No end of globalization: Digital technologies as a source of fragmentation of manufacturing
Florian Butollo
Apr, 2024
#Trade and FDI
#Manufacturing
#East Asia and Pacific
#EU / Western Europe
#North America
As the performance of digital devices increases by the minute and new digital base technologies like artificial intelligence (AI) or the internet of things (IoT) proliferate, economic relationships change. This not only accounts for processes within enterprises: the automation of tasks, tools for the integration of work processes, and collaboration through cloud infrastructures will also affect the geographies of production, i.e. the places at which work is performed.
In public discourse, one-sided expectations are widespread. The narrative of “Industry 4.0” is mostly associated with the prospect of economic de-globalization. The reasoning behind this is: digitalization makes the relocation ("reshoring") of manufacturing capacities feasible because labor costs become increasingly irrelevant. What is more, the use of AI and the IoT is increasing the flexibility of companies, and they are expected to manufacture customized products without significant losses in efficiency. This would encourage investments in geographical proximity to target markets because a quick response to customer requirements (rather than mere cost or quality considerations) would become the decisive competitive advantage. Even the labor-intensive apparel industry is in the process of relocating back, according to McKinsey & Company in its study "Is apparel coming home?" (Andersson et al., 2018).
Such assumptions are not entirely wrong, but flawed for two reasons: First, they depart from wrong assumptions about the scope of technology adoption and its consequences. Second, they omit the effects of digital network technologies that support a growing fragmentation of production – not reshoring. The main objective of this contribution is to counter this one-sided narrative by showing that digitalization does not solely entail opportunities for reshoring, but also facilitates further offshoring. Such a balanced view on the future of global production has important policy implications: neither should we expect an exodus of manufacturing from developing countries with potentially disastrous consequences for employment and development, nor are we witnessing the end of cost-driven offshoring and competition.
In public discourse, one-sided expectations are widespread. The narrative of “Industry 4.0” is mostly associated with the prospect of economic de-globalization. The reasoning behind this is: digitalization makes the relocation ("reshoring") of manufacturing capacities feasible because labor costs become increasingly irrelevant. What is more, the use of AI and the IoT is increasing the flexibility of companies, and they are expected to manufacture customized products without significant losses in efficiency. This would encourage investments in geographical proximity to target markets because a quick response to customer requirements (rather than mere cost or quality considerations) would become the decisive competitive advantage. Even the labor-intensive apparel industry is in the process of relocating back, according to McKinsey & Company in its study "Is apparel coming home?" (Andersson et al., 2018).
Such assumptions are not entirely wrong, but flawed for two reasons: First, they depart from wrong assumptions about the scope of technology adoption and its consequences. Second, they omit the effects of digital network technologies that support a growing fragmentation of production – not reshoring. The main objective of this contribution is to counter this one-sided narrative by showing that digitalization does not solely entail opportunities for reshoring, but also facilitates further offshoring. Such a balanced view on the future of global production has important policy implications: neither should we expect an exodus of manufacturing from developing countries with potentially disastrous consequences for employment and development, nor are we witnessing the end of cost-driven offshoring and competition.
Catch-up automation drives offshoring
A closer look at the effect of technological change shows that automation does not enhance the competitiveness of high-wage locations, but rather the one of emerging economies. It is a fact that the wage cost differential between the leading industrialized countries and the production locations in emerging economies, namely China, has been declining in recent years, Labour costs in China in 2005 were one-tenth of those in the US; today they are about one-third (Andersson et al., 2018). This is not related to technological change but, mainly a result of labor shortages, a rising tide of labor disputes, and general economic development. The attractiveness of offshoring manufacturing based on cost considerations thus has become much less attractive.
