Securing Semiconductor Supply Chains

Global supply chains, particularly in technologies of strategic value, are undergoing a remarkable reevaluation as geopolitical events weigh on the minds of decisionmakers across government and industry. The rise of an aggressive and revisionist China, a devastating global pandemic, and the rapid churn of technological advancement are among the factors prompting a dramatic rethinking of the value of lean, globally distributed supply chains.

These complex supply networks evolved over several decades of relative geopolitical stability to capture the efficiency gains of specialization and trade on a global scale. Yet in today’s world, efficiency must be recast in terms of reliable and resilient supply chains better adapted to geopolitical uncertainties rather than purely on the basis of lowest cost.

Indeed, nations worldwide have belatedly discovered a crippling lack of redundancy in supply chains necessary to produce and distribute products essential to their economies and welfare, including such diverse goods as vaccines and medical supplies, semiconductors and other electronic components, and the wide variety of technologies reliant on semiconductors. A drive to “rewire” these networks must balance the manifest advantages of globally connected innovation and production with the need for improved national and regional resiliency. This would include more investment in traditional technologies—for example, a more robust regional electrical grid in Texas, whose failure contributed to the supply disruption of automotive chips that Abigail Berger, Hassan Khan, Andrew Schrank, and Erica R. H. Fuchs describe in “A New Policy Toolbox for Semiconductor Supply Chains” (Issues, Summer 2023).

Of course, given its globalized operations, the semiconductor industry is at the forefront of these challenges. In particular, there is a need to distribute risks of single-point failures, such as those found in the global concentration of semiconductor manufacturing in East Asia. Taiwan and South Korea, which together account for roughly half of global semiconductor fabrication capacity, sit astride major geopolitical and geological fault lines, with the dangers of the latter often underestimated.

Recent investments to renew semiconductor manufacturing capacity in the United States are a key element of this rewiring. Through the CHIPS for America Act of 2021, lawmakers have authorized $52 billion to support restoring US capacity in advanced chip manufacturing, with $39 billion in subsidies for the construction of fabrication plants, or “fabs,” backed by substantial tax credits, and roughly $12 billion for related advanced chip research and development initiatives.

Berger and her colleagues argue cogently that it may also be possible to design greater resiliency directly into semiconductor chips. In some cases, greater standardization in chip architecture may allow some chips to be built at multiple fabs, reducing “foundry lock-in.” Such gains will depend on trusted networks among multiple firms as well as governments of US allies and strategic partners—although sorting the practical realities of commercial and national competition in a rapidly innovating industry that marches to the cadence of Moore’s Law will be challenging. The authors rightly point out that focusing on distinct market segments with similar use cases may offer win-win opportunities, but these, too, will require incentives to drive cooperation.

It is clear that global supply chains need a greater level of resiliency, not least through greater geographic dispersion of production across the supply chain. But whether generated by greater standardization, stronger trusted relationships, or through the redistribution of assets, the continued national economic security of the United States and its allies depends on a comprehensive, cooperative, and steady implementation of this rewiring. The authors propose a novel approach that should be pursued, but the broader rewiring will not happen quickly or easily. We still need to move forward with ongoing incentives for industry, more cooperative relationships, and major new investments in talent. We are not done. We need to think of semiconductors like nuclear energy, one involving sustained and substantial commitments of funds and policy attention.

Sujai Shivakumar