The Great Canadian Census Debate: The Economists Call It

It looks like we can call it a day on the Great Canadian Mandatory Long-Form Census Debate: the economists have weighed in, and they think the government's wrong, wrong, wrong!

Today’s Globe and Mail reports that 76 percent of economists surveyed by the Canadian Association for Business Economics say that it’s a bad idea to scrap the long-form census. Out of 252 economists surveyed, only 14 thought that it was a good policy. Of course one of these 14 was the Fraser Institute’s Niels Veldhuis, who has demonstrated a less-than-encouraging understanding of statistical analysis in his creative defence of the government’s position (Check out Stephen Gordon's merciless takedown of the Fraser Institute).

Does this mean we can tentatively conclude that about five percent of economists don’t understand statistics? Or, as Gordon might put it, that they are not part of "the community of evidence-based policy analysts"?

But the truly bizarre finding from this poll? That 30 economists surveyed didn’t know whether or not scrapping the mandatory long-form census was a good idea.

Really? In a field that worships numbers and statistical analyses, in a debate that has galvanized the Canadian research community, 30 economists didn’t know whether scrapping the long-form census is a good idea or not? That’s 30 economists who either haven’t been paying attention to the policy-wonk equivalent of a monthlong Lollapalooza festival, or who can’t be bothered to recall their first-year stats training.

Can anyone explain this? Are we witnessing the birth of a new subfield of economics that rejects the possibility of knowledge through statistical analysis? (I hope so; a postmodern turn in economics would be great fun.) Were these 30 economists actually sociologists in disguise? Enquiring minds, etc., etc.

What the Census debacle can tell us about governmental accountability

Jeffrey Simpson nails it today when he notes that the scrapping of the mandatory long-form census is a “temporary triumph over ideology.” (Well, one can hope that any such triumph would be temporary, but I’m feeling pessimistic today.)

This whole census mess raises another point that I haven’t seen discussed much. Namely, we’re about to find out what, if anything, can convince a Canadian government to change its mind on something.

Not just this government. The Harper Conservatives may be pushing the limits as to what is possible in our Parliamentary system (Exhibit A: choosing prorogation, rather than face a vote of confidence in the House), but they’re not breaking any laws. The powers that they’re using are available to any government, doubly so for a majority government. Custom and tradition are no match for someone with the ability and will to ruthlessly use the rules to their advantage.

On most issues, you can find reputable people supporting one side or the other (yes, even in copyright, despite the rhetoric). This census debate is different because of the nearly unprecedented diversity of voices opposing the decision: business groups and NGOs, provinces and territories, all statisticians, pretty much every economist and social scientist I can think of, Statistics Canada itself. All serious think tanks with even a basic understanding or respect for statistics and facts (which would exclude the Fraser Institute, based on the comment reported here, which would have gotten a failing grade in any introductory statistics course) are against the decision.

On the other side, you have Stephen Harper.

So, what might cause Harper to change his mind, especially if, as Simpson writes, this decision is based on ideology and not facts?

I can think of three things that could convince a government that prefers ideological arguments to rational, fact-based ones. (Hint: facts won’t do it.) The first is that the opposition could tie up any changes in committees, which are controlled by the opposition because they have the majority of seats in Parliament. Of course, all you have to do is introduce changes via regulation to get by that one, and that’s what we’ve seen here.

The second is a worry that the issue would hurt them in an general election. In theory, minority governments are susceptible to this type of pressure, but a vote of no confidence is like a nuclear bomb: the opposition can’t deploy it to thwart every thing they don’t like, and there’s always the possibility that that bomb might (pardon the pun) blow up in their faces if they lose the election. However, if all this unpleasantness rubs enough voters the wrong way, then the government might back down.

The third reason they might back down is if the party’s financial backers threaten to withdraw funding. However, I understand that the Conservative party’s funding now largely comes from individual donors. As a result, I think that any collapse in party revenues would be related to a drop in Conservative support.

Unlike the U.S. political system, which was designed to avoid concentrating excessive power in the hands of one person, the Canadian system has no such checks and balances. Previously, an independent public service was seen as a check on the government, as was the Governor General. But that’s tradition, not a hard and fast rule. At the end of the day, the only thing holding any Canadian government in check is fears about an upcoming election. If you’re the government, no fears = no worries.

What should give supporters of all parties pause is that these constraints will be much, much weaker for a majority government of any party: Liberals and New Democrats are no more or less virtuous than Conservatives. The farther you are from an election, the freer you are to do whatever you want, evidence and opposition be damned.

As someone who’s kind of a fan of popular control of one’s government, I find that even more worrying than the scrapping of the mandatory long-form census.

Post-script: As I write this, Donald Savoie is talking about this very issue on The Current (available here soonish). I’ll have to pick up his latest book when I finish my dissertation’s first draft.