Office Furniture (including Fixtures) Manufacturing (United States)

NAICS 2022 code 33721 — an investor's rollup primer

NAICS (the North American Industry Classification System) is the standard the U.S. government uses to sort businesses into industries. This is a NAICS industry (five-digit) that bundles four more specific child industries. This primer synthesizes across those four; each has its own detailed primer.

---

1. Overview

The federal label is a trap. "Office Furniture (including Fixtures) Manufacturing" sounds like the business of making desks and chairs — but by the government's own numbers, office furniture proper is a minority of it. Roughly two-thirds of the level's revenue comes from two things that are not office furniture at all: custom architectural woodwork/millwork (the paneled boardrooms, reception desks, and casework built into nice buildings) and store fixtures, shelving, and lockers (the gondola aisles in a grocery, the display cases in a jewelry counter, the lockers in a school). Actual office furniture — wood desks plus metal chairs, cubicles, and filing — is only about a third.[1][2]

That single fact reframes the whole industry. This is not one business; it is four adjacent manufacturing trades stapled together by the census because they share materials (wood, panels, steel, laminate) and factories, but which answer to three different demand cycles: the office cycle (return-to-office, corporate spending), the non-residential construction cycle (new buildings and interior fit-outs), and the retail capital-spending cycle (stores opening and remodeling). An investor who treats 33721 as "the office-furniture industry" will misread most of it.

The imbalance has also been widening. Two decades of census benchmarks show the two halves moving in opposite directions: wood office furniture had 569 establishments and 24,301 workers in 2002 against 299 and 13,196 today, while custom millwork went from 1,552 companies and $4.15 billion of shipments in 2002 to 2,227 firms and ~$10.0 billion — growing and fragmenting further at the same time.[3][6] The center of gravity has been drifting toward the craft and fixtures trades for a generation.

Why it matters to investors: at ~$28 billion of U.S. factory shipments and ~105,000 workers, this is a small, mature, deeply cyclical corner of American manufacturing — but the four children could hardly be more different in who owns them and how you'd buy in. Two of them (the office-furniture pair) have consolidated into a two-company oligopoly you can own on the stock market. The other two (millwork and fixtures) are thousands of small private shops whose listed representation exists only as undisclosed slivers inside much larger companies — private-market territory where the opportunity is to buy or roll up companies, not to buy shares. Reserve tickers and deal names for Sections 4 and 10.

---

2. What's inside — the four children and how they differ

The level splits into four NAICS industries. Lead with the contrast, because the differences across them are the real story:

Shares are each child's revenue and employment as a percent of the level totals in Section 3; concentration ratios and firm counts are 2022 Economic Census / 2023 County Business Patterns.[1][2][3][4][5][6]

How to read the table — five contrasts that matter:

1. Size is the opposite of what the name implies. The two "furniture" children (337211 + 337214) are only ~32% of revenue. The craft-and-fixtures children (337212 + 337215) are ~67%. The industry's center of gravity is millwork and store fixtures, not desks.

2. Concentration is bimodal, and the spread is enormous. The two office-furniture children are genuinely concentrated — top-four shares of 57.7% (337214) and 54.3% (337211), with HHIs of 1,075 and 944.9. The two craft/fixtures children are the opposite: HHIs of 150.5 (337215) and 24.4 (337212), the latter effectively zero and among the most atomized manufacturing industries in America. That is a roughly 44-fold spread in concentration inside one five-digit code. The level's own concentration score therefore lies (Section 3).

3. Ownership follows concentration. Concentrated segments are where the public companies live; fragmented segments are almost purely private. That is why "how you'd invest" flips completely across the row.

4. They ride different cycles. Wood and non-wood office furniture rise and fall with the office (corporate hiring, return-to-office). Millwork rides non-residential construction and fit-outs. Fixtures ride retail store spending plus warehouse/e-commerce build-out. A recovery in one does not imply a recovery in the others.

5. Even the federal definition of "small" differs. The SBA size standard is 500 employees in fixtures (337215), 1,000 in wood office furniture (337211), and 1,100 in non-wood office furniture (337214) — the government itself treats these as structurally different trades when it sets contracting thresholds.[38]

Where the code lines are drawn (so you don't double-count with neighbors): stock/catalog moldings and flooring are NAICS 321918 and factory wood windows and doors 321911, not 337212; wood kitchen cabinets and laminate countertops are 337110; institutional (school/lab/library/auditorium) furniture is 337127; refrigerated display cases are 333415; metal safes and vaults are 332999; on-site finish carpentry (238350) and on-site installation of modular systems and partitions (238390) are construction, not manufacturing; and the independent dealers who sell, deliver, and install contract furniture are wholesale trade (NAICS 423210), so their large service markup sits outside this level entirely.[3][4][5]

---

3. How big it is (the rollup)

Federal statistics for NAICS 33721 as a whole (our ground-truth figures):

Average pay works out to roughly $60,000 per worker per year (payroll ÷ employment),[2] and the children's own averages cluster tightly around it — about $55,000 in wood office furniture, $58,700 in fixtures, and $62,000 in millwork.[3][4][6] The four children reconcile to the level: establishments (299 + 2,371 + 189 + 789 = 3,648) and employment (13,196 + 45,382 + 17,471 + 28,674 = 104,723) tie out exactly, and receipts add to ~$28.4 billion.[1][2] Firm counts are the one line that does not tie — the children report 316 + 2,227 + 174 + 710 = 3,427 against 3,409 at the level — because a company operating in two of these industries is counted once here and twice below.

