OECD Supply Chain Resilience Review

2.  The state of international supply chains

Abstract

Supply chains are integral to production and trade, and today’s economies are more interlinked than ever. This chapter presents the state of play of global value chains, drawing on OECD data and indicators to map trends in trade patterns for countries and sectors, variations in trade interlinkages and potential exposure to shocks.

The 21^st century global economy is characterised by an unprecedented degree of international integration. Following the emergence of global value chains (GVCs) over the past few decades, it is now extremely common for the different stages of a production process (e.g. design, production, marketing, manufacturing, assembly, distribution) to be carried out in different countries (Gereffi and Fernandez-Stark, 2016[1]). The emergence of GVCs has been transformative in many ways. They have created new business opportunities, boosted productivity and reduced poverty (OECD, 2013[2]; Elms and Low, 2013[3]; World Bank, 2020[4]).

Economic globalisation is not new. In the mid-19^th century, for example, a significant increase in international trade and cross-border flows of capital and labour was already occurring. However, during this wave of globalisation (Williamson, 1996[5]) trade consisted largely of products fully manufactured in one country and exported to consumers in other countries. There were already trade flows of raw materials and commodities, but manufacturing was occurring in a single location.

The economic globalisation that has taken place in the past three decades is different, as production itself is now “unbundled” across countries (Jones and Kierzkowski, 2001[6]); (Baldwin, 2016[7]). This “unbundled production” has been facilitated by technological progress, falling costs of maritime transport (containerisation), long-distance communications, and trade and investment liberalisation. Such international fragmentation of production is reflected, for example, in the high volume of cross-border trade accounted for by products that are not destined for final use by consumers, but that serve instead as intermediate inputs or capital into further production processes. According to the OECD Trade in Value-Added (TiVA) database, intermediate products accounted for as much as 60% of global trade (goods and services) in 2020. Moreover, imports and exports of capital products whose purpose is investment (e.g. machinery, equipment, intellectual property) were responsible for an additional 12% of global trade.

This chapter explores the rise of GVCs, their benefits and risks, and recent trends in their reorganisation. It also investigates concomitant implications for trade policy and international trade integration.

2.1. International supply chains have boosted efficiency, created new opportunities and reduced poverty

The emergence of international supply chains has enabled real time co-ordination between highly specialised businesses located in different parts of the world. This has created new business opportunities and boosted productivity through the ability to achieve finer specialisation in a specific task or input (OECD, 2013[2]). Firms have also gained access to a larger and more diversified pool of capabilities and resources, helping them to innovate and produce more sophisticated products (Lema, Pietrobelli and Rabellotti, 2019[8]). Empirical evidence from industries and firms highlights the positive links between GVC participation and productivity (Criscuolo and Timmis, 2017[9]; Del Prete, Giovannetti and Marvasi, 2017[10]; Constantinescu, Mattoo and Ruta, 2019[11]).

While various factors explain the significant drop in global poverty since the 1990s, the emergence of GVCs has played an important role in increasing income in emerging and developing economies (Elms and Low, 2013[3]; World Bank, 2020[4]; WTO, 2024[12]). The rise in GVC trade1 is associated with a decrease in the percentage of the world population living in poverty, according to the World Bank’s societal poverty lines (Figure 2.1). In low-income and lower-middle income countries there is a clear correlation between increased GDP per capita and the share of trade in GDP (Figure 2.2).

 

2.2. Production and trade are more globalised than ever

The 21^st century started with historically high levels of globalisation. Those levels continue, notwithstanding episodes of disruption in global markets. Figure 2.3 shows that the degree to which world production depends on trade in inputs continues to rise. The “import intensity of global production” is the sum of all gross imports of intermediate inputs along the value chain expressed as a percentage of gross output. For each dollar of output in the world, this indicator tells us how many cents correspond to trade in intermediate inputs in GVCs.2

Although an initial rise in the import intensity of production can be observed as early as the 1970s – following the end of the Bretton Woods system and the shift to floating exchange rates – it was only in the mid-1980s that it began to increase sharply (Figure 2.3). Around this time, advances in information and communications technologies lowered the costs of co-ordinating geographically dispersed activities, enabling multinational enterprises (MNEs) to engage in vertical specialisation.

