The CHIPS and Science Act is bad law. It’s not all bad—more research spending is welcome and decent industrial policy is possible if it is properly focused. But the chips part isn’t focused and the science part leaves the door wide open to Chinese acquisition of US technological breakthroughs. If competing with China is truly a top national priority, the act must be improved, either in implementation or by revision in 2023.
Start with chips. Good law can be ruined in implementation, sometimes intentionally, and a dubious bill can be helped. President Joe Biden said he would personally ensure that larger grants to semiconductor companies are justified. This is encouraging, since the act says little about what our goals are. How advanced should chip plants be? Crucially, are we only subsidizing final manufacturing or the full supply chain?
If the act is enforced, firms receiving subsidies can’t expand manufacturing capacity for average and advanced technology semiconductors in China. But they can freely make improvements to the supply chain there. Beijing has made clear the chip chain is a priority. The results could be frighteningly similar to pharmaceuticals, where finished products are made everywhere but the world depends on Chinese chemical inputs.
In this regard, the act fails miserably. Politicians and CEOs endlessly talk up factories. Where will they get materials and equipment? Where will chips be assembled, tested, and packaged? There is no requirement or even guidance that subsidies ensure reliable supply chains—the word “or” dominates how money can be spent. Will our stab at industrial policy lead to shiny new fabs dependent on China’s shiny new supply chain?
Everything could work out. That currently depends on the executive going beyond the act to make the right allocation decisions—less money for manufacturers and more for the broader ecosystem. Early returns are discouraging—yesterday’s guidance from the Department of Commerce all but ignores the supply chain. And not only did chip-makers take every opportunity to press for subsidies, Commerce handed them a megaphone to do so.
Months of cheerleading for chip-makers make it hard to imagine Secretary Raimondo standing tough when companies inevitably say Beijing offers larger subsidies and they will downsize in the US if they don’t get more money. Or perhaps the threat will be to downsize here if Commerce finds their activities in China go beyond acceptable chip investment there. Expect to hear this starting in 2025.
Elevation of private over public interest is the chief industrial policy risk and Congress failed to be clear about what it wants for semiconductors, making it all too easy to ignore the public interest. The many members touting spending taxpayers’ money on semiconductors are now responsible for strict oversight.
The science side of the act avoids this problem because, at least prior to implementation, it’s not industrial policy. Supporting basic research is a core government function and doesn’t involve picking winners. Still, a number of agencies will be involved, which itself calls for an active congressional role.
With regard to competing with China, the text of the legislation is a very poor start. Research security provisions to prevent illicit acquisition of technology were present in the initial Senate bill, dropped by the House, and an attempt to restore them in the final bill was rejected. The reasons appear to be lobbying by universities and bizarre accusations of discrimination.
The universities’ motive is obvious. Just like companies, universities want handouts to do with as they please. Those suspicious of few-strings-attached chip subsidies to corporations are right and those suspicious of few-strings-attached research subsidies to universities are right; the mystery is why they didn’t unite to protect the taxpayer and national interest.
Universities add another element where much of the academic community remains willfully blind concerning the situation of their Chinese counterparts, pretending as if civil-military fusion doesn’t matter and Chinese researchers have a choice if approached by the Ministry of State Security. This ties to the bizarre accusations.
“We fear that it will be another way that the government will be able to target Chinese nationals.” Denying Chinese nationals access to federally-funded technology research constitutes “targeting”? Would denying Soviet nationals access in the 1970s or Iranian nationals now also qualify? The US should pretend that citizens of dictatorships are reliable with regard to advanced technology?
Even with ideal restrictions it would be difficult to prevent subsidized research from assisting Chinese technological progress. For the moment, there is a 100 percent chance the research will help China commercially, militarily, or in internal repression. When that happens, there should be serious consequences for enablers. But a far better path has the executive branch implementing better law than it received, or Congress fixing its mistakes.
The post Fix CHIPS and Science, or Else appeared first on American Enterprise Institute – AEI.