Beyond trade diversion: How the US-China trade war reshaped global production

Trade wars do not simply redirect exports; they also reshape the costs of production, disrupt supplier-customer relationships, and trigger adjustments within multinational firms’ global affiliate networks. This column uses data covering millions of manufacturing plants in 50 major economies to show that the US-China trade war reshaped manufacturing activity far beyond the two directly targeted economies, but not in one uniform direction. Industries in the intermediate stages of global value chains were squeezed from two sides: higher costs when they relied on US or Chinese inputs, and weaker demand from customers affected by the trade war. The most upstream and most downstream industries, by contrast, tended to gain on average.

As governments once again turn to tariffs to reduce strategic dependencies and shift production towards politically reliable partners, a familiar question has returned: where does production actually go when large economies raise trade barriers against one another?

The most common answer is that it moves to third countries. That answer is often right, but it captures only part of the adjustment. Trade wars do not simply redirect exports from one country to another. They also reshape the costs of production, disrupt supplier-customer relationships, and trigger adjustments within multinational firms’ global affiliate networks. 

The influential analysis by Fajgelbaum et al. (2023) showed that the first US-China trade war generated highly uneven effects among ‘bystander’ countries. Some countries expanded their exports in products targeted by tariffs, while others experienced export declines. This heterogeneity is a crucial starting point: the relevant question is not whether third countries benefit from a trade war, but which firms, industries, and countries benefit, and through which channels.

Trade diversion created visible winners

Several recent studies document important third-country gains from trade diversion. China’s retaliatory tariffs on US goods encouraged Chinese buyers to source substitutes from elsewhere. In Brazil, regions specialised in products targeted by Chinese retaliation experienced employment and wage-bill gains (Cavalcanti et al. 2025).

The effect worked in the other direction, too. US tariffs on Chinese goods created opportunities for Mexican firms to expand exports to the US (Utar et al. 2023). Vietnam, meanwhile, appears to have increased domestic value added in some strategic exports to the US, rather than merely serving as a transit point for Chinese goods (Schulze and Xin 2026).

These cases make trade diversion tangible. When US tariffs make Chinese products more expensive, firms in Mexico, Vietnam, or elsewhere may replace Chinese exporters in the US market

Yet the same Mexican evidence also points to the limits of a simple ‘third countries win’ narrative. Chinese retaliation had negative effects on some Mexican firms, and firms dependent on tariff-targeted Chinese inputs were adversely affected. The global consequences of a tariff therefore cannot be inferred from the fate of one export market alone.

Trade wars also disrupt production networks

The reason is that modern production is organised through global value chains. Goods cross borders repeatedly before reaching final consumers, and many products are simultaneously an industry’s output and another industry’s input.

Consider a car exported from the US to China. It may embody German engines and transmissions, Japanese electronics, and specialised components from Mexico and other European countries. If Chinese tariffs reduce US car sales in China, the impact does not stop with the US car producer. Demand can also fall for the third-country suppliers that provide complementary inputs and services.

Existing research has demonstrated these indirect effects. Mao and Görg (2020) show that the US-China tariff increases raised the cumulative tariff burden faced by downstream trade partners through supply-chain linkages. The effects were particularly important for US tariffs on Chinese inputs that were used in US production and then re-exported to third countries. 

These studies show why a bilateral tariff can have multilateral consequences. But they largely measure trade flows, tariff exposure, or input-output relationships. They cannot directly observe where production, employment, and establishment activity changed, or which firms were able to reorganise their operations in response.

A worldwide view of production and multinational adjustment

In our recent paper (Fadinger et al. 2026), we address this gap using establishment-level data covering millions of manufacturing plants in 50 major economies, including affiliates of more than 200,000 multinational enterprises. We combine this information with detailed data on US and Chinese trade-war tariffs on outputs and inputs.

Our data allow us to observe changes in sales, employment, and the number of establishments in third countries, rather than only changes in exports. They also allow us to distinguish multinational affiliates from domestic firms and to trace how activity moved across multinational affiliate networks.