The integration of production along GVCs accelerated during the 1990s with the collapse of the Soviet Union, the conclusion of the Uruguay Round and the creation of the World Trade Organization (WTO), a new wave of deep regional trade agreements, and market-oriented reforms in the People’s Republic of China (hereafter “China”) (and its subsequent accession to the WTO in 2001). At the onset of the financial crises in 2008, the cumulative value of trade in intermediate inputs at the world level had increased to 17 cents per dollar of output, up from 6 cents per dollar of output in 1965. The subsequent drop in the indicator reflects the collapse in trade that was part of the global financial crisis. By 2010 the import intensity of production had largely recovered to its pre-crisis level (Figure 2.3). In 2023, the global import intensity of production stood at a historically high level.

2.3. Trends vary across countries and industries: Slowbalisation?

In the 2010s, the fast pace of globalisation began to slow. After 2011 the import intensity of production decreased – as measured in current prices (dashed blue line in Figure 2.3 above). This has sometimes been interpreted as a sign that value chains are getting shorter (i.e. a process of diminished interdependence between countries in which production becomes more local) (Haugh et al., 2016[19]; James, 2018[20]; Livesey, 2018[21]; Antràs, 2020[22]).

However, the apparent decline in production dispersal has an alternative explanation. When computed from data in current prices, the indicator fails to account for important price movements that took place over that period. In the 2010s, prices of raw materials declined significantly (Figure 2.4). Everything else being equal, a drop in input prices naturally reduces the value of imported intermediates, so that the import intensity of production shrinks even if the volume traded stays unchanged.

Computing the import intensity of production from data in constant rather than current prices (Figure 2.3, dotted blue line) demonstrates that the dispersal of production did stop growing after the 2008 global financial crisis but remained stable afterwards until 2019. The decline in 2020-2021 related to the COVID-19 pandemic was followed by a strong rebound in trade leading to higher shares of trade in intermediate inputs in gross output. More recent data, when available, will shed further light on potential scenarios for the future (Altman, Bastian and Fattedad, 2024[24]).

2.3.1 Use of imported inputs varies across economies

When the import intensity of production in constant prices is examined at the country level, important differences emerge between major economies (Figure 2.5). In the case of China, use of imported inputs increased sharply between 1995 and 2005. This period was marked by rapid growth in manufacturing and large-scale shifts of production activities by MNEs to China. This fast industrialisation increased China’s imports of foreign-made inputs. Over time, domestic input suppliers became more competitive, and government support encouraged the development of more domestic value chains for selected high-end products. China upgraded its position in GVCs, shifting from processing trade and final assembly to the exports of branded products incorporating domestic inputs. Eventually, around the time of the global financial crisis (2008-2010), the import intensity of China’s production began to decline. Notwithstanding this clear trend towards greater domestic sourcing, China’s use of imported inputs remains high compared to OECD Member countries.

In OECD Member countries, on the other hand, the average import intensity of production increased relatively more slowly during the 1990s and 2000s, but it still trended upwards after the global financial crisis (Figure 2.5). The indicator did start to decline in the second half of the 2010s – reflecting diverging trends across the main OECD economies. The United States and Japan experienced a slight decline, while there was no slowdown for the European Union (analysed as a single economy). Imports of intermediate inputs (from non-EU economies) continued to account for a growing share of production after 2011, with a decline only observed during the COVID-19 crisis in 2020.

2.3.1. Some industries rely more on international supply chains than others

Looking at the import intensity of production for specific industries, there are differences both in changes over time and absolute levels (Figure 2.6). The import intensity of production is much higher in some industries, such as coke and petroleum, water transport and motor vehicles. In the coke and petroleum value chain, imported intermediate inputs accounted for about half of the value of output in 2019, which is more than two times the corresponding proportion for food products. These differences are explained by structural characteristics (e.g. perishable products that cannot be transported over long distances multiple times, proximity to raw materials, etc.); firms’ organisational choices (e.g. outsourcing vs. in-house production); and policy measures (e.g. high tariffs on inputs preventing the offshoring of their production).

