Construction of Buildings (United States) — NAICS 236

A Histometrics rollup primer for a general investing audience — relevant to both public-market and private investors. This is a NAICS subsector (the 3-digit level): the box that holds all the general contractors and builders who put up whole buildings, split into just two children — homes (2361) and everything that isn't a home (2362). The value of this page is the contrast between those two halves — nearly equal in revenue, but worlds apart in who builds them, who owns the builders, and how an investor touches each one. Figures are reported federal facts with citations; statements about the future are labeled as judgments.

---

1. Overview

The North American Industry Classification System (NAICS — the U.S. government's standard scheme for grouping businesses by what they do) uses code 236, Construction of Buildings, for the firms that assemble complete buildings [1]. It is one of three subsectors inside the broader Construction sector (code 23); the other two cover heavy and civil engineering (roads, bridges, utilities — code 237) and specialty trades (the plumbers, electricians, roofers, and framers who work under a general contractor — code 238). Subsector 236 is the "vertical building" box: general contractors (GCs — a GC takes responsibility for delivering a finished structure), merchant builders, and design-build firms who organize land, labor, materials, and a chain of subcontractors into a finished building. [1]

The subsector divides into exactly two children, and the split is the whole story:

• Residential Building Construction (2361) — the contractors and builders who put up, sell, and renovate American homes: custom houses, apartment buildings, for-sale ("spec") housing, and remodeling. [3]
• Nonresidential Building Construction (2362) — the GCs who build everything else: factories, offices, hospitals, schools, warehouses, hotels, stores, and the data centers now driving so much of the activity. [4]

These are roughly the same size by revenue — about half the subsector each — but they answer to different customers (homebuyers on one side, corporations and governments on the other), attract different owners (listed homebuilders and a sea of small contractors on one side, almost entirely private giants on the other), and are reached through completely different investment vehicles. Understanding 236 means understanding that fork.

---

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

The hierarchy narrows step by step: sector 23 (Construction) → subsector 236 (this page) → two 4-digit industry groups → the six-digit businesses beneath them. The meaningful division is here, at the top: residential versus nonresidential.

The contrast is the point. The two children are close to even in revenue, so neither dominates the rollup — but they are lopsided on almost every other axis. Residential is a cottage industry of enormous headcount and firm count; nonresidential is a contractor industry of fewer, larger, higher-paid firms.

Two things the size row hides.

• Equal revenue, very unequal firm counts. Residential and nonresidential book roughly the same receipts, yet residential does it with five times as many firms (203,762 vs. 42,367). The typical residential builder is tiny — often a single crew or a one-person remodeler — while the typical nonresidential firm is larger, pays better, and produces more revenue per worker. That is why residential holds 83% of the firms but only 49% of the money.
• The single most important fact for a public-equity investor. These two children are not equally reachable from a brokerage account. Residential contains the only large, listed, pure-play building companies in the entire subsector — the for-sale homebuilders. Nonresidential, despite equal size, has no publicly traded pure-play general contractor at all; its biggest builders (Turner, Bechtel, Whiting-Turner, DPR, Kiewit) are private. So even though the two halves are the same size, public-market depth is heavily skewed to the residential side, and nonresidential is an "indirect-only" story. [3][4]

---

3. Size (this level's rollup figures)

These are our ground-truth federal statistics for NAICS 236 [2]. Note the periods differ: receipts, firm count, and concentration come from the 2022 Economic Census (EC); establishments, employment, and payroll come from 2023 County Business Patterns (CBP) — so they are not a single-year snapshot, and they are the sum of the two children (the residential and nonresidential figures add cleanly to these totals).

At roughly $1.2 trillion of contractor revenue and 1.6 million paid workers, subsector 236 is one of the largest goods-producing corners of the U.S. economy. Average pay of about $82,000 (annual payroll ÷ employees) sits above the private-sector average, reflecting skilled trades, engineers, and project managers — though, as the child breakdown shows, it is really a blend of a lower-paid residential side (~$71k) and a higher-paid nonresidential side (~$97k) [3][4].

Concentration — more fragmented than either child. The subsector HHI is just 18.2 — against a 10,000 ceiling that would mean a single-firm monopoly, and a 1,500 line below which economists call a market "unconcentrated." The top four firms hold only 6.1% of revenue. A technical note worth understanding: the rollup HHI (18.2) is lower than either child's (54.3 residential, 21.3 nonresidential), because combining two already-fragmented industries whose leaders are different companies dilutes measured concentration further — the biggest homebuilder and the biggest office GC don't compete, so together they look even smaller. In plain terms: hundreds of thousands of firms compete, and no one is remotely dominant. [2]

Undercount and read-with-care caveats.

