With Beijing menacing Taiwan and its world-leading semiconductor manufacturing sector, American policymakers have converged on the view that the U.S. needs to make more of the chips that run its computers at home. The signature initiative emerging from this new industrial-policy consensus is the Creating Helpful Incentives to Produce Semiconductors (CHIPS) Act, passed by Congress and signed into law by President Joe Biden in 2022. The act promises more than $50 billion in direct grants and loans to chipmakers to set up shop in the U.S. and—with luck—lure a greater portion of the industry’s cutting edge back from Asia.
Already, though, the two highest-profile CHIPS recipients—America’s Intel and the Taiwan Semiconductor Manufacturing Company (TSMC)—have hit turbulence. Intel, for which the Department of Commerce slated $8.8 billion in March, posted a stunning net loss for Q2 2024 of $1.6 billion and announced a layoff of 15,000 workers in August. TSMC, meanwhile, has pushed back the production start date at its first CHIPS-funded Arizona plant from this year to next and at its second plant from 2026 to 2027 or 2028. This strikes an unflattering contrast with the ahead-of-schedule progress TSMC has made at its new fabrication plant in Japan’s Kumamoto prefecture.
Many factors dim the CHIPS Act’s prospects of propelling American fabs to the chip industry’s leading edge. Business types might note the Taiwanese management practices and workplace norms that TSMC (let alone Intel) cannot replicate stateside. Chistopher Miller’s 2022 book Chip War substantiates this thesis. Among other evidence, he cites an interview he conducted with Shang-yi Chiang, who headed TSMC’s R&D department and worked for American companies in Texas and California. Chiang told Miller that “people worked so much harder in Taiwan . . . If something broke at 1 a.m. in America, the engineer would fix it the next morning; at TSMC, they would fix it by 2 a.m.” The American political Right, alternatively, pins CHIPS slowdowns on the Biden administration’s diversity, equity, and inclusion (DEI) fixation. Matt Cole and Chris Nicholson, both of Vivek Ramaswamy’s Strive Asset Management, wrote for The Hill in May that CHIPS grant money is “so loaded with DEI pork that it can’t move.”
Another factor, though, deserves more attention amid the CHIPS Act’s early struggles: environmental “safetyism,” the philosophy that taking the utmost precaution ought to be a governing principle. Precaution can, of course, make sense in some contexts, but when it trumps evidence-based regulatory practices, the result is stagnation.
Removing safetyist hindrances to CHIPS is a lever of greater potential for accelerating productivity in the U.S. than addressing either workplace culture or DEI pork. Workplace culture is entrenched, extra-political, and vague; DEI pork is newfangled and already falling out of fashion. By contrast, safetyism is far more institutionalized than DEI, having created noticeable economic drag for decades, yet it is less sticky than cultural mores. Crucially, it can be reformed through standard regulatory practices.
While the 1970 National Environmental Policy Act (NEPA) has become the “abundance” movement’s bête noire, a lesser-known—and perhaps less comfortable—hurdle to the CHIPS Act’s success is the Toxic Substance Control Act. TSCA exhibits safetyism’s hallmark characteristics and yields its inevitable results.
Congress enacted TSCA in 1976 to give the U.S. Environmental Protection Agency the authority to regulate the entire lifecycle of chemical compounds. TSCA requires manufacturers to report on the chemicals they produce, use, or import and the potential worker health and environmental risks therein, and to comply with regulations on handling and disposal. The Lautenberg Chemical Safety for the 21st Century Act, signed into law by President Barack Obama in his final year in the White House, enhanced the EPA’s authority and tried to systematize its processes more effectively. President Biden’s EPA, however, has stretched TSCA risk evaluations beyond reason, resulting in bans and unachievable workplace standards for well-established industrial compounds, many of which are used in closed systems with little chance of exposure, and innovation-killing review lags for novel chemical creations.
Few think of semiconductor manufacturing as a dirty business—chips are made in “clean rooms,” after all—but it is without question a material-intensive process. Chipmaking requires specialized chemical compounds to create both the optimal conditions for their manufacture and the products themselves. According to the American Chemistry Council, producing advanced semiconductors requires more than 500 specialized compounds, including formaldehyde, PFAS, trichloroethylene, N-methyl pyrrolidone, and methylene chloride. These are some of the same chemicals that the Biden EPA’s TSCA implementation has relegated to purgatory.
As utilized by the Biden EPA, TSCA sets four designations for chemical compounds: “not likely to present an unreasonable risk,” “insufficient information,” “may present an unreasonable risk,” and “presents an unreasonable risk.” The first designation gives manufacturers the green light to use a chemical, the second and third a yellow light, and the fourth a red light.
Americans enjoy a presumption of innocence in a courtroom, but in TSCA regulatory proceedings, chemicals are presumed unreasonably risky until proven otherwise. The light is red, in other words, until the EPA changes it. Moreover, the EPA assumes that manufacturers will abide neither by their own internal protocols nor by other federal occupational safety and health regulations such as those that stipulate the use of personal protective equipment in certain circumstances. Instead of running analyses based on protocols that stipulate the wearing of gloves, for example, the EPA treats manufacturers as if their employees are bathing their bare hands in hazardous solvents for a whole shift. From the safetyist perspective, this makes perfect sense. From the perspective of a government that has committed billions to rapid production of crucial industrial inputs like chips, it’s a massive contradiction.
As of July 2024, 243 new chemicals have been in TSCA review for a year or longer, the American Chemistry Council reports. The result is that chipmaking businesses have more trouble getting and using the materials they need in the U.S. than they and their competitors do abroad. As the Ohio Chemistry and Technology Council noted in a May regulatory comment filing, TSCA’s new risk evaluation approach creates “significant economic damage” in the very states that are the epicenters of American industrial policy.
Taking into account lab innovation, industrial production, and commercialization, the United States has been the world’s undisputed technology leader for a century. The early semiconductor industry that blossomed in what we now know as Silicon Valley arguably represents the apex of American dominance across the technology value chain. Since the 1970s, however, America’s lead has narrowed in that field and others. Much of the narrowing can be attributed to other regions’ realization of their own potential and to the natural operation of global markets. The outcome, in broad strokes, has been that America designs, while Asia builds. But that’s not the whole picture. Environmental safetyism has imposed unjustifiable costs on chipmaking and many other industries.
As the EPA’s detractors would be quick to point out, TSCA is only one part of a suite of safetyist policies plaguing the country. To keep ahead of China—with no safetyist compunctions of its own—the U.S. must overturn the hyper-cautious norms that have locked down the nation’s productivity. Addressing TSCA’s problems is a great place to start.
Photo: sefa ozel / E+ via Getty Images