491,347
homes Oregon needs
over the next 20 years
~700
homes/year the new factory can build
Portland's Terminal 2 plant
~491
committed orders/year it needs
to survive (the missing piece)
6
mass-timber homes delivered so far
Hacienda's prototype 'Casitas'
Oregon built the supply side. The demand side is empty.
Oregon needs 491,347 homes over 20 years (Oregon DAS, Office of Economic Analysis), but its housing supply grew only about 5% from 2020 to 2025 (Willamette Week / Oregon Journalism Project). The idea behind mass timber: build homes in a factory — faster, with a smaller crew, out of Oregon's own wood.
And Oregon has gone further than anywhere else to set it up: the first statewide tall-wood building code in the country (Construction Dive), a national research institute (TallWood Design Institute (OSU / UO)), a $41.4 million federal grant (Oregon State University, College of Forestry), and a 39-acre factory campus on the Portland riverfront (Port of Portland).
The machine is built. It's just idling — because nobody has assembled the one thing a factory needs to survive: a steady stream of committed orders. That gap is this whole story.
What “mass timber” actually is
Mass timber means gluing smaller pieces of wood into big, strong components that can replace concrete and steel. It's not the flimsy 2×4 framing of a house — it's engineered wood strong enough to hold up an 18-story building.
Cross-laminated timber (CLT)
Layers of lumber glued at right angles into big structural panels.
Used for: Floors and walls — the flat slabs of a building.
Glulam
Glue-laminated beams and columns — many boards bonded into one strong member.
Used for: The posts and beams that hold a building up.
Mass plywood panel (MPP)
Veneer-based structural panels — an Oregon invention (Freres Engineered Wood).
Used for: A lighter, wood-efficient alternative to CLT panels.
Panelized
Panels and beams are made in a plant, shipped flat, and stood up on site like a giant flat-pack. The frame goes up fast with a small crew; everything else — plumbing, wiring, finishes — is still built on site. Lower risk.
Volumetric / modular
Whole finished rooms — walls, floors, wiring, plumbing, sometimes fixtures — are built in a factory, trucked to the site, and stacked like Lego. This captures the most factory efficiency and carries the factory-economics risk that is the central pitfall of this whole field.
It's not right for every building
Mass timber has a sweet spot, and outside it, cheaper options win. The rule: ordinary “stick-frame” wood is cheapest for low buildings; mass timber's edge shows up on large, repetitive, mid-rise buildings where labor is expensive (Construction Physics (Brian Potter), citing the 2023 Intl. Mass Timber Report). Try a few combinations:
How tall is the building?
How repetitive is the design?
What's the local labor market?
Verdict
A textbook fit for mass timber
The myth: wood is cheaper. The reality is harder.
This is the crux. The popular pitch — that mass timber is a cheaper way to build — is wrong in a way that, if you build on it, sinks the whole effort.
The myth
“Building with wood is cheaper than concrete and steel.”
The reality
The material costs the same or more — a median premium around 2%, up to 15% (Construction Physics (Brian Potter), citing the 2023 Intl. Mass Timber Report). Savings come from elsewhere — but only under the right conditions.
Lighter foundation
Up to ~60% lighter than concrete on some projects — so foundations shrink and seismic demand drops (a real edge in earthquake country).
Half the on-site crew
Prefabrication can cut the on-site labor force roughly in half — a big win where labor is expensive, and almost nothing where it's cheap.
Faster = cheaper financing
Building ~25% faster (sometimes more) cuts months of construction-loan interest. Speed is, in effect, a form of financing.
The single most important honest finding
Whether mass timber can actually make affordable housing cheaper is still unproven. It's the subject of an active 24-month federal study (HUD USER (PD&R)), and the industry's own affordable-housing guide avoids cost numbers entirely (WoodWorks – Wood Products Council). For affordable housing, any added cost lands straight on an already-tight public subsidy — with nothing to absorb it (Meyer Memorial Trust (hosted by Oregon DLCD)). The honest position isn't “it's cheaper.” It's “it can be — if specific things are fixed first.”
