Fifteen-plus years of BC mortgage experience, distilled into practical writing on buying, renewing, refinancing, and building wealth through real estate.
CMHC's 3.5 Million Homes Are Extra, and That Changes the Shortage Math
Canada started 259,028 new homes in 2025, and none of them counts toward the 3.5 million that CMHC says we're short. That sounds wrong until you read the full wording of CMHC's estimate, published in September 2023. Canada needs 3.5 million additional units by 2030. That means 3.5 million homes beyond what builders were already expected to put up anyway.
The difference between "units" and "additional units" decides how you read every soft quarter, every population dip and every builder press release between now and 2030. If you're buying in the North Okanagan, renewing a mortgage, or running numbers on a rental, it changes how much pressure you should expect to stay in the market. The shortage is a gap that sits on top of normal building. Normal building, even at a healthy pace, doesn't close it.
Read the word before the number
The common reading goes like this: we need 3.5 million homes, we build a few hundred thousand a year, so we're steadily chipping away at it. That reading holds for most targets people meet in daily life. If you need 100 bales of hay and you bring in 20 a week, you'll have them in five weeks.
CMHC's figure works differently. Its forecast already assumes a normal amount of construction happens. The 3.5 million is what's still missing after that. So a year of ordinary building leaves the gap where it was. Only building above the expected pace makes it smaller.
The math gets uncomfortable fast. Here's some rough illustrative arithmetic. Spread 3.5 million extra homes over the roughly seven years from the 2023 estimate to 2030, and you get about 500,000 extra homes a year. Last year's total starts were 259,028. The extra alone would be close to double everything Canada began building in 2025.
That's why newer projections have kept the number large. CMHC's 2026 outlook puts the need at up to 4.69 million homes by 2036, and it warns that construction could slow.
Why a soft quarter doesn't end the shortage
Here's where the usual instinct deserves some credit. When population growth slows or turns negative for a quarter, demand at the edge really does ease. Fewer new arrivals are competing for rentals. Some landlords offer a free month. Some sellers sit longer. Those are real signals, and ignoring them would be its own mistake.
The trouble starts when that easing gets read as the shortage resolving. A slower quarter reduces how many new households show up. It doesn't build any of the homes that were already missing. The backlog is a stock of absent housing. A pause in new demand stops it from growing for a while, but the homes still have to be built.
So prices and rents can stay firm while the headlines look soft. That outcome confuses people, but it fits the math above. When it happens, the market is mostly showing you the backlog.
Ask "total or extra?" before you decide
The idea worth keeping applies well beyond this one statistic. When you see a big housing number, ask whether it's a total or an amount on top of what's already happening. A total shrinks with every home built. An "extra" only shrinks when building beats the expected pace.
Put that question to the decisions in front of you:
Buying: a quiet season can be a good time to negotiate. Just don't assume the quiet will last long enough to reset prices.
Renewing: plan for a housing market that stays tight underneath, even if a soft patch arrives.
Investing: build your rental analysis around the size of the backlog, and treat any single quarter's population figures as a smaller factor.
A slower quarter tells you how many new households showed up. The shortage figure tells you how many homes were already missing, and right now that's the number that sets the floor under prices and rents.
Canada started 259,028 new homes in 2025, and none of them counts toward the 3.5 million that CMHC says we're short. That sounds wrong until you read the full wording of CMHC's estimate, published in September 2023. Canada needs 3.5 million additional units by 2030. That means 3.5 million homes beyond what builders were already expected to put up anyway.
The difference between "units" and "additional units" decides how you read every soft quarter, every population dip and every builder press release between now and 2030. If you're buying in the North Okanagan, renewing a mortgage, or running numbers on a rental, it changes how much pressure you should expect to stay in the market. The shortage is a gap that sits on top of normal building. Normal building, even at a healthy pace, doesn't close it.
Read the word before the number
The common reading goes like this: we need 3.5 million homes, we build a few hundred thousand a year, so we're steadily chipping away at it. That reading holds for most targets people meet in daily life. If you need 100 bales of hay and you bring in 20 a week, you'll have them in five weeks.
CMHC's figure works differently. Its forecast already assumes a normal amount of construction happens. The 3.5 million is what's still missing after that. So a year of ordinary building leaves the gap where it was. Only building above the expected pace makes it smaller.
The math gets uncomfortable fast. Here's some rough illustrative arithmetic. Spread 3.5 million extra homes over the roughly seven years from the 2023 estimate to 2030, and you get about 500,000 extra homes a year. Last year's total starts were 259,028. The extra alone would be close to double everything Canada began building in 2025.
That's why newer projections have kept the number large. CMHC's 2026 outlook puts the need at up to 4.69 million homes by 2036, and it warns that construction could slow.
Why a soft quarter doesn't end the shortage
Here's where the usual instinct deserves some credit. When population growth slows or turns negative for a quarter, demand at the edge really does ease. Fewer new arrivals are competing for rentals. Some landlords offer a free month. Some sellers sit longer. Those are real signals, and ignoring them would be its own mistake.
The trouble starts when that easing gets read as the shortage resolving. A slower quarter reduces how many new households show up. It doesn't build any of the homes that were already missing. The backlog is a stock of absent housing. A pause in new demand stops it from growing for a while, but the homes still have to be built.
So prices and rents can stay firm while the headlines look soft. That outcome confuses people, but it fits the math above. When it happens, the market is mostly showing you the backlog.
Ask "total or extra?" before you decide
The idea worth keeping applies well beyond this one statistic. When you see a big housing number, ask whether it's a total or an amount on top of what's already happening. A total shrinks with every home built. An "extra" only shrinks when building beats the expected pace.
Put that question to the decisions in front of you:
A slower quarter tells you how many new households showed up. The shortage figure tells you how many homes were already missing, and right now that's the number that sets the floor under prices and rents.
Read Next
Shared Roofs, Longer Stays: How Affordability Is Rewriting Canada's First Home Plans
Taxing B.C.'s Top Earners Harder Will Leave the Whole Province Poorer
Five Market Warning Signs to Watch Before Trouble Hits, Starting With Credit Default Swaps
Housing Targets Mean Little Without the Trades Workers to Build Them