CMHC Says Canada Needs 430,000 to 480,000 Units Built Annually: Construction Pace Headed the Wrong Direction

The federal housing agency projects that closing the country's shelter deficit requires nearly doubling the construction output Canada has sustained in any modern year. Between 2000 and 2023, annual completions averaged 205,000 units. The target now sits at 417,000 to 430,000 to 480,000 (by 2035, from June 2025 CMHC report) per year through 2030.

The gap narrowed slightly in the Canada Mortgage and Housing Corporation's 2026 modeling compared to earlier estimates, but the improvement came from revised economic projections and modestly lower demand forecasts rather than a surge in physical building. The agency still maintains the country needs 3.5 million additional units beyond what is already projected, a figure that dates to 2023 and has not moved much since. Slower population growth and higher baseline construction assumptions softened the delta, but the need itself remains unchanged.

The construction sector is slowing

Housing starts peaked in late 2021 and have been declining since. High interest rates, which the Bank of Canada began cutting only in mid-2024, froze financing for large-scale developments through most of 2024 and 2025. The lag effect means projects that would have broken ground in those years are still stuck in planning or awaiting capital. Labor shortages compound the issue. The construction sector employs a workforce that skews older, and the pipeline of new carpenters, electricians, and plumbers is not keeping pace with retirements.

Material costs have stabilized from their 2021, 2022 spike but remain elevated by pre-pandemic standards. Developers who might have penciled a project at 3.5% financing in 2019 are now underwriting at 6% or higher, even after the recent rate cuts. That spread changes the math on marginal projects, particularly in secondary markets where rents have not kept pace with Toronto or Vancouver.

The federal government introduced the Housing Accelerator Fund to push municipalities toward faster zoning approvals and density increases. Several cities have rezoned to allow fourplexes on single-family lots. Ottawa removed the GST on new purpose-built rentals, which shifted developer interest away from condominiums and toward rental construction. Both moves are directionally correct, but neither solves the bottleneck at the physical build stage. A streamlined approval process still requires someone to frame the walls, and that labor force does not expand quickly. The 2030 deadline remains a political commitment, not a forecast built on current construction capacity.

The baseline choice matters more than the headline number

CMHC frames its 3.5 million figure as what is needed to restore affordability to mid-2000s levels, when housing costs accounted for roughly 30% of gross household income. In Vancouver and Toronto, the mid-2000s were simply less unaffordable than today. The agency's definition also assumes stable immigration targets and no major demand-side interventions, both of which are politically uncertain heading into 2027.

The geographic concentration of the gap means the national figure obscures the real problem. Ontario and British Columbia account for the majority of the shortfall. Alberta and the Atlantic provinces have tighter markets than they did five years ago, but they are not facing the same structural crisis. A national target of 430,000 to 480,000 units per year implies that half will be built in just two provinces, and both are constrained by land-use rules, municipal permitting processes, and utility networks that federal policy alone cannot override.

Most industry analysts treat the 2030 target as aspirational rather than achievable. The math requires not just a policy shift but an expansion of the construction workforce and materials supply that would take years to achieve even if financing were unlimited. The housing minister can announce the target. The question is whether anyone can deliver it.