There are also differences over time when comparing the import intensity of production in 2011 and in 2019. The value chain with the highest level of international sourcing in 2019 – coke and petroleum – shows a clear shift towards more domestic supply chains, indicated by its data point lying in the bottom right triangle of Figure 2.6, meaning that the value was larger in 2011 than in 2019. This represents a decrease in international sourcing and marks a shift towards more domestic supply. This is also the case for basic metals. On the other hand, the supply chains for other industries, such as water transport, mining (non-energy) and pharmaceuticals, have become more international. Overall, manufacturing industries tend to be more on a trend towards lower import intensities, while services are tending towards higher import intensities (notwithstanding some heterogeneous trends within services and manufacturing industries).

While Figure 2.6 highlights the varying degrees to which different sectors rely on imported inputs, it does not shed light on the geographical distribution of these supply chains. Geography is important as firms strategically select suppliers from specific economies, guided by historical ties, supplier availability, expectations about growth paths and cost differences across markets, and geopolitical considerations. Box 2.1 offers insights into the geography of intermediate sourcing and foreign direct investments in the pharmaceutical supply chain.

 

Box 2.1. What guides sourcing decisions in pharmaceutical supply chains?

The COVID-19 pandemic highlighted the role of resilient pharmaceutical supply chains in safeguarding public health. Yet, medicine shortages are a recurring challenge beyond pandemic periods, motivating governmental efforts to bolster security of supply (OECD, 2024[25]).

This box explores the recent evolution of sourcing patterns in this sector, adopting a broad lens that includes both patented brand-name drugs and generic versions of medicines whose patent has expired.¹ Analysing data on pharmaceutical trade and investment flows provides two complementary perspectives on changes to the geography of pharmaceutical supply chains. Trade flows in pharmaceutical intermediate goods offer a glimpse into global sourcing patterns at given points in time, allowing changes in trading patterns to be monitored. A snapshot of intermediate trade flows encapsulates companies’ prior decisions on where to produce and source goods (reflecting past decisions on investing in subsidiaries, in relationships with suppliers, etc.). Implications of those past choices, together with the limited availability of suppliers, shape companies’ sourcing patterns. By contrast, foreign direct investment trends indicate corporate strategies for the future. Investments allow corporate leaders to reshape the geography of their company’s activities by establishing new subsidiaries or expanding existing ones.

Pharmaceutical trade trends

The Herfindahl-Hirschmann Index (HHI), a frequently used measure of market concentration, can help provide insight into the potential for diversification of supply chains. The HHI is calculated using detailed product-level data to assess the concentration of imports.² Ranging from 0 (fully diversified) to 1 (fully concentrated), it indicates whether sourcing of intermediate inputs relies on a small number of supplying countries. Recent OECD work uses a HHI value of 0.2 to indicate relatively high concentration. As shown in the left-hand graph of Figure 2.7, the level of concentration has fallen slightly since 2011, when it was 0.39. However, there has been no strong trend in recent years, with it staying relatively high at 0.31 in 2022 — within a range that part of the literature identifies as signalling critical concentration.³

The right-hand graph in Figure 2.7 links sourcing of inputs to two dimensions of distance. Geographical distance informs interpretation of a potential shift towards suppliers closer to the company’s home country (where the headquarters are located). Conversely, geopolitical distance relates to a potential reorientation of supply chains towards greater reliance on inputs from geopolitically aligned economies. As is customary in the related literature (Gopinath et al., 2025[26]), Figure 2.7 draws on a proxy for geopolitical distance based on a comparison of countries’ voting patterns at the United Nations General Assembly. This metric helps to gauge changes in supply chain linkages between politically misaligned countries (Gopinath et al., 2025[26]).⁴

In the right-hand graph in Figure 2.7, both measures of distance are normalised so that the level observed in 2004 corresponds to 100. A reduction in geographical distance compared to 2004, i.e. values lower than 100, would suggest a shift towards reorganising supply chains closer to home. For geopolitical distance, values below 100 could point to efforts to reduce ties with geopolitically misaligned countries.

Signs of a reduction in both types of distance appeared in the early 2010s, which were shaped by the aftermath of the global financial crisis and the introduction of new rules governing the international financial system (Brei and von Peter, 2018[27]). Trade patterns in the years since 2015 are ambiguous regarding geographical and geopolitical distance, with recent flows mirroring mid-2000s levels. While no clear shift is apparent, the most recent data points (since 2020) hint at increasing trade between distant economies.

