Public Accounts Committee on May 7th, 2014

[Michael Ferguson]

I think the user of this type of data would be more of a researcher or analyst trying to identify where there might be vacancies, perhaps at the municipal level and that sort of thing. But certainly we did identify that the information is really still limited to job vacancies at the national, provincial, and territorial level, and not at the local level. So anybody who is trying to do research about where there might be job vacancies from a policy point of view or understanding the economics of a particular situation wouldn't know where those job vacancies existed.

[Malcolm Allen, NDP, Welland, ON]

I appreciate the clarification. It's kind of at a macro policy level. But ultimately if the government is actually using that as a policy tool to encourage people to go somewhere, they really can't tell you where to go except to go to Alberta. I'm not asking you to comment on that. That's a conjecture that I'm making, but it's what I think I just heard, that I wouldn't necessarily use this stuff personally to go look for a job. But if the policy folks are writing and saying, “By the way, there are jobs available in Alberta”, that's about all they could tell me.

Let me go to paragraphs 8.53 and 8.52, because I was interested in them, especially where they talk about the national household survey data and about 25% of the census subdivisions not being any good due to quality concerns. The additional piece is that the data for an additional 686 census subdivisions, or 15%, was not released for confidentiality reasons. Is that a cumulative total, 25% plus 15%? Or is there a subset somewhere inside that 15%?

[Michael Ferguson]

It's an accumulative total. The 686 additional census subdivisions, where the information wasn't released for confidentiality reasons, were mostly places where there would be a very small population. That's why there is the confidentiality concern. In terms of census divisions, it is additive: 25% and 15%. However, in terms of the impact on the total 3% of the population for which information was not released, it wouldn't have moved that 3% very much at all because these are very sparsely populated areas.

[Malcolm Allen, NDP, Welland, ON]

I appreciate that statistically that wouldn't have moved the number of three, which is three times higher than when we used to have the long form census. That is a pretty extreme magnitude, quite frankly. We're not talking doubling or a half, but we're talking about three times as many. So, it was 1% in 2006 according to you when we had a long form census. It's 3%, which is three times the magnitude. That's a pretty high number. But, clearly for small subdivisions, and I used to live in a small community, data was important to us even if we were a small community when we were trying to make decisions. So even for a reeve in a small place, they still need data.

However, let me draw us back to the comment that's made in paragraph 8.52 which is that when we had the old census, 94% in 2006 were returned, versus the new one at only 69%. Not only do we get less data, it seems, if I'm reading correctly from the new one, but we also get three times more unusable data, the 3% of the population. Is that a fair and accurate description of those paragraphs?

[Michael Ferguson]

Well, I think I need to clarify that. Yes, the response rate went from 94% to 69%. Statistics Canada anticipated that there would be a decrease in the response rate. So they did increased the number of people they actually surveyed. So the 94% response rate was tied to three million surveys and the 69% rate was tied to 4.5 million surveys. So, at the end of the day, the number of responses actually received was pretty similar, because when you apply the response rate to the number of surveys sent out, it was similar. But what happened was that the response rate was not sufficient in certain areas and, therefore, they didn't have enough information on about 3% of the population. So the issue wasn't so much the number of responses they got back, but from where those responses came.

[The Chair, NDP, David Christopherson]

I'm sorry, but your time has expired. Thank you.

Over to Mr. Woodworth: pick up where you left off last time, sir.

[Stephen Woodworth, Conservative, Kitchener Centre, ON]

Thank you, Mr. Chair.

It's quite appropriate, because I'll also pick up where Mr. Allen just left off, because he's touching on exactly what I was trying to get at.

First of all, the 3% figure for the population with unavailable data is an increase from 2006, when it was approximately 1%. Correct?

[Michael Ferguson]

That's right.

[Stephen Woodworth, Conservative, Kitchener Centre, ON]

So even in 2006 there must have been some deficiency in the small area and sub-population data, correct?

[Michael Ferguson]

It was related to 1% of the population at that time.

[Stephen Woodworth, Conservative, Kitchener Centre, ON]

Again, just to be very clear—you just stated this, but I want to be very clear about it—in 2011 there were responses from about 3 million people across the country or about 200,000 more responses than in 2006. Correct?

[Michael Ferguson]

I think if you do the math, comparing 69% of 4.5 million to 94% of 3 million, you get a number that would indicate that there were roughly 200,000 more in 2011.

[Stephen Woodworth, Conservative, Kitchener Centre, ON]

All right.

So those percentages are different because, in fact, the surveys were received by 50% more households in 2011 than in 2006—in other words, by 4.5 million households rather than only 3 million households in 2006. Correct?

[Michael Ferguson]

That's right. Again, that was because they anticipated that there would be a change in response rate and knew that they had to increase the number of surveys they were to send out.

[Stephen Woodworth, Conservative, Kitchener Centre, ON]

You have not anywhere assessed the cost of providing that small area and subpopulation data, correct?

[Michael Ferguson]

That's correct. I think the only indication of cost in here other than the overall cost of producing data was that Statistics Canada did receive additional money to manage the change from the census to the national household survey.

[Stephen Woodworth, Conservative, Kitchener Centre, ON]

In terms of bridging that 3% gap, there's nothing in your report that would tell us how much it would cost to resource that.

[Michael Ferguson]

That's correct.

[Stephen Woodworth, Conservative, Kitchener Centre, ON]

On the other side of the cost-benefit equation, you also have not assessed what might be the economic value of that 3% subpopulation in small area data to users. Is that correct?

[Michael Ferguson]

That's correct. No, we haven't.

[Stephen Woodworth, Conservative, Kitchener Centre, ON]

In fact your recommendations suggest consulting with the users, and that might be one area where Statistics Canada should consult with the users just to determine the value to them of that data. Correct?

[Michael Ferguson]

Absolutely. It would be to determine what they are looking for, what the value of that is, and then they would need to determine the cost of collecting the data.

[Stephen Woodworth, Conservative, Kitchener Centre, ON]

As I understand it, and I don't know if this came from your report, or maybe it was the department's response, in November 2013, Statistics Canada did in fact initiate a more extensive dialogue with regional and local governments and agencies somewhat along this line. Am I reading that correctly?

[Michael Ferguson]

That is what the agency has said in their response, so I can't speak to exactly what they have done.

[Stephen Woodworth, Conservative, Kitchener Centre, ON]

But if they do that, that would be an appropriate ongoing further course of action in your view, would it not?