The concentration number is the trap to avoid. The level's Herfindahl-Hirschman Index (HHI — a standard concentration gauge that runs to 10,000) is just 109.7, with a top-4 firm share (CR4) of 17.5%, CR8 of 24.2%, CR20 of 33.8%, and CR50 of 45.2%.[1] Taken at face value that says "highly competitive, no dominant player." It is misleading at the level. The score is dragged down by the ~2,900 tiny millwork and fixture shops (337212 + 337215) that swamp the firm count. Inside the office-furniture children, concentration is genuinely high — HHIs of 944.9 (337211) and 1,075 (337214) with top-four shares of 54.3% and 57.7%[1] — and the recent mega-mergers (Section 8) pushed effective concentration higher still. Note that even those two sit near or below the 1,000 line that federal antitrust practice calls "unconcentrated," precisely because each has a genuine long tail beneath a top-heavy few.[6] Never quote the level's HHI as if it describes the office-furniture business. Aggregation here averages a concentrated oligopoly with an atomized craft trade and produces a number that describes neither.

Measurement caveats specific to the rollup:

• The children are not all measured the same way. The fixtures child has two federal revenue figures — ~$9.2 billion from the 2022 Economic Census (the vintage used in the level total above) and $8.807 billion from the 2023 Annual Integrated Economic Survey.[4] Its employment also differs by source: 28,674 in County Business Patterns (2023) against roughly 37,900 payroll jobs in BLS's December 2022 series, which after a redesign folded 337215 into a broader NAICS 3379 grouping and no longer breaks the industry out at all.[4]
• Price visibility is now uneven across the level. BLS discontinued the dedicated NAICS 337211 producer-price indexes with the October 2025 release, so wood office furniture no longer has a government six-digit price series.[6] The other two large children still publish: millwork's PPI stood at 187.0 in June 2026 (December 2003 = 100, about 87% cumulative), and the nonwood storage-rack index at 276.6 in June 2026, still ~41% above January 2020 after peaking 64% higher in May 2022.[3][20] You can separate price from volume in two children and no longer in a third.

Undercount and boundary caveats (they run in both directions):

• Millwork is genuinely undercounted. County Business Patterns counts only employer establishments, so the many one-person and informal custom-woodworking sole proprietors ("nonemployers") are excluded — we have no published nonemployer count for 337212, so we won't put a number on it, but the direction is clear. A large share of finish woodwork is also installed on site and booked under construction codes, and some custom work is done in-house by cabinet, furniture, or general-contracting firms classified elsewhere. The true craft footprint is larger than the ~45,000 millwork jobs shown.[3]
• Fixtures leak across the boundary too, in both directions. A firm building branded store environments in wood may be booked in millwork rather than fixtures; the line between 337212 and 337215 is blurry in practice and many "store-fixture" firms straddle both.[3][4]
• Everything here is U.S. production, not U.S. consumption. Shipments exclude imports, which are large in furniture and lighter fixtures — Vietnam and China supply a big share of what American offices and stores actually buy, and China is the world's largest office-furniture exporter. Heavy steel goods like gondola shelving and lockers are freight-protected and stay mostly domestic; lighter display items import freely. Domestic consumption exceeds the $28 billion factory line.[5][36]
• This is not the "office furniture market" you read about. BIFMA and S&P Global put the U.S. "business and institutional furniture" market at $15.27 billion (2023), $15.39 billion (2024, preliminary), and $16.07 billion (2025, preliminary) — a series BIFMA broadened in 2022 beyond office furniture into healthcare, education, hospitality, and other institutional verticals, spanning several NAICS codes and materials.[12] Don't compare it to the $28.38 billion here, which spans four trades but counts only domestic factory output. The nearest independent check on the level itself is IBISWorld's estimate of roughly $30 billion for office-furniture manufacturing as a whole — the same neighborhood as the census figure.[13]

Net effect: the office-furniture children are well-captured factory businesses whose market is understated by import exclusion; the millwork child is understated by both imports and its informal/construction-booked tail. The level modestly understates the true economic footprint of these trades.

---

4. The investable universe — where value concentrates across the children

The blunt reality: public-market value is concentrated in the two smallest children, and the two biggest children are almost entirely off the public markets. That inversion is the defining feature of this level.

Public exposure — a pure-play pair, then a scattering of slivers:

Further out: Berkshire Hathaway (NYSE: BRK.A/BRK.B) owns Marmon Retail Solutions (L.A. Darling, UNARCO) in retail fixtures and rack; Griffon (NYSE: GFF) holds 43% of Veritage Brands (ClosetMaid) after ONCAP took 57% in June 2026; and foreign-listed Quadient (Euronext Paris: QDT, 20,000+ Parcel Pending locker sites) and InPost (Amsterdam: INPST) play the parcel-locker theme. Leggett & Platt exited the fixtures business entirely, selling its store-fixtures operations to Lozier in 2014.[4][39]

Two things follow. First, there is no dedicated exchange-traded fund (ETF) for this level and still no public pure-play in the two largest children. Second — and this is a correction to the simple version of that story — the listed threads into millwork and fixtures are more numerous than a quick scan suggests, but in every single case the relevant revenue is buried inside a much larger parent and not separately disclosed. You can get exposure; you cannot get a clean read. Anyone marketing "the fixtures stock" or "the millwork stock" is selling a diversified or foreign proxy.

