While most analysts see the relocation of the chips industry to America as highly unlikely in the short term, Trump’s emphasis on reshoring reflects his transactional approach to trade and security.

A major pillar of U.S. President Donald Trump’s economic policy is the reshoring of what he considers to be heritage industries. The case par excellence is semiconductor manufacturing.

Having accused Taiwan of building its foundry model on the back of American know-how, Trump in August reiterated plans to impose steep tariffs on imported chips, pressing allies and partners to expand U.S. production.

Supporters of the policy argue that such pressure is necessary to restore U.S. technological leadership and manufacturing jobs, while also addressing vulnerabilities exposed during the Covid-19 pandemic and U.S.-China trade frictions. Critics counter that tariffs and pressure tactics are more politically motivated than practical, dismissing promises to “bring back” a wide range of factory jobs as unrealistic in an age of automation and AI. Even if production does return to the United States, they note, the jobs are likely to be fewer than advertised, harder to fill, and often low-paying.

Regardless, these tactics, including hefty tariffs, appear to have borne some fruit.

At a White House event in March with Taiwan Semiconductor Manufacturing Co. (TSMC) CEO C.C. Wei, Trump announced that the company, the world’s leading chipmaker and source of more than 90% of the most advanced semiconductors, would increase its Arizona investment by US$100 billion. The sum comes on top of the US$65 billion that TSMC had already committed to its first three Arizona fabs.

In a sign that its reshoring policy is more than just Presidential grandstanding, the Trump administration confirmed on August 22 that the government would take a 10% stake in Intel under a deal that converts government grants into equity. The U.S. administration is weighing whether to take equity stakes in other companies receiving funds under the 2022 CHIPS and Science Act, but has no similar plans for large companies that are expanding their U.S. investments, a White House official told Reuters in August. TSMC and Micron would fall into that category.

Despite vowing to dismantle the CHIPS Act established under his predecessor Joe Biden, Trump has quietly gone along with most of its provisions. In fact, both administrations have pursued similar goals, albeit with different rhetoric — Biden through the CHIPS Act’s incentive framework, and Trump through a mix of tariffs, pressure, and selective use of change to CHIPS funding mechanisms.

While Trump has insinuated that Taiwan built its foundry model on U.S. expertise, industry experts emphasize that the foundry concept — exclusively making chips on a contract basis for customers to incorporate in their own proprietary technology products — itself originated in Taiwan. In fact, as the historian of information technology Tinn Honghong has noted in her seminal work Island Tinkerers, TSMC founder Morris Chang approached Intel and other U.S. chip giants for investment in a foundry venture in the mid-1980s, only to be rejected. The foundry model was generally considered unworkable at the time.

More to the point, it is unclear just how serious Trump is about returning the United States to prominence in the semiconductor industry and what the effort would look like in practice.  

“I think a lot of this is empty talk,” says Bono Shih, a postdoctoral researcher in engineering ethics at Penn State University. He says the wholesale transplantation of the chips industry to the United States is unfeasible, or at the least would take many years to accomplish. Taiwan began producing 3-nanometer wafers in 2022, while the Arizona plant is not expected to reach that milestone until 2028. Shih says those hard facts should cut through the political rhetoric.

“Words will not change how the technology works,” says Shih, coauthor of Engineers and the Two Taiwans. “The basic science will not change, and so the reality remains the same.”

Others are not so sure. Having been at the forefront of the semiconductor industry in the United States and Taiwan for decades, Yang Guang-lei, also known as Konrad Young, has an insider’s insight into industry conditions, working cultures, and developments in both Taiwan and the United States. Although the United States is often depicted as a waning power, Yang believes Washington’s ability to shape the global agenda on cutting-edge technology remains undimmed.

He cites a 2024 report from the Boston Consulting Group — “the best publicly available report on the [semiconductor] industry” — as revealing the scale of U.S. ambitions. “They call it a forecast, not a goal,” he emphasizes. “But they plan to have a 28% share of advanced logic chips — under 10 nanometers — by 2032.”

Logic chips, used to perform calculations and processing, can be thought of as the “brains” of electronic devices. The Boston Consulting report projects that the United States, which now has negligible market share for these chips, will account for nearly a third of global capital spending in this market segment by 2032 — more than triple what it would have secured had investment continued at the pre-CHIPS Act pace.

