BUSI 365 Week 8 Global Industry Project Example

The Global Semiconductor Industry: Structure, Concentration and the Christian Manager

Introduction

Almost every product with a power switch now depends on semiconductors, from phones and cars to medical devices and power grids. The global semiconductor industry is also one of the most geographically concentrated industries in the world. A chip designed in California may be made with Dutch and Japanese equipment in a factory in Taiwan, then tested and packaged in Malaysia before it reaches an assembly line in China or Mexico. That structure delivered remarkable performance at falling prices for decades, but it has also become a source of risk and a focus of competition between governments.

This paper examines the industry's structure, its defining characteristics and the issues now reshaping it, and then considers how a Christian manager can work within it. Its thesis is that the industry's concentration is the product of economic forces, steep learning curves, immense fixed costs and specialized supplier networks, that cannot simply be legislated away, and that the industry's current turmoil places unusual ethical demands on the managers who work in it.

Structure: A Value Chain Divided by Stage

The semiconductor value chain has four main stages, each concentrated in different places and firms.

The first is design. Companies such as Nvidia, Qualcomm, AMD and Apple design chips but often own no factories; they are called fabless firms. Design depends on highly skilled engineers and on specialized software tools, and it is concentrated in the United States, which leads the world in the industry's design work and research spending (Varas et al., 2021).

The second stage is equipment and materials. A small number of firms supply the machines used to make chips. The Dutch company ASML is the only supplier of the extreme ultraviolet lithography machines needed for the most advanced chips, and U.S. and Japanese firms dominate deposition, etching and inspection equipment. Specialized materials, including ultra-pure chemicals, gases and silicon wafers, come from a further set of suppliers, many of them in Japan. Because each of these tools and materials embodies years of accumulated engineering, there are few substitutes, which gives equipment and materials suppliers considerable influence over the pace of the entire industry and makes them central to any government's attempt to control the flow of technology.

The third stage is fabrication, the manufacture of chips on silicon wafers in plants known as fabs. Leading-edge logic fabrication is concentrated in Taiwan, mainly at TSMC, a foundry that manufactures chips designed by other firms, while memory fabrication is concentrated in South Korea. Varas et al. (2021) estimated that all of the world's most advanced logic capacity, below 10 nanometers, was located in Taiwan and South Korea, with the large majority in Taiwan.

The fourth stage is assembly, testing and packaging, which is more labor-intensive and is spread across China, Taiwan, Malaysia, Vietnam and other Asian countries.

This division by stage reflects what Gereffi et al. (2005) call modular governance: lead firms can specify designs in great detail, and highly capable suppliers can produce them without the buyer controlling their operations. A fabless designer sends a digital design to a foundry much as a bicycle brand sends frame drawings to a frame maker, only with far higher stakes. Modularity allowed each stage to concentrate wherever it could be done at the greatest scale and skill.

Business Models: Integrated Makers, Foundries and Fabless Firms

The industry's structure is visible in the business models its firms follow. Integrated device manufacturers design and make their own chips. For decades this was the standard model, and some large firms still follow it, especially in memory, where Samsung, SK Hynix and Micron design and fabricate their own products. The integrated model gives a firm control over the link between design and process, which matters when the two must be developed together.

The foundry model separates the two. A foundry such as TSMC manufactures chips for many customers and does not compete with them by selling its own designs, which makes it a trusted partner for fabless firms. Because a foundry serves many customers, it can run its fabs at high volume and spread fixed costs widely, and because it gains experience from many designs, it moves down the learning curve faster than a single integrated firm can. Those advantages explain much of the foundry model's rise and of Taiwan's position in the industry.

The fabless model is the mirror image. Fabless firms concentrate their resources on design, software and customer relationships and buy manufacturing as a service. This model lowered the cost of entering the industry for new chip designers and helped produce waves of innovation in graphics, mobile and networking chips. It also means that the most valuable chip designers in the world depend on a few manufacturing partners they do not control.

Some integrated firms are now trying to combine models, opening their fabs to outside customers while continuing to make their own products. Whether that hybrid can win the trust of fabless customers, who may compete with the integrated firm's own products, is one of the open strategic questions in the industry.

Characteristics: Why the Industry Concentrated

Three characteristics explain the industry's concentration.

The first is the learning curve. Chipmakers improve yields, the share of usable chips per wafer, as they gain production experience. Irwin and Klenow (1994) studied memory chip generations and found strong learning-by-doing: unit costs fell substantially as cumulative output grew, and firms learned mostly from their own experience rather than from rivals. When the firm that produces the most learns the fastest, the industry tends toward a few large producers.

