Statistical analysis has been used to interpret the data for the three main components of migration (immigration, net interprovincial migration, and net intraprovincial migration). In this section, results from the statistical analysis are used to simulate migration trends in the future, based on economic assumptions. These simulations of the components of migration are used later in this report to project population for the 16 Census Divisions of the Greater Golden Horseshoe.
Simulations were performed for each of the Census Divisions of the Greater Golden Horseshoe, as well as for three aggregated areas (the Inner Ring, the Outer Ring, and the GGH as a whole). In this section, only the results for the three aggregated areas are discussed. Summaries of the simulations for each of the 16 Census Divisions are provided in Appendix A.1.
Table 1 shows the estimated sensitivities for the components to changes in the two economic variables. The analysis indicates that:
- A one-percentage-point increase in the Toronto CMA employment-to-population ratio is estimated to result in increases in all three components of migration, for all three geographic areas, with one exception -- net intraprovincial migration for the Outer Ring has a weak negative relationship to the employment ratio. This negative relationship, while somewhat unexpected, reflects the fact that stronger employment in the Toronto CMA could, depending on its nature, result in either more or less movement to the Outer Ring from elsewhere in Ontario.
- A $10,000 increase in the real (that is, inflation-adjusted) GTA house price is estimated to result in reduced in-migration for all three components of migration for the Inner Ring and for the Greater Golden Horseshoe. This is expected.
- For the Outer Ring, rising house prices in the GTA are estimated to result in increased immigration and net intraprovincial migration. This result is expected, as it would reflect economic rationing in the GTA. However, the estimates indicate that a GTA house price increase results in reduced net interprovincial migration. This result is not expected. It may be that when house prices increase in the GTA, prices also tend to increase in the Outer Ring, which deters people from moving there from other provinces. Alternatively, rising house prices in the GTA may lead people in other provinces to assume that prices are rising elsewhere in Ontario, which would generally discourage non-Ontarians from moving to the Outer Ring.
Table 1: Estimated Sensitivities of Migration Components to Two Economic Factors
| Inner Ring (%) | Outer Ring (%) | GGH (%) |
Immigration (% Share of Immigration to Canada) | |||
mpact of 1% Increase in Toronto CMA Employment Ratio | 0.0313 | 0.1097 | 0.1410 |
mpact of $10,000 Increase in GTA Real House Price | -0.0800 | 0.0047 | -0.0754 |
Net Interprovincial Migration (as a % of Canada's Population Excluding Ontario) | |||
mpact of 1% Increase in Toronto CMA Employment Ratio | 0.0169 | 0.0054 | 0.0223 |
mpact of $10,000 Increase in GTA Real House Price | -0.0206 | -0.0055 | -0.0261 |
Net Intraprovincial Migration (as % of Ontario Population) | |||
mpact of 1% Increase in Toronto CMA Employment Ratio | 0.0084 | -0.0001 | 0.0083 |
mpact of $10,000 Increase in GTA Real House Price | -0.0512 | 0.0195 | -0.0317 |
| Source: Will Dunning Inc. |
The results from this analysis were used to simulate how migration rates might be affected by future trends in economic variables.
First, it is necessary to develop assumptions for the employment-to-population ratio and real house prices. The employment ratio peaked in 2001, and it appears to have fallen slightly during the past three years. During the last economic cycle (which began with a peak in 1989 and ended with another peak in 2001), the average employment rate in the Toronto CMA was 63.3 percent. In the analysis that follows, it is assumed that at present the long-term average employment rate is 63.3 percent, a reduction from the rate of 64.4 percent seen over the last 12 months (ending in October 2005). The 63.3 percent rate is assumed to hold over the 2001 to 2011 period.
In addition, demographic change may affect the employment rate over time. The baby boomers are moving into middle age and retirement, and this change may affect the percentage of the population that chooses to work (or is able to work). This relationship can be investigated by combining detailed data on employment-to-population ratios by age group with data on the expected age structure of the future population. Table 2 displays this analysis, using data on the GTA employment-to-population ratio from the 2001 census and combining it with projections of population by age group for the GTA from Hemson Consulting's Growth Outlook for the Greater Golden Horseshoe.