Why a factory's costs depend on how full it is
A factory is a fixed-cost machine: rent, equipment, and payroll cost the same whether the line runs full or half-empty. Spread that overhead across few homes and each is expensive; across many and it nearly disappears. Drag the slider:
Cost per home
$406,429
vs $423,000 built on site
Compared to site-built
−4%
cheaper per home
Verdict
Break-evenCheaper than site-built — but not yet a sustaining business
The same factory builds homes that cost more than site-built when it's half-empty and less when it's full. That hyperbola — fixed cost divided by output — is why a half-empty factory is a money-loser, and why securing committed demand is the whole game.
Modeled on the briefing's default inputs ($423,000 site-built all-in; $355,000 modular marginal cost; $18M/yr fixed cost; 700-home capacity). The durable truth is structural: a half-empty factory loses money, so committed demand is the whole game.
Building it at scale
Say Oregon wanted to build a serious number of affordable, factory-built homes. What would it cost, and who would pay? Most of the money isn't public — federal tax credits and private debt carry the bulk.
Total cost to build
$340M
≈ $340,000 per home, all-in
Public share of that
$110M
32% of the total — about $110,000 per home
Who pays for it
≈ 1–2 years to actually build these. The real ceiling isn't money — it's how fast Oregon can hand out the rationed federal tax credits that anchor every deal. The factory could build faster than the financing pipeline can fill it.
Illustrative, scaled from the briefing's 5,000-unit model (≈$1.7B total, ~⅓ public). For comparison, Portland's 2016 housing bond was $258M and has opened 1,551 affordable homes (Portland Housing Bureau); Metro's 2018 regional bond was $653M (Metro).
What Oregon has already built
This isn't vaporware — most of the hard pieces exist and are funded, further along than most people realize. The honest caveat: the actual housing delivered so far is six prototype homes.
$41.4M federal EDA grant
Funded2022The Oregon Mass Timber Coalition won a $41.4M U.S. EDA Build Back Better Regional Challenge grant; Oregon & Washington later won the federal PNW Mass Timber Tech Hub designation (2023).
Oregon State University, College of ForestryTerminal 2 Innovation Campus
Planned2024–28A 39-acre 'Mass Timber and Housing Innovation Campus' on the Portland riverfront, run by the Port of Portland — the physical home of the industry.
Port of PortlandZaugg Timber Solutions factory
Under lease2025A century-old Swiss firm's North American affiliate signed a long-term ground lease at Terminal 2 to build a ~100,000-sq-ft factory making ~700 homes/year (interim ~2026, permanent ~2028). A lease — not a guaranteed order book.
Port of PortlandHacienda's six Mass Casitas
Delivered2023Oregon's largest Latino-serving housing nonprofit built six mass-timber modular homes at Terminal 2, delivered across the state — the clearest delivered link between the industry and affordable housing. Six homes: proof of concept, not scale.
Hacienda CDCModular Housing Development Fund
Funded2023HB 2001 put $20M into four $5M grants to factories (Blazer, Intelifab, Pacific Wall Systems, Zaugg). Supply-side capacity money — it commits no forward orders.
Oregon Housing and Community ServicesFactory-Produced Housing Initiative
Funded2025HB 3145 added $25M for up to five factory-produced-housing projects. A project-side set-aside — again, not a demand-aggregation lever.
Oregon Housing and Community ServicesThe graveyard
Factory-built housing has failed repeatedly for a century, and the failures are consistent — and avoidable only if you understand them. None of these died because the buildings didn't work. They died on economics.
FrameworkPortland, OR
A 12-story, all-mass-timber high-rise in the Pearl District — the first U.S. mass-timber high-rise to win a building permit (2017), with all 60 units affordable at ≤60% of area median income. Construction cost rose from $26M to over $34M (about a third) while the tax-credit equity behind it fell ~20% after the 2017 federal tax law.
The lesson: Not engineering, not code — both were cleared. A hard-cost premium is fatal for affordable housing because it lands on a tight subsidy gap with nothing to absorb it.
KaterraUSA · raised over $2 billion
A SoftBank-backed startup that tried to do design, manufacturing, and construction all at once, across a national market where every state regulates modular differently. It filed for bankruptcy in 2021.
The lesson: Doing too much, too fast, across inconsistent codes — never achieving the repetition that makes a factory pay.
Ilke HomesUnited Kingdom
A flagship UK modular factory that went into administration in June 2023 owing ~£319M and shedding ~1,150 jobs — while holding a roughly £1 billion order book / ~4,200-home pipeline.
