President Biden signed legislation in October exempting semiconductor projects from review under the National Environmental Policy Act. While the new Building Chips in America law will accelerate fab construction, and early Taiwan Semiconductor (TSMC) yields in Arizona are promising, the United States must also reform its high-skill immigration policies to stay on semiconductor manufacturing’s cutting edge.
Even if every environmental regulation were repealed, chipmaking firms in the U.S. would struggle to produce their wares at scale and over the long term, because the country lacks the skilled workforce to sustain such production. TSMC founder Morris Chang shared that assessment as the firm flirted with big investments in the U.S. ahead of CHIPS’s 2022 passage. “The United States stood out for cheap land and electricity when TSMC looked for an overseas site,” Chang said in an April 2021 symposium address, “but we had to try hard to scout out competent technicians and workers in Arizona because manufacturing jobs have not been popular among American people for decades.”
Affirming Chang’s point, the Semiconductor Industry Association anticipates a shortfall of more than 40,000 workers in engineering, computer science, and other high-skill chipmaking subfields in the United States by 2030. America’s education and technical-training systems likely would generate a viable labor supply in the long run if demand endures, but companies need skilled workers now. With China breathing down Taiwan’s neck and more than $50 billion in CHIPS Act funding heading out the door, the U.S. should court high-skill immigrants to meet the industry’s labor needs and fulfill the mission of de-risking its semiconductor supply chain.
One intuitive idea to plug this workforce gap is to make an explicit new immigration allowance for semiconductor-industry workers. The Economic Innovation Group in 2023 proposed “Chipmaker’s Visas,” which would be auctioned to industry firms and could be used to hire temporary foreign workers with experience in high-skill industry subfields.
The best argument in favor of such a policy is its likely bipartisan appeal. The CHIPS Act and the Building Chips in America Act proved that a broad swath of the legislature conceives of semiconductor manufacturing as a critical national interest. Many lawmakers would plausibly support a narrow immigration reform to further the CHIPS Act’s goals, even if doing so furthers sector-favoritism that rankles Hayekians.
The counterargument to creating a Chipmaker’s Visa whole-cloth is that existing immigration law, pending small regulatory adjustments, offers simpler solutions. As Institute for Progress senior immigration fellow Jeremy Neufeld explained to me, “there is a lot of untapped potential in some of the existing pathways.”
One available pathway Neufeld sees as ripe for expansion is the Immigration and Nationality Act of 1965’s Schedule A pre-certified employment category. While the INA broadly tasks the Secretary of Labor with ensuring that foreign workers do not displace qualified American workers, Schedule A excepted a list of occupational categories in which the American workforce was insufficient. Initially, the exemption covered foreign workers in categories such as engineering, physics, chemistry, and pharmacology. Lawmakers expected the Labor Department periodically to examine changes in the composition of the labor force and adjust its pre-certified categories accordingly.
The Labor Department, however, failed to uphold its end. Since 1991, the list has remained untouched and has included only two categories: physical therapists and nurses. To its credit, the Biden administration has tried to make amends. In December 2023, it announced that it would consider revisions to Schedule A to include occupations in the science, technology, engineering, and mathematics fields. This is an idea the incoming Trump administration should likewise pursue.
Another expandable visa category, Neufeld argues, is for workers at companies affiliated with American universities, who are thus exempt from the standard H1-B visa cap of 85,000. If the U.S. Customs and Immigration Service provided more clarity on qualifying affiliation status, companies involved with, for example, the National Semiconductor Technology Center, would be able to bring more talent to key CHIPS states’ college towns like Columbus and Tempe.
A third potential change would be to revise the Exchange Visitor (“J-1”) visa. The statute that inaugurated the program in 1970 requires some high-skill J-1 recipients, such as scientists and university researchers, to cycle back home for two years after five years in the United States if the State Department designates the recipient’s home country as “clearly requiring the services of persons engaged in the field of specialized knowledge or skill in which the alien was engaged.” In practice, it is the home countries themselves that set the list of fields. Neufeld estimates that 35,000 to 45,000 high-skill exchange visitors are sent back home to meet this requirement annually. He argues, however, that the requirement is outdated, having been undermined by newer research on human capital. “We now know,” Neufeld writes, “that global networks of skilled migrants are a crucial conduit for ideas, investment, trade, and technological advances for migrants’ countries of origin.” Revising the criteria for the home-leave requirement to align with modern economic insights could make this visa more attractive and keep many recipients in America for longer periods—all without harming recipients’ home countries.
Congress passed the CHIPS Act in response to two interconnected concerns: the pandemic-era chip shortage that slowed production of American goods, like cars, and China’s rising bellicosity toward Taiwan, home of the world’s cutting-edge chipmaker. American industrial policy’s central aim is to mitigate the threat China poses to the chip supply chain. Absent the tens of thousands of skilled workers that the Semiconductor Industry Association sees as necessary to sustain the U.S. chips industry, however, the most likely result of American industrial policy is that TSMC will build capacity stateside but leave its centers of innovation in Taiwan, where talent and experience are stronger.
A singular governmental focus on semiconductors is likely to result in excessive investment in the sector, but the industrial policy train has left the station. Putting American fabs at the forefront in the coming decades, and reducing China risk, will require inducing the world’s best talent to work here.
Photo by Cindy Schultz for The Washington Post via Getty Images