The US government’s attempt to stop the rise of Huawei may end up hurting our own technology industry:
One big unknown surrounding the new Huawei rule—which the chip industry’s lawyers are busily unpacking—is whether, under it, equipment manufactured at American firms’ overseas facilities counts as “American”. If so, advanced chipmaking factories that rely on such kit to fabricate cutting-edge chips for Huawei, as tsmc does, will need alternative suppliers. The American toolmakers’ Japanese rivals, such as Tokyo Electron and Hitachi High-Technologies, suddenly find themselves with a new geopolitical competitive edge. . . .
On May 18th the boss of Samsung Electronics toured his company’s new chip factory in Xian, a city in central China. The South Korean firm, which plans to invest $115bn in its chipmaking business over the next decade, has made it clear that it will not ignore China. America’s export controls may prompt it to kit out its foundries with equipment that will not fall foul of Sino-American geopolitics.
Chip-industry insiders report that semiconductor equipment is already being marketed inside China as “ear free”—meaning Chinese buyers need not worry about the “export administration regulations” that the Trump administration is using to attack Huawei. A person close to American toolmakers says some of them are thinking about moving their patents abroad to rebuild operations from scratch away from America’s jurisdiction, in order to circumvent present and future anti-Chinese restrictions. Mr Trump’s attempt to de-Sinify the semiconductor industry may do more to de-Americanise it instead.
PS. I’ve done several posts on the effect of lockdowns on the economy. This study caught my eye:
The problem with reopening is that while lockdowns are easing, customers aren’t returning. New research led by Harvard economist Raj Chetty finds that business reopenings aren’t enough to lure shoppers worried about the virus. In states that reopened early, such as Minnesota, spending hasn’t picked up more robustly than in states that stayed closed longer, like neighboring Wisconsin. And spending fell the most before lockdown orders went into effect, suggesting consumers were locking themselves down because of health concerns, a trend that seems to be continuing even as restrictions ease.
This doesn’t mean lockdowns have no economic effect, or are a good idea. But it does suggest they aren’t the main problem with the economy.
READER COMMENTS
Mark Z
Jun 24 2020 at 3:48am
Is there an equally valid counterargument that we should actually want to sell as many components as possible to Huawei so as to keep it as dependent on American technology as possible? Would that make it easier for “spooks in Washington” to spy on China? I don’t understand the tech here at all so maybe not, but it seems like a two-way street.
LC
Jun 24 2020 at 1:21pm
Mark Z:
This is a bigger disaster for American tech than being reported here, because the repercussions extend to all areas in semiconductor chain (Economist only lists fab equipment here) and will last longer than many realize. As someone who follows the Chinese tech industry closely, let me just give you 3 areas where American dominance is slowly being decimated:
Technical chip design software- typically referred as EDA, but cover wide variety from coding of designs, simulation of designs, verification of designs prior to fab, synthesis of netlist and layout tools-this is an area dominated by American Tech. 3 companies (Cadence, Mentor, Synopsys) control vast majority of market share (>85%). Since the campaign against Huawei started, Chinese and foreign companies has started to diversify away from American Tech. One of the biggest investments in Huawei R&D in the last 2 years has been EDA software. Chinese government is actively investing and assisting. Some of these losses will never come back.
Technical computing software – think Matlab, Mathematica – again there are alternatives being invested by Chinese academia, government and companies. Many open source alternatives are being added. Again, the competition for American firms will rise and some of the losses will never come back.
Semiconductor fab equipment, as outlined by Economist, will also spur challenges, including home grown Chinese companies.
Other areas of impact include tech personnel (H1-B visa restrictions will force many Indian contractors to look for opportunities in China. Already, the Indian tech presence in China is growing), American company reputations (many young Chinese engineers are increasingly turning toward Chinese government funded research institutions and companies because they can’t come to America or they don’t want to work for a country that’s seen as anti-China.)
