Tariffs vs. Subsidies in the Chip Industry
President Trump announced plans to impose 100% tariffs on imported computer chips unless the manufacturers are investing in production within the U.S. He stated, “if they’re building in the US, there’s no fee.”
This announcement coincided with a release of a working paper by two economists, Zhuo from the University of Michigan and Tiew from New York University. Their research suggests that the government’s focus on tariffs may be justified. By using a detailed model of the Global Contracted Semiconductor Market, they found that even a modest import duty of 10% could sway production decisions towards the U.S., making it more favorable than establishing similar manufacturing plants in Taiwan.
In contrast, capital subsidies and tax credits from President Biden’s Chips Act, while beneficial, do not show the same effectiveness. U.S.-based factories are predicted to be $1.6 billion to $1.8 billion less profitable over their lifetimes compared to overseas facilities, mainly due to higher labor, construction, and operational costs.
The research highlights challenges for the grant-first approach supported by the Biden administration, pointing out that market access plays a more crucial role in determining industry location than just financial assistance. This finding raises questions about the common belief among economists that subsidies always outperform tariffs.
The Challenge of Competing in Chip Manufacturing
Zhuo and Tiew simulated a U.S. version of Fab 14, a real facility operated by Taiwan Semiconductor Manufacturing Company (TSMC). The simulation revealed that the U.S. site would incur $1.2 billion in higher capital costs compared to Taiwanese facilities, with subsidies only partially alleviating the disadvantages.
Even with a 10% construction subsidy and a 25% investment tax credit, profitability for the U.S. facility was decreasing. Additional measures like banning advanced chip sales to China and export restrictions affected overall profits but did not fundamentally change the preference for overseas production.
This suggests that solely relying on subsidies cannot overcome the inherent disadvantages of certain locations. Without implementing more comprehensive policy measures, companies are likely to continue favoring international construction.
Impact of Tariffs on Domestic Production
An intriguing outcome of the paper is the effect of simulating a 10% tariff on imported chips. This relatively minor obligation—less than the 100% tariffs proposed by Trump—could boost demand for U.S. facilities, especially as foreign competitors become more aggressive.
With tariffs in place, U.S. fabs could become profitable by approximately $5.9 billion over their lifetime, reversing previous losses and benefiting sectors like defense, automotive manufacturing, and industrial machinery.
While the authors don’t delve into the broader economic implications of tariffs, they emphasize incentives that influence chip makers’ location decisions. Among various industrial policy strategies, customs policies are seen as the most effective lever for altering investment considerations.
Tariffs vs. The Chips Act
This paper provides academic support for a broader political shift. The Chips Act, which received bipartisan backing in 2022, has pledged over $50 billion to revitalize the U.S. semiconductor industry. Major companies, including Intel, Samsung, and TSMC, have announced expansions in the U.S., but some initiatives have slowed down due to rising costs and labor shortages.
The Biden administration often points to these announcements as proof that public investment can revitalize high-tech manufacturing. However, the slower-than-expected rollouts raise doubts about whether grants alone can facilitate significant structural changes.
Meanwhile, Trump has advocated for a more aggressive industrial policy that focuses on restructuring global trade rather than solely promoting private investment. While the Zhuo-Tiew paper doesn’t directly endorse this strategy, it provides a quantitative basis for Trump’s assertion that tariffs could be a more straightforward and effective tool in restoring domestic production capacity.
The key takeaway from the research is that location decisions are sensitive to relatively small changes in trade policy. This implies that even moderate tariffs could more swiftly influence global manufacturing trends than subsidies.
However, it’s still unclear whether Trump’s proposed 100% tariffs would yield similar results, as the study didn’t explore such extreme obligations. The potential impact on international supply chains and global semiconductor prices remains uncertain.
Nonetheless, the timing of the paper’s release alongside the new tariff announcement could shift discussions about how to bring semiconductor manufacturing back to the U.S. It suggests that public investment alone won’t suffice to shift corporate preferences, highlighting the importance of a well-structured customs policy in U.S. economic strategy.
