Production-network embedding and embodied carbon–energy footprints in semiconductor supply chains: An MRIO scenario analysis of Taiwan, Japan, and the United States
Using a multi-region input–output analysis of Taiwan, Japan, and the United States, this study reveals that the embodied carbon and energy footprints of semiconductor supply chains are driven primarily by upstream production-network embedding rather than by the direct efficiency or electricity mix of the host manufacturing site.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
When we buy a smartphone or a computer, we often think about the device itself, but the story of its creation begins long before it reaches a store shelf. It starts with a vast, invisible web of factories, mines, and power plants that work together to turn raw materials into the tiny chips that make modern life possible. This web is known as a supply chain. For decades, the world has relied heavily on one specific place to build these chips: Taiwan. However, as global tensions rise and nations seek to make their own supply chains more secure, there is a growing push to move this production to other countries like Japan and the United States. The big question for policymakers and environmentalists is simple: if we move the factories, does the environmental cost of making these chips go down, or does it stay the same? The answer is not as obvious as it seems, because the environmental impact of a product depends not just on the factory that makes it, but on the entire network of suppliers that feed it.
A team of researchers at Kyushu University set out to answer this question by looking at the hidden energy and carbon costs embedded in the production of semiconductors. They did not build new factories or run physical experiments. Instead, they used a powerful computer model that maps out how money and goods flow between different industries across the globe. This model allowed them to compare the actual production network in Taiwan with the existing production networks in Japan and the United States. They asked a specific "what if" question: if the same amount of semiconductor value produced in Taiwan were instead produced using the supply chains found in Japan or the US, how much less carbon and energy would be required? Their goal was to understand whether the location of the factory matters more, or if the hidden web of suppliers behind the factory is what truly drives the environmental footprint.
The researchers found that the answer lies deep within that invisible web. When they simulated moving the production to Japan or the United States using the existing supply chains in those countries, the results were dramatic. The amount of carbon dioxide associated with making the chips would drop by nearly 89 percent in Japan and by nearly 94 percent in the United States compared to the current situation in Taiwan. Similarly, the energy required to produce the same value of chips would fall by more than 96 percent in Japan and more than 97 percent in the United States. These numbers are not small adjustments; they represent a massive shift in environmental impact. However, the study makes it clear that these are not predictions that moving a single factory will automatically achieve these savings. Rather, they show the difference between the specific network of suppliers in Taiwan and the different networks in Japan and the US.
To understand why the savings were so large, the researchers broke down the numbers to see exactly where the emissions were coming from. They discovered that in Taiwan, the high carbon footprint is not caused by the chip factories themselves being inefficient, but by the electricity and heavy industries that supply them. In Taiwan, the power grid and the local heavy industries that provide materials are very carbon-intensive. When the researchers tried to isolate just the chip factory and swap out only the electricity source, the savings were much smaller, dropping the carbon footprint by only about 20 to 23 percent. This revealed that the massive difference seen in the full simulation comes from the entire upstream network. The suppliers in Japan and the United States are embedded in a different economic structure that is inherently less carbon-heavy, even before you look at the specific chip factory.
The study also looked at whether the savings were just an illusion caused by how the data was counted. They checked to ensure that the results were not hiding emissions in vague categories or "leftover" accounts. The data showed that the emissions in Taiwan were concentrated in real, identifiable sectors like electricity and heavy industry, while the lower emissions in Japan and the US came from a genuinely different mix of suppliers. In fact, the US semiconductor sector relies more heavily on foreign suppliers for its materials, whereas the Taiwanese sector is deeply integrated with its own domestic heavy industries. This structural difference is the key driver of the environmental gap. The researchers also noted that simply making the electricity cleaner in the host countries would not be enough to explain the massive drop in emissions; the entire way the supply chain is built matters more than just the power source.
Ultimately, this research suggests that the environmental consequences of moving semiconductor production are tied less to the physical location of the factory and more to the production network into which it is placed. If a country builds new chip factories but connects them to a supply chain that looks like Taiwan's—relying on the same types of carbon-intensive electricity and heavy industry—the environmental benefits will be much smaller than the simulations suggest. The study does not claim that moving production is a guaranteed solution, nor does it measure other factors like water use or the social impact on workers. Instead, it provides a clear map of the hidden costs. It shows that for nations planning to diversify their supply chains for security reasons, the most important step is not just building the factory, but understanding and reshaping the entire web of suppliers that will feed it. Without changing that network, the environmental gains of relocation may remain out of reach.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.