There are two things about the current American approach that’s particularly maddening and infuriating: 1.). This approach won’t work with American actions alone. Without a rest of world alignment in American camp, this will not work. There is no incentive for others to line up with an unpredictable America. (Even in the Cold War against Soviet Union, some countries explored the opportunity gaps unscrupulously.). 2.) This neatly dovetails with Chinese government’s MIC (Made in China) 2025 campaign. The Chinese government can spurn large investments in areas where they precisely can catch up to America and kick out American presence. Our own arrogance and stupidity is blinding us to the fact that we (America) already lost more than we ever could hope to recoup.
Michael Pettengill
Jun 25 2020 at 3:33am
You missed the patent situation.
Huawei has been far more aggressive designing and building 5G so its had to invent more 5G related solutions, so it has more 5G patents than any other entity. Which means by being an active builder, unlike Qualcom which emphasizes patents and outsources limited product production, Huawei is beating the US “we think for profit” model by putting product building first.
For the US to try to invalidate Huawei patents which are applied for in multiple international forums, China, India, others will counter by invalidating US patents, just as China is complying with the international system because its patent portfolio matches the rest of the world defensively.
Emerging 5G standards require Huawei patents, and all work on 5G standards requires working with Huawei.
Patents since circa 1980 have become a regulatory mess that both cost a lot of labor to get government monopoly power to cost competitors lots of labor to find substitutes to what’s mostly obvious. I remember the first patents on software which to we programmers seemed obvious, and likely not merely prior art, but common practice, but when coding rapidly and stealing from peers, having evidence of an exact date of code, it’s hard to invalidate a patent granted by an examiner who has not idea about the millions of lines of existing code, written with no thought to patents because they were “math”, and math couldn’t be patented. Chip circuit design is even more like math, but it’s now patented. A policy choice to both promote innovation by monopoly, and to win internationally by thinking, instead of making, given Asia could make with lower labor costs.
But China just put lots of workers into the patent monopoly game.
Huawei is the Chinese firm with the aggressive founder who got his workers to play the game to win, starting years ago.
Thomas Hutcheson
Jun 24 2020 at 9:11am
You only have to look at the catastrophic gap between TIPS expectations and the Fed’s supposed inflation target of 2% PCE to see that the recession is mainly a demand shock, triggered by a supply shock, perhaps, but a demand shock.
Tom DeMeo
Jun 24 2020 at 10:08am
Is digital tech generally defensible against bad actors? No, it isn’t. We are building a world we cannot defend.
David Henderson
Jun 24 2020 at 10:19am
Is the Wisconsin thing in Chetty et al a typo? Wisconsin opened up well before Minnesota.
Scott Sumner
Jun 24 2020 at 12:13pm
I’m not sure. I got that from a news source, so it’s second hand. But it does match other findings, such as Denmark/Sweden.
LC
Jun 24 2020 at 12:48pm
A version of Raj Chetty’s presentation slide is here: https://bcf.princeton.edu/wp-content/uploads/2020/06/chetty.pdf
As I recall from the talk, the team chose Wisconsin and Minnesota because they were very similar economically and closed around the same time, but Minnesota had an earlier opening. The data is in page 63 of the presentation.
David Henderson
Jun 25 2020 at 2:16pm
The Wisconsin Supreme Court opened up Wisconsin on May 13. Minnesota opened up on May 18. Hopefully Chetty et al’s data are better on the other states.
Philo
Jun 24 2020 at 1:18pm
Assuming you yourself choose the titles for your blog posts, I must complain about your liking for the nationalist use of ‘we’. As the post makes clear, we aren’t shooting ourselves in the foot: The U.S. government is shooting the U.S. technology industry in the foot. The agent and the patient are not at all the same, a point obscured by nationalist language.
Aladin
Jun 28 2020 at 10:04pm
Its funny to see Scott accused of nationalism. A cursory glance would reveal he is anything but.
Comments are closed.