News and Blogs


Evidence and decision-making: bend it like Harper

288292_335843203165010_170076704_o.jpg

The Broadbent Institute is pleased to present the second in a series of blog posts by a range of Canadian academics and thought leaders critiquing the record of the Conservative government. Read the first post here

Ideologues don’t like evidence. They know what the problem is and what to do about it.

Perhaps the most egregious example of this under Stephen Harper concerns the evidence about declining crime rates and the government’s insistence on the necessity of introducing harsher sentencing criteria as part of the much-derided Bill C-10 omnibus crime bill. 

But that was simply a matter of ignoring the evidence. In my experience, the left is equally wont to ignore evidence that doesn’t fit their narratives as the right. Psychologists tell us this is a common feature of the human mind: we are more likely to take note of evidence that “fits” our preconceived narratives about how the world works and ignore the facts that don’t.

Suppressing evidence is a more serious endeavour. As evidence is suppressed, our ability to monitor what governments are doing, or what is going on in our environment, or what is happening to Canadians living in villages, towns and cities across the country erodes.

The axing of the 2011 mandatory Census by the Harper government is the most prominent, but not the only example of this suppression at work. Apart from a few libertarians, no one supported this move – not business, not labour, not the polling firms or the provincial ministries of health, education, and so forth.  The head of Statistics Canada at the time, Munir Sheikh, resigned in protest. 

The mandatory Census was the lifeblood of almost all social and business planning.  It provided key data for studying things like income inequality and poverty since both low- and high-income households were required to report. The quality of Census data and the accessibility of the data to non-government users had improved exponentially since the 1970s. As a result, its importance as a tool for monitoring the effects of government policy was on the rise. Because it collected data on such a large sample (20 percent) of Canadians, it was able to shed a light into dark corners of Canadian society that no other data source could do. 

One of the most important functions of the Census was to monitor what was happening to Canadians over time. Are current governments doing a better or worse job than their predecessors? Breaking the Census series in 2011 means we can’t answer this question any longer. Transparency was the issue, and transparency lost.

But killing the mandatory census was not the only important entry point for data suppression. The 1990s were an important period of data innovation at Statistics Canada. Longitudinal studies – of children, young adults, and the labour force – were introduced. By the 2000s, Canadian researchers were just beginning to master the complex data these surveys produced on important questions such as the duration and consequences of poverty, unemployment, and the like. The longitudinal surveys are now gone, a result of budget-cutting.

Silencing the disadvantaged

Prime Minister Harper would no doubt accuse me of wanting to “commit sociology”. And I must oblige him. The social gulf between the “comfortable” and the “disadvantaged” in Canada has always been large. The “voices” of the disadvantaged in shaping our social fabric have been fading in recent decades. Good data did not necessarily provide the disadvantaged with “voice”. It did, however, make it more difficult for the “comfortable” to ignore the presence and life circumstances of the less comfortable. 

Killing data collection, like the mandatory Census, is only one form of “datacide”. Government bureaucracies collect all sorts of information on their program operations at both the federal and provincial levels. Despite rhetoric on the government’s embrace of ‘open data’, much of it has never been readily accessible to the public. 

One organization that sought to overcome the inaccessibility of important data was the National Council of Welfare (NCW), which produced widely-read annual reports on poverty and welfare income data for several decades. Canadians like to think of themselves as “kinder and gentler” towards the poor than our southern neighbours in the U.S. The NCW helped to remind us when and where we were neither so kind nor so gentle. The underlying data still exist but the Council itself was chopped in 2012. Fortunately, the Caledon Institute has committed to continuing the series.

Reports on old age security and similar programs traditionally produced by government ministries have likewise disappeared. The response - “Sorry, no documents with the given url were found in this archive” - has become increasingly common, the result of cutbacks at Archives Canada. You can still get the data if you know who to call, but not otherwise.

Message control

“Message control” from the Prime Minister’s Office (PMO) is perhaps the most egregious form of evidence suppression we’ve witnessed under Mr. Harper’s watch. “Message control” means the evidence is there, the analysis has been done, but researchers are forbidden from relaying crucial information to the media or the public. Canada’s scientists are so frustrated with this government’s overhaul of scientific communications policies that they took to the streets, marching on Parliament Hill in the summer of 2012 to decry the “Death of Evidence” and to protest what the journal Nature referred to as “muzzling”.

Their concerns, expressed on their protest banners, followed a precise logic: “No science, no evidence, no truth, no democracy.”

John Myles is Professor Emeritus of Sociology and Senior Fellow at the School of Public Policy and Governance, University of Toronto. From 2001-20011 he was Canada Research Chair in the Social Foundations of Public Policy. He spent several periods from 1986 to 2007 as Senior Research Fellow at Statistics Canada.

The opinions expressed in this paper are those of the author, and do not necessarily reflect the views of the Broadbent Institute.