Trump and Sheinbaum both say a bilateral deal is close, but reporting this week places Commerce Secretary Howard Lutnick between Mexico and the Section 232 steel and aluminum relief it has spent five months seeking.
WASHINGTON, Oct. 23, 2026
Two presidents have now said in the same week that a United States trade agreement with Mexico is nearly done. The negotiating record suggests otherwise, and the obstacle has a name.
Mexican President Claudia Sheinbaum told reporters on Tuesday, Oct. 20, that a deal with Washington is “very close” while acknowledging that “there are still some issues” outstanding. President Donald Trump had used similar language days earlier at a rally in Park City, Kansas, on Oct. 19, calling a prospective agreement “great” and telling the crowd, “We’re not going to lose 150 billion dollars with Mexico anymore,” according to an Oct. 22 account published by Latin Post.
Yet reporting from the same window describes talks that have cooled rather than converged. The Wall Street Journal reported on Oct. 15 that negotiations had effectively stalled. Bloomberg reported on Oct. 20 that the talks had hit an impasse, with Commerce Secretary Howard Lutnick exerting decisive influence over the American position. Mexico’s ambassador in Washington, Roberto Lazzeri, has said publicly that the two sides are “a little bit stuck on the details,” and identified Section 232 steel and aluminum tariffs as the principal sticking point.
At her Oct. 22 morning press conference, Sheinbaum did not revisit the metals question directly. She made a broader argument instead. Calling it “legitimate” for any country to want to strengthen its domestic manufacturing, she said the three North American economies “should see ourselves” and “work as a region,” adding that “it’s a lot better to work together in order to compete with other regions of the world than to isolate oneself,” and that “economic integration is very important for the three countries and it creates a very competitive region.”
That is the case Mexico has been making since May. The question is whether it is winning.
WHAT MEXICO IS ASKING FOR
Mexico’s objective in these negotiations has been consistent and narrow. It wants relief from three United States measures: the Section 232 tariffs on steel, the Section 232 tariffs on aluminum, and the Section 232 tariff on automobiles assembled in Mexico.
The rates involved are substantial. Vehicles produced in Mexico face a 25 percent tariff. Steel and aluminum carry 50 percent duties following the administration’s 2026 adjustments, which also broadened the derivative product lists and changed how the duty base is calculated for many articles. Mexican officials have floated a 15 percent ceiling as an acceptable landing zone across the affected categories, a figure that would align Mexico with the all in rates several other partners negotiated earlier in the administration.
The ask is not unreasonable on its face. Mexico is the United States’ largest goods trading partner, the single largest source of imported vehicles and parts, and a participant in a free trade agreement that remains in force. It has also aligned itself with Washington on third country issues that matter to the administration, maintaining a 50 percent tariff on Chinese automobiles despite pressure from Beijing and from Chinese manufacturers with plant investments in Mexico.
What makes the ask hard is that steel and aluminum are the two sectors where the administration has been least willing to grant exceptions to anyone.
THE LUTNICK FACTOR
The most specific reporting on why the talks stalled points to the Commerce Department.
Citing people familiar with the negotiations, the Journal reported that Lutnick has refused to support meaningful tariff reductions and has blocked proposals that would have lowered duties on finished steel and aluminum products. Domestic aluminum producers are described as having lobbied to preserve the tariffs, and the Commerce Department administers the Section 232 programs, including the inclusions process by which new derivative products are swept into the metals tariffs.
Mexican officials have been careful not to confirm that account. Economy Secretary Marcelo Ebrard said he did not get the impression that Lutnick opposed reducing the metals tariffs, a formulation that is diplomatic rather than contradictory. Ebrard has led the Mexican side through four bilateral rounds with United States Trade Representative Jamieson Greer since May 2026, and Greer remains the formal lead negotiator for Washington.
The structural point matters more than the personalities. Section 232 authority sits with Commerce, which conducts the investigations and recommends the remedies, while trade negotiations sit with the Office of the United States Trade Representative. When the concession a partner wants is relief from a Section 232 measure, the negotiator cannot deliver it without the agency that owns the program. That is a recurring feature of this administration’s trade architecture, and Mexico is not the first partner to run into it.
THE DEFICIT ARGUMENT
Trump’s rally framing, that the United States will no longer “lose” 150 billion dollars to Mexico, points at the number that has animated his approach to the relationship.
Census Bureau data cited by Latin Post show that from January through August 2026 the United States imported 419.344 billion dollars in goods from Mexico and exported 263.473 billion dollars, producing a goods deficit of 155.871 billion dollars for the eight month period. That is an increase of 21.2 percent over the same months of 2025.
A widening bilateral deficit during a period of aggressive tariff action is politically awkward for the administration, and it strengthens the hand of officials who argue that existing measures should be tightened rather than relaxed. It also complicates Mexico’s argument that integration benefits all three North American economies, because the headline number appears to show the benefit flowing one way.
