US-China trade war: which sectors are most vulnerable in the global value chain?
- What looked to be a relatively quiet summer on the US-China trade war front, has taken a new turn after the US announcement to up the ante at a later stage
- In this report, we focus especially on which segments of the Chinese and American economies are most vulnerable to a disruption of global supply chains caused by the US-China trade war
- We also assess the impact for the Netherlands as an example of a ‘third country’ where several industries will be harmed (indirectly) by these disruptions
- Roughly 50% of total gross US imports and exports consists of intermediate goods; this figure is even higher for China, for which 70% of gross imports and 62% of gross exports are intermediates
- Our analysis shows that despite the bigger negative macro impact on China, the US is more exposed to disruptions to bilateral intermediate trade flows than China
- For the US, computers and electronic products, electronic equipment and textiles/shoes are the most vulnerable segments, at risk of both direct and indirect effects
- For China, the wood and cork products industry is the most vulnerable to a disruption of supply chains caused by higher tariffs. Transport equipment, paper products and pork production also have some vulnerabilities
- The Dutch companies that are the most vulnerable in the supply chain between China and US are basic metals producers, manufacturers of computers and electronic products, paper products, and machinery products
- Consumers in the Netherlands are likely to see impact on prices in the coming period on computers and electronic products (such as smartphones), and electrical equipment (such as medical equipment) as the higher tariffs feed through in higher retail prices
Also available: a compact version in Dutch.
The US-China trade war: no turning back?
What first looked like a relatively quiet summer, with the prospect of new negotiations between the US and China in September, took a new turn in early August. On the evening of August 1 (European time), US president Trump announced by tweet the decision to further up the ante in the long-running trade dispute with China with another set of tariff hikes effective September 1. In the meantime, this has already been postponed to December 15, 2019, at least for some goods. The main reason for the decision was that Mr. Trump was not satisfied with how China had lived up to the agreements made during the late June G20 summit in Japan. These mainly concerned the limited purchases of agricultural products by China, and that China seemingly still exports fentanyl to the US. As a result, next to the 25% tariff on USD 250bn of imports from China already imposed, the US is about to increase tariffs by 10% on imports that were not yet targeted (see also Giesbergen et. al, 2019).
We have been flagging the risk of further acceleration of the trade war for quite a while. After the G20 meeting, we foresaw two major triggers that could lead to a renewed escalation in the period ahead: the situation regarding Huawei and (too) limited purchases of American agricultural products by China. But, in broader terms, we have also been arguing that it would be very difficult to reach a long-lasting and comprehensive deal on all the different topics that have been put on the table. Indeed, we felt that the best outcome would probably be a can-kicking exercise while staying in a stable-unstable equilibrium. Next to the different views both sides may have on trade and investments, it is simply impossible for the US and China to bridge the gap on structural issues that the US wants to see resolved, such as violation of intellectual property rights (IPR), forced technology transfer and exorbitant subsidies by the Chinese government to Chinese state-owned enterprises (SOEs). China is unable and unwilling to make commitments to alter these things. The question therefore was not if tensions would flare up again, but when. Well, that time has come, to put it mildly.
In this report, we focus especially on which segments of the Chinese and American economies are most vulnerable to a disruption of global supply chains caused by the bilateral trade war. In addition, we assess the impact for the Netherlands as an example of a ‘third country’ which might be harmed indirectly by these disruptions. Dutch exporters might be trapped in the US-China supply chain, whereas Dutch consumers could feel the pinch when goods imported from either China or the US become more expensive due to the tariffs.
Impact of the US-China trade war
In November 2018, we conducted an extensive scenario study in which we analyzed the macro economic impact of the current trade conflict under the different protectionist measures and in case of a full escalation. The results showed that the overall macroeconomic impact of the trade war is much larger on the Chinese economy than on the US. These effects are partly the result of the asymmetric trade relationship between both economic superpowers, where the US has the upper hand to target larger export volumes from China to US shores (USD 540bn in 2018) than the other way around (USD 120bn). In the full escalation scenario we assumed that both countries would target their bilateral imports in full by higher tariffs.
