Skip to main content

The Globe and Mail

Is census data usable? 'Our thinking has evolved,' chief statistician says

Wayne Smith, who replaced Munir Sheikh as Canada's chief statistician during the census controversy, gives an interview in his Ottawa office on Feb. 11, 2011.

Dave Chan/dave chan The Globe and Mail

The following is a full transcript of an interview with Wayne Smith, who was officially named Statistics Canada's new chief statistician last month. He took over from Munir Sheikh, who resigned last summer amid the controversy over the Harper government's decision to cancel the mandatory long-form census. The Conservatives defended their decision on grounds it was intrusive and coercive to force one-fifth of Canadian households to answer a detailed list of 40-plus questions on their home, work life and ethnicity.

The short-form census of 10 questions is still mandatory though and Statscan is moving ahead with voluntary long-form survey as a replacements. Critics say it will yield poorer data and will fail to accurately measure smaller sub-groups in the population who may just ignore the survey, leading to problems for users of the data, from governments to researchers to businesses.

Mr. Smith, however, said there's no scientific basis for predicting the optional survey will yield poorer data, in part because it will be sent to a wider group of Canadians: one-third of households instead of one-fifth.

Story continues below advertisement

The interview took place at the agency's Ottawa headquarters on Feb. 11, 2010.

STEVEN CHASE: I suppose if you asked your users to vote whether they wanted the long-form census or the NHS, they'd pick the long form census. I guess we're curious about what we're going to end up with now given we've got a survey that is not a census.

What do you expect will be the best you can hope for in terms of the response rate [for the NHS]

WAYNE SMITH: I guess the answer is it depends. And it really depends on our efforts to engage Canadians and to convince them of the importance of the census and the National Household Survey.

We've used the planning assumption of 50 per cent [response rate] Our plans have been based on achieving a 50-per-cent response rate.

We've never done a survey on this scale, on a voluntary basis before. We haven't needed to do a survey on a voluntary basis in a relatively charged environment, where people have taken strong views and positions in the past.

So we're in unexplored territory.

Story continues below advertisement

Fifty per cent therefore was our planning assumption but it is possible to do – in our ongoing voluntary surveys, we typically see response rates significantly higher than that.

We decided not to be ambitious and set the target higher than the 50 per cent. We'll see what happens.

What happens is going to depend ...

The procedures and everything else are in place. We're using essentially the same infrastructure, the same kind of approach we used in 2006.

The census, in terms of the content, the length, the methods that we're using, the systems that we're using, the staff that we are using, the method of collection we're using, are pretty much equivalent.

The only difference that's different in 2011 for the National Household Survey is that the survey is voluntary and the sample is bigger.

Story continues below advertisement

What was the response rate for the long form census [in 2006]

Ninety-four [per cent]

But 50 per cent I would emphasize is a planning assumption and possibly we will do much better.

But we've been using that [50 per cent target]in establishing our plans.

So we're actively engaged ...

To some extent it is now in the hands of Canadians.

Will Canadians participate? Will they respond voluntarily?

And that, to a significant extent, is in the hands of Canadian institutions.

Will municipal governments, provincial governments ... will Canadian businesses, will ethnic organizations ... will health organizations across Canada come out in support of not only the 2011 census but also the National Household Survey?

When you say institutions need to support it, do you mean doing their own advertising, like community ...

It can be as simple as ... You can think of the employees of those organizations, you can think of the stakeholders in the organization, you can think of the members in those organizations; you can think of the clients of those organizations.

Organizations have methods of communicating with all of those people. So it might be in an employee newsletter carrying an article about how important the census is for their particular organization, how the data is used.

We've had municipalities issue proclamations --already there are proclamations being issued by municipalities across the country, indicating their support of the census.

In the past, support has been expressed in the form of census beer. It's been painted on the side of trucks: census messages.

All of those things are part of getting that message across that says 'This is important: we need your support'."

When the minister announced the change – to the NHS – or at least answered questions about the change, he said that $30-million had been allocated to help advertise ... to help spend money on getting better response rates. That money seems to have gone to other things including postage and printing.

So my question is how much are you spending [on boosting the response rate] on extra efforts: on those second visits or third visits to the door?

