"The essence of US-China strategic competition is a struggle between economic development and domestic governance rather than a traditional hegemonic or military-security struggle."
A useful reminder that Chinese analysts are at least as prone to slanting things favorably as their American counterparts:
These quotes consistently underestimate how much value, IP, and technological capability is still within American hands when it comes to the development and fabrication of chips.
They overestimate goodwill towards and dependence on China in Taiwan and the Netherlands, doubly so in South Korea, whose sitting government seems to remember the Lotte debacle well, and triply so in Japan.
They obviously misunderstand the reality that the GOP will be more than willing to follow CHIPS up with further direct subsidies to industry, even if perhaps dressed up as tax cuts.
They still proclaim that state-led allocation of capital is more conducive to bleeding-edge technological development than market-led.
Most importantly, however, they continue to delude themselves that the US strategy plays into the hands of China's "strengths," namely its investment-led development model. For the last 3-4 years, I've seen Chinese policymakers and think-tankers desperately grasping for a reason as to why that model will continue to be world-beating, instead of already being a decade past its expiration date.
They *know* that the Party lacks the institutional capacity to shift China out of the channel that a legacy of rapid development has carved in the bedrock, and they understand that even if the Party had the capability, Xi's conception of comprehensive national power will not brook any such attempt. So everyone whistles past the graveyard, hunting for a rationale as to why it's actually a *good* thing that investment is 45% of GDP in 2022, when Wen Jiabao said consumption should take a leading role in 2011.
Let's see how that goes; I think Pettis is more or less on-point when he says the hangover from existing overinvestment has already ensured China faces a Japan-style Lost Decade (or two), but if they continue down this road further I think the collapse of the Brazilian military regime in the crisis of the 1980's is a more relevant model.
No objections to format, there's value in a clean translation.
Just commenting that I think they've drank a bit too much of their own kool-aid; anyone with any credibility within Xi's government is no longer *able* to do fully objective analysis.
isn't "defense spending" by China's gov also considered "consumption"? So they can pivot out to a consumption led economy by simply stocking weapons missiles ships jets?
Lol, you are right, I believe. If non-welfare spending by the government gets subtracted from consumption, the household consumption bit of the economy is under 40%.
"They obviously misunderstand the reality that the GOP will be more than willing to follow CHIPS up with further direct subsidies to industry, even if perhaps dressed up as tax cuts."
The biggest problem for US build up chip section is NOT money, but talent (i.e. lack of talent).
I don't know what's your background, but from my engineering background, literally NOBODY is studying the computer engineering, while everyone can code a little bit get equally well paid software/data science field and less cyclical industry.
Just go to the Arizona TSMC job forum, the turnover over there is huge. Who is going to work in those crazy oncall schedule TSMC setup while there are pretty good alternative jobs out there?
I want US chip succeed in US, but I am afraid the whole 50 billions chip act and whatever money government throw after that is completely waste of taxpayer's money. There are reasons why chip sectors center around Asia, not US and some of those are not just money factor.
I mean, those are still, at heart, money factors. It's hardly surprising that the wage structure required to get someone with a sought-after MS or Ph.D to accept that sort of employment is significantly higher in the US, where GDP per capita is close to twice that of South Korea or Taiwan.
That, too, is a cost gap that could be dealt with by the Federal government throwing enough money at the problem to make up for the higher private sector wage structure.
That TSMC is having some staffing-related teething issues operating in the US for the first time is hardly surprising, let's see how they play out.
Here are some of not money factors involved. 1) Hardware engineer in US has no respect compared to software engineer, while in Asia Hardware engineer is treated as superstar. 2) There are always plenty of engineers in Asia due to overeducated population in general. As for US, there are plenty of alternative choice for those same engineer type 3) Process engineering (and in particular) are required long hour (like 12 hours shift) and crazy hour. Maybe it is opportunity cost or proud. They can handle it but I have doubt US engineers can handle it.
It is not just my view. That's same view from TSMC CEO's view.
Ehh, every Chinese and Taiwanese firm is littered with senior personnel who have unevidenced opinions about how Americans can't hack it. Color me unimpressed, singularly and completely unimpressed.
I went to graduate school at a top Mainland university for an engineering course of study with Mandarin as the language of instruction, I've worked in engineering fields in both the US and in China.
Things are different in China or Taiwan but not better in any meaningful sense of the word. The "last to leave" culture is rampant and huge amounts of time are wasted on pointless bullshit to fill out 9-9-6 schedules that make people miserable and produce per-hour productivity figures that look like someone scraped them out of the very bottom of the latrine.
Even in Taiwan and South Korea, the tide is turning among my age cohort. Hell, even on the Mainland! No one looks at Japan's put-upon lower-tier office workers and thinks that's a viable model, except maybe the CEO of TSMC, who can't envision anything else.
The US has plenty of sophisticated manufacturing with process engineers working crazy hours for good money; I know more than a few.