We also anticipated a weaker Chinese currency (CNY) against the USD in case the Trump administration announced another round of tariffs on all Chinese exports, given that China’s options to strike back with equal measures are far more limited. Given that the USD/CNY currency pair breached the psychological level of 7 recently, we have now decided to switch to our full escalation scenario. Not only do we have the prospect of renewed hikes in tariffs and a ‘signal’ by China that it is willing to use its currency as a means to mitigate or offset the impact of tariffs on its exporters, but we also have increasingly belligerent language on both sides that suggests that it will become increasingly difficult to reverse course in this trade dispute. Another important factor related to the disproportionate impact on China on a macro level is the fact that China is much more dependent on US technology than the other way around. These factors together result in disproportionate negative effects for China.
In contrast to the results of our November 2018 study, we now zoom in on disruptions to global supply chains where the US is more exposed than China. This is one of the caveats of our previous analysis which we want to research, next to potential effects from international firms offshoring their production from China.
Caveats previous scenario study
Firms stepping up offshoring from China
Our 2018 scenario study has the upside that we combine the direct impact on trade (using an econometric trade model) with dynamic productivity effects (using two productivity side models) for China and the US. Nevertheless, the study also has a number of caveats. First, many foreign firms are planning to step up their effort to shift production lines out of China. Examples are Apple, toymaker Hashbro, Nintendo and Samsung. Some of these firms already had plans to shift production to other Asian low-wage countries, such as the Philippines, India and Vietnam, given rising labor costs in China. However, uncertainty caused by the trade war is putting this offshoring trend in overdrive. An exodus of foreign activity from China will not only hurt Chinese production and employment, but will in the long run also reduce China’s capacity to benefit from foreign knowledge spillovers (Wang and Wu, 2016). See Hayat (2019) for an extensive coverage of RaboResearch on this topic.
Negative impact of global supply chain disruption
A second caveat of our previous scenario study is the fact that it does not include the negative impact of the trade war via a disruption of global value chains. Over the past decades, multinational firms have increasingly been exploiting international comparative advantages by relocating parts of their production processes abroad (see OECD (2013a), Baldwin and Lopez-Gonzalez (2015) and Auer, Borio and Filardo (2017)). In this way, foreign firms in more advanced economies can, for example, benefit from relatively low labor costs in emerging and developing Asia for the assembly of goods, while marketing and R&D are located at the home base, which enables them to produce more efficiently and improve their competitiveness. The additional effect of these ‘sliced-up value’ chains is that multinationals have become more reliant on intermediate products or commodities from abroad and are therefore also more vulnerable to certain price shocks caused by protectionist trade measures, such as import tariffs.
Nowadays, roughly 50% of total gross US imports and exports consists of intermediate goods; this figure is even higher for China, for which 70% of gross imports and 62% of gross exports are intermediates. These figures highlight the importance of taking global value chain integration into account when analyzing potential trade war effects. Therefore, in this report we aim to map the vulnerabilities for different industries caused by a disruption of global supply chains by the US-China trade war.
Global value chain integration and protectionism: the mechanisms
Higher import tariffs on intermediates can directly and indirectly hurt companies in countries that have imposed tariffs. Amiti, Redding and Weinstein (2019) show that especially domestic consumers are bearing the brunt of the higher US tariffs implemented in 2018 and that the tariffs have already altered the supply networks of US firms. Gawande, Hoekman and Cui (2015) conclude that the increasing fragmentation of production across global value chains (initially) may have prevented countries from raising trade barriers and inducing protectionist measures in the aftermath of the global financial crisis of 2008. The complex relations between the different fragments within global value chains imply that any disruption in these supply chains generates amplified effects throughout industries and economies on a higher scale. Frohm and Gunella (2017) estimate that a 1% change in economic activity in an industry’s global value chain translates into 0.3ppts impact on activity of the sector.
Ultimately, firms confronted with higher costs for certain inputs will start to look for alternative suppliers who are not subject to higher tariffs. However, a supply shift cannot be organized overnight and involves transaction costs since other suppliers generally cannot meet proper requirements and know-how straightaway. Furthermore, if one compares the integration of global value chains nowadays to decades ago, there has been a dramatic change. The level of differentiation in value chains, which can be expressed by looking at trade in value added sub-components compared to gross trade, has clearly been on the rise (Blanchard, 2019). In terms of the trade war impact on integrated value chains, we make a distinction between direct and indirect impact for the countries involved, as well as impact on third countries that are not directly part of the trade conflict.
