Plaudits to CMHC's Economics Team for Its Contribution to Market Insights
By: Frank Clayton, Ph.D., Senior Research Fellow
April 20, 2026
(PDF file) Print-friendly version available
1. Executive summary
CMHC's Spring 2026 Housing Supply Report gets the fundamentals right on how housing markets work and the sources of new supply. It segments the supply by new structural groupings, broadens the supply lens beyond traditional housing starts, and highlights the role of filtering in creating more affordable housing over time. The report covers six metropolitan markets.
Our appraisal of the report鈥檚 contents focuses on metropolitan Toronto. Elements of the Toronto data raise questions about measurement and interpretation. These include the sizable number and composition of "conversions", the lack of accessory suite creation in new houses, and the focus on counting buildings rather than units for small versus very large projects. The report would be even more useful with fuller data disclosure and clearer treatment of the various supply components by unit types and geographies.
Highlights
- CMHC's economics team is to be commended for pursuing a market-based treatment of housing supply;
- Segmenting the market by updated housing types (ground-oriented and missing middle) provides a more realistic view of supply and demand pressures than the traditional split in housing starts by type;
- Filtering is correctly identified as a key mechanism through which new, higher-end supply improves affordability in older stock over time;
- Metropolitan Toronto's reported dominance of conversions in its missing middle supply and lack of any new accessory suites is not credible and reflects data gaps;
- Counting structures rather than units for 3鈥�5-unit versus 100+ unit apartments understates the contribution of large projects to total supply; and
- Future reports should consider consolidating the national and regional reports into a single report, publishing base data for percentages, disaggregating conversions, tracking losses from the existing housing stock, and distinguishing between supply by unit type added to built-up and greenfield areas within metropolitan markets.
2. CMHC's contribution to housing market understanding
Canada Mortgage and Housing Corporation (CMHC) should be proud of its economics team. Their work helps Canadians understand how housing markets function and what drives housing prices and rents, which is essential for designing effective policies to expand supply and improve affordability.
This market understanding is showcased in CMHC's Spring 2026 Housing Supply Report, which highlights several important features of Canada's major housing markets.[1]
3. Housing supply by type
3.1 From traditional to policy-relevant breakdowns
Historically, CMHC has published housing starts using a structural breakdown of single-detached, semi-detached, row, and apartments/other, with a condensed split between single-detached and all other units (called "multiples"). In today's market and policy environment, this condensed breakdown is much less useful; regrouping structural types into ground-oriented, missing-middle, and higher-rise apartment categories is more relevant for understanding current supply and affordability challenges.
3.2 Ground-oriented (ground-related) homes
Ground-oriented (what CUR calls "ground-related") homes include single-detached, semi-detached, and row homes. These are typically family-sized ownership units, a market segment that continues to face structural supply shortages and persistent affordability constraints, especially in higher-cost metropolitan areas like Vancouver and Toronto.
Among the seven census metropolitan areas (CMAs) examined, Toronto was the only one where 2025 ground-oriented starts fell well below their 10-year average, a pattern also observed for apartments. This comparison underscores the severity of Toronto's supply challenges in both family-oriented and multi-unit forms.
3.3 Missing middle housing
The report defines "missing middle" housing as accessory suites, multiplexes, row homes, stacked townhouses, and low-rise apartments (4 storeys or less). It focuses on new projects containing these unit types, while explicitly excluding the addition of accessory suites within existing houses from its standard housing starts series.
Importantly, the report treats conversions as part of the missing middle supply. Conversions are curtly defined as additions to supply by repurposing existing buildings. All conversions are included, regardless of the number of storeys in a structure, because CMHC only has data on the total units created 鈥� not the type of structure. Thus, for example, a 25-storey office building converted to residential use falls under the missing middle label in the CMHC report.
Metropolitan Calgary and Edmonton lead with about 55% of combined total starts and conversions in missing middle forms, followed by Toronto at 37%. Most missing middle units in Calgary and Edmonton are townhouses, while more than half of Toronto's reported missing middle units are conversions.
4. Filtering and affordability
Many land-use and housing planners either misunderstand or overlook the process economists call filtering. CMHC's economics team correctly highlights filtering as a central mechanism through which new supply improves affordability in older stock over time.
When new homes are completed, higher-income households tend to move into them, freeing up existing homes for others down the income ladder, according to CMHC. This chain reaction 鈥攆iltering 鈥� means that building more market housing for higher- and middle-income households expands the stock of more affordable housing over the long term across the entire system.
5. Toronto: data issues and oddities
The Toronto CMA analysis reveals some oddities in how the report measures and classifies new supply.
5.1 Conversions dominating Toronto鈥檚 missing middle supply?
CMHC estimates that in the Toronto CMA in 2025, conversions accounted for 21.2% of combined missing middle starts and conversions, while missing middle starts accounted for 15.5% of the combined supply. The role of conversions was much higher in Toronto than in any of the other six metropolitan markets examined.
As noted, CMHC includes all units created in existing buildings as missing middle homes, since it has information only on the total number of units created. Unknown are the units created in existing office buildings with more than four storeys, which are not missing middle units. Moreover, it is not clear whether conversions in the supply report measure gross additions to new supply or are net of losses in the existing housing stock through demolitions and deconversions.