The counterargument, which Mexican officials and much of the United States manufacturing sector make, is that the bilateral goods balance is a poor measure of value capture in an integrated production system. A vehicle assembled in Mexico and shipped north typically contains a large share of United States origin content, and the gross import figure attributes the full customs value to Mexico. Tariffing that vehicle at 25 percent taxes the American content inside it as well.
That argument has not visibly moved the administration’s framing, though it has shaped the technical work. Commerce issued guidance in mid October, published in the Federal Register on Oct. 14, on how to assess United States content for Section 232 goods eligible for preferential treatment under the United States Mexico Canada Agreement. That guidance is the mechanism by which American content inside a Mexican assembled good can reduce the duty base, and it is one of the few areas where the two sides have made concrete progress.
THE CANADA SHADOW
Mexico is negotiating in the shadow of what happened to Canada.
The Canadian track collapsed in late August 2026. Washington imposed 50 percent duties on Canadian goods under Section 338 of the Tariff Act of 1930, collected from Aug. 22. After Canada retaliated, the administration issued five further proclamations on Sept. 8 that modified the product lists effective Sept. 15 and imposed outright import bans on certain Canadian alcoholic beverages and dairy products for goods entered on or after Sept. 29. Importers have been told that preferential origin under the North American agreement does not exempt the affected goods, and the first court challenge to the Section 338 action was filed at the Court of International Trade on Oct. 20 by a food importer seeking class treatment.
Mexico is now the only North American partner still in formal bilateral negotiations with Washington. Sheinbaum has pursued what her government describes as a cool headed approach, avoiding retaliation, keeping channels open and absorbing tariff increases rather than matching them. The approach has preserved the negotiation. It has not yet produced the tariff relief that is its entire purpose.
The Canadian example cuts both ways in Mexican internal debate. It demonstrates the cost of escalation, which supports continued patience. It also demonstrates that patience is not automatically rewarded, which strengthens voices arguing for a harder line.
THE USMCA CLOCK
Running underneath the bilateral talks is the larger question of the agreement itself.
On July 1, 2026, the administration declined to extend the United States Mexico Canada Agreement at its scheduled six year joint review, opting instead for annual reviews. Canada and Mexico had both supported a 16 year extension. The agreement remains in force until July 1, 2036, but without an extension at the joint review its long term status becomes a recurring annual question rather than a settled fact.
The formal review machinery is now turning. USTR published a Federal Register notice on Oct. 5, 2026, opening consultations and scheduling a public hearing ahead of the 2027 joint review, with comments due by 11:59 p.m. Eastern on Jan. 12, 2027. That notice sets no rates and changes no duties, but it establishes the record on which the next review will be argued.
For Mexico, the interaction between the bilateral tariff talks and the review process is delicate. Securing metals and automotive relief through a bilateral deal would strengthen Mexico’s position going into the review. Failing to secure it, and then entering a review in which the United States holds both the tariff leverage and the agenda, is a materially worse starting position.
THE ELECTION CALENDAR
Both governments were understood to want an announcement before the United States midterm elections on Nov. 3, 2026. According to people familiar with the talks cited in the Journal’s reporting, American officials have since grown comfortable with letting an announcement slip past that date.
That shift is significant for anyone trying to time commercial decisions. A deal announced before Nov. 3 would have been driven partly by political need, which usually means accepting a narrower and faster package. A deal negotiated after the election is more likely to be comprehensive, and more likely to be slow.
Public opinion in the United States has moved against tariffs over the course of the year, and the administration has nonetheless kept them central to its economic message. That combination argues against pre election concessions on the most visible measures, which is exactly what steel and aluminum relief would be.
WHAT IT MEANS FOR IMPORTERS AND EXPORTERS
For companies with North American supply chains, the practical guidance in this moment is about preserving optionality rather than betting on an outcome.
Classification and origin discipline is the first priority. The value of any eventual metals relief, and the value of the existing United States content mechanism, both depend on being able to substantiate origin and content at the line level. Companies that have been filing on estimates should expect that a negotiated settlement will come with documentation requirements at least as demanding as the current ones.
Second, model the three plausible outcomes rather than one. The first is a bilateral package that caps the relevant rates at or near 15 percent, which would materially reduce landed cost on Mexican steel, aluminum and vehicles. The second is a narrower deal that addresses automotive content rules and leaves the metals rates intact, which is the shape most consistent with Commerce’s reported position. The third is no deal before the 2027 review, leaving current rates in place through at least the first half of next year.
Third, watch the derivative lists. Commerce continues to run an inclusions process that adds products to the steel, aluminum and copper tariffs. A company whose Mexican sourced article is not currently a covered derivative can become covered without any change in the headline negotiation. That process has added hundreds of product categories since 2025 and has not paused for the bilateral talks.
Fourth, for exporters, consider the retaliation question honestly. Mexico has not retaliated. Its restraint is a deliberate policy choice that could change if the talks fail outright, and the sectors most exposed in any Mexican response are the agricultural and consumer goods categories that have featured in previous Mexican countermeasure lists.