It's not an easy question to answer. Ultimately it depends on what various pieces of the process cost and how much money is available. But let me do an effort at it.

We had $630-million to conduct the census and the National Household Survey.

The government has give us $30-million additional dollars. Of that amount, $15-million is planned to be a contingency to deal with any shortfall in response rates in either census or the National Household Survey.

The purpose of that money is to send people to doorsteps. It's not for advertising. It's not for communications. It's to actually send enumerators to the door trying to get questionnaires.

There will be an advertising campaign in support of the census portion of the exercise. We can't advertise for the NHS at the same time because the collection periods overlap. And the messages are different: this one's mandatory and this one's voluntary.

So we're going to do a limited amount of advertising mainly in ethnic newspapers ...

With the NHS the effort is going to be vested primarily in the doorstep.

How much have you budgeted in advertising for the NHS?

I don't know the money. It's not a large amount of money, but I don't know it offhand. I don't even know if we've got an estimate right at the moment about what the precise amount is.

The other money – the other $15-million is largely to address additional costs associated with the change in plan from going to an integrated mandatory census to moving to ... that's what it was given to.

But effectively we have the global budget [$660-million]and the budget will be used to accomplish the mission ... That entire amount of money is available to ensure the success of the National Household Survey and the census of population.

What are you doing for 2016 [the next census]

We've been asked for 2016 to look at options and to come back to the government with options for 2016 and to look at them very comprehensively and do it in collaboration with the national statistics council.

If you think of the debate over the last few months, a number of options have been out there and discussed. Could Canada do a census based entirely on administrative records. Could Canada do a census like France where they do a part of the country every year and they don't do the country in one year. Could we do something like the United States where they have a decennial census once every 10 years and run a very large survey every year in the interim that is capable of producing smaller area data. That's another model that's been talked about.

So data mining?

Data mining is a version of using administrative data. Basically you think of any administrative file that exists. What some countries are doing – for example, Denmark or Finland – if you live in those countries, you have to register your address. If you move you have to register. If you want to have a job you have to be registered. They actually have a basic register that says: here are the people living in the country and here is where they are living and this information is current and accurate.

And once they have done that, then they have all these other files: these people are in the education system; these people have cars, these people are in the health system ... You could think of all the files the government holds: they link those file to the basic population register and achieve something that is a very rich data base. There are certain types of data it doesn't contain obviously but it contains a very rich set of data.

That is their method for conducting a census.

Is that viable in Canada? Would Canadians even conceive of doing something like that? Those are issues that have to be addressed.

Back to the census we have now. This process [the NHS]will obviously be more labour intensive and that's where the $15-million comes in.

Maybe. Remember the $15-million is a contingency amount. If we get to the response rates that we need we may never spend that money.

What are the response rates you need then?

The targets are 98 per cent on the census and 50 per cent on the National Household Survey.

But the 50 per cent ... one national number doesn't do it for us. We need the response to be fairly uniform across all geographic areas and fairly uniform across all population groups.

So we're not aiming necessarily for exactly 50 per cent and then we stop when we get there. If the response rate in B.C. is 25 per cent and it's 75 per cent in Quebec, we will need to continue.

So it's the notion of trying to get to a relatively uniform rate of at least 50 per cent. If the response and participation is very high it is possible we may go well beyond 50 per cent without ever spending the $15-million.

As can you can imagine, we got a crash course in census response rates [last summer]and methodology while combing through all those documents [released under access to information] One of the documents last year said we expect a 50 per cent [initial response rate]and said if we go back [to the doors] we could bump it to 65 to 73 per cent or something like that.

Oh these things were in the access to information [release to media]... When we do an access to information request, we put a lot of documents out, and there are things that are well founded and there are things that are not ...

This was a brief to the minister's office.

A brief to the minister's office.

It was a March 2010 e-mail to Mr. Clement's office, a senior StatsCan official advised that a self-administered voluntary survey would yield an initial response rate of only 50 per cent. She added that with follow up work and sufficient resources the agency could push the response rate up to 65 or 70 per cent.

I just want to be clear on that.