Private exposure (this is where most of the industry actually lives):

• Office furniture (337211/337214): large family- or employee-owned makers — Haworth (Holland, Michigan; $2.5 billion of global sales in 2024, rising to $2.7 billion in 2025, up 8%), KI (Krueger International) (100% employee-owned), Teknion, Global Furniture Group, Humanscale, OFS, and the Jasper, Indiana family cluster (Indiana Furniture, Jasper Group, Versteel) that calls itself the "Wood Office Furniture Capital of the World." Foreign brands — Vitra, Okamura, Kokuyo — compete through imports and U.S. subsidiaries.[5][6][30][37]
• Millwork (337212): thousands of family shops plus a consolidating layer — Stevens Industries (wholly employee-owned; large commercial casework), JBI Interiors (~$150M), Beaubois (Norea Capital owns 40%+ since late 2024; bought Four Daughters Millwork in January 2026), Adams Group (acquired Mill-Rite Woodworking), Fetzer, Merritt Woodwork, Glenn Rieder, Mission Bell, Imperial Woodworking, Woodwork Corporation of America, Modern Woodcrafts, and Fairfield County Millwork (sold to Ashford Sato in 2026). Note the change: idX, long cited as a private platform, now sits inside UFP Industries.[3][24][28][29]
• Fixtures/shelving/lockers (337215): the largest names are private or employee-owned — Lozier (the biggest U.S. retail-shelving maker; the fixtures primer puts it at ~$500–700 million and the FDMC-300 ranking at $705 million), Madix, Streater, Interlake Mecalux, Spacesaver (ESOP — Employee Stock Ownership Plan), Lyon, Penco, Republic, List/Hallowell, Edsal, Frazier, Steel King, Ridg-U-Rak, DeBourgh, Hollman (the largest U.S. locker maker), Salsbury, and Metro/InterMetro (owned by Italy's Ali Group since 2015).[3][4][27]

One number captures how far below the radar these two trades sit: across the roughly 50 largest tracked North American architectural-woodwork and retail-fixture manufacturers, combined 2025 sales were only about $3.5 billion (~$1.5B millwork, ~$2.0B fixtures) — against the roughly two-thirds of a $28.38 billion level those trades represent.[27] The overwhelming majority of the revenue is in shops nobody ranks.

Bottom line: to own the office-furniture third of this industry, buy HNI or MillerKnoll. To own the millwork-and-fixtures two-thirds, you generally have to buy a private company — there is no ticker.

---

5. How the money works

Despite the four different end-markets, the children share one economic engine: cyclical, fixed-cost manufacturing where profit comes from keeping expensive plant busy and defending the spread between input cost and selling price. The levers:

• Orders and backlog are the pulse. Most of this is project-based — a corporate relocation, a building fit-out, a retail rollout — so orders lead revenue by weeks to months. Backlog and book-to-bill are the leading indicators across all four children; MillerKnoll carried roughly $0.76 billion of unfilled orders at fiscal year-end 2025, and Steelcase's 6% Americas order growth in fiscal 2025 was read as an early recovery signal.[5][10] In millwork specifically, revenue is recognized over time (percentage-of-completion) like construction; in fixtures, multi-year retail "programs" supply the recurring base.[3][4]
• Capacity utilization drives margins. Factories (CNC — computer-numerical-control — routers and panel saws for wood; presses, roll-formers, welders, and paint booths for metal) carry heavy fixed costs. Full plants spread overhead and expand margins; idle plants between projects destroy profitability. This operating leverage is why the whole level is so cyclical, and it works in reverse fast: showrooms, design teams, and dealer-support infrastructure do not shrink as quickly as orders.[5]
• Margins are thin and sit in a narrow band across very different businesses. MillerKnoll's North America Contract segment ran a 35.7% gross margin and 6.2% GAAP operating margin in fiscal 2025; LSI's Display Solutions segment, the closest listed read on fixtures/millwork, ran 18% gross and ~8% operating; custom woodwork shops typically run low-to-mid double-digit gross and single-digit net margins.[3][10][25] (The children cite different figures for HNI's fiscal-2025 workplace-furnishings GAAP operating margin — 8.5% in one, 9.5% in the other, with 10.5% non-GAAP — so treat the precise number with care.[5][6])
• Input costs split by material — and both are tariff-exposed. The wood children (337211, 337212) live on hardwood, veneer, plywood, MDF (medium-density fiberboard), and HPL (high-pressure laminate). The metal children (337214, 337215) live on steel and aluminum, where Section 232 tariffs (Section 7) flow straight into cost and get passed on as surcharges of 25–50% during tariff episodes.[4][19] Contract pricing lags, so an input spike can squeeze margins for a quarter or two before list prices catch up — the nonwood storage-rack PPI's 64% climb from January 2020 to May 2022 is the sharpest recent illustration.[20]
• Stock vs. custom is the margin story. Commodity product (basic HON desks, catalog gondola shelving, standard lockers) sells high-volume at thin margins and competes with imports. Custom, design-led work (premium seating, branded store environments, seismic rack systems, spec-driven architectural woodwork) is stickier and higher-margin. The best operators mix both.
• Freight quietly draws the map — in one child more than the others. Shelving, racking, and lockers are heavy and bulky with low value per pound, so shipping is a large share of delivered cost. That favors regional plants near customers, protects domestic makers of the heaviest goods from imports, and means a national retail rollout is won partly on logistics.[4] The office-furniture children face no comparable protection at the value end.
• Two profit pools you don't see in these codes. The independent dealer channel captures the design/delivery/installation markup on contract furniture (NAICS 423210, wholesale) — MillerKnoll put 53.7% of its fiscal-2025 sales through independent dealers, which is also a credit exposure, since dealers often hold the end-customer receivable.[6][10] And millwork shops carry slow working capital — they buy materials and pay craftsmen up front, then bill on milestones while the customer holds back retainage of 5–10% until the whole building is done.[3]
• Labor is a real cost and, increasingly, a constraint — ~$60,000 average wage across the level,[2] with skilled craft woodworkers repeatedly flagged as among the scarcest trades. BLS projects U.S. woodworking employment to decline 2% through 2034 while still generating ~21,400 annual openings, almost all replacement hiring.[3]

