NAICS 3311 — Iron and Steel Mills and Ferroalloy Manufacturing

A rollup primer. NAICS = North American Industry Classification System, the U.S. government's standard code for industries. This is a four-digit "industry group" that contains exactly one five-digit industry, 33111. For a general investing audience — public-market and private investors alike.

1. Overview

NAICS 3311 covers the businesses that make steel from scratch: reducing iron ore, making pig iron, converting iron into steel, and casting and rolling it into the semi-finished and finished shapes the rest of the economy builds with — plus ferroalloy production, the manganese-, silicon-, chromium- and molybdenum-bearing additives that give steel its chemistry [1]. The classification is process-based: a plant that only rolls, draws, or forms steel it bought from someone else generally sits in NAICS 3312, not here [1]. This is heavy, capital-intensive commodity manufacturing — a small number of companies running a small number of very large plants — and it sits at the top of the domestic steel supply chain. It is a classic cyclical, rising and falling with construction, autos, machinery, and energy.

2. What's inside — and why this level equals its one child

A four-digit NAICS industry group normally splits into several five-digit industries. This one does not: 3311 contains a single child, 33111 (also named "Iron and Steel Mills and Ferroalloy Manufacturing"), which in turn contains a single national industry, 331110. All three codes describe the same set of companies, plants, and products. So everything true of 33111 is true of 3311.

The one structural fact worth stating at this level is how lopsidedly the industry splits by production route. At the start of 2025, two companies operated integrated blast-furnace steelmaking at eight active locations, while 47 companies ran electric-arc-furnace (EAF) "minimills" at 102 sites [2]. The U.S. Department of Energy reports that EAFs now make 70% of domestic steel, and that remelting scrap in an EAF takes less than half the energy of making steel from ore through a blast furnace [3]. An earlier version of this page called the country merely "majority-EAF"; the revised research puts a number on it, and the number is decisive.

This page is a short pass-through. The full detail — route-by-route economics, company-level metal spreads and input ratios, trade and environmental rules, and the investable names — lives in the 33111 primer. Read that for the deep dive; this page adds no companies or economics beyond its single child.

3. Size (this level's rollup figures)

Federal ground truth for 3311 (which equals 33111). These measures come from different programs and different years — they are not alternative estimates of one number:

On the County Business Patterns basis that works out to roughly $116,000 of payroll per worker [5] — high pay reflecting skilled, heavily unionized, capital-intensive work. Physical output was about 82 million metric tons of raw steel and 21 million tons of pig iron in 2025, against roughly 105 million tons of annual capacity and 95 million tons of apparent consumption, with net imports supplying 13% of what the country used; the USGS values 2025 raw steel sales at about $149 billion, a different scope from the Census receipts figure above [2].

The sources disagree on employment. County Business Patterns reports 78,532 workers for 2023 [5]; the Census Bureau's own six-digit industry profile reports average employment of about 86,000 [6]. That roughly 10% gap is not reconciled — hold one series constant across years rather than mixing them.

Two counts of "how many companies" that should not be reconciled. The Census firm count (73, in 2022) covers firms whose establishments are classified in this industry; the USGS count (two integrated producers plus 47 EAF producers, at the start of 2025) covers companies actually operating raw-steel furnaces [2][4]. Likewise the 359 establishments are all payrolled operating locations, a wider net than the eight integrated locations and 102 minimills that melt raw steel. Do not subtract or divide across the two.

Undercount caveat (reversed here). The usual small-shop or nonemployer undercount that plagues fragmented industries does not apply — with 73 firms and 359 plants, federal data captures this industry almost completely. The previous version of this page asserted that public stock-market value understates the industry because so much of it is private or foreign-owned. That direction remains plausible, but the revised child research could not establish a public-versus-private share of output from available sources, and could not establish the share of revenue attributable specifically to ferroalloys either. Treat both as open questions rather than settled facts. One timing note: receipts and concentration are 2022 Economic Census figures, while employment, plants, and payroll are 2023 County Business Patterns.

4. Investable universe (where value concentrates)

Because 3311 has just one child and no sub-industries beneath it, value does not split across "children" — it concentrates across a short list of dominant firms. On 2022 Economic Census data the top 4 firms take 59.5% of revenue, the top 8 take 75.1%, and the top 20 take 94.4% [4]. Read those as a 2022 snapshot: a current four-firm concentration ratio could not be established from available sources.

