Key takeaways:
- There has been no trend in the ratio of housing completions to population growth over the past forty years
- There have been changes in the types of homes that are being built
- Over the past twenty years, increasingly fewer single-detached houses have been built, while apartment units have been completed in their place
Canadian Housing Crisis A significant rise in Canadian real estate prices beginning in 2002 that has caused housing to become increasingly unaffordable.
It is often claimed that the Canadian housing supply is not keeping up with population growth. Factors such as post-2016 increases in immigration are frequently cited for contributing to rising real estate prices by growing the Canadian population faster than houses can be built. Under analysis of Canada Mortgage and Housing Corporation (CMHC) data, however, this effect is not so clear-cut.
Population Change The net change in population due to net natural increase and net migration.
Housing Completions The number of new housing units made ready for occupancy.
Though immigration has increased in recent years, Canada’s low birth rates and aging population (along with emigration) have served to moderate population growth. Previous decades have even sustained higher population increases than current rates.
| Decade | Average Annual Population Change |
|---|---|
| 1950-1960 | 405.5 K |
| 1960-1970 | 343.3 K |
| 1970-1980 | 317.5 K |
| 1980-1990 | 320.8 K |
| 1990-2000 | 307.4 K |
| 2000-2010 | 330.9 K |
| 2010-2020 | 388.7 K |
Note: Population numbers account for temporary residents, including students, workers, and refugee claimants.
Similarly, the raw number of housing completions in Canada has remained above historic averages in recent years, with increases mirroring population growth.
| Decade | Average Annual Housing Completions |
|---|---|
| 1950-1960 | 112.7 K |
| 1960-1970 | 150.2 K |
| 1970-1980 | 224.3 K |
| 1980-1990 | 180.3 K |
| 1990-2000 | 151.3 K |
| 2000-2010 | 192.0 K |
| 2010-2020 | 188.1 K |
Putting these two metrics together, a more informative ratio of housing completions to population change can be obtained.
As shown in the above graph, Canada’s ratio of new housing completions to population growth has remained close to historic averages in recent years. In fact, the ratio of new housing completions to population growth doesn’t appear to have followed any clear trend since the 1980s.
Such evidence might indicate that recent increases in real estate prices are not simply caused by a shift in the ratio of completions to population growth, as this ratio has remained relatively constant for the past forty years.
| Decade | Housing Completions per 1,000 Population Increase |
|---|---|
| 1950-1960 | 277.8 |
| 1960-1970 | 437.6 |
| 1970-1980 | 706.5 |
| 1980-1990 | 562.0 |
| 1990-2000 | 492.4 |
| 2000-2010 | 580.2 |
| 2010-2020 | 483.8 |
The second part of the graph tells an additional story. Despite the fact that housing completions seem to have kept up with growth, there are substantive changes in which types of homes are being completed.
Though completion rates for semi-detached houses and row houses, commonly referred to as “missing middle housing”, have remained essentially the same over the past twenty years, increasingly fewer single-detached houses have been built, while apartment units have been completed in their place. In 2014 the rate of apartment completions even outpaced the rate of single detached house completions — and the trend hasn’t reversed since.
| Decade | Single-Detached Completions per 1,000 Population Increase |
|---|---|
| 1950-1960 | 214.5 |
| 1960-1970 | 218.7 |
| 1970-1980 | 365.7 |
| 1980-1990 | 325.2 |
| 1990-2000 | 291.0 |
| 2000-2010 | 326.3 |
| 2010-2020 | 200.2 |
| Decade | Apartment Completions per 1,000 Population Increase |
|---|---|
| 1950-1960 | 75.3 |
| 1960-1970 | 184.1 |
| 1970-1980 | 287.0 |
| 1980-1990 | 179.1 |
| 1990-2000 | 117.5 |
| 2000-2010 | 163.1 |
| 2010-2020 | 228.3 |
How can the ratio of housing completions to population growth be interpreted?
Understanding housing completions in terms of houses built per population increase can be somewhat unintuitive. To produce a more accessible metric, the ratio can be flipped to show the population increase per housing completion. For example, instead of saying that 483.8 houses were built for every 1,000 people added to the Canadian population between 2010 to 2020, it can be said that one house was built for every 2.07 people that were added to the population.
When put into these terms, houses appear to be built at a higher rate than the average household size, reported to be 2.47 people in 2016 — as would be necessary to meet housing needs and replace demolished or repurposed units.
| Decade | Population Increase per Housing Completion |
|---|---|
| 1950-1960 | 3.60 |
| 1960-1970 | 2.29 |
| 1970-1980 | 1.42 |
| 1980-1990 | 1.80 |
| 1990-2000 | 2.03 |
| 2000-2010 | 1.72 |
| 2010-2020 | 2.07 |
Can trends in housing completion ratios be quantified?
Though it is possible to visually examine patterns in housing completion, the actual trend in the ratio of population growth to housing completions can also be quantified. However, this type of analysis has the potential to be heavily biased by which range of years is chosen to be examined as cutting out specific time periods may hide parts of the trend. Here are some possible ways to measure the year-to-year trend in the housing ratio:
| Time Period | Annual Change in Completion to Population Change Ratio | Significance Level |
|---|---|---|
| 1948-2021 | + 3.43 houses per 1,000 population increase each year | ≈0 |
| 1970-2021 | - 3.77 houses per 1,000 population increase each year | <0.01. |
| 1980-2021 | + 0.27 houses per 1,000 population increase each year | Not significant |
| 2000-2021 | - 0.78 houses per 1,000 population increase each year | Not significant |
| 2010-2021 | + 14.58 houses per 1,000 population increase each year | Not significant |
These values can be interpreted as the trend of the housing ratio for the specified time period. For example, Between 1948 and 2021, the number of annual housing completions per 1,000 increase in population grew, on average, by 3.43 houses per year. In other words, on average, Canadian housing completions grew at a faster rate than the population during this time period. This relationship is statistically significant.
Notably, analyzing different time periods will yield different trends. Between 1970 and 2021, on average the number of annual housing completions per 1,000 increase in population decreased. Yet from 1980 to 2021, 2000 to 2021, and 2010 to 2021 there were no statistically significant changes in the ratio. In fact, the only way to claim that the ratio of housing completions to population increases has been declining is to compare current data with data from the 1970s.
These trends remain the same even when 2020 and 2021 are excluded, indicating that pandemic-related deviations do not affect the overall trend.
What does this mean for the Canadian housing crisis?
The ratio of housing completions to population change has not followed any detectable trend since the 1980s, meaning that recent intensification of the housing crisis cannot be attributed housing completions failing to keep up with popoulation growth any more than they have over the past forty years.
Instead, other factors may contribute to a perceived scarcity in housing, including:
- Regional differences in housing completion ratios
- Increased ownership by investors compared to first-time home buyers
- Trends in the types of homes that are being built
The shift in which types of homes are being built possibly contributes to a public perception that the housing supply is failing to meet population needs since if people are primarily seeking single-detached homes, they would indeed be correct in observing that this type of housing is increasingly scarce.
These factors can be further investigated through CMHC data to obtain a better understanding of the Canadian housing crisis.
It is important to be cautious about reducing the Canadian housing crisis to a crisis of supply. If the core problem is the allocation of housing, then adding more supply will not solve the housing crisis and may even exacerbate existing allocation issues.
All data and code for this analysis can be found here.
Edited to clarify that population estimates include temporary residents.