Why a 50% Steel Tariff Becomes a 1% Project Cost

A 50% steel tariff does not make a finished building cost 50% more. The steel mill products index reached 374.2 in July 2026, up 22.5% from 305.5 in July 2025, yet material is less than one-third of a typical erected frame and framing is around 12% of total project cost. On those assumptions, the mill-price movement adds roughly 0.9% to the building budget—not 50% (FRED WPU1017; AISC structural-steel economics).

That 0.9% is a first-order estimate, not a claim that every project will land there. Steel-intensive industrial work, specialized imports, constrained fabrication, and critical-path delays can produce larger effects. The 22.5% index increase also cannot be attributed entirely to tariffs; demand, capacity, scrap, energy, freight, inventories, and purchasing behavior also move mill prices.

Enter your project budget and cost shares; the calculator shows which number wins for your inputs.

The Consensus View Gets the Procurement Risk Right

The received wisdom is that a 50% steel duty will produce tariff-sized steel quotes, make steel framing unaffordable, and justify switching structural systems. That position is understandable. Importers pay duties on covered goods at entry, suppliers may pass those costs through, and contractors facing uncertain quote validity may add contingency.

The consensus is also right about specialized products and schedule exposure. A narrowly specified imported section, crane component, façade assembly, or piece of industrial equipment may have few qualified substitutes. A customs or sourcing delay can then disrupt fabrication, erection, enclosure, commissioning, or other critical-path work. On a constrained project, that indirect cost can exceed the duty itself.

What does not follow is that the tariff percentage should be applied to the steel subcontract—or the entire project. The duty applies to qualifying customs value. It does not automatically apply to domestic steel, fabrication labor, detailing, erection, concrete, mechanical systems, land, or every other budget line.

The economic burden is not fixed by who writes the customs check. Importers, foreign suppliers, distributors, fabricators, subcontractors, contractors, owners, and end users may absorb different portions depending on bargaining power, inventory, demand, substitutes, procurement timing, and contract language. AGC accordingly describes tariffs as charges paid by importers while warning that affected construction-material prices can fluctuate as costs move through the market (AGC tariff resources).

One Tariff Dollar Shrinks at Each Budget Level

Three percentages must remain separate:

1. The tariff rate applied to the customs value of a covered import.
2. The steel-price change that reaches the purchaser’s quote.
3. The total project-cost change after that increase passes through the installed package and complete budget.

For a market-price estimate, the project impact is the frame share multiplied by the material share of the frame and the mill-price change. Using the generalized AISC figures, 12% × 33% × 22.5% equals approximately 0.9% of total project cost.

AISC describes material as less than one-third of a typical framing system’s cost, with fabrication and erection accounting for around 70% of the steel package. It places framing at around 12% of project cost and says a 5% steel-price increase represents less than 0.2% of total project cost under its typical assumptions.

A separate educational estimating reference places raw material at approximately 30%–40% of an installed structural-steel package. It gives approximate installed costs of $2.00–$4.50 per pound compared with $0.70–$1.20 per pound for material. The publisher labels those figures preliminary and not for construction, so they are conceptual checks rather than current bids (Steel Calculator estimating reference).

The dilution is mechanical. If raw material is 35% of an installed package and its price rises 10%, the first-order package increase is 3.5%. Fabrication, engineering, delivery, cranes, erection labor, management, and other package costs do not automatically rise by 10%.

The same arithmetic applies from package to project. If the frame is 12% of the budget, a 3.5% frame increase adds 0.42% to the total project before secondary effects.

Direct Customs Exposure Requires Different Math

The 0.9% estimate starts with an observed mill-price movement. A direct-duty estimate instead starts with the value actually exposed to customs:

Direct tariff scenario equals total relevant purchase value multiplied by the tariff-covered import share, tariff rate, and assumed pass-through.

If covered customs value is already known, multiply that value only by the tariff rate and pass-through. Applying an import share again would double-discount the exposure.

Consider the draft’s $2 million raw-steel purchase. If 20% is covered imported value, the tariff rate is 50%, and pass-through is 100%, the covered value is $400,000 and the modeled direct increase is $200,000. Against a $50 million project, that is 0.4%.

Sensitivity around the same purchase shows why the assumptions belong in the estimate:

The respective direct increases are $50,000, $150,000, and $300,000. These are hypothetical sensitivity cases, not forecasts.

A historical illustration reached a similar scale by another route: a 25% duty on $700,000 of raw steel produced a $175,000 direct increase on a $40 million project, or approximately 0.44%. It excluded market responses and indirect costs (CFMA tariff-cost illustration).

Coverage Depends on the Product, Not the Word “Steel”

Steel is not one customs category. Primary metal, derivative articles, fabricated assemblies, and equipment containing steel can receive different treatment.

AGC’s August 2026 summary reported 50% for items made entirely or mostly from covered steel, aluminum, or copper; 25% for substantial derivatives; and 15% for certain industrial or electrical-grid equipment containing those metals. Those are broad orientation categories, not shipment-specific determinations.

