With the prospect of far-reaching decoupling of goods trade between China and the US, fears of deflection of Chinese exports to third markets have returned. This column examines the factual record on the frequency of, and foreign government response to, Chinese trade deflection during the first US-China trade war. The authors find that trade deflection occurred 42% of the time in big ticket export items. Where deflection happened, importing governments took defensive measures in only half of cases. There is huge variation across importing nations in the scale of Chinese trade deflection and in their appetite for absorbing extra imports without taking defensive action. The authors relate these findings to the current scale of Chinese exports to the US and to the likely determinants of Chinese trade deflection to third markets going forward.
Decoupling of trade between China and the US has been much discussed over the past ten years. The tariff escalation witnessed in the days since the 2 April 2025 announcement by the US to impose 34% additional duties on Chinese goods makes far-reaching decoupling highly likely in the months and years ahead. In turn, this has raised concerns that Chinese products destined for the US market will be deflected to third markets, putting import-competing firms in the latter under pressure.
Already certain industry associations and policymakers are calling for the pre-emptive raising of import barriers against Chinese made and other foreign goods. If such defensive measures were implemented on a sufficient scale, then the current largely bilateral trade wars (between the US and its trading partners) would widen – increasing the risk of a 1930s-style implosion of the world trade system.
Trade deflection may sound like a technocratic footnote. In fact, properly identifying prior incidents of trade deflection and documenting the willingness of governments to absorb more Chinese imports may help calibrate current fears about trade deflection. Fortunately, the first US-China trade war provides a relevant precedent case to ‘scale’ trade deflection.
What is trade deflection? Must it happen?
The mental model of those afraid of deflection of Chinese goods exports to third markets is that the share of Chinese exports absorbed by other nations rises as Washington singles out Chinese goods for higher import barriers. Trade deflection is, therefore, the reaction of foreign exporters to a unilateral restriction on imports abroad – and should not be confused with trade diversion, which follows from the implementation of a regional trade agreement.
A prerequisite for trade deflection in the present context is that a given product exported from China faces a new import restriction in the US. We will exploit this requirement because some Chinese products were exempted or omitted from the tariff increases imposed by the US during the first trade war with China, which began in 2018.
A second observation is actually a hypothesis: Chinese exporters have stronger incentives to redirect products to third markets the larger the total value of export sales into the US before the trade war. On these grounds, we will identify the six-digit HS product codes where China exported more than £1 billion to the US in 2017. It turns out there are 76 such products, 16 of which did not face import tariff increases during the first trade war.
An important implication of these observations is that they caution against blanket assertions about the inevitability of broad-based Chinese trade deflection in the months ahead. As has become evident this past weekend – and as earlier US Executive Orders made plain – not every Chinese product faces the same tariff hike and so the incentive to deflect exports to third markets may be weaker for some products than for others.
Furthermore, the existence of such incentives does not guarantee that they were acted upon and that deflection occurred. We let official trade data reveal how often trade deflection actually occurred during the first US-China trade war.
Is all Chinese trade deflection bad?
To the extent that Chinese trade deflection results in more supplies being made available in third markets – possibly at lower prices — then in some cases it may be welcomed by an importing nation’s government. This may come to pass when there are few or no import-competing rivals to the deflected imports, where there are currently shortages, and where domestic producers have excessive pricing power.
Here we take no position on whether trade deflection is good or bad. We note, however, that from their public statements some appear to assume that trade deflection must be bad and so merits raising import barriers. This strikes us as jumping too quickly to a policy recommendation.
The related empirical question is: During the first US-China trade war, when trade deflection occurred, how often did the importing government take defensive measures? Those defensive measures could involve imposing tariffs or awarding corporate subsidies to import-competing firms. Entries on implemented commercial policy interventions from the Global Trade Alert database were used for this purpose.
Identifying trade deflection in granular import data
Our empirical analysis sought to identify the frequency of Chinese trade deflection after the first US-China trade war began and before the COVID-19 pandemic scrambled cross-border trade. Our approach detects trade deflection from the US market to other G20 members’ and Switzerland’s economies during 2019. This extends an earlier analysis we did of Chinese trade deflection to the EU published last November (Evenett and Martin 2024).
We used the finest-grained product data available in the BACI database to identify those product categories (six-digit HS codes) where China exported more than $1 billion in 2017, the year before the first trade war started. There were 76 such billion dollar-plus product categories. We focused on these products because there was a lot of trade that could be deflected in the first place.
For each of these 76 products, we calculated the cumulative average annual growth rate of nominal bilateral exports from China to each of the import destinations (the rest of the G20 and Switzerland) from 2012-2017. Call this the pre-war trend rate of import growth.
