Pages

19 December 2020

Would China Invade Taiwan for TSMC?

By John Lee and Jan-Peter Kleinhans

Our previous article explained the importance of the Taiwanese firm TSMC as a critical link in the global semiconductor supply chain. Although it is not the only firm with the ability to manufacture cutting-edge logic chips, TSMC is the only viable choice for chip design companies in many situations, and under normal market conditions is likely to remain so for years to come. Control of TSMC’s foundries in Taiwan might thus appear a decisive factor both in Beijing’s readiness to risk attempting unification through force, and for other states deciding whether to take a strong stance against this.

Widening the lens, however, it becomes apparent that TSMC is in fact unlikely to tip the balance. First, invasion likely remains at the bottom of the list for China’s leaders when sizing up solutions to the “Taiwan problem.” Beijing’s military options are components of a long-term political strategy predicated on basic stability across the Taiwan Strait and relying upon China’s growing economic gravity to create the conditions for peaceful unification. This situation has favored China’s economic and technological rise, in which Taiwanese firms including TSMC have played a significant part.

While this strategy is likely being adjusted in response to rising hostility among Taiwan’s younger generations toward China, Beijing retains enough levers of influence within Taiwanese society that it will still see a politics-first approach as viable. The continuing importance to Chinese industry of skilled Taiwanese labor, especially in the semiconductor sector, means outright hostilities with Taiwan would have serious economic consequences, amplifying the effects of U.S. decoupling measures.

Some experts claim that China now has the military capacity to quickly overwhelm Taiwan. Even if this is correct, invasion remains a high-risk endeavor that, even if successful, would still entail major negative ramifications for China. It can be expected only in conditions under which China’s leaders see the immediate political stakes outweighing the military risks, implying a narrow range of scenarios.

If the Chinese government resorts to armed conflict, there are various measures short of invasion that would offer a credible pathway toward an acceptable political outcome for Beijing. For China’s leaders, the imperative would not be for quick resolution but rather to create the conditions for an eventual victory, defined in political terms and not necessarily by direct military occupation. In this context, seizing TSMC’s foundries would be a collateral benefit following from choices made for other reasons.

Second, Beijing cannot presume access to TSMC’s cutting-edge capabilities even after a successful invasion. TSMC’s workforce is increasingly being recruited abroad, as Taiwan’s education system struggles to supply adequate numbers of specialists. Workers with foreign citizenship are unlikely to remain by choice in a Chinese-occupied Taiwan, while keeping them by force would impose even more political costs on China. And even discounting the sabotage of TSMC’s foundries by Taiwanese authorities in the face of Chinese occupation, Beijing must allow for the facilities being damaged in fighting.

Even if a critical mass of TSMC’s workforce and installations was captured intact, the firm’s ability to stay at the cutting-edge under post-invasion conditions would likely be compromised. Foundries depend on intellectual property, machinery, and chemicals that are almost entirely supplied by firms in the United States and U.S.-allied countries, with TSMC’s market leadership dependent on continually upgrading processes and equipment in close cooperation with these suppliers. Such cooperation would likely be drastically curtailed following a Chinese invasion, not just for a hypothetical Chinese-controlled TSMC but for mainland Chinese firms like SMIC, which are already struggling to upgrade their foundry capabilities.

This leads to the most important reason why TSMC is unlikely to be a critical factor in Beijing’s calculations toward Taiwan. The Chinese Communist Party under Xi Jinping has made it clear that achieving independence in “core technologies” like semiconductors is strategically vital for China’s future. This means escaping dependence on a transnational supply chain dominated by firms from the U.S. or allied nations. Taking control of TSMC would not achieve this goal, as TSMC’s role in the global value chain depends on inputs from U.S., European, and Japanese firms.

The whole thrust of China’s industrial and technology policy is now toward achieving greater self-reliance. This includes a broad-based approach to climbing up the semiconductor value chain, by supporting Chinese firms to upgrade their own capabilities. In response to U.S. export control measures, Huawei and SMIC are both working to establish chip-manufacturing lines free from U.S.-origin technology. Although the obstacles are immense, the massive resources being applied by the Chinese state and concentration of key industries in China means that over the long term, Chinese chip design companies can expect to depend progressively less on TSMC.

At this stage however, China’s leaders openly acknowledge that the nation’s high-tech industries still depend on foreign inputs, and hence prioritize keeping open international channels in order to obtain them. With these channels already under pressure from U.S. decoupling measures and increasingly adverse attitudes in other advanced economies, Beijing is unlikely to jeopardize them further through high-risk gambles that would at best yield only short-term advantages.

For third states calibrating their Taiwan policy, long-term considerations are also likely to outweigh the immediate imperative to maintain access to TSMC’s foundries. From the viewpoint of U.S. policymakers, the continuing shift of the cross-strait military balance in China’s favor means that a defense guarantee to Taiwan cannot substitute as a risk mitigation strategy for re-shoring leading-edge foundry work.

The United States is already taking significant steps to achieve this, including pressuring TSMC to relocate facilities away from Taiwan. The extant capabilities of U.S.-based firms like Intel and GlobalFoundries likely provide an adequate basis to achieve re-shoring within an acceptable time frame, given enough willpower and money. For the most critical vulnerabilities, namely military uses, the task is more manageable given that these applications typically do not rely on cutting-edge chips.

Likewise, Japan’s government has reportedly approached TSMC to build a fab in Japan. The European Union, under its drive for “technological sovereignty” and “strategic autonomy,” is also promoting development of cutting-edge fabs within the EU. Seventeen member states have declared a commitment to establishing leading-edge semiconductor production facilities in Europe.

Among other links in the global semiconductor value chain, the nations of Southeast Asia have repeatedly chosen to subordinate a range of political and economic risks to the imperatives of greater economic integration with China. For these countries, TSMC and other Taiwanese firms are links in supply chains that increasingly center on China, a situation that is being reinforced by economic trends around the region. As for South Korea – a U.S. ally that is already actively debating strategic neutrality toward China – the outsize importance of Samsung, TSMC’s only competitor, means that a perceived interest in the elimination of Taiwanese chip-making competition cannot be ruled out.

The hard reality for Taiwan is that even if decoupling from China continues, the mainland will remain a more important economic and security actor across the board for other countries. Governments worldwide are trying to mitigate the risks inherent in over-concentration in the value chain for key technologies by diversifying supply locations and building local manufacturing capacity. From the viewpoint of supply chain security, this is a more feasible approach than maintaining a commitment to Taiwan’s independence in the face of Beijing’s resolve to achieve unification.

There are a range of arguments for why other states should help Taiwan to maintain its de facto independence from China. But TSMC’s undeniably critical role in the semiconductor supply chain is unlikely to tip the balance for foreign governments. Even with TSMC on the scales, it is hard to see third parties perceiving their economic interests in Taiwan to outweigh those in China. If a case is to be made for supporting Taiwan against subordination by the People’s Republic, it must lie in the same type of considerations that are foremost for Beijing’s leaders themselves – political ones.

John Lee is a senior analyst at the Mercator Institute for China Studies (MERICS).

Jan-Peter Kleinhans is the project director for technology and geopolitics at Stiftung Neue Verantwortung (SNV), focusing on semidconductors as a strategic asset.

No comments:

Post a Comment