Governing in the Dark: Good policy-making requires reliable statistical data
TheStar.com – opinion/editorialopinion
October 27, 2012. Susan McDaniel
How to lie with statistics is a popular phrase that went viral before going viral became a concept. Truth is, lying, or at least misrepresenting, is made easier when there are fewer statistics. Machiavelli understood this well, as do his modern political offspring — much to our peril as citizens.
Time was, back before 2006 when the current federal government came to power, that evidence-based policy-making, as well as evidence-based practice in medicine, nursing, law, management and most other endeavours, was the guiding principle. It’s simple really: Try to base policy and practice on what we know about the problem at hand and what we know about what works to solve it. No point prescribing something in medicine or policy for a problem that is disappearing on its own, for example, or prescribing something that makes a problem worse.
Even as a minority government, the Harper Conservatives gave little welcome to evidence, scientific or statistical. In July 2010, before the majority, the long-form census hit the dustbin. Why? The government argued that the questions asked infringed on the privacy of Canadians. In fact, there had been only a handful of official complaints about privacy invasion, by asking questions such as how many bathrooms or bedrooms you have in your house, information readily available on the Internet for all houses for sale to anyone interested.
In the to-and-fro of who said what, the chief statistician in 2010 felt he had to resign. His two-word response when asked, after he left office, whether a voluntary survey could replace the long-form census, was: “It cannot.”
Some statements at that time suggested Statistics Canada thought it was more independent of government oversight than it indeed was, or should be. That and subsequent double cuts that hit Statistics Canada have led to beliefs that the truths statistics tell may not be so truthful. Statistics Canada was doubly cut in that it faced a slice from its budget as large as several other government departments. However, added to those cuts were surveys funded by other departments whose budget cuts no longer allow them to fund surveys.
Why does this matter for Canadians? Just as scientific evidence matters to our well-being, so does statistical data and analysis. Statistical data does three things:
• It tells us whether we are making progress or not.
• It identifies the kinds of changes that make things better for us as individuals and for society.
• When government policy is being considered or evaluated, it tells us about policy effectiveness.
The long-form census debacle plus the double cuts to Statistics Canada have raised questions for Canadians about statistical data. One question is unwarranted but still asked: Can Statistics Canada data be trusted as much now as in the past if the agency is seen as just another government department? The answer clearly is yes. Statistics Canada remains an unbiased, trusted, arms-length from politics official statistical agency. What data it collects is reliable and dependable.
That said, the data it collects and its capacity to analyze that data has been diminished, and Canadians are the worse for it. We can no longer assess our progress on either the economic or social front to the extent we should in 2012.
This means we cannot identify the kinds of changes we as individuals or as a society might make to improve. It also is increasingly clear that its statistical analyses do not inform policy as much as they should. One glaring example is the crime trend analysis that shows crime declining, yet the policy agenda includes building more prisons. Instead of evidence-based policies or practices, we seem to have the image at least of policy-based evidence-making, where anecdotes, assumptions and power take precedence over careful analysis and statistical trends.
All said, Statistics Canada, rated as the best statistical agency in the world, is diminished but still strong. It can still tell us about ourselves, but not in as fulsome and cutting edge a way as it could.
The diminishment of Statistics Canada diminishes Canada and Canadians.
Susan McDaniel, Canada Research Chair in Global Population and Life Course and Professor of Sociology, University of Lethbridge, is a long-time user of Statistics Canada data. She has previously been on faculty at the Universities of Alberta and Waterloo.
< http://www.thestar.com/opinion/editorialopinion/article/1278417–governing-in-the-dark-good-policy-making-requires-reliable-statistical-data >