Another factor is cyclical nature of chip industry many young engineers against join this field.
When the new fab build, a lot of times old engineers from old fab got layoff since they are out of touch with newer technology. That's another reason many older engineers in firm like Texas Instrument against career in Chip industry. I saw plenty of engineers in early 50s are forced to early retire and has to find another professional
But... TSMC and Samsung (and TI) are running fabs that were built before I was born and people are still running them!
Look, while strategic, with exceedingly long lead times, and finicky as hell... at the end of the day the fabrication of even pretty cutting-edge wafers is on the lower end of the profitability and value-added side of the sector, let alone pure commodity chips.
There is a reason that Nvidia has, in a typical year, over half of TSMC's profitability on revenues one-fifth as great and a quarter the invested capital, for example.
If the US genuinely sees this as a strategic need, it will subsidize its way to solving the associated economic bottlenecks, and it has far more fiscal slack to do so than any other actor and a better technological starting point as well because about 40-50% the value add in the semiconductor sector is still coming from American firms.
I do not see a world in which the US still permits this strategic vulnerability by 2028 or '29, even if it requires another $50 billion every other year for two decades to make it pencil out.
EDIT: I am, in fact, vastly more concerned about the trailing edge applications. Aside from existing foundries, it truly makes no sense and would require vast subsidy to make a profitable 45nm fab in the US aside from the fully depreciated ones that are slowly winding down. Even TSMC will not build another.
The US needs a near-shoring strategy to get American firms investing for trailing-edge fab capacity in at least one of Mexico, the Dominican Republic, or Colombia.
Allowing China to produce 50% of commodity chips (as they will by 2030) doesn't matter at all from an economic or innovation perspective, but from a geostrategic one it's deeply worrisome.
I was wondering if a "super strong dollar" policy, if it could be pursued discretionarily on purpose, is a part of the US' economic warfare v China toolkit? A lot of the supply chains which China claims to be heavily integrated into are likely low margin supply chains, and for China's economy to maintain its earnings, would necessitate RMB devaluation in step with other suppliers like ASEAN S. Korea etc. And then, given that the US would be exporting inflation to the global South with it's rate hikes and super strong dollar, would require/test price-increase power among the supply chain links, at a time of US import disruption. Not to mention the debt crises that might unfold for the heavily leveraged, overseas.
However, this strong dollar policy is drawing blood in the OECD countries and home in the US as well!
A useful reminder that Chinese analysts are at least as prone to slanting things favorably as their American counterparts:
These quotes consistently underestimate how much value, IP, and technological capability is still within American hands when it comes to the development and fabrication of chips.
They overestimate goodwill towards and dependence on China in Taiwan and the Netherlands, doubly so in South Korea, whose sitting government seems to remember the Lotte debacle well, and triply so in Japan.
They obviously misunderstand the reality that the GOP will be more than willing to follow CHIPS up with further direct subsidies to industry, even if perhaps dressed up as tax cuts.
They still proclaim that state-led allocation of capital is more conducive to bleeding-edge technological development than market-led.
Most importantly, however, they continue to delude themselves that the US strategy plays into the hands of China's "strengths," namely its investment-led development model. For the last 3-4 years, I've seen Chinese policymakers and think-tankers desperately grasping for a reason as to why that model will continue to be world-beating, instead of already being a decade past its expiration date.
They *know* that the Party lacks the institutional capacity to shift China out of the channel that a legacy of rapid development has carved in the bedrock, and they understand that even if the Party had the capability, Xi's conception of comprehensive national power will not brook any such attempt. So everyone whistles past the graveyard, hunting for a rationale as to why it's actually a *good* thing that investment is 45% of GDP in 2022, when Wen Jiabao said consumption should take a leading role in 2011.
Let's see how that goes; I think Pettis is more or less on-point when he says the hangover from existing overinvestment has already ensured China faces a Japan-style Lost Decade (or two), but if they continue down this road further I think the collapse of the Brazilian military regime in the crisis of the 1980's is a more relevant model.
yea--maybe next time i do this i'll insert parentheticals on where I think they're missing stuff or getting facts wrong
No objections to format, there's value in a clean translation.
Just commenting that I think they've drank a bit too much of their own kool-aid; anyone with any credibility within Xi's government is no longer *able* to do fully objective analysis.
isn't "defense spending" by China's gov also considered "consumption"? So they can pivot out to a consumption led economy by simply stocking weapons missiles ships jets?
Lol, you are right, I believe. If non-welfare spending by the government gets subtracted from consumption, the household consumption bit of the economy is under 40%.
"They obviously misunderstand the reality that the GOP will be more than willing to follow CHIPS up with further direct subsidies to industry, even if perhaps dressed up as tax cuts."
The biggest problem for US build up chip section is NOT money, but talent (i.e. lack of talent).
I don't know what's your background, but from my engineering background, literally NOBODY is studying the computer engineering, while everyone can code a little bit get equally well paid software/data science field and less cyclical industry.
