Separate the Tariff Statute From the Tariff Rate Before Reading Inflation Data

A tariff is usually described by its headline rate. That number is the least informative thing about it. Two tariffs at an identical 15 percent will produce different price paths, different corporate hedging behavior, and different monetary policy responses depending on which statute authorized them — because the statute, not the rate, determines how long the tariff can survive without further action.

This distinction stopped being academic in the United States on February 20, 2026, when the Supreme Court held that the use of the International Emergency Economic Powers Act to impose reciprocal tariffs exceeded the authority granted by that law. That single ruling removed the legal basis for the majority of the tariff structure built during 2025 — an estimated $142 billion in IEEPA duties had been collected in that year alone. The tariff wall did not disappear. It was rebuilt out of different statutory material, and the new material has different expiry dates, different rate ceilings, and different renewal requirements.

Anyone reading tariff policy through a single headline rate, or through the headline consumer price index, is currently reading the wrong instrument.

The Data: What the Tariff Wall Actually Collects

Customs duty receipts are the cleanest available read on tariff intensity, because they are an accounting fact rather than an estimate. The Monthly Treasury Statement reports gross customs duty receipts on a fiscal-year basis running from October through September. The series below shows the buildup, the peak, and what happened after the February ruling.

Month	Gross customs duty receipts	Context
October 2024	$7.95 billion	Pre-escalation baseline
February 2025	$7.68 billion	Still at baseline
June 2025	$27.15 billion	Escalation fully in collections
September 2025	$30.49 billion	Fiscal year 2025 close
October 2025	$33.09 billion	Series peak
February 2026	$27.17 billion	Month of the ruling
May 2026	$21.93 billion	Post-ruling trough
July 2026	$24.83 billion	Partial recovery

First, the escalation was real and large. Monthly receipts moved from under $8 billion to over $30 billion inside roughly eight months. Fiscal year 2025 closed at $202.36 billion in gross customs duties. Fiscal year 2026 had already collected $269.16 billion through July, against $141.73 billion over the same October-to-July window a year earlier.

Second, the ruling did not unwind the wall. Receipts fell from a $33.09 billion peak to a $21.93 billion trough — a decline of roughly one third — and then began rising again. A statutory authority was removed and substantially replaced within a single quarter. That is the operative fact, and it is more informative than any announcement.

Third, gross receipts are not net revenue. Refunds arising from the ruling flow through separate accounting channels on their own timeline. The gross series measures what was assessed at the border, not what is ultimately retained, and the difference is not yet determinable from published monthly data.

Where the Tariff Deliberately Does Not Appear

A recurring error is to look for tariff effects in the import price index. They are not there, and that is by design.

The Bureau of Labor Statistics states plainly that "the prices for the items used to calculate the Import/Export Price Indexes exclude duties," and that the majority of import prices are quoted free on board at the foreign port. The stated rationale is that a primary purpose of the series is deflating the foreign trade components of the national accounts, which exclude taxes when measuring gross domestic product.

The July 2026 release showed all-import prices down 0.4 percent on the month and up 5.9 percent over twelve months, with import prices excluding fuel up 0.4 percent on the month and 4.5 percent over twelve months — the largest annual advance in that series since June 2022. Fuel imports were down 7.2 percent on the month but up 25.2 percent year over year.

None of those movements are tariff pass-through. They are movements in the price the foreign seller charges, measured before any duty is applied. If a foreign supplier were absorbing tariff cost by cutting the invoice price, that absorption would appear in this index as a decline. The 4.5 percent twelve-month increase in non-fuel import prices is therefore evidence against broad exporter absorption — not evidence of pass-through, which this series is built not to capture.

The Statutory Clock Is the Variable That Matters

With the emergency-powers route closed, tariff policy reverted to the ordinary trade statutes. Each of these carries its own duration limit, and those limits are what determine whether a tariff behaves as a one-time price-level shift or as a persistent inflation input.

