Industrial Clusters as Capital Infrastructure: The Restructuring of China’s Manufacturing Geography and the Repositioning of Global FDI

Industrial Clusters as Capital Infrastructure: The Restructuring of China's Manufacturing Geography and the Repositioning of Global FDI

Global foreign direct investment is undergoing a silent change in its unit of measurement. For the past three decades, multinational companies' site-selection reports used countries as rows and costs as columns; today, the same site-selection report is more likely to take "clusters" as its basic unit—the smart terminal belt on the eastern bank of the Pearl River Delta, the integrated circuit manufacturing and packaging/testing corridor in the Yangtze River Delta, the electronic information grouping in Chengdu-Chongqing, and the panel and photovoltaic industrial belt in Anhui. The state provides the institutional framework; clusters provide the real production function.

This change is not unique to China, but China's particularity lies in this: it has turned clusters from spontaneous market evolution partly into a conscious supply of public goods. According to public compilations, China currently has more than 2,000 industrial clusters, most concentrated in eastern and central economic hubs, and their core functions can be summarized as three things—lowering transaction costs, accelerating knowledge flows, and securing talent supply. Understanding these three things brings one closer to the real logic of global capital allocation than understanding any specific investment promotion policy.

I. Why the Decision-Making Unit Has Shifted Down from the "Country" to the "Cluster"

The classic explanation from agglomeration economics is externalities: when upstream and downstream firms are densely distributed within the same geographic radius, the time costs of intermediate goods procurement, process debugging, and sample iteration are compressed to a level that other regions find difficult to replicate. For industries whose product cycles are measured in quarters or even months—consumer electronics, new energy components, some medical devices—this time compression itself constitutes a competitive barrier, and its value often exceeds the price differential in labor costs.

The second variable is the change in risk structure. In recent years, global FDI has generally become more selective and more concentrated: efficiency-seeking investment has given way to market-seeking and resilience-seeking investment. But resilience does not equal dispersion. What firms truly seek is to "build redundancy among several clusters with depth," rather than to cut capacity into pieces and scatter them to the lowest-cost locations. Low-cost destinations without supplier density provide only low cost, not low risk.

The third variable is the geographization of policy tools. National high-tech zones, economic and technological development zones, and pilot free trade zones form a tiered institutional container; fiscal, land, energy, and talent policies stack within the container, forming an investment environment that firms can anticipate and compare. When policies are deployed on a park-by-park basis, the unit of capital assessment naturally becomes park-based as well.

II. Electronic Information: From a Single Coastal Pole to Multi-Centered Division of Labor

The center of gravity of the electronic information industry remains in the Pearl River Delta. In 2022, Guangdong's electronic information and communications industry generated operating revenue of about RMB 4.67 trillion (about US$642.25 billion), accounting for 26% of the province's industrial revenue. The industry is clustered along the east bank of the Pearl River, with strengths in smart terminals, information and communications, and integrated circuit design. This density was not formed in a short period, and it is difficult to replicate in the short term.

But “the center of gravity being in the Pearl River Delta” does not mean that “other regions are merely supporting players.” The Yangtze River Delta has a deep foundation in notebook computers, semiconductors, and mobile phone manufacturing, with Shanghai serving as a hub for IT enterprises; the Bohai Rim Economic Circle, anchored by the Tianjin area, specializes in communications, software, and home appliances, while Beijing plays a dual role as an R&D and distribution center; Chengdu, Xi'an, and Wuhan are more involved in the production of home appliances and military electronics.

What truly merits investors’ attention are two new growth curves.

The first is Anhui. Electronic information and communications have become the province’s second-largest industrial sector and the main driver of industrial growth. Anhui has a relatively complete display panel industry chain, accounting for about one-tenth of global capacity, and its integrated circuit industry covers fairly complete capabilities from design to testing. More representative is photovoltaics: according to relevant data from the National Development and Reform Commission, the province’s photovoltaic industry revenue grew from 26.8 billion yuan (about US$3.7 billion) in 2012 to 882.8 billion yuan (about US$121.4 billion) in 2021. This is not the success of a single company, but the completion of vertical closure from raw materials to modules by an entire industry chain within a specific geographic area.

