Effect of COVID-19 on Global Value Chains and Future Prospects

I would like to discuss the effects of COVID-19 on GVCs and their policy implications, focusing on the new debate about whether something is wrong with global value chains and whether new policies should be enacted to fix them. First, COVID-19 has had a huge impact of on global value chains. This is first and foremost a health crisis that has become a serious economic crisis. Looking at a leading indicator—global manufacturing export orders—we can expect a trade collapse larger than the one observed in 2008. This will have a huge impact on GVCs, mainly on the demand side.

The debate on GVCs relates to vulnerabilities—or dysfunctions—related to some very specific value chains. The shortage in face masks triggered some new concerns about GVCs. GVCs for essential goods had to continue to function and these were tested during this crisis. At the OECD, we see that these GVCs were resilient—particularly the food supply chains. Major shortages of supplies like face masks resulted from unanticipated surge demand. I think this issue needs to be addressed in the future. Interestingly, this shortage has been addressed mainly through GVCs. This debate was triggered by the fact that half of world face mask production was concentrated in China, which ceased exporting masks due to domestic need, but in the end China has vastly increased production of face masks.

Reshoring or shortening supply chains fails to address the key questions raised by COVID-19. The first is how to address a surge in demand. I believe that strategies like maintaining buffer stocks, stockpiling and improving the agility and reactivity of supply chains would be helpful. A broader question is how to diversify supply and address bottlenecks. Maybe the solution is improving supply chain transparency and raising risk awareness. If we focus on domestic supply, we reduce the capacity to adjust to economic shocks, limiting our options particularly during the recovery phase of the crisis.

In relation to GVCs, robustness is the capacity to continue to produce during a crisis whereas resilience is the ability to return to normal operations after a crisis. What we are seeing from the literature coming from the 2011 earthquake and other crises is that complex value chains are not robust and are easily disrupted by a crisis, but tend to be resilient because having a large network of suppliers helps you recover more quickly after a crisis.

How can we improve the robustness of GVCs? The traditional risk management steps are information and risk assessment, choice of risk management strategies, reactivity and agility. Companies can organize their responses to disruptions and their management can determine how to prepare. What is now being discussed is whether greater supplier redundancy and diversification are needed to improve robustness. This would involve some offshoring and some international value chains.

What should governments do to help companies build more robust GVCs? Governments have the very important role of ensuring that key infrastructure continues to function—particularly the trade transport network. They may also consider regulatory changes to help firms to cope with emergencies and to adjust to production changes. However, governments have an important role to play post-crisis by working to anticipate and prevent future crises. Governments should be as prepared as companies and should have contingency plans. For example, if plans had been made to be able to ramp up production of ventilators and face masks, it would have helped greatly during the COVID-19 crisis.

How can governments help companies recover more quickly after a crisis? The management literature suggests that the answer is visibility in the value chain. Visibility means knowing the entire supply chain to be able to anticipate where bottlenecks may arise. It also means having the visibility on the demand side to anticipate demand fluctuations. Companies that do this very seriously have very deep information systems enabling them to assess the recovery time for each of their suppliers. However, these companies don't generally diversify, preferring long-term relationships with single suppliers which are thought to engender a greater commitment from suppliers towards rapid recovery.

Governments have the important role of creating a stable regulatory environment—and the policy risk companies face is within a government's control, so this responsibility should be taken seriously. Government can also try raise risk awareness to prepare for new and unfamiliar risks, such as those posed by the digital economy. Much wider use of stress tests, like those performed on financial institutions during the financial crisis, would be one way of doing this.

The first issue with this idea is that value chains are international. Everything upstream in a value chain—research, R&D, design—is done in innovation clusters in cities and it's really difficult to change the locations of these. The availability of raw materials is geographically dependent. Each processed input has its own value chain and countries specialize in these inputs. Looking downstream within a value chain, proximity to consumers is important, particularly with the new trend toward "servicification" of manufacturing. Reshoring may only be possible for final assembly and first-tier suppliers, meaning that some part of the value chain will always be international.

Reshoring also inevitably means compelling companies to adopt business models that they naturally would not. Promoting domestic production would inevitably involve erecting trade and investment barriers in addition to subsidies and maybe even government control, leading to economic distortion and possibly triggering retaliation across countries. I think it would be a bad idea to go in this direction as we recover from the crisis.

We need to be sure that the benefits of new GVC policies outweigh the risks. Given that we are still facing the COVID-19 pandemic, now may not be the best time to experiment with new trade and investment policies; we should really focus on recovery. It is possible to imagine some kind of softer reshoring strategy. Rather than forcing companies to adopt production models they don't want, they could be induced to find a business rationale for reshoring if governments step in to improve domestic business conditions and promote investment in education, skills, innovation and infrastructure. I also imagine more international cooperation in the form of private/public platforms where governments and companies could discuss regulation and value chain bottlenecks. This kind of cooperation could also help prepare companies and GVCs for crises.

Commentator:
I agree with Sebastien on the importance of clearly distinguishing between GVC robustness and resilience. I likewise agree with him that promoting reshoring of production in developed countries could be very dangerous. I share many of Sebastien's views on global value chains and my remarks will mainly recontextualize his comments for the situation in East Asia and raise a few additional points about global value chains.

