Decoupling US-China supply chains: High tech on the move
Co-author: Ralph van Mechelen during his rotation at RaboResearch
- The COVID-19 pandemic and the US-China trade war have increased pressure on firms to consider diversifying their supply chains away from China
- We find clear indications that US supply chains have started to shift away from China
- US manufacturing imports from China dropped by 17% or USD 88bn in 2019, resulting in a decline of China’s share in US imports by 4ppts. This marked decline stands out as a break in the series
- The shift away from China is especially apparent in the computer & electronics sector (e.g. semiconductors and wireless equipment)
- It is no surprise that we find the strongest signs of relocation in this sector given its strategic importance (in terms of security and key technological inputs) and the pressure put on firms by the Trump administration to keep Chinese firms out of their digital infrastructure
- Vietnam, Mexico and Taiwan are the main beneficiaries of the shift in US imports, but the US is also benefiting as we find clear indications of reshoring to the US
- While US imports from China fell by USD 88bn, imports from the rest of the world increased by USD 68bn - meaning that at least USD 20bn worth of production could have moved from China to the US
- Going forward, we see the expected rise in geopolitical tensions as the most important reason for a further acceleration of supply chain relocation in a wide range of sectors
Supply chain relocation is gaining attention
The major disruptions to global trade caused by the COVID-19 pandemic have brought the issue of supply chain relocation to the forefront. However, this issue is not new. The sharp rise in Chinese wages since the early 2010s combined with the perceived risk of an overdependence on China had already triggered firms to rethink their supply chain strategy. In addition the rising trade tensions between the US and China have increased both the cost and uncertainty of producing in China. In an attempt to decrease their reliance on China yet maintain a foothold in the Chinese market, firms have already started to adopt a ‘China plus one’ strategy. With the tensions between the US and China dominating the headlines and the disruption caused by Covid-19 fresh in everyone’s mind, more firms may consider relocation as a means to diversify and thereby boost the resilience of their supply chains.
In a previous Special, we already touched upon the question of supply chain relocation, but were unable to determine whether this development was actually taking place on a larger scale. In this report we use US import data to answer the question: are supply chains shifting away from China, and if so where to? We also shed light on the considerations underlying the decision to relocate activities away from China.
To assess whether supply chains have started to shift away from China, we analyzed disaggregated monthly US Census import data (NAICS3-6) for a selection of countries between 2013 and 2019. We focus on manufacturing imports since manufacturing comprises more than 85% of total US imports (see Figure 1), and China is the US’ most important import partner in this sector. A relocation of supply chains should show up in the mix of countries from which US imports originate. Specifically, in case of relocation, one would expect China’s share in total US imports to decline and other countries to gain at China’s expense.
Gross import data versus value added
We would have preferred to use ‘trade in value added data’, but this data is only available up to 2015. In our view, this is an insurmountable shortcoming since it seems reasonable to expect that the recent trade tensions are an important driver of any potential relocation. That said, an important caveat when using gross import data is that we cannot observe whether part of the value added was generated in a different country than the country of origin. As such, we cannot analyze how much US value added is reimported in the US – a situation which may occur when intermediates produced in the US were first exported to the country where the final product was assembled. In that case, we are likely to overestimate the degree of reshoring.
Similarly, we might overestimate the degree of supply chain relocation from China if Chinese value added is re-exported through, for instance, Vietnam. Chinese manufacturers could simply be trying to avoid tariffs through Vietnamese re-exports, i.e. transporting final goods to Vietnam so that it becomes the country of origin. These US imports would then contain very limited Vietnamese value added. There is some indication of relabeling of Chinese goods in Vietnam, most notably involving textiles, seafood, agricultural products and steel and aluminum (see here and here). Although this is a shortcoming in our analysis, we assume relabeling is not taking place on a massive scale as the US has been vigilant on tariff evasion via this route. It has already warned Vietnam that it can expect tariffs on its exports (see here and here). More importantly, relabeling in itself is an indication of a pending supply chain shift. In other words, relabeling to avoid tariffs does not seem a viable long-term solution and sooner or later the supply chain will shift anyway.
Where could production move to?
Besides data for China and the US, we have collected US Census data on imports from Thailand, Malaysia, Vietnam, Taiwan, India, South Korea and Mexico. We have chosen this selection of Asian countries based on a previous Special, where we examined which Asian countries were most likely to be the beneficiaries of a shift of supply chains away from China. We have also included Mexico in our sample, because of its already large share in US manufacturing imports. Combined, these countries represented 50% of total US manufacturing imports in 2018 (see Figure 2).
Using the data on imports originating from these countries, we can investigate where supply chains may have relocated to. We do this by looking for significant changes in the mix of countries from which US imports originate. For instance, a decline in US imports from China in appliances in combination with an increase in US imports from Mexico could be indicative of a supply chain shift from China to Mexico.