This analysis suggests that changes in the age structure of the population could cause the employment-to-population ratio in the GTA to fall by 5.75 percentage points between 2001 and 2031. Since the analysis holds employment rates constant for each age group, this change would be entirely due to demographic change; the effects of economic cycles are ignored in this simulation.
Table 2: Estimation of Potential Impacts of Population Change on the Employment-to-Population Ratio in the GTA
Age Group | Hemson Population |
| Implied |
| ||
2001 | 2031 | Emp-Pop Ratio, 2001 Census, GTA (%) | 2001 | 2031 | ||
15-19 | 347,200 | 438,700 | 41.53 | 144,190 | 182,189 | |
20-24 | 358,700 | 498,100 | 69.00 | 247,488 | 343,668 | |
25-29 | 392,700 | 585,800 | 80.43 | 315,835 | 471,139 | |
30-34 | 441,900 | 597,200 | 80.76 | 356,895 | 482,321 | |
35-39 | 492,400 | 614,500 | 81.99 | 403,719 | 503,829 | |
40-44 | 456,100 | 579,900 | 82.62 | 376,816 | 479,095 | |
45-49 | 391,700 | 543,400 | 82.86 | 324,563 | 450,261 | |
50-54 | 347,100 | 486,400 | 79.26 | 275,111 | 385,520 | |
55-59 | 253,100 | 436,500 | 67.96 | 171,998 | 296,630 | |
60-64 | 202,400 | 422,200 | 45.68 | 92,453 | 192,854 | |
65+ | 585,100 | 1,450,300 | 9.76 | 57,130 | 141,610 | |
Total | 4,268,400 | 6,653,000 |
| 2,766,197 | 3,929,116 | |
Total Employment-to-Population Ratio |
|
|
| 64.81% | 59.06% |
Sources: Statistics Canada, Hemson Consulting Ltd, calculations by Will Dunning Inc.
Admittedly, it is likely that the age-specific employment rates will vary over time. In the analysis that follows, an optimistic assumption is made: that demographic change will cause the employment-to-population ratio to fall by one percentage point in the 2011-to-2021 period and again in the 2021-to-2031 period -- this is only about one-half of the reduction indicated by this analysis of the 2001 census data.
The following assumptions were used in these simulations of migration rates:
- The Toronto CMA employment-to-population ratio averages 63.3 percent during 2001 to 2011. This is the average rate over the last business cycle.
- For both 2011 to 2021 and 2021 to 2031, the average employment rate is assumed to fall by one percentage point (to 62.3 percent between 2011 and 2021 and to 61.3 percent between 2021 and 2031). These reductions are the consequence of the aging population.
- For real house prices, it is assumed that between 2001 and 2011, the average real house price is 5 percent higher than the 2001 figure ($221,833 versus the 2001 figure of $211,270; all figures are expressed in 1992 dollars). This assumed average figure for the 2001-2011 decade is lower than the figure for 2004 ($243,611), and slightly lower than the average for the past four years ($227,938). This assumption implies that prices will fall in real terms during the next few years (in other words, that actual prices would increase less rapidly than overall inflation).
- For the subsequent two decades (2011 to 2021 and 2021 to 2031), the average real house price is assumed to be unchanged, at $221,833. While it is assumed that house prices will be unchanged in real terms, during 1980 to 2004 the average real rate of increase was 2.2 percent per year.
- If house prices were assumed to increase in real terms, the consequence would be even greater negative impacts on migration than is shown below.
The simulation results are shown in Table 3 and discussed in detail below. The table also shows the actual results for 2001 in comparison to the simulated results for the same year.
