On March 3, the Taiwan Semiconductor Manufacturing Company (TSMC) announced, together with U.S. President Donald Trump, its decision to invest massively in the United States. Per TSMC CEO C.C. Wei, this investment will direct $100 billion to the construction of three new fabrication facilities featuring the company’s most advanced process nodes, two advanced packaging plants, and a research and development center in Arizona. The construction process will take place over the coming years, reportedly to counter potential tariffs, ranging from 25 to 100 percent, that the Trump administration might impose on Taiwan. The investment plan will bring TSMC’s critical semiconductor production closer to the company’s U.S. clients.
Commenting on the agreement, Trump declared, “We must be able to build the chips and semiconductors that we need right here. It’s a matter of national security for us.”
This Taiwan-U.S. investment plan aligns with the broader U.S. industrial policy objectives of reshoring vital supply chains and reducing dependency on semiconductor chokepoint economies in Asia. These objectives are especially relevant to Taiwan, given its dominant position in the semiconductor field and escalating tensions between the two sides of the Taiwan Strait.
In one sense, the investment decision reflects the TSMC’s adaptability, and other Asian companies such as South Korea’s Samsung and LG are also reportedly considering moving plants to the United States. However, such moves transcend their face value, as they raise fundamental challenges to local regulatory frameworks concerning foreign investment reviews, core key technology protection, and, particularly in the case of Taiwan, national security law. Indeed, some have characterized TSMC and its surrounding ecosystem as the foundation of Taiwan’s “Silicon Shield.” Although many assess this narrative differently, TSMC’s massive expansion in the United States has raised concerns about Taiwan’s ability to maintain its technological supremacy, strategic centrality, economic prosperity, and national security amid shifting geopolitical dynamics.
Lessons From the 1980s Japan-U.S. Semiconductor Saga
The strategic importance of the semiconductor industry is undeniable, and this is not the first time that the U.S. government has acted so aggressively to intervene in and reshape this critical sector. A similar moment occurred in the 1980s, when the Reagan administration, under mounting pressure from U.S. semiconductor firms, took decisive action against Japan in response to its growing influence in the industry.
While the United States, at the time, was still the world’s dominant manufacturer generally, Japanese firms managed to catch up to and surpass their U.S. counterparts in the specific field of memory-chip production. Between 1978 and 1986, Japanese firms essentially dominated the production of dynamic random-access memory (DRAM) chips – then the most popular type. The U.S. global market share fell from 70 percent to 20 percent, while Japan’s surged from 30 percent to 75 percent (though interestingly, Japan’s share of the U.S. market remained insignificant during this period).
Washington, then still committed to neoliberalism, initially hesitated to act against Japan. However, the decline of U.S. DRAM chip manufacturers raised national security concerns in both economic and military terms, which were amplified by industry lobbying. Under pressure, the Reagan administration abandoned laissez-faire policies for protectionism, culminating in the 1986 Japan-U.S. Semiconductor Agreement. This deal was hardly reciprocal: it imposed market access requirements, managed production, and price controls, partly backed by threats of antidumping and Section 301 investigations. Despite its asymmetry, it marked an example of East Asia-U.S. intergovernmental negotiation to strike policy arrangements in the semiconductor industry through dialogue and coordination.
Problems With Informal, Ad Hoc, Company-Specific “Silicon Statecraft”
Trump’s interventionist approach to the global semiconductor industry, framed under national security and “America First,” to a certain extent echoes Reagan’s “Managed Trade” playbook from decades past. Like Reagan, Trump responded to foreign dominance in critical technology by adopting an aggressive governmental intervention involving unilateral tariffs. Both administrations viewed technological leadership as essential to United States economic resilience and military supremacy.
However, a key distinction between the two lies in how each administration attempted (or is attempting) to reconfigure a key tech industry. Under Reagan, the reconfiguration proceeded chiefly through government-to-government negotiations and formal agreements, backed by tools like antidumping measures and tariffs. The framework expanded U.S. market access while curbing Japanese firms’ dominance domestically and abroad, benefiting U.S. chipmakers.
By contrast, the Trump administration has largely worked outside of the intergovernmental model. The Trump administration bypassed the Taiwanese government, directly engaging TSMC to reshape the global semiconductor supply chain – a move with potentially lasting implications. Unlike the Japanese government in the 1980s, today’s Taiwanese government has been notably absent from the discussions. Indeed, not until days after the TSMC investment had been agreed upon did Taiwan’s President Lai Ching-te, accompanied by C.C. Wei, hold a joint press conference providing a belated explanation mainly to the Taiwanese public regarding the matter.
