Ranked by citation count.
The debate
Daniel: TSMC produces over 90 percent of the world's most advanced chips on a single island 100 miles from a hostile power. The RAND Corporation ran a wargame on this—the US had no good options in any scenario. That is the definition of a critical national security vulnerability.
Kevin: I understand the concern, but this concentration exists for a reason. Semiconductor fabrication is the most complex manufacturing process on Earth. TSMC spent decades building expertise that cannot be replicated by writing a check. The CHIPS Act invested 52 billion dollars and we are still years from producing cutting-edge chips domestically.
Daniel: That is exactly the problem. We have created a dependency we cannot quickly escape. Farrell and Newman's work on weaponized interdependence shows that when critical network nodes concentrate in one location, they become chokepoints that adversaries can exploit. Taiwan is a chokepoint with a Chinese missile pointed at it.
Kevin: But that same interdependence cuts both ways. China imports over half of Taiwan's semiconductor exports. A blockade or invasion would devastate China's own tech industry and economy. The economic cost of disrupting TSMC is so catastrophic for all parties that it creates a powerful deterrent.
Daniel: Academic analysis of the Silicon Shield theory finds it is a hopeful narrative, not a credible deterrent. History shows that nations go to war against their economic interests all the time. Russia invaded Ukraine despite massive economic costs. We cannot bet national security on the assumption that rational economic calculation will prevail.
Kevin: Fair point about the limits of deterrence. But the proposed solutions have their own problems. Full self-sufficiency would cost over a trillion dollars and raise chip prices by 35 to 65 percent. That would slow AI development, weaken competitiveness, and ultimately harm national security in different ways. The efficient global supply chain is itself a source of strength.
Daniel: Nobody serious is proposing full autarky. The goal is strategic diversification—building enough allied capacity in the US, Japan, South Korea, and Europe so that no single disruption is catastrophic. TSMC's Arizona fab, the Japanese facility, the Dresden plant—these are steps in the right direction, but they will take years to reach scale.
Kevin: And during those years, the existing system continues to function. The vulnerability is real but manageable. Overreacting with aggressive reshoring could damage the US-Taiwan relationship, drive up costs, and potentially weaken the very deterrent that has kept the peace for decades.
Daniel: The vulnerability is not manageable if it cannot survive a single bad month. A Chinese quarantine of Taiwan before 2027—before new fabs come online—would cripple every industry from defense to healthcare. We are in a race against time and the evidence says we are behind.
Kevin: Then the priority should be accelerating diversification while strengthening military deterrence in the Taiwan Strait, not pretending we can quickly replace 50 years of Taiwanese manufacturing excellence. The vulnerability is real, but the solution requires patience and partnership, not panic.
Frequently asked questions
Why is the world so dependent on Taiwan for chips?
Taiwan, mainly through TSMC, produces over 90 percent of the world's most advanced semiconductors. This geographic specialization has delivered large efficiency gains but concentrates production in one place.
What is the Silicon Shield theory?
It is the idea that Taiwan's semiconductor dominance deters Chinese aggression because the global economy depends on it. An academic analysis concluded the idea lacks rigorous grounding and functions more as a hopeful story than a credible deterrent.
Could the US make all its own advanced chips instead?
Evidence suggests full self-sufficiency would be very costly. One estimate cites $1 trillion in upfront costs and 35-65 percent price increases, and research on US manufacturing decline points to workforce and infrastructure gaps that cannot be rebuilt quickly.
What is being done to reduce reliance on Taiwan?
The CHIPS Act and allied investments aim to diversify chip production across the US, Japan and Europe. The period before these new fabs become operational may be the window of greatest vulnerability.