This resonates with contributions highlighting the role international trade played in the global response to the pandemic (World Bank and WTO, 2022[28]). Trade in medical goods expanded during the pandemic. Faced with a surge in demand, supply chain managers likely needed to source products from farther afield.

Overall, the relatively stable pattern emerging from Figure 2.7 may point to underlying factors that limit supply chain managers’ capacity to alter the geography of sourcing in the short run. Constraints such as past investments in manufacturing sites and supplier relationships, lengthy regulatory approval processes, complexities related to intellectual property rights, and a scarcity of alternative suppliers likely restrict the room for rapid reconfigurations of supply chains.

 

Pharmaceutical investment trends

Contrasting with the picture emerging from trade data, information on foreign direct investment (FDI) suggests companies are changing their supply chain strategies. Based on all the investments of pharmaceutical companies worldwide, the top-left graph in Figure 2.8 illustrates a marked shift towards investment destinations that are geographically and geopolitically closer, especially since 2016. During the period shown here (2004-2024), average geopolitical distance was greatest in the mid-2000s, while geographical distance peaked in the mid-2010s. A shift towards greater proximity on both measures is visible after 2016. The tendency is even more pronounced for investments made since 2022.

When focusing on subsets of investments by business function (HQ functions, R&D and production projects), two observations stand out: (1) projects linked to headquarter functions (upper-right graph) – encompassing strategic activities, training, and IT infrastructure – as well as projects focused on research and development (lower left graph), display a trend towards a reduction of both dimensions of distance, mirroring the overall pattern (upper-left graph); (2) the picture is remarkably different for production-focused investments (lower-right graph). While geopolitical distance is also trending downwards, geographical distance has remained relatively stable since 2017 and even saw an increase in 2023.

This pattern aligns with a scenario where pharmaceutical companies try to minimise exposure to risks by curtailing investments in geopolitically distant countries. Importantly, the absence of a clear drop in geographical distance for production-focused investments suggests that rather than nearshoring, companies might seek to maintain a varied global production network while strategically decreasing stakes in geopolitically tense economies.

In summary, synthesising the insights emerging from trade and investment data yields two principal observations: (1) the geography of sourcing evolves only slowly, reflecting multi-faceted constraints; and (2) companies appear likely to weigh the benefits of geopolitical risk mitigation against the costs of abandoning an established, historically shaped supply chain structure.

 

Notes to Box 2.1

1. Trade data rely on the Base pour l’Analyse du Commerce International (BACI) database (Gaulier and Zignago, 2010[29]) and encompass 71 6-digit intermediate products assigned to the pharmaceutical sector based on a correspondence table created for the OECD BTIGE database (https://www.oecd.org/en/data/datasets/bilateral-trade-in-goods-by-industry-and-end-use-category.html). Information on FDI is sourced from the fDi Markets database of the Financial Times (fDi Markets, a service from The Financial Times Limited 2024. All Rights Reserved). This comprehensive resource tracks greenfield investments worldwide. For each investment project, the database provides a description of the corresponding business function. The analysis presented in this section is based on counts of all investment projects assigned to the pharmaceutical sector in this database.

2. The Herfindahl-Hirschmann Index (HHI) was calculated through a three-step process: First, a granular HHI was computed by summing the squared market shares for each pharmaceutical product, importer, and year. Second, these preliminary HHI values were adjusted by the proportion of each product in the importer’s total pharmaceutical intermediate imports to derive an importer-specific HHI. Finally, the importer-specific HHIs were weighted by the importer’s contribution to global pharmaceutical production, using OECD data, to compute the global HHI for a given year.

3. While Figure 2.7 presents a weighted average HHI across all importing countries and pharmaceutical intermediates, this aggregate view may mask higher concentration levels and differing trends at the individual country level. Similarly, calculating concentration at the level of individual firms or incorporating indicators of a product’s substitutability would yield a different perspective. Moreover, there is no universally accepted threshold defining critical concentration. Regarding concentration of imports, Baur & Flach (Baur and Flach, 2022[32]) use 0.33, whereas Arriola et al. (2024[33]) and the European Commission (European Commission, 2021[34]) employ a threshold of 0.4. For more details on HHI and related analysis, please see section 3.3.