That said, the effects of automation are exactly the opposite as proponents of the reshoring argument assume: In Germany, the US, and Japan, for instance, automation is predominantly incremental. This means that there are hardly any leaps in automation (against the background of already highly automatized production) (Krzywdzinski, 2021). Science-fiction-like tales about “manless factories”, as always, ignite public imagination, but such approaches are scarce. There are many frictions concerning the actual return on investment (advanced automation is expensive), the need to adapt and readjust processes (which is burdensome), and the need for workers’ intuition and experience that becomes particularly relevant in the context of highly developed production models. ‘Industry 4.0’, if taken seriously beyond the fairy tale, is not about the substitution of work, but about data-based optimization of processes that help to raise productivity and to enhance the flexibility and responsiveness of enterprises.
In many emerging economies, however, automation is gaining speed. China, in particular, is by far the main recipient of industrial robots, accounting for roughly two-thirds of world consumption (see figure). The bulk of such investments is not advanced automation of the ‘Industry 4.0’ type at the technological frontier (which is much harder to implement). Rather, catch-up automation is the low-hanging fruit: devices that are up and running in advanced industrial countries for many years but have played a minor role in emerging economies because they were not yet affordable (Butollo & Lüthje 2017). The reason such automation equipment is being implemented now on a grand scale lies in the changing cost structure: automation equipment can now be bought relatively cheap, whereas the cost of labor has risen. The variables concerning the economic feasibility of automation have reversed!
The result is relative productivity gains on the part of the emerging economies – because the use of technology there meets operating costs that remain comparatively low. This particularly concerns locations with an intermediate cost structure close to the US, Europe, and Japan. In Eastern Europe, German lead firms are pushing suppliers to invest in automation equipment of a similar level of sophistication as in Germany (Schwarz-Kocher, Krzywdzinski, & Korflür 2018). Under these circumstances, from a German perspective, the pressure to relocate is not decreasing but increasing. As a result of automation, companies in emerging markets can manufacture more productively while production costs remain comparatively low.
That said, the effects of automation are exactly the opposite as proponents of the reshoring argument assume: In Germany, the US, and Japan, for instance, automation is predominantly incremental. This means that there are hardly any leaps in automation (against the background of already highly automatized production) (Krzywdzinski, 2021). Science-fiction-like tales about “manless factories”, as always, ignite public imagination, but such approaches are scarce. There are many frictions concerning the actual return on investment (advanced automation is expensive), the need to adapt and readjust processes (which is burdensome), and the need for workers’ intuition and experience that becomes particularly relevant in the context of highly developed production models. ‘Industry 4.0’, if taken seriously beyond the fairy tale, is not about the substitution of work, but about data-based optimization of processes that help to raise productivity and to enhance the flexibility and responsiveness of enterprises.
In many emerging economies, however, automation is gaining speed. China, in particular, is by far the main recipient of industrial robots, accounting for roughly two-thirds of world consumption (see figure). The bulk of such investments is not advanced automation of the ‘Industry 4.0’ type at the technological frontier (which is much harder to implement). Rather, catch-up automation is the low-hanging fruit: devices that are up and running in advanced industrial countries for many years but have played a minor role in emerging economies because they were not yet affordable (Butollo & Lüthje 2017). The reason such automation equipment is being implemented now on a grand scale lies in the changing cost structure: automation equipment can now be bought relatively cheap, whereas the cost of labor has risen. The variables concerning the economic feasibility of automation have reversed!
The result is relative productivity gains on the part of the emerging economies – because the use of technology there meets operating costs that remain comparatively low. This particularly concerns locations with an intermediate cost structure close to the US, Europe, and Japan. In Eastern Europe, German lead firms are pushing suppliers to invest in automation equipment of a similar level of sophistication as in Germany (Schwarz-Kocher, Krzywdzinski, & Korflür 2018). Under these circumstances, from a German perspective, the pressure to relocate is not decreasing but increasing. As a result of automation, companies in emerging markets can manufacture more productively while production costs remain comparatively low.