The listed exposure is genuinely narrow, and the names differ by route more than by product — EAF producers with downstream integration on one side, the integrated ore-to-steel producer with its automotive book on the other, plus foreign-listed owners of U.S. assets. The specific public and private names, with segment shares and tonnages, are catalogued in the 33111 primer (§4). One caution belongs at this level: company tonnage figures include internal transfers, downstream processing, and sometimes Canadian output, so they cannot be divided into the USGS U.S.-only totals in §3 to manufacture market shares [2].

5. How the money works

At this level the economics are simply those of 33111: a commodity manufacturer whose margin is a spread — realized selling price minus the cost of metallic inputs — levered by how full the plant runs. Steel Dynamics reports that metallic inputs are normally 55%–65% of steel-mill manufacturing cost [8]. On the integrated side, Cleveland-Cliffs estimates a net ton of raw steel consumes roughly 1.4 net tons of iron-ore pellets or equivalent, 0.3–0.4 tons of coke, 0.3 tons of scrap, and 5–6 MMBtu of natural gas, against about 1.1 tons of metallic input per ton in its EAFs [9]. Integrated mills carry very high fixed costs and expensive shutdowns; EAFs start and stop more readily but sit closer to scrap and electricity prices [3][9]. Utilization is the second lever — Nucor's steel-mill utilization rose from 76% in 2024 to 83% in 2025, more tonnage over the same asset base [7].

There is no single "industry margin," and 2025 is the clearest evidence. In the same year Cleveland-Cliffs reported a negative 5% steelmaking gross margin while Steel Dynamics' steel operations earned $1.428 billion of operating income on $13.413 billion of sales — about a 10.6% operating margin [8][9]. Any number quoted as "the" margin for this industry is really a statement about one route and one product mix. Working capital reinforces the cycle: mills and service centers build costly inventory into rising prices, and the destock that follows cuts orders harder than end demand falls. Full mechanics are in 33111 (§5).

6. Demand drivers

Federal data now gives the breakdown directly. Of 2025 U.S. net steel shipments by market: construction 31%, service centers and distributors 26%, automotive 15%, converting and processing 11%, and all other uses 17% [2]. The 26% sitting with distributors is worth dwelling on — a quarter of shipments go to intermediaries whose restocking and destocking can temporarily dominate real end-market demand.

Underneath that, demand is nonresidential and infrastructure construction, autos, machinery and appliances, energy pipe and tube, and — newer — data centers, grid and transmission investment, and reshored semiconductor and battery plants; in 2025 the large producers described infrastructure, data centers, energy, and advanced manufacturing as strong, with automotive and residential construction softer [7][8]. Because these are capital-goods and building end-markets, the industry rises and falls with interest rates, construction starts, and vehicle production. Over a longer horizon the mix keeps moving toward EAFs and lower-carbon iron; the Department of Energy names hydrogen-based direct-reduced iron, iron electrolysis, electrified reheat furnaces, and carbon capture as priority technologies, several still needing cost reduction and commercial-scale demonstration [12]. See 33111 (§6).

7. Regulation

Three forces dominate, all covered in 33111 (§7). First, trade policy: Section 232 of the Trade Expansion Act lets the President tax imports on national-security grounds, and as of April 2026 the U.S. applied a 50% rate to articles made entirely or almost entirely of steel and 25% to qualifying derivative articles [13]. The U.S. International Trade Commission's study of the earlier Section 232 measures found they cut affected imports 24%, raised U.S. steel-product prices 2.4%, and lifted domestic production 1.9% on average over 2018–2021 — while raising downstream prices and reducing downstream production [14]. Protection transfers economics to the mills at the expense of the customers that buy from them.

Second, environmental rules, which are route-specific. The U.S. Environmental Protection Agency's amended hazardous-air-pollutant standards (NESHAP) for integrated iron and steel mills require maximum-achievable-control technology on sinter plants, blast furnaces, and basic-oxygen shops [15]; EAFs escape most of the blast-furnace burden but generate particulate emissions and hazardous furnace dust and are exposed to grid power prices. The spending is real but not ruinous at the top of the industry — Nucor expects roughly $250 million of safety and environmental spending in 2026 and puts environmental-compliance capital expenditure at existing facilities below $100 million a year in both 2026 and 2027 [7].