Our planning assumption is 50 per cent and our target is 50 per cent. Is it possible that we could get to 65, 75 per cent? Absolutely. That statement was early days in the thinking about what we felt we could say competently.

So if you have a response rate of 50 per cent versus 94 per cent before and you're asking largely the same questions – give or take the shift in the language questions – what is the erosion?

How much do you lose in terms of richness of data?

We should talk about that fairly thoroughly, actually. As I talked about a minute ago, the major difference between the National Household Survey and the 2006 census long form is essentially that the survey is voluntary and the sample is bigger.

The one thing we know with absolutely certainty is the response rate going to fall from making the survey voluntary.

We haven't done it before this way so we don't know by how much. Fifty [per cent]is a very conservative number. We're likely to get there relatively easily I would think -- but you never know. It depends on the environment.

If the response falls to 50 per cent, there's one inevitable consequence of that. And that's exactly why the sample has been bumped to one in three from one in five. To compensate for that effect to the extent possible.

At a 50-per-cent response rate with a one in three sample, the sampling error would still be somewhat worse in the National Household Survey than it was in the 2006 census, the long form.

If the response is actually 60 per cent on the NHS it's a wash. They are basically just as accurate, one as the other.

And if it goes to 70 pct actually the NHS sampling error is smaller than it was in 2006.

So it depends on the outcome.

If we can get Canadians to participate and get the response rates higher -- from the sampling error perspective, we can do as well or better.

Is that even for small populations [subgroups]

The same principle applies even in small populations. If the response rate goes up, with a higher sample, I get just as many people responding at 60 per cent in any given area as I did in the 2006 census.

But you're assuming the behaviour of the respondents is similar across different demographics.

That's the other big issue. That's the next piece. The next piece is non-response error. I mean think about it. People publish survey data all the time and you've probably never heard of non-response error until this discussion.

You learn about it in polling. Right?

In polling ... their actual response rate is like 6 per cent or 10 per cent or 25 per cent. The exposure to non-response error is very high in these polls and very high in other non-survey research – the risk of it.

I invite you to talk to other experts in statistics, but the reality is that there is no necessary link between the response rate and having a degree of non-response bias that actually makes the data less usable and unfit for use.

No but it was the compulsory nature of the other one [the mandatory 2006 census long form]that helped.

The only way the compulsory nature helped ...

If you think about it we had 94-per-cent response rate in 2006. That means 6 per cent of the population didn't respond. That's 2 million people. You talk about low income people. That's a very sizeable piece of potential non-response bias. If you look at the 2006 census technical evaluation papers, there are populations with significant non-response bias. Young males are traditionally undercounted in the census. Recent immigrants are traditionally undercounted.

The only survey that isn't exposed to non-response bias is one with a 100 per cent response rate.

You make it seem like there's no reason why Mr. [Munir]Sheikh, in his parting note, said...

No. I'm ...

He said that the National Household Survey is not a substitute for the census long form.

The issue is risk. If we could go in with a guarantee of a 96-per-cent response rate, the risk, the exposure is substantially reduced.

If we have a 50 per cent response rate, there is greater risk.

But there is no guarantee this data will not be usable. There is no guarantee it will be subject to major non-response bias beyond the levels we've traditionally seen in the census. We'll see. It may happen but you can't say before it starts that it's going to happen.

Are you telling me that we're going to get just as many Inuit, just as many unemployed, just as many people who don't speak English and are immigrants.

I am saying that either I nor anybody else can tell you today that we're not.

But it's a fair bet.

No it's not a fair bet. There is no scientific reason why you would say that before it even starts, before I see the results, that there's going to necessarily be a significant problem with the count of Inuit or Métis or immigrants beyond the levels we've already seen in the 2006 census.

So you're saying there is no reason to believe right now that there might be a poorer reading of small sub-population groups?

There's no necessary reason why that would be the case. There's a heightened level of risk and that's the most you can actually say. Then the rest of it will depend on what happens and that's why it's so vitally important we get the support.

You're hoping that business groups, community groups [support this] How do you ensure they do?

We've hired a very large communications staff regionally. And they're already out there, they're actually out there meeting with associations, we've met with Indian reserves, we've met with municipalities, we've met with a wide variety of organizations. And so far, I can tell you that the reaction has been very favourable. The indications are that we're going to get support.