China’s share in total US imports has dropped
Our analysis shows that China’s market share in US imports dropped markedly in 2019 (see table 2 in the appendix). US imports from China dropped by 17% or USD 88bn, resulting in a decline of China’s share in US imports by 4ppts.
This might be the watershed moment in terms of firms decreasing their dependence on China that till thus far has not been visible in the hard data. From the beginning of the 2000s, the Chinese share in US manufacturing imports rose sharply as labor was abundant and relatively cheap given its quality. Despite manufacturing wages increasing in China and firms adopting a ‘China plus one’ sourcing strategy, China’s share in US imports remained relatively constant between 2013 and 2018. The marked decline in 2019 therefore clearly stands out as a break in the series (Figure 3). The timing of the break seems no coincidence when considering that the Trump administration launched a trade spat which saw tariffs lobbed back and forth between the US and China in 2018 and 2019 (see here for a detailed timeline of the US-China trade war).
Who took over that share?
While US imports from China fell by USD 88bn, imports from the rest of the world (ROW) increased by USD 68bn (see Figure 4). This suggests that some production has shifted from China to other countries. As Figure 5 shows, the main beneficiaries have been Vietnam, Mexico and Taiwan. The fact that only part of the drop in US imports from China has been absorbed by ROW suggests that some production was reshored to the US – a matter we will return to at a later stage.
Vietnam captured Chinese market share mostly in apparel and communications equipment, specifically wireless communication equipment (see Figure 13 in the Appendix). The increase in the latter is especially noteworthy as US imports from Vietnam in this industry doubled in 2019. Even though the Vietnamese share in this sector has been gradually rising for some years, the rapid increase in 2019 could indicate some relabeling of Chinese manufacturing. We have no clear evidence of this at this time though.
Mexico saw an increase in US market share in a broad spectrum of sectors, but mainly in transportation equipment and computer & electronics (see Figure 14 in the Appendix). Mexico has already been eating away at China’s market share for the last couple of years in a number of subsectors, most notably computer equipment (see Figure 6). It’s highly likely that the US-China trade war has accelerated this process. The most important reasons why Mexico has become increasingly attractive as a manufacturing hub for US firms are its proximity to the US market and its cost competitiveness relative to China. The new USMCA trade deal will further enhance Mexico’s appeal.
Taiwan mainly benefited from a shift in printed circuit assembly supply chains, a semiconductor subsector (see Figure 15 in the Appendix). The Chinese share in US imports in this product group halved in 2019, mainly because the trade war has induced Taiwanese firms to relocate parts of their supply chain back to Taiwan. The Taiwanese government is also actively stimulating this move (see here). In addition, the US is trying to persuade the Taiwanese Semiconductor Manufacturing Corporation (TSMC) to build a new facility in Arizona (see here). If successful, this would further dampen US imports from China in this sector.
India and Thailand
Surprisingly, two countries that do not seem to be profiting much from the drop in US imports from China are India and Thailand. Despite India’s close diplomatic relationship with the US and its ‘Make in India’ strategy, India only saw a slight increase in its share of US imports, mostly in machinery. Moreover, we had expected Thailand to have benefited more given its relatively similar export mix and competitive manufacturing wages.
The most important reason why India and Thailand have not been able to profit more of the fall in US imports from China is that the major shifts have occurred in computer & electronic products sector; an industry that is relatively small in both countries (see Figure 7 and Figure 8).
Zooming in on computer & electronics manufacturing
It is not surprising that activities in the computer & electronics industryare showing the largest shift away from China. In a previous study (see here) we examined which US industries are most vulnerable to disruptions in the global supply chain: we looked at the share of Chinese intermediates in US production and the share of US value added in Chinese exports. The main findings of that study show that the computers & electronics and electronic equipment sectors are especially vulnerable to a disruption of the international supply chain. Consequently, firms active in these industries are more likely to relocate activities abroad on the back of rising uncertainty caused by the US-China trade war.
High tech product groups contribute heavily to the shift away from China
When we zoom in to the subsectors of computer & electronics manufacturing, it becomes apparent that three high-tech product groups in particular are the main drivers of the shift away from China: namely wireless communication equipment, printed circuit assemblies and semiconductors (Table 1). These product groups are of strategic importance to the US and have received specific attention in the new US Global Economic Security Strategy (GESS) which was proposed in 2019. A part of GESS is the ‘CHIPS for America Act’, which is specifically aimed at incentivizing firms to reshore advanced semiconductor manufacturing. Additionally, US Congress is considering a USD 25bn fund to assist US technology firms to move production out of China (see here).
Are US firms reshoring production?
So is the production that has shifted away from China moving back home? US import data suggests it is. While US imports from China fell by USD 88bn, imports from the rest of the world (ROW) increased by USD 68bn. This means that at least USD 20bn worth of production could have moved from China to the US.