The delayed response highlighted a lack of transparency, as the Taiwan government’s role remains unclear. This is especially concerning given that the National Development Council, a major TSMC shareholder with board representation, should have been aware of such a significant decision. Beyond the mixed and sometimes conflicting information, TSMC’s decision and the government’s alleged prior knowledge do not align with the outbound investment approval process overseen by the Ministry of Economic Affairs’ Department of Investment Review. One point is almost certain: TSMC’s investment announcement at the White House left little, if any, room for proper review by Taiwanese authorities after the fact.
This “silicon statecraft” is troubling for several reasons. For one, the process’ informal, ad hoc, and company-specific nature exploits the power asymmetry between the U.S. government and foreign corporations. The Trump–TSMC dynamic is arguably unprecedented and carries significant implications beyond Taiwan and the semiconductor industry. Consider, for example, the implications for Panama Canal ports or SoftBank. Directly “negotiating” (for lack of a better term) with the U.S. president – who publicly threatens tariffs and other punitive measures – over the reported TSMC-Intel joint venture puts TSMC at a severe disadvantage. TSMC lacks the bargaining power and political leverage a government would wield in high-stakes negotiations.
Taiwan’s government also faces constraints in navigating the complexities of China-U.S. relations under Trump 2.0, especially given the lack of formal diplomatic relations, which complicates any government-to-government interaction. Still, bypassing formal channels strips away even the limited protections and diplomatic leverage intergovernmental engagement could offer. Conversely, East Asian governments like Japan and South Korea have engaged directly with the United States on projects such as Alaska’s natural gas pipeline, underscoring the state-to-state dialogue absent in Taiwan’s case.
To some extent, the arrangement may make long-term business sense for the TSMC as it considers operations diversification, energy and water resilience, workforce shortages, and overall production capacity. That said, concerns about tariffs reportedly drove TSMC to embrace the investment arrangement; however, there is no guarantee that the Taiwanese chip industry will be free from future U.S. tariffs or other economic pressures. This absence of a guarantee leaves the industry vulnerable to shifting U.S. policies.
Worse still, the financial incentives previously granted to TSMC under the CHIPS and Science Act of the Biden administration might no longer be available, further limiting TSMC’s ability to offset the considerable costs of its U.S. expansion. There are, theoretically, legal avenues that the TSMC can pursue to challenge the possible revocation of promised benefits under U.S. law, but in practice, political pressure could make such a move unfeasible. These concerns will likely be compounded if TSMC ends up having trouble recovering from sunk costs and lost opportunities, particularly if the company’s planned expansion fails to materialize or encounters unexpected obstacles such as a shortage of experienced engineers or cultural clashes between Taiwanese management and local U.S. staff.
Paying attention to all of these potential pitfalls are Japan, South Korea, the Netherlands, and other key players in the global semiconductor supply chain. The broader concern is that if Trump’s unilateral, strategic moves prove effective in the case of Taiwan, they will only reinforce the belief that tariffs are an effective and even legitimate tool to pressure trading partners generally. Even more troubling is the precedent this informal, ad hoc, and company-specific silicon statecraft sets: if Trump can selectively target specific companies and pressure them directly rather than engage their home governments, the global balance of economic power may shift unhealthily toward Washington.
TSMC’s proposed expansion in Arizona presents no evident trade-off for Taiwan’s socio-economic and political costs – there are no U.S. concessions, no economic incentives, and certainly no security assurances. As for such guarantees, obviously, they should rest on far more than a Taiwanese firm’s commitment to investing in the U.S. Southwest. Indeed, even through the lens of the hard-bargaining transactional worldview of Trump, Taiwan’s dominant role in semiconductor supply chains should at least provide it with some leverage to have a dialogue to secure meaningful reciprocal benefits, particularly robust security assurances. In the TSMC saga, however, there appears to be little, if any, reciprocity or strategic benefit for Taiwan as a whole, other than a possible (but not guaranteed) short-term avoidance of coercive tariffs.
The Merits of a Taiwan-US Semiconductor Agreement
The interests of U.S. firms are deeply intertwined with TSMC and the broader semiconductor ecosystem in Taiwan. Hence, the success or failure of TSMC’s expansion in the United States will have significant implications not only for Taiwan but also for U.S. companies. Any disruption to TSMC – whether through the sudden imposition of tariffs or other measures – would inevitably distress Nvidia, Apple, Qualcomm, and similar U.S. firms closely partnered with the TSMC, not to mention the defense industry. The Trump administration should carefully factor these potential costs into its semiconductor policy decisions.