The IoT and AI make distributed manufacturing even easier
Recent technological breakthroughs do not mainly concern mechanical robotics anyway. It is mainly about connecting the internet with offline processes and using the data that can be detected by sensors to optimize them. Such methods can be used to improve the efficacy of distributed work processes in fragmented manufacturing networks, but also in distributed knowledge work. Not the only field of application, but probably one of the most relevant is the combination of e-commerce and logistics that has revolutionized retailing in recent years. The IoT and AI play a significant role in this. This can be observed, for example, in the way Amazon operates: the company can deliver products to consumers within the blink of an eye because it can anticipate the future behavior of customers based on the analysis of present-day buying behavior. Accordingly, the goods are distributed in advance to local department stores, from where they can be delivered quickly (Ulanoff 2014).
While there are major differences between online retailers serving private customers and the management methods used by large companies to coordinate their supply networks, elements of the approaches described above are also taking hold in the supply chain. Contract manufacturers in the electronics industry, which act as world market factories for the major brand-name companies in the IT sector, have, for example, been organizing their manufacturing since the early 2000s in such a way that hubs for configuring products are set up near the target markets, while the actual manufacturing of the hardware takes place in Asia (Hildrum, Ernst, & Fagerberg 2011). The secret of these companies' success lies not only in the combination of high-tech and low wages but also in their sophisticated logistics networks, which minimize warehousing costs based on data monitoring and predictive analysis.
Both the combination of e-commerce/logistics nexus and the practices of digitized supply chain management are examples of how the objective of rapid responsiveness to differentiated customer demand does not necessarily require production networks to be located close to customers but is compatible with a global structure of production. And the story is not over: Alibaba founder Jack Ma claims that soon any product on earth can be delivered to any place on earth within 72 hours by his company (Hu, 2020). This might reflect the hubris that is typical for the shooting stars of the new tech companies, but it also should be taken seriously. It shows that a quick response to customers does not necessitate locating production facilities close to consumers: advanced logistics does the trick.
While there are major differences between online retailers serving private customers and the management methods used by large companies to coordinate their supply networks, elements of the approaches described above are also taking hold in the supply chain. Contract manufacturers in the electronics industry, which act as world market factories for the major brand-name companies in the IT sector, have, for example, been organizing their manufacturing since the early 2000s in such a way that hubs for configuring products are set up near the target markets, while the actual manufacturing of the hardware takes place in Asia (Hildrum, Ernst, & Fagerberg 2011). The secret of these companies' success lies not only in the combination of high-tech and low wages but also in their sophisticated logistics networks, which minimize warehousing costs based on data monitoring and predictive analysis.
Both the combination of e-commerce/logistics nexus and the practices of digitized supply chain management are examples of how the objective of rapid responsiveness to differentiated customer demand does not necessarily require production networks to be located close to customers but is compatible with a global structure of production. And the story is not over: Alibaba founder Jack Ma claims that soon any product on earth can be delivered to any place on earth within 72 hours by his company (Hu, 2020). This might reflect the hubris that is typical for the shooting stars of the new tech companies, but it also should be taken seriously. It shows that a quick response to customers does not necessitate locating production facilities close to consumers: advanced logistics does the trick.
A multi-directional geographic reshuffling
As argued above, the reshoring narrative is flawed. At the same time, even if this is the case, there are reasons that make greater geographic integration of manufacturing and consumer markets seem likely. The interruption of supply chains during the early stages of the COVID19 pandemics demonstrates the risks of an overly complex, globalized, and time-sensitive production network. Currently, management strategists strive for enhanced resilience of operations, although it remains questionable whether this will lead to a significant retreat from global sourcing (Butollo, 2020). More relevant are geopolitical tensions: the current shortage of computer chips shows the perils of excessive dependence on imports (Tyson & Zysman 2021). And in the face of more severe trade restrictions, it might simply become a necessity for manufacturers to maintain a physical presence close to end-markets. In a multipolar world economy that is less polarized between rich countries as consumers and developing countries as manufacturing hubs (which never was that evident), the heydays of offshoring and globalization might be over. But the future most likely will see a complex reshuffling of the geographies of production in which tendencies of a geographic integration of manufacturing and consumption coincide with the opposite tendency of global fragmentation. New digital technologies play an important part in making the latter possible.