Third, and new to this page, labor is a regulatory-grade constraint rather than just a cost line. More than 90% of Cleveland-Cliffs' hourly workforce is union-represented, with several major site agreements up for negotiation in 2026, alongside shortages of electricians and other experienced industrial workers [9]. Work stoppages, wage escalation, and retiree obligations bear hardest on integrated mills.

8. Consolidation

The industry is already concentrated — a Herfindahl-Hirschman Index (HHI, a standard measure of market concentration) of 1,071.9, just inside the "moderately concentrated" band, on only 73 firms [4] — though as with the ratios in §4 that is a 2022 reading, and no current HHI could be established. Nippon Steel of Japan completed its acquisition of U.S. Steel on June 18, 2025 [10][11], taking a marquee American name off the U.S. public market. Nucor assembled its position through acquisitions, and integrated capacity clustered under Cleveland-Cliffs through its combinations with the legacy AK Steel and ArcelorMittal USA assets. Further consolidation may be limited by antitrust review, particularly for deals combining integrated capacity serving automotive customers. Private and foreign ownership — BlueScope's North Star mill, numerous regional minimills, and specialty and ferroalloy producers — sits alongside the listed names; how much of output that represents is not established (§3).

The larger competitive pressure is external. The OECD estimates global excess steelmaking capacity reached 640 million tonnes in 2025 and projects 745 million tonnes by 2028 [16]. Even where tariffs block direct imports, displaced foreign steel depresses world prices, reroutes through third countries, and provokes further trade actions.

9. Risks

The same risks as 33111 (§9): metal-spread compression, when input prices rise before selling prices reset or selling prices fall before input costs do; deep cyclicality tied to construction and auto capital spending, amplified by distributor destocking; utilization volatility, where high fixed costs turn a demand dip into severe margin damage and capacity additions can overshoot; global overcapacity at 640 million tonnes and rising [16]; trade-policy reversal, since Section 232 rates are changeable by executive action, exemption, or litigation; environmental compliance and decarbonization capital expenditure [15]; labor constraints — union negotiations, skilled-worker shortages, and legacy pension and retiree obligations, concentrated at integrated mills [9]; customer concentration among large automakers, builders, and distributors; inventory losses on high-cost metal bought near a peak; technology-transition risk, if hydrogen DRI or carbon capture investments miss on cost or timing [12]; and substitution by aluminum, composites, engineered wood, and concrete.

10. How to invest & outlook

For public-market investors, the first decision at this level is route, not ticker: EAF-focused exposure with downstream integration versus integrated ore-to-steel exposure with automotive concentration and higher operating leverage — the two behaved very differently in 2025 (§5). One large pure-play left the U.S. public market with the June 2025 Nippon–U.S. Steel deal [10][11]; a steel-industry exchange-traded fund gives a diversified basket. Valuation should be built on mid-cycle earnings rather than peak metal spreads. For private investors, the action is in regional and specialty mills and ferroalloy operations, where diligence turns on metallic-input formulas, customer qualifications, union agreements, environmental liabilities, power and freight arrangements, and the cost of surviving a low-utilization year; note that no authoritative industry-wide private valuation multiple or transaction median could be established. Verify that a target actually makes steel rather than processing purchased metal — that is the line between this industry and NAICS 3312 [1]. Tickers, multiples, and dividend detail are reserved for the 33111 primer (§4, §10), which is the right place to go for anything actionable — this four-digit level adds no companies or economics beyond its single child.

Outlook: demand is supported by infrastructure spending, manufacturing reshoring, data-center construction, and energy investment, and near-term pricing by the 50% Section 232 rate [13] and firm domestic output (~82 million tons of raw steel in 2025) [2]. Margins are the uncertain half: 640 million tonnes of global excess capacity heading toward 745 million by 2028 [16], volatile scrap, ore, and energy costs, and reversible trade policy can each overwhelm demand growth. The multi-year story remains the shift to lower-carbon electric-arc production. Full outlook in 33111 (§10).

---

Sources

Drawn from the revised child 33111 primer and the primary federal and company sources it cites, plus our own ingested federal ground truth for this level (stats-3311). Sources [4] and [5] were recorded without links and are listed here without one; every other entry links to the primary source.