“In consequence, we will see meaningful shifts in the global distribution of chipmaking capacity,” the report reads. It goes on to predict the diversification of wafer fabrication capacity “markedly beyond Taiwan and South Korea to include the U.S., Europe, and Japan.”

This trend, says Young, is the thinking behind the U.S. government’s outreach to TSMC. “How do you make it from zero to 28% when everybody is growing?” he says. “The best way is to get TSMC over there.”

Transfer barriers

At the same time, replicating Taiwan’s advanced chip manufacturing in the United States is likely to take much longer than the U.S. president imagines, says Ian Chong, an associate professor of political science at the National University of Singapore. He cites the highly specialized, sophisticated, and integrated manufacturing ecosystem created over decades in Taiwan. 

“When you create a wafer fab, it’s not like an apartment block when you can just come up with these prefabricated bits and get it up in a few months to a year,” says Chong, who is also a nonresident scholar at Carnegie China. Adding to the challenge are serious environmental concerns such as pollution control, maintaining a dust-free environment, and assuring sufficient water supply. “Those kinds of things take time to set up,” he notes.

This view is widely supported by other analysts, including Island Tinkerers author Tinn, who highlights the highly technically skilled workforce Taiwan has cultivated to ensure the uppermost standards in its foundry model.  

“Taiwan’s success is in part down to its quality control engineers, who help with maintaining the functioning of the factory,” she says. She questions whether this capability can be reproduced in the short-term outside Taiwan.  

Another potential problem is the ability to attract top talent from Taiwan to help ensure that the U.S. operations are up to standard. Despite President Trump’s assertions, a career in the United States is no longer the draw it once was. As outlined in Engineers and the Two Taiwans, the cross-Pacific struggle to compete for skilled workers has persisted for decades.

Once Taiwan began to gain competitive advantage in the 1990s, a divide arose between U.S.-based professionals and those who chose to remain in or return to Taiwan. “I think many Taiwanese engineers would be reluctant to work in the U.S. right now,” says Chang Kuo-chi, who coauthored Engineers and the Two Taiwans with Bono Shih and Virginia Tech professor Gary Downey.

“Since the 2000s, local engineers have become more influential than in the past,” says Chang, a professor and research fellow at National Taiwan University’s Institute of National Development. He indicates that a move to America might nowadays be perceived as a step down on the career ladder for some engineers and that ample remuneration would be required as an incentive to counter that view. Lifestyle and cultural differences could also play a part.

“TSMC’s engineers in Arizona might face challenging issues in their daily and professional lives,” he says. He notes that when the company was setting up in Arizona, local unions objected to certain of its working practices and management bemoaned the shortage of qualified personnel.

However, by far the biggest concerns for Taiwan are the implications for national security. While not an absolute protection, many observers have regarded the near-total global reliance on TSMC’s advanced semiconductors as a form of “silicon shield.” Disruption would pose serious problems for U.S. and European tech companies — and their national economies.

Taiwanese opposition politicians and other critics of TSMC’s U.S. investments have suggested that the move risks lessening the American incentive to come to Taiwan’s aid in case of a military attack. But TSMC and Taiwan government officials have stressed that the companies’ most cutting-edge technologies and R&D activities will remain rooted in Taiwan.

As a result, the Arizona buildup needn’t put a much of a dent in that silicon shield, at the same time as it helps alleviate Washington’s sense of vulnerability about the potential impact on chip supply of any cross-Strait conflict.

Given the President Trump’s mercurial and transactional approach to policymaking, however, many analysts conclude that nothing can be taken for granted. “The actuality is that [the U.S.-Taiwan relationship] is pretty much mutually beneficial,” says Chong. Yet Trump appears to regard every interaction as a zero-sum game. “The actuality is that [the relationship] is pretty much mutually beneficial,” says Chong. “But if Trump sees the other side is somehow doing well, it suggests that perhaps the United States is losing,” he says. In the best-case scenario, Trump may well just be laying out extreme positions to get more than what he started with, says Chong, rather than preparing to sell Taiwan down the river. For her part, Tinn is optimistic that the “long-term collaborative relationship” that has defined connections between the U.S. and Taiwan tech sectors will continue.