The scorecard: orders/backlog, capacity utilization, gross and operating margin, input-cost pass-through, and (for millwork) working-capital discipline. Mature office-furniture leaders like HNI also generate steady free cash flow and pay dividends.

---

6. What drives demand

The level's distinctive feature is that its four children answer to three separate demand cycles — so there is no single "demand driver" for 33721:

• The office cycle (337211 + 337214). Corporate capital spending, white-collar hiring, and — the swing factor of the decade — return-to-office versus hybrid work. Office attendance had recovered to roughly 72.6% of pre-pandemic levels by 2025, and the 2025–26 tone has tilted toward tighter mandates.[5] Hybrid is not purely negative: it shrinks assigned-desk and filing demand but drives redesign toward collaboration zones, touchdown stations, acoustic/privacy products, and reconfigurable layouts. The children disagree on how bad the real-estate backdrop is — NAIOP reported U.S. office vacancy at 11.8% in its second-quarter 2025 forecast while CBRE cited ~20.7% in the third quarter of 2025, a gap driven by methodology and coverage. Either way, new office construction is near its lowest since the 2008–09 financial crisis, partly offset by "flight to quality" refurbishment.[6][16][17]
• The non-residential construction and fit-out cycle (337212). Offices, hospitals, schools, courthouses, hotels, labs, and data centers buy architectural woodwork during build-out and refresh. The aggregate hides sharp divergence: U.S. non-residential construction spending reached $1.226 trillion in 2024, up 7.0%, but education rose 8.5% and healthcare 5.1% while commercial fell 10.6% and lodging 5.6%; private non-residential was running at a $738.7 billion seasonally adjusted annual rate in May 2026, down 0.3% from April.[3][18] The leading indicator is the AIA (American Institute of Architects) Architecture Billings Index (ABI) — a monthly diffusion index where 50 is breakeven; architects bill 9–12 months ahead of construction spend. Update from the prior read: the ABI did edge toward breakeven in early 2026 (49.8 in March) but slipped back to 47.3 in June 2026, with architecture firms now 41 months without a majority seeing billings growth. Institutional (healthcare, education, government) remains the firmest segment.[3][14][15]
• The retail capital-spending cycle (337215). Store openings and remodels are the biggest driver — grocery, drug, discount, and dollar chains refresh interiors on a cycle and keep spending (Target ~$1 billion/year on stores; Dollar General planned 4,700 real-estate projects for 2026). Physical retail has proven resilient (~80% of shopping is still in-store, with shopping-center vacancy near a two-decade low), and e-commerce has created new demand for warehouse racking and parcel lockers even as it pressures traditional fixtures: e-commerce was 16.9% of U.S. retail sales in Q1 2026, growing 9.8% year over year against 3.9% for total retail.[4][21][22] Warehouse automation is shifting the mix further toward firms that can supply structural engineering and integration rather than commodity beams and uprights.
• Cross-cutting floors under all four: a replacement cycle (furniture, fixtures, and casework wear out and are refreshed every ~7–15 years regardless of new construction) and government/institutional procurement through GSA (General Services Administration) schedules, with steadier healthcare, education, and government verticals cushioning the corporate cycle.[33]

---

7. Regulation

Light-touch overall, and mostly standards- and trade-driven rather than heavily licensed. The common threads across the children:

• Formaldehyde emissions (the key product rule wherever panels are used). Composite wood panels (particleboard, MDF, hardwood plywood) and the furniture, casework, and wood fixtures made from them must meet strict formaldehyde limits under TSCA Title VI — Title VI of the Toxic Substances Control Act, enforced by the U.S. Environmental Protection Agency (EPA) — harmonized with California's CARB (California Air Resources Board) Phase 2 standard, and be certified and labeled. This binds three of the four children, not just the wood pair.[3][4][31]
• Voluntary but near-universal product standards. For furniture, the ANSI/BIFMA standards (American National Standards Institute / Business + Institutional Furniture Manufacturers Association) govern strength, durability, and stability, with ANSI/BIFMA e3 underpinning the four-tier BIFMA LEVEL certification and M7.1/X7.1 governing low-VOC (volatile organic compound) emissions used in LEED credits and GREENGUARD. For architectural woodwork, the AWI Standards (Architectural Woodwork Institute) and its QCP (Quality Certification Program) are effectively mandatory for institutional and high-end projects — a genuine competitive moat, since a QCP licensee must control a manufacturing facility and supervise the labor it certifies. For racking, ANSI/RMI MH16.1 is the principal design specification.[3][4][5][32]
• Trade policy — the swing variable, and it now cuts finer than tariffs alone. Section 232 tariffs on imported steel and aluminum (raised to 50% in 2025) hit the metal children directly, and Section 301 tariffs on Chinese goods plus newer furniture-specific actions reshape import competition for all four.[19][36] On top of that sit product-specific antidumping and countervailing duties in the fixtures child: Commerce imposed measures on Chinese steel racks, the USITC maintained the boltless-steel-shelving orders on China after a sunset review, and further cases have covered prepackaged boltless shelving from India, Malaysia, Taiwan, Thailand, and Vietnam. Changes in duty scope or country of origin can move domestic price competition quickly.[4][35]
• Government procurement. Federal buying flows through GSA schedules and the Trade Agreements Act (TAA) — which bars China-made goods and advantages domestic makers — plus Buy American content rules and, for some furniture, mandatory AbilityOne sourcing (made by people who are blind or have significant disabilities). Government work also carries procurement-integrity, audit, and suspension/debarment exposure.[5][33]
• Factory, worker, and building safety. OSHA (Occupational Safety and Health Administration) rules on woodworking dust (a combustible and carcinogen risk), metal fabrication, finishing-line VOCs, and secured storage of materials; EPA's wood-furniture and metal-furniture-coating NESHAPs (National Emission Standards for Hazardous Air Pollutants) at major sources; ADA accessible-aisle and checkout dimensions for fixtures; and fire/flame-spread building codes on finished interior products. Recorded injury rates are low — 2.2 cases per 100 full-time workers in both office-furniture children.[3][4][5][6]
• Timber sourcing. The Lacey Act requires declarations that imported wood was legally harvested; USDA APHIS began Phase VII in December 2024, expanding covered products, and from January 2026 declarations must be filed electronically.[3][6][34]

---

8. Consolidation

The parent-level story used to be "two halves consolidating in opposite styles." The revised children support a sharper version: three distinct consolidation styles, and the difference is where the investment thesis lives.

• The office-furniture children (337211 + 337214): dramatic, top-down mega-mergers. In five years the historic "big five" collapsed into a two-company race: Herman Miller + Knoll → MillerKnoll (2021, initially valued at ~$1.8 billion); HNI + Kimball International (2023); and HNI + Steelcase (closed December 10, 2025), creating a ~$5.8 billion leader with a targeted ~$120 million of eventual synergies.[5][7][8][9][10][11] Two of those price tags are reported differently by the children, and the difference is definitional rather than factual: Kimball at ~$485 million as announced versus ~$504 million of total accounting consideration, and Steelcase at ~$1.9 billion in cash and stock per HNI's SEC filing versus ~$2.2 billion of enterprise value in press accounts. Use the filing figure for equity consideration and the press figure only as an enterprise-value shorthand. The result at the top is essentially two U.S. public giants, one large private maker (Haworth), and a long tail of specialists — with effective concentration materially higher than the 2022 census snapshot shows.
• The millwork child (337212): slow, bottom-up private-equity roll-up. With a top-four share of 6.7% and an HHI of 24.4 — more fragmented than in 2002, when CR4 was 10.1% — this is a pure fragmentation play. Private equity acquires profitable regional shops, pools purchasing and back-office, adds CNC automation, and cross-sells across geographies: Beaubois (Norea Capital-backed, buying into the U.S. with Four Daughters Millwork in January 2026), Adams Group (Mill-Rite Woodworking), and a steady stream of shop sales to investors such as Fairfield County Millwork to Ashford Sato in 2026.[3][28][29] No national brand, no wave — one shop at a time.
• The fixtures child (337215): strategic industrial buyers, not just financial ones. This is the update the children now support. Consolidation here has been led as much by large operating companies as by private equity — Nucor rolling up Hannibal Industries ($370 million, 2021) and Elite Storage Solutions ($75 million, 2022) to build Warehouse Systems on its own steel; Ali Group buying Metro/InterMetro (2015); UFP Industries absorbing idX; LSI adding EMI Industries; Lozier buying Leggett & Platt's store-fixtures operations (2014); and, on the consumer edge, ONCAP and Griffon forming Veritage Brands around ClosetMaid in June 2026.[4][24][25][26][39] Expect ongoing, deal-by-deal gathering rather than a wave.

The barrier to consolidation is the same everywhere: skilled craft and fabrication labor. You can buy shops and brands, but you cannot easily buy master woodworkers, and project-by-project work resists standardization. Automation favors acquirers with enough throughput and engineering discipline to keep expensive equipment utilized — which in turn requires preserving the local estimating and customer relationships that made the target profitable in the first place.[3]