You can't tell me the budget you've had to add to your communications budget to promote the NHS?

We don't actually have to add anything because we were going anyway, we were going for the census. We were going to do this for the census. So all this extra cost of going to this municipal government and saying, we need your support for the census and for the NHS and sending them just for the census is trivial.

So you've got $660-million budget.


And I assume then by rule of thumb that about a fifth of that would be for the NHS then?

Remember, it doesn't work that way, because we're sharing infrastructure. This enumerator that I'm going to send to the door, the training of that enumerator, it's the same enumerator, they'll be out there in certain periods, they will be collecting both, how do I decide who that belongs to?

But there must be incremental costs for going out again four weeks later.

Yes, but I could potentially tell you what our estimated cost is from the time that we stopped the census...but that won't tell you the cost of the NHS, it will only tell you how much did we spend after we stopped, how much did we think we were going to spend.

The whole infrastructure is shared. And that's why, we're not trying to be difficult or hide anything. The infrastructure is actually shared between the census and the NHS. You can't really pull apart the costs of the two. Certain costs, yes, we can tell you how much when we're finished. We'll be able to tell you how much we spent on printing a questionnaire, because there really are distinct costs. The communications program, I'm not going to be able to separate it because it's the same people going to one place and talking about both of them together. I'm not going to be able to tell you how much the Internet infrastructure costs because we're using the same...

So you're not putting in any extra effort in trying to reach those small population groups, which typically had low response rates.

They've always had traditionally low response rates, so we've always been going there.

Who are they, by the way?

The ones that we worry about are youth – young men particularly. Recent immigrants, big problem. The low income – we have to pay attention to areas where there is a high concentration of low income. That tends to be... if you think of Ottawa, that tends to be urban core. Therefore urban core municipalities, things that we can do in the core. Those are the kinds of groups that we traditionally have problems with.

Non permanent residents – people who aren't citizens of Canada, but are counted, they're another group we have problems with. We can approach them through areas where they traditionally work. A lot of them are students

But you're not putting a lot of extra efforts..

No. Because the same problem has always existed. So we're using the same mechanisms and the same processes to do that.

So the short form is out in early May and then it's the household survey, except for the North?

In the North we're doing both together. We've already started. I was briefed the other day. In the North in remote areas, we do 100 per cent on the NHS, it's not a sample, we do every household. In the 700 that we've done already we have a 96-per-cent response rate on both the census and the NHS.

When you say you do 100 per cent, why is that?

If we sampled in these small communities, we wouldn't be able to publish data. It's also done by enumerators who actually do an interview, we're going in and we're flying people in and out, so we've started.

In the beginning of May, we'll invite Canadians to respond by Internet, we'll send the information they need to be able to respond. We've already chosen the sample for the NHS, so if you're in the NHS sample, when you go on our website to respond to the census, we will also offer you the opportunity to complete the NHS at the same time. So as early as the beginning of May, we'll start getting some responses to the NHS. But the major mail-out for the NHS is at beginning of June, that's when we start the intense operations.

You've said in your own internal newsletter in July you said the results of the NHS will "of course never be comparable to census data." So what do you expect are the differences?

The major difference is that potentially we're going to see a larger sampling error, and we're at great risk of non-response error, because of the change in methodology between the two, we may see some difficulties in comparability between the two. But we've put a document on our website to kind of illustrate those issues and the magnitude of those issues potentially. And now we're doing everything that we can possibly do to ensure that...based on that information. The reason we did those studies was to inform our planning so that we could improve relative to what those numbers show.

I understood what they were trying to do there, they were trying to set the baseline for what the problems were. And then move backwards on that.

We're doing our best to, but at the end of the day, I don't know, you don't know, nobody knows what's going to happen until we've done it.

But we do know it's significant or else you would have trusted the language question to the long-form census.

The issue there was that there was an explicit legal requirement that those questions be in the census.

The government of Canada, though, fought and made the opposite case. Their legal arguments before the court were that they could fulfill those on the long-form survey.