Table 3: Simulated Migration Rates, Based on Two Economic Factors
| Inner Ring (%) | Outer Ring (%) | GGH (%) |
Immigration (% Share of Immigration to Canada) | |||
001 Actual | 48.41 | 2.97 | 51.38 |
001 Simulated | 44.44 | 3.50 | 47.95 |
rror in 2001 | 3.96 | -0.53 | 3.43 |
2011 | 42.16 | 3.53 | 45.69 |
021 | 42.12 | 3.42 | 45.55 |
031 | 42.09 | 3.31 | 45.40 |
Net Interprovincial Migration (as a % of Canada's Population Excluding Ontario) | |||
2001 Actual | 0.0464 | 0.0095 | 0.0559 |
2001 Simulated | 0.0475 | 0.0126 | 0.0601 |
Error in 2001 | -0.0010 | -0.0032 | -0.0042 |
2011 | -0.0040 | -0.0037 | -0.0077 |
2021 | -0.0209 | -0.0091 | -0.0300 |
2031 | -0.0377 | -0.0145 | -0.0522 |
Net Intraprovincial Migration (as % of Ontario Population) | |||
2001 Actual | -0.1126 | 0.1080 | -0.0045 |
2001 Simulated | -0.1553 | 0.1520 | -0.0033 |
Error in 2001 | 0.0427 | -0.0440 | -0.0012 |
2011 | -0.2168 | 0.1655 | -0.0514 |
2021 | -0.2252 | 0.1656 | -0.0597 |
2031 | -0.2337 | 0.1657 | -0.0680 |
Source: Will Dunning Inc.
Implications for immigration:
- In 2001 the shares of immigration for the Inner Ring and the GGH were somewhat higher than the simulated rate. In 2001 and 2002, the Inner Ring and Greater Golden Horseshoe shares of immigration were elevated. For the years on either side, the actual and simulated results are closer together. It is unclear at this point whether the 2001 and 2002 results were temporary aberrations or represent permanent shifts in the relationships.
- Based on the relationships estimated using 18 years of data, it appears that the Inner Ring and the Greater Golden Horseshoe will experience gradual erosion in their shares of Canada's immigration. During the 2001-to-2031 period, the Inner Ring share would be similar to the rate seen in the mid-1990s.
- These projected rates can be used to project actual numbers of immigrants. The 2005 Hemson report Growth Outlook for the Greater Golden Horseshoe (page 20) assumed that immigration to Canada will average 218,000 people per year from 2001 to 2006, and just under 210,000 for the remainder of the projection period. Using this assumption, and applying the projected shares shown above, it is possible to calculate anticipated average annual immigration figures for the Inner Ring, Outer Ring, and Greater Golden Horseshoe, as shown in Table 4. For the Inner Ring, immigration is projected to total almost 2.7 million over the 30-year period, or slightly less than 90,000 people per year.
Implications for emigration:
- The rates of emigration (as percentages of the populations of each area) are assumed to remain at the 1997-2004 level. As the population increases, therefore, the number of emigrants will also rise. For the entire 30-year period, about 550,000 people are expected to leave the GGH for other countries, an average of about 18,000 people per year.
Implications for net immigration (immigration minus emigration):
- The total contribution to population growth in the Greater Golden Horseshoe over the 30-year period would be about 2.33 million. Annual net contributions from immigration would fall over time, as immigration falls (slightly) and emigration increases. From 2001 to 2011, the contribution would be about 80,700 people per year; between 2021 and 2031 it would be about 75,300 people per year.
Implications for net interprovincial migration:
- The simulated migration rates for 2001 were close to the actual rates for both the Inner Ring and the GGH, but less close for the Outer Ring.
- The actual data indicate that net interprovincial migration has deteriorated with the economic slowdown in the GTA. In the Inner Ring, net interprovincial migration has fallen from a peak of about 10,100 in 1999-2000 to less than zero (about -3,200) in 2003-2004. For the GGH, the numbers have fallen from a peak of +13,100 in 1999-2000 to -4,400 in 2003-2004. Based on the assumption of a normal economic performance (on average) in the years up to 2031, the projections indicate that central Ontario could lose population to other provinces.
