In our work at REAPontario, we are collecting and analyzing Foundational Data to provide context for the Living Data collected from immigrant entrepreneurs, service providers, and other project stakeholders. To this end, we have purchased data from the 2006 Census and 2011 National Household Survey for the Province of Ontario.
“The data is broken down into 11 economic regions as defined by Statistics Canada.”
Ottawa, Ontario 
Kingston-Pembroke, Ontario 
Muskoka-Kawarthas, Ontario 
Toronto, Ontario 
Kitchener-Waterloo-Barrie, Ontario 
Hamilton-Niagara Peninsula, Ontario 
London, Ontario 
Windsor-Sarnia, Ontario 
Stratford-Bruce Peninsula, Ontario 
Northeast, Ontario 
Northwest, Ontario 
REAPontario further groups these 11 regions across the province into 5 larger economic development regions. These align more readily with the strategic and operational decisions being made in the market-place by businesses, industry associations, public institutions, governments and funders.
#1 Golden Horseshoe  
#2 Southeast Ontario   
#3 Southwest Ontario   3570] 
#4 Northwest Ontario 
#5 Northeast Ontario 
In addition, we have data on the following variables:
- Immigrant Status and Period of Immigration
- Class of Worker
- Industry (NAICS 2007 codes)
- Labour Force Status
Our next series of blog posts will present our preliminary, high-level analysis of this foundational data. We will be looking at the above variables as they manifest in each region, and also across Ontario as a whole.
A caveat about data comparison
Readers may be aware that Canada’s last full Census occurred in 2006 and that the Census was replaced with the National Household Survey (NHS) in 2011. Opinion is quite divided on whether or not the 2006 census data can be compared to 2011 NHS data.
Statistics Canada, for the most, implies that the two years can be compared, as stated here.
Some regional and municipal government sites (such as Peel Region) have released data briefs in which such comparisons have been made, along with a disclaimer about the difficulties of doing so.
Others state outright that the comparison can’t be made, including The Institut de la statistique du Québec. See for example the following two articles, which sum up the debate:
Why the controversy?
The non-response rate is much higher for the NHS than it ever was for the census. Given that non-respondents tend to be different than respondents on some key characteristics, the sample may be biased. For example, very rich and very poor people were the most likely to not respond. Surveys like the Labour Force Survey have smaller samples than the NHS, they also have higher response rates and are more representative. Some people say that this problem can be overcome (sort of) at high levels of geography (eg, province, large cities). But at lower levels of geography, or with smaller population groups, you are more likely to have unrepresentative samples and answers.
How is REAPontario handling this issue?
Our regions are fairly large in terms of population. Nonetheless, in view of the above concerns, we are erring on the side of caution and not making direct comparisons of the 2006 and 2011 data. Unless trends identified by using other data sources (such as Labour Force Survey) are similar, any comparisons should be treated with caution.
Sarah Wayland, Principal Investigator