Treasury Secretary Scott Bessent calls the court-ordered return of 166 billion dollars in invalidated IEEPA duties a form of corporate welfare, as thousands of small importers report that the money they are owed has not arrived
WASHINGTON, Sept. 23, 2026 – Seven months after the Supreme Court held that the International Emergency Economic Powers Act did not authorize the tariffs imposed under it, the government is still returning the money, the Treasury Secretary is openly unhappy about having to do so, and a large number of the importers entitled to refunds say the process is not working for them.
The dispute came back into public view this week. In remarks to reporters outside the White House, Treasury Secretary Scott Bessent described the refunds as corporate welfare and rejected suggestions that the funds should instead be routed to consumers. “The American people had the money in the U.S. Treasury, and we were forced to give it back,” Bessent said, in comments carried by multiple outlets. A Bloomberg analysis published Sept. 22 examined where the refunds actually stand.
The figures at the center of the argument are large. Approximately 330,000 importers paid or deposited an estimated 166 billion dollars in IEEPA duties across more than 53 million individual customs entries, according to Customs and Border Protection data and court filings. That is the pool the Supreme Court’s Feb. 20 ruling made refundable.
What has actually been repaid is a moving and somewhat opaque number. As of July 10, 2026, CBP had accepted roughly 121.75 billion dollars in claims for processing and repaid about 86.3 billion dollars to importers including statutory interest. A later court-ordered filing with the Court of International Trade reported that as of July 31 roughly 100 billion dollars had been paid out of about 128.7 billion dollars accepted for processing, which is around 60 percent of the total collected. The government has not published a comprehensive figure for what Treasury has actually disbursed.
The gap between the amount collected and the amount returned, and the gap between the amount accepted for processing and the amount paid, are where the current political fight lives.
How the money came to be owed
The Supreme Court ruled on Feb. 20, 2026 that the International Emergency Economic Powers Act did not confer authority to impose the tariffs that had been levied under it. The statute, enacted in 1977, grants the president broad powers to regulate international commerce in response to declared national emergencies. The question before the Court was whether that grant extends to the imposition of across-the-board import duties. The Court held that it does not.
The ruling did not disturb tariffs imposed under other authorities. Section 232 duties on steel, aluminum and copper, grounded in national security findings by the Commerce Department, remain in force and have in fact been expanded repeatedly through 2026. Section 301 measures against specific unfair trade practices are likewise unaffected. What the decision invalidated was a particular legal theory, not the tariff project as a whole.
That distinction matters enormously for importers trying to work out what they are owed. A single entry may have carried IEEPA duties, Section 232 duties and ordinary most-favored-nation duties simultaneously. Only the first category is refundable. Disentangling them across 53 million entries is the administrative problem that has consumed CBP for most of the year.
CAPE and its discontents
CBP’s response was to build a system. The Consolidated Administration and Processing of Entries, universally shortened to CAPE, launched its first phase on April 20, 2026, giving importers and authorized customs brokers a direct pathway to file refund claims.
The mechanics are straightforward in description. An importer or its broker files a declaration through CAPE listing the entries for which a refund is sought. CBP validates the claim, recalculates the duties owed on those entries with the IEEPA component removed, and reliquidates the entries at the corrected amount. Reliquidation triggers repayment of the difference, plus statutory interest. CBP has said that valid refunds will generally be issued within 60 to 90 days after a claim is accepted, with more complex cases taking longer.
In practice, importers have encountered friction at every stage, and the pattern of complaints has been consistent enough that it has drawn congressional attention.
The first problem is account approval. Filing through CAPE requires an approved account, and CBP has in some cases failed to approve importer accounts within 90 days of liquidation. Because refund eligibility is tied to liquidation timing, an importer whose account approval is delayed past a deadline can find itself locked out of money it is plainly owed through no fault of its own. This is not a marginal technicality. It is the difference between recovering a duty payment and not recovering it.
The second problem is data. Reconstructing IEEPA duty exposure across years of entries requires entry-level records that many smaller importers never maintained in a form suitable for this purpose. Companies that relied on a customs broker to handle filings and never systematically archived entry summaries are now trying to rebuild a duty history from bank records and commercial invoices. Brokers, dealing with the same crunch across hundreds of clients simultaneously, have limited capacity to help.
The third problem is scale and sequencing. With 330,000 importers in the queue, the practical effect of a first-come, first-served processing model is that sophisticated filers with dedicated trade compliance staff submitted early, clean, complete claims and were paid, while smaller companies without that capability are further back in a line that is still moving.
In September, members of Congress urged CBP to move quickly to ensure that American importers receive full refunds, citing reports of significant logistical hurdles and noting specifically that many of the importers at risk of missing out are small businesses operating on thin margins. Senator Maria Cantwell had earlier written to Bessent on the subject in February, shortly after the ruling.
