Where the truth lies…
Like most of you, I have for years spent dreary summer vacations at the beach, reading bodice-ripping page turners and discussing just how early is too early for cocktail hour. This summer, however, I have been at my desk reviewing first-year elementary statistics and the reliability 0f voluntary survey data. What a refreshing change!
Honestly, I have been absolutely shocked at the sustained outcry over the Harper government’s cancellation of the long-form census. I could hardly rouse myself to care about statistics on the eve of my first-year exam in the subject. And yet, the announcement has blown up into a political debacle that has dominated headlines and discussions for weeks. When it comes to arcane policy discussions, methods of obtaining reliable statistical data might be tough to beat. Surely Harper’s strategists figured that such a change, announced during the height of summer doldrums, would pass unnoticed. I mean, who cares?
Well, it turns out LOTS of people care. People from a cross-section of Canada – business leaders, policy makers, academics. All of whom rely on the detailed data of the long form census to make crucial decisions in their sectors. The result has been nearly universal – and vocal – condemnation. The decision is such bad policy, and undermines Statistics Canada’s mandate so severely, that our Chief Statistician was compelled to resign.
Even still, why should we care?
There is the obvious reason that the long-form census provides unrivaled data for decision making by industry and policy-makers, and a fount of information to academics and other research groups across the country. Despite the Industry Minister’s insistence to the contrary, a voluntary survey cannot achieve the statistical rigour of a mandatory census. And good data is essential, especially in the much-vaunted knowledge economy. Data mining is a cornerstone of this new economy, and its effective performance requires good, reliable data. And it didn’t get much more reliable than StatsCan long survey census data.
The government’s specious argument that the survey is too intrusive is undermined by the data – only two Canadians complained to the privacy commissioner about filling out the 2006 census. As reported, that’s 0.0000067 of Canadians, which even with my elementary knowledge, is pretty statistically insignificant. The numbers don’t back up the policy.
Indeed, there’s the rub. Numbers don’t always back up policy. Justifying policy decisions with data is what we should expect from a transparent, accountable government. Unfortunately, it doesn’t always work that way. For instance, Statistics Canada (again! so damn inconvenient!) released numbers last week showing that the rates for both crime and violent crime specifically have declined consistently during the last decade – dropping 17% and 22% since 1999 respectively. Nonetheless, the government has introduced Bill C-13, “tough on crime” legislation that will reportedly nearly double the cost of administering prisons, at an additional cost of nearly $5-billion (!) annually. How does Minister of Public Safety Vic Toews justify the urgency of such a bill? By citing an abstract need to “keep Canadians safe”.
This isn’t simply a case of ideology trumping data. Indeed, as Paul Saurette, Assistant Professor of Political Studies at the University of Ottawa, argues persuasively (sorry – forgot the link. It’s a great article which you should read here), it is evidence of this government’s embrace of “epistemological populism”. As he writes:
[epistemological populism] 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.
This strain of thinking has been increasingly evident in politics. This is especially true in the United States, where conservative radio talk show hosts rant about “elites” and “eggheads” who don’t understand what regular people just know to be true. Here in Canada, those who cite the need for accurate data are, in the conservative think-tank Fraser Institute’s estimation, “vested interest groups” and “elites”. Satirical conservative pundit Stephen Colbert captured the essence of epistemological populism in his White House correspondents’ dinner speech:
That’s where the truth lies, right down here in the gut. Do you know you have more nerve endings in your gut than you have in your head? You can look it up. Now, I know some of you are going to say, “I did look it up, and that’s not true.” That’s ’cause you looked it up in a book. Next time, look it up in your gut. I did. My gut tells me that’s how our nervous system works.
This is why the long form census decision is big news. This is why people are up in arms. The long form census is the best data we have in this country, and is essential to understanding our society. But relying on data encourages – indeed, should necessitate – evidence-based decision making. But data can be inconvenient, and may contradict our assumptions (just ask any disappointed graduate student). Instead of changing assumptions, epistemological populism encourages our prejudices and assumes that the data is wrong – we can just believe our guts. And that allows a government to do whatever it feels is right.
As a modern democracy, we must expect better from the government. We should be beyond governing by hunch and feel. Science policy works in two directions. This blog is mostly about policy for science, but science for policy is also essential. Ignoring the data when crafting policy is unconscionable. Eliminating the data altogether is unforgivable.