Just go to the Arizona TSMC job forum, the turnover over there is huge. Who is going to work in those crazy oncall schedule TSMC setup while there are pretty good alternative jobs out there?
I want US chip succeed in US, but I am afraid the whole 50 billions chip act and whatever money government throw after that is completely waste of taxpayer's money. There are reasons why chip sectors center around Asia, not US and some of those are not just money factor.
I mean, those are still, at heart, money factors. It's hardly surprising that the wage structure required to get someone with a sought-after MS or Ph.D to accept that sort of employment is significantly higher in the US, where GDP per capita is close to twice that of South Korea or Taiwan.
That, too, is a cost gap that could be dealt with by the Federal government throwing enough money at the problem to make up for the higher private sector wage structure.
That TSMC is having some staffing-related teething issues operating in the US for the first time is hardly surprising, let's see how they play out.
Here are some of not money factors involved. 1) Hardware engineer in US has no respect compared to software engineer, while in Asia Hardware engineer is treated as superstar. 2) There are always plenty of engineers in Asia due to overeducated population in general. As for US, there are plenty of alternative choice for those same engineer type 3) Process engineering (and in particular) are required long hour (like 12 hours shift) and crazy hour. Maybe it is opportunity cost or proud. They can handle it but I have doubt US engineers can handle it.
It is not just my view. That's same view from TSMC CEO's view.
https://appleinsider.com/articles/22/10/24/tsmc-says-efforts-to-rebuild-us-semiconductor-industry-are-doomed-to-fail
Ehh, every Chinese and Taiwanese firm is littered with senior personnel who have unevidenced opinions about how Americans can't hack it. Color me unimpressed, singularly and completely unimpressed.
I went to graduate school at a top Mainland university for an engineering course of study with Mandarin as the language of instruction, I've worked in engineering fields in both the US and in China.
Things are different in China or Taiwan but not better in any meaningful sense of the word. The "last to leave" culture is rampant and huge amounts of time are wasted on pointless bullshit to fill out 9-9-6 schedules that make people miserable and produce per-hour productivity figures that look like someone scraped them out of the very bottom of the latrine.
Even in Taiwan and South Korea, the tide is turning among my age cohort. Hell, even on the Mainland! No one looks at Japan's put-upon lower-tier office workers and thinks that's a viable model, except maybe the CEO of TSMC, who can't envision anything else.
The US has plenty of sophisticated manufacturing with process engineers working crazy hours for good money; I know more than a few.
We'll do fine.
Another factor is cyclical nature of chip industry many young engineers against join this field.
When the new fab build, a lot of times old engineers from old fab got layoff since they are out of touch with newer technology. That's another reason many older engineers in firm like Texas Instrument against career in Chip industry. I saw plenty of engineers in early 50s are forced to early retire and has to find another professional
But... TSMC and Samsung (and TI) are running fabs that were built before I was born and people are still running them!
Look, while strategic, with exceedingly long lead times, and finicky as hell... at the end of the day the fabrication of even pretty cutting-edge wafers is on the lower end of the profitability and value-added side of the sector, let alone pure commodity chips.
There is a reason that Nvidia has, in a typical year, over half of TSMC's profitability on revenues one-fifth as great and a quarter the invested capital, for example.
If the US genuinely sees this as a strategic need, it will subsidize its way to solving the associated economic bottlenecks, and it has far more fiscal slack to do so than any other actor and a better technological starting point as well because about 40-50% the value add in the semiconductor sector is still coming from American firms.
I do not see a world in which the US still permits this strategic vulnerability by 2028 or '29, even if it requires another $50 billion every other year for two decades to make it pencil out.
EDIT: I am, in fact, vastly more concerned about the trailing edge applications. Aside from existing foundries, it truly makes no sense and would require vast subsidy to make a profitable 45nm fab in the US aside from the fully depreciated ones that are slowly winding down. Even TSMC will not build another.
The US needs a near-shoring strategy to get American firms investing for trailing-edge fab capacity in at least one of Mexico, the Dominican Republic, or Colombia.
Allowing China to produce 50% of commodity chips (as they will by 2030) doesn't matter at all from an economic or innovation perspective, but from a geostrategic one it's deeply worrisome.
I was wondering if a "super strong dollar" policy, if it could be pursued discretionarily on purpose, is a part of the US' economic warfare v China toolkit? A lot of the supply chains which China claims to be heavily integrated into are likely low margin supply chains, and for China's economy to maintain its earnings, would necessitate RMB devaluation in step with other suppliers like ASEAN S. Korea etc. And then, given that the US would be exporting inflation to the global South with it's rate hikes and super strong dollar, would require/test price-increase power among the supply chain links, at a time of US import disruption. Not to mention the debt crises that might unfold for the heavily leveraged, overseas.
However, this strong dollar policy is drawing blood in the OECD countries and home in the US as well!