The second is Chongqing. Relying on the policy orientation of guiding high-tech industries toward the west, Chongqing has become one of western China’s electronic information hubs. According to data released by the Chongqing municipal government, in 2022 the city’s electronic information and communications manufacturing output exceeded 700 billion yuan (about US$96.3 billion); its industrial structure covers smart terminals, integrated circuits, and instrumentation, and it has gathered more than 800 manufacturing enterprises above designated size, including several major manufacturers in the integrated circuit field. Areas such as Xiyong Microelectronics Park, Liangjiang New Area, and Western Science City are strengthening clusters around new display, new energy, advanced energy storage, and integrated circuits, with the goal of forming a trillion-yuan-level industrial scale.

What appears here is a structural change: China’s electronic information geography is no longer a hierarchical relationship of “coastal manufacturing—inland supporting production,” but a parallel division of labor among several clusters with different specialties.

III. Semiconductors: Zoned by Function, Not Administrative Divisions

The spatial pattern of integrated circuits better illustrates the nature of clusters—they are zoned by industrial function, not distributed by administrative divisions.

The Yangtze River Delta is the center of gravity for manufacturing and packaging/testing. The region gathers cities such as Shanghai, Nanjing, Suzhou, Wuxi, and Hefei; its integrated circuit manufacturing and testing scale exceeds half of the country’s total, and its industry chain extends from design to materials. Beijing-Tianjin-Hebei, with Beijing at its core, has advantages in design and some manufacturing links, with university resources and high-end design parks forming its talent base. The Pearl River Delta leads in design; Shenzhen alone accounted for 88% of Guangdong’s integrated circuit and semiconductor revenue in 2020, directly related to downstream demand from its information industry.Meanwhile, the central and western regions are establishing differentiated positioning, focusing on wide-bandgap semiconductors and automotive-grade chips, with cities such as Xi'an and Wuhan positioning themselves toward storage. This "differentiated competition" is not entirely determined by planning; it is more the result of the combined effect of resource endowments, existing industrial foundations, and local policies. For investors, this means that when making semiconductor arrangements in China, the question is not "whether to enter China" but "which segment of the value chain to enter and which cluster to locate in."

IV. Biopharmaceuticals: Cluster Forms Shaped by Policy Parks

The cluster formation mechanism of biopharmaceuticals is markedly different from that of electronics manufacturing. Its core elements are not supplier density but the accessibility of clinical resources, regulatory channels, talent, and capital; therefore, it is highly dependent on platform-type spaces. China's biopharmaceutical market is expected to grow from RMB 345.7 billion (approximately USD 47.6 billion) in 2020 to RMB 811.6 billion (approximately USD 111.76 billion) in 2025, more than doubling over five years; this growth rate itself constitutes the demand foundation for cluster expansion.

Spatially, biopharmaceutical clusters are generally located within national-level high-tech zones or economic and technological development zones, and there are also independently established biopharmaceutical parks. Major distribution areas include the Yangtze River Delta, the Bohai Rim, the Pearl River Delta, and parts of Northeast China. Central provinces such as Henan, Hunan, and Hubei, as well as the Sichuan-Chongqing region in the west, are also forming growing clusters. Policy documents such as the "14th Five-Year Plan for the Development of the Bioeconomy" emphasize the integration of biotechnology and information technology, providing a medium- to long-term policy anchor for such clusters.

For multinational pharmaceutical and medical device companies, the evaluation dimensions for biopharmaceutical clusters are completely different from those for manufacturing: distance to approval and clinical resources, availability of R&D talent, and whether the local innovation ecosystem has the capacity to undertake collaborative R&D are often more important than land and energy prices.

V. Three Drivers Behind the Rise of Inland Clusters

The emergence of inland clusters is often simply attributed to "industrial transfer policy," but there are at least three layers of actual drivers.

The first layer is policy rebalancing. Coordinated efforts by central and local governments have led western and northern provinces to begin cultivating new clusters oriented toward modern industries.