The direct impact is felt by domestic producers who immediately face higher costs for foreign components needed in their production processes (Figure 1). For example, auto manufacturer Tesla announced that the higher tariffs on Chinese parts pushed up costs for the company by USD 50 million in Q4 2018. The higher cost of intermediates means that these companies will face either a deterioration in competitiveness due to higher retail prices (as well as higher export prices and a lower global market share) or an absorption of the higher costs, which will hurt their profitability. Another example related to the trade war is Huawei components. In May, the US already blacklisted Huawei. As a result, key US suppliers (such as Xilinx, Broadcom and Google) are unable to sell components to the Chinese telecom giant. Although the ban on Huawei mainly has a negative impact for Huawei itself, US and foreign (e.g. NXP) suppliers that operate in the Huawei supply chain are affected as well.
China, however, also has some leverage over the US in this respect: namely rare earth metals or rare earth elements (REE). These metals are used for applications in the high-tech industry, for instance in mobile phones, lasers and defense material. China provided 80% of these rare earths to the US between 2014 and 2017 (see USGC). Although some studies expect that China’s dominance in this field will decrease in the longer term (e.g. Habib et al., 2016), in the short term the US high-tech industry remains vulnerable and an export restriction on REE by China is an important chip on the bargaining table. But it should be noted that president Trump at the same time already seeks to increase domestic production of these metals.
The indirect impact is related to domestic producers and consumers who face higher costs or prices due to imported final goods that have been exposed to higher tariffs. In this case an American company that has outsourced parts of its assembly to China and then reimports these as goods for its final processing or for direct sale to its US customers is exposed to higher tariffs twice (see Figure 1). This is for instance the case with Apple, which ships components to China for the assembly of its products and reimports the final products for distribution on the US and the global market. The indirect impact on Chinese firms is most likely significantly lower, simply because US firms use China as a manufacturing hub, and not the other way around. The more an individual company is exposed to an indirect impact of higher tariffs, the more likely it will ultimately choose to move assembly lines out of China and seek alternative locations (Hayat, 2019).
Indirect impact on third countries: the case of the Netherlands
Exporters and consumers in third countries might also feel the pinch from a disruption in supply chains between China and the US. Exporters in third countries that are not directly involved in the trade conflict may be caught up in the supply chain between the US and China and face price increases on their intermediates. Take, for instance, a Dutch supplier for the semiconductors industry that ships its intermediates to China. In China these parts are used in the production of mobile phones and then shipped as final goods to the US (see Figure 2). Consumers in third countries will also bear the brunt of disrupted supply chains: think of more expensive electronic products, such as laptops and computers that use more expensive Chinese intermediates and are shipped from the US to European shores (see Figure 3). Or the other way around: Huawei customers in Europe that will be unable to receive updates of their operating systems.
Extent of value chain integration between the US and China
The central question in this report is which industries and related product groups of the US and Chinese economies are most susceptible to higher bilateral trade barriers which have a disruptive impact on supply chains. We already presented an overview of gross trade flows and related product groups in our previous studies. It should be noted that not all these groups have been targeted with higher tariffs yet. Besides, gross trade flows do not reflect the importance of internationally sliced-up value chains. That is why an assessment of the dependency on particular intermediate goods is also essential. Below, we discuss the vulnerabilities for the individual countries separately.
We map two dependencies to determine to what extent value chains between US industries are exposed to a disruption of the supply with China. First, we calculate the share of Chinese intermediates used by US industries in their production (direct impact, vertical axis in Figure 4). Second, we assess what share of US value added is incorporated in gross China export to US shores (indirect impact, x-axis in Figure 4). Finally, the size of the circles represents the share of industries in total value added, to illustrate the relative importance for the economy. Ultimately, the more remote an industry is from the origin, the more dependent this US industry is on Chinese intermediates and the more exposed the industry is to price hikes caused by higher import tariffs.
Figure 4 shows that computers and electronic products, electronic equipment and textiles and shoes are the most vulnerable US sectors in the current trade war. These product groups are most likely to be subject to price increases due to more expensive intermediates: more than 20% of US production in these industries consists of Chinese intermediates. For the computers and electronic products category, the dependency is related to the supply of Chinese rare earth elements, mentioned earlier in this report. As a result, from a business perspective it is not surprising that, for example, footwear companies such as Nike and Adidas are warning about the potential effect higher tariffs will have on the US economy and consumers.
