At The Hill Mychael Schnell reports that Congress has enacted its bill intended to encourage the manufacture of semiconductors in the United States:
The House passed a $280 billion bill on Thursday to strengthen the domestic chip manufacturing industry and finance scientific research in a bid to boost the United States’s competitiveness on the global stage, sending the measure to President Biden’s desk for final approval.
The legislation, titled the CHIPS and Science Act, cleared the House in a 243-187-1 vote. Twenty-four Republicans supported the measure, and one Democrat voted “present.â€
and
The Senate approved the measure in a bipartisan 64-33 vote on Wednesday, receiving support from Minority Leader Mitch McConnell (R-Ky.). The bill’s passage through both chambers marks a significant congressional achievement and the culmination of more than a year of negotiations over legislation to increase the U.S.’s competitive edge against China.
The House and the Senate initially each passed their own China competition bills, referred to as “America COMPETES†and “USICA,†respectively, but conference negotiations between the two chambers had stalled for weeks.
Lawmakers ultimately came to a consensus on the CHIPS and Science Act, which will allocate $54 billion for chips and public wireless supply chain innovation, including $39 billion that will go towards financial assistance to build, expand and modernize semiconductor facilities in the U.S. It also includes $11 billion for research and development by the Department of Commerce.
The measure seeks to establish a 25 percent tax credit for investment in semiconductor manufacturing and funnel $81 billion to the National Science Foundation (NSF), $20 billion of which will go towards an NSF directorate.
The editors of the Wall Street Journal refer to it as the “Schumer-Manchin Tax and Subsidy Pact”.
So, here’s my question. Two years from now how many more microchips will be manufactured in the U. S. than at present and what proportion of our total consumption does that represent? Five years? Ten years?
If the reporting on the bill is accurate to my eye it appears long on carrots and very short on sticks and has no provision for evaluating whether the bill has succeeded in its objective or not.
I don’t oppose subsidies as a strategy but I think they need to be more directed than these will be. If subsequent to 1962 Congress had passed bills that called broadly for space flight, we would never have landed on the moon.
My kneejerk reaction to this bill is that in two years it will have no material impact on our semiconductor manufacturing and in five years will only have a minor impact. Our dependence on imported semiconductors will continue to grow. I honestly don’t see how we can justify this on a security basis. It would be as though we were dependent for oil on Japan on December 7, 1941.
New fabs take years to construct, so I don’t think this will have any effect in the short term. And I also don’t think this measure will roll back the dominance of Taiwan and TSMC. Even with a more dedicated strategy, significantly reducing our reliance on foreign chip supplies would take a decade or more.
Since it will take 2 years to build a plant that is pretty safe bet. Intel, TSMC and someone else had already announced intentions to build new state of the art plants int he US so guessing we see a significant change in 5 years. The question is could the money let a Global Foundries also develop state of the art factory? How many of those do we need? We still need places cranking out the 20-40 nanometer chips I think.
Steve