Industry Committee on Nov. 2nd, 2016

[Anil Arora]

First of all, Statistics Canada puts out on the web all its publications that are coming up in the future, so you can go on the website today and be able to tell what we are going to release tomorrow, what we are going to release the next day, and so on. The schedule of releases is transparent. It's there. All those releases get posted as of 8:30 in the morning, every single morning. If there is a technical issue—and it could be for all sorts of reasons, I suppose—that release wouldn't come out on that particular day. I think Canadians would know that there was some sort of an issue, and that does happen from time to time, for various reasons. There could be power outages, there could be technical issues, there could be—

[Ben Lobb, Conservative, Huron—Bruce, ON]

One question I had for you is on one of the other concerns Mr. Smith and his team had. They were concerned about the number of projects, in his words, going red. Is that the same system that you would use, maybe a colour coordination for issues, and are there currently any issues, any projects, between you and Shared Services Canada that are currently in the red at this time?

[Anil Arora]

I bring in a certain way of looking at things and managing things. For example, these are just made-up numbers. On a weekly basis, let's say, I need another terabyte or what have you, and then there are certain peaks when, let's say, we're going to get international trade data from our colleagues in the U.S., which is going to increase the requirement by a petabyte of space, for example. We've tracked those requirements for the amount of space and the associated capacity in terms of server power and so on very carefully, and we've said, okay, at this rate of growth, with the programs as we anticipate, with the demand that we have and given our current capacity, at what point do we start to become exposed to risk?

[Ben Lobb, Conservative, Huron—Bruce, ON]

Is it your position today that there are no projects behind or in the red between Stats Canada and Shared Services Canada?

[Anil Arora]

If we don't do what we're doing now, which is increasing the capacity, we see that the further out we go, as more data comes in and the capacity isn't there, obviously we would start to now see some increased risk.

[Ben Lobb, Conservative, Huron—Bruce, ON]

Yes, I think people who operate data centres or corporations that are data centres would have a pretty sharp criticism of Shared Services Canada, if that's their issue. In the real world, that is not an issue. You can ramp up and ramp down as you see fit. I'll leave that as my own edification there on that. We understand there are no projects currently in the red, and we may follow up with that later on.

Does Shared Services Canada have an effective veto on any of the information that is in their storage right now for Stats Canada to pull out as they see fit? Is there any issue right now with the data that is stored and your ability to pull it out and use it or retrieve it at any time you like?

[The Chair, Liberal, Dan Ruimy]

Time is up, but I will allow you very briefly to answer the question.

[Anil Arora]

They have absolutely no say. As I've said up front, we've actually put it in the agreement that they have no say in terms of what we have on there or what happens to it.

[Ben Lobb, Conservative, Huron—Bruce, ON]

Prior to that, did they? Prior to the agreement, did they?

[Anil Arora]

I would suspect not. I have no evidence to suggest that they did.

[The Chair, Liberal, Dan Ruimy]

Thank you.

We're going to move to Mr. Masse. You have seven minutes.

[Brian Masse, NDP, Windsor West, ON]

Thank you, Mr. Chair.

Thank you for being here.

One of the first things I want to touch on is a statistic that came up in terms of the response rate. I was part of the complete count in the year 2000 as a city councillor, because my riding, Windsor West, had a 54% response rate to the previous censuses.

When you say “response rate”, are those censuses that were completed to the full, and then they were satisfactorily input into data, in terms of the 80% that you said was the response rate?

[Anil Arora]

Yes. The response rate is calculated based on the total number of questionnaires that are distributed to occupied dwellings, and from that frame, the number of forms that are received.

[Brian Masse, NDP, Windsor West, ON]

Okay, “received” meaning completed, done, statistically entered, so 98% of the people of Windsor West whom I represented, despite our languages and the cultural differences, responded to the request that was sent.

[Anil Arora]

You have to break that down. The 98% is a global response rate for all of Canada for both the short form and the long form. There's a differential between the long form and the short form in this census, and it is marginal—

[Brian Masse, NDP, Windsor West, ON]

Okay, but you're here saying it's a 98% response rate.

[Anil Arora]

The second thing is, of course, that the response rate varies by geography. I wouldn't know off the top of my head what the response rate was for Windsor. For many communities, the response rate may be 100% and in other cases it may be a bit lower, but overall the response rate was 98%.

[Brian Masse, NDP, Windsor West, ON]

I'd like to have a breakdown by riding. I think that the stats should be provided to members of Parliament to show the interest of the area where the people are. I know Windsor was targeted. I have a hard time believing in that 98%, so I'm assuming that in testifying here today, you're saying that people replied to the census and it was entered into the data system as a 98% overall response rate of good, usable data. That's not just responding in general, but this is good, usable data that can be equated equally.

[Anil Arora]

I repeat, the response rate that is being quoted by Statistics Canada as 98% is essentially the questionnaires we got back within the universe of the questionnaires we sent out. There are differences, as I said, between the long form and the short form and within the different geographies.

Even within the questionnaires themselves, of course, there is item non-response, and that comes out as we put out all the different statistics on labour mobility or immigration or language or whatever. What happens is that for certain cells you can have fewer responses than you do for others. When those data are released, the quality identifiers for each of those.... There's also sampling, as you know, for the long form; not every Canadian household fills that out. All the responses, as well as that sampling, are taken into account, and a quality identifier is released with the information.

[Brian Masse, NDP, Windsor West, ON]

I think that's rather interesting for a scientific approach to data. The use of a response rate at that number is quite misleading, and the usable data, as well.

I will move, Mr. Chair, that the clerk invite former chief statisticians Mr. Munir Sheikh and Mr. Wayne Smith to testify before the Standing Committee on Innovation, Science and Technology.

[The Chair, Liberal, Dan Ruimy]

Do you have that notice of motion? We'll pass it on to the clerk. Thank you.

[Brian Masse, NDP, Windsor West, ON]

It's a motion, Mr. Chair. It doesn't require 48 hours. I move the motion.

[The Chair, Liberal, Dan Ruimy]

Just one second.

[Brian Masse, NDP, Windsor West, ON]

Well, I'm moving the motion. I said I'm moving the motion, and I'm moving the motion.

[The Chair, Liberal, Dan Ruimy]

You can say it, but it doesn't relate to what we're here for today, so this is—