US Secretary of Commerce Howard Lutnick brought new attention to Taiwan’s “silicon shield” in a September 28 interview, during which he said Taiwan should move half of its world-leading semiconductor manufacturing capability to the United States.
Taiwan produces 60% of the world’s semiconductors and 90 of the most advanced chips – not to mention that the industry accounts for 15%of Taiwan’s GDP.
Taiwan Vice-Premier Cheng Li-chiun, who led Taiwan’s trade talks with Washington, on October 1 dismissed Lutnick’s idea, saying, “Our negotiating team has never made any commitment to a 50-50 split on chips … nor would we agree to such conditions.”
Cheng’s response reflects the belief of Taiwan’s people that global reliance on Taiwan-made chips helps protect the island from a Chinese military assault. That notion has some merit, but it also has limits.
The term “silicon shield” is attributed to a 2000 article by Journalist Craig Addison, who argued that the US would militarily defend Taiwan to “protect its supply of information technology products from Chinese aggression,” just as the US military intervened to expel invading Iraqi forces from controlling oil supplies in Kuwait in 1991.
The updated version of the theory is that China won’t attack Taiwan because this would interrupt the supply of semiconductors upon which China’s economy depends. China gets about one-third of its semiconductors from Taiwan.
Furthermore, as Addison argued, the reliance of other countries on semiconductors from Taiwan increases the likelihood they would oppose (in the US case, militarily intervene against) China’s attack.
Accordingly, moving chip production out of Taiwan would make Taiwan less secure by weakening China’s disincentive to attack.
The full story, however, raises doubts about the efficacy of the shield.
To begin with, there is an opposite counter-theory: Taiwan’s semiconductor production may give Beijing an additional and decisive incentive to forcibly annex Taiwan, which the Chinese government already wants to do. China may decide it needs to ensure China’s access to advanced chips, or cut off the supply of chips to the US, or both.
According to this theory, the race between China and the US to achieve artificial general intelligence – the condition in which AI equals or exceeds human intelligence in most fields of endeavor – could force China’s hand even if there is no other compelling political, economic or military reason for Beijing to attack.
Even if the shield works now, what about in the medium-term? Taiwan may not be able to sustain its dominance of global semiconductor manufacturing given that the Chinese, American, Japanese and South Korean governments are heavily investing in their own indigenous semiconductor industries, partly to escape the vulnerability of over-reliance on a supplier that is under threat of military attack. The 2022 CHIPS and Science Act, for example, provided almost $53 billion to expand America’s indigenous chip industry.
Taiwan is seeing a decline in the availability of qualified engineers who produce chips. On top of a shrinking general population, smaller percentages of Taiwan college students are choosing to major in STEM fields. Some also take overseas jobs that offer higher pay and more favorable work-life balance. Moreover, climate change is worsening chronic water shortages in Taiwan. Chip manufacture requires heavy water usage.
Perhaps most importantly, it is doubtful that Taiwan’s semiconductor manufacturing industry will be the decisive factor in Beijing’s deliberations over whether or not to attack Taiwan. Xi Jinping or any other Chinese Communist Party (CCP) general secretary would pull the trigger if convinced that Taiwan were so decisively moving toward permanent political separation from China that he would lose his mandate to rule, and that war seemed the only available recourse.
China will not take on the huge risks and losses of a Taiwan Strait war over an economic or technological issue.
During the early 1980s, many observers argued that the Chinese government would renew Britain’s lease over the New Territories because Hong Kong under British administration was a money-maker for China. But Beijing took Hong Kong back anyway in 1997, and then dismantled Hong Kong’s civil liberties starting in 2020. Although these actions threatened Hong Kong’s economic profitability, they were consistent with the more importanr objective of preserving regime security – in this case, by demonstrating that the CCP government could reassert full control over a formerly lost territory.
Another example was the 1989 crackdown on protestors in and near Tiananmen Square while international media looked on. The government was willing to suffer global reputational damage and economic sanctions in order to send the message to the Chinese public that Beijing does not tolerate people power.
The silicon shield issue pulls on the threads of three larger conditions that frame Taiwan’s security situation.
Taiwan’s domestic politics
The first involves Taiwan’s domestic politics. Although both of Taiwan’s main political parties oppose transferring half of the island’s semiconductor production to US soil, they disagree more generally over to how manage semiconductor production as a strategic asset
That disagreement is a manifestation of their fundamentally different approaches to Taiwan’s grand strategy: peace with China through partial integration versus eschewing China to maximize linkages with friendly countries.
Members of the opposition Kuomintang (KMT, Chinese Nationalist Party) have criticized the Democratic Progress Party (DPP) government at every step of the ongoing drama involving US-Taiwan trade – blaming the DPP for US President Donald Trump’s tariffs against Taiwan, calling for retaliatory tariffs against the US and saying that the Taiwan Semiconductor Manufacturing Company (TSMC) should suspend its investments in the US.
In 2020, in Taiwan’s response to previous US demands, TSMC built semiconductor production facilities in Phoenix, Arizona. KMT politicians said that the DPP had “caved in to pressure from the US government” and that “TSMC will surely become USMC.”