---

9. Risks

• Cyclicality — three cycles, one operating-leverage-heavy cost base. Downturns in corporate spending, non-residential construction, or retail capex each hit part of the level hard, and the fixed-cost factory base turns thin margins negative fast.
• The secular office question (337211/337214). Durable hybrid/remote work and shrinking corporate footprints are a structural headwind to seat counts, only partly offset by redesign demand. The right question is not how many employees came back but how much occupied space is being consolidated or reconfigured, and which categories get the spending.[6]
• The retail e-commerce shift (337215). A long-term migration online caps growth for traditional store fixtures — partly offset by warehouse racking and parcel-locker demand. Substitution from used racking, plastic and composite systems, and automated storage that replaces static shelving adds to the pressure.[4]
• Input-cost and tariff whipsaw. Lumber and panels for the wood children; steel and aluminum for the metal children. Tariffs cut both ways — raising import prices (a help) while raising the cost of imported panels, hardware, and metal (a hurt) — and the fixtures child's price protection depends on trade remedies that can be narrowed or circumvented by shifting country of origin.[4][19][35][36]
• Import competition at the value end from Vietnamese and Chinese factories, capped mainly by tariffs, freight economics on heavy goods, and TAA/Buy-American rules on the government side.[5][36]
• Skilled-labor scarcity, most acute in custom millwork, where an aging craft workforce and thin training pipelines lengthen lead times and lift wages — with BLS projecting the woodworking occupation to shrink 2% through 2034 even as ~21,400 openings a year need filling.[3]
• Fixed-price estimating and working-capital risk (337212). Most work is fixed-price, so a blown labor estimate or uncompensated scope creep comes straight out of profit; slow milestone billing plus 5–10% retainage strains cash; and a general contractor's insolvency or a stalled project can leave a subcontractor unpaid. A large backlog can mislead if it holds underpriced jobs or projects without reliable start dates.[3]
• Product-liability and warranty exposure (337215). Pallet rack carries the highest-consequence product risk in the level — seismic calculations, load plaques, anchoring, installation quality, forklift damage, and unauthorized field modifications all create liability.[4]
• Integration risk. HNI must digest both Kimball and Steelcase, including overlapping dealer networks and brands; merger synergies can disappoint and distract.[5][7]
• Dealer failure. A financially weak dealer may stop paying for delivered product, take local relationships with it, and leave a manufacturer without installation and service coverage that takes time and money to replace.[6]
• Customer concentration. Heavy reliance on corporate, government, and — for fixtures — a handful of national retail buyers ties fortunes to a few budget cycles; losing a national program can idle a plant overnight.[4]
• Thinning public data. With the 337211 PPI discontinued and BLS's fixtures employment series folded into a broader grouping, the level is measurably harder to track than it was two years ago.[4][6]

---

10. How to invest and the outlook

Match the vehicle to the child — they are not interchangeable.

• Public-market investors effectively get only the office-furniture third of this level cleanly. The cleanest listed exposures are HNI (NYSE: HNI) — the scaled, dividend-paying leader — and MillerKnoll (NASDAQ: MLKN) — the premium-design pairing of Herman Miller and Knoll. Both are diversified (HNI also makes fireplaces, about a quarter of sales; MillerKnoll has a large retail and international business), so you are buying a broad workplace-furnishings cyclical, not a clean segment bet — and with Steelcase inside HNI, the public menu is narrower than it was a year ago. Virco (VIRC) offers small-cap, school-tilted exposure. For the millwork-and-fixtures two-thirds, the listed threads are UFP Industries (UFPI) via idX, LSI Industries (LYTS) via Display Solutions, Nucor (NUE) via Warehouse Systems, Stanley Black & Decker (SWK) via Vidmar/LISTA, Berkshire Hathaway (BRK) via Marmon Retail Solutions, and Global Industrial (GIC) as a distributor — with foreign-listed Quadient (QDT) and InPost (INPST) for parcel lockers. In every one of those cases the relevant revenue is not separately disclosed. There is no ETF and no public pure-play in the two largest children. Follow orders, backlog, price-cost, dealer health, plant utilization, and segment margins rather than consolidated revenue.
• Private-market investors get the millwork-and-fixtures two-thirds — the larger and more direct opportunity set. The theses: buy a profitable regional shop (real cash flow, but key-person and cyclicality risk), or back a roll-up platform consolidating fragmented millwork or fixture shops (Beaubois, Adams Group, and others), capturing purchasing scale, automation, and cross-selling — noting that in fixtures the competing bidder is often a strategic industrial buyer, not another sponsor. What to underwrite: backlog and book-to-bill (and whether backlog pricing still holds), shop/plant utilization, AWI-QCP or BIFMA certification, craft-labor retention, estimating discipline and change-order realization, customer/GC diversification, steel and panel exposure, warranty and engineering-responsibility history, and (for millwork) retainage and WIP accounting. Availability is a real constraint — many of the best fixtures targets are family- or employee-owned. The dealer, installation, and workplace-services layer between factories and buyers is almost entirely private, is itself an investable surface, and can earn on renovation work even when new-furniture unit demand is weak.

Near-term drivers to watch (forward-looking):

• Office side: the stickiness of return-to-office mandates against ~72.6% of pre-pandemic attendance, a recovery in corporate capital spending, whether HNI realizes the ~$120 million of targeted Steelcase/Kimball synergies without integration missteps, and "flight-to-quality" refurbishment — against an office-vacancy backdrop the sources put anywhere between ~12% and ~21%.[5][6][16][17]
• Construction/millwork side: whether the ABI can hold above the low-47s after its March 2026 brush with breakeven — a sustained move through 50 would feed woodwork demand about a year later — with non-residential spending projected to grow only ~1.7% in 2025 and ~2.0% in 2026 and institutional (healthcare, education, government) the firmest segment.[3][14][15]
• Retail/fixtures side: retail remodel budgets (healthy near-term — Target ~$1 billion a year, Dollar General 4,700 projects planned for 2026) plus the faster-growing warehouse-racking and smart-parcel-locker pockets; independent studies put fixture and metal-shelving sub-segments on ~4.5–7.5% annual growth paths into the early 2030s, with the locker niche going from ~$990 million (2024) to ~$1.5 billion (2030).[4][21][23]
• Across all four: lumber/panel and steel/aluminum prices and tariff policy — the single line most likely to make or break a given year's margins — plus, in fixtures specifically, whether the antidumping orders on shelving and racking hold.[4][19][35]

Bottom line. NAICS 33721 is not one industry but four, joined by shared materials and split across three demand cycles and two ownership worlds. The concentrated office-furniture children are a small, consolidated, publicly investable oligopoly — and shrinking, at roughly half their 2002 employment. The fragmented millwork and fixtures children are the larger, private, cash-generative, still-consolidating majority, and they have been growing while the furniture half shrank. Read the level's headline HHI with suspicion, size each child on its own cycle, treat every "listed exposure" to millwork or fixtures as an undisclosed sliver of a bigger company, and pick the vehicle — a stock or a shop — that actually matches the segment you want.