No, no. There were two questions that were moved, if you look at official language act, under the official languages act, there are regulations and the regulations try and determine where federal departments have to provide services bilingually. Those regulations specific a method of calculation, the method of calculation explicitly refers to decennial census data from basically these three questions: the mother-tongue question, the official languages question, the home language question.

So in those regulations, that information is required from the decennial census, not from the national household survey, the decennial census in order for those regulations to be applied. And those were the reasons for moving the questions is because there was an explicit legal requirement, not a notion that people felt they were entitled, because of some issue, something under law that entitled them to more information. There was explicit reference in law to those questions, and that's why it was felt - we would have had to change the official languages act and its regulations if we didn't put those questions on the census.

Why did the government's lawyers go to court and fight it?

That wasn't the fight. If you recall, before the court case ever went to court, we'd already removed those questions. There would have been nothing to argue about. The court case was about the people who were contesting, saying, that's not enough.

What are your instructions to staff on people who refuse to fill out the short form?

Our instructions are to basically, we are going to apply the law in the same way we've applied it in every previous census in whatever form the law is at that time. There are amendments before the House of Commons that may modify it. Nobody is proposing any amendment that would remove a penalty of some kind for refusal to respond to the census -- not the NHS. In our communications to Canadians and our communications to our staff, our approach has always been, to try and persuade people on the merits of the case to participate in the census.

We don't go to the doorstep and say, 'hi, I'm from the census, you haven't sent your form in, and you're going to going to be going to jail or be fined if you don't do that.' You go to the door and say, 'I'm from Statistics Canada, we've noticed you haven't yet returned your census form, I'd like to explain to you why this is so important.' We try and persuade them, and only if the person is absolutely determined to refuse, will we make them aware that there penalties under the Statistics Act. The only change that we're making is that we're going to refer to 'penalties', because we don't know what those penalties will be.

I understand the government hasn't actually introduced those changes yet, so we're still waiting for that. In some of the discussions last summer, there was talk of how we should change the way we report this information so there is no possibility of comparison to previous numbers because it's apples and oranges.

Our thinking has evolved.

The question is, how will you report things and how will you allow people to make comparisons to previous areas? How will you do that?

A lot of things are written in emails in an organization. Our intention is, and we've had some discussion recently, there's no reason to conclude there's a problem with the data. Our intention is to publish the data in essentially in the same way we published the data in 2006 census, we'll publish for the same geographic areas, we'll publish for the same variables, subject to the questions being the same. We'll have similar services and products and services. The only situations in which Statistics Canada will not proactively publish information, is where we know beyond the shadow of a doubt there is a problem that makes the data unusable. Otherwise, we will publish the data and we will tell people everything that we know about the quality with the data. So if we think there may be some problems or some issues, we will provide that information.

And the sampling error?

We have never traditionally provided the sampling error from the census. In fact that was one of the big surprises that we had recently was discovering how high the sampling error was on some of the data we published in 2006. We'll certainly give some indications on the sampling error. But if we've done comparisons to 2006 or have other information that would guide us, to suggest that this data is simply wrong and not only wrong, but massively wrong, we won't publish it. It will still be available if somebody asks for it, because under Access to Information, if the data isn't confidential, we have to provide it anyway. We will not publish information that we have clear evidence presents a problem. And you may be surprised to learn that we have had problems like that in the past. In the 1986 census we had asked a question about Aboriginal origin, where the data was fundamentally flawed, and we ultimately didn't disseminate it because we had problems with the response. So it happens.

To the extent that we know we've got a problem that in our view makes the data unusable, we won't publish it. But otherwise we will publish it and people will be able to make comparisons to the 2006 census and previous censuses based on the data.

I'm sure users would have preferred to keep the long-form census. I'm sure you're aware of the concerns raised by different groups, whether it's researchers or it's health researchers. In some of the briefs we got from Access to Information, HRSDC told StatsCan that the less reliable data would compromise their ability to determine EI eligibility. Indian and Northern Affairs says it would effectively hurt their ability to effectively manage and evaluate performance in areas of health and housing. The province of Ontario, saying we need this for our tax and budget decisions.

Can you assure these people that they have no reason to worry?

