Hidden Exposure: Measuring U.S. supply chain reliance

Date December 19, 2023
Speaker Richard BALDWIN (Professor of International Economics, IMD Business School, Lausanne)
Moderator URATA Shujiro (Chairman, RIETI / Professor Emeritus, Waseda University)

The global perception of supply chains has shifted from productivity drivers to potential vulnerabilities. De-fragmentation replacing production fragmentation and a trend towards localization contribute to this change. Analyzing countries’ industrial inputs challenges the belief that increased links to foreign sources create vulnerability. The shift from idiosyncratic to systemic shocks adds complexity and increases the potential for disruption. Supply chains should be viewed as networks rather than chains. Indicators based on gross trade emphasize the importance of considering the entire flow. Key distinctions include using “look through” exposure and analyzing imported intermediates. Such indicators can reveal deeper dependency when considering hidden exposure. Policy considerations involve understanding the risk wedge, where private sector optimization aligns with societal interests. Historical practices in various sectors demonstrate that governments intervene to safeguard broader resilience. Resilience discussions should move beyond simple reshoring to adaptive inventory strategies. Nuanced indicators and ongoing research are crucial for effective future policy formulation.


Global framing of supply chain disruption issues

Historically, global supply chains were perceived as drivers of productivity and growth, benefiting both developed and developing nations. However, the contemporary perspective has shifted, now characterizing them as potential sources of vulnerability. The recent G7 Communique underscored the significance of supply chain resilience, acknowledging it as a critical concern. Heads of state are increasingly vocal about supply chain disruptions and the inherent vulnerabilities they introduce.

What led to the shift from a positive perception to one of uncertainty and vulnerability? The focus lies in the concept of supply chains as “links” and disruptions as “shocks.” The problem doesn't stem from the links themselves. Rather, there has been a noticeable trend toward de-fragmentation in the global supply chain. Between 1995 and 2013, production fragmentation increased, but in the last decade, this trend has reversed. Presently, world supply chains are experiencing a shift towards localization: they are becoming less “involved.”

Analyzing Japan's reliance on imported industrial inputs from 1995 to 2020 reveals that the percentage of Japan's industrial inputs has remained relatively stable since 2015. There is no discernible escalation in the extent of dependence on foreign imports. Additionally, the proportion of Japan's imports of industrial inputs in relation to all industrial imports demonstrates a declining trajectory. It becomes evident that the challenge posed by foreign shocks does not arise from an increase in the links to foreign sources.

The crux of the issue lies in the nature of the shocks themselves. In the past, shocks were predominantly idiosyncratic, confined to specific sectors or nations, such as earthquakes or strikes. However, the contemporary landscape presents more systemic and prolonged shocks, concurrently impacting numerous sectors and nations simultaneously. Instances like U.S. tariffs, Brexit, the COVID-19 pandemic, and events like the Russian invasion of Ukraine and now the war in the Middle East exemplify this shift. While businesses could navigate idiosyncratic shocks, systemic shocks prompt governmental involvement due to their widespread and enduring impact.

We have classified these shocks into six distinct combinations based on their source: supply, demand, or the connectivity between supply and division. Furthermore, we classify the shocks as either idiosyncratic or systemic in nature. It's crucial to emphasize that these shocks are not mutually exclusive and in fact can have causal interlinkages. While the media recently attributes these shocks almost exclusively to supply-related issues, we contend that the problem is more intricate and extends beyond such simplistic characterization.

We identify three primary sources of impending systemic shocks. Firstly, geopolitical tensions, exemplified by the ongoing U.S.-China dynamics. Secondly, the impact of climate change, illustrated by potential trade disruptions in the Panama Canal due to low water levels or severe storms affecting vital trade ports. Lastly, the digital realm, particularly cyber-attacks, poses a significant threat to critical infrastructure such as pipelines, airports, or shipping facilities.

Measurement issues and new indicators

The U.S. government, particularly national security agencies and the commerce department, is significantly concerned regarding supply chains, particularly in the realm of semiconductor production. Because these agencies view the approach through the lens of a conventional business value chain, their focus centers on individual firms engaged in the entire process of purchasing, manufacturing, and selling products.