---

Sources

1. U.S. Census Bureau — 2022 Economic Census, Concentration/Comparative Statistics, NAICS 33721 and children 337211/337212/337214/337215 (receipts, firms, CR4/CR8/CR20/CR50, HHI) (2022). https://www.census.gov/programs-surveys/economic-census.html
2. U.S. Census Bureau — County Business Patterns 2023, NAICS 33721 and children (establishments, employment, annual and Q1 payroll) (2023). https://www.census.gov/programs-surveys/cbp.html
3. Child primer 337212 — Custom Architectural Woodwork and Millwork Manufacturing (firm/establishment counts, HHI 24.4, 2002 benchmark, fragmentation, private ownership, roll-ups, retainage, ABI, AWI-QCP, PPI, labor scarcity) (2026), and its underlying sources.
4. Child primer 337215 — Showcase, Partition, Shelving, and Locker Manufacturing (HHI 150.5, AIES vs Census revenue, BLS series change, private makers, public proxies, freight economics, trade remedies, parcel lockers, retail capex) (2026), and its underlying sources.
5. Child primer 337214 — Office Furniture (except Wood) Manufacturing (scope, adjacent-code exclusions, HHI 1,075, dealer channel, BIFMA series, deal values, synergy target, office attendance) (2026), and its underlying sources.
6. Child primer 337211 — Wood Office Furniture Manufacturing (HHI 944.9, 2002 benchmark, PPI discontinuation, dealer share, Jasper cluster, vacancy dispute) (2026), and its underlying sources.
7. Steelcase Inc. — "HNI Corporation Completes Acquisition of Steelcase Inc." (closed Dec 10, 2025). https://www.steelcase.com/press-releases/hni-corporation-completes-acquisition-of-steelcase-inc/
8. HNI Corporation — Form 10-K, fiscal 2025 (SEC filing). https://www.sec.gov/Archives/edgar/data/48287/000004828726000084/hni-20260103.htm
9. Woodworking Network — "HNI completes $2.2B acquisition of Steelcase, creating $5.8 billion furniture giant" (2025). https://www.woodworkingnetwork.com/management/fdmc-300/hni-completes-22b-acquisition-steelcase-creating-58-billion-furniture-giant
10. MillerKnoll, Inc. — Form 10-K, fiscal 2025 (SEC filing). https://www.sec.gov/Archives/edgar/data/66382/000006638225000069/mlkn-20250531.htm
11. Woodworking Network — "HNI to acquire Kimball International in $485 million transaction" (2023). https://www.woodworkingnetwork.com/news/woodworking-industry-news/hni-acquire-kimball-international-485-million-transaction
12. BIFMA (Business and Institutional Furniture Manufacturers Association) — "Industry Market Size" ($15.27B 2023; $15.39B 2024 prelim; $16.07B 2025 prelim; series broadened in 2022) (2025). https://www.bifma.org/page/Industrymarketsize
13. IBISWorld — "Office Furniture Manufacturing in the US — Industry Analysis" (~$30B) (2026). https://www.ibisworld.com/united-states/industry/office-furniture-manufacturing/870/
14. American Institute of Architects (AIA) — Consensus Construction Forecast (Jan 2026) and Architecture Billings Index (ABI 45.3 Nov 2025; 49.8 Mar 2026; nonresidential ~1.7% 2025 / ~2.0% 2026; institutional firmest) (2025–2026). https://www.aia.org/resource-center/consensus-construction-forecast/january-2026
15. American Institute of Architects (AIA) — Architecture Billings Index, June 2026 (ABI 47.3; 41 months without majority billings growth) (2026). https://www.aia.org/resource-center/abi-june-2026-billings-remain-weak-architecture-firms
16. CBRE — "U.S. Real Estate Market Outlook 2026 — Office" (office vacancy ~20.7% Q3 2025; lowest new supply since 2008–09) (2025). https://www.cbre.com/insights/books/us-real-estate-market-outlook-2026/office
17. NAIOP — "Office Space Demand Forecast, Q2 2025" (11.8% vacancy) (2025). https://www.naiop.org/research-and-publications/research-reports/reports/office-space-demand-forecast-2q25
18. U.S. Census Bureau — 2024 Annual Value of Construction Put in Place (nonresidential $1.226T, +7.0%; end-market breakdowns) and Current Construction Spending (private nonresidential $738.7B SAAR, May 2026). https://www.census.gov/construction/c30/pdf/pr202501.pdf
19. Wipfli — "Manufacturers tariffs update: what's ahead for the industry" (Section 232 steel/aluminum at 50% in 2025) (2025). https://www.wipfli.com/insights/articles/manufacturers-tariffs-update-whats-ahead-for-the-industry
20. Federal Reserve Bank of St. Louis (FRED) / BLS — Producer Price Index: Nonwood Storage Racks (PCU337215337215E) (2026). https://fred.stlouisfed.org/data/PCU337215337215E
21. LightBox — "Retail 2024 Wrap and 7 Trends in Store for 2025" (Target ~$1B/yr store capex; Dollar General ~4,700 real-estate projects) (2025). https://www.lightboxre.com/insight/retail-2024-wrap-and-7-trends-in-store-for-2025/
