Taiwan pushes back on US plan for equal chip split over economic concerns

A US proposal that Taiwan produce only half of America's semiconductors has proved controversial in Taiwan.

When US Secretary of Commerce Howard Lutnick suggested in late September that Taiwan and the United States should each manufacture 50% of the semiconductors consumed in the American market, it instantly reverberated across both shores of the Pacific. The remark was considered a step towards reducing America's dependence on Taiwan's concentrated chipmaking prowess.

But in Taipei, the notion was met with skepticism, unease, and even outright rejection. Opposition Kuomintang (KMT) party lawmaker Hsu Yu-chen (許宇甄) called Lutnick's proposal exploitative.

The rationale in Washington

At its core, the “50-50” concept represents Washington's latest aim to recalibrate the geography of semiconductor supply chains. For decades, Taiwan has been an indispensable workshop of the digital economy, with TSMC currently commanding over 90% of the world's most advanced chip production.

This concentration has long been touted, by both Taiwanese and those from around the globe, as a “Silicon Shield,” a deterrent against Chinese aggression, since Taiwan's centrality renders disruption catastrophic for global markets. 

Lutnick said a “Big deal” with Taiwan could materialize soon.

Vice Premier of Taiwan Cheng Li-chiun (鄭麗君), who was responsible for the trade negotiations between Taiwan and the US, said that Taiwan did not make such a commitment concerning this issue. 

Taiwan's reactions  to the proposal

In Taiwan, however, the reception was decidedly frosty. Lawmaker Hsu Yu-chen from the major opposition party Kuomintang (KMT) castigated Lutnick's comments as tantamount to hollowing out Taiwan's strategic asset. She warned that Taiwan's leverage in international security could be eroded due to this initiative.

Another KMT lawmaker, Huang Chien-hao (黃健豪), said that establishing robust chip production in the US could require as much as US$500 billion in investment, with a significant share of the equipment inevitably sourced from Taiwan. He said from an economic perspective, Washington should view any short-term trade deficit not as a liability that warrants punitive tariffs, but as an investment.

In mid-October, former deputy minister of Taiwan's Ministry of Foreign Affairs (MOFA) Roy Lee Chun (李淳) stated on a podcast that the 50-50 proposal would be counterproductive for the long-term goals of the US. It may be a negotiation tactic rather than an unyielding goal for the US, he added

Aside from politicians, economists in Taiwan are also skeptical of the idea. Many swiftly pointed out the monumental obstacles embedded in the 50-50 vision. For instance, Liu Pei-chen (劉佩真), an economist at the Taiwan Institute of Economic Research (TIER), argued that the proposal is not only technically arduous but also economically challenging.

She told Taiwan's CNA that costs for labor, plant construction, and regulatory compliance in the US are substantially higher than in Taiwan, and recent increases in visa fees have further inflated the expense of deploying skilled engineers on American soil. Such factors would, she argued, render the quest for a  “50-50” partition fraught with difficulties in cost, efficiency, and workplace integration.

For Taiwan, she said the greater challenge would lie in safeguarding its technological primacy under any prospective agreement, and ensuring that intellectual property and operational control remain firmly anchored at home even as overseas fabs come online.

Balancing the demands of sustaining next-generation R&D onshore while simultaneously supplying elite talent to foreign facilities, Liu added, will test the industry's capacity to allocate its scarce resources wisely.

Eric Wu (吳金榮), president of Taiwan's fabless electronics company Explore Microelectronics, wrote in Business Next that TSMC's first Arizona fab has a monthly capacity of around 20,000 wafers at the 4-nanometer node, while the second could reach 30,000 wafers per month.

He argued that even if all six TSMC US sites scaled to about 30,000 wafers each, their combined output would reach only 30% of TSMC’s projected advanced-node capacity in Taiwan by 2026.

Moreover, Wu pointed out in the same article that enthusiasm among American workers for front-line semiconductor production remains limited, particularly in positions that demand grueling shift work. In essence, he concluded that the idea of a 50-50 chip split was neither feasible nor practical.

Strategic implications

For Taiwan, the 50-50 debate raises a stark question: Will it retain its semiconductor dominance, or be forced into a shared model that erodes its strategic edge?