Because the nominal value of observed trade tends to rise due to price inflation unrelated to trade deflection, we do not conclude that such deflection has occurred every time we find higher recorded values of bilateral exports from China to a third market after the trade war started. 2 Instead, for a product, only when the growth rate of the value of Chinese exports to a third market during the trade war exceeds the pre-war trend growth rate is there a potential case of trade deflection. Insisting on a benchmark based on a pre-war growth rate also takes account of other natural reasons for the expansion of Chinese exports, such as exporter innovation.
Furthermore, we examine whether the annual growth rate of imports into a destination third market exceeded the pre-war trend growth rate by 1%, 2%, and 5%. Clearly, the higher this latter percentage, the higher the hurdle and the fewer the number of cases of trade deflection will be observed. For each destination market for Chinese goods, the number of cases of trade deflection and the number of cases on “no trade deflection” are recorded.
Treating the EU as one import destination, including Switzerland, means there are 16 third markets that Chinese exports can be potentially deflected towards. As there were 60 products where more than $1 billion of Chinese exports to the US existed in 2017 and where the US levied higher tariffs, then a total of 16×60, or 960, instances of trade deflection could be found.
With the 1% threshold of excess import growth, a total of 494 instances of Chinese trade deflection were found, just over half (51.4%) of the potential number of cases (960). With a 2% excess import growth threshold, the cases of trade deflection fell to 474. With a 5% excess import growth threshold, a total of 401 cases were found – implying that trade deflection occurred 41.8% of the time. Such findings imply that, as far as the rest of the G20 and Switzerland were concerned, the first China-US trade war did not see an immediate surge of Chinese exports to their markets across the board. A cautious estimate is that Chinese exports accelerated to these third markets just over 40% of the time.
Benchmarking trade deflection cases against other big ticket Chinese exports
Recall that there were 16 $1 billion-plus product lines shipped by China to the US in 2017 that did not subsequently face higher import tariffs during 2018 and 2019. There is no incentive for Chinese producers of these products to divert exports to third markets — and they serve as a benchmark.
Consequently, we can check whether the share of Chinese exports of these goods sent to Switzerland or to G20 members behaved differently from the changes observed in the products where trade deflection was possible. Surely, if Chinese trade deflection is pervasive, then the change in the shares of exports destined to third markets should rise more for goods where American tariffs are levied compared to cases where there is no incentive to divert exports?
Table 1 presents evidence on the shares of Chinese exports going to each import destination for all goods, for the 60 goods where US tariffs were increased, and for the 16 products where there is no incentive to divert exports.
Table 1 Shares of Chinese exports shipped to G20 members and Switzerland: for all goods and for big-ticket imports where trade deflection is possible and where it is not
As the last row of Table 1 shows, for those products where American tariffs were levied, the percentage of Chinese exports absorbed by the rest of the G20 rose from 43.4% in 2017 (pre-trade war) to 47.6% in 2019 and higher still to 52.3% in 2023. These findings reinforce concerns about Chinese trade deflection.
However, for the 16 products where there is no incentive to engage in trade deflection, the percentage of Chinese exports shipped to the rest of the G20 fell then rose — from 44.7% in 2017 to 44.2% in 2019 and up to 47.9% in 2019. Therefore, properly benchmarked, trade deflection of the order of 4.7 percentage points of total Chinese exports occurred from 2017 to 2019. Between 2017 and 2023, such Chinese trade deflection had grown further to 5.7 percentage points.
There was considerable variation across the rest of the G20 and Switzerland in the changes in their Chinese export shares in those products where there was an incentive for Chinese trade deflection. The EU stands out — the percentage of Chinese exports it absorbs rose 2 percentage points from 2017 to 2019, an increase greater than any other importer considered here.
Every other country saw their percentage of Chinese exports rise less than 0.6 between 2017 to 2019 — and in the case of Argentina and Türkiye, the percentage of Chinese exports they absorbed actually fell. The EU’s experience of trade deflection is the exception, not the rule.
What lessons then for the frequency and scale of Chinese trade deflection during the first trade war? Trade deflection occurred between 40% and 50% of the time. In the products where trade deflection was incentivised, the share of Chinese exports shipped to the rest of the G20 rose 4.2 percentage points. But the huge variation across the G20 in the changes in the percentage of Chinese exports absorbed cautions against overgeneralising about the extent of trade deflection (see the final columns of Table 1).
Frequency of Chinese trade deflection varies markedly across the G20 and Switzerland
Further evidence of the diversity of experience with Chinese trade deflection after the onset of the first trade war is borne out in Table 2. This table reports not only the number of instances of trade deflection detected by importer when the conservative 5% extra import growth threshold was used, but also the number of cases where deflection was followed by defensive measures.