The emergency in East Asia and the need for various kinds of social distancing restricted both production and consumption. We initially had some supply shocks emanating from China—particularly medical equipment and related goods. The production system for other goods is almost completely intact but now faces low demand. We will face two obstacles in the coming "exit" stage. The first is the huge negative demand shock; we will likely have a prolonged recession. The second is the need to lift restrictions on the domestic and international movement of people. The situation has resulted in claims that the era of GVCs is over and that production should return to developed countries—so-called reshoring—to make value chains more robust/resilient. Sebastien rightly challenges these claims and I share his views.

I'd like to mention in terms of GVC robustness/resilience that building in robustness is much more costly than resilience. Because of this, it may only be practical to bolster the robustness of GVCs for essential goods.

The large amount of empirical data we have from previous East Asian crises is basically consistent with Sebastien's arguments. Global value chains can be a transmission channel for an economic shock. However, network trade is less likely to be disrupted and more likely to revive. We have found network trade as a whole to be both robust and resilient. The key is the nature of global value chains, particularly in the context of East Asia. For example, GVCs include natural resources in which trade is very slow and transactions are basically market transactions. By contrast, international production networks in the machinery industries and other industries—those subject to what Richard Baldwin has called "the second unbundling"— include relation-specific transactions that are probably more robust and also more resilient than other types of transactions.

East Asian countries other than China first perceived COVID-19 as a supply shock when imports from China were interrupted in January and February. This triggered claims that we are too dependent on China despite the fact that production in China seems ready to resume. The initial worries are likely different from what we are talking about now. Japan is probably not too dependent on China. South Korea's case is different given their much tighter economic relationship with China. Geopolitics have also begun to intrude, with the relationships between the US and other developed countries and China being currently strained, giving rise to discussions concerning economic decoupling. The resulting reshuffling of GVCs will depend on the extent of this "decoupling." Will it be limited to some IPR- or private data-related goods, or will it extend to high tech goods more generally, or even go beyond that? The extent of this commitment to decoupling is very important to the possible reshuffling of GVCs in East Asia.

I believe GVCs can still be a source of efficiency and competitiveness in East Asia. I agree with Sebastien that pursuing policies that are contrary to market mechanisms could be very costly and I think that deeper involvement in GVCs can still work in East Asia. Digital transformation is another very important issue. ICT could strengthen both GVCs and also traditional industries like agriculture, transport and tourism. It could also help upgrade government services. ICT is generating new businesses, including service outsourcing—the so-called "third unbundling, which will Provide us with different types of global value chains.

I think a huge demand shock will come and restrictions on people's movements will remain. I believe we’re looking at a prolonged recession at all levels: the country level, the regional level and the global level, and while international production networks are robust and resilient to short-run shocks, a long downturn in the world economy may trigger substantial reshuffling. The transportation/logistics sector has already been badly damaged and SMEs and local firms involved in global value chains and production networks may be affected. More generally, business failures will accelerate in both developed and developing countries, unemployment may go up, and a financial sector crisis may occur—possibly the collapse of asset/foreign currency markets. Health policy and macroeconomic policy are important, but in addition to these, what kinds of policies are needed for production networks? I’d like to hear Sebastien’s opinions on this.

MIROUDOT:
Dr. KIMURA raised many important points, particularly on the question of cost. It’s true that what I described as the robustness is a very costly strategy. The management literature reveals widespread skepticism about supply redundancy and supplier diversification due to how costly and time-consuming it is to find and build trust with suppliers. The possibility of doing this also depends on the type of input. A very specialized input requires a long-term relationship with a supplier whereas more basic inputs that are more homogeneous offer greater potential for supplier diversification. Improving robustness is also costly because of the necessity of somehow addressing crisis response at every stage of the production process. The potential cost of a disaster or risk must somehow be balanced against the cost of the risk management strategy. This is something that companies can only determine themselves.

Governments can play a role by asking companies in specific sectors that produce essential goods (e.g., face masks and ventilators) whether they can ensure surge capacity production in the event of a crisis. This would force companies to examine the cost of implementing strategies that would allow them to fulfill these commitments. Some kind of government support may also be possible on a case-by-case basis. I think this could be achieved through some type of public-private partnership platform. One difficulty will be distinguishing essential goods from other goods when we don’t know what the next crisis will be.

Whether the "just-in-time" strategy should be abandoned in favor of "just-in-case" strategies depends on the type of input and the phase of the production process. Companies will identify inputs and phases of the production process that are essential and target them for risk reduction investments while retaining the just-in-time strategy for lower-risk parts of the value chain. This is another area in which I don’t think it’s reasonable for a government to dictate what is best for companies.

In some cases, companies fully control a value chain and can implement a comprehensive resilience strategy for it. This is not the case for value chains that involve many companies in different countries. How can we improve resiliency in these value chains? Sophisticated goods and products (e.g., semiconductors and mobile phones) cannot be fully domestically produced. By contrast, for simple products (e.g., basic pharmaceuticals like painkillers), reshoring of production is conceivable. It is impossible to produce complex goods today without some kind of international value chain. That does not mean that these value chains will not change and evolve.

