Canada's Modular Gap Was Never a Factory Problem

About 2 per cent of Canadian homes are built off site. In Sweden the figure is roughly 45 per cent; in Japan, 15. Those three numbers are the most quoted statistic in Canadian construction, and they are almost always told the same way — as a story about a country that has not adopted a technology.

They are better read as a story about a country that never placed an order.

That distinction matters commercially, because the two readings imply opposite strategies. If Canada has a technology gap, the money is in factories, automation and engineering. If Canada had a customer gap, the money was always in the order book — and the moment the order book appears, the constraint moves somewhere else entirely, to the people who supply what goes inside a module rather than the people who weld its frame.

The arithmetic of a surplus

Start with capacity, because the capacity number is not close.

ATCO Structures & Logistics operates five plants in Canada. At its Calgary facility it runs two production lines, and one line turns out one to two modules a day — around 500 a year at full rate. Over the past seven years the company has delivered about 3,000 housing units, which is roughly 430 a year. Read those two figures together and something odd appears: one line at one of five plants, running at rate, would out-produce the company’s average housing book across all five.

The national arithmetic says the same thing. Our 2025 volume estimate for the Canadian market is 17,850 modules. At 500 modules per line-year, the entire country is about 36 production lines running flat out — against a trade association, the Modular Building Institute, with 72 Canadian members, and against a single manufacturer that already runs five plants.

Canada did not have a factory problem. It had a customer problem. Nobody in this market had ever ordered like a manufacturer’s customer — same design, repeated, contracted years ahead, in volume — because Canadian housing has been procured building by building, by owners who buy one building.

What changed was a signature

Build Canada Homes has prioritised up to 4,000 Direct Build units across six federal land sites: Dartmouth, Longueuil, Ottawa, Toronto, Winnipeg and Edmonton. More important than the number is where it sits in the paperwork. Proponents are asked to demonstrate experience and innovation in prefabrication, modular construction and mass timber. Selected design-builders must prioritise Canadian-sourced materials. At least 40 per cent of homes on each site must be delivered below market. Behind it sits the federal Housing Design Catalogue, which publishes standardised, region-specific designs for rowhouses, fourplexes, sixplexes and accessory dwelling units.

Every one of those is a rule about who may bid. Taken together they do something no building code has managed: they make off-site manufacture a qualification criterion rather than a cost decision. A fabricator no longer has to win an argument about whether modular is cheaper. It has to be on a shortlist.

Value the pipeline at the CAD 440,000 a door that the market’s own multifamily line implies — the derivation is set out below — and 4,000 units is about CAD 1.76 billion of addressable work: 32.9 per cent of one entire year of the Canadian modular market, delivered across a handful of years. That is the first multi-year, standardised order book this industry has been offered in public.

The full market path from that starting point is set out in our Canada modular construction market forecast: CAD 5,355.0 million (USD 3,826.9 million) in 2025 to CAD 8,649.9 million (USD 6,181.6 million) by 2033, 6.18 per cent a year. The published announcement of those figures is here.

Edmonton is a test, not a volume event

The Edmonton site is the Village at Griesbach, on the final quarter section of a master-planned community built on the former Canadian Forces Base: 64 hectares, 158 acres, with about 2,500 further homes planned alongside it. Two parcels in the northeast corner are being delivered by Direct Build and the programme caps them at 355 units — 55 townhomes and 300 apartments in a building of 23 metres, about six storeys. The Request for Qualifications closed on 22 December 2025; the Request for Proposals follows in early 2026.

One number in the public record should be thrown out before it gets quoted. A construction-intelligence listing for the 300-unit parcel carries a CAD 20 million value against a 36-month contract — CAD 66,667 a door. That is not a construction cost at any plausible input price, and the programme publishes no project value of its own. Treat it as site works or a partial figure.

That CAD 440,000 is not an assumption, and it is worth showing where it comes from because everything downstream of it depends on it. Two per cent of Canada’s 245,367 housing starts in 2024 is 4,907 off-site homes. The multifamily modular line two years ago was about CAD 2.160 billion. Divide the second by the first and the market implies roughly CAD 440,000 a door — a plausible all-in cost for a mid-rise apartment unit in this country, which is exactly why the cross-check holds. It is also why the CAD 5.100 billion total is never divided by off-site housing: that would produce CAD 1.04 million a home, and the absurdity of that number is the clearest proof that this market is not a housing count.

Size it from the market instead. At CAD 440,000 a door, 355 units is about CAD 156 million: 6.9 per cent of Canada’s 2025 multifamily modular spend, or 2.3 per cent a year across a 36-month term. And 1.66 per cent of Edmonton’s 21,337 housing starts in 2025.

That last figure is the one to hold. The flagship federal modular project in one of the two fastest-growing large housing markets in the country is under 2 per cent of one city’s annual starts. Griesbach cannot move a national number. Its significance is entirely in what it establishes: a client that specifies once and builds 355 times, in a building type that repeats. It is the first entry in a book, and the book is what the industry never had.

