Home | Login | Search
 
166 Bedford Highway, Halifax, N.S., B3M 2J6  donald.shiner@msvu.ca

Community Profiles

Assumptions and Limitations of the Model

Population Projections: This model begins with the actual 2001 Census count in each FSA. The projections are based on the population projection factors developed by Statistics Canada to produce Population Projections for Canada, Provinces, and Territories 2000-2026 (Cat. No. 91-520-XPB). The key difference between the StatsCan projections and the ones used in the model is the starting point: StatsCan projections use the 2000 preliminary population estimates which are based on the 1996 census, while the model uses the 2001 FSA Census count.

The result of this difference is that our model projections are consistently lower than those using the StatsCan data. This is because the approach based on the 1996 Census typically overestimates the 2001 Census actual results by 4%.

In selecting which StatsCan projection scenario to use as the basis of our model projections, we have chosen the Low Growth assumption projections. This choice was made as the result of our analysis of variance between the StatsCan 2001 population projection for each of the four Atlantic Provinces and the actual 2001 Census counts, as follows:

  Low Growth Scenario 2001 Population Projection 2001 Census Report % Difference
NL 536,600 512,930 4.6%
NB 757,100 729,500 3.8
NS 942,500 908,010 3.8
PEI 139,400 135,295 3.0

By starting with the 2001 Census actual count and using the low growth scenario, we anticipate that our projections will present a reasonable picture of what may happen.

Our population projections begin with the actual 2001 Census count in each FSA  then use the Provincial level projection factors of the low growth scenario to produce the projections for each FSA.

Our model lets you select an FSA to examine the 50+ population (male and female) by age cohort and see the likely impacts of health and wealth on these relatively small populations. Because of the nature of population sizes in each FSA and age cohort, we are necessarily dealing with small numbers. When a particular result is less than 5 individuals, our model will replace the number with an asterisk (*).

Health States: Our four health states are factors produced by Statistics Canada based on the National Population Health Survey and other sources, and described in an article in Canadian Social Trends in 2000. These factors are calculated at the national level as the data are not available at the provincial level.

Wealth States: Our four wealth states are factors based on the 1999 Survey of Financial Security (Catalogue no. 13F0026MIE-no.002). These factors have been developed at the level of the four Atlantic Canadian provinces.

Copyright 2009. ASHRA. All rights reserved. Contact Us | Legal Notice | Site Map