Population Analysis, Global to Local
Part 2: A Comparison of Population Projections for BC
by William Warren Munroe, March 21, 2013

Population forecasting benefits from taking into consideration, migration. For places within Canada, estimates of interprovincial migration are made.

Let’s take a look at forecasts of Net Interprovincial Migration for British Columbia...

Figure 1: Net Interprovincial Migration observed (1981/1982 to 2008/2009) and projected (2009/2010 to 2035/2036) according to seven scenarios, British Columbia (Chart 1.13)

Source(s): Statistics Canada, Demography Division. Date Modified: 2010-06-18 http://www.statcan.gc.ca/pub/91-520-x/2010001/ct043-eng.htm


Notice that the projections do not show the ups and downs seen in the historical numbers. The historical numbers rise and fall above and below zero; however, the forecasts tend to be relatively straight and are above zero. These forecasts are meant to provide an average interprovincial migration. While averages of the past may provide a convenient way to forecast the long term future, we may benefit from considering the possibility that interprovincial migration may become negative in the short to medium term (5 to 10 years) future.

As the Population Analyst for BC Stats, I suggested creating a fourth scenario (after low, medium, and high) that reflects the fluctuating migration seen in the historical migration numbers for BC.

Let’s take a look at BC’s Net interprovincial numbers referred to in my April 2012 article "BC's Annual Net Interprovincial Migration is negative for the first time since 2002"

Figure 2: Net Interprovincial Migration, BC, 1971 to 2011

From "BC's Annual Net Interprovincial Migration is negative for the first time since 2002" by W. W. Munroe, April 2012 at http://www.wminfomatics.com/WManalytics/Articles/120406/PopBC2011a.html


Notice the difference? Now let's take a look at BC’s total population forecast.

Figure 3: Population observed (1981 to 2009) and projected (2010 to 2036) according to six scenarios, British Columbia (Chart 3.36)

Source(s): Statistics Canada, Demography Division. Date modified:2012-12-20 http://www.statcan.gc.ca/pub/91-520-x/2010001/ct023-eng.htm


With mortality rising and fertility below replacement in so many areas within BC, why is the low-growth scenario so high? For example, deaths have exceeded births in the Nanaimo Census Division (referred to as a Regional District in BC) since 1999.

Now, let's take a look at population projections created using the well known cohort change ratios method, referring to the 1996, 2001, 2006, and 2011 census counts ...

Figure 4: British Columbia, Population Estimates 1996 to 2011, and Projections, to 2016 to 2051

If population change seen between 1996 to 2011 inclusive were to continue into the future, BC's population total would be approximately 5 to 5.5 million people by 2051.

Low fertility (below replacement) and increasing mortality means that the relatively high population growth seen in Figure 3 can only come from rising, positive net migration - increasingly higher than historical numbers.

As we have seen in Figure 1, this assumption is not reflected in the Figure 2 estimates.

With fertility below replacement, migration becomes more important to consider. Projections that include fluctuating migration can be beneficial.

More open public debate about Population Statistics, and "best practices" of Official Statistical Agencies would also be beneficial.


NOTE TO THE READER:

The method used to create the population projection (2016 to 2041) portion of Figure 4 was developed to address questions raised by people across BC about how population projections, used to justify recommendations to close public schools permanently, were created.

See the Service webpage for the write up "Population Projection Package" describing how the Figure 4 projection was created.


In preparation for presentations to Population Specialists (Applied Demographers), a more detailed methods paper is currently being written. Thank yous go out to David A. Swanson, PhD, (University of California, Riverside) for having written "Using cohort change ratios to estimate life expectancy in populations with negligible migration: A new approach" (Swanson and Tedrow, 2012, Canadian Population Society publication) and for sending the methods paper "Forecasting the Population of Census Tracts by Age and Sex: An Example of the Hamilton-Perry Method in Action " (David A. Swanson, Alan Schlottmann, Bob Schmidt 2009).

These and other readings show that variants of the Population Signature method have been used for many years and continues to be refined. Referring to a 2002 report from the UN, Dr. Swansen et al., (2012) state "we prefer to use the more general term "cohort change ratios" (CCRs)." Agreed! The Population Signature charts are simply a graphic representation of the cohort change - the difference between one census aged to compare with the next. The differences by age group are used to calculate ratios that are applied to the latest census age sex counts, to create projections. (November 10, 2012).


Statistics Canada Statement "It is also worth noting that the accuracy of any projection is conditional on the reliability of the base population estimates, the component data, and the degree to which the underlying assumptions correspond to future trends. Projections are not predictions; they are instead an effort to create plausible scenarios based on assumptions regarding the components of population growth, which are themselves subject to uncertainty. Accordingly, it cannot be claimed that the values observed in the coming years will always remain within the range implied by the low-growth and high-growth scenarios."

Source: the 1996 to 2011 census counts are from numeric tables from Statistics Canada found online via http://www12.statcan.ca/census-recensement/index-eng.cfm.


Warren Munroe’s Population Analysis Qualification: The 2016 to 2041 projections are created by Warren Munroe with reference to the 1996 to 2011 census counts from Statistics Canada. Of course the future is open to variation; therefore, projections are still a guess, but at least, with reference to the 1996 to 2011 census counts, this projection can reasonably be considered an educated guess.

While this median scenario is based on the 1996 to 2011 census counts from Statistics Canada, the United Nations and Statistics Canada did not examine, test, review, comment nor contribute to Warren Munroe's projection numbers, 2016 to 2061.



Back to Top



Website content, code, and design by W.W. Munroe. Copyright 2013.