The corporate welfare argument
Bessent’s framing deserves examination on its merits, because it is not simply a political talking point and it is not simply wrong.
The argument, as he has put it, is that the refunds deliver a windfall to large corporations. There is a version of this that is analytically serious. Tariffs are collected from importers of record, but their economic incidence is distributed across the supply chain. To the extent that an importer passed the duty cost forward into wholesale and retail prices, the ultimate payer was the consumer. Refunding the duty to the importer therefore returns money to an entity that may not have borne the cost, and there is no mechanism by which those funds flow back to the households that paid higher prices.
That is a genuine problem in the economics of tariff remedies, and it is not unique to this episode. It is the standard difficulty with any refund of an indirect tax.
There are, however, several points on the other side.
The incidence question is empirical and varies enormously by product, market structure and the competitive position of the importer. Research on the tariff rounds of the past several years has found substantial pass-through in some categories and substantial margin absorption in others. Firms selling into competitive markets with price-sensitive customers, which describes most small importers, frequently could not pass the full duty forward and ate part of it. For those firms the refund is compensation for a cost actually borne, not a windfall.
There is also the matter of working capital. Even where a duty was eventually passed through, the importer financed it in the interim. A company that paid millions in duties at the port and recovered the cost over subsequent quarters through pricing carried that cost on its balance sheet, often on borrowed money. Statutory interest on the refund addresses part of this but does not capture the opportunity cost or the credit consequences for firms that drew down lines of credit to pay duties.
And there is the legal point, which is the one that ultimately governs. The Supreme Court held that the money was collected without authority. Money collected by the government without legal authority is returned to the party that paid it. Whether the recipient is sympathetic, whether it passed the cost along, and whether a policymaker would prefer to spend the funds differently are not considerations that bear on the obligation. The remark that the government was forced to give it back is, as a description of the legal position, accurate.
The distributional concern Bessent raises is real, but the remedy for it would have been not to collect the duties unlawfully in the first place.
Who is actually affected
The composition of the 330,000 importers matters to how this story should be read, and it is frequently misrepresented in both directions.
The dollar concentration is at the top. A relatively small number of very large importers, national retailers, automotive manufacturers, electronics distributors and industrial firms, account for a disproportionate share of the 166 billion dollars. These companies have dedicated customs compliance departments, retained trade counsel, and complete entry records. They filed early and have largely been paid.
The headcount concentration is at the bottom. The overwhelming majority of the 330,000 are small and mid-sized businesses whose individual refunds run from a few thousand dollars to a few hundred thousand. For a company with 4 million dollars in annual revenue and a 6 percent operating margin, a 180,000 dollar refund is not a windfall. It is roughly nine months of profit and, in many cases, the difference between servicing debt taken on during the tariff period and defaulting on it.
This is the population that congressional letters have focused on, and it is the population least well served by a self-service claims portal that requires clean entry-level records and an approved account.
What importers should do now
For any company that paid IEEPA duties and has not yet been made whole, the operative advice from customs practitioners is consistent.
Verify account status in CAPE first. Because the 90-day approval issue can foreclose eligibility, confirming that an account is approved and in good standing is the highest-value single action available. An importer discovering an approval problem now has time to escalate it; one discovering it after a deadline may not.
Reconcile independently rather than relying solely on broker records. Pull entry summaries, match them against duty payments, and identify the IEEPA component line by line. Where records are incomplete, CBP’s own ACE data can be requested and used to reconstruct the history.
Separate the refundable from the non-refundable with care. Section 232 duties on steel, aluminum and copper derivatives, which have been expanded several times through 2026 and now apply to full customs value on a widening list of derivative articles, are not affected by the ruling. Claiming them produces a rejected or delayed filing.
Account for the interest. Statutory interest accrues and is payable with the refund. It is taxable income in the year received, which has consequences for companies that have already closed books on the periods in which the duties were paid.
Treat the refund as a balance sheet event, not a revenue event. Refund timing is uncertain enough that building it into a cash forecast as a dated receipt is a mistake. Several importers have reportedly made commitments against expected refunds and found the money slower to arrive than the 60 to 90 day guidance suggested.
The interest question
One aspect of the refund program that has received less attention than it deserves is statutory interest, and it is financially significant enough to warrant its own analysis.
Refunds of unlawfully collected customs duties carry interest under federal law, running from the date of payment or deposit. Over a period in which IEEPA duties were collected across roughly two years, and with refunds now being issued seven months and counting after the Supreme Court ruling, the accrued interest on the 166 billion dollar principal is a substantial sum in its own right.
