Oh no! Not the Census Again!

In his testimony before the Industry Committee today, Ivan Fellegi, the former and universally respected head of Statistics Canada, expressed pleasant surprise at how strongly people in almost every sector of Canadian society have come out in support of the census and the importance of the information produced through the mandatory long form.  For many, this must seem like a beltway issue, technical and remote from their everyday lives.  If Canadians not directly involved are attending at all during these summer days, many must be asking, “what’s the fuss all about?”

Some of the defenders of the decision to scrap the long-form census will try to portray those opposed as pursuing their own narrow interests or as elites who are out of touch with what really matters to most Canadians.  But the issue continues to play and the loose and disparate alliance trying to save the long form continues to grow because this decision does touch issues that matter for Canada, issues at the heart of the role of government, the nature of knowledge, and the integrity of our public institutions (although I will save this latter issue for a separate post).

Here is a bit of a breakdown of what’s at play:

1) Is the mandatory long-form questionnaire unduly intrusive and coercive?

The long form census is mandatory and therefore by definition depends on the coercive power of the state.   Statisticians will explain that a census must be mandatory if it is to be useful.  This is everywhere the case.  Is it reasonable and proper to search for and even test options that are less coercive?  Of course.  I expect that this is ongoing.  But let’s  be clear, no workable alternative has yet been  proposed.

The voluntary approach doesn’t cut it.   Yes, Statistics Canada uses sample surveys for some purposes but they can do so only because their census data allow them to assess the selection bias and correct accordingly.  We know that some categories of people are more likely to fill out questionnaires than others and without the census base we cannot judge the level of distortion this creates.  This cannot be corrected by sample size.  Whatever the sample size we would not know how to assess the selection bias, putting the data in serious doubt.  No statistician has yet, as far as I know, disputed this nor can I see how they could.  That means that the census must be mandatory and that means the threat of sanctions.

Are the penalties unduly coercive?  I could do no better than ask you to read the analysis of the distinguished economist and Professor Emeritus of The University of British Columbia., W.T. Stanbury who concludes that the threat of fines and up to 3 months jail time – the kind of penalties associated with”summary” or administrative offences – is “mild” and appropriate to the purposes at hand.    In fact, no one has ever been jailed for failure to fill out the survey.  Others have done an excellent job of  pointing to the many areas of our life in which governments seem fully willing to threaten fines or even jail, often in the face of significant opposition, and for disputable ends.  But the ends in this case do not really seem much in dispute – vital information for governing, managing the economy and accountability   There is a lot of room to dispute what penalties would constitute the least restrictive approach necessary to ensure high compliance but what is clear right now, given the options at hand, is that only a mandatory census, with penalties for failure to comply with the law, can deliver the goods.

What about the intrusiveness of the questions themselves and the privacy right?  Few have complained to the Privacy Commissioner who has worked closely with Statistics Canada to address any privacy issues and this Agency, charged with protecting our right to privacy, seems satisfied.  Apparently more have complained through political channels and some to Statistics Canada directly, though a very small proportion of those who have had to fill out the census.   We should not dismiss such concerns.  It’s not hard to imagine that some who have come here from oppressive regimes would be wary.  No doubt some others hate the questions or the process for a range of reasons.  I, like many,  dislike forms and answering surveys and this kind of reaction is one of the reasons that response rates to voluntary surveys are going down.  And, as trust in government generally is in decline, some would say sharp decline,  the anxieties about such surveys and how the information might be used probably increase.  That is precisely why Statistics Canada and the government should do all they can to assure Canadians that here in Canada we can take great pride that the privacy of this information is extremely well protected and that there has not been one single breach of a citizen’s privacy in the entire history of the census.

The Scandinavian countries which no longer use a census have a much more powerful and more intrusive approach.   Denmark, for example, has three major registers, one for population, one for business and one for housing.  These are impressive and detailed  – and require significant compulsory reporting.  If such registers existed in Canada I have no doubt Statistics Canada would happily rely on them but creating such registers would be an enormous and expensive task, especially in a federation, and I am not convinced that Canadians would, in the end, prefer this.  As for Britain, which is apparently considering alternatives, they have given themselves over a decade to sort out what might work, if anything.

Of course, it is legitimate to review the questions to make sure that each can be justified. That is why the government must always approve the questions before Canadians are surveyed. Some questions which may, on their face,  seem useless or unnecessary,  may in fact address important needs.  For example, while it may seem strange to ask at what time one leaves for work, this information is crucial for assessing and addressing traffic challenges.  Other questions, however, may have outlived their usefulness.