The computers and electronic products industry is also indirectly vulnerable to higher tariffs: 7.5% of Chinese exports consists of US value added incorporated in these products. This is not surprising, as US manufacturers use China as a manufacturing hub for assembling laptops, mobiles, etc. This vulnerability is consistent with findings by Lovely and Liang (2018) who conclude that 87% of the tariff hikes on computers and electronic products (as part of the US protectionist package on USD 50bn of Chinese export goods in 2018) are borne by foreign firms.
To determine to what extent value chains between Chinese industries are exposed to a disruption of the supply with the US, we also map two dependencies in a similar way as we measured the US vulnerability before. When we shift our focus to China, the wood and cork products industry is the most vulnerable to a disruption of supply chains caused by higher tariffs: 7% of Chinese production in this industry depends on US intermediates and 1.5% of US exports shipped to the China in this category consists of Chinese value added. These results are in line with recent observation by Bloomberg that there is increasing interest of the furniture industry operating in China to shift production lines to Vietnam. It is no coincidence that Vietnam is among the highest ranked countries on our Where Will They Go index (Hayat, 2019).
Other industries that show some vulnerability are transport equipment and paper and printing products. These industries are relatively less dependent on US intermediates, but there is a relatively large share of Chinese value added incorporated in the US export goods from these sectors which are shipped to China. Think of cars, planes and helicopters. Finally, there are vulnerabilities in the agricultural sector, which mostly relate to the import of US soybean for the Chinese pork industry, which our colleagues from Food & Agri Research (FAR) have repeatedly reported (see for example here). A final industry worth mentioning is computers and electronic products. This category for example relates to Huawei, which is dependent on US integrated circuits and operating systems. Although the dependency of this product category is not large in terms of share in total production (1.8%), US intermediates are currently vital for Chinese high-tech manufacturers. Although China is making efforts to decrease reliability on foreign components and become self-sufficient, Chinese chip manufacturers are still years behind the technological leaders. This is in line with our previous scenario study, which also mentions the asymmetric relation between the US and China in terms of technological spillovers.
The results shown in Figures 4 and 5 (see also Appendix B) indicate that the US is more exposed than China to disruptions in the US-China supply chain. US manufacturers are much more reliant on Chinese intermediates than the other way around. US firms are especially vulnerable in the computers and electronic products categories since there is also a major part of Chinese export shipped to US shores that consists of US value added incorporated in these products. Against the backdrop of these vulnerabilities, it makes sense that US companies are putting relocation of their production lines out of China into overdrive (Hayat, 2019).
One aspect that is not properly considered in this analysis is the qualitative dependency within the supply chain. As mentioned, the dependency of the Chinese high-tech industry on US know-how incorporated in US intermediates may not be large in terms of the share in total production (1.8%), but these components are still vital to this industry and it could take decades for China to become self-sufficient in these areas. In the end, China is much more reliant on US technological knowledge in order to propel productivity growth than the other way around. Another caveat is related to foreign exchange developments. This study is based on the assumption of static FX developments but changes can of course affect bilateral trade flows. Take, for example, the weakening of the CNY against the USD since the most recent trade war escalation (begin August).
Potential impact on ‘third countries’: the case of the Netherlands
In order to assess the impact of a disruption of supply chains on ‘third countries’, we use the Netherlands as an example. As said, there are two conduits by which the Dutch economy could be indirectly affected. First, Dutch exporters might be trapped in the US-China supply chain, while Dutch consumers could feel the pinch when goods imported from either China or the US become more expensive due to the tariffs. Given that now more US consumer products will be targeted by higher tariffs, the chances for this have risen. See Figures 2 and 3 for an illustration of the potential impacts on Dutch consumers and producers. For consumers, this mainly relates to higher prices for imported goods, which also holds for producers together with a potential loss of market share in these countries. For a more detailed description of our methodology and results for the Netherlands, please see also Appendix A and B of this report.