- In Table 4, the projected rates for net interprovincial migration are applied to the projected population of the rest of Canada, to determine migration for the period up to 2031. The results indicate that from 2001 to 2031, the Greater Golden Horseshoe would lose about 187,000 people to other provinces. The average annual rate of out-migration would be low from 2001 to 2011 (averaging -1,500 people per year) and would increase to about -11,100 people per year from 2021 to 2031.
Implications for net intraprovincial migration:
- The Inner Ring simulation for 2001 overestimates net out-migration by about 40 percent, although the simulations for both 2000 and 2002 are more accurate. Similarly, for the Outer Ring, the simulation for 2001 is about 40 percent above the actual amount. For the GGH as a whole, where net out-migration was low, the simulation was quite close to the actual amount.
- Net intraprovincial migration for the GGH began to deteriorate in 1999, before the economic slowdown, apparently as the consequence of rising real house prices. Unfortunately, the last available data point is for July 2003, which precludes an evaluation of the consequences of the developing economic downturn and the continued real increases in house prices.
- The analysis of the available data suggests increasing migration out of the Inner Ring and into the Outer Ring. Over the 30-year period, the Inner Ring would experience net intraprovincial out-migration of 922,000, or about 31,000 people per year; in the Outer Ring, intraprovincial in-migration would total 930,000, or 31,000 people per year. For the entire GGH, net intraprovincial migration would be close to nil -- a total of about 8,000 over the 30 years.
Combining these components, it is possible to project total net migration:
- Total net migration is projected to add 2.15 million people to the population of the GGH during the coming 30 years, or an average of about 71,700 people per year. The contribution of net migration would fall over time, from just over 80,000 people per year from 2001 to 2011, to about 63,500 people per year from 2021 to 2031.
- About 55 percent of the total net migration would be to the Inner Ring (about 1.17 million people); 45 percent (977,000 people) would go to the Outer Ring.
- In the Inner Ring, net migration would fall sharply during the 30-year period, from an average of 48,800 people per year between 2001 and 2011 to an average of 30,000 people per year between 2021 and 2031. On the other hand, for the Outer Ring, net migration would increase slightly over time, from about 31,300 people per year between 2001 and 2011 to 33,600 people per year between 2021 and 2031.
Table 4: Projections of Components of Migration, Annual Averages by 10-Year Period and Total Migration, 2001 to 2031
| Inner Ring | Outer Ring | GGH |
Immigration | |||
001-2011 | 89,775 | 7,521 | 97,296 |
011-2021 | 88,445 | 7,185 | 95,629 |
021-2031 | 88,379 | 6,954 | 95,333 |
otal Over 30 Years | 2,665,989 | 216,596 | 2,882,585 |
Emigration | |||
001-2011 | -13,291 | -3,277 | -16,568 |
011-2021 | -14,728 | -3,792 | -18,520 |
021-2031 | -15,750 | -4,253 | -20,003 |
otal Over 30 Years | -437,688 | -113,216 | -550,905 |
Net Interprovincial Migration | |||
001-2011 | -785 | -718 | -1,503 |
011-2021 | -4,289 | -1,870 | -6,158 |
021-2031 | -8,001 | -3,082 | -11,082 |
otal Over 30 Years | -130,741 | -56,693 | -187,434 |
Net Intraprovincial Migration | |||
001-2011 | -26,866 | 27,813 | 947 |
011-2021 | -30,798 | 31,326 | 528 |
021-2031 | -34,607 | 33,939 | -668 |
otal Over 30 Years | -922,717 | 930,790 | 8,073 |
Total Net Migration | |||
001-2011 | 48,833 | 31,340 | 80,173 |
011-2021 | 38,629 | 32,849 | 71,479 |
021-2031 | 30,022 | 33,559 | 63,580 |
otal Over 30 Years | 1,174,842 | 977,476 | 2,152,318 |
Source: Will Dunning Inc.