The Industry Committee heard some proposed changes that would still produce reliable data, including dropping some questions, and reconsidering the penalties.  The census is not immune from change.  It has been changed in the past.  But the survey should be mandatory and most of the questions continue to serve important needs.   Requiring  twenty percent of Canadians to answer these questions every five years means, for most of us,  thirty minutes of irritation once in our lifetime.  We require relatively little of our citizens – is this too much?

2) Are the data really that important?

Every level of government uses the survey results to determine the needs of Canadians, to make sure that resources are spent wisely and to assess what works and for whom.  Statistics are essential for good government and this is the crux of the issue.  Some libertarians, critical of the government in other respects,  have come out in support of the government’s decision here because, as they explain it,  statistics enable governments to intervene  and they simply do not want government doing what they do.  They want as little government as possible as a matter of ideology.  They do not want public health care.  They do not want social services.  They do not want public intervention on the environment.  They do not want the government interfering in the economy.  But most Canadians do want some, much or all of this and will resist any direct attack on such services. The libertarians, then, generally welcome any measure that chokes off the tax revenues and information necessary for governments to act in these areas.

But even those who might be sympathetic to the cries for less government will want whatever government is left standing to work well and efficiently and surely we can all agree that, in a democracy, we must be able to hold our governments accountable for what they choose to do or not to do, for how they spend and for how they cut.  Any government policy – more government, less government, different government – needs to be tested against its consequences for the country and its citizens and, for this, we need reliable data.

I would add that many of those who want less government will concede that local, private and voluntary initiative will have to fill the gap to the extent that it can.  Not surprisingly, business, voluntary and philanthropic organizations too are nervous about how they will make their investment decisions absent these data.  And would we want to limit the capacity of provincial and municipal governments?

3) Is knowledge important in charting our course?

It seems rather strange to be asking this question at all, but in a fascinating article in The Mark News,  political scientist Paul Saurette points to an anti-expert, anti-knowledge strand in the conservative movement.  This view, which he labels “epistemological populism”, can be traced to the oft-cited economist hero of libertarians, Friedrich Von Hayek.  Saurette writes, “I call it epistemological populism because it is a theory of knowledge that assumes that the most reliable and trustworthy type of knowledge is the direct individual experience of ‘common’ people – the lessons of which can be unproblematically universalized. In such a theory, the more numerical, general, and statistical the analysis, the less trustworthy it is. For as we all know, our own eyes never lie but numbers can say whatever they want them to say.”  Of course, experience, conviction and common sense matter.  And a healthy skepticism is a good thing. No one owns certain knowledge.  But dismissing experts as elites and denying the value of systematic knowledge is dangerous  to democracy, to accountable government and to our everyday lives.   Our health practitioners, for example, have been among those most vocal about their dependence on high quality census data.  More generally, in an increasingly competitive world with new crises seemingly coming faster, do we really want to trust our future entirely to intuition, fate or the workings of the free market?  The truth may be elusive and facts may only be one path, but we are in big trouble when we treat the truth – and the facts – as the enemy.

The government has indicated that it understands  the importance of the information and has found a “balanced” approach.  The experts say it won’t work and have offered up a compromise that might.  The issue is too important not to find a solution.

3 Responses to “Oh no! Not the Census Again!”
  1. Beijing York says:

    Thanks for some excellent and informative posts about the census debacle, Alex. The government’s framing of this issue has me gobsmacked. There is an ugly element of fear mongering going on: punishment; invasion of privacy; and to hear Stockwell Day, opportunity for racist attacks.

    Having worked as a policy analyst with the feds and a consultant with both non-profits and private sector businesses, I can attest to how valuable census data is for any sound analysis. Both my last two projects, one on youth crime prevention and one on socio-economic impact assessment related to a resource development project, have required use of census data to understand and describe past and present conditions. You cannot make projections without a baseline.

    What was somewhat shocking to discover is how many people I had to work with dismissed stats in general or the concept of ensuring reliability when producing statistical analysis. I think there are a frightening number of Canadians who dismiss statistics and social sciences as being some kind of elitist voodoo at worst or a replication of the intuitive obvious at best. Harper has managed to tap into that sentiment to make this fly and the fact that he would knowingly trump government efficiency to pander to such votes and destroy the government’s ability to operate effectively is extremely distressing.

  2. johannes says:

    Hi Alex

    Fear not – it seems leaked documents suggest a solution.



  3. Thank you johannes you give us a new information about the new census form

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: