2011 census

[cybercoma]

As usual, you're making arguments I never said.

The section you pointed to is irrelevant with regards to my post that you quoted, because it covers what those with access to the data are and are not allowed to share publicly, it doesn't stop them from seeing it without my permission.

And therein lies the problem with what you're arguing. If you have to give permission to every single researcher, who is bound be the confidentiality laws that I posted, to use your data, then you would be getting contacted a ridiculous number of times. The problem with this is that the researchers don't even know who you are. They don't have your name, phone number, or address. More to the point, they don't care who you are personally. They're only interested in analyzing the information in the data sets.

The bottom line is this. What you're arguing leads to the government, policy analysts, and researchers making blind decisions. More to the point, it's nearly impossible for a researcher working with a large dataset (NHS has over 13,000,000 respondents) to get permission from every single individual before conducting his/her research. Even smaller surveys like the Labour Force Survey (LFS) has a sample of roughly 100,000 people. The Survey of Household Spending has a sample of 21,000 households. Aside from not having access to respondents names, addresses, or telephone numbers, it would be a nightmare trying to contact and get permission from that many people individually, if they had to be notified every single time the data set was accessed.

And that's all I'm going to say to you about this, because I'm not interested in edifying you about sampling and survey design, nor quantitative methodologies. You're not interested in learning and I'm not going to waste my time.

[WWWTT]

As usual, you're making arguments I never said.

The section you pointed to is irrelevant with regards to my post that you quoted, because it covers what those with access to the data are and are not allowed to share publicly, it doesn't stop them from seeing it without my permission.

Good debate!

But only to prove why Harper wanted to get rid of the long form census.

The conservative government is always looking for minor wedge issues to win life time support from certain voters. You fit that target class.

In a very big way, you have become the "stat" that you so desperately wish to avoid!

WWWTT

[Bryan]

Aside from not having access to respondents names, addresses, or telephone numbers, it would be a nightmare trying to contact and get permission from that many people individually, if they had to be notified every single time the data set was accessed.

Too bad for you. Find another line of work.

[On Guard for Thee]

Why do they even need a census these days? Harper just collects the info illegally from the phone companys anyway.

[Michael Hardner]

They have all of that data already, they don't need the census.

...

That's all it does.

Which one is it ?

[Bryan]

Which one is it ?

What is the contradiction you think you are seeing?

[segnosaur]

One does have to wonder why the NHS was more expensive, when you have significantly less expenses attached to compliance. What actually added to the cost in 2011?

In addition to the issue of training on a new system/procedures that was mentioned earlier, you probably have the following additional costs:

- More forms are sent out (with the hope that the government would receive roughly the same number of voluntary responses as if they sent a smaller number of forms, but made responses mandatory). This drives up printing costs.

- More effort must be made to analyze the data. Its not simply a case of tossing all the numbers into a spreadsheet and calling it a day... analysis can be fairly complex. (e.g. if a particular city or demographic has a lower response rate, they'll have to 'skew' the data for that city when figuring out national numbers.)

[Michael Hardner]

What is the contradiction you think you are seeing?

"The census collects data they don't already have" - which you agreed to.

"They already have that data" - which you posted.

I'm stuck on this - I think government collecting data is a potential revenue source, and a valuable service.

[segnosaur]

"The census collects data they don't already have" - which you agreed to.

"They already have that data" - which you posted.

Not to put words in Bryan's mouth...

But it might be possible for a person to recognize that the government already has the data they think it needs, while the data that it doesn't have (that would be collected from the census) is stuff it doesn't need.

For example, its possible for a person to think "It would be useful to have statistics on household income... so they could (in theory) get that from Revenue Canada. The government should not care about my race or religion or how many bathrooms in my house (which might only be captured from the census). Its data they don't have, but I don't WANT them to have it."

[Michael Hardner]

But it might be possible for a person to recognize that the government already has the data they think it needs, while the data that it doesn't have (that would be collected from the census) is stuff it doesn't need

Aggregating data is something that it doesn't need, per se, but it can use in the future with a little vision.

For example, its possible for a person to think "It would be useful to have statistics on household income... so they could (in theory) get that from Revenue Canada. The government should not care about my race or religion or how many bathrooms in my house (which might only be captured from the census). Its data they don't have, but I don't WANT them to have it."

A company might want to know where the most bathrooms are in a postal code, which government could sell back to them. It's an idea.

[segnosaur]

For example, its possible for a person to think "It would be useful to have statistics on household income... so they could (in theory) get that from Revenue Canada. The government should not care about my race or religion or how many bathrooms in my house (which might only be captured from the census). Its data they don't have, but I don't WANT them to have it."

A company might want to know where the most bathrooms are in a postal code, which government could sell back to them. It's an idea.

Oh yes, its quite possible for a company to be interested in that type of information. It might even bring in valuable income.

But then, there are the following issues:

- Should the government have the right to force people to answer a survey just because it can earn money from the results? People have pointed out that the survey takes less than an hour, but that's still time that I might have preferred doing something (anything) else... pulling the wings off flies, plotting my revenge on society, building a bunker for when society collapses. There are cases where the government needs to dictate how we spend our time in order to keep society functioning (e.g. jury duty); whether this is one of those cases is questionable.

- While the survey produces data that is less accurate than a mandatory census, would that really be significant when the data is used for commercial purposes? I can imagine (for example) a plumbing company being interested in the number of bathrooms in a community, but if the numbers are off by a few percentage points, it probably won't make a difference in the company's business strategy. (Not to mention the fact that data would already be years out of date even if the census was mandatory.)

[Michael Hardner]

- Should the government have the right to force people to answer a survey just because it can earn money from the results?

That's a fair question. An hour is too much time for everybody to be required to do this IMO.

if the numbers are off by a few percentage points, it probably won't make a difference in the company's business strategy. (Not to mention the fact that data would already be years out of date even if the census was mandatory.)

Turnaround time is another fair question. These things are polls, so there's a certain measure of inaccuracy but overall I think it would be useful.

[cybercoma]

More on the problem.

"Canada has a serious data deficit. And the dearth of facts threatens to cause a host of policy mistakes, affecting everything from who gets unemployment benefits, to where governments spend their limited training dollars and whether employers can import foreign workers."

Read the rest: http://www.theglobeandmail.com/report-on-business/economy/why-canada-has-a-serious-data-deficit/article18598633/

[segnosaur]

More on the problem.

"Canada has a serious data deficit. And the dearth of facts threatens to cause a host of policy mistakes, affecting everything from who gets unemployment benefits, to where governments spend their limited training dollars and whether employers can import foreign workers."

Read the rest: http://www.theglobeandmail.com/report-on-business/economy/why-canada-has-a-serious-data-deficit/article18598633/

Minor point... while that article does mention the elimination of the mandatory long-form census, it seems like the majority of the article is discussing other problems with Stats Can, totally unrelated to the census... it talks about cuts to stats can in general (including things like labor market research).

In fact, I can't see how any of the issues raised in the article relate to the census at all... things like job vacancy rates are probably not going to be gathered from a census (since the delays in collecting and analyzing the data would probably limit its usefulness.)

[cybercoma]

The long-form is a part of a larger problem with StatsCan and CPC policy