THE SECTORS CARRYING THE COST
The burden of the current tariff structure is not spread evenly across the bilateral relationship, and understanding where it concentrates explains why Mexican negotiators have prioritized metals and vehicles over everything else.
Automotive is the largest single exposure. Mexico is the leading source of imported light vehicles and of automotive parts for United States assembly plants, and the production system is genuinely binational, with components crossing the border several times before a finished vehicle emerges. A 25 percent duty on the finished vehicle, applied on top of metals duties captured earlier in the chain, compounds through the structure in ways that a single rate comparison understates. Assemblers have responded with a mix of price increases, margin absorption and selective shifts in build location, and the effects show up in United States consumer vehicle prices with a lag.
Steel and aluminum are the second exposure, and the one that generates the most political heat on the American side. The 50 percent rate applies not only to primary metal but to a long and growing list of derivative articles, which means the tariff reaches fabricated products, structural components, appliance parts, fasteners, electrical housings and packaging. Mexican producers of finished metal goods have found themselves competing against American producers who receive the protection while buying inputs at tariff inflated prices, a dynamic that has generated complaints from United States manufacturers as well.
Agriculture and food products sit in a third category. They are largely outside the Section 232 structure, which is why they have not featured in the negotiation, but they are the categories most exposed if the relationship deteriorates and Mexico abandons its policy of restraint.
Finally, there is the services and logistics layer. Cross border trucking, warehousing and customs brokerage volumes track goods flows, and the uncertainty premium in the current environment has pushed shippers toward shorter commitments and more buffer inventory. Those costs are real even though they never appear as a duty line on an entry summary.
A NEGOTIATION WITHOUT A DEADLINE
One structural weakness in Mexico’s position deserves attention. There is no forcing event.
The tariffs are already in place and are being collected. There is no expiry date that would restore the previous rates absent an agreement, no statutory clock that compels a decision, and no scheduled proceeding before the 2027 joint review that requires either side to move. The midterm elections provided a soft deadline, and reporting indicates that the American side has already let go of it.
Negotiations without deadlines tend to settle at the convenience of the party that is comfortable with the status quo. On the current rate structure, that is Washington.
Mexico’s available responses are limited. Retaliation would invite the Canadian outcome. Litigation is not realistically available against a Section 232 measure. Appeals to regional solidarity, which is the theme Sheinbaum returned to on Oct. 22, are a long term argument rather than a lever. What remains is the slow accumulation of technical agreement on content rules, origin verification and third country coordination, in the hope that enough of it eventually makes the headline rates easier to concede.
THE SHAPE OF A DEAL
If an agreement does emerge, its likely architecture is becoming visible from the pieces already in place.
Tariff rate quotas are the most probable instrument for metals. They allow the administration to claim that the Section 232 measure remains intact while giving a partner duty free or reduced duty access up to a historical volume. They also give Commerce a continuing administrative role, which addresses the institutional objection reported to be at the heart of the impasse.
Automotive relief is more likely to come through content rules than through rate cuts. The United States content guidance issued in October provides a mechanism to reduce the effective duty on a Mexican assembled vehicle without changing the 25 percent headline rate, by excluding qualifying American content from the duty base. That is a politically cheaper concession than a visible rate reduction.
A third element, present in several of the administration’s other 2026 agreements, is coordination against third country transshipment and overcapacity. Mexico’s existing 50 percent tariff on Chinese vehicles is the kind of commitment that fits naturally into such a package, and extending similar treatment to Chinese steel and aluminum inputs would address the leakage concern that domestic producers raise most often.
THE RISK OF DRIFT
The most likely failure mode here is not a rupture. It is drift.
Both governments have strong reasons to keep saying a deal is close. For Trump, an imminent agreement with the largest trading partner is a useful answer to a widening deficit and softening domestic support for tariffs. For Sheinbaum, the prospect of relief justifies a policy of restraint that has cost Mexican exporters real money for more than a year.
Neither incentive requires the deal to actually close. And the longer the negotiation runs without a result, the more the 2027 joint review becomes the forum in which these questions are settled, on a timetable and with an agenda that Washington controls.
One further variable could change the arithmetic quickly. The administration’s tariff program has been reshaped repeatedly by litigation this year, and the Section 338 action against Canada drew its first court challenge on Oct. 20. Section 232 measures have proved far more durable in court than emergency authorities, so Mexico cannot expect judicial relief on the metals tariffs. But a ruling that constrains one instrument tends to increase the administration’s reliance on the others, which argues for the steel, aluminum and automotive duties becoming more entrenched rather than less.
Lazzeri’s formulation, that the talks are stuck on the details, is accurate but incomplete. The details in question are not technical. They are the rates themselves, and they are owned by an agency that has spent the year making them higher rather than lower. Until that changes, “very close” is a description of sentiment rather than of distance.