As economists, we contend that this perspective is incomplete, and that supply chains are not simple, linear chains but intricate networks of firms that are buying and selling goods throughout. In the 1990s and 2000s, policymakers posed the question, “where is the work actually done?” This inquiry prompted a focus on “value-added trade” and gauging the extent of value contributed by a specific country to the subject country’s exports. Consequently, indicators such as backward and forward linkages emerged. Fast forward to the 2020s, and policymakers began asking, “how vulnerable are my supply chains?” This question askes where the production occurs for the subject country’s inputs. In response to this evolving query, we developed indicators based on gross trade, instead of value-added trade.

Gross trade is a measure of all international trade, while value added trade removes imported intermediates to ascertain the specific contribution of Japanese value added. The Organization for Economic Cooperation and Development (OECD) has incorporated our indicator into the 2023 Trade in Value Added (TiVA) database update.

Estimations of impacts of shocks based on value added trade exclusively have been shown to be incorrect in recent years. In a simulation of the inflationary impacts in the UK stemming from COVID-related shocks in China, the Bank of England enormously underestimated the impacts. Their estimation was based only on backward linkages, or value-added trade originating from China. However, disruptions did not just halt Chinese value added; they affected the entire flow. Consequently, the actual price impact turned out to be significantly larger than anticipated. When evaluating disruptions, it is imperative to consider the entirety of the flow, or gross trade, rather than isolating specific value-added components.

The primary differentiator of our indicators lies in the reliance on gross trade rather than value-added trade. A noteworthy illustration of the value of gross trade comes from the Ambassador Bridge strike in 2022. During this event, the Teamsters obstructed the main bridge linking Canadian and American auto industries for six days. This not only impeded Canadian value added but impacted the entire gross value added. If the significance of this blockade were assessed using only value-added metrics, it would lead to an underestimation of the actual impact experienced. Therefore, our approach, centered on gross trade, provides a more comprehensive understanding of the real repercussions in such scenarios.

The second notable distinction in our indicators involves differentiating between “face value” exposure to foreign production, which refers to direct purchases from another country, and “look through” exposure, which refers to indirect purchases of foreign goods with intermediate production stages in a third country. In the case of U.S.-China trade, this approach reveals a more intricate and interconnected dependency of U.S. industry on China than it previously seemed.

The third critical distinction lies in our inclusion of imported intermediates used in production for domestic consumption and not only exports. Unlike approaches such as backward linkages, which solely focus on the content of exports, our methodology considers all imports of intermediate goods. In the context of disruptions, the emphasis is on the impact on domestic production, whether or not the good is exported later. Our presentations to government mostly focus on indicators on the import side because that tends to be the focus of their attention, but globally, where systematic embargoes and counter-retaliation exist, the reliance on foreign production as a demand factor may also become increasingly significant.

U.S. Global Supply Chain Engagement

Expanding on how look-through-based indicators are important is the fact that they reveal hidden sources of exposure. as the data from 2018 indicates. There is a hidden source of exposure. At face value, the U.S. has significant exposure to Chinese shocks, but other trade partners like Mexico or Canada also represent significant, if not smaller inputs. However, when employing a look-through approach, China emerges as the overwhelmingly dominant player, and U.S. production is reliant on them in every sector except pharmaceuticals, because often, the other trade partners were also using Chinese intermediate parts. This nuanced analysis underscores that the U.S. exhibits a far higher level of exposure to China than conventional statistics may imply.

A reason for the fact that this exposure has been hidden is simply the speed with which this exposure has shifted to China. When comparing face value versus look-through value in 1995 and 2018, a notable transformation becomes apparent. In 1995, neither face value nor look-through value featured China; instead, Japan held a significant position, with over half of the U.S. exposure due to Japanese production. However, by 2018, China had become the dominant player, occupying over 90% of the U.S. exposure to foreign shocks, showcasing the accelerated shift in the geographic concentration of sourcing over this period. This evolution highlights how the dynamics of international trade have rapidly transformed, underscoring the need for a comprehensive understanding of hidden exposures.