A general indication of US firms moving manufacturing activity to the home base is the drop in the US import ratio, which is calculated by dividing total manufacturing imports by gross manufacturing output. The US import ratio decreased by 0.6 ppts between 2018 and 2019. This may seem like an insignificant drop, but it is clearly a break in the series since the import ratio consistently increased in the six years before 2019 (Figure 9).
When we zoom in to the specific sectors, the computer & electronics sector again stands out (see Table 2 in the Appendix). The majority (USD 24bn) of the USD 40bn drop in US imports from China in this sector was not imported from elsewhere (see Figure 8). The USD 24bn drop in US imports amounts to 6% of total imports of computers and electronics. Unsurprisingly, wireless communication equipment, printed circuit assemblies and semiconductors are the main contributors to this shift (Figure 10). While imports in the computer & electronics sector dropped, gross manufacturing output in the United States actually increased by 4.9% (Figure 11), meaning that the total value of new locally-sourced production is probably even higher.
Another indication of reshoring of computer & electronics manufacturing activity can be found in employment data (see Figure 12). Employment growth in this sector and some subsectors outperformed the private sector as a whole after years of underperformance. Firms have clearly been ramping up jobs in this segment, or at least were reluctant to lay off workers during the onset of the COVID-19 outbreak.
While the drop in imports in computer and electronic manufacturing is the largest by far in absolute numbers, other sectors stand out if we look at percentage changes (see Table 2 in Appendix). Wood products and primary metals manufacturing both show a double-digit drop in imports. However, as China’s role in these sectors is relatively limited, we have excluded them from our analysis. In other sectors the drop in US imports is much less significant. The reshoring of production seems to be predominantly concentrated in the computer & electronics manufacturing sector. We have therefore focused predominantly on this sector.
What can we expect?
With the supply chain disruption from COVID-19 still top of mind, firms are increasingly considering diversification of overseas activity away from China (see here and here). Supply chain relocation might therefore speed up considerably in the coming period. In this respect, it is important to bear the following considerations in mind as the gains from relocation must clearly outweigh the benefits of producing in China.
Relocation decision will be based on cost-benefit analysis
Firms aiming to serve the vast domestic market may prefer to keep production in China as this minimizes transportation costs. Many firms that initially came to China motivated by low wages ended up staying in China to serve the large Chinese consumer market – also called the ‘In China, For China’ strategy. These firms will not be inclined to leave China, especially when taking into account that the Chinese market is bound to grow further as the Chinese economy develops.
Other benefits from producing in China include the country’s experience of hosting international supply chains and the ample availability of relatively cheap but high-quality labor. These perks may be hard to find in other countries that hope to lure production from China.
Supply chain relocation is not a decision that is taken lightly. Firms have often spent many years setting up a production base and developing protocols to make safe and high-quality products for low prices. Relocation of the production base would involve significant costs. This is especially true for investment decisions, like relocating production facilities, but also for the decision about which import partner to use. It may take time to build relationships with suppliers in other countries, which is important since an abrupt change of partners could lead to logistical problems or a loss of quality.
Meanwhile, some firms contemplating relocation may be strapped for cash and therefore more occupied with surviving rather than scrambling together the investment outlays required to relocate. In addition, government support for relocation, e.g. in the form of subsidies or tax breaks, is only available for a select number of industries (mainly high tech).
Governments will likely ramp up efforts to bring production back home
That said, we do expect that the US administration (either red or blue) will continue to persuade firms to move back home by using regulatory incentives. Recently, president Trump signed an executive order on securing information and communications technology. In addition, US Congress is likely to vote the ‘CHIPS for America Act’ into law this summer. It’s not only US firms that are facing increasing political pressure; countries like Japan, South Korea, Taiwan (see here, here and here) already have legislation in place to persuade firms to reshore production.
All in all, due to the regulatory incentives, the fallout from COVID-19 and the expected re-emergence of trade tensions between the US and China, we see the process of supply chain relocation in US manufacturing continuing in the future. Moreover, as the tension between China and the US rises, other countries will be forced to pick a side—as we are already seeing in the case of Huawei. This will lead to a more polarized world where picking one side will automatically adversely affect the relationship with the other. This polarization will further incentivize countries that have not picked China’s side to reduce their dependence on China. As such, we see the expected rise in geopolitical tensions as the most important reason for a further acceleration of supply chain relocation in a wide range of sectors.
 This strategy means that firms are diversifying their operations by adding another sourcing location in Asia.
 We have used the revised US Census import data from 2 July 2020.
 A NAICS Code is a classification within the North American Industry Classification System.
 Countries were ranked on four indicators: i) similarity to the Chinese export basket, ii) competitiveness of manufacturing wages, iii) attractiveness of the investment climate, and iv) institutional quality.
 Creating Helpful Incentives to Produce Semiconductors (CHIPS).