Over the past two months, there have been considerable media articles and commentary regarding an under-count of Manitoba's population. Manitobans should understand the issue and why it is important.
As chief statistician of Manitoba, I categorically state Statistics Canada has substantially underestimated Manitoba's population. The issue is about statistical methodology which has serious consequences for Manitobans. An under-counted population has significant negative impacts on fiscal transfers, resulting in the province not receiving its rightful share.
Every five years, Statistics Canada conducts a census. Before population estimates are finalized, they complete a follow-up study to estimate how many people were missed by the census. MBS's concern lies with that follow up study and statistically unusual results, which have led to an under-count of Manitoba's population.
Statistics Canada's new population estimates were released Sept. 26, 2013, and effectively delete approximately 18,000 Manitobans. Before September, Statistics Canada had stated Manitoba's 2011 population estimate to be 1,251,690. Now, they reduced it to 1,233,728.
The following is an overview of the situation.
Prior to Statistics Canada finalizing its population estimates in September, the Manitoba Bureau of Statistics along with other provincial and territorial statistical agencies were involved in a six-month population evaluation process with Statistics Canada.
This evaluation process occurs every five years. The latest process was to determine the new population series based on the 2011 Census counts adjusted for estimates of the number of persons not completing a Census form. Due to these discussions, Statistics Canada's preliminary estimates for Manitoba and the other provinces and territories were revised a number of times.
Statistics Canada's own evaluation analysis identified their estimates of Manitoba's counted and missed persons were statistically very unusual and stood far apart from estimates for other provinces and territories. In fact, the Manitoba results were extreme, having never occurred before for any province or in any time period.
MBS raised concerns on the statistical challenges of the Manitoba estimates at meetings of federal-provincial-territorial statistics officials. These were followed up with written position papers that identified the issue, as well as recommending options to correct the statistical "errors."
MBS's position, based on Statistics Canada's own evaluation analysis, concludes the likely cause of the statistical estimation concerns lies with the samples of Manitobans selected by Statistics Canada to determine the number of individuals missed by the 2011 Census.
There is strong evidence these Manitoba samples were biased and not representative of Manitobans, in that they contained too many of the individuals who tend to participate in the census and an insufficient number of persons who tend to get missed in the census (e.g., immigrants, First Nations, Metis, young adults, inner city and northern residents and non-permanent residents). This resulted in a serious underestimate of Manitoba's population.
Statistics Canada's new population estimates for Manitoba are not consistent with available independent evidence. Before the deletion of 18,000 individuals from the population estimates, Statistics Canada had estimated Manitoba's population grew by 5.7 per cent between 2006 and 2011. This is fairly close to the 6.7 per cent increase in the number of individuals filing income tax returns over the same period.
It is also consistent with the strong economic growth Manitoba experienced over this period. Statistics Canada's revised numbers for Manitoba now estimate the population grew at only 4.2 per cent over that period.
While Statistics Canada has acknowledged there are statistical issues with their population estimate for Manitoba, they have not been able to isolate the issues or "errors" and decided to make no adjustments at all to the province's 2011 population estimate. Again these "errors" are unique to Manitoba.
MBS strongly disagreed with Statistics Canada's decision. Our position is that Statistics Canada's revised population estimates for Manitoba have statistical errors and have underestimated our province's population and a resolution is required to ensure a more accurate estimate.
An MBS evaluation that brings Manitoba statistically in line with other provinces indicates that the under-count is 16,200, worth about $100 million in federal transfers, a significant amount of money.
MBS continues to have discussions with Statistics Canada on its underlying methodology. We trust there can be a resolution to Manitoba's population estimates that is reflective of the actual number of persons resident in our province.
Wilf Falk is chief statistician, Manitoba Bureau of Statistics.