Product descriptions, customs classifications, country of origin, metal content, entry dates, exclusions, and applicable legal measures determine treatment. Visually similar products can receive different treatment.

Policy proposals must also be separated from final rules. BIS requested comments in August 2026 on adding 14 derivative articles, including specified cranes and lifting equipment, trailers, welding-machine parts, heat-exchanger parts, and filled steel containers. Most would face a 25% tariff if included, but the cited action was a proposal rather than an established final expansion (KPMG report on the BIS proposal).

Trade treatment can change quickly. A qualified customs or trade professional should confirm the classification, origin, customs value, rate, exclusions, entry date, and interaction with other measures for each shipment. The planning calculations here are not customs or legal advice.

Four Cost Channels Belong on Separate Lines

The direct customs duty is the most auditable channel. Its worksheet should identify the product, origin, classification, covered customs value, rate, entry timing, and pass-through assumption.

A market-wide steel response is different. Domestic steel does not pay an import duty, but its price can rise as import competition, supply, utilization, and buyer behavior change. Because scrap, energy, freight, demand, and production conditions also affect quotes, a domestic increase should not automatically be labeled a direct tariff cost.

Installed-system changes form a third channel. Alternative sourcing can change freight, coatings, testing, certifications, connection details, submittals, fabrication methods, and erection sequencing. A cheaper section can still cost more after redesign or if it misses the required grade, size, certification, or delivery window.

Indirect project costs form the fourth channel. These can include customs review, supplier qualification, inventory financing, storage, additional handling, design review, longer lead times, extended general conditions, carrying costs, contract administration, and delay.

Combining all four channels into one escalation percentage prevents anyone from seeing which assumptions can be mitigated. It can also count the same exposure twice.

Switching Structural Systems Is Not a Tariff-Rate Comparison

The supplied evidence does not provide a universal price for switching from steel to concrete, timber, or another structural system. A numeric break-even premium therefore cannot be claimed for every building.

The relevant comparison is nevertheless clear. Under the typical 12%, 33%, and 22.5% assumptions, the tariff-era mill movement creates about 0.9% of project-level exposure before indirect effects. A proposed structural switch must beat that dollar amount after engineering, foundations, fire protection, floor depth, enclosure interfaces, schedule, permitting, procurement, and lost design work are included.

That is why the 50% headline is a poor trigger for redesign. The alternative is not competing against 50% of the building budget or even 50% of the erected frame. It is competing against the much smaller amount that reaches the project after material and frame shares are applied.

There are projects where switching can still win. A steel-intensive industrial frame, long-span building, bridge, high-rise, or design dependent on specialized imports may sit well above the generalized shares. A system change may also deliver independent benefits in schedule, carbon, fire resistance, vibration, architectural expression, or local trade availability. Those benefits should be evaluated on their own merits rather than credited entirely to tariff avoidance.

Historical Prices Show Volatility, Not Causation

The WPU1017 observations supplied for December were 186.9 in 2017, 223.0 in 2018, and 187.4 in 2019. That represents an approximately 19.3% rise followed by an approximately 16.0% decline.

The sequence shows why procurement timing matters and why a one-year increase should not automatically become a permanent escalation assumption. It does not prove that tariffs caused the movement. The series alone cannot separate duties from demand, capacity, energy, scrap, freight, exchange rates, inventories, or product mix.

A separate historical comparison reported that the steel mill products index increased 12.3% after the March 2018 measures while the new-construction price index increased by less than 4.3%. The gap is consistent with dilution and incomplete immediate pass-through, but it does not establish a specific contractor margin loss.

Broad construction estimates require the same caution. Cushman & Wakefield estimated that the combined tariffs in effect as of April 7, 2026 would add 6% to construction-material costs and 3% to total U.S. commercial project costs relative to a 2024 baseline. That is a combined-tariff model, not an observed steel-only increase (Cushman & Wakefield tariff estimate).

The Defensible Budget Is a Dated Range

A project estimate should record steel tonnage, raw-material value, installed-package value, product descriptions, classifications, origins, covered shares, applicable rates, quote dates, exclusions, pass-through assumptions, and the total project budget. Domestic market response, redesign, storage, financing, and schedule allowances should remain separate.

Contract allocation matters as much as arithmetic. Fixed-price, guaranteed-maximum-price, and reimbursable arrangements distribute exposure differently depending on allowances, exclusions, change-in-law language, notice requirements, substitution rights, documentation, and change procedures. No clause automatically grants recovery; project-specific legal advice is necessary.

Early purchasing can reduce later price exposure but adds financing, storage, insurance, handling, damage, cancellation, and specification-change risk. Domestic sourcing can reduce direct customs exposure without eliminating domestic price movements or capacity constraints. Supplier diversification can improve resilience while adding qualification and coordination work.

The useful answer is therefore not “50%.” It is a dated project range that traces covered value through pass-through, material share, installed frame, and total budget. For typical inputs, that chain lands near 1%. Projects move away from that result only when their actual steel intensity, import exposure, market response, or schedule consequences justify it.