22. U.S. Census Bureau — "Quarterly Retail E-Commerce Sales, 1st Quarter 2026" (16.9% of retail sales; 9.8% YoY growth vs 3.9% total retail) (2026). https://www.census.gov/retail/ecommerce.html
23. GlobeNewswire — "Smart Parcel Locker Market Research Report 2025–2030" (~$990M in 2024 to ~$1.5B by 2030; Quadient/Parcel Pending 20,000+ locker sites) (2025). https://www.globenewswire.com/news-release/2025/11/24/3193274/28124/en/Smart-Parcel-Locker-Market-Research-Report-2025-2030.html
24. UFP Industries Inc. — Form 10-K, fiscal 2025 (idX Holdings: custom interior fixtures, casework, architectural millwork) (SEC filing). https://www.sec.gov/Archives/edgar/data/912767/000110465926019567/ufpi-20251227x10k.htm
25. LSI Industries Inc. — Fiscal 2025 results (net sales $573.4M; Display Solutions ~$325M; EMI Industries and Canada's Best acquisitions) (2025). https://www.businesswire.com/news/home/20241107886924/en/LSI-Industries-Reports-Fiscal-2025-First-Quarter-Results-and-Declares-Quarterly-Cash-Dividend
26. Nucor Corporation — "Nucor Completes Acquisition of Hannibal Industries, Inc." ($370M, 2021). https://nucor.com/news/?article=nucor-completes-acquisition-of-hannibal-industries-inc.-122649
27. Woodworking Network — "FDMC 300: Architectural woodwork and retail fixture firms" (~$3.5B combined 2025 sales across the ~50 largest tracked firms; Lozier, idX, JBI, Beaubois rankings) (2025). https://www.woodworkingnetwork.com/management/fdmc-300/fdmc-300-architectural-woodwork-and-retail-fixture-firms-share-outlooks
28. Wood & Panel — "Beaubois Millwork completes the acquisition of Four Daughters Millwork" (Norea Capital 40%+ stake; U.S. expansion) (2026). https://www.woodandpanel.us/news/article/beaubois-millwork-completes-the-acquisition-of-four-daughters-millwork-to-broaden-its-u-s-presence/
29. Woodworking Network — "Adams Group acquires Mill-Rite Woodworking" (2023–2025). https://www.woodworkingnetwork.com/management/fdmc-300/adams-group-acquires-mill-rite-woodworking
30. Haworth — "Haworth Group Global Sales Top $2.7 Billion" (2025). https://www.haworth.com/na/en/about/about-haworth/media-room/haworth-group-global-sales-top-2-7-billion.html
31. U.S. Environmental Protection Agency — "Formaldehyde Emission Standards for Composite Wood Products" (TSCA Title VI; CARB harmonization) (2025). https://www.epa.gov/formaldehyde/formaldehyde-emission-standards-composite-wood-products
32. BIFMA — "Standards Overview" and e3/LEVEL sustainability standard; Architectural Woodwork Institute — "A Guide to the AWI Standards" and Quality Certification Program (2024–2025). https://www.bifma.org/page/standardsoverview; https://awiqcp.org/a-guide-to-the-awi-standards-of-architectural-woodwork/
33. U.S. General Services Administration — "Furniture and Furnishings" (GSA schedules; Trade Agreements Act; AbilityOne) (2025). https://www.gsa.gov/buy-through-us/products-and-services/office-management/furniture-and-furnishings
34. USDA APHIS — "File a Lacey Act Declaration" (Phase VII from December 2024; electronic filing from January 2026) (2025). https://www.aphis.usda.gov/plant-imports/file-lacey-act-declaration
35. U.S. International Trade Commission — "USITC Maintains Antidumping Duty Order on Boltless Steel Shelving from China" (sunset review) (2021). https://www.usitc.gov/press_room/news_release/2021/er0420ll1757.htm
36. IndexBox — "New U.S. Tariffs on Furniture Imports Target China, Vietnam" (imports; tariffs) (2024). https://www.indexbox.io/blog/new-us-tariffs-on-furniture-imports-target-china-vietnam/
37. Jasper Chamber of Commerce — "Work Here" (Jasper, Indiana, "Wood Office Furniture Capital of the World") (2025). https://jasperin.org/work-here/
38. U.S. Small Business Administration — "Table of Size Standards" (337211: 1,000 employees; 337214: 1,100; 337215: 500) (2023). https://www.sba.gov/document/support-table-size-standards
39. Griffon Corporation — "ONCAP and Griffon Corporation Announce Launch of Veritage Brands" (57% ONCAP / 43% Griffon ownership of ClosetMaid) (June 2026); PR Newswire / Leggett & Platt — "Leggett & Platt Announces Divestiture of Store Fixtures Operations" (to Lozier) (2014). https://ir.griffon.com/news-releases/news-release-details/oncap-and-griffon-corporation-announce-launch-veritage-brands; https://www.prnewswire.com/news-releases/leggett--platt-announces-divestiture-of-store-fixtures-operations-281294141.html