The United States is pouring money into microchips, but will this fix the supply chain problem? The newly-passed CHIPS Act provides more than $75 billion for advanced fabrication of microchips in the U.S. and is the centerpiece of a strategic effort to boost domestic high-tech manufacturing.
Money alone, however, is not enough. If policymakers don’t correct an array of obstacles across the high-tech supply chain, the CHIPS effort and America’s broader high-tech manufacturing strategy could falter.
Over the past four decades, most of the world’s advanced semiconductor manufacturing capacity gravitated to Taiwan and nearby Asian locales, while the U.S. share of leading-edge capacity fell to around 12 percent. Low-wage labor was a factor, but enormous government subsidies were more important. Now, China and India are pushing additional subsidies and further undercutting the U.S. on a host of industrial regulatory burdens.
Sensing this growing geographic concentration risk, a bipartisan legislative coalition agreed to support American chip manufacturing with $52 billion in grants, $24 billion in tax credits, and a range of other incentives. The administration also launched a new office to distribute funds to qualified candidates.
A mere focus on the big chip firms themselves, however, will not secure the objective. Semiconductors and other finished high-tech products do not magically emerge from fabs, the giant $10 billion to $20 billion factories owned by Intel, Samsung and TSMC.
Making a chip is a long and unimaginably complex undertaking involving thousands of steps, beginning with humble sand that is refined into polished silicon wafers the size of a medium pizza. Multimillion-dollar machines then shine powerful ultraviolet light through a series of masks and lenses to create atomic-scale maps of what will become switches and wires. Chemical liquids and vapors then wash away unwanted materials and deposit new nano-structures to create mind-blowing 3D grids of up to 80 billion transistors on a single tiny chip. To accomplish this, fabs and the nano-precision equipment inside require hundreds of specialty materials.
Among the most important are fluoropolymers, a category of inert and slippery substances based on carbon-fluorine bonds, first discovered in 1938. You know the first fluoropolymer as TeflonTM. Part of a larger category known as PFAS, fluoropolymer chemistries are essential across the range of high-tech and precision manufacturing.
Fluoropolymers are highly resistant to corrosion and extreme temperatures and are used throughout the chip-making process, which requires ultra-purity. From the infrastructure of a fab to the photolithography process itself (chemically etching patterns into silicon), these chemistries are vital. You simply cannot build a fab or a microchip — or most other high-tech products, such as lithium batteries, pharmaceuticals and medical devices — without fluoropolymers.
There are troubling signs that policymakers may overreach when it comes to these essential compounds. In June, as part of its ongoing approach to PFAS, the Environmental Protection Agency (EPA) set recommended levels for PFAS so low that technology does not exist to detect them. Sen. Shelley Moore Capito (R-W.Va.) described the levels as “impossible” and the policy move as “ideological.” The American Chemistry Council said it was “a failure by the agency to follow its accepted practice for ensuring the scientific integrity of its process.”
The EPA’s action is an indication of what’s to come if the government reaches beyond a science-based process to establish unworkable standards, creating regulatory and permitting hurdles for major new investments — just as the U.S. is seeking to ramp up capacity. China and India, of course, are doing just the opposite.
We’ve seen over the past few years how interruptions of crucial inputs throw delicate production webs into chaos. As we emerged from the pandemic, a chip shortage itself led to long waits for automobiles and washing machines. Now, natural gas disruptions and price spikes threaten to shut down major industries across Europe.
For too long, Americans and Europeans, staring at our smartphones, took for granted the seemingly automatic supply chains of the physical world. Much of the manufacturing migration to Asia would have occurred no matter what. But the U.S. also compounded this outflow with our own mistakes, pushing industry away with a punishing corporate tax code and self-harming regulation.
No more. If we want the digital magic of the internet cloud and the effortless acceleration of electric vehicles — and if we want to make them at home — we need to re-engage every link in the value chain. That means producing more energy, mining more materials, and allowing scientists and engineers to transform these raw resources into a million miraculous substances and devices.
It doesn’t mean retreating from the world or cutting off trade. Certainly, it does not mean total self-sufficiency or autarky. Nor does it mean environmental harm. We are better than ever at doing these things cleanly. Richer economies yield cleaner environments.
If U.S. policymakers are serious about the strategic goal of CHIPS, they will adopt a complementary regulatory approach so that our policies are not frustrating one another but pushing in the same direction. All the money in the world can’t produce a microchip if you don’t let people build it.
Bret Swanson is president of the technology research firm Entropy Economics, a nonresident senior fellow at the American Enterprise Institute, and a visiting fellow at the Krach Institute for Tech Diplomacy at Purdue. Follow him on Twitter @JBSay.