Skip to main content

An Evaluation of the Accuracy of Statistics Canada’s Annual Population Growth Estimates for Ontario’s Greater Golden Horseshoe and Component Municipalities

By: Dr. Frank Clayton with research assistance provided by Jodee Ng

August 31, 2017

Bottom Line

We cannot give a grade higher than a ‘B-’ at this time to Statistics Canada for the accuracy of its estimates of the distribution of annual population growth for municipalities located within the Greater Golden Horseshoe of Ontario between 2011 and 2016. While the estimates of the distribution of population growth among municipalities are close to the actual counts from the 2016 Census of Canada for many municipalities, there are some notable disparities. These include (a) overestimating population growth in the city of Toronto and Peel region; (b) underestimating growth in the regions of Waterloo, Niagara, York and Halton; and (c) underestimating the share of GGH growth in the Outer Ring.

1.    Background

For those of us who monitor economic, demographic and real estate market trends at a local level, the annual population estimates prepared by Statistics Canada are an essential tool. The Centre for Urban Research and Land Development (CUR) has released a number of analyses of demographic patterns in the Greater Golden Horseshoe (GGH) and its economic heartland, the Greater Toronto Area (GTA), primarily relying on the population estimates for the years since the population counts had been released from the 2011 Census of Canada. [1]

Statistics Canada prepares estimates of population (total, by age, and components of growth) each year (as of July 1) for census divisions across Canada.[2] These postcensal population estimates are preliminary and are revised every five years once the results from the next census become available. Within the GGH, census divisions include two upper-tier cities, and fourteen upper-tier regions and counties (separated cities within a county are treated as part of the county). For the purpose of this blog post, census divisions are referred to as municipalities.

The annual estimates incorporate estimates for population undercount in the Census of Canada (for various reasons, a small proportion of the population do not respond to the census questionnaire). The population counts released to date from the 2016 Census of Canada are actual counts and not adjusted for population undercount.

What we do in this post is compare the growth numbers from the 2016 Census of Canada and the annual estimates for 2011-2016 in terms of percent distributions of where population growth occurred.[3] Implicit in this approach is the premise that the undercount rates (enumerated population as a percent of the estimated total population) are the same in all municipalities in 2016 as they were in 2011 which may get challenged when Statistics Canada releases its adjusted census counts. This comparison provides an approximation of the accuracy of the annual estimates when compared to the benchmark counts from the quinquennial Census of Canada.

In this post, the population growth comparisons are done at two geographic areas used by the Province in its land use planning: the Greater Toronto and Hamilton Area (GTHA) and the GGH. Our focus is on the differences between the estimated and actual growth that are 1.5 percentage points or more in size.

2.    Greater Toronto and Hamilton Area (GTHA)

2.1  Annual estimates significantly overstated the city of Toronto’s share of GTHA growth

  • Statistics Canada had estimated that the City of Toronto accounted for 33.1 percent of the GTHA’s growth between 2011 and 2016, whereas the Census counts showed it to be lower at 30.6 percent.

2.2   Annual population estimates also significantly overstated Peel region

  • Statistics Canada had estimated that Peel region accounted for 25.3 percent of GTHA growth – turns out the share was only 22.3 percent.

    The shortfall in Peel region’s actual share of the GTHA’s population growth may be in part due to the result of Mississauga exhausting its supply of land for ground-related housing and delays in bringing more serviced lands onto the market in Brampton.

2.3    Conversely, the annual estimates understated the share of growth for Halton and York regions

  • The annual estimates underestimated regions of Halton and York at 10.1 percent and 17.8 percent rather than 12.3 percent and 20.3 percent, respectively.

3.   Greater Golden Horseshoe (GGH) 

3.1   Annual estimates overestimated population growth in the GTHA and underestimated growth in the Outer Ring

  • The Outer Ring’s share of GGH population growth in 2011-2016 was considerably higher than the annual estimates indicated – 21.8 percent rather than 16.6 percent.

    It may well be that the dispersion of population growth in the GGH was greater than estimated due to homebuyers moving further afield in the search for more affordable ground-related homes, especially single-detached houses.

3.2  Annual population estimates were too low for Waterloo and Niagara regions in particular

  • Waterloo region’s share of GGH population growth is 5.8 percent of the GGH’s growth, 1.7 percentage points higher than what the annual estimates showed.
  • Niagara region also had a higher share of the GGH growth than the annual estimates indicated – 3.4 percent rather than 1.8 percent.
  • Simcoe and Wellington counties were also understated in the annual estimates (1.2 percentage points differences).


[1] See Frank Clayton, Demographic Update – Millennials in the Greater Golden Horseshoe in Mid 2015, Centre for Urban Research and Land Development, Ryerson University, July 6, 2016; Frank Clayton, Population Dynamics in the Greater Golden Horseshoe – Millennials vs. Baby Boomers, Centre for Urban Research and Land Development, Ryerson University, November 19, 2015; and Frank Clayton and Justin Shin, A Look at Reasons Why the Growth Plan Population Forecasts May Be Off-Target, Centre for Urban and Land Development, Ryerson University, November 3, 2015.

[2] Statistics Canada. Table 051-0001 - Estimates of population, by age group and sex for July, CANSIM (database). Accessed August, 2017,

[3] Percent distribution of the population growth is used so that preliminary annual growth estimates including undercount can be compared to the population growth per the 2011 and 2016 Censuses which excludes population undercount.