As for the Taiwanese government, its relative absence from the TSMC arrangement, though perhaps superficially convenient to certain elements of the U.S. government and to the TSMC itself, will, in all likelihood, undermine the rules-based international system generally and, more specifically, place great strains on U.S. credibility, U.S. relations, and U.S. interests as they pertain to decades-long alliances in the Indo-Pacific. Many in this region may reasonably draw parallels between the TSMC saga and the recent upheaval in U.S. policy toward Ukraine. The end result will be abiding doubts as to the reliability and consistency of strategic partnerships with the United States, a country that crafted the post-1945 global order.
The informal, ad hoc, and company-specific silicon statecraft that has characterized the TSMC saga should cede place to a more formal, inter-governmental approach that includes the Taiwanese government, the U.S. government, Taiwanese firms, and U.S. firms.
Unlike U.S. and Japanese high-tech firms operating in the 1970s and ‘80s, U.S.-led global supply chains and Taiwan’s semiconductor industry are largely complementary. This fact makes cooperation between the two a natural and mutually beneficial path forward. Be it hard law or soft law, a Taiwan-U.S. semiconductor agreement would help ensure that clear, enforceable rules govern intellectual property protection, regulatory harmonization, supply chain resilience, energy resilience, technology transfer protocols, workforce shortage solutions, healthy talent flows, and dispute resolution mechanisms driven by coordination between policymakers and industry leaders on both sides. Improvements in each of these areas would help the U.S. tap into and benefit from Taiwan’s central role in the global semiconductor supply chain.
More crucially, such an agreement could streamline the inbound and outbound investment reviews and approval processes on both sides, ensuring greater policy coherence and predictability. Notably, in the TSMC case the Taiwanese government arguably had to regauge its overseas investment restrictions at the cost of technological leadership.
An agreement would also promote the idea of “reciprocity and fairness” that Trump has constantly championed. For example, from the U.S. perspective, Taiwan’s industry-development fund and regulatory streamlining support for TSMC’s partner companies will aid in establishing a comprehensive semiconductor ecosystem in the United States. And from the Taiwanese perspective, a well-structured public–private, Taiwan-U.S. partnership would help prevent future incarnations of the current TSMC saga, in which the Taiwanese government has found itself unpleasantly caught between pressures from domestic stakeholders and external demands from Washington.
A formalized bilateral framework would hopefully enable stronger Taiwan-U.S. coordination on export controls, addressing the enforcement limitations of ECRA/EAR’s extraterritorial reach – as evidenced by ongoing restricted chip trafficking through various channels. Such structured cooperation would improve enforcement effectiveness, serve as a model for international coordination, and extend mutual benefits beyond semiconductor production to address shared challenges in onshoring costs, infrastructure, energy security, supply chain resilience, and AI governance. Also, through the bilateral semiconductor agreement, Taiwan’s government may help reduce compliance burdens on smaller local firms by enabling regulatory support and resources to level the playing field, thus more effectively achieving mutual policy goals.
Such an agreement may seem idealistic, but it ultimately rests on both governments’ political will and wisdom to reconsider and prioritize long-term strategic stability over short-term maneuvering. The institutionalization of the Taiwan-U.S. semiconductor partnership, even at a minimal level, would have far-reaching geopolitical implications. It would send a clear signal to U.S. allies – including South Korea, the Netherlands, and Japan – that they are not merely utilitarian suppliers of high-end technology to the United States but essential strategic partners in global semiconductor governance. Consequently, the U.S. would enjoy greater supply chain resilience, fewer unilateral vulnerabilities, and a reasonably balanced, strategic alternative to fragmented, deal-by-deal negotiations.
As we attempt to discern a path forward in Taiwan-U.S. semiconductor relations, Reagan’s words from that 1987 Japan-U.S. semiconductor agreement offer a telling contrast to the recent events surrounding the TSMC: “We said from the very beginning that when they [the Japanese government] returned to abiding by the agreement that we thought we had, we would lift the sanctions…. And the amount of tariff that we have removed is just proportionate to the extent that they have so far returned to abiding by the agreement.”
This structured framework – where policy-driven, proportionately guided government-to-government negotiations were paramount – differs starkly from the informal, ad hoc, company-specific silicon statecraft embraced by the second Trump presidency. Reagan’s acknowledgment that “there are people in Japan, like Prime Minister Nakasone, who have worked very hard” underscores the value of engaging, not bypassing, governmental counterparts.
The intensification of semiconductor geopolitics shows no signs of relenting. But perhaps there is wisdom in revisiting not just Reagan’s policies but how he and his cabinet dealt with U.S. allies diplomatically: an approach that balanced firmness with fairness and unilateralism with some level of reciprocity, all within an institutional framework that deserves to stand the test of time.
The authors would like to thank Jieh-Min Wu and Hui-Heng Hong for their comments on an earlier draft of this article.