Figure 6 shows the Dutch export sectors which are most active in the supply chain between China and US. The vulnerabilities are especially present for manufacturers of computers and electronic products, paper products, chemicals/pharmaceuticals, electrical equipment, machinery and basic metals. The impact, however, should not be overstated, as in general we are talking about a small proportion of total value added in these industries: between 0.2 and 0.3%. Then again, for individual multinational Dutch companies the impact could be quite substantial and some of them, such as Philips, among several other multinational operating companies who are trapped in sliced-up value chains, are taking precautionary measures to mitigate the negative effects of the tariffs. Philips reportedly expects the trade war will shave off USD 69mln from core profits this year.
Figure 7 shows which product groups are prone to price hikes, for example for Dutch consumers. Particularly computers and electronic products -for instance Apple products, but also electronic medical products and instruments- shipped from the US to the Netherlands will be more expensive in the upcoming period, as higher tariffs feed through into higher retail prices. In addition, the Dutch consumer market might see higher prices for furniture shipped from China to the Netherlands. But the effect on the latter is expected to be lower compared to the first two categories if one takes into account the figures presented in table B.4 of the appendix.
Conclusions and what’s next?
In an extensive scenario study in November 2018 we analyzed the macroeconomic impact of the current US-China trade conflict. These results showed that the trade war impact on a macro level is much larger on the Chinese economy than on the US. In contrast to the results of this scenario study—which thus showed disproportionately negative effects for China—we now zoom in on disruptions to global supply chains, to which the US is more exposed than China. This was one of the caveats of our previous analysis which we wanted to research.
Our latest results indicate that the US is more exposed than China to disruptions in US-China integrated supply chains. US manufacturers are much more reliant on Chinese intermediates than the other way around. US firms are especially vulnerable in the computers- and electronic products categories since there is also a major part of Chinese export shipped to US shores that consists of US value added incorporated in these products. Against the backdrop of these vulnerabilities, it makes sense that US companies are putting relocation of their production lines out of China into overdrive (Hayat, 2019). For China, the wood and cork industry is most vulnerable to a disruption of supply chains caused by higher tariffs. Transport equipment, paper and printing products and pork production also have some vulnerabilities.
One aspect that we are unable to examine in this analysis is the qualitative dependency within the supply chain. As mentioned, the dependency of the Chinese high-tech industry on US know-how incorporated in US intermediates may not be large in terms of the share in total production (1.8%), but these components are still vital to this industry and it will take decades before China may be self-sufficient in these areas. In the end, China is much more reliant on US technological knowledge in order to propel productivity growth than the other way around, which is one of the key reasons why the Chinese economy disproportionally bears the brunt in case of full trade war escalation between both countries.
Our analysis furthermore shows that Dutch companies that are the most vulnerable in the supply chain between China and US are producers of basic metals, manufacturers of computers and electronic products, paper and printing products, and machinery. The impact, however, should not be overstated, as this concerns a small proportion of total value added in these industries: between 0.2 and 0.3%. Nevertheless, individual multinational companies, including Dutch ones, could feel the pinch of the trade war, especially if they are active in both countries. Finally, from an import perspective, Dutch consumers can expect price hikes of computers, electronics, medical products and instruments if the trade war intensifies further.
After the most recent announcement by the US to further up the ante by increasing tariffs by 10% on imports that were not yet targeted and China’s subsequent retaliatory policies, it seems highly likely that we need to brace ourselves for a period of prolonged tension between both countries (see also Giesbergen et. al, 2019). Therefore, this study provides valuable insights on the trade war impact from the perspective of integrated-value-chain analysis.
 Under the scenario of the protectionist packages up and until August 2019, we find a cumulative loss in GDP growth for China of 1.5ppts up to 2030 compared to a benchmark scenario without a trade war and even 5.7ppts in case of a full escalation. For the US, the estimated impact would be much smaller, with an estimated cumulative varying from 0.9ppts and 1.6ptts in case of full escalation.
 What might mitigate the negative impact of the trade war is accommodative fiscal and/or monetary policy by China and/or the US or a patriotic shift of Chinese consumer preferences from US to Chinese manufactured items; our analyses did not include these factors either.
 For a total overview of the methodological approach to determine the bilateral dependencies in terms of value chain integration, see appendix A. For a more detailed overview of underlying data of these figures and industries included, see appendix B.