• Small operators are undercounted — heavily on the residential side. CBP counts only employer establishments with paid employees; it excludes the self-employed and nonemployer businesses. Residential construction is thick with one-person contractors and remodelers who carry no payroll — counted separately by the Census Bureau in the millions industry-wide — so the true population of home-building firms is far larger than these figures show. The effect is much milder on the nonresidential side, where work is larger-scale and firms are bigger. Read the ~246,000 firm count as "employer firms," not everyone building a building. [3][4]
• Receipts are the builders' slice, not the value of the buildings. Under the general-contractor model, most project dollars pass straight through to subcontractors, specialty trades, and suppliers who are classified in other NAICS codes (heavy-civil 237, specialty trades 238, engineering services 541330). Payroll is only about 11% of receipts, so the builders' own value-add is a thin coordinating layer on a much larger flow of money. Treat the $1.223 trillion as revenue booked by building general contractors, not the total value of all U.S. buildings put in place — the installed value is several times larger. [4]
• No official margin or backlog. Federal data publish no NAICS-236 operating margin, order backlog, or win rate. Do not infer any from a single company's results.

---

4. Investable universe — where value concentrates across the children

The subsector's investable map has one dominant feature: public depth lives almost entirely on the residential side, and even there it is concentrated in one business model.

• Residential (2361) — the only direct public route in the whole subsector. A dozen-plus listed homebuilders trade on U.S. exchanges, but nearly all of them run the for-sale / merchant model (own the land, carry the finished house as inventory, sell it with the lot). Custom home building, apartment development, and remodeling — the other three-quarters of residential activity by breadth — have no pure-play listed vehicle and are reached indirectly: apartment REITs (Real Estate Investment Trusts — companies that own income-producing property and pass most of their income to shareholders) for rental multifamily, home-improvement retailers and building-products makers for remodeling, and homebuilder ETFs (exchange-traded funds — baskets of stocks that trade like a single share) for broad exposure. [3]
• Nonresidential (2362) — no pure-play at all. The biggest builders — Turner, Bechtel, Whiting-Turner, DPR, Kiewit, Mortenson, Clark, McCarthy — are private, family-owned, employee-owned via ESOPs (employee stock ownership plans), or foreign-parented. Public investors reach this half only through: specialty mechanical/electrical contractors (the cleanest listed way in, with the most direct data-center and fab leverage and steadier margins than GCs), sitework/civil names, and diversified EPC (engineering-procurement-construction) firms. [4]

The through-line for value. In both children, defensible margin sits away from plain general contracting. On the nonresidential side that means the specialty trades and EPC firms; on the residential side it means the capital-intensive for-sale builders (who earn a gain on owned inventory) rather than the fee-for-service custom builders and remodelers. Where the work is bespoke, local, and relationship-driven — custom homes, remodeling, small GC work — margins are thin and no one scales; where land, capital, marketing, or specialized engineering dominate, scale and public listings appear. [3][4]

---

5. How the money works

Subsector 236 runs on two fundamentally different profit models, split almost exactly along the child line — which is why no single margin benchmark describes it.

• The pass-through / fee model (dominant in nonresidential, and in custom homes and remodeling). The builder collects a large contract value, pays most of it out to subcontractors and suppliers, and keeps a thin margin for coordinating the work and absorbing risk. Revenue is recognized over time (percentage-of-completion), so accurate cost estimating drives reported earnings and a single mis-bid surfaces as a margin write-down. Backlog — signed, not-yet-built work — is the most-watched forward gauge on the nonresidential side, though it can be delayed, resized, or canceled. Non-official benchmarks put GC net margins near 5–6%. [4]
• The inventory / gain-on-sale model (the listed homebuilders). For-sale builders own the land, carry the home as inventory, and earn a gain on selling finished land-and-house. This is capital-intensive and highly cyclical, valued on price-to-book and normalized through-cycle earnings rather than a single-year multiple. Apartment developers run a third variant — a build-to-a-yield spread, the gap between a project's stabilized yield-on-cost and the market capitalization ("cap") rate at which it can be sold. [3]

What unites all of it: thin, cyclical margins; volume as the primary margin lever; and working-capital timing — builders lay out cash for land, labor, and materials well ahead of payment, so retainage and slow payment can sink a paper-profitable firm on liquidity even in a good year. [3][4]

---

6. Demand drivers

Both children are deeply cyclical and interest-rate sensitive, but they are wired to different customers and transmit rate moves through different channels — and right now they are on different clocks.