No, I told you a minute ago. Nobody knows what's ultimately going to happen. And I told you there's heightened risk because of the lower response rate. There's not an a priori reason, that doesn't mean it can't happen. So we won't know until we're out the other end. But neither does anybody else. All I'm saying is that nobody can draw firm conclusions that this data is going to be fundamentally flawed. There's no scientific basis for that. We will not know until we come out the other end. I understand that people perceive the higher risk, and they would rather that the risk wasn't there. And the perception is that if the long-form had remained mandatory, that risk would not be there, or would be smaller actually. And I understand that I don't know either and I can't give assurances because I won't know what I've got until we've actually gone thru this process. We've never done it this way before. All I'm saying, all I'm asking Canadians to do is to suspend judgment because there's no scientific basis for saying that the data is going to be fundamentally flawed.

No scientific basis in that nobody's a fortune teller, is that what you're saying.

I can sit down, and we can work out for you, for any given variable, for any given city, based on various response rates, what happens to the sampling error if the response rate falls. There's a necessary consequence. If the response rate is this, this is the effect. If the response rate is that, this is the effect. That's science. But people saying there is going to be massive under-coverage of certain populations, to such an extent that the data is ... there's nothing in statistical theory that supports that argument. You won't know until it's done.

You've said you hope StatsCan will still be the world's leading statistical agency we've collectively built it to be. In the interim, we've seen international groups weigh in -- the European statisticians [who criticized the scrapping of the mandatory long-form]

The group you're referring to ... it's the advisory committee to Eurostat. They wrote a general report, and in the context of that report they made a couple of statements regarding Statistics Canada. There were maybe two or three sentences in total about Statistics Canada -- and two of them contained errors of fact.

I pointed that out to Mr. Akerholm [chair of the European Statistical Governance Advisory Board]and suggested he might want to inform people who received the report because he was in error. He was in error because he said the government had used dormant powers, to make a decision. That wasn't true. Not only did they always have the power, it was never dormant, they've always actively used it. Every government who approved the census, has taken it very seriously and more than one has intervened in ways to ask for changes to be made. This isn't new. To say it's a dormant power is completely misleading.

Second thing is, he said that Dr. Sheikh had resigned because he disagreed with the government's decision, that's not what Dr. Sheikh said.

Their point is that countries should look at this and make sure that statistical agencies are legislatively independent. That was their point.

They're very unclear on their point, unless the argument is that the government should have no role in determining whether the census should be conducted, and what the content of the census is. And is that a realistic view of the world?

I don't know, it's not my world, it's your world. I'm just saying that they said it should be a lesson that other countries should make sure they are legislatively independent.

The international community has expressed views without being well informed. I think that what's really quite surprising is that the world has upheld Statistics Canada as a model statistical agency even though this has been the way censuses have been approved since this agency has been created. Nothing new happened this year.

Well, they moved from a long-form census to an NHS, to a survey.

So what you're saying is Statistics Canada was a model until the government actually did something and exercised its power in a way that some people didn't like.

I agree, except that I can't find one user that actually likes this.

I'm not saying you can. I'm not trying to convince you that there are people out there who are users of the data who would say that we would prefer this change or process or this method. I'm not saying that, I'm not trying to convince you of that. All I'm saying is that I think some of the statements that have made that have dismissed ... as fundamentally flawed and unusable, the results of the national household survey before we've even conducted it, are not based on sound science. They're not based on sound science.

Point taken. Just to go back to the beginning, 2016, you were asked to look into this, what's the point of this, are you thinking of moving entirely away from the census.

I think the government wants to step back, there was an interesting debate last fall and we had other people saying Canada should be looking seriously about a register-based census, and other people talking about the U.S. model, and other discussions about, that there are ways to even improve on the model that we're operating in 2011, and saying that it behooves us and the government wants to step back and say, okay, let's look at those other models, what is possible in Canada?

People have suggested that if we could make a register census work in Canada, we could save buckets of money and avoid annoying a whole bunch of Canadians by asking them to fill out forms. They want us to look at those models.

The American model seems to promise more timely data, every year you're getting new data. They run a huge survey called the American Community Survey. And they run it in such a way that if you add up the data, they disperse the sample across the entire country, if you add up the data for five years you can actually publish estimates for very small geographic levels, and since they're doing it on a continuous basis, every year, you can publish new estimates based on the most recent five years of data. It's a rolling census in a way.