Authority	Statutory rate ceiling	Duration limit	Continuation requirement
Section 122, Trade Act of 1974	15% ad valorem surcharge	150 days	Extension requires an Act of Congress
Section 201 safeguard	No statutory ceiling	4 years initial; 8 years in aggregate	Affirmative Commission determination; relief exceeding one year must be phased down at regular intervals
Section 301	No statutory ceiling	Terminates automatically after 4 years	Written continuation request from the benefiting domestic industry during the last 60 days of the 4-year period
Section 232	No statutory ceiling	No automatic expiry	Procedural sequence: report within 270 days of initiation, determination within 90 days, implementation within 15 days, written statement to Congress within 30 days

The asymmetry is the point. Section 122 is fast, capped, and short; Section 232 is slow to originate but has no expiry clock at all. A structure leaning on Section 122 has a hard 150-day horizon that only legislation can extend. A structure that migrates into Section 232 has converted a temporary measure into a standing one.

This is why the same headline rate carries different information depending on its source. A 15 percent surcharge with a 150-day statutory life is a bounded cost shock importers can hedge or wait out. An equivalent rate embedded in a sectoral determination with no expiry is a permanent change to the cost base, and will be treated as such in contract renegotiation, sourcing, and capital allocation.

Reading Consumer Prices Without Over-Attributing Them

The July 2026 consumer price data shows a large tariff structure coexisting with an aggregate price picture that tariffs did not cause.

Index	12-month change
All items	+3.4%
All items less food and energy	+2.5%
Commodities less food and energy commodities (core goods)	+0.8%
Services	+3.0%
Energy	+14.7%
Gasoline	+24.6%
Food	+3.0%
Shelter	+3.2%
Apparel	+3.9%
Household furnishings and operations	+2.2%
New vehicles	+0.5%

Core goods is the category where import duties land most directly. It rose 0.8 percent over twelve months while the headline index rose 3.4 percent. Energy rose 14.7 percent and gasoline 24.6 percent. The gap between headline and core goods in this print is overwhelmingly an energy story, and energy is the category least exposed to the tariff schedule.

Within core goods the dispersion fits a structure broad in legal scope but concentrated in incidence: apparel 3.9 percent, household furnishings 2.2 percent, new vehicles 0.5 percent. Those are the categories independent modelling flags as most exposed, but the magnitudes are modest and new vehicles sits near flat.

Estimates of the aggregate effect are correspondingly contained. Independent modelling as of April 2026 put the average effective tariff rate at 11.8 percent before substitution — the highest since the early 1940s, setting aside 2025 itself — with a short-run price level effect of roughly 0.7 percent if the Section 122 surcharge lapses, or about 1.1 percent if extended. Annual household cost estimates were $940 and $1,500 in 2025 dollars, ranging from $517 for the bottom decile to $2,175 for the top under the expiry case. Long-run real output was estimated 0.11 to 0.18 percent smaller, near $30 billion a year.

Level Shift Versus Inflation Rate

The central analytical point is that a tariff is a level effect, not a rate effect, unless something extends it.

A duty imposed once raises the price of affected goods once. Twelve months later the year-over-year comparison is made against the already-elevated price, and the contribution to measured inflation falls toward zero even though the price level remains permanently higher. A tariff produces sustained inflation only if it is repeatedly escalated, if it triggers second-round wage and margin responses, or if it sits in intermediate inputs that ripple downstream over several years.

A Section 122 surcharge with a 150-day life produces a level shift and then, absent congressional action, a reversal — which registers as a negative contribution on the way out. A Section 232 measure with no expiry produces a level shift that never reverses and that firms amortise into long-lived cost structures. Same rate, opposite implications for the inflation path two years out.

What Would Change the Conclusion

This framework has real limits, and several of them are strong enough to invert the conclusion under specific conditions.

Statutory expiry is not the same as economic reversal. A tariff can lapse on schedule and leave prices where they are. Once a supply chain has been re-routed at cost, once contracts have been renegotiated at higher levels, and once a domestic producer has priced up to the protected level, removing the duty does not automatically restore the prior price. Downward price rigidity is well documented. The clean symmetry implied above — level shift up, level shift down — is the optimistic case, not the base case.

Substitution can hide the effect rather than eliminate it. The gap between the pre-substitution average effective rate of 11.8 percent and post-substitution estimates around 8.2 to 10.5 percent is not free. It represents buyers moving to goods they previously did not prefer. The consumer price index measures the price paid, not the quality or suitability of the item bought. A low core goods reading is consistent with genuine absorption, and it is also consistent with substitution toward cheaper and lower-quality alternatives, which is a welfare loss that never appears in the index.

Legal uncertainty has its own price, and it is in none of these series. Between an authority being struck down and a replacement being implemented, importers face a distribution of possible duties rather than a known one, typically priced through inventory buffers, dual sourcing, and wider quoted margins. None of that appears in customs receipts or the consumer price index.

