In a knowledge-based economy, we need more, not less, evidence-based policy design.
Only rarely does the Canada West Foundation as an organization express a view on a policy issue. The norm is to speak through our research reports and their authors, and to avoid endorsing any particular policy options or views. However, the debate over the long-form census is an exception, a case where the issues seem so important that they call for an institutional response. Thus in a variety of forums, the Foundation has joined with other Canadian think tanks in calling upon Prime Minister Stephen Harper’s government to reverse its decision to make the long-form census voluntary.
Why has the Canada West Foundation taken up this crusade? In part, the answer comes from our methodological experience and expertise; we know that a voluntary census will not produce reliable data. This is not an opinion; it is as close to a fact as one can get in the complex world of statistical analysis. The government’s argument that a voluntary census will work if mailed to more people and backed up by a massive advertising campaign is simply wrong. We will pay more for less, which seems an odd strategy in this fiscal environment.
But does it matter if data quality is corrupted? I would argue that in a knowledge-based economy, paying more for poorer data makes no sense. We need more rather than less evidence-based policy design, and the movement away from the current long-form census is movement towards policy impoverishment. It is also a signal to the world that we don’t take ourselves seriously, that we’re content to rely on hearsay and guesswork.
All of this, of course, may seem rather abstract, so let me give you a concrete illustration from the Foundation’s work. Part of what we do involves mapping patterns of demographic change in a highly dynamic, growth-driven region of the country. We know that the region has been transformed fundamentally since the census in 1971, the year the Foundation was created, and we believe that understanding and mapping that transformation is fundamentally important in coming to grips with the future of the West in Canada and within an increasingly competitive global economy. Self-knowledge is not everything, but it is an important start.
However, if we move to the short-form census, either alone or coupled with a deeply flawed voluntary long-form census, we will no longer be able to map the demographic evolution of the West. When asked to describe how our linguistic and ethnic diversity is changing, to map the migration of Aboriginal peoples into urban centres, to understand patterns of integration, assimilation, and migration for new Canadians, we will have to fall back on guesswork and assumptions. I can’t believe that we will settle, that we should settle, for “by guess and by golly” when we can do so much better.
Of course, none of this means that the status quo should be exempt from criticism. We could certainly eliminate the threat of jail for failing to complete the census, a threat that has never been carried out in any event. We can and should review the content of the long form. And, if we’re smart, we can further insulate the census operation from political influence.
In short, we can do better, but the long-form census is one baby that should not be thrown out with the bath water.