That promises more timely data but there's also some serious cost implications in doing that.

So everyone's saying we've heard some really interesting stuff about possible alternate scenarios for the census, so let's step back, look at what happens in 2011, what other countries are doing, other scenarios that people can imagine and make a decision for 2016 on what we want to do.

Is this a request from the minister's office, or from a minister?

It's from the government.

So we could move away entirely from a census then on the short form, that's one of the options?

Any option that's looked at is potentially an option that could be adopted. Realistically, from a statistician's point of view, the register-based census is enormously attractive. You get all this data, and you don't have to spend any money, and you get it every year.

But the reality in Canada is that unless you have this population register that people have to keep up to date, you always know where they're living, you can't make it work in Canada.

And the privacy issues around linking all these data are huge. So yes, that's an option that we need to understand and look at and consider the pros and cons. Is it likely? Not very likely.

Do you think it was overblown, what happened last summer?

I think we had a very important debate. The government made a decision based on its values and priorities and Canadians reacted to it and raised all kinds of issues and concerns. And the government heard that and has made a decision to maintain that course of choice. But it was a very important debate and a lot of information was put on the table that will inform future decision-making regarding the census.

I wouldn't say it's overblown. Some elements of the debate are somewhat misleading in the way they've been presented to people. Mainly the non-response biases, that's there's no way this can succeed, that's unsupportable scientifically. The notion that's out there that people who are poor or recent immigrants refuse to respond. The problem with those groups isn't that they refuse to respond, the problem is that we never get to them in the first place.

Let me be more precise, refusal maybe is part of the problem, but it's not the entire problem, and in my view it's not even the largest piece of the problem. The problem is we don't find them. If you think about, we have a problem with young males. Two guys are sharing an apartment, we send the questionnaire, one guy fills it out, dashes it off, sends it in, and we never get the other one. Recent immigrant just arrived from their home country, they move in with another family while they're looking for a place to live for a while. Our enumerator shows up, we wind up getting a questionnaire from one of the two households and not the other one.

In some cities there's a problem with illegal basement suites. These are people renting out their basements...obviously people living in these suites are likely to be people who are poor or recent immigrants, students, young males. We go to the door, knock on the door, the people in the house don't want to tell us about the basement suites.

So it's true that in voluntary surveys, young males, recent immigrants, low-income people are underrepresented. We have not thoroughly assessed that in the 2006 census, I suspect that's equally true in the 2006 census that these people were under-represented. But the reason isn't because that we're going to the door, and they're saying go away, I don't want to answer your questions. The reason is because we're not finding them in the first place.

Remember I was talking about these two guys living in the apartment together. If I had a question back from that household, it's over, I'm not going back there. So the mechanisms that cause those problems are only partially based on whether the survey's voluntary or not. And you can't really conclude that the problems are going to be any worse in 2011 than they were in 2006 based on any scientific theory. There's a higher risk because of the lower response rate, but that's it.

I was going to ask you the obvious question: Would you prefer the old system to this one?

Obviously, I'm a public servant...

Has it made your job harder?

It's challenging, obviously. We have changed course and changing course is obviously more complex than staying on the course you were initially on. But we have changed course very nimbly and we were ready. In terms of our ability to conduct this, we are ready. We have the systems, we've tested the systems, we're hiring the people, we've got the logistics in place, we're already operating, there's no problem from my perspective in terms of us being able to conduct this census. Relative to 2006, we had huge problems, we had huge problems hiring in some parts of the country in 2006, we're not seeing that this time around.

The issue now, is, to a significant extent, in the hands of Canadians. Will they participate? If they participate in large numbers, if people encourage them to participate, if they participate across the country and relatively uniformly, we can get very good data out of this survey.

Report an error Editorial code of conduct Licensing Options
As of December 20, 2017, we have temporarily removed commenting from our articles. We hope to have this resolved by the end of January 2018. Thank you for your patience. If you are looking to give feedback on our new site, please send it along to If you want to write a letter to the editor, please forward to