Methodology

Urban areas were divided into four types, which were defined according to the census metropolitan area (CMA) and census agglomeration (CA) concepts. CMAs and CAs consist of one or more municipalities concentrated around an urban core. In order to be deemed part of a CMA or CA, a municipality must be strongly integrated into the urban core, in other words, there must be a great deal of commuting between the two areas.²  The following types of urban areas were used for this project:

1) The three CMAs whose populations surpassed 1,100,000 in 2001, i.e., Montréal, Toronto and Vancouver;³

2) CMAs that had between 500,000 and 1,100,000 inhabitants in 2001, i.e., Québec, Ottawa-Gatineau, Hamilton, Winnipeg, Edmonton and Calgary. We often refer to these types of regions as “large urban areas”;

3) CMAs and CAs with 100,000 to 499,999 inhabitants in 2001, i.e., the 18 smaller CMAs as well as larger CAs such as Barrie, Cape Breton, Chatham-Kent, Guelph, Moncton, Peterborough and Kelowna. These are the “medium-sized urban areas”; and

4) CAs with 10,000 to 99,999 inhabitants in 2001. There are 106 census agglomerations in this type of metropolitan region, which we will refer to as “small urban areas”.

Rural areas were grouped into four categories based on the concept of census metropolitan area and census agglomeration influenced zone (MIZ). The purpose of the MIZ concept is to classify non-metropolitan areas, that is, municipalities which are not part of a CA or CMA, according to whether they are subjected to a 1) strong; 2) moderate; 3) weak; or 4) no metropolitan influence (McNiven, Puderer and Janes, 2000). The classification is based on commuting volumes. When more than 30% of a non-metropolitan municipality’s population works in the urban core of a CMA or of a CA, it is categorized as a strong MIZ, whereas when between 5% and 30% do so, it is categorized as a moderate MIZ. Municipalities where this proportion is lower than 5% are categorized as a weak MIZ, and when it has fewer than 40 commuters to CMAs or CAs, it has no MIZ. It should be noted that all of the territories (with the exception of Yellowknife and Whitehorse) are considered as having no MIZ for our purposes.

Municipalities differ in size and surface area as the laws that define them vary from province to province or from territory to territory. These differences mean that some more heavily populated and urbanized provinces (like Ontario) have fewer municipalities than others (like Saskatchewan). As well, it would appear that close to 30% of municipalities (1,635 out of 5,600) fall into the no MIZ category and more than 80% are classified as rural (4,605 out of 5,600). Finally, it should be noted that some categories are not represented in certain provinces. For instance, because Prince Edward Island does not have any census agglomerations with more than 100,000 inhabitants, three of the urban categories are not represented there. Also, all municipalities in Nunavut are classified as no MIZ. These differences reflect significant provincial features related to their degree of urbanity/rurality. If we analyze the proportion of the provincial population that is part of each of the eight categories included here (Table 1.1), we see that Newfoundland and Labrador is the most rural province (53% of its population lives in a rural municipality), whereas Ontario is the most urban (87% of its population lives in a metropolitan region).

Historical analysis: choice of a constant rather than variable geography

Many of the analyses presented in this document look at the evolution over time of certain demographic characteristics according to the urban/rural typology that we have just presented. These historical analyses are based on a constant geographic breakdown according to the geographic structure of the 2001 Census, which was applied to the 1971, 1981 and 1991 censuses. To create this constant geography for the analysis in this study, we adjusted Enumeration Areas for each Census prior to 2001 to correspond to municipal boundaries in 2001. In total, eight geographic regions along the urban-to-rural gradient were based on these reconstituted municipalities.

It should be noted that this approach does not allow us to measure (as would a geography with variable borders that changed from one Census to the next) the phenomenon of urbanization in its entirety because it does not take into consideration the reclassification of certain rural areas into urban areas over time. A significant number of municipalities that had been classified as rural in 1971 have since been classified as urban, thereby contributing to the country's urbanization. Such reclassification can occur:

1) When a municipality near an existing metropolitan area merges with it; or,

2) When one or more non-metropolitan municipalities grow to the point that they meet the criteria for a new metropolitan area.

This is why the population in urban areas included in the censuses prior to 2001 seems higher when using a constant geography than with a variable geography and that the difference between these two methods becomes more pronounced the more we recalculate over time.

On the other hand, the use of a variable geography would significantly complicate the historical comparisons. With a variable geography, one cannot determine whether the demographic changes observed are attributable to real demographic changes or to the impact of the many geographic reclassifications that occur from one Census to the next. In this study, however, we want to isolate the regions' demographic component from their growth.

This reason is essentially why, in this document, the historical comparisons will use a geography with constant borders rather than one that is variable over time.

Notes

1. In this analysis, 1971, 1981, 1991 and 2001 census sample data (questionnaire 2B) were used. These data do not include institutions starting from 1981. As this population accounts for a very small proportion of the total population, this exclusion does not have any effect on the conclusions.
2. In general, more than 50% of the employed labour force has to work in the urban core or vice versa. Some municipalities that do not meet this criterion may nonetheless be included because of spatial delimitations or historical comparability (2001 Census Dictionary).
3. The population thresholds that define the categories are based on 2001 data. For example, Vancouver is classified as a CMA with more than 1.1 million inhabitants for each census year analyzed in this study even though its population was actually less than 1.1 million inhabitants in 1971.