In a previous report, CMHC provides a more detailed definition of conversions.[2]
- Conversions 鈥� non-residential to residential: A non-residential space is considered as being converted to residential units if the newly created units are being used for residential purposes;
- Conversions 鈥� residential to residential: Refers to the addition or removal of units from a residential structure on an existing foundation. There are 2 types of residential-to-residential conversions: a positive conversion and a negative conversion;
- Positive conversion: Refers to the conversion of a single unit structure to a multi-unit structure or to the addition of multi residential units to an existing multi-unit structure. In either case, the conversion results in a net increase in the number of dwelling units;
- Negative conversion: Refers to the removal of units from an existing residential structure. This could include the removal or deconversion of residential units within a multi-unit structure or a multi-unit structure being converted to a single-unit residential structure. In either case, the conversion results in a net loss in the number of dwelling units.
The report also includes another definition of conversions that appears to overlap with the residential-to-residential conversions definition.
- Infill conversion: Refers to a conversion that occurs within an established neighbourhood where infrastructure to support residential development is already in place, rather than in a greenfield or new development area.
In the report, CMHC estimates the supply of new units from starts and conversions for 2018-2025 (first half) for six central cities. The 2024 data for the city of Toronto show:
- Toronto had 2,942 missing middle housing starts in 2024, far fewer than Calgary (14,526), Edmonton (6,190), and Ottawa (4,036), with only Vancouver lower at 1,002; and
- Within Toronto's 2,942 missing middle combined starts and conversions, the breakdown was 42% conversions, 27% stacked townhouses, 17% row, 11% low-rise apartments, 3% multiplex, and zero accessory suites.
The estimate of zero accessory suites (secondary suites, basement apartments, "granny" suites) in starts of Toronto houses in 2024 is an anomaly, especially when Vancouver, Edmonton, and Calgary have a significant proportion of the missing middle housing in the form of accessory suites in new houses started.
5.2 Structures versus units for small vs. large projects
The Toronto CMA report presents an annual series from 1988 to 2025 comparing the share of apartment structures started with 3鈥�5 units to those with 100+ units, expressed as percentages of total apartment structures. It concludes that market conditions are pushing developers toward smaller projects, noting that 2025 was the first year in which 3鈥�5-unit buildings outnumbered 100+-unit buildings.
Counting structures rather than units is a weak way to gauge supply contributions. It takes 50 three-unit buildings to match the total units in a single 150-unit apartment building, so unit counts are the more informative metric for policy and market analysis.
6. Strengthening future CMHC housing supply reports
CMHC's work on housing supply represents a major advance over the traditional focus on starts alone, and the economics team deserves accolades for pushing this agenda. Several enhancements would further increase the value of future supply reports.
6.1 Consolidate housing supply report files into a single file
The Spring 2026 release consists of a national report plus separate reports for each of the seven CMAs. Providing a consolidated file that brings all reports together would improve accessibility for analysts and policymakers working across regions.
6.2 Provide base data for percentage graphics
The report's graphs are illuminating, and CMHC's practice of including the plotted data is commendable. Adding the base counts behind percentage figures in the downloadable data would make the graphics more transparent and usable.
6.3 Expand and disaggregate supply components to cover all units created in existing structures
CMHC has tracked new housing units created in new structures by unit type for decades. It is time to compile similar data on housing units created within the existing stock of residential and non-residential structures (what CMHC calls "conversions") by unit type. The end product would provide a comprehensive database of gross additions to the housing stock by unit type.
CMHC has now broadened its measures of new supply beyond conventional housing starts by incorporating conversions. Next steps should include splitting conversions by unit type and building height (number of storeys), and the quality of the conversion components of supply.
6.4 Track losses from the existing housing stock
New gross supply is only part of the story: losses from the existing housing stock through demolitions, fires, and deconversions (e.g., merging multiple units into a single home) also matter. Including estimates of these losses would allow CMHC to report net supply, a more powerful series for both market and policy analysis.
6.5 Distinguish gross and net supply of housing by location (built-up urban areas versus greenfield areas)[3]
Finally, distinguishing between new housing supply added in built-up areas and that on previously undeveloped (greenfield) land would greatly enhance the usefulness of CMHC's supply estimates for urban policy and growth management purposes. For planners and decision-makers, understanding where new units by type are located is as important as knowing how many there are.
CMHC already releases its housing starts data in larger urban areas by census tract. What is proposed here is that the census tracts be consolidated into totals for built-up urban areas and greenfield areas with regular releases.
End Notes
[1] CMHC. 鈥淪pring 2026 Housing Supply Report.鈥� I also consulted two other CMHC reports while preparing this blog. CMHC. 鈥淗ow Common is 鈥楳issing Middle鈥� housing development in Canada?鈥� October 23, 2025, and CMHC. 鈥淯nderstanding Filtering: A Long-Term Strategy to New Supply and Housing Affordability.鈥� June 2024.
[2] CMHC. 鈥淗ow Common is 鈥楳issing Middle鈥� housing development in Canada?鈥� October 23, 2025.
[3] Built-up urban areas versus greenfield areas is planning terminology in Ontario, elsewhere different terminology may be used to describe these differing development patterns. City of Edmonton planners distinguish between developing and developed areas.
References
Canada Mortgage and Housing Corporation (2026). 鈥淪pring 2026 Housing Supply Report.鈥� March 11, 2026. [Online]. Available: .
Canada Mortgage and Housing Corporation (2025). 鈥淗ow Common is 鈥楳issing Middle鈥� housing development in Canada?鈥� October 23, 2025. [Online]. Available: .
Canada Mortgage and Housing Corporation (2024). 鈥淯nderstanding Filtering: A Long-Term Strategy to New Supply and Housing Affordability.鈥� June 27, 2024. [Online]. Available: .