4. Data on geographical distance come from the CEPII gravity database (Conte, Cotterlaz and Mayer, 2022[30]). The geopolitical distance metric is derived from Bailey, Strezhnev, & Voeten (2017[31]) based on UN General Assembly votes. The original variable was first converted into a five-year moving average, beginning two years before the year in question. For example, for the analysis referring to 2020, the average from 2014-2018 was used. For limitations of this data source, see Voeten (2023[35]).

2.3.2. Exposure to upstream and downstream production shocks varies across industries

The import intensity of production, with its emphasis on an industry’s reliance on inputs sourced from abroad, can be construed as an indicator of exposure to shocks from foreign input suppliers. Firms, however, are also exposed to shocks that might affect their foreign customers. Figure 2.9 presents two indicators of exposure to foreign production, foreign product exposure for imports (FPEM) and foreign product exposure for exports (FPEX) (Baldwin, Freeman and Theodorakopoulos, 2022[36]). FPEM indicates the percentage of an OECD sector's inputs that are supplied from abroad. Similarly, FPEX captures the sector’s reliance on sales to foreign markets.

The figure shows that exposure to foreign markets tends to be high for upstream sectors such as raw materials. This is also the case for sectors supplying services to foreign markets, in particular infrastructure services which are essential for cross-border trade, such as transport and distribution. Non-strategic manufacturing, which focuses on consumer goods, is the only sector in the OECD with higher upstream dependencies (on foreign inputs, FPEM) than for foreign sales (FPEX). On average, the most exposed sector both upstream and downstream is strategic manufacturing (that includes industries such as petroleum and electronics) and reflects the complexity of the industry’s supply chains which require more production steps and border crossings.

The section is linked to the country-by-country statistical annex on international supply chain interdependencies and trends. It shows the percentage of a country’s sector that could be exposed to foreign disruptions, both upstream (FPEM) and downstream (FPEX) for each OECD Member country (Annex B).

2.3.3. Multinational enterprises account for a large share of trade in GVCs

MNEs’ investment decisions have profound effects on the structure of supply chains. For a firm aiming to reach its foreign customers, exporting its products across country borders may not be the only available option. In many sectors, firms choose to serve their foreign markets by setting up local branches operated through subsidiary companies, known as affiliates.3 From the point of view of their host countries, affiliates are domestic firms. Nevertheless, they operate under the control of their foreign parent company.

In fact, the growing number of companies that have set up foreign branches over the years has been driven by a variety of motives beyond serving local markets. Often, MNEs are drawn to a country by a combination of factors, including its location, natural resources, labour costs, specialised knowledge and the regulatory environment.

MNE groups abroad are nowadays among the most important actors in the global economy. At the world level, the foreign affiliates of MNEs accounted for 11% of global gross output in 2020 (Figure 2.10). Following a period of rapid expansion in the 2000s, this share has been stable since 2011. These figures do not account for the activities of MNEs in their home country, i.e. the activities of headquarters and other home country establishments. According to OECD estimates, the domestic activities of MNEs account for an additional 20% of global gross output. As a result, the activities of MNEs represent around one-third of global gross output.

The activities of MNEs tend to both originate and be concentrated in OECD Member countries. Of all the gross output produced by foreign affiliates globally, 71% is produced by affiliates (regardless of location) of MNEs that are headquartered in OECD Member countries and 57% is produced by foreign affiliates located in OECD Member countries.

The share of foreign affiliates’ sales in world gross output has remained roughly constant since the 2008 global financial crisis. While FDI flows have shown some sign of decline since 2016 (IMF, 2023[37]), there is no similar trend when it comes to the activities of foreign affiliates.4 Multinational production has instead become more dynamic in the past decade with the rise in MNEs from emerging economies and the developing world (Buckley et al., 2023[38]).

References

[18] ADB (2024), Economc Globalization (database), https://kidb.adb.org/globalization.

[24] Altman, S., C. Bastian and D. Fattedad (2024), “Challenging the deglobalization narrative: Global flows have remained resilient through successive shocks”, Journal of International Business Policy, Vol. 7/4, pp. 416-439, https://doi.org/10.1057/s42214-024-00197-0.