Where the money actually moves

Here is the part of the forecast that most readers skip, and it is where the strategic content is.

Modules grow from 17,850 to 24,366 over the period — 3.97 per cent a year. Market value grows at 6.18 per cent. The gap between those two numbers is realised value per module, which rises from CAD 300,000 to CAD 355,000: 2.13 per cent a year of pure content.

Canada is not mainly going to build more modules. It is going to build modules that contain more building — taller, more heavily serviced, and further finished before the truck arrives. That is what a standardised order book does to a fabricator’s product: once the design is fixed and repeated, the rational move is to push labour and risk back up the chain into the factory, because the factory is the only place in the delivery sequence where repetition reduces cost.

So the demand that this programme creates lands first on the component suppliers whose products are consumed per module rather than per building:

A module’s bathroom arrives as a pod, so the sanitary ware specification is decided once per programme and then repeated hundreds of times. The volumes are trivial in national terms; the specification lock-in is total.

Glazing behaves the same way. A standardised window schedule is the difference between a plant that can hold a takt time and one that cannot — which is why the building and construction glass package sits inside the envelope decision rather than downstream of it.

Envelope, sealant and fire-stopping products are consumed per module rather than per building, and the domestic supply chain is thinnest exactly there — the waterproofing and construction chemicals layer that a Buy Canadian priority is intended to reach.

And fire separation between modules is the most inspection-intensive part of closed construction, which is where the method’s practical ceiling sits: above six or seven storeys, the fire protection requirement is what effectively caps the product — and, incidentally, what makes the vertical transportation package in a six-storey building a different commercial proposition from the one in a four-storey walk-up.

The counter-intuitive consequence: the headline market size understates the opportunity for these suppliers. A 6.18 per cent market growing with flat module count is a market where the money is being spent inside each unit, not on more units.

The risk is a signature, not a technology

Which brings us to what should genuinely worry anyone underwriting this market.

Growth is one segment. Multifamily modular carries CAD 2,268.0 million of the CAD 5,355.0 million base and 70.4 per cent of all value growth to 2033, compounding at 9.20 per cent. Commercial grows at 1.80, education at 2.20, healthcare at 1.00. A seven-line market in which one line does seven-tenths of the work is not really a seven-line market.

And that one line is levered to one buyer. Our multifamily rate of 9.2 per cent is deliberately above the 7.4 per cent in the published industry forecast, and the reason is documentary rather than bullish: that forecast was constructed in the first half of 2025, Build Canada Homes launched in September 2025, and its Direct Build requests for qualifications went out in January 2026. The largest demand programme in this market’s history is not in the number.

So state the sensitivity plainly. Hold multifamily at the source’s own 7.4 per cent and the market reaches CAD 8,078.9 million in 2033 at 5.27 per cent — 6.6 per cent below our base case, with every other segment unchanged. Anyone who thinks the programme disappoints should not shade the whole market down. They should move one line and accept that they are making a judgement about one public agency’s delivery capacity.

The evidence that it is delivering so far is timing, not volume: 2029 at CAD 6,805.9 million against a published CAD 6,400 million, a 6.3 per cent premium, all of it programme.

Why the regional split is not a rounding error

Western Canada holds 45.1 per cent of the market and grows at 6.80 per cent. Eastern Canada holds 54.9 per cent and grows at 5.65 per cent. The smaller half is the faster one, and the reason is in the starts data rather than the modular data: Alberta recorded 53,184 housing starts in 2025, up 14 per cent, and Edmonton a record 21,337, up 16 per cent — in a year when Toronto fell 31 per cent and Vancouver 3 per cent. Alberta carries 49.5 per cent of the four western provinces’ starts.

Modular capacity does not travel. A plant is economic within a haul radius, and the demand that justifies Alberta capacity is Alberta demand. That is the single most consequential structural fact in this market, and it is the one assumption a reader should test hardest, because the transport radius is the only link in the chain that is not published anywhere in a citable form.

It also explains something a national forecast obscures: the Griesbach programme is not a Canadian event. It is an Alberta fabricator event, and it will not draw a single module from Ontario.

What to watch

If you sell into Canadian residential construction, the demand signal that now matters is not housing starts. It is programme awards.

Specifically: whether Build Canada Homes converts Requests for Qualifications into executed contracts on two or more of its six sites within the same year. One site at a time is a project. Two or more in one year is an order book, and an order book is what converts 36 lines of latent capacity into a supply chain that has to be specified, priced and scheduled — which is the moment the component demand in this analysis becomes someone’s revenue line rather than a forecast.

For the neighbouring case — the site-assembly counterpart to this analysis, where the constraint is a provincial planning rule rather than an order book — see British Columbia’s small-scale multi-unit housing market. The two markets share a supply chain and almost nothing else.