Trade integration and domestic value capture

The OECD country notes provide the following observations on a common 2022 basis. Foreign content is measured against gross exports of goods and services. Foreign final-demand exposure is measured against domestic value added, a different denominator.

Sources: OECD country notes for the United States and Mexico, based on the 2025 ICIO database. The reports identify the observation year separately from their publication date.

The first row indicates a larger imported production contribution within Mexico's export basket. The second indicates a greater share of Mexican production income connected to demand abroad. Neither establishes that the average Mexican export has a smaller economic benefit than an American export. The size of domestic markets, industry composition and the scale of trade differ. A content share and a contribution to income answer distinct questions.

Using the rounded foreign-content shares, the complementary domestic shares are 91.7 percent and 66.2 percent. These are economy-wide export-content measures, not estimates for bilateral U.S.–Mexico trade, the automotive industry or a particular company. Applying them to every shipment would replace product-specific sourcing with an aggregate assumption.

Consider an entirely hypothetical chain. A U.S. producer creates an intermediate worth forty-five entirely from U.S. value added and exports it to Mexico. A Mexican assembler also imports ten of Canadian value and twenty from other economies, then adds twenty-five through resident production. The finished product is exported to a U.S. final purchaser for one hundred.

Assume no other inputs, taxes, transport margins, inventories or transactions. The Mexican shipment is one hundred, but Mexican value added embodied in it is twenty-five. North American value added is eighty. U.S. value added is forty-five, even though the finished product arrives as a gross import from Mexico.

Under the same restricted assumptions, U.S. gross bilateral exports to Mexico are forty-five and imports from Mexico are one hundred, giving a bilateral gross balance of minus fifty-five. The Mexican value added absorbed in U.S. final demand is twenty-five. The rest of the import value is attributable to U.S., Canadian and other production. The difference is a reconciliation of origins, not evidence that either trade record is fabricated.

The example deliberately returns the U.S. intermediate to U.S. final demand. It is therefore different from a chain in which Mexican households consume the U.S. input. Final-demand attribution must follow the actual destination rather than assume the immediate export partner is the ultimate purchaser.

In the illustration, Mexico's domestic share is twenty-five percent while North America's share is eighty percent. Both are correct under their respective boundaries. An input from Canada is foreign to Mexico but domestic to a three-economy regional account. A regional comparison should remove internal transactions consistently rather than merely add three national export totals.

Rules of origin pose a separate question. A trade agreement can define qualifying content through specified legal tests, product classifications and calculation methods. Those tests are not automatically identical to an input-output estimate of domestic or regional value added. This study does not assess legal origin or tariff eligibility for any product.

The statistical and legal accounts can both be useful. A firm needs the applicable origin rules for a transaction and production-origin information for its economic exposure. A customs certificate does not supply a complete income-distribution account, while a model-based TiVA share does not establish shipment-level legal compliance.

For strategic analysis, the regional perspective can reveal production relationships hidden by a gross bilateral deficit. It cannot establish that every regional input is readily substitutable or that relocating a stage has no cost. The location and capability of the specific producer remain operationally important even where the regional value total is unchanged.

INEGI's VAEMG publication defines its export-value-added measure through global manufacturing production less imported intermediate inputs. Its methodology also expresses the measure as manufacturing gross value added plus intermediate inputs of national origin. That stated construction should be retained; the label alone does not establish equivalence to OECD's full inter-country tracing of direct and indirect value-added origins. INEGI, VAEMG2023 methodological note.

A supplier invoice of national origin can contain imported value farther upstream. Whether and how a particular statistical compilation traces that content depends on its method. For a U.S.–Mexico comparison, the researcher should reconcile production coverage, valuation and indirect-import treatment before placing the national indicator beside an OECD domestic-content share.

This is not a reason to discard the national series. It can describe its selected manufacturing population and developments at the boundary it measures. Its industry detail may be valuable for domestic industrial analysis. The limitation arises when a figure is assigned a different meaning without performing the bridge between methods.

The same caution applies to comparing a manufacturing series with an economy-wide export series. Services and primary products alter the aggregate, while the manufacturing account may select particular enterprises or production arrangements. A difference can reflect coverage as well as sourcing. Common labels do not remove the need to identify the actual units included.