[22] Antràs, P. (2020), De-Globalisation? Global Value Chains in the Post-COVID-19 Age, National Bureau of Economic Research, Cambridge, MA, https://doi.org/10.3386/w28115.

[6] Arndt, S. and H. Kierzkowski (eds.) (2001), A framework for fragmentation, Oxford University Press.

[33] Arriola, C. et al. (2024), “Towards demystifying trade dependencies: At what point do trade linkages become a concern?”, OECD Trade Policy Papers, No. 280, OECD Publishing, Paris, https://doi.org/10.1787/2a1a2bb9-en.

[35] Badache, F., L. Kimber and L. Maertens (eds.) (2023), https://www.jstor.org/stable/10.3998/mpub.11685289.47, University of Michigan Press, https://www.jstor.org/stable/10.3998/mpub.11685289.47.

[31] Bailey, M., A. Strezhnev and E. Voeten (2017), “Estimating Dynamic State Preferences from United Nations Voting Data”, Journal of Conflict Resolution, Vol. 61/2, pp. 430-456, https://doi.org/10.1177/0022002715595700.

[7] Baldwin, R. (2016), The Great Convergence: Information technology and the new globalization, Belknap Press.

[36] Baldwin, R., R. Freeman and A. Theodorakopoulos (2022), Horses for Courses: Measuring Foreign Supply Chain Exposure, National Bureau of Economic Research, Cambridge, MA, https://doi.org/10.3386/w30525.

[32] Baur, A. and L. Flach (2022), “German-Chinese trade relations: How dependent is the German economy on China?”, ifo Institute EconPol Policy Report, Vol. 38, https://www.cesifo.org/en/publications/2022/working-paper/german-chinese-trade-relations-how-dependent-german-economy-china.

[40] Beugelsdijk, S. et al. (2010), “Why and how FDI stocks are a biased measure of MNE affiliate activity”, Journal of International Business Studies, Vol. 41/9, pp. 1444-1459, https://doi.org/10.1057/jibs.2010.29.

[41] Blanchard, O. and J. Acalin (2016), What does measured FDI actually measure?, Policy Briefs 16-17. Peterson Institute for International Economics, https://www.piie.com/publications/policy-briefs/what-does-measured-fdi-actually-measure.

[13] Borin, A. and M. Mancini (2019), Measuring What Matters in Global Value Chains and Value-Added Trade, World Bank, Washington, DC, https://doi.org/10.1596/1813-9450-8804.

[27] Brei, M. and G. von Peter (2018), The Distance Effect in Banking and Trade, Elsevier BV, https://doi.org/10.1016/j.jimonfin.2017.10.002.

[38] Buckley, P. et al. (2023), “The evolution of emerging economy multinationals”, Journal of Business Research, Vol. 160, p. 113746, https://doi.org/10.1016/j.jbusres.2023.113746.

[11] Constantinescu, C., A. Mattoo and M. Ruta (2019), “Does vertical specialisation increase productivity?”, The World Economy, Vol. 42/8, pp. 2385-2402, https://doi.org/10.1111/twec.12801.

[30] Conte, M., P. Cotterlaz and T. Mayer (2022), “The CEPII Gravity database”, CEPII Working Paper No. 2022-05.

[9] Criscuolo, C. and J. Timmis (2017), “The relationship between global value chains and productivity”, International Productivity Monitor, Vol. 32, https://ideas.repec.org/a/sls/ipmsls/v32y20174.html.

[10] Del Prete, D., G. Giovannetti and E. Marvasi (2017), “Global value chains participation and productivity gains for North African firms”, Review of World Economics, Vol. 153/4, pp. 675-701, https://doi.org/10.1007/s10290-017-0292-2.

[3] Elms, D. and P. Low (eds.) (2013), Global value chains in a changing world, World Trade Organization, Geneva.

[34] European Commission (2021), “Strategic dependencies and capacities”, Commission Staff Working Document No. 352.

[29] Gaulier, G. and S. Zignago (2010), “BACI: International Trade Database at the Product-Level. The 1994-2007 Version.”, CEPII Working Paper No. 2010-23..

[1] Gereffi, G. and K. Fernandez-Stark (2016), Global Value Chain Analysis: A Primer, 2nd Edition, Center on Globalization, Governance & Competitiveness, Duke University, Durham, NC.