Phoenix celebrates TSMC’s big investment. Photo: LocalToday
In March, TSMC pledged to invest an additional $100 billion to further expand its production in the US. After that announcement, Former ROC President Ma Ying-jeou accused the ruling government of “selling TSMC” as a “protection fee” to Washington.
PRC propaganda has joined the fray to criticize the DPP and promote the messages of yi Mei (doubting America) and the inevitability of unification. “Now is the time for the island to abandon illusions that it could court the US by investing in the US,” said one Chinese “expert” quoted in Global Times. “The only choice [for Taiwan] is to integrate with the chipmaking ecosystem of the mainland.”
Another Global Times article said “experts have slammed the DPP authorities for their near-complete pandering to the United States, and … forfeiting the island’s benefits.”
The current Washington administration
Second, Taiwan’s centrality in semiconductor production may not be enough reason for the US to intervene militarily.
Previous US governments considered Taiwan worth defending as a linchpin of US strategic leadership in Asia. Forcible absorption of Taiwan into the PRC would expand China’s territory, economic capacity and ability to project power. US allies and friends in the region would be at risk of Chinese domination, and neutral governments would be more likely to align with Beijing.
While many members of Congress still hold this view, it is not clear that the current White House does. Trump seems to believe the US need not be the pre-eminent strategic power in the Asia-Pacific, but rather can achieve American security and prosperity through hemisphere-based economic leverage rather than by influencing foreign affairs via the forward deployment of US military forces.
As for Taiwan specifically, based on his public statements Trump does not regard it as a US strategic asset. Rather, he criticizes Taiwan for “stealing” America’s semiconductor business and free-riding on US military protection. (Both accusations are unfair. US companies chose to leave most chip production to countries such as Taiwan that could do it more efficiently. There is no official US commitment to defend Taiwan, and Taipei pays for the weapons it gets from the US.)
If Taiwan is a strategic asset for other reasons, its importance as a supplier of semiconductors is additive. If not, chip supply becomes an isolated problem for US leaders, the solution of which may not involve defending Taiwan from forcible PRC annexation. Republican Party presidential candidate Vivek Ramaswamy gave voice to this idea in August 2023, when he suggested America should abandon Taiwan once it was no longer useful as a chip factory. Lutnick’s proposal was a milder rendering of this idea. Both versions have the indirect goal of undermining the silicon shield.
Is Taiwan militarily indefensible?
Finally, Washington is gravitating toward the belief that Taiwan is militarily indefensible, which means US forces would not intervene to oppose a PRC attack, Shield or no shield. Perhaps the most significant part of Lutnick’s statement about Taiwan was this:
It’s 80 miles from Mainland China. And the Chinese have said, “We’re gonna take Taiwan” – like they’re not even shy about it.
Lutnick and Trump. Photo: The White House
Lutnick was echoing a similar statement made by his boss Trump:
It’s 68 miles away from China. . . . they could just bombard it. They don’t even need to – I mean, they can literally just send shells.
In both cases, the implication is that defense is futile if the PRC decides to force annexation.
Lutnick went on to argue, “if you have 95 percent [of the world’s chip-making capacity], how am I going to get it to protect you?” He seems to mean that if China attacks Taiwan and cuts the supply of chips to the US, the American armed forces will not be able to intervene to defend Taiwan. But this doesn’t make sense.
US weapons systems do indeed rely heavily on Taiwan-sourced semiconductors. An F-35 takes 18 months to build; a Tomahawk cruise missile, two years; and a destroyer, five years. An interruption of the chips supply would affect US capacity to field new equipment months or years later, but would not prevent the US military from intervening to stop a PRC invasion attempt.
US creeping toward abandonment?
It sounds, therefore, like Lutnick is making a disingenuous argument, which reinforces the suspicion the US may not be serious about intervening.
The talk of mitigating the damage caused by a possible PRC takeover of Taiwan comes amidst other indications that the US might be creeping toward abandonment. The White House’s current pursuit of a bilateral trade deal with China raises the real possibility that the US might bargain away its support for Taipei. Trump has already reportedly withheld a $400 million package of military assistance for Taiwan in an attempt to get a better deal with Beijing.
According to the Wall Street Journal, Beijing is trying to entice Washington to change its official stance from “not supporting” Taiwan independence to “opposing” Taiwan independence. That might seem like a small change. Recall that President George W. Bush was paraphrased as saying “I’m not a nuance guy. `Do not support.’ `Oppose.’ It’s the same to me.”
Xi, however, surely appreciates the nuance.
“Not supporting” is consistent with the more basic US policy of respecting Taiwan’s right to choose its own political relationship with China, as reflected in numerous statements by US officials. By contrast, opposing independence denies Taiwan freedom of choice.
Rather than saving Taiwan to protect the chip supply, Elbridge Colby, now under secretary of defense for policy, said in 2023 that if China takes Taiwan the US should destroy TSMC facilities in Taiwan so China could not have the chips. That would make the shield worse than ineffective.
For now, the silicon shield disincentivizes a Taiwan Strait war. But as a small country threatened by a much larger close neighbor, and with the resolve of its protector perhaps weakening, Taiwan remains fundamentally insecure.
Denny Roy is a senior fellow, East-West Center, Honolulu.
