The redirection of Chinese exports away from the US has fuelled concerns about its spillovers to other countries. This column shows why trade-basket similarity can be a misleading measure of exposure and introduces four simple, theory-consistent indices: export competition, import competition, consumer gains, and sourcing gains. They produce different country rankings and reveal how competitive losses can coexist with gains from cheaper goods and inputs. Thus, the response to spillovers from the surge in Chinese exports should be guided by where the shock actually lands, not by which countries look most like China.

When two giants fight, smaller economies are not flies on the wall. China’s share of US imports fell from 22% in 2017 to about 8% in 2026, a level not recorded since China joined the World Trade Organisation (WTO). Over the same period, the share of Chinese exports going to the US halved, from around 20% to about 10%. Given the scale of Chinese exports, this contraction and associated reorientation have substantial implications for other economies.

Indeed, since ‘Liberation Day’ in April 2025, concerns that Chinese goods blocked by US tariffs will be redirected to other markets have intensified. The European Commission set up a dedicated import-surveillance tool to monitor potentially harmful trade diversion. In parts of Asia, concerns about displacement and import surges led to new or extended actions against Chinese products. Yet the implications are not one-sided. Diversion towards the EU appears limited to a relatively small tail of highly exposed products (Schulte et al. 2026). At the same time, cheaper Chinese goods may lower euro area inflation (Corsello et al. 2025) while hurting producers who compete with China at home and abroad (Aprigliano et al. 2026). 

The key question is therefore which countries are most exposed, and how. A common answer relies on how closely a country’s trade basket resembles China’s. The Finger–Kreinin (1979) export similarity index has been used to document China’s growing resemblance to advanced economies, particularly the euro area (de Soyres et al. 2025). ECB staff have used it to identify which exporters could replace Chinese products in the US market (Gunnella et al. 2025). But resemblance does not tell us how a shock affects producers and buyers. In a recent paper (Antràs et al. 2026), we develop four indices to measure these effects through export competition, import competition, consumer gains, and sourcing gains. We show why trade-basket similarity can be a misleading guide and provide a more informative assessment of exposure.

Four indices of exposure

We focus on how the redirection of Chinese exports affects third countries in markets outside the US, distinguishing four conceptually distinct, and not mutually exclusive, margins:

• Export competition: China redirects sales from the US to third destinations and reduces other exporters’ market shares there.
• Import competition: redirected Chinese goods increase import penetration and competitive pressure in domestic markets.
• Consumer gains: cheaper imports from China reduce prices for consumers in destination markets.
• Sourcing gains: cheaper Chinese intermediate inputs reduce costs and strengthen the competitiveness of downstream firms.

The intuition behind the indices is straightforward. When Chinese exporters can no longer sell as much to the US, they must cut prices elsewhere. The larger the share of Chinese output previously sold in the US, the more exports must be redirected, and the larger the price decline in other markets.

How much a country feels the shock depends on two further factors: how important China already is in the relevant markets, and how much those products matter to the country. A fall in Chinese prices puts more pressure on countries where China is already a close competitor. The four indices trace how the same fall in Chinese prices affects a country’s exporters, domestic producers, consumers, and firms using Chinese inputs. Crucially, the importance of each product differs across these groups. Cheaper imports pose little competitive threat if a country does not produce the goods concerned, but they can still benefit buyers. Similarly, gains from cheaper inputs depend on whether domestic firms use those inputs in production.

The shock lines up with the data. Products for which the US was initially a more important market for Chinese exporters later saw faster growth in Chinese exports to other destinations. They also saw sharper falls in Chinese exports to the US. 

Similarity is not exposure

Similarity describes how closely two trade baskets resemble each other, not how a Chinese supply shock affects an economy. First, similarity indices give no role to the size of the shock for each product. Second, they do not trace how that shock reaches a country’s producers and buyers: how important China is in the affected markets, how readily buyers switch suppliers, and how much those products matter to the economy. Their symmetry illustrates this distinction: China is as impacted by competition from Honduras as Honduras is impacted by competition from China, but the competitive pressures they exert on one another need not be equal. Finally, similarity alone cannot distinguish losses for exporters and domestic producers from gains for consumers and firms using cheaper inputs. 

