In the first week of February 2015, the House of Commons will vote on Kingston MP Ted Hsu's private member's bill calling for restoration of Canada's long-form census. Although research methods and data analysis are not ordinarily popular or newsworthy topics, there has been considerable media and public attention to the details of how Canada has turned its data clock back to the year 1956, or 1871, or 1666, depending on how you look at it.
The deficits of the voluntary National Household Survey (NHS) with which the Harper government replaced the respected long-form census are now glaringly apparent. Thus, Hsu's bill, An Act to Amend the Statistics Act, is supported by many commentators from health care, urban planning, poverty, business, the labour force, economics and other fields as well as experts in statistics and data analysis.
As someone who has worked on early childhood policy since Stephen Harper was in high school, I decry the new data deficit and support the call to restore the long-form census without delay.
In light of the inadequate data from the methodologically flawed NHS, today it would be hard to find anyone who knows much about data collection and analysis who would argue that the abolition of what was Canada's primary data collection instrument has been anything other than foolhardy. Indeed, many consider the abolition of the long-form census to be ideologically driven, especially in light of the clear evidence of the NHS' weaknesses.
As a result of the cancellation, Canadian policymakers, labour force forecasters, researchers and advocates are flying blind whether they're concerned with housing, public health, immigration, education, business, employment or the labour force. I've been among those monitoring the situation, shaking my head over what (several) columnists at Maclean's called the "statistical illiteracy" of the minister responsible and in despair about the implications of the new data deficit.
Last week child care was identified in the Globe and Mail by Tavia Grant as one of the policy areas damaged by Canada's post-2010 data void. She observed that "It's [now] more difficult to plan subsidized child care." I'm not sure that Grant knows (as many don't) how rare systematic planning is vis-à-vis Canadian child care where a free market model rules. Moving to a publicly planned approach at all levels of government from reliance on the market is a key piece of the national child-care policy for which my colleagues and I have been advocating for donkey's years.
Ironically, as a support to the coherent planning that would be part-and-parcel of the more publicly managed child-care system Canada demonstrably needs, we've also been advocating for more, better, and more focused data for many years. Thus, abolition of the long-form census in 2010, which produced all sorts of key demographic data supporting such child care planning as has been practiced by some local and provincial/territorial governments was a real shock and -- as Grant identifies -- damaging.
Data at the community and neighbourhood level about income and family composition is fundamental for child care planning as is reliable data about families with children in rural communities, young children with disabilities, Aboriginal children in remote communities and so on. Without this kind of fundamental information, it's difficult to identify need or demand for services, grapple with affordability issues or understand whether existing policy and services are "working" or not, thus whether public dollars are well spent or adequate.
As well, it should be noted that numbers aren't just numbers. Data becomes useful and meaningful when it is considered in context and can yield comparisons -- groups with other groups, regions with other regions, neighbourhoods with neighbourhoods. It is only when reliable data in which analysts have confidence is available to "slice and dice" that meaningful information emerges to support policymaking, research and advocacy. Thus, as US top executive Carly Fiorina has observed, "The goal is to turn data into information, and information into insight".
One good illustration of the new data deficit for child care: data no longer available about children with disabilities came from a survey called the PALS (Participation and Activity Limitation Survey). This "postcensal" survey was linked to the long-form census and was valuable for shaping, assessing and advocating for initiatives aimed at inclusion of children with disabilities in child-care services. StatsCan now guardedly admits that the survey that replaced the PALS -- based on the new National Household Survey -- may have issues of "data accuracy," "measurement error" and "under coverage." Oh, and as it isn't comparable to the PALS, there's no data continuity to help assess or understand change. Oh, and children aren't included in the new survey in any case.
Canadian conceptualization about child care has evolved a great deal in the last two decades. Today there is considerable knowledge to support the idea that families and children will continue to be badly served unless there is a commitment to robust public policy that builds a real system. A resurrected long-form census isn't the whole of the data package needed to inform building this system but it is an important fundamental source for some of the key information needed for good child-care policy-making.
Its cancellation leaves multiple gaps in our knowledge and nothing to build on if -- when -- a federal government turns its hand to working with provinces/territories to build the national child care program families from coast-to-coast-to-coast desperately need.
One reason to support resurrecting the long-form census is that -- depending on the outcome of the next federal election -- we could be as close as 2015 to finally beginning to build that national child care program. For this, we'll need good data, and good data should begin with a reliable, well-designed national census that includes all Canadians.