The second characteristic is fixed cost. A leading-edge fab now costs tens of billions of dollars, and the equipment inside it must be replaced as technology advances. Only firms with very large volumes can spread those costs. This is a cost driver of globalization in the sense the course's first module described: the product must be sold to the whole world to pay for its development and production.

The third characteristic is the industry's history of specialization. Macher et al. (1998) examined the recovery of the U.S. semiconductor industry in the late 1980s and 1990s, after Japanese producers had taken the lead in memory chips, and found that it rested heavily on the rise of specialized design firms, new product segments such as microprocessors and the separation of design from manufacturing. That separation, which made the fabless model possible, also made it possible for fabrication to move to the firms and places that did it best.

Logistics supports this geography. Chips are light and extremely valuable for their weight, so they travel by air with little cost relative to their value. Hummels and Schaur (2013) showed that firms pay high premiums for speed when goods are valuable and demand is uncertain, which describes chips exactly. Distance therefore does little to discourage concentration in the semiconductor industry, unlike heavy goods such as vehicles.

Finally, the industry illustrates the argument of Ghemawat (2003) that the world is only partly integrated. Some stages are global in reach, but each is rooted in particular places with particular skills, institutions and supplier clusters, and those places cannot be replicated quickly.

Issues: Risk, Rivalry and Subsidy

Supply chain concentration and risk

The concentration that made the industry efficient also made it fragile. During the pandemic, a shortage of relatively simple chips halted automobile production around the world, showing how a disruption in one stage can stop industries far downstream. Baldwin and Freeman (2022) argue that concentration in a few locations exposes global supply chains to shocks that individual firms cannot control, and the semiconductor industry is an extreme case. A natural disaster, a power failure or a political crisis affecting Taiwan would disrupt the supply of advanced chips for nearly every industry.

Export controls and geopolitical rivalry

Semiconductors have also become central to competition between the United States and China. Bown (2020) traced how the U.S. trade conflict with China came to focus on semiconductors, as concerns about technology transfer and national security led to restrictions on sales of chips and equipment to Chinese firms. Since then, the United States has broadened its controls on advanced chips and chipmaking tools, and allied countries with key equipment makers have adopted related measures. For firms, this means that customers, suppliers and products must be screened for export rules, and that access to the large Chinese market is uncertain.

Subsidies and the effort to rebuild

Governments are responding with subsidies. In the United States, the CHIPS and Science Act of 2022 provided about $52 billion in funding, most of it for incentives to build fabrication plants on U.S. soil, and Europe, Japan, South Korea and China have announced their own programs. The goal is resilience. But research suggests that relocating production is not automatically the most resilient choice. Grossman et al. (2023) showed that the value of reshoring depends on where shocks are likely to occur and that diversifying production across several countries can be a more efficient form of protection. In an industry with learning curves and enormous fixed costs, spreading production across many locations also raises costs, so the industry faces a real tradeoff between efficiency and resilience.

Talent and the limits of subsidy

Money alone cannot build a semiconductor industry. Fabs require large numbers of engineers and technicians trained in process engineering, equipment maintenance and quality control, and the learning that Irwin and Klenow (1994) documented lives in those people and their routines. Regions trying to build new fabs face shortages of exactly these workers, and new plants have reported delays and higher costs as they recruit and train staff. A subsidy can pay for a building and its equipment; it cannot instantly supply the accumulated know-how that gives established producers their yields. This is why new fabs outside the established clusters often take years to reach the efficiency of plants in Taiwan or South Korea, and why governments are pairing subsidies with workforce programs and university partnerships.

Customers and demand cycles

The industry's customers add another layer of risk. Demand comes from a few large sectors, smartphones, personal computers, data centers and, increasingly, vehicles and industrial equipment, and it moves in cycles. When demand surges, as it did during the pandemic, capacity cannot be added quickly because new fabs take years to build. When demand falls, the industry's high fixed costs turn shortfalls into heavy losses. Rapid growth in demand for chips used in artificial intelligence has recently concentrated profits among the few firms able to design and produce the most advanced processors, intensifying both the concentration of the industry and governments' interest in it.

The Christian Manager: In the World but Not of It

In his prayer before his arrest, Jesus said of his followers that they are not of the world, just as he is not of the world, yet he did not ask the Father to take them out of the world but to keep them from the evil one (English Standard Version Bible, 2001/2016, John 17:14-16). The phrase "in the world but not of it" captures that tension: Christians are sent into ordinary work, but their loyalties and values come from God rather than from the world's priorities. The semiconductor industry gives that tension specific forms.