I think what Dr. KIMURA said about this issue being driven more by geopolitical factors and by the China factor is important. The servicification of manufacturing can be expected to produce some changes in global value chains. They may become shorter term because of the increased importance of producing as close to the consumer as possible, but they will remain very international because specialization is the main economic mechanism driving competitiveness and efficiency. It’s difficult to conjure up this kind of specialization for an entire economy within a single country or region. I believe global value chains will change, but we cannot really do without them.

Commentator:
The definition of emergency goods is very important. If we include food, we risk the exploitation of these policies by agricultural protectionists, particularly in Japan and South Korea. For this reason, I don’t think we should include all food; I think we have to be very careful. Also, I’m very glad that Sebastien shares my optimism about the continued viability of global value chains for economic development in East Asia.

MIROUDOT:
Visibility in the value chain is often asymmetrical. Big companies like Toyota have greater access to supply chain information than smaller companies. When we look at the overall data at the OECD, we don’t find a high degree of dependency on China, but when you look at specific products and go upstream in the value chains, you find some raw materials and products that are very China-dependent. Governments can promote supply chain transparency and set up tools to create visibility.

ICT gives us many new tools, and in particular, tools that can improve visibility. By placing sensors on inputs, companies can create what they call a "control tower": a full ICT system where they can track stocks, inventories and the movement of inputs in real time. Blockchain can also play a role by promoting mutual trust. An integrated company can trust its own information, but when working with other companies, trust becomes an issue as does confidentiality. Governments can promote these new tools and create a regulatory environment that encourages their adoption.

Commentator:
Global value chains include various kinds of transactions and international divisions of labor. In the case of textiles, garments and wholesalers, for example, they are relatively simple and easy to separate. Just as an example, in one case where demand declined in New York, orders were canceled immediately throughout the supply chain. This is impossible in the case of relatively complicated machinery industries. Some transactions are relation-specific and parties are incentivized to avoid damaging the relationships they’ve established.

The international division of labor is very complicated in East Asia. ICT is a very important element in strengthening communication inside networks. Right now, people cannot physically travel and ICT makes it possible to sustain production networks and maintain existing relationships. We really have to maintain the location advantages of East Asia, particularly the relatively complicated division of labor that exists here.

MIROUDOT:
There is now a realization that demand will stay low also because even if manufacturing resumes, some sectors, like the transport sector and many services, will not begin to recover until the pandemic ends.

I agree with Dr. KIMURA that the global value chain is not just manufacturing; services link manufacturing activities. Restrictions on the movement of people impact manufacturing value chains. If a vaccine is found within a matter of months, it is conceivable that the recovery could begin by the end of the year. If not, and we have to continue to live with the virus indefinitely, firms will have to adjust their production processes and value chains will be more significantly negatively impacted.

Commentator:
I think the demand shock will be huge; it could be much larger than even the global financial crisis. The transportation sector and the service-related logistics sector will be at immediate risk. I also worry about SMEs as they are local firms integrated into value chains. Multinationals and key production sectors may survive the coming years, albeit with great difficulty. I think there will be a huge impact on human capital issues too, particularly in newly developed and developing countries. In the past, in the case of GVCs, many multinationals and other big companies tried to maintain employment despite low production by doing a lot of training and some reshuffling of production lines. In the current situation, there would have to be a certain amount of breathing room to make that possible and I worry about whether such breathing room exists with this pandemic. Governments need to do more than provide subsidies; they should act to maintain production networks, particularly in newly developing countries in East Asia.

MIROUDOT:
We don’t really see data indicating widespread reshoring even if some companies are reshoring in the US or EU. We see evidence of GVC restructuring, however, particularly with companies moving production to countries like Vietnam, India and Mexico due to the tensions with China. Opportunities for these emerging economies exist. Some African economies are also starting to participate in global production networks.

In terms of policy, we have recently developed at the OECD what we call the ABCs of GVC-oriented policies. A is for attraction: promoting trade investment openness, regulations and institutions to attract foreign companies. B is for buzz. Richard Baldwin has said that cities are the factories of the 21st century, so you want to create cities where you have knowledge spillover, that attract talent with universities etc., and create knowledge clusters. C is connectedness, with foreign countries. Even if it sounds like an old-fashioned policy recommendation, we believe the ABCs will help developing countries increase their participation in global value chains.

Commentator:
I think this is a very important moment for many countries—ASEAN member states, for example—to strengthen their locational advantages and attractiveness for investment in a more organized way. The shock is huge and value chains will be reshuffled. Geopolitics is increasingly complicated and some countries can take advantage of the resulting trade and investment diversion effects. Vietnam and Mexico seem to be doing this. In any case, they really have to take GVCs very seriously.

MIROUDOT:
Right now, the mood favors unilateral policies rather than international cooperation, but I think this GVC topic can be an area in which countries should talk to each other in the future, whether through regional trade agreements or old or new forums, and with businesses in order to move to new types of regulation and discipline.