On a global scale, when assessing the percentage of manufactured intermediate production in the world, China has experienced a swift and substantial increase, establishing actual global dominance. Since approximately 2014, China's production of manufactured intermediates has surpassed that of all developed countries combined. This fact is largely unknown.

Creating Policy

The crucial consideration is determining when policy intervention is justified, especially as firms are continually optimizing their supply chain risk to the degree that they are able. There are different perceptions of risk from the private sector (firms) and governments. Perhaps because firms are already engaged in supply chain risk management, their perception of risk is a better approximation of when action should be taken than the perception of governments. As firms seek more cost savings, they concentrate production in low-cost regions, simultaneously increasing risk. From a private sector standpoint, the optimal scenario involves maximizing cost savings while minimizing risk, resulting in an ideal diversification of supply chain risk. If the public sector has a higher perception of risk than the private sector, it might be a reason that the government would intervene in supply chain diversification and resilience.

Governments have recently decided to intervene in many sectors so we should carefully examine for which sectors this is appropriate.

Governments around the world have consistently implemented costly, persistent, and intrusive policies to diversify supply risk in both farming, the defense sector and the financial sector. In all of these sectors the public sector perception of risk is more cautious, resulting in government action.

The financial sector exemplifies another domain where private sector perception of risk is not accepted by the public sector. It is important to determine if semiconductors or medical supplies have gained those criteria.

Mapping shocks to remedies

We aim to highlight a broader perspective on the discussions around enhancing supply chain resilience, as the conventional discourse often revolves around reshoring, friendshoring, or diversifying the supply chain.

To address concerns about geolocated supply shocks, such as the shutdown in China during the COVID-19 pandemic, geo-diversification of supply and stockholding are both appropriate responses, while reshoring may not reduce risk.

On the other hand, demand shocks, such as the example seen during the COVID-19 pandemic when the surge in demand for electronics created a ripple effect in the auto industry, geo-diversifying supply would not have resolved the problem. Only increased stockholding would have been able to address the demand shock.

It is important to think carefully about what shocks are possible when determining appropriate policy responses.

The key takeaway is that accurately measuring foreign supply chain exposure necessitates careful consideration and the selection of appropriate indicators. Recognizing that there isn't a one-size-fits-all solution, which underscores the need for greater theoretical and empirical research on supply chain disruptions, instead of supply chains or global value chains themselves.

Comment and Q&A

URATA Shujiro:
Utilizing appropriate indicators is essential for addressing specific questions. One noteworthy example is trade value added, commonly employed by economists. However, diverse production-based indicators may also be relevant. Professor Baldwin's proposed indicators offer valuable insights into the impacts of supply chain disruptions on the economy and production.

Examining Japan's exposure to China, the face value is 3.0%, while the look-through value is 4.2%. Notably, Korea demonstrates an increased exposure, with a face value of 5.1% and a look-through value of 7.0%. Mexico's exposure is even higher.

Analyzing look-through values reveals intriguing insights. In the case of the U.S., despite a rise, exposure to China remains relatively modest. Professor Baldwin's data indicates the U.S.'s overall exposure to foreign countries is 12.3%, compared to 3.5% for China. Depending on the perspective, the aggregate exposure is not overwhelmingly significant.

On another note, differentiating between robustness and resilience is vital. It raises questions about whether the focus, for both businesses and governments dealing with supply chain disruptions, should be on robustness or resilience. Understanding if these aspects are complementary or substitutes is crucial, and consideration should be given to what governments or businesses prioritize.

Concerning government policies and subsidies, persistent support may lead to issues like overspending. To address this, governments need to strategize and implement effective measures to manage subsidies appropriately, so what would you recommend on that topic?

In a paper by Yasuyuki Todo, he found large impacts from potential supply disruptions from China and especially that within Japan, inter-business linkages and transactions intensify the impact of supply chain disruptions. Exploring strategies to navigate this intricacy becomes imperative for a resilient business ecosystem.

