What the Modular Building Institute's $20.5 billion tells you — and what a trade body can't.

If you have read anything about modular construction in North America this year, you have read something that traces back to the Modular Building Institute. Its July release — a $20.5 billion US permanent modular market in 2025, growing 6.5 percent a year to 2030 — is the figure in every deck, every trade story and, we should say, on the front page of this site.¹ MBI also runs the World of Modular conference, the Awards of Distinction, the Modular Advantage magazine, the Inside Modular podcast, the government-affairs desk that tracks code adoption state by state, and, as of August, a federal lawsuit against the State of Oregon.⁵

That is a remarkable amount of infrastructure for a sector at five percent of construction activity, and the industry is better for it. It is also a set of roles — statistician, promoter, convenor, publisher, litigant — that pull in different directions. This piece is not a takedown. It is a reading guide: what to take from modular.org at face value, what to read with the by-line in view, and the questions a trade association is structurally unable to ask, which happen to be the questions this publication exists for.

The number

Start with the statistic, because it is the most useful thing MBI produces and the least examined. The 2026 report was developed with FMI Consulting and describes "updated methodologies, enhanced regional analysis, and robust data from both primary and secondary sources." It identifies multifamily as the largest segment, flags office and data centres as emerging, and names the US West as the leading region.¹ The full PMC report, the Canadian report and the Relocatable Buildings report are available free to members and for purchase by everyone else; the public gets the press release.

What the public release does not disclose is the number of firms surveyed, the response rate, how "permanent modular construction" is bounded at its edges (bathroom pods? panelised hybrids? data-centre skids?), or how the estimate is reconciled year over year when methodology changes. We looked for an independent critique of the methodology and found none, which is itself telling: nobody outside MBI has the underlying data to critique. That does not make the number wrong. It makes it unaudited, and an unaudited number from an organisation whose members benefit from a larger one should be quoted with its provenance attached, which is what we do.

What MBI does well

Three things, and they are not small. First, the government-affairs desk is the only place tracking off-site construction regulation across every US state in something like real time: which states have adopted ICC/MBI 1200 and 1205 (Virginia first, Colorado second with 1210 as well), which have prevailing-wage bills moving, which have new third-party inspection rosters.⁶ Our own Reciprocity Tracker would be thinner without it, and we say so on the page.

Second, the standards work. The 1200-series exists because MBI and the International Code Council put years into it. Whatever one thinks of the association's advocacy, a consensus fabrication and inspection standard that a state can adopt by name is the precondition for everything else in this industry, and no one else was going to write it.

Third, the magazine has improved. The August 2026 issue of Modular Advantage carried pieces on the on-site/off-site interface, DfMA in healthcare and mass timber, factory testing, Canadian contract structures and a project profile of Alora at Dry Creek in Colorado — 329 units across five buildings, built on standardised unit prototypes.³ These are practitioner topics, treated at practitioner depth. Read them.

Read the by-line

Read them, though, with the by-line in view, because most are contributed by member companies about their own work. The Alora profile is a case in point: it is a useful account of standardisation and early collaboration, and it is written from inside the project team, with no cost figure, no schedule variance against the original plan, and no comparison to what a site-built equivalent would have cost in the same submarket. That is not a criticism of the authors, who wrote the article they were asked to write. It is a description of the genre. Trade-magazine project profiles are the industry's record of what it is proud of, and they are indispensable for that; they are not the record of what it learned.

The same applies to the flagship research pieces. Rebuilding Hope, MBI's account of the Lahaina wildfire response, credits Fading West, Guerdon, ProSet and Marex with delivering more than a hundred permanent modular homes on a timeline of days per unit in the factory and roughly two weeks to barge to Maui.⁴ It is a good story and probably a true one. We could not find an independent assessment of outcomes, unit cost or resident experience, and the one cost figure in the piece is a benchmark an executive offered, not an audited actual. An independent publication's job is to go and find out.

And the "What is Modular Construction?" explainer — the page most first-time visitors land on — is one-sided by design: speed, quality, sustainability, no trade-offs.² That is what an association's front door should say. It is also why a developer who reads only that page and then commissions one building is surprised by the deposit schedule.

Seven questions a trade body can't ask

None of what follows is a failing of MBI's staff. It is what happens when the organisation that measures an industry is funded by the industry it measures.

Which projects failed, and why? The failure record — 461 Dean Street's litigation, Toronto's 63 percent overrun,⁹ Katerra, Veev, the Harbinger asset sale in Vallejo¹⁰ — is where the industry's most useful lessons sit. You will not find a post-mortem on modular.org, because the parties are members or were.

What does the capital stack actually look like? The most recent financing resource in MBI's research library is a 2020 toolkit produced with Fannie Mae and NIBS. Nothing since on deposits, loan-to-cost haircuts, unsecured factory work or in-transit insurance — the frictions that, on our reading, explain slow adoption better than any code does.

Do the approvals travel? MBI tracks adoption state by state. It does not publish a synthesis of which states accept which others' approvals, because the answer — mostly they don't — is an argument against the scale story.

What does each factory actually produce? No plant-level capacity, utilisation or output figures exist anywhere in the public record, MBI's or otherwise. Our Factory Ledger makes that absence visible, one "none published" at a time.

Is it greener, or is it usually greener? MBI's research page links to third-party carbon studies. It does not engage with the ones finding higher upfront embodied carbon in steel-framed towers.

What happens in Canada? A Canadian PMC report exists (with gaps in the series) and the August magazine carried one article on Canadian contracts. Against the depth of US state coverage, the second-largest market in the association's territory is an afterthought.

What about the workers? The Oregon lawsuit and the New York and Pennsylvania prevailing-wage fights are covered on modular.org as advocacy — accurately, from the industry's side.⁵ The underlying question of what factory jobs pay relative to site jobs, and whether the productivity gain is shared, is not covered at all. We take it up in the next piece.

A trade body's job is to make the case. Someone else's job is to check it. Both are necessary; only one of them is currently well funded.Volumetric Building editorial position

Who fills the gaps

Partially, others do. McKinsey's 2019 report remains the strategic frame, and it is more candid than its citers about cost savings being "the exception, not the norm."⁷ The Terner Center at UC Berkeley is the best policy-academic source in North America; its March 2026 Potential Pathways report set out seven policy themes and forty actions for California, drawn from lenders and trades as well as manufacturers, and is the closest thing to an independent agenda the sector has.⁸ Construction Dive does actual reporting on financials and M&A. The Canadian Home Builders' Association and CSA Group cover the A277 regime. A handful of practitioner blogs — Offsite Straight Talk broke the Harbinger WARN story — do the unglamorous work of watching filings.¹⁰ We keep an annotated list on the Landscape page, including what each source is for and what it will not tell you.

What none of them do is cover the sector continuously, across the border, with the failures and the financing in the frame. That is the gap.

How to use this

Read modular.org for the regulatory tracker, the standards and the practitioner articles; it is the industry's best record of itself. Attach provenance to every number you take from it. When a project profile lacks a cost, a schedule variance or a comparison, assume the omission is informative. Then come here, or to Terner, or to the auditor's report, for the other half. The industry does not need less MBI. It needs a second source, and it has never had a durable one.