1. U.S. Census Bureau, 2022 NAICS Manual, 2022, https://www.census.gov/naics/?details=33&input=33&year=2022 — industry scope and the process-based boundary with NAICS 3312.
2. U.S. Geological Survey, Mineral Commodity Summaries 2026: Iron and Steel, 2026, https://pubs.usgs.gov/periodicals/mcs2026/mcs2026-iron-steel.pdf — 82 Mt raw steel and ~21 Mt pig iron (2025), ~$149B sales value, ~105 Mt capacity, 95 Mt apparent consumption, 13% net import reliance, plant and company counts by route, and 2025 market shipment shares.
3. U.S. Department of Energy, Iron and Steel Manufacturing, 2026, https://www.energy.gov/cmei/ito/iron-and-steel-manufacturing — EAFs produce 70% of domestic steel; scrap remelting uses less than half the energy of the ore route.
4. U.S. Census Bureau, 2022 Economic Census — receipts $129.8B; CR4 59.5%, CR8 75.1%, CR20 94.4%, CR50 99.7%; HHI 1,071.9; 73 firms for NAICS 3311/33111. (No link recorded.)
5. U.S. Census Bureau, County Business Patterns 2023 — employment (78,532), establishments (359), and annual payroll ($9.1B) for NAICS 3311. (No link recorded.)
6. U.S. Census Bureau, NAICS 331110 Industry Profile, 2023, https://data.census.gov/profile/331110_-_Iron_and_steel_mills_and_ferroalloy_manufacturing?codeset=naics~331110 — 359 employer establishments; average employment ~86,000.
7. Nucor Corporation, 2025 Form 10-K, 2026, https://www.sec.gov/Archives/edgar/data/73309/000119312526071575/nue-20251231.htm — steel-mill utilization 76%→83%, ~$250M 2026 safety and environmental spending, environmental capex below $100M/yr, 2025 end-market commentary.
8. Steel Dynamics Inc., 2025 Form 10-K, 2026, https://www.sec.gov/Archives/edgar/data/1022671/000110465926021395/stld-20251231x10k.htm — metallics 55%–65% of mill cost; $13.413B steel-operations sales and $1.428B operating income.
9. Cleveland-Cliffs Inc., 2025 Form 10-K, 2026, https://www.sec.gov/Archives/edgar/data/764065/000076406526000025/clf-20251231.htm — per-ton input ratios by route, negative 5% steelmaking gross margin, >90% union-represented hourly workforce.
10. United States Steel Corporation, Form 8-K: Completion of Acquisition, 2025, https://www.sec.gov/Archives/edgar/data/1163302/000110465925062586/tm2518973d1_8k.htm
11. Nippon Steel Corporation, Completion of U.S. Steel Acquisition, 2025, https://www.nipponsteel.com/en/newsroom/news/2025/20250618_100.html — closed June 18, 2025.
12. U.S. Department of Energy, Industrial Decarbonization Roadmap, 2022, https://www.energy.gov/sites/default/files/2022-09/Industrial%20Decarbonization%20Roadmap.pdf — hydrogen DRI, iron electrolysis, electrified reheat, and carbon capture as priority technologies.
13. U.S. Department of Commerce, Section 232 Tariff Summary, 2026, https://www.trade.gov/press-release/what-they-are-saying-president-trump-strengthens-us-steel-aluminum-and-copper — 50% on steel articles, 25% on qualifying derivative articles as of April 2026.
14. U.S. International Trade Commission, Economic Impact of Section 232 and 301 Tariff Actions, 2023, https://www.usitc.gov/press_room/news_release/2023/er0315_63679.htm — imports −24%, steel-product prices +2.4%, domestic production +1.9% over 2018–2021.
15. U.S. Environmental Protection Agency, Integrated Iron and Steel Manufacturing NESHAP, 2025, https://www.epa.gov/stationary-sources-air-pollution/integrated-iron-and-steel-manufacturing-national-emission
16. Organisation for Economic Co-operation and Development, OECD Steel Outlook 2026, 2026, https://www.oecd.org/en/publications/2026/06/oecd-steel-outlook-2026_a79fb861/full-report/international-efforts-to-address-the-steel-crisis-are-intensifying_b6ca4947.html — global excess capacity 640 Mt in 2025, projected 745 Mt by 2028.