CBP’s July figures explicitly noted that the roughly 86.3 billion dollars repaid as of July 10 included statutory interest, which means the principal component was somewhat smaller than that headline number. The distinction matters for importers trying to reconcile what they have received against what they are owed, because an importer comparing a refund check against its own record of duties paid will find the check larger and may not immediately understand why.
It matters more for tax treatment. Interest received on a duty refund is generally taxable income in the year of receipt, while the underlying duty refund reverses a cost that was deducted in an earlier period. The accounting treatment differs by circumstance and by whether the company expensed the duties or capitalized them into inventory, and importers receiving large refunds in 2026 for duties paid in 2024 and 2025 face a genuine tax planning question that many have not addressed.
Practitioners have advised affected companies to engage tax advisers before rather than after the refund arrives, on the straightforward logic that the options available narrow considerably once the money is in the account and the fiscal year has closed.
Small business and the capacity gap
The congressional intervention in September framed the refund problem explicitly as a small business issue, and the framing is accurate for reasons worth setting out in some detail.
Customs compliance is a specialized function. A large importer employs licensed customs brokers in-house, maintains entry-level records in a dedicated trade management system, and retains outside trade counsel on standing arrangements. When a refund opportunity of this magnitude appeared, such a company could assign staff to it within days.
A small importer has none of that infrastructure. It typically uses an outside broker for filings, keeps commercial invoices and bills of lading but not systematically archived entry summaries, and has no employee whose job description includes customs. When CAPE opened in April, that company’s options were to attempt the filing itself with incomplete records, to ask its broker to do it in competition with the broker’s larger clients, or to hire a consultant whose fee might represent a meaningful fraction of the expected refund.
Brokers themselves have been a bottleneck. The customs brokerage industry was not sized for a one-time event requiring detailed retrospective reconstruction of entry histories across 53 million entries. Brokers prioritized, rationally, their largest and longest-standing accounts. Smaller clients waited.
The account approval delays compound this. An importer with a dedicated compliance function noticed immediately when an account approval did not come through and escalated it. An importer without one may not have noticed at all, and may still be unaware that its claim is stalled rather than merely queued.
None of this is malicious, and CBP has been operating under considerable pressure to build and run an unprecedented system on a court-imposed timetable. But the cumulative effect is a program whose burdens fall most heavily on the participants least able to bear them, which is the point the congressional letters have made.
What the numbers do and do not show
Some care is warranted in interpreting the published figures, because they measure different things and have been conflated in public discussion.
The 166 billion dollars is the estimated total of IEEPA duties paid or deposited. It is an estimate, not a reconciled ledger, and it covers entries some of which may not ultimately qualify for refund for reasons unrelated to the merits, including entries that were never liquidated, entries subject to other proceedings, and entries where the importer of record no longer exists.
The amount accepted for processing, reported as roughly 121.75 billion dollars as of July 10 and about 128.7 billion as of July 31, measures claims that have passed initial validation. Acceptance is not payment, and the gap between acceptance and disbursement represents claims in the reliquidation pipeline.
The amount repaid, about 86.3 billion as of July 10 and roughly 100 billion as of July 31, is the figure closest to money actually returned, though it includes interest and the government has not separately published a principal-only total. Treasury’s own disbursement figures have not been comprehensively released.
The implication of the July 31 filing, that around 60 percent of the collected total had been paid out, is therefore a reasonable summary but not a precise one. What can be said with confidence is that a substantial majority of the dollars have moved and a substantial majority of the importers have not yet been fully paid, because the dollars are concentrated among filers who were quick and the headcount is concentrated among filers who were not.
The wider policy consequence
The refund episode is shaping trade policy in a way that will outlast the payments.
The most immediate effect is a shift in the legal architecture of American tariffs. With IEEPA foreclosed as a basis for broad duties, the administration has leaned harder on Section 232, which requires a Commerce Department national security finding but is considerably more robust to judicial challenge, and on Section 301, which requires an investigation into specific foreign practices. The tariff program has not shrunk. It has been re-founded on sturdier statutory ground. Importers hoping the Supreme Court ruling signaled a general retreat have been disabused of that expectation by the steady expansion of Section 232 coverage through 2026.
The second effect is on how businesses model tariff risk. A duty that might be refunded years later after litigation is a different financial instrument from a duty that is simply a cost. Some importers have begun treating contested duties as contingent assets and structuring supplier and customer contracts to allocate the recovery explicitly, which is a sophistication the market did not previously require.
The third effect is fiscal and political. A 166 billion dollar refund obligation is a significant line item, and Bessent’s public irritation reflects a real budgetary reality. It also establishes a precedent: the scale of the exposure created by legally aggressive tariff action is now documented, and future administrations weighing the same trade-off will have a number to look at.
For the small importers still waiting, none of that is the point. The point is that a court said the money was theirs, seven months have passed, and a portal has not yet given it back.