He's at it Again...

Professor Gordon, that is. This time it's the National Post:

Economists Niels Veldhuis, of the Fraser Institute, and Stephen Gordon, of Laval University, join host Chris Selley for a surprisingly spirited debate on the elimination of the mandatory long-form census.

Also, as Canada's biggest Hogan's Heroes fan (seriously - I do a pretty good Col. Klink imitation, but only after a couple drinks), I got a huge laugh out of this headline, even if I don't think it's entirely fair: Tories try Hogan’s Heroes defence in census feud.

Edited to Add: Messrs. Veldhuis and Gordon spend a fair bit of time in the podcast discussing a U.S. study which compared the effectiveness of a voluntary census vs. a mandatory one. The report is available as a PDF here: Testing the Use of Voluntary Methods. One of the key findings:

The use of voluntary methods had a negative impact on traditionally low response areas, that will compromise our ability to produce reliable data for these areas and for small population groups such as Blacks, Hispanics, Asians, American Indians, and Alaska Natives (see page 13). The current ACS design, based on mandatory methods, results in reduced levels of reliability for some areas and population groups due to low levels of mail response. This is a direct consequence of the additional number of cases that are subject to subsampling. In a voluntary setting, the problem is magnified. The proportion of completed interviews after mail and telephone attempts decreased to less than 50 percent for Black and Hispanic households. This has a significant impact on the total number of interviews and thus, on the reliability of estimates for these population groups.

Comments

[Rick]

Way to keep your composure Stephen-- WOW that guy was a dick.

The real issue is that without reliable census data every nonsense claim the Fraser Institute makes would degenerate into a he said she said debate rather than than being disproved directly after someone checked their data and data collection method. Worse, the conservatives could justify just about every nonsense policy they could think up using those same Frazer studies, studies that are solicited by the conservative government itself. All those "think" tanks for hire benefit directly from eliminating the mandatory census.... I wish you would have brought that up, but it would have sounded too partisan.

Regardless, this is worth the fight.... keep it up!!!

[Winston]

Niels: [The Fraser Institute] surveys every single Canadian doctor, asks them how long Canadians are waiting for care. We get a 30% response rate, and we're able to estimate how long people are waiting for care.

Stephen: How do you know you got the right answer if you only got 30%?

Niels: Excuse me?

Stephen: How do you know you have the right answer if you only have a 30% response rate?

Niels: We have every doctor in Canada that we send a survey to, and get a 30% response rate.

Stephen: How do you know you have a random sample?

Niels: Again, we send it to every single---

Stephen: Right, but only 30% respond.

Niels: That's right!

[Stephen Gordon]

The next sound you hear is my head exploding.

[Dean]

Can this 'gentleman' be any more disingenuous? I refuse to believe he's downright clueless -- a considerably more likely explanation is that he knows full well what the problem is, but he is banking on the fact that most people listening will not. Disgraceful.

[ThomasL]

@Stephen

Your point is basically that the reliability of the results demands mandatory participation.

For the sake of argument, lets grant that. Are you also contending that the value of every question demands mandatory participation?

Because the justification that "the results will be compromised otherwise" includes no upper-bound -- it is equally fit to justify a 5,000 page census form mailed to every Canadian citizen compelling them to detail every possession and experience from childhood.

[ThomasL]

I can put that better:

Arguing from the confidence of the results, even if perfectly right, admits no possibility of a question too trivial, insulting, or invasive to compel a person to answer.

[Mike Moffatt]

Stephen discussed that in the podcast - there is, in fact, a process to ensure that questions that are 'too trivial, insulting, or invasive' do not appear on the census.

[ThomasL]

@Mike

Quisnam constituo?

Certainly not me by this logic. It hardly helps if someone else decides for me what is too invasive or timewasting, since their subjective valuation of my privacy and my time versus their curiosity at the answer to any individual question is entirely beyond my control.

It treats these questions as if they were objectively so: "This question is not invasive." As if that assertion had the same quality as pointing at a blue sky and saying, "The sky is not yellow."

There is no (and can never be) such objectivity. There is no single definition of privacy where one can say, with moral truth at their back, "This question is not invasive and anyone who claims that it is is a bald liar."

Without such a thing -- which simply cannot exist -- we are back to the question. Who decides?