 The value added share of business services, a sector which is hardly exposed, is included to achieve an adequate scaling of each individual industry.
Amiti, M., S. Redding, and D. Weinstein (2019). The Impact of the 2018 Trade War on U.S. Prices and Welfare. CEPR Discussion Paper 13564.
Auer, R., C.E.V. Borio, and A.J. Filardo (2017). The Globalisation of Inflation: The Growing Importance of Global Value Chains. CEPR Discussion Paper 11905.
Baldwin, R., and J. Lopez‐Gonzalez (2015). Supply‐chain trade: A portrait of global patterns and several testable hypotheses. The World Economy, 38(11), 1682-1721.
Blanchard, E. (2019). Trade wars in the global value chain era. VoxEU.
Erken, H.P.G., M. Every, B.C.J. Giesbergen and E. de Groot (2019). The US-China trade war in the rerun. Rabobank.
Erken, H.P.G., B.C.J. Giesbergen and I. de Vreede (2018). Re-assessing the US-China trade war. Rabobank.
Frohm, E., and V. Gunnella (2017). Sectoral interlinkages in global value chains: spillovers and network effects. ECB working paper series no. 2064.
Gawande, K., B. Hoekman and Y. Cui (2014). Global supply chains and trade policy responses to the 2008 crisis. World Bank Economic Review, 29(1), 102-128.
Giesbergen, B.C.J., E. de Groot and M. Every (2019). US-China trade war: no turning back. Rabobank.
Habib, K., L. Hamelin and H. Wenzel (2016). A dynamic perspective of the geopolitical supply risk of metals. Journal of Cleaner Production, 133, 850-858.
Hayat (2019). Leaving China: Which Asian countries can benefit from a relocation of manufacturing activities?. Rabobank.
OECD (2013a), Global value chains (GVCs): United States, Paris.
Wang, C. and A. Wu (2016). Geographical FDI knowledge spillover and innovation of indigenous firms in China. International Business Review. Volume 25, Issue 4, August 2016, Pages 895-906.
Appendix A: Methodology and data
In order to analyse value chain integration between the US and China and the extent to which the Netherlands is exposed to a disruption of these chains, we use data from the 2018 release of the OECD Trade in Value Added (TiVA) database. This database offers country-level trade flows data for 34 industries (two-digit ISIC). Using this static dataset implies that we assume no substitution effects. We distinguish four different flows of goods and services which we will discuss separately below.
Flow 1: Foreign intermediates used in domestic production
First, to assess the dependency of domestic production on intermediates, we use the following equation:
where represents the share of intermediates of trading partner j in the production of industry i of country m in year t, the exports of intermediates and Y is the gross production.
Flow 2: Reimported value added
The reimported value added of both China and the US is calculated by applying equation 2:
where is the reimported value added of country m from trading partner j in sector i and year t, is the exported foreign value added and X represents export. Here we assume that the exported value added per sector from j to m has the same weight as the relative exports per sector between j and m.
Flow 3: The impact on the Netherlands
Impact on Dutch producers
Employing the origin of value added indicator in the TiVA database helps us to measure the impact on Dutch producers, which we estimate with the next equation:
where is the Dutch value added in sector i, exported from trading partner j to country m in year t. The term: represents the exported Dutch value added, X represents gross exports and is the total Dutch value added. We assume that the relative export flows between j and m have the same weight as the value added flows.
Impact on Dutch consumers
To estimate the impact on Dutch consumers we use equation (1) and proxy the outcome to a part that flows to the Netherlands, see equation (4):
Where is the imported production in sector i, from country m, in year t that needs intermediates from trading partner j. Furthermore, are Dutch imports.
Different approach in calculating flow 2
The method used to calculate flow 3 can also be adopted for flow 2, with the equation becoming:
However, as the proxy is used to distinguish the country of destination instead of identifying the country’s value added in total foreign value added, it underestimates the actual flow of value added when bilateral exports are large. This method works best when there is a third country involved which has a relatively lower amount of exports compared to the other countries. The gross exports of the Netherlands are small compared to the total gross exports of the US and China. This makes it difficult to use the same approach for both types of analysis. Hence, we used this approach to calculate the impact for the Netherlands, and equation (2) for calculating the reimported value added.