The lesson: The decisive lesson: a paper pipeline is not survival. If you can't convert orders into delivered, profitable volume fast enough to carry the factory's overhead, you're toast.
L&G Modular HomesUnited Kingdom
Backed by one of the UK's largest investors, it halted production at its Leeds-area factory in May 2023, ultimately booking £279M in cumulative losses over the life of the business.
The lesson: Deep pockets don't save a factory that can't fill its line — fixed cost runs whether the line moves or not.
House by Urban SplashUnited Kingdom
A modular venture co-owned by Urban Splash, Japan's Sekisui House, and the UK government's Homes England — into administration in 2022.
The lesson: Even with a government co-owner and a deep-pocketed industrial partner, the factory economics still bit.
The decisive lesson from the UK, which ran the closest thing to a national demand push: demand is necessary but not sufficient. Diffuse requirements didn't save the factories — and the UK's new 2026–2036 program dropped its modular target entirely (GOV.UK). The demand has to be guaranteed, standardized, and big enough to hold a factory above break-even.
The missing piece: someone to aggregate the demand
The demand for these homes exists many times over. The problem is that it's fragmented, mistimed (it clumps around tax-credit deadlines), and stuck in a standoff: no buyer commits to an unproven factory; the factory can't prove itself without commitments.
A demand aggregator
One entity signs framework deals with many buyers — 'when you build, source through us' — then smooths their lumpy orders into a steady factory calendar, because one buyer's slow year offsets another's busy one.
A public demand floor
Government commits to buy a floor number of homes if private demand falls short. It can't go bankrupt and always needs housing — so it can hold the line through downturns. It spends creditworthiness, not cash.
A private operator runs the line
Government guarantees the demand; a disciplined private firm runs the factory. Like public transit: the public commits to the service, a contractor operates it.
No one has proven this yet. The most-promoted U.S. example — a California master agreement for 560+ homes — is an operational framework, not a binding guaranteed purchase, and has delivered zero units so far (Guerdon Modular Buildings). Oregon wouldn't be copying a proven model; it would be building one. That's both the opportunity and the risk.
The other two payoffs — honestly
Mass timber is rare among economic levers because it can serve housing, jobs, and climate at once. Each of those payoffs is real — and each has a catch worth stating plainly.
A feedstock wrinkle: the federal government manages ~60% of Oregon's forestland but supplies only ~10% of the harvest, and lumber is the single biggest cost in a CLT panel — so the “rural jobs” story runs into federal forest policy (Oregon Forest Resources Institute).
The climate case (both sides)
Wood stores carbon and avoids the heavy emissions of concrete and steel. A 2025 study found scaling mass timber could net 25.6–39 billion tons of CO₂ by 2100 (Nature Communications). The honest catch: most of that is forest storage, not the panels, and it comes bundled with more logging — natural forests shrink by 8.1–18.9 million hectares in the model. Real and contested at once.
The jobs case
The sector could support up to 17,000 jobs in Oregon (Oregon Mass Timber Coalition) — factory work ($50,000–$80,000), forestry and mill jobs in rural towns, and union construction trades — most without needing a four-year degree, spread across the urban–rural divide.
Fix what blocks it — don't just “scale it”
Mass-timber housing in Oregon is a rare triple win if it's done right, and another well-funded grave if it's done wrong. What decides it isn't the technology — it's whether committed demand shows up before the factory has to carry its overhead.
Demand is guaranteed, standardized, and big enough to keep the line above ~70% full — through downturns.
The factory stays asset-light; the public absorbs the ramp-up risk but never owns or runs the line.
The product is standardized and repeatable, not a bespoke building every time.
Financing fits the front-loaded cost curve, stacking tax credits + bonds + a green layer + land.
The state pulls its own levers — rewarding factory-built homes in how it hands out tax credits.
Everyone respects the honest economics: ~10% cheaper at full tilt at best, and only at full tilt.
The window is now: the Portland factory reaches early production around 2026 and full operation around 2028. The demand-side answer has to exist before the line must carry its overhead — the 2026–2028 window. That timing, not the technology, is what makes this urgent.
Where these numbers come from
Built from a deeply-researched briefing, with each load-bearing claim re-checked against a primary or credible source in June 2026. A few figures are modeled estimates (the factory cost curve, the scale illustration) and are labeled as such. Where a popular figure didn't hold up, it was corrected, not repeated.