Evidence of meeting #61 for Industry, Science and Technology in the 40th Parliament, 3rd Session. (The original version is on Parliament’s site, as are the minutes.) The winning word was data.
A recording is available from Parliament.
Former Chief Statistician of Canada, Statistics Canada, As an Individual
--unpaid and unremunerated, and I make myself available to advise on technical issues--non-management, non-directional, just purely technical issues--to anybody who wants to pick my brains since I have--
Conservative
Former Chief Statistician of Canada, Statistics Canada, As an Individual
--50-some years of experience.
Conservative
Mike Lake Conservative Edmonton—Mill Woods—Beaumont, AB
Fair enough; is there anything specific you're doing to ensure the success of the national household survey or the census?
Former Chief Statistician of Canada, Statistics Canada, As an Individual
I think Statistics Canada at this point is doing everything that in my view it can reasonably do to make the census as successful as possible under the circumstances.
Conservative
Mike Lake Conservative Edmonton—Mill Woods—Beaumont, AB
Do I have less than a minute left, Mr. Chair?
I'm going to go back to Mr. Smith, if I could.
Maybe you could elaborate in terms of the national household survey what steps for follow-up are involved in the process?
Chief Statistician, Statistics Canada
There will be an initial mailing. We will be doing follow-ups as well. Ultimately, if don't get a response from the household, we will be taking a sample of non-respondents and going to the door hoping to speak to the respondent. If possible, we might interview them at the door. We will return multiple times until we actually succeed in speaking to the respondent. If the respondent is there and initially refuses to complete the NHS, we will go back one more time in an effort to persuade them. If at that point they refuse again, we will stop any follow-up efforts.
We're hoping through this technique that we will be able, if you will, for what I might describe as a core sample, to push the response rates to a very high level, which will help us deal with some of the issues of response bias.
Conservative
The Chair Conservative David Sweet
Thank you very much, Mr. Smith and Mr. Lake.
Now we'll go to the Bloc.
Monsieur Cardin, vouz avez cinq minutes.
Bloc
Serge Cardin Bloc Sherbrooke, QC
Thank you, Mr. Chair.
Good afternoon, gentlemen, and welcome.
Let me say first that I sympathize with you about what I call the mess created by the Harper government, not the government of Canada. There is no other word for it.
During the debate on the census--it was extensively debated and we heard witnesses about it--, the Conservatives stated that it would be pointless to impose a mandatory long-form census where fines, and even jail terms, might be inflicted to single-parent mothers. However, we know very well that the short-form census is also mandatory and that people might also be jailed if they refused to answer. So, their explanation does not hold water. Otherwise, they would have made the same decision for the short-form census by making it voluntary and withdrawing the penalties.
Previously, the short-form census was sent to 80% of households, and the long-form one, to 20%. Now, the voluntary long-form census will be sent to 30% of households. To what percentage of households will the short-form census be sent? I missed the figures you gave in your statement.
Chief Statistician, Statistics Canada
The census questionnaire, which we call the short-form census, will be sent to 100% of the population.
Chief Statistician, Statistics Canada
Yes, it will be mandatory and will be sent to 100% of the population. Actually, to be more precise, it will be sent to 100% of households.
Bloc
Serge Cardin Bloc Sherbrooke, QC
The short-form census will be sent to 100% of households and, on top of that, the voluntary census will be sent to 30% of the population, will it not?
Chief Statistician, Statistics Canada
About one third of households will receive the national household survey, composed of 65 questions. Answering that survey will be voluntary. It will include 65 questions and will be received by one household out of three, not by 100% of the households.
Bloc
Serge Cardin Bloc Sherbrooke, QC
This means that you will hit 100% of households once, and that it will be optional for 30% of households.
Some witnesses of the conservative persuasion appeared before the committee. They were quite shocked by the compulsory nature of the long-form census and wanted it to disappear. However, when questioned, we realized that none had ever received it. They even said quite clearly that they would not fill it. This means that those people will not respond to the voluntary questionnaire. They said it was an infringement of their privacy, on top of many other reasons.
How much does it cost to encourage Canadians to fill the voluntary questionnaire? Since people will not be forced to fill it, they might be less tempted to do so. You will therefore have to spend some money to inform Canadians and to encourage them to fill the questionnaire if you want to collect useful data.
Chief Statistician, Statistics Canada
I do not agree with you. Statistics Canada has been doing voluntary surveys for many years and we regularly get response rates of 75% to 85%. Canadians do cooperate with us.
We recently surveyed Canadians to see how much cooperation we could expect from them and 85% said that they would probably respond to a survey from Statistics Canada if asked to do so.
So, I do not believe one can say there will automatically be a problem with the response rate. The results will depend somewhat on the sociopolitical context at the time of the survey. However, it is quite possible that Canadians will cooperate with us and that we will get a good response rate even with the voluntary survey.
Serge Cardin Bloc Sherbrooke, QC
What could be the effects of the changes made to the census on the international reputation of Canada relating to our ability to collect reliable data?
Chief Statistician, Statistics Canada
I was recently in New York City and had the opportunity to speak to some colleagues. I found that what happened in Canada was extremely misunderstood. I had the opportunity to explain exactly what had happened. Of course, none of my colleagues endorsed the idea of doing a census on a voluntary basis. However, they got a better understanding of what had happened. They understood that everything happened in compliance with Canadian legislation and that the inherent roles of government and of Statistics Canada had been respected. Now, they are anxious to know the results of our 2011 experience. So, I cannot say that this will really have a negative effect on the reputation of Statistics Canada, after having explained clearly what really happened.
Conservative
The Chair Conservative David Sweet
Thank you, Mr. Smith and Mr. Cardin.
Mr. Généreux, you have the floor for 5 minutes.
Conservative
Bernard Généreux Conservative Montmagny—L'Islet—Kamouraska—Rivière-du-Loup, QC
Thank you, Mr. Chair.
Thank you to the witnesses.
Mr. Fellegi, I would like you to explain why, in the census, the answers provided by some Canadians might be less good to a voluntary questionnaire than to a mandatory questionnaire? You said earlier that you will have to guess the results of the census since you have never seen such a situation in your 50 years of experience.
Let us suppose that, as a Canadian and a Quebecker, I receive the voluntary census and decide to respond. The answers I will provide will not be any different than those I would provide to a mandatory questionnaire, I believe. Do you really think the answers provided in the next census will be different than those provided in the previous mandatory census?
Former Chief Statistician of Canada, Statistics Canada, As an Individual
I didn't say that those who respond to the voluntary survey will give poorer responses. What I was saying is that the overall result is very likely to be worse.
Conservative
Former Chief Statistician of Canada, Statistics Canada, As an Individual
Because fewer people will have responded, and the ones who will have responded will not necessarily be representative of those who didn't respond. It's well known in the survey literature that certain groups of people—underprivileged mostly, but also people with very high incomes—are much less likely to respond in a voluntary survey than in a compulsory survey.
Conservative
Bernard Généreux Conservative Montmagny—L'Islet—Kamouraska—Rivière-du-Loup, QC
But how do you know that fewer people will respond?
Former Chief Statistician of Canada, Statistics Canada, As an Individual
Because that's the experience everywhere, and Statistics Canada's own planning assumption is a substantial—perhaps overstated, but substantial—deterioration of response rate. The historical response rate to the long form was 94%. The current planning assumption of Statistics Canada is 50%. That's a huge deterioration of response rate. Even if they exceed it and they get 70%, that's still a huge deterioration of response rate from 94%.