A few weeks ago, Canadian scientists went to Parliament Hill to hold a protest rally that they dubbed “the Death of Evidence“. The speakers at the event outlined the effects of the federal government’s research funding cutbacks. They pointed out that research supported by government funding actually saves money, because it produces reliable evidence that helps government make sound decisions. They also emphasized the difficulty or impossibility of restarting their work if it ends.

My own research isn’t in the “hard” sciences, so the cutbacks protested at the Death of Evidence rally don’t directly harm my work. But my colleagues whose research will be affected are mad and very frustrated, because they know the value of the work they do.
And personally I’m angry about the disrespectful way that these researchers learned they were losing their funding. Many had no advance notification of the government’s intentions, and simply received a written notice informing them of the date in the near future that their funding would end. The short notice gave them almost no time to arrange alternative sources of support – if there were any  – because getting adequate research funding from any source is a lengthy and labour-intensive process.

And for those working on collaborative projects – which Canada’s federal granting agencies have encouraged researchers to get involved in – the sudden loss of funding also causes a huge problem for the other research participants, who now have to decide whether the entire project can go on without the contributions of Canadian data and expertise. In some cases, it can’t. So these decisions by Canada’s federal government are not only hurting the scientific community in Canada, but also making Canada look short-sighted and ignorant internationally.

Statistics Canada (Statscan) is the federal government agency that many researchers in my discipline rely on for data. Statscan has had several well-publicized problems in the last few years, mostly around the federal government cancelling the “long form” collection of census data. Allegedly, Canadians were complaining about this form being too intrusive –  they actually weren’t – and the chief statistician resigned in protest against this blatant interference with Statscan’s professional judgement and autonomy.

Now, Statscan has announced that over 30 of its surveys and data collection projects will be cancelled or severely cut back “to meet budgetary saving target[s]”. A lot of people think that all Statscan produces is the Canadian census, so I want to discuss two of the data collection programs being eliminated – the Survey of Labour and Income Dynamics and the Salaries and Salary Scales of Full-time Teaching Staff at Canadian Universities – to demonstrate the immense value of Statscan’s other work. I have seen in my own work how useful these particular sets of information are, and am very worried about the implications of losing both.

An example from 2009 of the data collected by the Survey of Labour and Income Dynamics, showing the differences in employment status between individuals with and without disabilities. (Credit: Statistics Canada)

The Survey of Labour and Income Dynamics (SLID) has been running since 1970. As the name suggests, it collects information on income, employment experiences, education, and demographics (including households as well as individuals). As you can imagine, having data from as far back in 1970 is incredibly useful in being able to track patterns and movement in these areas across time. Another really useful feature of these data is that the same respondents are interviewed for a period of six years. So identifying meaningful relationships between variables like education and employment is even easier, because the same people are being followed.

Statistics Canada assures us that it will “will continue to conduct a survey to produce annual estimates on income” in place of SLID. But income is only one of the three very broad areas that SLID addresses. What will replace the SLID data on employment and education? For my own research interests, I particularly fear the (apparent) loss of data on union membership – which, given the federal government’s negative attitude toward unions and information on unions, maybe isn’t too much of a surprise. Perhaps the reasoning is that if people don’t have information about unions, they won’t know enough about them to want to join one.

The Salaries and Salary Scales of Full-time Teaching Staff at Canadian Universities data collection has been going since 1970. In addition to salaries, it collects information on “age, gender, previous employment, education, citizenship, type of appointment, subject taught and year of appointment”. Now this is a data set that I am biased toward because of where I work, but this information is valuable in so many ways. For example, universities and faculty unions, or individual faculty members, can use it to have informed contract negotiations. Also, students considering academic careers can get an idea of salaries and employment opportunities in the disciplines they want to study. And, as with SLID, the length of time over which these data have been collected make it possible to track patterns in these data as they develop and change, as well as enabling researchers to take “snapshots” of what’s happening during individual periods or dates.

Statistics Canada indicates that these data will no longer be collected. At all. So where will this information now come from? There are some “industry” organizations, such as the Canadian Federation of Business School Deans, that collect some of these data for their particular discipline. But there certainly aren’t enough organizations doing this sort of work to be able to replace all of the extensive data across disciplines that Statscan now collects. And since the Bureau of Labor Statistics (Statscan’s American counterpart) has data showing that post-secondary teaching is one of the fastest-growing occupations in the US, I’m surprised that a federal government that’s all about job creation wouldn’t want more, rather than less, information about an area with great potential for job growth.

In making these criticisms, I want to be clear that I am not criticizing the staff at Statistics Canada. I have used Statscan’s services many times during my research, and I have always had the pleasure of dealing with intelligent, enthusiastic people who are really excited by their work and are eager to help others use Statscan’s data thoughtfully and fully. What I am criticizing is whatever level of authority in the federal government made the decision that Statscan should discontinue these 30-plus programs – I can’t honestly imagine a statistician not realizing the value of the data that will be lost – or that placed such restrictive financial constraints on Statscan to force it into downsizing or discontinuing  important work. These  cuts at Statistics Canada are, sadly, further evidence of the federal government’s reckless disregard for well-informed decision-making – which we should all be very concerned about.