Richard BALDWIN:
Firstly, concerning gross flows and the emphasis on production disruption, the focus shifts to value added when addressing employment shocks. In fact, there is an entire family of measures, including foreign production exposure to value added which is appropriate for examining employment, and the correct indicator should be chosen for the appropriate target, as outlined in the paper. We also explain how to derive other indicators to address other shocks, like connectivity etc.

Every large manufacturing economy is relatively self-sufficient in intermediates, ranging from 80-85% in the U.S., Japan, Germany, China, etc. Smaller countries may have even higher self-sufficiency figures, so the focus of our paper was that the figure was larger than previously thought

Distinguishing between robustness and resilience is crucial. Robustness refers to the ability to sustain processes despite various shocks, often involving costly policies. Resilience, on the other hand, focuses on swift recovery and adopts more cost-effective strategies. While the public discourse in the U.S. often incorrectly uses “resilience” for both, differentiating between these strategies is essential.

Our measures, while offering valuable insights, only provide insights in the aggregate, as an early warning, and supply shocks are often specific.

Subsidies are a nuanced topic, necessitating an open and data-based discussion. Hasty policies on supply chain resilience might prove ineffective and costly. A more thoughtful approach is essential.

In terms of international responses to shocks and disruptions, it would be extremely beneficial as a public good to establish a multilateral cooperative initiative through which data from all supply chains could be shared, allowing for better responses to shocks, as is done with weather monitoring.

In terms of overspending, increasing global interest in some of these areas may have other beneficial effects and eliminate free riding.

URATA Shujiro:
How is risk related to supply chain disruptions actually measured?

Richard BALDWIN:
Consulting industry experts with a deep understanding of the risks associated with a specific sector is essential for this, as each industry faces unique risks.

URATA Shujiro:
What is the role of international cooperation and rulemaking in addressing supply chain issues and disruptions?

Richard BALDWIN:
Coordinated efforts, like friend-shoring, exist, particularly in the face of systemic shocks, as they are international issues. While the rationale for multilateral coordination is evident, implementing it in today's complex world presents challenges. Initiatives are underway in the G7 and Indo-Pacific region. There is a clear argument for this as it is truly an international issue.

URATA Shujiro:
Where is the data shown in the presentation accessible?

Richard BALDWIN:
It is available via the OECD website. It is under TiVA, which stands for Trade in Value Added.

URATA Shujiro:
Given the dominance of Chinese intermediaries, is it feasible to diversify supply chains away from China, not just in face value but also in the look-through sense, using industrial policies? If so, what strategies could be employed to achieve this?

Richard BALDWIN:
China is currently producing 35% of the world’s industrial inputs, and so the global supply chain is deeply interconnected, making a true decoupling unfeasible, at least for the foreseeable future. The U.S. government has restricted decoupling efforts to advanced semi-conductors, quantum computing, and advanced AI tools, while maintaining openness to other industries and products.

While interdependence can allow a competitor to weaponize the relationship, but such actions also then are more costly. In general, I think extensive trade reduces the likelihood of conflicts between nations.

URATA Shujiro:
The risk wedge model implies that when the government is more risk-sensitive than private companies, it results in a diminished reward. Do you believe that the risk assessments of the public and private sector would align if they shared information?

Richard BALDWIN:
A challenge in supply chains is that large companies also lack a comprehensive understanding of their entire supply chain links, potentially underestimating the risk associated with their rewards, because their production involves such large numbers of parts etc. Another issue is the fact that most companies view their partnerships as private information, but universal sharing of such information would definitely be an overall benefit. Sharing the information on shocks would be a very beneficial development as it would allow for better responses.

URATA Shujiro:
I am of the opinion that the government has a responsibility to share information with the private sector if they want the private sector to behave appropriately.

Richard BALDWIN:
The Bank of England has identified supply chains as a threat to financial stability, prompting them to request stress testing of companies by banks. Both the loaning banks and the companies involved are resisting such efforts, as that could jeopardize their businesses. Exploring avenues for responsible information sharing in this context could prove advantageous, as such events can have macro implications.

*This summary was compiled by RIETI Editorial staff.