The US says it shipped $334 billion of goods to Mexico in 2024; Mexico says it received $263 billion — a $71 billion gap that runs the wrong way, and has widened every year since 2021

The United States and Mexico ran roughly $840 billion in goods across their shared border in 2024, by the US government's own count — enough to make Mexico the United States' largest trading partner. Both governments file that trade to the United Nations' Comtrade database. The two filings do not match. We pulled every leg of the bilateral flow for 2021 through 2024. On goods moving north — Mexico to the US — the two sides agree almost exactly. Mexico reported exporting $475.2 billion in 2023 and $503.4 billion in 2024; the US reported importing $480.1 billion and $510.0 billion. The importer's figure sits about 1% above the exporter's in every year — roughly what freight and insurance add to a shipment's value. On goods moving south — the US to Mexico — the two ledgers come apart. For 2024 the US reported exporting $334.0 billion; Mexico reported importing $262.7 billion from the US. That is a $71.3 billion gap, and it runs backward: the importer's figure, which normally carries freight and should be the higher of the two, is a fifth *lower* than what the exporter says it shipped. It is not a one-year blip. The southbound gap was $55.1 billion in 2021, $57.5 billion in 2022, $61.4 billion in 2023 and $71.3 billion in 2024 — widening every year.

The following is analysis, not fact. The asymmetry is directional, and that is the whole finding. The same two statistical offices, reading the same border in the same year, agree to within about 1% in one direction and split by a fifth in the other. Whatever is happening is specific to goods the US declares as bound for Mexico. The direction rules out the usual first explanation — the FOB/CIF freight convention. Trade statistics value exports at the seller's border (FOB) and imports with freight and insurance added (CIF), so an importer's number is normally the larger of the two — as it is northbound. Southbound it is smaller, so freight cannot account for it: the gap is there despite the accounting convention, not because of it. What the gap *is* remains an open question these two ledgers cannot settle. National statisticians have documented several candidate causes for exactly this kind of US-Mexico asymmetry: goods that pass through the US and are re-exported onward but recorded as US exports to Mexico; exporters coding a final destination less precisely than importers, who must declare origin to clear customs; and goods entering Mexico under temporary manufacturing-import programs that the two countries book differently. ^sspeculative Which of these carries the $71 billion — or how much is plain measurement error — cannot be isolated from the totals alone. We did not decompose the flow by commodity or firm. ^sspeculative

No goods went missing. A bilateral gap of this kind is a measurement seam, not evidence that $71 billion of exports vanished or that anyone miscounted on purpose. Statisticians expect mirror data to disagree, and the US Government Accountability Office flagged the specific limitations of US-Mexico trade figures — the maquiladora treatment, the weaker reliability of export-destination coding — as far back as 1993. On a value-added view, which nets out the imported content embedded in what crosses the border, the balance can look different again. The honest reading is narrow: the two governments' headline southbound goods numbers have not reconciled for at least four years, and the size of the 'US trade deficit with Mexico' depends on whose ledger you read. It is not a claim about who is right.

**View from the statistical agencies.** Neither the US Census Bureau nor Mexico's INEGI would call this a contradiction. Each measures what crosses its own frontier under its own rules, and the OECD treats reconciling the two — chiefly by stripping out re-exports — as one of the hardest standing problems in trade measurement, not a scandal. On that reading the $71 billion is the visible seam between two honest accounting systems that were never built to produce the same number. It is a lens on the ledgers, not a verdict they hand down.