References
Andersson, J., Berg, A., Hedrich, S., Ibanez, P., Janmark, J., & Magnus, K. H. (2018). Is apparel manufacturing coming home. McKinsey Apparel, Fashion & Luxury Group.
Butollo, F. (2020). COVID-19 and Global Value Chains: Trigger for a sound economic order? Wissenschaftszentrum Berlin Für Sozialforschung. Retrieved 12 June 2020 (https://www.wzb.eu/de/node/67239).
Butollo, F., & Lüthje, B. (2017). Made in China 2025': Intelligent manufacturing and work. In K. Briken, S. Chillas, M. Krzywdzinski, & A. Marks (Ed). The New Digital Workplace: How New Technologies Revolutionise Work, London: Red Globe Press.
Hildrum, J., Ernst, D., & Fagerberg, J. (2011). The complex interaction between global production networks, digital information systems, and international knowledge transfers. In Handbook on the Economic Complexity of Technological Change. Edward Elgar Publishing.
Hu, M. (2020). Alibaba's logistics arm Cainiao to speed up delivery times to meet the boom in online shopping. South China Morning Post.
Krzywdzinski, M. (2021). Automation, digitalization, and changes in occupational structures in the automobile industry in Germany, Japan, and the United States: a brief history from the early 1990s until 2018. Industrial and Corporate Change.
doi: 10.1093/icc/dtab019.
Schwarz-Kocher, M., Krzywdzinski, M., & Korflür, I. (2018). Standortperspektiven für die produktionswerke der automobilzulieferindustrie. Endbericht an die Hans Böckler Stiftung 1–149.
Tyson, L., & Zysman, J. (2021). America's vital chip mission. Project Syndicate.
Ulanoff, L. (2014). Amazon knows what you want before you buy it. Machine Learning Times.
World Robotics (2020). IFR presents world robotics report 2020. International Federation of Robotics
Andersson, J., Berg, A., Hedrich, S., Ibanez, P., Janmark, J., & Magnus, K. H. (2018). Is apparel manufacturing coming home. McKinsey Apparel, Fashion & Luxury Group.
Butollo, F. (2020). COVID-19 and Global Value Chains: Trigger for a sound economic order? Wissenschaftszentrum Berlin Für Sozialforschung. Retrieved 12 June 2020 (https://www.wzb.eu/de/node/67239).
Butollo, F., & Lüthje, B. (2017). Made in China 2025': Intelligent manufacturing and work. In K. Briken, S. Chillas, M. Krzywdzinski, & A. Marks (Ed). The New Digital Workplace: How New Technologies Revolutionise Work, London: Red Globe Press.
Hildrum, J., Ernst, D., & Fagerberg, J. (2011). The complex interaction between global production networks, digital information systems, and international knowledge transfers. In Handbook on the Economic Complexity of Technological Change. Edward Elgar Publishing.
Hu, M. (2020). Alibaba's logistics arm Cainiao to speed up delivery times to meet the boom in online shopping. South China Morning Post.
Krzywdzinski, M. (2021). Automation, digitalization, and changes in occupational structures in the automobile industry in Germany, Japan, and the United States: a brief history from the early 1990s until 2018. Industrial and Corporate Change.
doi: 10.1093/icc/dtab019.
Schwarz-Kocher, M., Krzywdzinski, M., & Korflür, I. (2018). Standortperspektiven für die produktionswerke der automobilzulieferindustrie. Endbericht an die Hans Böckler Stiftung 1–149.
Tyson, L., & Zysman, J. (2021). America's vital chip mission. Project Syndicate.
Ulanoff, L. (2014). Amazon knows what you want before you buy it. Machine Learning Times.
World Robotics (2020). IFR presents world robotics report 2020. International Federation of Robotics