• Residential (2361) — consumer-driven, mortgage-rate-gated. The dominant lever is mortgage and interest rates, which set buyer affordability, apartment-developer financing, and big-remodel demand. Reinforcing structural drivers: the mortgage "lock-in" effect (owners with sub-6% loans won't sell, freezing resale inventory and pushing buyers toward new homes), a structural housing shortage estimated at roughly 3.7–4.9 million units, an aging housing stock (median U.S. home now over 40 years old — a remodeling tailwind), and demographics and migration. [3]
• Nonresidential (2362) — derived demand from capital budgets. Demand follows corporate, institutional, and government decisions to add capacity or space. The swing factors now are the AI (artificial-intelligence) data-center boom (data-center construction spending jumped ~32% in 2025 and is masking weakness elsewhere), the reshoring / semiconductor build-out behind the 2022 CHIPS Act (Creating Helpful Incentives to Produce Semiconductors — ~$39B of fab incentives) and clean-energy credits, and interest rates and financing for debt-heavy projects. Institutional budgets (hospitals, schools, universities) follow demographics and appropriations; corporate offices, retail, and hotels remain structurally soft post-pandemic. [4]

Judgment. The common master variable is the cost of money, but the two halves are out of phase. Residential is rate-gated and depressed, waiting on a mortgage-rate decline that would re-accelerate all its segments at once. Nonresidential is two-speed — the reshoring/factory surge appears to have peaked (though record backlogs keep the biggest industrial builders busy well into the back half of the decade), while the AI data-center surge is still climbing. A decisive fall in rates would lift both children together. [3][4]

---

7. Regulation

There is no single federal regulator for building construction; the binding rules apply at the point of work, and the same stack reaches both children — with different segment-specific overlays.

• State and local rules are the biggest swing factor: zoning, land use, density and permitting, building and energy codes, and contractor licensing. The U.S. has no single national building code. On the residential side, local zoning and entitlement is the single biggest determinant of whether housing gets built at all. [3][4]
• Federal overlays reach the job site: worker safety under the Occupational Safety and Health Administration (OSHA, 29 CFR Part 1926); environmental permitting under the Environmental Protection Agency (EPA) and the National Environmental Policy Act (NEPA — which can add months or years to a large factory or plant); and prevailing-wage rules (Davis-Bacon) plus domestic-materials requirements (Buy America) attached to public and subsidized work. [3][4]
• Segment-specific rules: the EPA lead-paint Renovation, Repair and Painting (RRP) rule and warranty/lien law on the residential-remodeling side; the Low-Income Housing Tax Credit (LIHTC) for affordable multifamily; the Fair Housing Act across housing; and surety bonding on nonresidential work, which effectively caps how much a contractor can take on and is a real barrier to scaling. [3][4]

---

8. Consolidation

The subsector's very low average concentration (HHI 18.2) hides real divergence in how — and whether — each child consolidates.

• Residential (2361): the for-sale builders are consolidating fastest — the ten largest captured a record ~44.7% of new single-family closings in 2024 — because land, capital, and marketing reward scale. But custom home building stays stubbornly fragmented, and remodeling is so fragmented that roll-ups barely dent it (the bigger money there is one layer up, in building-products distribution). [3]
• Nonresidential (2362): general contracting has resisted roll-ups (low barriers, people-and-process businesses, local reputation), so most merger activity sits in the specialty trades (mechanical, electrical, HVAC, roofing), where recurring service revenue and higher margins support consolidation. ESOPs are the dominant succession tool (construction ESOP firms grew from roughly 700 to over 1,100 in a decade), and foreign ownership sits at the very top (e.g., Turner via Germany's HOCHTIEF/ACS). Private-equity roll-ups are accelerating, often as the fastest way to acquire experienced crews amid a labor shortage. [4]

The pattern across both: scale compounds where land, capital, marketing, or specialized engineering dominate; it stays elusive where the work is bespoke and locally relationship-driven. [3][4]

---

9. Risks

The two children share a risk set but tilt differently within it.

• Cyclicality and rate risk — building construction is among the most cyclical activities in the economy; it turns with financing conditions. Residential turns on mortgage rates and buyer affordability; nonresidential on corporate/institutional capital budgets and project financing. [3][4]
• Theme concentration (nonresidential) — recent growth leans on two narrow themes, reshoring/fabs and data centers/AI; if either cools, much of the growth reverses. [4]
• Thin, volatile margins and fixed-price execution risk — one mispriced job or a cost overrun on a lump-sum mega-project can erase profit, and cost inflation and tariffs erode already-thin margins across both halves. [3][4]
• Financing and inventory risk — land exposure for for-sale builders and the yield-to-cap-rate spread for apartment developers on the residential side; working-capital and payment risk (retainage, slow pay) across both. [3][4]
• Skilled-labor shortage — a structural, binding constraint on capacity across the whole subsector; the industry must attract on the order of 349,000 additional workers in 2026. [4]
• Policy reversal — subsidies, tariffs, immigration, and prevailing-wage rules are politically contingent and bear directly on the nonresidential (especially industrial) side. [4]
• Roll-up and leverage risk in the private-equity-backed tiers of both children. [3][4]
• Measurement risk — federal employer statistics omit a large nonemployer tail (heaviest in residential), so top-down sizing is approximate. [2][3]