The final product's customs category does not capture all the industries generating its value. Design, logistics, wholesale services, finance and other business services can contribute upstream. A factory-centered comparison that attributes all domestic content to assembly misses those production links.

Services can also cross a border independently or be supplied by a resident affiliate. Their treatment in production and trade accounts depends on the transaction and residence of the units. The nationality of the corporate group does not by itself locate the value added. A U.S.-owned plant resident in Mexico can generate Mexican domestic value added while part of its resulting primary income accrues to nonresident owners.

The distribution between wages, taxes, depreciation and operating income requires an additional reconciliation. A domestic-content percentage is not a profit margin or the share reaching households. For questions about income retained by residents, payments to nonresident owners and receipts from abroad must be considered separately.

This distinction is relevant to supplier development. A resident supplier can create local wages and capabilities regardless of its shareholder nationality. Conversely, a domestically owned distributor can sell largely imported goods. Procurement classified only by ownership cannot establish the underlying domestic production contribution.

A small value-added share can correspond to an indispensable component. If a low-cost input has no qualified alternative, its interruption can stop the entire chain. The value weights identify income origins, not the operational probability or severity of disruption.

A business comparison should therefore examine supplier substitutability, qualification lead times, inventories and transport reliability alongside value-added accounts. Increasing the domestic share can diversify some exposures while concentrating others. A single national percentage cannot establish the resilience of a production system.

The reverse is also possible: using imported intermediates can support specialization and larger domestic output. A higher imported share is not automatically evidence of reduced domestic value in absolute terms. If exports rise from one hundred to one hundred fifty while the domestic share falls from sixty to fifty percent, domestic content rises from sixty to seventy-five. The share falls, but the level increases by twenty-five percent under the stipulated example.

Those quantities are hypothetical and distinct from the official country table. They show why both the level and share of domestic content belong in an assessment. A policy targeting a share alone could discourage productive specialization without establishing a larger domestic-income contribution.

A producer can import inputs for domestic sales and exports during the same period. Deducting every import from the export total would assign all foreign inputs to the exported product, even when some serve local customers. Firm or industry accounts need an allocation consistent with the production being measured. Inventories create a further timing difference because inputs imported this year can be embodied in a later shipment. The sourcing account should track those uses before a net shipment value is described as locally created income.

Input-output estimates incorporate national accounts, trade data and balancing assumptions. The observation year can lag the publication date. A2026 release describing 2022 production links is not evidence that the same shares prevailed after every subsequent trade-policy or investment change.

Recent customs and firm data can provide additional information on shipments and sourcing. They should not be treated as a direct update of all historical input-output coefficients without a stated method. Revisions to national accounts and product classifications also require a consistent series rather than a splice of unrelated vintages.

An investment announcement adds another timing issue. Announced capacity does not establish completed production, and completed assembly capacity does not establish a deeper domestic supplier network. Evidence of new local sourcing, the supplier's own inputs and actual sales is required to identify that change.

For U.S.–Mexico analysis, this suggests a layered account: current gross transactions for the immediate trade environment, comparable value-added estimates for production origins, and firm or project evidence for subsequent changes. Each layer provides information the others do not independently supply. Their differences should be reconciled rather than concealed in one headline.

The common OECD evidence identifies materially different foreign-content and final-demand exposure in the two economies'2022 trade structures. It does not allocate every bilateral shipment or determine the welfare effect of integration. Mexico's national manufacturing indicator contributes additional domestic detail at a distinct methodological boundary.

The worked regional chain demonstrates why a U.S. import from Mexico can contain U.S. value, why national and North American content differ, and why a gross bilateral balance is not a complete measure of the division of production income. These are central features of integrated production rather than statistical exceptions.

A defensible assessment of value capture therefore identifies the origin of value added, the final demand sustaining it and the distribution of resulting income. It retains gross trade for the shipment questions it answers and adds operational evidence for resilience. This framework makes the economic consequences of U.S.–Mexico integration clearer without expecting one trade total to describe production, income and risk simultaneously.