These differences matter in practice. Take Germany. On export similarity, it ranks sixth among more than 100 economies: a textbook case of a country that ‘looks like China’ and should therefore be highly exposed. Yet on our export competition index, Germany does not make the top 20. The same is true of Italy, France, and the UK. Our index depends on where a country sells and how strong Chinese competition is in those markets. It is high for economies specialised in product–destination markets where China is already a dominant supplier, even if their overall export structure does not resemble China’s. The most exposed include Vietnam, Cambodia, and Bangladesh, as well as Central European manufacturers such as Hungary, Czechia, and Poland. Across countries, the correlation between the two measures is only 0.49 (Figure 1). 

Figure 1 Export competition versus export similarity with China

Notes: Each dot is an economy. The export competition index is the predicted percentage decline in exports caused by the redirection of Chinese exports. For economies that serve as transshipment hubs, measured bilateral trade flows — and therefore their estimated exposure and ranking — may partly reflect rerouting rather than domestic production. The export similarity index is the Finger–Kreinin index relative to China, calculated across products and destinations. Source: Antràs et al. (2026).

On the import side, the key distinction is between exposure to cheaper Chinese goods and competitive pressure on domestic producers. Import similarity cannot distinguish between the two. Our import competition index instead captures whether the affected products matter for domestic production: even a large import surge poses little competitive threat if a country does not produce those goods.

Losses and gains can coexist

Competitive pressure and gains from cheaper Chinese goods are distinct dimensions of exposure. Countries facing stronger import competition are not necessarily those benefiting most from lower import prices (Figure 2a). Similarly, cheaper Chinese inputs can hurt domestic input producers while reducing costs for downstream firms, with the strength of these effects differing across economies (Figure 2b). We do not aggregate the four indices into a single measure of net exposure. They capture changes in different economic outcomes, so their magnitudes cannot simply be added or subtracted to establish whether a country gains or loses overall. Instead, they provide a multidimensional assessment of where competitive pressures and gains arise within each economy.

Figure 2 Competitive pressures and gains from cheaper Chinese goods

(a) Consumer gains versus import competition

(b) Sourcing gains versus import competition in intermediate inputs

Notes: Each dot represents an economy. Panel (a) compares the consumer gains index, measuring the predicted percentage decline in import prices, with the import competition index, measuring the predicted percentage decline in domestic sales. Panel (b) compares the sourcing gains index, measuring the predicted percentage increase in exports from cheaper Chinese inputs, with the import competition index restricted to intermediate inputs. For economies that serve as transshipment hubs, measured bilateral trade flows — and therefore estimated exposure — may partly reflect rerouting rather than domestic production or final demand. Source: Antràs et al. (2026).

The magnitudes are economically meaningful. For several of the world’s 20 largest economies, the export competition index predicts export declines of 0.4% to 0.9%. Among the most exposed economies, import competition implies falls in domestic sales of more than 3%. For some countries, consumer and sourcing gains each exceed 1%.

Country-level exposure can also conceal striking product-level concentration. Mobile phones alone account for almost a third of India’s import competition exposure and a fifth of South Korea’s. In Vietnam, mobile phones account for about a third of export competition exposure but only 14% of import competition exposure: the same product matters very differently across channels. Mobile phones also account for about a third of Paraguay’s consumer gains index, illustrating how a product associated with competitive pressure elsewhere can be an important source of gains for buyers. All four exposure indices, their alternative variants, and country-by-product breakdowns are freely available on the accompanying .

A portable diagnostic

We implement the framework using the product-level shock created by the US–China tariff conflict, but its logic goes beyond this episode. The key input is a product-level change in Chinese export prices, separate from the country-specific weights that determine who bears it. The same weights can therefore be combined with price or supply shocks from other sources, such as Chinese industrial policy, productivity growth, exchange-rate movements, or weak Chinese domestic demand. In these applications, the economies exposed need not be those caught in the middle of the US–China conflict. More broadly, the approach offers a portable set of diagnostics for identifying where a supply expansion by a large exporter creates competitive pressure and where it generates gains.

The policy implication is that the response to spillovers from the surge in Chinese exports should be guided by where the shock actually lands, not by which countries look most like China.

Source : VOXeu