Lag structure may simply not be complete. Duties assessed in the first half of 2026 are still travelling through inventory that was purchased earlier. A core goods reading of 0.8 percent may reflect pre-tariff inventory being sold down rather than absorption of the tariff. If that is the explanation, the effect is deferred, not avoided, and the appropriate expectation is a larger core goods contribution over the following two to four quarters. Distinguishing between these two explanations from published aggregate data alone is not currently possible, and claiming otherwise would be overreach.

Energy is doing something that could reverse sharply. The 24.6 percent twelve-month gasoline increase is what pushed headline to 3.4 percent, but the monthly energy index fell 1.5 percent in July after a 5.7 percent decline in June. If those monthly declines persist, headline inflation will fall toward core while the tariff structure is unchanged — an outcome that would be misread as evidence tariffs were never inflationary, when it would only be evidence that energy base effects rolled over.

What to Watch Next Week

• The Section 122 clock. The statutory limit is 150 days with extension only by Act of Congress. Whether the surcharge is allowed to lapse, replaced by sectoral determinations, or made the subject of legislation is the single largest fork in the near-term tariff path, and it is the difference between the roughly 0.7 percent and 1.1 percent price level scenarios.
• New Section 232 initiations. The sequence runs 270 days from initiation to report, so an investigation opened now sets a cost floor well beyond any current political horizon. Initiations are a leading indicator of tariff permanence; rate announcements are not.
• Monthly customs duty receipts. The relevant question is whether the recovery from the $21.93 billion May trough continues. Sustained monthly readings back above $28 billion would indicate the replacement architecture has fully substituted for what was struck down.
• Core goods, not headline. The next consumer price release is scheduled for September 11, 2026, with producer prices on September 10 and import and export price indexes on September 16. Core goods is the series that carries tariff information; headline is currently an energy series wearing an inflation label.
• The September 15-16 policy meeting. It carries a Summary of Economic Projections. How a level-shift tariff effect is treated in projections — as a one-time adjustment to be looked through, or as a persistent input — is the transmission point from trade policy into rate policy.

Concrete Framework

A monitoring routine that survives changes in the underlying legal architecture:

1. Classify by authority before anything else. For any tariff under discussion, identify the statute. Section 122 implies a 150-day horizon and a 15 percent ceiling. Section 201 implies four years with a mandatory phase-down and an eight-year absolute cap. Section 301 implies a four-year automatic termination unless a continuation request arrives in the final 60 days. Section 232 implies no expiry at all. Any analysis that skips this step is estimating the wrong duration.
2. Separate the invoice price from the duty. Import price indexes are quoted free on board at the foreign port and exclude duties. Movements there describe foreign supplier pricing. Tariff cost appears only in customs receipts and, partially and with lag, in producer and consumer price indexes.
3. Track gross customs receipts monthly and treat them as a floor, not a total. Refunds from vacated authorities are processed separately and are not netted out in the monthly gross figure. Use the series to measure direction and magnitude of change, not to size fiscal contribution.
4. Anchor on core goods and its components. Compare the core goods twelve-month change against the exposed subcategories — apparel, household furnishings, new vehicles. Wide dispersion within core goods with a low aggregate suggests targeted incidence. A rising aggregate with narrowing dispersion suggests broadening pass-through.
5. Set expiry-date checkpoints rather than watching headlines. Statutory deadlines are known in advance. Calendar them: 150 days from a Section 122 invocation, 270 days from a Section 232 initiation, the 60-day window preceding a Section 301 four-year anniversary. These are the dates on which the tariff path actually branches.
6. Hold both scenarios with equal weight. Scenario one: the level-shift interpretation holds, core goods stays near 1 percent, the price effect is bounded near the 0.7 to 1.1 percent estimates, and expiries reverse part of it. Scenario two: inventory lag is masking incomplete pass-through, downward rigidity prevents reversal, and the effect arrives over the following four quarters rather than never. The published aggregate data as of August 2026 is consistent with both, and saying so is the accurate answer rather than an evasive one.

The headline rate will keep being the number that gets quoted. It is the number that tells the least about what happens next.

This analysis is for informational purposes only and is not investment, financial, or legal advice. Figures are drawn from published government releases and independent research as of August 2026 and are subject to revision.