

Wascally Wabbit

It wouldn't have been an issue jarrid / wilson – except for the fact that your leader (isn't there a gruff German word meaning that?) decided last fall to make it one!
Think of it – we could have all gone to Rodeos or diving into rivers in spate – without having to worry our heads about such things…had he not decided to have a brain f*rt!

wilson

Yah WW, Harper is 'allowed' to govern until Liberals find a real good leader,
but only if he doesn't change ANYTHING Liberal!

Funny thing is, not intruding on citizen's private lives is a liberal-progressive thing,
and it has become obvious to Canadians that it the anti-Anything Conservative is what is driving this outrage.

hosertohoosier

The response rate on the mandatory census is not necessarily the same kind of problem as the response rate on the voluntary census. The key issue is whether the census is systematically more likely to be filled out by some people than others. The current non-responders are small in number, and may often reflect mistaken addresses and things like that. This isn't going to discriminate against any particular group (although Stats Canada has generally had serious problems counting aboriginals). In a random sample mistakes and misrepresentations often cancel out other mistakes and misrepresentations. For instance, some people may overstate their income, just as others understate it.

The first problem is that non-responders in a voluntary survey would be a sample of people that are don't want to take the census. It is not hard to imagine young people, French Canadians and… anti-government folks making up a disproportionate number of those refuseniks. The second problem is one of orders of magnitude. According to one survey I saw, 19% of Canadians say they would not take the long form. The number is probably higher than that in practice, because people usually overstate their civic-mindedness on surveys.
tono-bungay

Not having a response does not imply refusing to respond. Out of all the census forms that don't come back, StatsCan looks into the reason for the lack of response for many but not all of them, and has "waves" of follow-up to determine which ones can't respond and which ones won't. Those that won't get cajoled, helped, threatened, and then charged, and if they still refuse they go to court. So far for 2006, 2 have flat out refused, and gone to court. One was convicted and fined, but the judge declined to order that the census be filled out. In virtually all cases that are followed up, the census answers are entered by census staff who have other means to determine what the answer would have been. A lot of non-respondents are dead, illiterate, or have dementia or are abandoned homes, or are duplicates of people who are actually elsewhere. Look at the stats for people who speak neither English nor French. How did they fill out their forms? They didn't. Somebody helped them.

Proper methodology does not require that every person be forced to reply. You just need a statistically valid sample of every type of non-respondent so that you can appropriately weigh the answers of each type. It's not cost-effective to go after every single person. But for the people who don't feel like cooperating or those who have have something to hide (too rich or too poor or whatever) or have political objections, you need a sufficient sample of the characteristics of each group to be able to confidently figure out what their answers would have been. For that, you need the power of the law to convince the reluctant.

There is a good number of those who actually refused to answer: two. Everyone else didn't need to be pushed that far – the mere threat of a fine (usually $100) and face-to-face help with the form was enough to get their answers.