The first is integrity under competitive pressure. The industry's value lies in intellectual property, designs and process knowledge worth billions of dollars, and the temptation to take shortcuts with others' property, or to look away when a supplier or customer does, is real. A Christian manager respects intellectual property and insists on honest dealing even when rivals do not.

The second is faithfulness in compliance. Export controls are complex, and violations may be invisible to anyone outside the company. A manager who is not of the world treats the law as something to obey faithfully, not something to evade cleverly, while also speaking honestly within the company about the costs and wisdom of particular rules.

The third is stewardship. Fabs consume vast amounts of water and energy and use hazardous chemicals. A Christian manager treats the environment and the communities around a plant as part of the creation entrusted to human care, supporting investments in water recycling and safety even when they reduce short-term profit.

The fourth is the treatment of people across the chain. Assembly and testing employ large numbers of workers in countries where labor protections are weak. Being not of the world means refusing to treat distant workers as mere inputs, insisting on fair conditions in the company's own operations and its suppliers'.

The fifth is fairness in scarcity. When chips run short, as they did during the pandemic, suppliers must decide which customers receive limited output. Those decisions can favor the largest or most profitable buyers and leave smaller customers, such as makers of medical devices or small manufacturers, unable to operate. A manager who is not of the world weighs not only who pays most but who depends most, honors existing commitments and explains allocation decisions honestly, even when a customer is disappointed. Scripture's concern for the weak and for keeping one's word applies directly to these choices, which are made quietly but affect whole communities.

Finally, a Christian manager in an industry so tied to national rivalry must remember that a citizenship higher than any nation shapes his or her loyalties. That does not mean indifference to one's country, but it does mean treating colleagues, customers and competitors in other nations as neighbors rather than enemies.

Conclusion

The global semiconductor industry is a product of powerful economic forces. Learning curves, enormous fixed costs and modular relationships between designers, equipment makers and foundries concentrated each stage of production in a few firms and places, while the light weight and high value of chips meant distance offered little protection to would-be competitors. That structure delivered extraordinary technological progress, but it has also produced fragile supply chains and made chips a focus of geopolitical rivalry and government subsidy. The industry's next decade will be shaped by the tradeoff between efficiency and resilience. The most likely outcome is not a return to national self-sufficiency, which the industry's economics make prohibitively expensive, but a gradual spreading of some capacity across more countries, with the most advanced production remaining concentrated for years to come. Firms that understand both the economic logic of concentration and the political pressure to diversify will be best placed to manage the transition. For Christian managers, the industry is a demanding place to work faithfully: amid intense competition, complex law and global stakes, they are called to be fully engaged in the world's work while remaining rooted in loyalties that are not of it.

References

Baldwin, R., & Freeman, R. (2022). Risks and global supply chains: What we know and what we need to know. Annual Review of Economics, 14, 153-180. https://doi.org/10.1146/annurev-economics-051420-113737

Bown, C. P. (2020). How the United States marched the semiconductor industry into its trade war with China. East Asian Economic Review, 24(4), 349-388. https://doi.org/10.11644/KIEP.EAER.2020.24.4.384

English Standard Version Bible. (2016). Crossway. (Original work published 2001)

Gereffi, G., Humphrey, J., & Sturgeon, T. (2005). The governance of global value chains. Review of International Political Economy, 12(1), 78-104. https://doi.org/10.1080/09692290500049805

Ghemawat, P. (2003). Semiglobalization and international business strategy. Journal of International Business Studies, 34(2), 138-152. https://doi.org/10.1057/palgrave.jibs.8400013

Grossman, G. M., Helpman, E., & Lhuillier, H. (2023). Supply chain resilience: Should policy promote international diversification or reshoring? Journal of Political Economy, 131(12), 3462-3496. https://doi.org/10.1086/725173

Hummels, D. L., & Schaur, G. (2013). Time as a trade barrier. American Economic Review, 103(7), 2935-2959. https://doi.org/10.1257/aer.103.7.2935

Irwin, D. A., & Klenow, P. J. (1994). Learning-by-doing spillovers in the semiconductor industry. Journal of Political Economy, 102(6), 1200-1227. https://doi.org/10.1086/261968

Macher, J. T., Mowery, D. C., & Hodges, D. A. (1998). Reversal of fortune? The recovery of the U.S. semiconductor industry. California Management Review, 41(1), 107-136. https://doi.org/10.2307/41165978

Varas, A., Varadarajan, R., Goodrich, J., & Yinug, F. (2021). Strengthening the global semiconductor supply chain in an uncertain era. Boston Consulting Group and Semiconductor Industry Association.