U.S. Semiconductor Manufacturing Tax Credits Need To Be Extended and Broadened

The "Creating Helpful Incentives to Produce Semiconductors" (CHIPS) Act has been a smashing success. The legislation has helped stimulate $540 billion in U.S. semiconductor supply chain investments through more than 100 projects spread across 28 U.S. states, including 17 new fabs and eight new supply chain and advanced packaging facilities. As a result, analysts expect the United States will increase its global share of advanced logic (sub-10 nanometer (nm)) chip manufacturing to 28 percent of global capacity by 2032, up from virtually nil in 2022.

But while most focus on the $39 billion in financing for loans and grants, an even more important, yet comparatively unsung, element of the legislation was its inclusion of a 25 percent investment tax credit (ITC), which is due to expire on December 31, 2026. In the reconciliation legislation now before Congress, policymakers should extend the ITC, and extend its applicability to the semiconductor design sector.

Section 48D of the CHIPS Act provided a 25 percent tax credit for investments toward constructing advanced manufacturing facilities focused on semiconductor manufacturing or related equipment. The ITC was particularly attractive to semiconductor manufacturers for several reasons, beyond simply offsetting 25 percent of the cost of constructing new semiconductor production facilities. First, while companies could not be sure exactly what loan/grant package they might receive from the CHIPS Program Office—or indeed how long it might take to conclude those negotiations or to actually receive disbursements therefrom—they knew they’d be able to claim the 25 percent ITC as soon as they put a shovel in the ground.

Second, that 25 percent played a crucial role in ensuring the economics of U.S. semiconductor manufacturing can work, vitally important when the cost to build a leading-edge fab can reach $28 billion. Yet it costs at least 30 percent more to build and operate a U.S. semiconductor fab (over 10 years) than it does in Asian competitors such as Taiwan, South Korea, and Singapore (and 37 to 50 percent more than in China). And while certainly costs (such as labor, land, regulatory compliance, etc.) tend to be greater in the United States, as much as 40 to 70 percent of the aforementioned cost differential can be directly attributed to government incentives. Thus, the ITC (alongside the CHIPS Act’s loans/grants) have played a critical role in closing that cost gap and making the economics of semiconductor manufacturing in the United States viable.

Overall, the CHIPS Act has played the foundational role in reversing America’s ignominious slide in its share of global semiconductor manufacturing, which had fallen from 37 percent in 1990 to 10 percent in 2022. Analysts project the United States will realize a 203 percent increase in its fab capacity in the decade from 2022 to 2032. In addition to the aforementioned increase in logic chip production, the United States is projected to increase its global share of leading-edge DRAM memory chip production to 10 percent by 2035. Projects announced after the CHIPS Act passed are expected to create and support over 500,000 American jobs: including 68,000 facility jobs in the semiconductor ecosystem, 122,000 construction jobs, and over 320,000 additional jobs (through indirect and induced effects).

In short, the advanced manufacturing investment credit has been a key part of the CHIPS Act’s success. With it due to expire at the conclusion of 2026, Congress should extend it at least through the end of this decade, as proposed in the bipartisan Building Advanced Semiconductors Investment Credit (BASIC) Act co-sponsored by Representatives Claudia Tenney (R-NY) and John Mannion (D-NY). That legislation would go even further by increasing the credit from 25 to 35 percent.

Just as the cost of constructing a semiconductor fab continues to increase, so does the cost of designing leading-edge chips themselves. In fact, the design costs for a complex 5 nm-class system on a chip (SoC) were an estimated 80 percent higher than design costs of a 7 nm-class SoC, and totaled over $540 million. The cost to design a chip increased 18-fold from the 65 nm chip generation to the 5 nm generation. And now analysts estimate the costs to design the newest, leading-edge 2 nm chips have reached upwards of $725 million. In total, the U.S. private sector will likely invest $400 to $500 billion over the next 10 years in semiconductor design-related activities, including research and development (R&D) and workforce development.

As such, U.S. leadership in semiconductor design—in everything from artificial intelligence to communications to digital and analog chips—could greatly benefit from a similar tax credit. For this reason, Congress should pass the Semiconductor Technology Advancement and Research (STAR) Act, which would provide a 25 percent tax credit for semiconductor design R&D expenditures.

The reconciliation bill currently moving through Congress provides an opportunity for policymakers to revise the tax code to stimulate greater levels of private sector investment in the United States, such as by restoring full first-year expensing for companies’ R&D costs. In this process, policymakers should extend and expand the investment tax credits that have been so vital in restoring U.S. semiconductor leadership.