Table 2 Instances of trade deflection in 2019 where the bilateral annual growth rate of imports from China were 5% above pre-trade war trend
Recall there are sixty products where there is an incentive for trade deflection to third markets. The second column of Table 2 reveals that the number of instances of trade deflection in 2019 varied from 5 (Türkiye) to 42 (Saudi Arabia). The average number of instances of Chinese import surges faced by these economies was 25 — implying that in more than half of the products (35) there was no Chinese trade deflection
Figure 1 represents this finding graphically — in only four G20 members was Chinese trade deflection found in more than half of the 60 big ticket Chinese export categories considered here.
Figure 1 Only four G20 members experienced trade deflection in 30+ products that China exported more than $1 billion to the US before the first China-US trade war
Revealed willingness to absorb Chinese imports
In just under half of cases of Chinese trade deflection did Switzerland and the rest of the G20 accept the additional imports arising from Chinese trade deflection without taking defensive action to support import-competing firms (see column 3 of Table 2). This proportion is high and presumably reflects a level of comfort or acquiescence to Chinese imports as to challenge any current claims that importing governments always move against trade deflection.
In 212 of the 401 cases where trade deflection was detected, importing governments took defensive measures (see column 4 of Table 2). When doing so, governments leant more heavily on import tariff increases than bailouts or subsidies to import-competing firms (see columns 5-7 of Table 2).
Moreover, there is little evidence of these importing nations singling out products where trade deflection occurred for more defensive measures. Comparing the findings in columns 4 and 10 of Table 2 reveals that the rest of G20 and Switzerland were as likely to take defensive measures against imports where Chinese trade deflection had not occurred then when it had (52.5% of instances compared to 52.8%).
In trade policy deliberation there are always cause celebres — examples of salient import-competing sectors that highlight every unwelcome increase in foreign competition no matter the reason. Evidence from the first US-China trade war reveals that importing governments did not single out for special protection local firms facing increased competition on account of Chinese trade deflection. For example, the EU took defensive measures in 91% of cases where Chinese trade deflection happened — and 95% of the time when it did not.
Still, as Figure 2 shows, there was significant variation across importing governments in both their willingness to take defensive measures in cases of trade deflection and in the mix of defensive measures taken. Argentina, India, and Indonesia relied more heavily on tariff increases — whereas the EU and UK resorted to a blend of subsidies and tariffs, for example.
Figure 2 By their actions, many G20 governments revealed they were prepared to absorb unimpeded many cases of Chinese export deflection from the US market
Going forward: The 2% challenge
Before closing, let’s scale the challenge facing the world trading system of absorbing Chinese exports originally destined for the United States.
China exported $439 billion to the US in 2024. The total value of goods imports by countries other than the US was close to $22 trillion in 2024, according to UNCTAD. Therefore, if all of China’s exports to the US were deflected to third markets, overall, buyers outside the US would have to absorb just 2% more imports by value.
For sure, there may be certain products and importers that might experience import growth from China in excess of 2%. But since the 2% statistic is an average, correspondingly, there must also be cases that would face less growth.
Furthermore, it is worth bearing in mind that UNCTAD estimated that during 2024 world goods trade expanded $500 billion. As $310 billion of that import growth occurred outside the US, it would take less than 17 months of current trade growth to absorb $439 billion.
Finally, by 2023, the total number of products for which more than $1 billion was exported from China to the US fell to 64, reducing the range of goods where there is plenty of trade to deflect in the first place (for the list of these products, see the table in the Annex of the Zeitgeist Series Briefing. All these considerations caution against exaggerating fears of Chinese exports swamping world markets in the near term.
A surgical, evidence-driven approach is needed
US-China decoupling poses systemic risks. “The biggest flashpoint lying ahead is the potential for Trump’s vertiginous tariffs on China…to provoke a flood of cheap Chinese products pouring into other markets,” contended the Financial Times over the weekend. Uppermost are fears of Chinese trade deflection, which are fanned with little evidence, despite the Trade War precedent.
In this column, we presented evidence that Chinese trade deflection occurred within 18 months of the first trade war starting, but it was patchy. Moreover, when deflection happened, in almost half of cases importing states didn’t take defensive steps against accelerating Chinese imports. By their actions, governments revealed that some Chinese trade deflection was desirable.
This is a moment for surgical, evidence-driven deliberation that weighs the costs and benefits of temporary defensive measures — not for sweeping import tariff hikes against imagined threats.
Source : VOXeu
