« CSET Foretell Blog

Four Possible Scenarios for U.S.-China [De]Coupling in the Semiconductor Industry

Dec 15, 2020 12:26PM UTC

Saif M. Khan and Michael Page

Foretell is CSET's crowd forecasting pilot project focused on technology and security policy. It connects historical and forecast data on near-term events with the big-picture questions that are most relevant to policymakers. This post is part of our Scenario series, where we break down a big-picture scenario into a collection of predictors and metrics we are monitoring on Foretell.

The production of semiconductor chips relies on complex global supply chains. China's ability to produce advanced semiconductor chips depends on imports of U.S., Japanese, and Dutch semiconductor manufacturing equipment (SME) to make chips. Although China is increasing its chipmaking capacity using imported SME, it still relies heavily on imported chips. China especially relies on the United States, Taiwan, and South Korea for imports of the most advanced chips. If China cannot import SME, it will remain dependent on imported chips.

Imagine four possible scenarios based on the above industry dynamics. Each scenario is a function of two predictor variables: (i) China’s access to foreign chips, and (ii) China’s access to foreign SME. 

Which of these four scenarios is most likely? Semiconductor trade and industrial policy could influence and hinge on the answer to that question. For example, increased U.S. and allied export controls on SME could produce a shift toward scenarios II and IV, while increased controls on chips could mean a shift toward scenarios IV and III. China’s industrial policy could also produce scenario shifts: Chinese chipmakers obtaining increased Chinese state subsidies might buy larger quantities of SME; while expansion of China’s technology sector could increase its chip purchases. As trade and industrial policy drives the world toward one of these scenarios, new policy options will become available and are described in more detail in the subsections below.

One way to monitor the relative likelihood of these scenarios over time is to identify metrics for the two predictor variables and track those metrics. China’s access to foreign chips can be measured by three metrics: 1) all Chinese chip imports; 2) U.S. chip exports to China; and 3) changes in chip export policy. Together, these metrics also inform changes in the share of chip imports China obtains from the United States. China’s access to foreign SME can be measured by four metrics: 1) all Chinese SME imports; 2) U.S. SME exports to China; 3) percentage of revenue of the top five SME producers (all based outside China) from Chinese customers; and 4) changes in SME export policy, such as new controls affecting Chinese chipmakers. The first two metrics together inform changes in the share of SME China imports from the United States.

Scenario I: China imports chips while indigenizing

In this scenario, the United States and its allies do not impose stricter export controls on SME and chips. As a result, China continues to import chips in large quantities to meet present demand, while also importing large amounts of SME to build its domestic chipmaking industry. China’s global chipmaking capacity share is projected to increase from 15 percent in 2020 to 24 percent by 2030. While building this capacity takes time, the Chinese market slowly weans itself off of foreign chips as it indigenizes production. As a result, the United States and its allies eventually lose some leverage over Chinese access to chips. For example, the United States and its allies would lose some ability to prevent Chinese state actors from accessing chips. However, even with imports of SME, Chinese chipmakers only have an outside chance at successfully building advanced fabs capable of manufacturing the most advanced chips.

In the near term, this scenario sustains U.S. and allied industry revenues. To maintain a technological and competitive edge, chip companies invest their high revenues in substantial R&D to “run faster” to maintain their ability to produce chips more advanced than Chinese counterparts. U.S. and allied governments also consider expanding public semiconductor R&D funding to supplement private efforts.

In the long term, as China continues to indigenize chip production, non-Chinese chip companies risk losing some business to Chinese counterparts. To counter China’s heavy subsidies, U.S. and allied governments consider increasing financial incentives for local construction of advanced fabs to slow their decline in global chipmaking capacity share relative to China.

Scenario II: Chinese chip dependence

The United States and its allies decline to impose stricter export controls on chips, but expand export controls on SME. With reduced access to foreign SME from the United States, Japan, and the Netherlands, China’s fledgling SME industry struggles to fill the gap and Chinese chipmakers’ plans to build new advanced fabs stall. China’s growth in global chipmaking capacity share slows, or even reverses, as new fabs are built outside China. China struggles to maintain its most advanced existing fabs, which can't operate consistently without continual servicing from foreign SME companies. As a result, China remains dependent on chip imports in the longer term, giving the United States and its allies sustained leverage to apply export controls to limit China’s chip access in the future. China increases efforts to indigenize its SME industry, but extensive subsidies are no substitute for their limited access to experienced talent with the requisite tacit know-how.In the near term, U.S. and allied SME companies suffer some revenue loss, but recover as chipmakers build relatively more fabs outside China to meet consumer demand. U.S. and allied chipmakers benefit from China’s inability to indigenize chip production and continued access to the Chinese market.

Scenario III: China indigenizes chip production

In this scenario, the United States and its allies limit export controls on SME, but expand export controls on chips. With reduced access to imports of advanced chips, China suffers setbacks to its economic, technological, and military ambitions. This spurs the country to expand its already substantial subsidy programs supporting Chinese chipmakers, including for increased purchases of foreign SME. China’s past efforts at semiconductor industrial policy have not succeeded in producing domestic chipmakers competitive with world leaders. China’s new efforts likewise struggle. Still, China’s efforts do substantially increase overall domestic chipmaking capacity to serve its growing chip market, even if large proportions of that capacity use legacy technologies. Therefore, U.S. and allied leverage over China’s access to chips declines.

Meanwhile, by losing access to the vast Chinese market, U.S. chip companies lose revenue, slowing their R&D and reducing their technological edge. U.S. and allied governments therefore consider increasing public R&D for chip design and manufacturing to fill the innovation shortfall, but the prospects of matching lost private R&D are daunting given that private investment accounts for the vast majority of total R&D in the space. By contrast, U.S., Japanese, and Dutch SME companies flourish, given heavy Chinese subsidies to chipmakers purchasing SME.

Scenario IV: China loses access to advanced chips

The United States and its allies expand export controls on both SME and chips. With reduced access to advanced chips, China’s economic, technological, and military development all suffer harm. Without access to foreign SME, China’s plans to indigenize advanced chip production to make up for lost chip imports also hit a wall. China attempts massive, desperate efforts to create fully indigenous chip supply chains, but struggles with anything more than legacy chip technologies. Global chip supply chains ultimately split into two—one for China, and one for the rest of the world.
U.S. and allied chip and SME companies suffer long-lasting harm, as they are no longer able to reap the revenues provided by global markets. Innovation in the global semiconductor industry slows as a result—with downstream effects on the development of other advanced technologies that depend on cutting-edge chips. U.S. and allied governments consider increasing public R&D across the semiconductor supply chain to recoup losses from export controls, a tall order given the scale of lost innovation throughout the supply chain.

You can forecast these metrics at cset-foretell.com.

To stay updated on what we’re doing, follow us on Twitter @CSETForetell.


Saif M. Khan

Saif M. Khan

CSET Research Fellow
Michael Page

Michael Page

CSET Research Fellow
Tip: Mention someone by typing @username