In a week, the United States will mark an anniversary that no American should want to celebrate: It was last June 10 and 11 that companion bills were introduced in both the House and Senate to increase greatly the U.S. government’s support for domestic semiconductor manufacturing. Since I’m a strong backer of such efforts, why am I so downbeat? Because despite the importance of strengthening the American footprint in this sector for both national security and future prosperity, and despite seemingly strong bipartisan support for this effort (at least in principle) nearly a year later, not a single penny has been been spent.
It would actually be reasonable to argue that the federal government took way too long to take even that preliminary step. After all, as I documented in this article last October, America’s global leadership in producing (as opposed to designing) the microchips increasingly crucial to so many defense-related and civilian products and services – and indeed, entire industries – had been waning for decades, and was finally lost in 2017. That’s the year when U.S.-owned Intel became unable to keep up with Taiwan’s Taiwan Semiconductor Manufacturing Company in turning out semiconductors featuring the world’s smallest circuit sizes – the main indicator of a chip’s capabilities.
So it’s not terribly impressive that American political leaders took two years to begin responding in a serious way. (And P.S. – the executive branch, under President Trump, clearly wasn’t johnny-on-the-spot, either, in using the bully pulpit to sound the alarm and generate support for action.)
Still, the bipartisan nature of the legislative effort – at a time of heated partisanship on virtually every other national issue – seemed cause for encouragement. Even better: Just a month later, the House and Senate passed their respective semiconductor bills.
Since then, however, progress has been sluggish. The Representatives and Senators didn’t manage to get their acts together before that session of Congress ended in order to draft and pass the consensus bill needed to go to the President’s desk for signing. Therefore, the measures died, and work needed to begin all over again this past January, when the new Congress convened.
Semiconductor work was proceeding along another track in late 2020, and resulted in key provisions of the expired bill being incorporated into legislation authorizing the Defense Department’s levels and kinds of spending for this fiscal year. That bill became law this New Year’s Day (over a Trump veto for unrelated reasons), but according to Congress’ procedures, authorizing bills can’t trigger any spending. That requires an appropriations bill – which also must be passed in identical form by both chambers before enactment.
Six months later, there’s still no money flowing. The story is excrutiatingly difficult to follow, but it appears that Senate Majority Leader Chuck Schumer of New York tried to speed up the process in May with an emergency funding measure. Passage seemed likely at month’s end, before the Senate’s scheduled Memorial Day recess, but was stymied at the last minute by a sadly typical array of political shenanigans from both the minority Republicans (whose support was needed because of the Senate’s filibuster provision requiring super-majorities to pass most legislation) and Democrats. (See here and here for good accounts.)
Passage of a similar measure by the House looks to be easier, because of the Democrats’ slightly bigger majority. But there the process is less advanced, since the House Democrats’ own technological competitiveness proposals were only introduced in committee May 25.
It’s not like the U.S. private sector has been standing still. Intel, most significantly, seems determined to reemphasize manufacturing again, and has committed to put lots of money where it’s mouth is. But without a major helping hand from Washington, this campaign is sure to be swamped by the massive amounts of foreign government subsidies for promoting advanced semiconductor manufacturing that have been announced lately. (Here’s a useful summary.)
I’m generally a fan of the cautious approach to policymaking fostered by the U.S. Constitution’s separation of powers and checks and balances principles. And I wouldn’t be so fast, like so many Democrats, to junk the Senate’s filibuster rule (which is not found in the Constitution). Yet time is not America’s friend when it comes to regaining lost ground in a fast-moving industry like semiconductors, and if Washington continues its business-as-usual approach on this issue, history will likely conclude that the American political system failed a big test.
Full disclosure: I own a not-trivial number of shares of TSMC common stock.
RealityChek 