[26] Gopinath, G. et al. (2025), “Changing global linkages: A new Cold War?”, Journal of International Economics, Vol. 153, p. 104042, https://doi.org/10.1016/j.jinteco.2024.104042.

[19] Haugh, D. et al. (2016), “Cardiac Arrest or Dizzy Spell: Why is World Trade So Weak and What can Policy Do About It?”, OECD Economic Policy Papers, No. 18, OECD Publishing, Paris, https://doi.org/10.1787/5jlr2h45q532-en.

[37] IMF (2023), World Economic Outlook, April 2023: A Rocky Recovery, International Monetary Fund, https://www.imf.org/en/Publications/WEO/Issues/2023/04/11/world-economic-outlook-april-2023.

[23] International Monetary Fund (2024), Primary Coomodity Price System (database), https://data.imf.org/?sk=471dddf8-d8a7-499a-81ba-5b332c01f8b9.

[20] James, H. (2018), “Deglobalization: The Rise of Disembedded Unilateralism”, Annual Review of Financial Economics, Vol. 10/1, pp. 219-237, https://doi.org/10.1146/annurev-financial-110217-022625.

[8] Lema, R., C. Pietrobelli and R. Rabellotti (2019), “Innovation in global value chains”, in Handbook on Global Value Chains, Edward Elgar Publishing, https://doi.org/10.4337/9781788113779.00032.

[21] Livesey, F. (2018), “Unpacking the possibilities of deglobalisation”, Cambridge Journal of Regions, Economy and Society, Vol. 11/1, pp. 177-187, https://doi.org/10.1093/cjres/rsx030.

[39] OECD (2024), Analytical Activities of MNEs (AMNE) (database), https://www.oecd.org/en/data/datasets/multinational-enterprises-and-global-value-chains.html.

[16] OECD (2024), Inter-Country Input-Output tables (ICIO), https://www.oecd.org/en/data/datasets/inter-country-input-output-tables.html.

[25] OECD (2024), Securing Medical Supply Chains in a Post-Pandemic World, OECD Health Policy Studies, OECD Publishing, Paris, https://doi.org/10.1787/119c59d9-en.

[2] OECD (2013), Interconnected economies: Benefiting from global value chains, OECD Publishing, Paris, https://doi.org/10.1787/9789264189560-en.

[17] University of Groningen (2023), Long-run World Input–Output Database (WIOD), University of Groningen, https://www.rug.nl/ggdc/valuechain/long-run-wiod.

[5] Williamson, J. (1996), “Globalization, convergence, and history”, The Journal of Economic History, Vol. 56/2, pp. 277-306, https://doi.org/10.1017/s0022050700016454.

[15] WITS (2025), World Integrated Trade Solutions (dataset), https://wits.worldbank.org/.

[14] World Bank (2025), World Development Indicators, https://databank.worldbank.org/source/world-development-indicators (accessed on  2025).

[4] World Bank (2020), World Bank Development Report 2020: Trading for development in the age of global value chains, World Bank, Washington, DC.

[28] World Bank and WTO (2022), Trade Therapy: Deepening Cooperation to Strengthen Pandemic Defenses, The World Bank and the World Trade Organization, https://doi.org/10.1596/978-1-4648-1885-1.

[12] WTO (2024), World Trade Report 2024: Trade and Inclusiveness: How to make trade work for all, World Trade Organization, Geneva, https://www.wto.org/english/res_e/publications_e/wtr24_e.htm.

Notes

← 1.  GVC trade refers to trade flows associated with the vertical specialisation of economies that import inputs in order to export. For a more detailed definition, see note under Figure 2.1.

← 2.  Relying on gross trade flows, this indicator double counts the value added in inputs that cross borders several times, thus accounting for the level of segmentation in world production and not just the foreign value added included in gross output.

← 3.  For statistical purposes, a firm is designated as a foreign affiliate if more than half of the voting power is controlled directly or indirectly by a foreign entity.

← 4.  FDI flows or stocks are often regarded as a biased measure of foreign affiliates’ activity as they are affected by mergers and acquisitions, profit shifting and financial restructuring (Beugelsdijk et al., 2010[40]; Blanchard and Acalin, 2016[41]).