We find that the US-China trade war reshaped manufacturing activity far beyond the two directly targeted economies, but not in one uniform direction.

First, US tariffs on Chinese final goods increased production in third countries. This result is consistent with trade diversion: Chinese products became less competitive in the US market, creating opportunities for producers located elsewhere. It is also consistent with multinational firms shifting production away from China and using third-country affiliates as export platforms.

Second, Chinese tariffs on US final goods reduced third-country manufacturing activity. This may seem counterintuitive if one expects all third countries to benefit from substitution away from US exporters. But many third-country firms are linked to US producers through supply relationships and complementary demand. When US exports to China decline, so can demand for the inputs and services supplied by firms in third countries.

Third, tariffs on intermediate inputs matter at least as much as tariffs on final goods. US tariffs on Chinese inputs reduced activity in third countries, consistent with higher input costs weakening the competitiveness of US producers and reducing demand for complementary foreign production. A tariff intended to protect an upstream domestic industry may therefore impose costs on downstream domestic firms and on foreign firms connected to the same production network. By contrast, Chinese tariffs charged on inputs sourced from the US benefited manufacturers in third countries: as Chinese plants became less competitive due to more expensive specialized intermediate inputs, third country suppliers replaced them as exporters to the U.S and in third markets. 

Figure 1 brings the four tariff channels together. It shows their predicted effects on sales separately for multinational establishments (red diamonds), domestic establishments (yellow squares), and all establishments (blue circles) in third countries. The effects differ sharply by tariff direction and by whether the tariff applies to final goods or inputs. The sales response is most pronounced for multinational establishments, which is consistent with these firms using their cross-border affiliate networks to shift activity across locations. The overall effect masks these offsetting forces.

Figure 1 Predicted effects of the four US-China trade-war tariff channels on third-country sales

Multinationals cushion shocks, but not everyone benefits

The adjustment was driven primarily by multinationals. Firms with affiliates in several countries could reallocate production, sourcing, and sales across locations thereby avoiding trade-war tariffs on their outputs and inputs; domestic firms responded much less. Chinese multinationals, in particular, very actively shifted activity and affiliate links towards third countries. 

This flexibility can cushion the immediate impact of tariffs for multinational firms. But it also means that the gains from trade diversion need not be broadly shared. In our data, the clearest losses occurred in intermediate stages of global value chains. These industries were squeezed from two sides: higher costs when they relied on US or Chinese inputs, and weaker demand from customers affected by the trade war. The most upstream and most downstream industries, by contrast, tended to gain on average.

The regional pattern is also uneven. Third countries in Asia and North America were, on average, more negatively exposed, while Europe was broadly insulated on average. But these averages conceal substantial variation across industries and firms. A European supplier deeply integrated into a US or Chinese production network can be highly exposed even when the regional average is modest.

Figure 2 shows the same four tariff channels separately for North America excluding the United States, Asia excluding China, Europe, and the rest of the world. The regional panels underline that there is no single third-country effect of the trade war. The balance between trade diversion, higher input costs, and disrupted supplier-customer relationships differs markedly across regions. Asia and North America were, on average, more negatively exposed, while Europe was broadly insulated on average; each regional average nevertheless contains substantial industry-level variation.

Figure 2 Predicted effects of the four US-China trade-war tariff channels on third-country sales, by region

Resilience requires looking beyond bilateral trade

The policy lesson is not that countries should withdraw from global value chains. International production networks can create flexibility and help firms absorb shocks. But resilience requires understanding the full network of dependencies, not merely direct imports from a particular country.

For firms, this means mapping exposure across suppliers, customers, and foreign affiliates; diversifying critical sources of inputs and export markets; and retaining the capacity to shift production when disruptions occur. For policymakers, it means evaluating tariffs not only through their bilateral effects, but also through the costs they impose on downstream users and foreign partners.

Trade wars do not simply bring production home. They reorganise global production through substitution, input costs, complementarities, and multinational networks. The visible winners from trade diversion are real. But so are the less visible losses that arise when tariffs disrupt the production relationships on which firms across the world depend.

Source : VOXeu