The second layer is factor structure. Energy prices, availability of green electricity, land supply, and labor stability are gaining weight in some energy-intensive segments (such as photovoltaics and lithium-battery materials), which will systematically change site-selection rankings.

The third layer is demand geography. The central and western regions themselves are large and growing consumer markets. Combined with land logistics corridors such as the China-Europe Railway Express, "being close to the market" and "being close to exports" no longer necessarily point to the coast.

The result is that China's manufacturing geography is evolving from a "single-pole coast" toward "multi-polar specialization." This process is not uniform, nor is it anywhere near complete, but the direction is already clear.

VI. Global Comparison: What Exactly Has "China+1" Relocated?Only by placing China's clusters in a global coordinate system can we clearly see the true boundaries of supply chain restructuring. Penang's electrical and electronics belt, northern Vietnam's electronics assembly zone, southern India's manufacturing corridor, and northern Mexico's nearshoring belt have all become options for multinational companies' dispersed layouts. What these regions share is: friendly policies, controllable costs, and geographic proximity to end markets.

But their common weakness is equally clear—supply chain depth. A mature industrial cluster includes not only assembly plants but also tooling, precision machining, specialty materials, testing services, equipment maintenance, and an engineering talent market. These elements need to be accumulated on a ten-year timescale. Therefore, what happens more often in reality is node migration, not chain migration: companies move final assembly or a particular process step out of China while retaining core suppliers, materials, and equipment stages in China.

This does not mean that China's cluster advantages are unshakable. Tariff structures, export controls, and customers' mandatory dual-sourcing requirements are all genuinely reshaping order flows. A more accurate judgment is: the depth of China's clusters raises the cost of relocation, yet lowers the thresholds for relocation other than urgency—once an alternative cluster crosses critical scale, relocation will accelerate.

VII. Investment Due Diligence Framework from a Cluster Perspective

If clusters become the basic unit of capital allocation, then the granularity of due diligence must adjust accordingly. The following metrics are more explanatory than macro GDP:

• **Supplier density**: the number of locally substitutable suppliers for the same process, which determines bargaining power and shutdown risk;
• **Engineering talent supply**: the alignment between local universities' disciplinary structure and companies' needs, as well as the stability of talent mobility;
• **Upstream and downstream completeness**: whether missing links require cross-border procurement, which directly affects delivery lead times;
• **Energy and logistics conditions**: electricity tariff structure, green electricity share, and port or block-train corridor costs;
• **Policy stability**: continuity of industrial park policies, rather than the scale of incentives in a single year;
• **Cluster maturity stage**: whether it is in the initial agglomeration stage, rapid expansion stage, or mature competition stage, corresponding to completely different entry strategies.

One fact that needs to be repeatedly emphasized is: differences between different clusters within the same province may be greater than differences between countries. Investment analysis using “China” as the unit has already lost precision in most industries.

VIII. Long-term Judgment and Risks to Watch

In the long run, the competitive focus of China's industrial clusters will shift from “investment incentives” to “factor quality”—talent, energy structure, institutional predictability, and the innovation ecosystem. This means the role of local governments will shift from subsidy providers to platform operators, and multinational companies' evaluation weights will adjust accordingly.At the same time, the risks are real and should not be obscured by cluster narratives. In some segments (such as solar manufacturing), there are already signs that capacity expansion is outpacing demand absorption, and downward price pressure may squeeze corporate profits and reinvestment capacity within clusters. Cluster homogenization is also a medium-term risk: when multiple regions target the same track at the same time, dispersed resources will weaken the efficiency of any single cluster in reaching critical scale. In addition, energy constraints, land quotas, and competition for engineering talent may become practical ceilings for some inland clusters.

For global capital, the conclusion is not complicated. The depth of China's industrial clusters remains the main source of supply chain stickiness, but it is being repriced: allocations will be more dispersed, more node-based, and more reliant on empirical judgment at the cluster level, rather than broad judgments about the overall market. Understanding this is more valuable than forecasting FDI data for any single quarter.