---

10. How to invest and outlook

How to invest. The subsector splits investors cleanly by market:

• Public-market investors have genuine direct depth in only one place — listed homebuilders (the for-sale model), which are the sole large pure-play building companies in the whole subsector; value them on price-to-book and normalized through-cycle earnings, not a single-year price-to-earnings multiple near a peak. Everything else is indirect: apartment REITs, home-improvement retail and building-products names, and homebuilder ETFs for the residential themes; and, for the entire nonresidential half, the specialty and diversified contractors — mechanical/electrical names such as EMCOR (ticker EME), Comfort Systems USA (FIX), and IES Holdings (IESC); sitework/civil such as Sterling Infrastructure (STRL) and Granite Construction (GVA); and diversified EPC such as Fluor (FLR), Jacobs (J), and AECOM (ACM). Weigh backlog quality and margin trajectory over headline revenue for the contractors. [3][4]
• Private-market investors find the operating industry directly in everything the public market can't reach: custom home building, apartment development, and remodeling on the residential side, and the private building giants and biggest fabs and data centers on the nonresidential side — through private equity, private credit / project finance, ESOP participation, and real-assets and digital-infrastructure strategies. Returns come from operating skill, backlog discipline, and cost control, not index beta. [3][4]

Outlook (mid-2026; forward-looking judgment). The subsector is a fragmented, thin-margin, execution-driven $1.2-trillion business, and its two halves are on opposite clocks. Residential is rate-gated and margin-pressured, held up by a multi-million-unit housing shortfall, frozen resale inventory, and aging stock — waiting on a mortgage-rate decline that would re-accelerate it broadly. Nonresidential is genuinely two-speed: data centers, AI infrastructure, healthcare, and renovation are strong; offices, retail, hotels, and winding-down factory mega-projects are weak — the American Institute of Architects' (AIA) July 2026 consensus expects nonresidential building spending to fall 0.3% in 2026 and rise 3.0% in 2027, carried almost entirely by data centers. The single macro variable that would move both children at once is the cost of money: lower rates would lift homebuyers and project financing together. The durable lesson holds on both sides — money is made through backlog and inventory discipline, risk control, and cash conversion, and, on the public side, through the listed homebuilders and the specialty/diversified contractors rather than the private general-contracting giants. [3][4]

For the full segment detail — company rosters, contract economics, and the six-digit breakdowns — see the child primers: [2361] Residential Building Construction and [2362] Nonresidential Building Construction.

---

Sources

This is a two-child rollup. Subsector-level figures come from our ground-truth federal statistics for NAICS 236; all synthesis, contrast, and forward judgments are drawn from the two child primers (2361 and 2362), which themselves aggregate the six-digit primers and their federal and industry sources. The residential and nonresidential figures sum cleanly to the subsector totals.

1. U.S. Census Bureau. 2022 NAICS Definition — 236 Construction of Buildings (subsector scope; children 2361 Residential and 2362 Nonresidential Building Construction; sibling subsectors 237 Heavy and Civil Engineering and 238 Specialty Trade Contractors). 2022. https://www.census.gov/naics/?input=236&year=2022
2. Histometrics ground-truth federal statistics — NAICS 236 (receipts $1,223,192,386 thousand and firms 246,042, Economic Census 2022; establishments 254,630, employment 1,621,439, annual payroll $133,659,908 thousand, Q1 payroll $32,391,996 thousand, County Business Patterns 2023; CR4 6.1%, CR8 8.9%, CR20 14.7%, CR50 22.3%, HHI 18.2, Economic Census concentration 2022). The two children sum to these totals.
3. Histometrics rollup primer — NAICS 2361, Residential Building Construction (child; receipts $601.4B, firms 203,762, establishments 212,178, employment 905,357, annual payroll $64.4B, Q1 payroll $15.6B, HHI 54.3, CR4 11.4%; full synthesis of the four six-digit residential businesses — custom single-family, rental multifamily, for-sale builders, and remodelers — plus demand drivers, regulation, consolidation, risks, the listed-homebuilder roster, and the mid-2026 outlook). 2026.
4. Histometrics rollup primer — NAICS 2362, Nonresidential Building Construction (child; receipts ~$621.8B, firms 42,367, establishments 42,452, employment 716,082, annual payroll ~$69.3B, Q1 payroll ~$16.7B, HHI 21.3, CR4 6.4%; full synthesis of the two six-digit businesses — industrial (fabs, factories, plants) and commercial/institutional (offices, hospitals, schools, warehouses, data centers) — plus contract economics, the specialty/EPC investable roster, demand drivers, regulation, consolidation, risks, and the AIA-anchored outlook). 2026.

For the underlying company-level detail, contract economics, and full source lists, see the child primers (NAICS 2361 and 2362) and the six-digit primers beneath them.